Metabolism and analysis of cysteinyl leukotrienes in the monkey.

Predominant hepatobiliary elimination from blood and subsequent enterohepatic circulation of cysteinyl leukotrienes is demonstrated in the monkey Macaca fascicularis. From intravenous [3H]leukotriene C4, about 40% were recovered as metabolites in bile and about 20% in urine within 5 h. [3H]Leukotriene E4 was a predominant metabolite of defined structure in blood plasma, bile, and urine. From intraduodenal [3H]leukotriene C4, about 5% were recovered as metabolites in bile and about 8% in urine within 8 h. Endogenous cysteinyl leukotrienes generated in vivo were measured after implantation of a subcutaneously looped biliary bypass. Tapping of the loop allowed access to bile and prevented interference by leukotrienes produced by surgical trauma (Denzlinger, C., Rapp, S., Hagmann, W., and Keppler, D. (1985) Science 230, 330-332). Endogenous cysteinyl leukotrienes were analyzed in bile, urine, and blood plasma by the sequential use of high-performance liquid chromatography and a radioimmunoassay that was optimized for leukotriene E4 as a predominant metabolite detected in the tracer studies. Biliary leukotriene E4 rose from less than 0.2 to 9 nmol/liter, when leukotriene synthesis was elicited in anesthesized monkeys by staphylococcal enterotoxin B administered intragastrically. This study provides an approach to the analysis of cysteinyl leukotrienes in primates and serves to define the role of these mediators under pathophysiological as well as physiological conditions in vivo.

gastrically. This study provides an approach to the analysis of cysteinyl leukotrienes in primates and serves to define the role of these mediators under pathophysiological as well as physiological conditions in vivo.
Evaluation of the function of cysteinyl leukotrienes in pathophysiological and physiological processes requires information on their generation and metabolism in uiuo. Until now, data on endogenous cysteinyl leukotriene production have been limited (10)(11)(12)(13)(14)(15)(16)(17). Several measurements were per-* This work was supported by the Deutsche Forschungsgemeinschaft, Bonn, through SFB 154, Freiburg, and the Fraunhofer Gesellschaft, Munchen, Federal Republic of Germany. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Cysteinyl leukotrienes are formed by conjugation of the 5lipoxygenase product leukotriene A4 (LTA;) with glutathione yielding LTC,. This conjugate is converted in the mercapturic acid pathway to LTD, and LTE, by consecutive release of the y-glutamyl and the glycinyl moieties (1-3,10,23-25). The Nacetylation product of LTE,, N-acetyl-LTE,, represents the mercapturic acid derivative and is the predominant metabolite in rat bile (14) and feces (26).
The aim of the present study was to develop a method for determination of cysteinyl leukotrienes generated in primates i n uiuo. Tritium tracer studies were performed using I3H] LTC, to determine the relative amounts and the pattern of LTC4 metabolites occurring in biological fluids such as blood plasma, bile, and urine. [3H]LTC4 was injected intravenously or into the duodenum, i.e. into the upper part of the small intestine. Intraduodenal administration served to study the possibility of intestinal reabsorption of cysteinyl leukotrienes eliminated via the bile into the intestine.
Based on the results of the tracer studies, a sensitive and selective method for determination of cysteinyl leukotrienes was elaborated and served to measure these mediators after stimulation of their production by staphylococcal enterotoxin B. This enterotoxin is a major cause of food-borne enteric intoxications and is known to elicit shocklike reactions in man (27).
A preliminary report on minor parts of this work was presented earlier (28).
N-Acetyl-LTE, and N-a~etyl-[~HlLTE~ were synthesized from LTE, and [3H]LTE,, respectively, as described previously (14). Reversedphase high-performance liquid chromatography (RP-HPLC) separation (14) was used to control the purity of the leukotrienes and to purify the unlabeled leukotrienes employed as standards for the radioimmunoassay (RIA). The concentration of unlabeled leukotrienes was determined by absorbance measurements at 280 nm using a molar absorptivity of 40,000 1 X mol" X cm" (29  Ref. 48. One week before bile sampling for analysis, a subcutaneously looped biliary bypass was implanted under ketamine anesthesia (16 mg/kg) in a procedure analogous to the one described previously for rats (11,13). By choledochotomy a catheter (Silastic, 1.6-mm inner diameter, reduced with a polyethylene tube, 1.1-mm inner diameter, both from Dow Corning Corp., Midland, MI) was inserted into the bile duct and oriented toward the liver. A subcutaneous loop was then created and the catheter reinserted into the same choledochotomy site and advanced toward the duodenum. The gall bladder was removed. After this operation, dehydrocholate was administered at a dose of 14 mg/ kg every day. On the day of the experiment, normal liver function was ascertained by measurement of enzyme activities in blood plasma. The plasma concentration of C-reactive protein was also in the normal range. Experiments on leukotriene metabolism were performed under pentobarbital anesthesia (initial dose, 26 mg/kg). For bile sampling, as well as for intraduodenal administration of [3H] LTC,, the subcutaneous loop of the biliary bypass was opened. I3HJ LTC, was injected by bolus at a concentration of 0.3 pmol/liter in phosphate-buffered 20% ethanol (pH 6.8) into the cubital vein or into the duodenum. The tracer dose was 10 pCi/kg, body weight, corresponding to 0.25 nmol/kg. Staphylococcal enterotoxin B at a dose of 10 pg/kg was administered into the stomach in 5 ml of phosphate-buffered saline by gastric tube. Urine that was collected from a catheter introduced into the urinary bladder and bile were sampled continuously into ice-cold 90% aqueous methanol, containing 1 mM HTMP, 0.5 mM EDTA, pH 7.4, and stored at -20 "C under argon. Blood was withdrawn from a venous catheter or by puncture of the femoral artery and mixed immediately with a cold solution containing (at final concentrations) heparin (3300 units/liter), EDTA (0.3 mM), indomethacin (2 p~) , HTMP (70 p~) , pH 7.4. Blood plasma, obtained after centrifugation of the blood samples, was diluted in 4 volumes of cold 90% aqueous methanol containing 1 mM HTMP, 0.5 mM EDTA, pH 7.4, and stored at -20 "C under argon.
High-performance Liquid Chromatography-Before application to HPLC, blood plasma, bile, and urine were deproteinized the biological samples were made 80% of aqueous methanol, containing 1 mM HTMP, stored at -20 "C for at least 3 h, centrifuged at 8000 X g for 10 min, and the supernatant was evaporated to dryness under a gentle stream of nitrogen. The dried samples were again resuspended in 80% cold methanol with HTMP, stored at -20 "C, recentrifuged, and the supernatant evaporated. This procedure was repeated up to 2 times until a supernatant without visible protein pellet was obtained. The samples were then evaporated again and resuspended in 30% aqueous methanol. HPLC was performed on a C,,-Hypersil column (4.6 X 250 mm, 5-pm particles, Shandon, Runcorn, United Kingdom) with a C,, precolumn (Waters, Milford, MA). The mobile phase consisted of methanol, water, and acetic acid ( Radioimmunoassay for Cysteinyl Leukotrienes-A 400-or 600-p1 aliquot of the neutralized HPLC fractions was evaporated to dryness and resuspended in 200 pl of assay buffer (0.9% NaCI, 0.1% gelatin, 10 mM EDTA, 0.1% sodium azide in 10 mM phosphate buffer, pH 7.4). Standards for evaluation of cross-reactivities were dissolved in 200 pl of assay buffer. Standards used as references for determination of cysteinyl leukotriene concentrations in HPLC samples contained, in addition, neutralized and evaporated leukotriene-free HPLC eluate, in a volume corresponding to the respective HPLC samples. Antiserum (100 pl), diluted 1:4000 in assay buffer, was added to the samples and standards. After mixing, the tubes were preincubated for 30 min at room temperature before addition of labeled leukotriene.
[3H]LTE4 (3.4 nCi) was added in 100 p1 of assay buffer, mixed, and incubated at 4 "C for 16-20 h. Unbound [3HJLTE4 was precipitated by addition of 500 pl of charcoal suspension (0.5% charcoal, 0.5% dextran in 10 mM phosphate buffer, pH 7.4) and subsequent centrifugation at 1400 X g for 15 min at 4 "C. The supernatant was added to 10 ml of scintillation fluid. In the absence of evaporated HPLC eluate, the lower detection limit of the assay system for LTE, was at about 4 fmol; the relative percent cross-reactivities at 50% binding of LTE4, N-acetyl-LTE,, LTD,, LTC,, 11-trans-LTE,, and LTB, were 100,320,280,90,30, and less than 0.005 on a molar basis, respectively. In an experiment where 600 pl of evaporated HPLC eluate were used with each sample, the lower detection limit for LTE, was at about 7 fmol, and the relative percent cross-reactivities of LTE4, LTD4, and LTC, were 100, 260, and 90, respectively.

Hepatobiliary and Urinary Elimination of Leukotriene Radioactivity after Intravenous or Zntraduodenal Injection of
[3H]LTC4-After intravenous injection of [3H]LTC,, radioactivity was rapidly eliminated from the circulating blood, and increasing amounts of leukotriene radioactivity were detected in bile and urine (Fig. l). The rate of leukotriene secretion into bile exceeded secretion into urine especially during the first 2 h after [3H]LTC, administration. Within 5 h, about 40% of the administered radioactivity was recovered in bile and about 20% in urine.
After intraduodenal injection of [3H]LTC4, total blood plasma contained less than 0.05% of the administered tracer dose at all times investigated by analysis of small aliquots of blood plasma. However, following a lag period of about 1 h, leukotriene radioactivity appeared in bile and urine (Fig. 2). This indicates intestinal absorption and consecutive biliary as well as urinary secretion of [3H]leukotriene metabolites. Secretion of radioactivity into urine predominated over secretion into bile during the later time periods. Within 8 h, about 8% o f administered radioactivity was recovered in urine and about 5% in bile. The time course of elimination suggests that

FIG. 1. Elimination of radioactivity from intravenously administered ['H]LTC4 into bile and urine.
[3H]LTC4 (10 pCi/kg of body weight) was injected into the cubital vein of anesthetized monkeys. Blood samples of 0.2 ml were collected at the times indicated. Bile and urine were sampled continuously. Data are expressed as percent of injected tritium circulating in blood or accumulated in bile or urine. Mean values from 3 animals are given. The total amount of leukotriene radioactivity recovered in bile and urine within the experimental period ranged from 41 to 89% of the injected dose. The relative proportion of biliary elimination ranged from 50 to 70% of the recovered radioactivity.

FIG. 2. Elimination of radioactivity from intraduodenally administered ['H]LTC4 into bile and urine.
[3H]LTC4 (10 pCi/ kg of body weight) was injected into the duodenum of anesthetized monkeys. Metabolites formed in the duodenum included LTD, and LTE,. Data are expressed as percent of injected tritium accumulated in bile or urine. Mean values from 3 animals are given. The relative proportion of biliary elimination ranged from 39 to 66% of the recovered radioactivity.

FIG. 3. RP-HPLC separation of 'H-labeled metabolites in bile and urine after intraduodenal or intravenous ['H]LTC4
administration. Bile and urine were collected during the indicated time periods following intraduodenal or intravenous administration of [3H]LTC4. Samples were deproteinized in 80% methanol and separated by RP-HPLC at pH 5.6 as described under "Experimental Procedures." The arrows indicate retention times of [3H]leukotriene standards determined immediately before or after the respective separation of bile or urine, since retention times showed minor variations after several runs. secretion continued beyond the 9-h time period of sample collection shown in Fig. 2. [

3H]Leukotriene Metabolites in Biological Fluids after Intravenous Injection of
[3H]LTC4-More than 90% of the radioactivity in venous blood plasma co-eluted with standard [3H]LTE4 as analyzed by RP-HPLC of samples collected at 1 and 5 min after intravenous injection of [3H]LTC4. The low concentration of tracer in blood collected at later times after injection (Fig. 1) did not allow proper identification of metabolites.
The predominating metabolites found in bile and urine after intravenous injection of [3H]LTC4 were more polar than [3H]LTC4 and eluted with mean retention times of 0.25 (LT 0.25) and 0.45 (LT 0.45) relative to [3H]LTC4, respectively. The relative amount of L T 0.45 diminished with time, while the more polar L T 0.25 metabolites increased (Fig. 3, panels on the right; Fig. 4). The polar [3H]leukotriene metabolites are heterogenous; urinary L T 0.25 could be separated on RP-HPLC into 2 major and at least 3 minor components using a linear gradient of 5-65% of aqueous methanol. Each of the 2 major components amounted to about 30% of the LT 0.25 radioactivity; the 3 minor components comprised 10-20% each.

r3H/Leukotriene Metabolites in Biological Fluids after Intraduodenal Administration of [3H]LTC4-The low
concentration of 3H radioactivity present in blood plasma after intraduodenal [3H]LTC4 precluded identification of metabolites. The biliary pattern of [3H]leukotriene metabolites obtained after intraduodenal [3H]LTC4 (Fig. 3 (Fig. 3, lower left panel), is present only in urine and not identical with ll-truns-LTE4. The percentage of metabolites is expressed relative to total [3H]leukotriene metabolites in the respective sample. Mean values from 3 animals are given.
The polar LT 0.25 metabolites predominated in urine collected after intraduodenal administration of [3H]LTC4 at all sampling times (Fig. 3, lower left panel; Fig. 5  Determination of Endogenously Generated Cysteinyl Leukotrienes in Biological Fluids-Endogenous cysteinyl leukotrienes were measured by the sequential use of RP-HPLC and RIA. The concentrations of LTE,, LTD,, or LTC, in blood plasma were all below 15 pmol/liter, when blood was obtained under conditions of minimal tissue and cell injury by puncture of the femoral artery. By contrast, when plasma was obtained after suction of venous blood associated with cell and tissue injury, it contained LTD, immunoreactivity in concentrations up to 3.5 nmol/liter; LTE, and LTC, were below 250 and 200 pmol/liter, respectively. Bile and urine sampled under control conditions contained less than 200 and 50 pmol/liter, respectively, of immunoreactive cysteinyl leukotrienes. Endogenous leukotriene production was elicited by intragastrically administered staphylococcal enterotoxin B and yielded LTE, as the major metabolite detected in bile at a concentration of 9 nmol/liter (Fig. 6). Taking into account their relative cross-reactivities, LTC,, LTD,, and N-acetyl-LTE, were all below 0.4 nmol/liter after the enterotoxin dose of 10 Fg/kg. Under the same conditions, in urine, a single immunoreactive metabolite was found that eluted shortly after LTE, with a retention time of 2.4 relative to LTC, (Fig. 6). This compound reached a concentration of 1 nmol/liter of LTE, immunoreactivity and co-eluted with the radioactive L T 2.4 metabolite observed in urine after intraduodenal administration of [3H]LTC4 (Fig. 7). The endogenous urinary metabolite shown in Figs. 6 and 7 resisted chemical N-acetylation, excluding its identity with ll-truns-LTE, that eluted at a similar retention time in the RP-HPLC system a t p H 5.6.

DISCUSSION
Elucidation of the metabolism of cysteinyl leukotrienes and determination of their concentrations i n vivo is necessary for an improved understanding of their role in pathophysiological and physiological processes. The present study was performed in monkeys 1 week after implantation of a subcutaneously looped biliary bypass, allowing the sampling of bile without extensive traumatic manipulations on the day of the experiment. This prevented interference due to endogenous cysteinyl leukotrienes induced by the surgical trauma (11).
[3H]LTC4 served as the precursor for determination of the compartmental distribution and the metabolite pattern of cysteinyl leukotrienes in blood plasma, bile, and urine. Elimination of intravenously injected [3H]LTC4 from the circulating blood was comparably fast in the monkey (Fig. 1) and in the rat (10,11). However, appearance of tracer in bile and urine proceeded considerably slower in the monkey (Fig. 1). This indicates prolonged processes of organ storage and metabolism preceding hepatobiliary and renal elimination in the primate. Hepatobiliary relative to renal elimination predominates in all species investigated up to now (Fig. 1;Refs. 10,14,and 25); however, the relative quantity of cysteinyl leukotrienes secreted with the urine is much larger in the monkey ( Fig. 1) than in the rat (10,14).
Leukotriene metabolites were also detected in bile and urine after intraduodenal injection of [3H]LTC4 (Fig. 2), indicating intestinal absorption, transport in blood, and subsequent hepatobiliary as well as renal secretion. The sequence of biliary secretion, intestinal absorption, transport to the liver via the portal vein, and resecretion of a substance with bile is known as enterohepatic circulation. This process conserves several endogenous substances as well as drugs in the mammalian organism, such as bile acids, vitamins, antibiotics, and cardiac glycosides (31). Our results (Fig. 2) demonstrate that cysteinyl leukotriene metabolites are to be considered as novel members in the list of substances undergoing enterohepatic circulation. Intestinal reabsorption of [3H]LTC4 metabolites excreted via bile into duodenum may explain, in part, the observation that elimination of radioactivity with urine predominates over elimination with feces after intravenous injection of [3H]LTC4 in humans (32). The elimination routes of cysteinyl leukotrienes in the monkey are depicted in Fig. 8.
RP-HPLC analyses of blood plasma obtained after intravenous injection of [3H]LTC4 indicated rapid metabolism of tracer to [3H]LTE4. This agrees with earlier studies on [3H] LTE3 formation from [3H]LTC3 in blood circulation in the monkey (33). Metabolism of LTC, in blood plasma in uitro proceeds much slower (34) than in circulating blood i n vivo. This may be explained by the fact that y-glutamyl transferase and dipeptidase, which produce LTD, and LTE,, respectively, are plasma membrane-bound ectoenzymes (23,35) present in blood plasma at low activities only (34).
RP-HPLC analyses of extracts from bile and urine, obtained after intravenous or intraduodenal injection of [3H] LTC,, indicated rapid conversion to a series of metabolites . Among these, LTD, and LTE, have been structurally identified (1,25). Several of the metabolites were more polar than LTC,, with retention times on RP-HPLC of 0.25-0.8 relative to LTC, . The radioactive cysteinyl leukotriene metabolite eluting shortly after standard LTE, with a retention time of 2.4 relative to LTC, (Figs. 3 and 5) was found exclusively in urine and only after intraduodenal administration of [3H]LTC4. This indicates involvement of intestinal metabolization and renal elimination in the formation and appearance of this relatively unpolar metabolite.
Of the radioactivity from intravenously administered [3H] LTC,, about 10 and 1% were recovered as [3H]LTEI during the 5-h experimental period in bile and urine, respectively (Fig. 4). By contrast, a much higher percentage of intravenously administered [3H]LTC4 was reported to appear as [3H] LTE, in urine of man (32). This apparent discrepancy might be explained in part by the lack of enterohepatic circulation under our experimental conditions where bile was collected continuously, thereby preventing intestinal metabolization and reabsorption of LTE, and its derivatives. Moreover, the differences in species and dose may play a role.
As a result of the tracer studies, LTE, appears as a predominant metabolite among the cysteinyl leukotrienes of defined structure . Therefore, the method for sensitive determination of endogenously generated cysteinyl leukotrienes was optimized for LTE, analysis. The sequential application of RP-HPLC and RIA has been employed successfully in previous measurements of cysteinyl leukotrienes in biological samples (6-14, 16, 36, 37). Deproteinization and RP-HPLC separation of biological fluids prior to RIA is required to remove leukotriene-binding material that interferes in the RIA. In addition, RP-HPLC separation enables detection of individual metabolites and contributes thereby to the selectivity of the assay system. Most of the previous cysteinyl leukotriene RIAs have been optimized for sensitive detection of LTC, rather than LTE, (38-44). Appearance of significant amounts of LTC, in biological fluids i n vivo is, however, unlikely, as indicated by the rapid metabolism of [3H]LTC4 in the monkey Ref. 33) as well as in the rat (12,14).
Endogenous cysteinyl leukotrienes in blood plasma were below detectability when blood was obtained under conditions of minimal tissue and cell injury. This is in line with the tracer studies indicating that cysteinyl leukotrienes are rapidly eliminated from the blood whenever they appear in the circulation (Fig. 1). High concentrations of LTD, detected in blood plasma after suction of venous blood are likely to be due to cell injury induced during sample collection (11). LTD,, formed from endogenous LTC,, may accumulate due to inhibition of LTD, dipeptidase by the EDTA in the anticoagulant mixture (45). Artifactual formation of other eicosanoids during collection of blood is recognized as an obstinate pitfall, e.g. in the determination of thromboxane B, concentrations in blood plasma (46). Our results suggest that this applies to cysteinyl leukotrienes as well. Caution is particularly indicated when LTC, and/or LTD, are detected as major metabolites, in contradiction to the rapid conversion of [3H]LTC4 to [3H]LTEa in circulating blood i n vivo. LTE, was the major detected metabolite in bile (Fig. 6) after stimulation of endogenous leukotriene generation by staphylococcal enterotoxin B. This predominance of LTE, is in agreement with the metabolite pattern in bile after administration of [3H]LTC4 (Fig. 3). N-Acetyl-LTE,, the end product of the mercapturic acid pathway of leukotrienes and the major endogenous cysteinyl leukotriene metabolite in rat bile (14), was not found in significant amounts in monkey bile, a result consistent with the absence of N-a~etyl-[~HlLTE, in biological fluids of the monkey following [3H]LTC4 administration. LTE, measurement in bile represents a useful approach to the analysis of cysteinyl leukotriene generation in the primate in uiuo. Recently we have analyzed human bile, obtained during enteral retrograde cholangiography, and found highly increased concentrations of LTE, in patients with acute pancreatitis (unpublished measurements). Our studies in the monkey will facilitate further investigations on the generation of cysteinyl leukotrienes under pathophysiological as well as physiological conditions in man.