European Jo~wmtlof Pi~a~acolo~,

205 f 1991)259-266 0 1991 Elsevier Science Publishers B.V. All rights reserved 0014-2999/91/$03.50

WP 52161

OSY

esisi

Patrick R. Young, Randy L. Bell, Carmine Lanni, James B. Summers, Dee W. Brooks and George W. Carter Immwtosciences Research Area, Abbott Laboratories, Abbott Park, IL 60064. U.S.A.

Received 27 June 1991, revised MS received 3 September 1991, accepted 10 September 1991

In the search for a model of leukotriene (LT) production to provide a method to determine in vivo 5lipoxygenase (S-LO) inhibitory astivity by various compounds, a passive anaphylactic reaction in the rat peritoneal cavity was examined, refined and characterized. The reaction, produced by passive sensitization with an i.p. injection of ratbit anti-bovine serum albumin (anti-BSA) followed by an i.p. injection of DSA, resulted in the biosynthesis of large amounts of sulfidopeptide LTs measurable by immunoassay or by reversed phase high performance liquid chromatography. The oral activity of several 5LO inhibitors has been examined using this model. An example of these is zileuton (Abbott-64077), a potent 5lipoxygenase inhibitor now under clinical evaluation. Zileuton inhibited sulfidopeptide LT biosynthesis in the rat peritoneal cavity in a dose-dependent manner (ED,, = 3 mg/kg). WY-49,232, MK-866, BW A4C and phenidone also produced good activity with ED,, values of 6,8, 11 and 17

mg/kg, respectively. This modified rat peritoneal anaphylaxis model appears to be a valuable tool for establishing in vivo activity of S-LO inhibitors. Leukotriene biosynthesis; 5-Lipoxygenase inhibition; Zileuton; Peritoneal anaphylaxis model (rat); (In viva activity)

1. Introduction

Bioconversion of arachidonic acid by 54ipoxygenase (5-LO) leads to the production of several very potent biological substances believed to be invoived in various pathophysiological states. Among these putative mediators are the sulfidopeptide leukotrienes (LTC,, LTD, and LTE,; slow reacting substance of anaphylaxis; SRS-A) which are potent constrictors of airway smooth muscle and are believed to be involved in asthmatic conditions (Lewis and Austen, 1981). LTB, is also generated by the 5-LO pathway and has been shown to irfluence inflammatory cell locomotion and secretory responses (Samuelsson, 1983; Sirois, 1985; Garcia er al., 1987). Orally effective inhibitors of 5-LO could represent a new class of therapeutic agents to treat allergic and inflammatory diseases. Because of the complexity of the diseases such as asthma and arthritis and the inadequacies of experimental animal models designed to mimic them, a model capable of demonstrating in vivo inhibition of LT formation would be a useful, primary method to define the oral 5-LO inhibitory activity of compounds.

Correspondence to: P.R. Young, Department

47K, AP-9, Abbott Laboratories, Abbott Park, IL 60064. U.S.A.. Tel. 1.708.937 3812 or 937 9841, fax. 1.708.938 5034.

SRS-A was shown by bioassay procedures to be released in the rat peritoneal cavity in response to an antigen-antibody reaction (Rapp, 1961; Orange et al., 1968; Smith et al., 1980). With the technology now available to measure the specific LTs by radioimmunoassay (RIA) (Hayes et al., 1983), enzyn,e immunoassay (EIA) (Miller et al., 1985) or by reversed phase high performance liquid chromatography (RPHPLCI (Murphy et al., 1979; Lewis et al., 19801, this type of experimental model is appropriate to demonstrate acute in vivo 5-LO inhibitory activity. The characteristics of the model designed to evaluate LT biosynthesis inhibitors are described in the following report as well as the oral activity of several compounds. Also described is the activity of zileuton, a potent 5-LO inhibitor (Carter et al., 1991) which was discovered at Abbott Laboratories and is now undergoing clinical trials.

2. Materials and methods 2.1. Rat peritoneal anaphykxis

model

The general methodology described below was found to consistently produce high levels of immunoreactive leukotrienes in the rat. Several variations in timing,

and antigenobjected and the oB‘ ~~~~~~t~~ of su~fido~ptid~ leukotriene ism \~erc‘ studied. The effects of modifying parameters will be described in the appropriate

was withheld for about 18 h ey derived rats (SASCO inc., we~~b~ng a~p~imately 200 g. Each rat sensitized by an i.p. injection of rabbit SA or the IgG fraction of this antisera A) prepared at Abbott Laboratories as deby Csrter et al. (1952). Three hours after ~~~iti~at~o~. the rats were injected i-p. with 4 mg BSA tion V.. ICN ~mmu~QBioiogi~ls. Costa Mesa, CA). BxR sensitization and challenge injection volumes were 5 ml in 2Q PM potassium phosphate buffered saline 1, pH 7.1. ~~rn~~~~~tested for inhibitory activity were a~~~istere$ by gavage (p.0.) 1 h prior to antigen chafknge tmiess othemise stated. The rats were killed with carbon dioxide 15 min after chalienge. the peritoneal cavity opened and the fluid Lu)ntents collected with a plastic trocar and disposabIe pipets. The cavities were rinsed with 5 ml of ewld PBS containing 0.1% gelatin, 0.1% sodium azide and 10 mM t~~ta~~urn EDTA. The peritoneal fluid and the wash were transferred directly to tubes containing 20 ml of ice cold methanoi, allowed to stand for 20 min. vortexed and then centrifuged at 1000 X g for 15 min. The fluid volumes were recorded and the supernatants stored frozen until analyzed for Ieukotrienes by RIA or EIA. Exceptions to this were the studies in which histamine and thromboxane (Tx.B,) were measured. in these studies the peritoneal fluids were ~ent~f~ged prior to methanol precipitation to remove cellular content. ln a study to evaluate concurrently the ex vivo activity of zileuton, 3-4 ml of blood were collected by cardiac puncture with heparinized syringes from each rat immediately following the cotlection of peritoneal fluids. Within 15 min, aliquots of each blood sample were incubated at 37 o C for 30 min with 50 p M calcium ionophore, A23187 (Sigma Chemical Co., St. Louis, NOI. A set of samples from the vehicle control group was incubated with dimethylsulfoxide (DMSO) to serve as a non-stimulated control. incubations were terminated by placement in an ice bath for 5 min. The samples were centrifuged and the plasma added to 4 VO~IIIES of cold m~thanoi to precipitate proteins. The methanol extract was diluted with assay buffer and LTB, measured by ETA. to testing,

frc@Q

2-L lmmmoassuys The antibody (Ab-1) incorporated in the LTC, RIA kit (New England Nuclear, Boston, MA) used in many of our studies had a cross-reactivity of 100, 55 and 9%

for LTC,, D_, and E,, respectively. L-Cysteine was added to the BSA challenge solution and to the gelatin wash buffer at a concentration of 30 mM in the studies which were assayed using this antibody (method 11. This essentially prevented metabolic conversion of LTD, to LTE, so that the products measured were predominantly LTC, and LTD,. in later studies, a RIA kit (Amerskam Corp., Arlington Heights, IL) that contained an antibody (Ab-2) with a cross-sensitivity of 64% to LTE, was used and L-cysteine was not added to the fluids (method 2). Most recently, we have generated a rabbit antisera to LTD, (Ab-3) for RIA evaiuations which has 100, 100 and 88% cross-reactivity to LTC,, LTD, and LTE,, respectively. Specific LTE,, LTB, and TxB, EIA kits available from Cayman Chemical, Arm Arbor, Ml and a [‘ZSI]RiA for histamine (AMAC, inc., Westbrook, ME) were also used in several of the studies. The methanolic supernatants were diluted 1: 3 (v/v) with PBS containing 0.1% w/v gelatin, 0.1% sodium azide and 10 mM EDTA for assay. Methanol (22%) was included in the assay standards to compensate for the amount of alcohol in the biological samples. The RiAs or EiAs were conducted as recommended by the manufacturers with the exception that volumes were adjusted to accommodate automation using a Cetus Propette (Cetus Corp., San Francisco, CA). in LT RIA studies, radioactivity was quantitated with a liquid scintil!ation counter (Packard Instrument Co., Downers Grove, IL) and the immunoreactive LT levels in the biological samples were calculated from a LTC, standard curve with a useful range of 0.2516 ng/ml and expressed as ng of LTC,-like activity Imethods 1 and 2 with Ab-1 and 2). in recent studies with Ab-3, values were calculated from a standard curve with LTE, in the same range of sensitivity and expressed as LTE,-like activity. in EIA studies, color development in 96 well plates was determined by a microplate reader (Molecular Devices Corp., Menlo Park, CA) comparing samples to LTE,, LTB, or lldehydro-Tag standard curves. The histamine assay was evaluated using a ANSR gamma counter (Abbott Laboratories, North Chicago, IL).

2.3. Sumpfe extraction and RFHPLC analysis The individual rat methanolic lavage samples tapproximately 30 ml) were diluted with 20 ml of 0.2% phosphoric acid (pH 1.9) and aspirated through 1 ml Cl8 Bond-Elute~ #lumns (~alytjchem Intern., Harbor City, CA) with vacuum. The columns were washed with 10 ml of 35% methanol in 0.04% phosphoric acid (pH 2.4) followed by elution of products with 1 ml of 70% acetonitrile in water. Using a mobile phase of 40% acetonitrile in 5 mM triethylamine (TEA) acidified to pH 3.3 with formic acid at a flow rate of 1.0

ml/min, the products were separated on a Cl& 5~ Radial-Pako cartridge (Waters Assoc., Milford, MA) and quantitated by UV absorbance at 280 nm (tritiated LT standards were obtained from New England Nuclear and the cold LT standards were obtained from either Bio-Mega, Inc., Laval, Canada or Biomol Research Laboratories, Philadelphia, PA). 2.4, Leukotriene biocon rlersion The profiles of the sulfidopeptide leukotrienes in rat peritoneal fluids receiving various treatments of the aminoipeptidase inhibitor, L-cysteine, and the yglutamyl transpeptidase inhibitor, serine borate, were examined by uv RP-HPLC following the described procedures. These enzyme inhibitors were administered either separately or in eombination by incorpo-

rating them in the antigen and lavage solutions. 2.5. Compound abbreviationsand preparation The following compounds were prepared by Abbott research chemists or purchased from Sigma Chemicals: N-all-benz~b~hien-2-ylethyl)-N-hydro~rea (zileuton; A-64077); N-(3-phenoxy-cinnamyl)-acetohydroxamic acid (BW A4C); 3-amino-l-(m-tri~uoromethyi)phenyl~2-pyrazoline (BW 75X); l-phenyl-3-pyrazolidinone (pllenidone~ 3-(3-(~chiorobe~l~-3-t-bu~lthi~5-i~propyhndol-2-yl)-2&dimethylpropanoic acid (MK-886; L-663,536); nordihydroguaiaretic acid (NDGA); LYmethyl-6-~2-quino~inylmetho~~-2-naphtha~ene acetic acid (WY-49,232). 2,3,5-Trimethyl-6-(12-hydroxy-5,10d~ecad~yl)-1,4-~nz~uinone (AA-861) was a gift from Takeda Chemical Industries and (Z)&chloro2,3-dihydro-3-(hy~o~-2-thieny~methyiene~2-ox~lHindole-1-carboxamide, sodium salt (Tenidap; CP66,248) was a gift from Pfizer Central Research. Agents evaluated for 5-LO inhibitory activity were prepared just prior to use as suspensions in 0.2% methylcellulose in water for oral administration or 0.2% methylcellulose in PBS for i.p. injection and vigorously mixed with a motorized homogenizer equipped with a Teflon pestle (Cole-Parmer Instrument Co., Chicago, IL). Qn occasion, a drop of Tween 80 was added to prepare a homogenous suspension.

3. Results 3.1. Characterizationof the model The production of SRS-A by an IgG-based reaction as described by Orange et al. (1968) was utilized as the starting point for the development of a model to evaluate 5-LO inhibitors. The reaction was examined in a number of different ways. Once we established that

q 10

line

20

30

40

After Challenge (mln)

1. Effects of varying the amount of antigen (panel A) and time peritoneal fluid collection after challenge with 4 mg antigen (panel on mean +S.D. LTCJike levels 0%1) generated in the rat ritoneai cavity (no L-cysteine was used in these earIy studies). The :s were passively sensitized with 1 mg rabbit anti-BSA i.p. (n = 4-j) and challenged 3 h later with BSA i.p.

$.

rabbit anti-BSA could be substituted for homologous rat hyperimmune sera for passive sensitization, we evaluated the amount of antibody, the time interval for sensitization and the time of sampling after challenge to optimize conditions for drug testing. In addition, several studies were conducted to determine the profile of sulfidopeptide LTs present in the peritoneal fluids to optimize immuno~say sensitivity. Initial studies indicated that consistent immunoreactive sulfidopeptide LT levels were obtained using a 3 h interval between the i.p. injections of antibody and antigen. The effects of the amount of antigen injected and the time of peritoneal fluid colfection after challenge appeared to be critical (fig. 1). The peak reaction reached a plateau between challenges of 4-8 mg of BSA in rats sensitized with 1 mg anti-BSA. The peak LT levels were seen between 15 and 40 min after challenge. In a subsequent study, LT levels 2 h after challenge (63 ng/rat) were approximately 53% of those seen at 15 min (115 ng,/rat). Following these early studies, we found that rabbit anti-BSA sera with a titer of greater than 5 mg antib~y/m~, diluted 1:30 with

rata

~tisesfmmthe scquttntial hydrolysis of

4

0

LTE4

q

LT04 LTC4

g&q1 r&dues from LTC, and LTD,, ~s~c~~~~ly~ 0~ would expea that the peritconcal cavcq.?acity FQ tIXF&Oiize LTC, once The high levelsof LT present in the

ity hxi ~~~~~~~~~~

ie was ~~~~. ItmeaE FfUids

made it pussibte to examine the proof the ~~ti~Ieukotrieues by RP-MPLC analysis of ~~~~~~~~ tumid e~Fracts from individual rats. The pre~~i~~~~ LT present 15 min after the a~~phyiacFic LTE, (fig. 2). This was confirmed by ILK, into a rat’s peritoneal cavity, solIecting the periFonea$ fluids 15 min later and analyzing be ~~~c~ prsing idiomatic RI’-HPLC ~Radiomatic nsmen~s. Tampa. FL). There was a 64% conversion to ~~~~T~~ in the sample retrieved from the rat c~~~cd to materid placed in peritoneal fluids immediaFely prior to the addiFion of cold methanol. No

5

10 Retcnlion Tii

15

20

25

(minmes)

Fig. 1 Reversed phase HPLC analysis of the rat peritoneat fluids 15 tin after i.p. challenge with antigen (panel B). Synthetic standards of the ~fido~~t~de Ieukotrienes (recovery of tritiated standards thnougb extraction was approximately 70-85%“c)are shown in panel A After protein precipitation, the sample was extracted with a solid phase Cl8 cartridge and injected on to a Cl8 column using a mobile phase of 4% acetonitrife and 5 mM TEA-formate, pH 3, at a flow rate of 1 ml/min. The predominant peak seen in the lavage fluids was LTE,.

Fig_ 3. The addition of L-cysteine and serine borate to the chaiienge and to the wash fluids significantly altered the LT product profile at the 15 min time period as assayed by RP-HPLC, Without inhibitors (A), the major product was LTE,. L-cysteine, 30 mM. effectively prevented the conversion to LTE, (B) while 10 mM serine borate partially prevented the breakdown of LTC, (0. This activity was also visible when both enzymatic inhibitors were added fD) at these respective con~ntrations.

[ -‘Y]LTD, was detected suggesting that the conversion of LTDJ to LTE, occurs very rapidly. Since there was a low cross-reactivity (9%) of the RIA antibody fAb-1) to LTE,, studies were conducted FO examine the effects of L-cysteine and serine borate, two inhibitors of sulfidopeptide LT metabolism (Anderson et al., 1982; Krilis et al., 1983). Figure 3 shows the effects of the metabolic inhibitors L-cysteine and serine borate complex on the product profiles when included in the antigen injection and the wash fluids. Although prevention of the conversion of LTC, to the other products was accomplished with serine borate, it was not complete, whereas, L-cysteine appeared to prevent the majority of LTD, from converting to LTE, consistently, as reported by Wei et al. (i986). The addition of serine borate complex to the antigen solution also appeared to cause considerable irritation when injected into the rats. It is not known whether this was a response to the increased levels of LTC, or due to the complex itself but it appeared to bz an uncontrollable variable. For these reasons, evaluation of test agents using the RIA Ab-1 was done in the presence of L-cysteine in the challenge and wash solutions to safely optimize the amount of immunodetectable LTs measured in this model. In studies using RIAs or EIAs containing antibodies that have greater cross-reactivity for LTE, or are specific for this leukotriene, L-cysteine has been eliminated from the procedure without a significant change in inhibitory activities. 3.2. inhibition of leukotriene biosynthesis Zileuton is a highly effective inhibitor of 5-LO with an IC,, of 0.5 FM against the enzyme from rat basophilic leukemia cells (Carter et al., 1991). This was approximately 4-6 times greater than the potency of phenidone and BW 75X. When tested for oral effec-

263 TABLE I Comparison of the inhibitory effects of zileuton and reference agents in the rat peritoneal anaphylaxis model. Compound

Oral EDs, mg/kg a

Immunoassay

Zileuton WY-49232 MK-866 BW A4C Phenidone BW755C Tenidap NDGA Nafazatrom AA861 Hydrocortisone In~methacin Ibuprofen Cyproheptadine

3 (2- 4) 6 (4-10) 8 (2-141h I1 (8-161 17 (10-281 36 ( 14-521 52% at 100 mg/kg N.A. at 100 mg/kg ccf N.A. at 100 mg/kg d N.A. at 100 mg/kg d N.A. at 30 mg/kg e N.A. at 20 mg/kg N.A. at 100 mg/kg N.A. at 10 mg/kg

RIA method EfA-LTE, RIA method RIA method RIA method RIA method EIA-LTE, RIA method RIA method RIA method RIA method RIA method RIA method RIA method

1 2 2 1 1 1 1 1 2 1 1 2

’ Dose calculated by linear regression to produce 50% inhibition of LT biosynthesis with 95% confidence limits in parenthesis or effect at highest dose tested. b Compound was administered 4 h prior to challenge. ’ N.A. = produced no significant effects (P > 0.05). d NDGA, nafrazatrom and AA861 produced significant inhibition of 40, 51 and 86%, respectively, when administered at 1 mg/kg i.p. e Produced no significant inhibition at 30 mg/kg administered 1 h or two doses at 20 and 1 h prior to challenge.

tiveness as inhibitors of LT generation in the rat peritoneal model, zileuton, phenidone and BW 755C produced dose related inhibition of the levels of LTC,-like material (method A). Linear regression analysis calculated the 50% inhibitory doses (ED,,) to be 3, 17 and 36 mg/kg for zileuton, phenidone and BW 755C, respectively. Table 1 compares the effect of these compounds and several reported inhibitors of LT synthesis in the rat peritoneal model. Carlson et al. (1989) reported that WY-49,232 was an orally active S-LO inhibitor and effectively inhibited LTC, biosynthesis (ED,, = 12 mg,/kg p.o.1 in mice using a zymosan peritonitis model. In the rat, this compound reduced the amount of LTE, present in the peritoneal cavity after an anaphylactic reaction with an ED,, of 6 mg/kg p.o. BW A4C, a compound reported to reduce LTB, levels in inflammatory exudates in rats (Higgs et al., 19881, also effectively inhibited LTE, levels (ED,, = 11 mg/kg p.o.1 in the rat peritoneal anaphyl~is model. MK-866 was reported to be a potent leukotriene biosynthesis inhibitor, not a direct enzyme inhibitor, with a long duration of activitY (Gillard et al., 1989). Initial studies in the rat model using a 1 h pretreatment time gave variable results, however, when a longer pretreatment time (4 h) was used, the compound exhibited good potency (ED,, = 8 mg/kg). Tenidap, which was shown to inhibit cyclooqgenase and 5-LO (IC,, = 0.01 and 18.0 FM, respectivelyf in purified cell populations, but not 5-LO in whole blood, was recently tested in 10 patients with rheumatoid

arthritis at a daily dose of 120 mg (Blackbum et al., 19911. The ex vivo production of LTB, by synovial exudates was inhibited in six patients, however, when the observed decrease was corrected for neutrophils in the fluids (LT13/106 PMNI, the change was not sign& cant. It is likely that tenidap reduced exudate Lm, pr~uction by reducing the influx of PMN into the synovial space rather than directly inhibiting 5-LO. Since the rat peritoneal reaction is acute and mainly dependent on resident cells, we felt it of interest to test tenidap in this model. Tenidap produced a mild effect on LT biosynthesis in the rat (table 1). Significant inhibition (52%) was produced at 100 mg/kg p-o. but the compound was inactive at 30 mg/kg, in fact, an interesting obse~ation at this dose was that the peritoneal LTE, levels were significantly higher (43%) than those of the control group. A&861, one of the first 5-LO inhibitors described, was reported (Mono et al., 1986) to inhibit an eariy pleural Arthus reaction in the rat when injected intrapleurally with the antiserum. AA-861 and the standards NDGA and nafazatrom, were not orally effective in the peritoneal model but did produce significant inhibition when injected i.p. (table 11. This possibly reflects poor bioavailability in the rat by the oral route. The cyclooxygenase inhibitor, indomethacin, and the steroid, hydrocortisone, did not si~i~cantly effect the LT levels in the peritoneal cavity when administered at relatively high doses compared to the antiedema activity produced in the pleural carrageenan iRfl~ation rat model (Haviv et al., 19881. Although most of the described drug evaluations were based on the inhibition of the sulfidopeptide leukotrienes, LTB, levels also were increased in the peritoneal cavity during the anaphylactic reaction and were about 3-fold higher than in rats that were not passively sensitized. The amounts of LTB, were appro~mately one third that of LTE,. The peritoneal fluid in rats in the 15 min anaphylactic reaction also contained levels of TxB, and histamine which were significantly higher (approximately 1.5 fold) than those of non-sensitized rats injected with BSA (fig. 4). The oral inhibitory activity of zileuton on levels Of LTB, and LTE, in the same rats is shown in fig. 5. BY linear regression analysis, no significant difference was evident between the inhibition of LTB, or LTE, in the peritoneal fluids (ED,, = 4 and 7 mg/kg, respectiveb) by the compound. Conversely, the ex viva inhibition by zileuton on LTB, levels in the plasma of ionophorestimulated whole blood was si~i~cantly more potent with an ED,, of 1.4 mg/kg (fig. 5). It is reasonable to assume that the increase in potency in the ex viva study is due to higher concentrations of drug circulating in the blood as compared to levels reached at the tissue site responsible for the leukotriene synthesis ain the peritoneal cavity. Zileuton also showed specificity of

0.5 Lx?3

LTB4

TxBZ?

Histamine

1

2

4

8

Pretreatment Time (Hrs)

Fit. 4_ A proftk of LTE,. LTB,. thromboxane B2 (determined by EPA) and histamine Ervels (determined by RIA) in rat peritoneal fluids after a chrtlknge injection of BSA to non-sensitized rats (solid bars) or to passiveiy sensitized rats (striped bats).

inhibitoq activity in the rat peritoneal mode! in another s&r&. Increases in thromboxane or histamine were not effected by pretreatment with 70 mg/!cg p-o. of zileuton whi!e ED, values for peritoneal levels of LTE, and LTB, (determined by specific EIA! were 5 and 6 mg/li_g. respectively. The rat peritoneal mode! provided a simple technique to establish the duration of the inhibitory effects of compounds after a single on! dose (fig. 6). By varying the time of administration of zileuton prior to the antigen challenge. significant inhibitory activity was obtained for at least 4 h. The mode! can also be used to determine if inhibition is maintained after daily oral dosing. Table 2 shows that there was no difference in

Fig. 6. The duration of inhibitory activity by zileuton was studied by varying the time the drug was gavaged (10 mg/kg) prior to the antigen challenge. Each bar represents the mean and standard error of a different group (n = 7-S) pretreated with either 0.2% methylcellulose (stippled bars) or zileuton (striped bars). Significant differences between the means of LTC, levels (method 1) of the treated groups versus the controls occurred at all but the 8 h time (*P < 0.05).

TABLE 2 Activity of zileuton following daily U-day) dosing in the rat peritoneal anaphylazis model. Oral treatment

N i-Leukotrienes Inhibi(mean, ng/rat) a tion

Vehicle (5 doses) 4 ml/kg per day 10 195.7*21.4 Zileuton (5 doses) 10 mg/kg per day 10 77.0+ 11.2 b 10 85.7* 8.8 b Zileuton (1 dose) ’ 10 mg/kg

60.7% 56.2%

a Immunoreactive LTC,-like levels determined by method 1; mean +_S.E.M. b Significantly different from vehicle control but not from each other, P < 0.05, using analysis of variance and Duncan’s multiple range test. c Rats received four doses of vehicle, 4 ml/kg per day, prior to the test.

the inhibitory effectiveness of zileuton between acute dose and a series of five daily doses.

an

4. Discussion

-20 0.1

1

10

100

OralDosam@g Fig. 5. inhibitory effects of zileuton on the biosynthesis of LTE, (triangles) and LTB, (squares) in the rat peritoneal anaphylaxis model (both had similar s.d. and only those ior LIB, are shown for clarity). The ccmpound was administered 1 h prior to the i.p. antigen challenge and the peritoneal fluids collected 15 min after the challenge. Also shown are concomitant ex vivo effects (circles) on LTB, synthesis generated by ioaophore stimulation of the whole blood collected at sacrifice (the s.d. of the high inhibitory values are within the symbols). LT levels were determined by speciftc EIA.

We have found the rat peritoneal anaphylaxis reaction resulting from BSA challenge to passively sensitized rats to be a simple and rapid method to evaluate putative 5LO inhibitors for oral in vivo activity. In a rapid response to antigen challenge, substantial amounts of leukotrienes are synthesized with the major measurable product being LTE,, presumably resulting from a rapid metabolism of LX? and LTD,. The IgG-mediated reaction appears similar in product formation and peak appearance of the sulfidopeptide LTs as described for an IgE-mediated reaction. The white blood cell counts of the peritoneal fluids 15 min after challenge are low and the cells obtained are mainly mononuclear cells with a small percentage of mast cells. This could mean that the products are coming from resident macrophages and mucosa! mast

265

cells as suggested for the IgE response (Wei et al., 1986). Several models have been used to demonstrate in vivo 5-LO inhibitory activity by compounds. Some of these center on physiological parameters such as arachidonic acid induced mouse ear edema (Jones et al., 1986) and in~ammato~ exudate leukocyte accumulation in the rat (Higgs et al., 1988). In a rat Arthus pleurisy model, Berkenkopf and Weichman (1991) compared the effects of several agents, including zileuton, on inflammatory exudate volume and cell influx. The rat peritoneal model described provides another, excellent method to examine 5-LO inhibitory effects by agents, in vivo. Relative potencies can be compared by demonstrating biochemically, the suppression of LT synthesis in the intact animal. This model and the others directed at evaluating the ability of 5-LO inhibitors to attenuate pathophysiological effects of the LTs should play important roles in the discovery of potent and selective S-LO inhibitors to identify candidates for clinical development. Zileuton demonstrated potent in vivo activity by the oral route using the rat peritoneal model. The compound produced inhibition of both LTE, and LTB, synthesis but was inactive against the increase of levels of th~m~xane and histamine. Phenidone, BW 755C and more recently discovered agents also produced oral inhibitory activity of LT synthesis while other 5-LO inhibitors did not, presumably because of a lack of bioavailabilty in the rat by this route of administration. Zileuton was absorbed rapidly and produced peak plasma levels between 15 min and 2 h at a dose of 20 mg/kg p.o. in the rat (Carter et al., 1991). Plasma levels could be detected for over 6 h and the duration of inhibition of LT biosynthesis in the peritoneal anaphylaxis model, shown in fig. 6, is consistent with this. As for therapeutic utility, early clinical trials with zileuton have shown a reduction in LTB, concentration in the rectal dialysis fluid from patients with ulcerative colitis (Laursen et al., 1990), amelioration of the asthmatic response to cold dry air (Israel et al., 1990) and attenuation of allergen-induced nasal congestion (Knapp, 1990). These reports are encouraging and imply that continued in-depth clinical studies will link 5-LO inhibition with therapeutic anti-inflammatory activity in a variety of diseases (Samuelsson, 1983; Drazen and Austen, 1987; Garcia et al., 1987; Lewis et al., 1990).

Acknowledgements The authors would like to thank .I. Barlow and E. Roberts for their excellent assistance, R. Dyer and D. H. Albert for their suggestions during the development of the model and the following individuals for their technical expertise during drug evaluations: S.

Majest. W. Hinz. J. Helms, D. Wilcox, J.G. Martin. 3. Bouska, E. Otis, R. Maki and B. Gunn.

References Anderson,ME., R.D. Allison and A. Meister, 19%2,Intercom,ersion of leukotrienes catalyzed by purified gamma-glutamy transpepti. &se: Concomitant formation of leukotriene DJ and gammaglutamY aminoacids,Proc. Sot. Acad. Sci. U.S.A., 79, itxig. Berkenkopf, J.W. and B. M. Weichman, 1991, Comparison of several new 5-lipoxygenase inhibitors in a rat Arthus pleurisy model, Eur. J. Pharmacol. 193, 29. Blackburn, W.D.J., L.W. Heck, L.D. Loose, J.D. E&a and T.J. Car& 1991. Inhibition of 5-lipoxygenase product formation and polymorphonuclear cell degranulation by Tenidap sodium in pa. tients with rheumatoid arthritis, Arthritis Rheum. 34, 201. Carison,R., w. Calhoun, L. O’Neill-Davis, L. Tomchek, J. Lt~gay,A Kreft, J. Musser. J. Hand, J. Chang and L. Marshall, 1989, Biochemical and antiinflammatory activity of an orally selective 5-lipoxygenase Inhibitor. WY-49.232 and its (s) and (r) enantiomers, Pharmacologist 31, 195 (Abstr. 414). Carter, G.W., M-K. Martin, R.A. Krause and P.R. Young, 1982, The effects of anti-inflammatory and other agents on the rat dermal Arthus reaction, Res. Commun. Chem. Pathol. Pharmacol. 35, 189. Carter, G.W., P.R. Young, D.H. Albert, J. Bouska, R. Dyer, RL Bell, J.B. Summers and D.W. Brooks, 1991, 5.Lipoxygenase inhibitory activity of zileuton, J. Pharmacol. Exp. Ther. 256, 929. Garcia, J.G.N., T.C. Noonan, W. Jubiz and A.B. Malik, 1987. Leukotrienes and the pulmonary microcirculation, Am. Rev. Respir. Dis. 136, 161. Gdlard, J.. A.W. Ford-Hutchinson, C. Chan, S. Charleson, D. Denis. A. Foster, R. Fortin. S. Leger, C.S. McFarlane, H. Morton. H. Piechuta, D. Riendeau, C.A. Rouzer, J. Rokach. R. Young, D.E. Maclntyre. L. Peterson, T. Bach. G. Eielmann, S. Hopple, J. Humes, L. Hupe, S. Luell, J. Metzger, R. Meurer, D.K Miller, E. Opas and S. Pacholok. 1989, L-663,536 fMK-8661t3_(1-(4-cblorobenzyl~-3-t-butyl-thio-5-isopropylindol-2yl~-2,2-dimethylprop~oic acid), a novel orally active leukotriene biosynthesis inhibitor, Can. J. Physiol. Pharmacol. 67. 456. Haviv, F.. J.D. Ratajczyk, R.W. DeNet, F.A. Kerdesky. R.L. Walters, S.P. Schmidt, J.H. Holms. P.R. Young and G.W. Carter. 1988, 3-[ I-(2.Benzoxazolyl)hydrazino]propanenitrile derivatives: Inhibitors of immune complex induced inflammation, J. Med. Chem. 31, 1719. Hayes. E.C.. D. Lombardo, Y. Girard, A. Maycock, J. Rokach, A. Rosenthal, R. Young, R. Egan and H. Zweerink 1983, Measuring leukotrienes of slow reacting substance of anapl-ylaxis: development of a specific radioimmunoassay. J. Immunol. 131. 429. Higgs, G.A., R.L. Follenfant and L.G. Garland, 1988. Selective inhibition of arachidonate 5.lipoxygenase by novel acetohydroxamic acids: effect on acute inflammatory responses, Br. J. Pharmacol. 94, 547. Israel, E., R. Dermarkarian, M. Rosenberg, R. Sperling. 0. Taylor. P. Rubin and J.M. Drazen, 1990, The effects of a 5lipoxygenase inhibitor on asthma induced by cold, dry air, New En@. J. Med. 323, 1740. Jones. G.H., M.C. Venuti, J.M. Young, D.V.K. Murthy. B.E. Lee. R.A. Simpson, A.H. Be&s, D.A. Spires, P.J. Maloney, M. Kiuscman, S. Rouhafza, K-C. Kappas, CC. Beard. S.H. Unger and P-S. Cheung, 1986, Topical nonsteroidal antipsoriatic agents. 1. 1,2,3,4_tetraoxygenated naphthalene derivatives, 5. Med. Chem. 29, 1504. Knapp, H.R., 1990, Reduced allergen-induced nasal congestion and leukotriene synthesis with an orally active 5-hpoxygehase inhibitor, New Engl. J. Med. 323, 1745.

zymr -linked immunosorbent assays for measurement of Ieukrrtrienes and prostaglandins, J. Immunol. Meth. Xl. 169. Murphy. RX., S. Hammarr!rom and B. Samuelsson, 1979, Leukotriene C: a slow reacting substance (SRS) from murinr mastocytoma cells. Proc. Natl. Acad. Sci. 76. 4275. Orange. R.P.. M.D. Valentine and K.F. Austen, 1968. Antigen-induced rele::se of slow reacting substance of anaphylaxis (SRS-A) in rats prepared with homologous antibody. in J. Exp. Med. 127, 767. Rapp. H.J., 1961, Release of a slow reacting substance (SRS) in the peritoneal cavity of rats by antigen-antibody reaction. J. Physiol. (London) 1%. 35~. Samuelsson. B.. 1983, Leukotrienes: mediators of immediate hypersensitivity reactions and inflammation. Science 220, 568. Sirois, P.. 1985. Pharmacology of the leukotrienes in: Advances in Lipid Research. eds. R. Paoletti and D. Kritchevesky (Academic Press. New York). Smith. H.. J.W. Ross and B.A. Spicer, 1980, Rat passive peritoneal anaphylaxis following sensitisation with guinea pig antiserum: an immediate hypersensitivity reaction without histamine release, Int. Arch. Allergy Appl. Immunol. 62, 195. Wei, Y.. K. Heghinian, R.L. Bell and B.A. Jakshik. 1986, Contribution of macrophages to immediate hypersensitivity reaction, J. Immunol. 137. 1993.