Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Leukotriene (LT) B₄ is an inflammatory mediator that is synthesized by a number of cell types, including mast cells, neutrophils, and macrophages (Gorman, 1989). It is a potent activator of neutrophils, inducing chemotaxis, aggregation, degranulation, adherence, and priming of inflammatory cells for induction by other agonists, such as prostaglandins and cysteinyl leukotrienes (Ford-Hutchinson and Evans, 1986). Excessive production of LTB₄ has been implicated in the pathogenesis of various immune and inflammatory diseases, including psoriasis and inflammatory bowel disease (Ford-Hutchinson and Evans, 1986).

The rate-limiting step in the biosynthesis of LTB₄ is the specific epoxide hydrolase, LTA₄ hydrolase (E.C. 3.3.2.6), which converts the unstable allylic epoxide LTA₄ to LTB₄ (Fig. 1). LTA₄ hydrolase is a soluble monomeric enzyme, ubiquitous in mammalian tissues and often found in cells that do not contain 5-lipoxygenase (5-LO) (Lindgren and Edenius, 1993). Although sequentially not homologous to other epoxide hydrolases (Funk et al., 1987), LTA₄ hydrolase shares homology with puromycin-sensitive aminopeptidase and is a member of the peptidase-M1 family. Sequence comparisons with metalloproteases allowed the further identification of LTA₄ hydrolase as a bifunctional zinc-containing enzyme. In addition to hydrolyzing the triene epoxide LTA₄ to LTB₄, it also displays an aminopeptidase activity. The aminopeptidase activity accepts a variety of substrates from unnatural single amino acid p-nitroanilides to opioid peptides, including the dynorphins and enkephalins, but is thought to preferentially hydrolyze tripeptides with N-terminal arginine (Orning et al., 1994). Compounds that selectively inhibit LTA₄ hydrolase would preferentially block the formation of LTB₄ and thus be an attractive pharmacological target.

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Reaction catalyzed by LTA4 hydrolase.

Early initiatives in the discovery of LTA₄ hydrolase inhibitors were largely confined to compounds that resemble the natural substrate/inhibitor LTA₄ (Djuric et al., 1992; Labaudinière et al., 1992). After the discovery of the zinc-containing nature of LTA₄ hydrolase, bestatin and captopril, inhibitors of Zn²⁺-containing aminopeptidases and angiotensin-converting enzyme, were also shown to inhibit LTA₄ hydrolase (Orning et al., 1991a,b). A series of nonpeptide transition-state analog inhibitors of LTA₄ hydrolase were designed on the basis of their ability to inhibit the aminopeptidase activity (Yuan et al., 1993), and more recently a series of inhibitors that incorporate potential zinc-chelating moieties have been described (Hogg et al., 1995, 1998; Ollmann et al., 1995). Some of these inhibitors display time-dependent kinetics and most of them preferentially inhibit the peptidase activity over the epoxide hydrolase activity (Wetterholm et al., 1995; Hogg et al., 1998). The hydroxamic acid-containing peptide kelatorphan and related analogs were among the most equipotent inhibitors of both the epoxide hydrolase and peptidase activities of LTA₄ hydrolase described to date (Penning et al., 1995). Recently a potent series of inhibitors around a 1-[2-(4-phenylphenoxy)ethyl]pyrrolidine nucleus have been described (Penning et al., 2000; Penning, 2001).

In the present study, we describe a chemically novel and simple nonpeptide compound designated SC-57461A (3-[methyl[3-[4-(phenylmethyl)phenoxy]propyl]amino]propanoic acid HCl) as a potent and selective inhibitor of LTA₄ hydrolase (Fig. 2). The pharmacology of its free acid, SC-57461, is shown for comparison. Inhibition was investigated on several species of purified enzyme and ionophore-stimulated whole blood, and the effects compared.

View larger version (5K): [in this window] [in a new window]   Fig. 2.   Structure of SC-57461A (3-[methyl[3-[4-(phenylmethyl)phenoxy]propyl] amino]propanoic acid HCl).

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Recombinant Human LTA₄ Hydrolase (rhLTA₄) Assay [Enzyme-Linked Immunosorbant Assay (ELISA)]. rhLTA₄ was prepared as previously described (Gierse et al., 1993) and stored at 20°C as a stock solution in 50 mM Tris buffer, pH 8.0, containing 150 mM NaCl, 2.5 mM -mercaptoethanol and 50% glycerol (specific activity = 650 nmol/min/mg). LTA₄ substrate was prepared from the methyl ester in tetrahydrofuran (BIOMOL Research Laboratories, Plymouth Meeting, PA) by room temperature incubation with 30 M equivalents of LiOH for 18 h, and stored at 80°C until used. Enzyme was diluted into assay buffer [0.1 M potassium phosphate, pH 7.4, 5 mg/ml fatty acid free bovine serum albumin (BSA), and 10% dimethyl sulfoxide (DMSO)] and 250 ng (18 nM final assay concentration) in 25 µl was incubated with test compound, also in assay buffer, for 10 min at room temperature. LTA₄ substrate was diluted in assay buffer without DMSO to a concentration of 350 ng/ml and 25 µl (8 ng) was added. The reaction (total volume 200 µl) was allowed to proceed at room temperature for 10 min. A 25-µl sample was added to 500 µl of assay buffer without DMSO to stop the reaction. LTB₄ was quantified in the diluted sample by a commercially available ELISA (Cayman Chemical, Ann Arbor, MI) by using an in-house generated anti-LTB₄ polyclonal antibody, R11 (Gierse et al., 1993).

LTA₄ Hydrolase Kinetic Assay (HPLC). LTA₄ hydrolase kinetics was determined using an RP-HPLC assay to quantify LTB₄. The assay contained 1 µg of enzyme (143 nM final concentration) in 100 µl of 50 mM HEPES buffer, pH 8.0, containing 1 mg/ml fatty acid-free BSA. The reaction was initiated by the addition of 25 µM LTA₄ free acid in ethanol (to a final ethanol concentration of 2%) and incubated at 25°C for 15 to 30 s. The reaction was stopped by the addition of 2 volumes of cold buffer (3:1, 200 mM citrate, pH 3.5/methanol) containing 250 ng of prostaglandin B₁ as an internal standard. Samples were centrifuged for 5 min at 15,000g. The supernatant (200 µl) was applied to a Nova-Pak C18 RP-HPLC column (3.9 × 300 mm) using a computerized Waters HPLC system (Waters, Milford, MA). The column was equilibrated at 1 ml/min with 60% buffer A (60% acetonitrile, 40% methanol, 0.1% acetic acid), 40% buffer B (0.1% acetic acid in water), and eluted over 20 min with a concave gradient to 90% buffer A, 10% buffer B. The eluate was monitored at 270 nm, and the LTB₄ quantified from the ratio of LTB₄/PGB₁ in comparison with a known standard curve.

Peptidase Assay. Aminopeptidase activity was determined spectrophotometrically by the release of the colorimetric product 4-nitroaniline from L-leucine-p-nitroanilide (Leu-pNA) or L-arginine-p-nitroanilide (Arg-pNA) in 50 mM Tris-HCl, pH 7.5, containing 1 mg/ml fatty acid-free BSA in 200 µl. The reaction was started by the addition of recombinant human LTA₄ hydrolase. Absorbance was continuously monitored at 405 nm and the reaction allowed to proceed at room temperature. For the IC₅₀ determinations, 1 mM substrate was used with 1 µg of protein (72 nM final concentration) and the reaction allowed to proceed for 15 min. For the kinetic determinations, the substrate was varied from 0.083 mM to 1 mM, and 50 ng of protein (3.6 nM final concentration) was used with a total reaction time of 60 min. Inhibition kinetic constants (K_i values) were calculated using the program k-cat by BioMettalics (Princeton, NJ) or by the graphical method of Dixon (1972).

Epoxide Hydrolase Assay. Recombinant human cytosolic epoxide hydrolase was assayed using LTA₄ as the substrate. The assay was conducted as described for the LTA₄ hydrolase kinetic assay except the reaction was allowed to proceed for 1 min before quenching. 5(S),6(R)-Dihydroxyeicosa-7E,9E,11Z,14Z-tetraenoic acid was quantitated by RP-HPLC exactly as described for LTB₄.

Assays for Arachidonic Acid-Metabolizing Enzymes. Assays for other enzymes known to metabolize arachidonic acid into inflammatory mediators were performed as previously described. These include human recombinant 5-lipoxygenase (Smith et al., 1995), human recombinant cyclooxygenase-1 and cyclooxygenase-2 (Gierse et al., 1995), and leukotriene C₄ synthase activity in THP-1 cells (Welsch et al., 1994).

Whole Blood Eicosanoid Production. Human blood collected in heparin containing Vacutainer tubes was diluted 1:4 with RPMI-1640 (Invitrogen, Carlsbad, CA) and 200 µl was added per well in 96-well microtiter plates. Compounds diluted in 1% DMSO were added to the blood in duplicate and allowed to incubate for 15 min at 37°C in a humidified incubator (5% CO₂). Calcium ionophore A23187 Calcimycin (20 µg/ml, 1% DMSO final concentration) was added and the incubation continued for 10 min. The incubation was terminated by centrifugation (600-1300g, 10 min, 4°C). Supernatants were analyzed for LTB₄ and thromboxane B₂ (TxB₂) with ELISA by using commercially available reagents (Cayman Chemical) and the in-house-generated anti-LTB₄ polyclonal antibody R11.

Materials. SC-57461 and SC-57461A (>95% purity), zileuton {N-(1-benzo[b]thien-2-ylethyl)-N-hydroxy urea}, and A-78773 {N-[3-[5-(4-fluorophenoxy)-2-furanyl]-1-methyl-2-propynyl]-N-hydroxyurea} were synthesized by the Chemistry Department at Pharmacia Research and Development, Skokie, IL.

Enzymes. Purified enzymes from nonhuman species were cloned, expressed, and purified as described for the human. Clones were obtained from a mouse lung cDNA library for the murine, and a rat spleen cDNA library for the rat. After purification, the specific activities of the epoxide hydrolase were 420 nmol/min/mg for the murine (>95% by SDS-polyacrylamide gel electrophoresis), and 653 nmol/min/mg for the rat (89% pure by SDS-polyacrylamide gel electrophoresis). Recombinant human cytosolic epoxide hydrolase was a generous gift from Bruce D. Hammock (Department of Entomology, University of California, Davis, CA) (Beetham et al., 1993).

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Inhibition of rhLTA₄ Hydrolase. SC-57461A and SC-57461 are both very potent, nonpeptide inhibitors of rhLTA₄ hydrolase when either LTA₄ or peptidase substrates are used (Table 1). No inhibition of rhLTA₄ hydrolase was observed with either the first- (zileuton) or second-generation (A-78773) 5-LO inhibitors (Carter et al., 1991; Harris et al., 1995), demonstrating selectivity of these inhibitors for the 5-LO enzyme. In contrast to previously described LTA₄ hydrolase inhibitors (Wetterholm et al., 1995; Hogg et al., 1998), SC-57461A is equipotent as an inhibitor of the hydrolase and peptidase activities. SC-57461A is equipotent with kelatorphan as an inhibitor of LTA₄ hydrolase (Penning et al., 1995).

View this table: [in this window] [in a new window]   TABLE 1 Inhibition values for rhLTA4 hydrolase IC50 = nanomolar concentration ± S.E. Most values were generated from three or more experiments as indicated in parentheses. IC50 values for LTA4 were determined using the ELISA assay. The Ki was determined using the HPLC assay. IC50 values were calculated using a four-parameter log fit. The Ki values were fitted using a nonlinear weighted regression routine to fit primary velocity vs. substrate data or by the graphical method of Dixon (1972) as discussed in the text.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Competitive inhibition of rhLTA4 hydrolase by SC-57461. One microgram of rhLTA4 hydrolase was incubated with increasing amounts of SC-57461 (, 0 nM; , 50 nM; , 100 nM; , 200 nM; , 300 nM; , 400 nM) as described under Experimental Procedures. The inhibition constant derived from the graphical method of Dixon for SC-57461 was Ki = 23 nM.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Time-independent inhibition of rhLTA4 hydrolase by SC-57461A. rhLTA4 hydrolase was preincubated with varying concentrations of inhibitor for the indicated periods of time. LTA4 was added to start the assay or the enzyme was added last for time 0, and the reaction quenched as described under Experimental Procedures. LTB4 was quantified by RP-HPLC. Preincubation times: , 0 min; , 30 s; and , 5 min.

Specificity of Enzyme Inhibition. SC-57461 was tested for its specificity to inhibit other enzymes in the arachidonic acid cascade. It did not inhibit recombinant human 5-lipoxygenase, recombinant human cyclooxygenase-1, or recombinant human cyclooxygenase-2 at 100 µM or LTC₄ synthase activity in THP-1 cells at 100 µM (data not shown). SC-57461 was also tested as an inhibitor of recombinant human cytosolic epoxide hydrolase (Haeggström et al., 1986; Beetham et al., 1993). By using LTA₄ as the substrate, the IC₅₀ was determined to be 300 µM (Fig. 5). This >1000-fold selectivity of SC-57461 for LTA₄ hydrolase over an unrelated cytosolic epoxide hydrolase is further demonstration of its specificity. In addition, SC-57461 at 100 µM did not inhibit other metalloproteases, including human leukocyte elastase, human liver cathepsin B, rabbit lung angiotensin-converting enzyme or bovine spleen cathepsin D (data not shown). A structurally related LTA₄ hydrolase inhibitor, SC-54581, did not inhibit neutral endopeptidase at 100 µM (data not shown).

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Specificity of inhibition of SC-57461 on LTA4 hydrolase versus cytosolic epoxide hydrolase. Enzyme used: , recombinant human cytosolic epoxide hydrolase, IC50 = 300 µM; , recombinant human LTA4 hydrolase, IC50 = 0.10 µM.

Inhibition of Whole Blood Eicosanoid Production. The activity and selectivity of SC-57461A and SC-57461 were also demonstrated in whole blood under conditions in which both the 5-LO and cyclooxygenase pathways were stimulated simultaneously with calcium ionophore. SC-57461A and SC-57461 were potent inhibitors of LTB₄ production in whole blood from a variety of species (Table 3). Both compounds (IC₅₀ values in human blood, 49 and 65 nM, respectively) are similar in potency to A-78773 in human and rhesus blood. SC-57461A is less potent than A-78773 in the mouse, rat, and dog. In contrast to the purified enzyme, SC-57461 was 3-fold less potent on murine blood compared with human, but the rat/human ratio of activity was similar.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Previous investigators have followed a two-pronged approach in the search for potent LTA₄ hydrolase inhibitors. The first approach consisted of inhibitors based on the LTA₄ substrate. Djuric et al. (1992) synthesized compounds based on an oxabicycloheptene nucleus in an attempt to mimic the vinyl epoxide of the natural substrate. These compounds did not inhibit the isolated enzyme but inhibited LTB₄ biosynthesis in the HL-60 cell line. Labaudinière et al. (1992) synthesized a series of -arylalkanoic acid derivatives in an attempt to mimic the allylic epoxide of LTA₄. The most potent of these analogs displayed IC₅₀ values in the low micromolar range against porcine leukocyte LTA₄ hydrolase.

A second approach focused on the peptide isostere after the discovery that LTA₄ hydrolase is a zinc-containing enzyme displaying aminopeptidase activity. Two series of compounds evolved from the backbones of bestatin and captopril (Yuan et al., 1991, 1992, 1993). In the first series, the investigators used the norstatine backbone of bestatin to build potential peptide transition state analogs. This series eventually evolved to incorporate -keto--amino esters, which in an aqueous environment hydrated to the gem diol, capable of coordinating with the Zn²⁺ in the active site. The most potent of these compounds were shown to be relatively selective for the aminopeptidase activity of LTA₄ hydrolase.

Other zinc binding motifs have also been explored as LTA₄ hydrolase inhibitors. Hydroxamic acids have been shown to be potent inhibitors of LTA₄ hydrolase. Hogg et al. (1995, 1998) reported a series of inverted hydroxamates to be low nanomolar inhibitors of the aminopeptidase activity, but less potent for the epoxide hydrolase activity. Kelatorphan has been shown to be equipotent as an inhibitor of both of the LTA₄ hydrolase activities, but it is not selective for LTA₄ hydrolase (Penning et al., 1995). A very potent -mercaptoamine (Yuan et al., 1993; Ollmann et al., 1995) has been described, but it also inhibits thromboxane B₂ production in human whole blood. Again, it is assumed that part of its inhibitory activity is due to the zinc-thiol interaction.

SC-57461A is the first reported compound with a structure unrelated to the natural substrate LTA₄ or a peptide isostere to possess potent inhibitory activity against LTA₄ hydrolase. Its IC₅₀ of 2.5 nM for the epoxide hydrolase activity makes it one of the most potent inhibitors reported to date. It is at least 50-fold more potent against the epoxide hydrolase activity than the most potent inhibitors reported by Yuan et al. (1993), Hogg et al. (1995, 1998), and Labaudinière et al. (1992); as well as captopril and bestatin (Orning et al., 1991a,b; Baker et al., 1995). It is also unique in that its potency was maintained against the aminopeptidase activity, as exhibited by a 2 nM K_i by using arginine-pNA as the substrate. In contrast to the previously published LTA₄ hydrolase inhibitors, SC-57461A represents a truly selective class of inhibitors that does not inhibit other zinc-containing aminopeptidases, including human leukocyte elastase and rabbit lung angiotensin-converting enzyme or other enzymes in the arachidonic acid cascade. Please see Penning (2001) for additional information (chemistry, biology, toxicity) regarding this series of LTA₄ hydrolase inhibitors.

The mechanism by which SC-57461A inhibits LTA₄ hydrolase is competitive as illustrated in Fig. 3. To properly characterize the inhibition, the epoxide hydrolase data were analyzed by the graphical method of Dixon (1972) rather than the more conventional Lineweaver-Burke analysis. This was necessary because of the high enzyme concentration (143 nM) necessary to run the assay relative to the potency of the inhibitor. To minimize this effect, the K_i in the aminopeptidase assay was determined using an enzyme concentration of 3.6 nM. Data analyzed by both conventional methods and the graphical method of Dixon (1972) provided similar results. SC-57461 was also shown to act as a time-independent inhibitor because no enhanced inhibition of enzyme activity was evident, even after extensive preincubation of the inhibitor with the enzyme (Fig. 4).

Multiple isozymes of LTA₄ hydrolase have been reported. Erythrocytes and B-lymphocytes have each been reported to contain a second subtype that distinguished itself by its divergent kinetic properties (McGee and Fitzpatrick, 1985; Orning et al., 1990; Odlander et al., 1991). Bigby et al. (1994) reported that epithelial cell-derived LTA₄ hydrolase present in bronchoalveolar lavage fluid lacked aminopeptidase activity and had unique kinetic properties. They found that the profile of inhibition of epithelial cell-derived LTA₄ hydrolase by certain metalloproteinase inhibitors differed from that of the neutrophil enzyme (Baker et al., 1995). Partial N-terminal sequence determinations of the airway epithelial-derived enzyme failed to show any sequence divergence over the amino terminal 17 amino acid span (Ned Seigel, Pharmacia, personal communication). The recent report of an alternative splice site in the LTA₄ hydrolase gene would support these findings (Jendraschak et al., 1996). The specificity of SC-57461A for the postulated subtype is as yet unclear.

SC-57461A retained good potency in inhibiting LTB₄ production in human and rhesus monkey whole blood. These data suggest that SC-57461A readily penetrates the appropriate cells and interacts with ubiquitously expressed LTA₄ hydrolase. These data also show that SC-57461A is fairly metabolically stable in blood and binding to plasma proteins does not negate efficacy. SC-57461A is slightly less potent in rat and dog whole blood. This may reflect a species-dependent decrease in intrinsic activity against the enzyme as supported by the decrease in potency against the isolated rat enzyme, rather than increased instability in whole blood of different species.

SC-57461 and SC-57461A have also been assessed in several animal models. SC-57461 was shown to be orally active with a pharmacodynamic half-life exceeding 24 h. SC-57461 was shown to inhibit LTB₄ production in a rat model of ionophore-induced peritoneal eicosanoid production, in a rat reversed passive Arthus model, and in an arachidonic acid-induced ear edema model in mice (Kachur et al., 2002). SC-57461 and SC-57461A displayed selective inhibition of LTB₄ in a dose-dependent manner in these models.