Characterization and Biologic Properties of 5,12-Dihydroxy Derivatives of Eicosapentaenoic Acid, Including Leukotriene B5 and the Double Lipoxygenase Product*

Leukotriene Ba (LTBa) and three stereoisomers were prepared biosynthetically from eicosapentaenoic acid and compared with the analogous derivatives of arach- idonic acid for their chemotactic and aggregating effects on human neutrophilic polymorphonuclear leu- kocytes. Leukotriene B4 (LTB4), LTBa, and the 6-trans-diastereoisomers of each were generated by ac-tivating polymorphonuclear leukocytes with the cal- cium ionophore A23187 in the presence of 14C-labeled and unlabeled arachidonic acid or 14C-labeled and un- labeled eicosapentaenoic acid, respectively. The double lipoxygenase products, (5S,12S)-6-trans-8-cis-LTBr and (5S,12S)-6-trans-8-cis-LTB6, were generated from 5s-hydroxyeicosatetraenoic acid and racemic 5- hydroxyeicosapentaenoic acid intermediates by incubation with platelet sonicates. The products of each reaction were isolated by reverse-phase-high performance experiments with a previously described competition assay (28). Serial dilution of the unlabeled synthetic or natural 5J2-dihydroxy derivatives of arachidonic acid and EPA, [14,15-3H]LTB4, and rabbit im-buffer for 60 min at 37 "C before the addition of 100 p1 of goat anti- mune plasma were incubated in a final volume of 250 p1 of Isogel Tris rabbit IgG plasma (previously titrated to antibody excess with respect to the rabbit IgG in the immune plasma and in the control nonim-mune plasma). The mixtures were shaken and the immune precipi- tates were formed overnight at 4 "C and sedimented at 1500 X g for 60 min at 4 "C. The supernatants were discarded and the immuno- precipitates were dissolved in 200 p1 of 0.1 N NaOH and mixed with 2.5 ml of scintillation fluid for measurement of radioactivity. Syn- thetic LTB, was detectable on the linear portion of its radioligand binding inhibition curve over a dose range from 0.05-1.0 ng. Biological Properties of the 5,12-Dihydroxy Derivatives of Arachidonic Acid and EPA-The chemotactic activity of each of the natural 5,12-dihydroxy derivatives was compared with synthetic LTB, over a concentration range from 0.01-50 or 100 ng/ml using PMN from 3 normal donors in a modified Boyden assay. In two further experiments with different donors, the chemotactic activity of synthetic LTB, was compared to synthetic LTB,. As determined by UV absorbance, known amounts of the synthetic or natural products stored in Solvent I1 were dried under reduced pressure and resuspended in HBSS containing 30 mM Hepes, pH 7.35 (HBSS/Hepes), at defined concentrations. Eight hundred-pl samples were placed into the lower wells of chemotactic chambers. One ml of 2 X lo6 PMN in HBSS/ Hepes buffer containing 0.4% ovalbumin (w/v) was added to the upper well of each chemotactic and expressed as the number of PMN/10 high power fields after subtraction of the back- ground migration in the chambers without a chemotactic stimulus The in duplicate and the intra- coefficient for was 19.5%. Human PMN were aggregated in a Model 300 aggregometer NY) One-half ml of a PMN suspension of 1 X lo' cells/ml of HBSS was added to a cuvette containing a stirring bar and the suspension was warmed for 2 min at 37 "C. Dose-aggregation response relationships were assessed by the addition of each natural or synthetic product in 1-40 p1 of HBSS to a final concentration of 0.1-40 ng/ml. Addition of the agonist produced a rapid initial increase in light transmission, followed by a second more gradual increment. Since only the second phase increment in light transmission was associated with microscopic evidence of PMN aggregation, as also reported by ultrastructural analysis (30), this second phase response, recorded as millimeters deflection from base-line, was used to compare the response to each product with that to synthetic LTB,.

Leukotriene Ba (LTBa) and three stereoisomers were prepared biosynthetically from eicosapentaenoic acid and compared with the analogous derivatives of arachidonic acid for their chemotactic and aggregating effects on human neutrophilic polymorphonuclear leukocytes. Leukotriene B4 (LTB4), LTBa, and the 6trans-diastereoisomers of each were generated by activating polymorphonuclear leukocytes with the calcium ionophore A23187 in the presence of 14C-labeled and unlabeled arachidonic acid or 14C-labeled and unlabeled eicosapentaenoic acid, respectively. The double lipoxygenase products, (5S,12S)-6-trans-8-cis-LTBr and (5S,12S)-6-trans-8-cis-LTB6, were generated from 5s-hydroxyeicosatetraenoic acid and racemic 5hydroxyeicosapentaenoic acid intermediates by incubation with platelet sonicates. The products of each reaction were isolated by reverse-phase-high performance liquid chromatography and identified by their retention times relative to the appropriate totally synthetic standards, ultraviolet absorption spectra, immunoreactivity in a radioimmunoassay for LTB4, and, for all but the double lipoxygenase products, by incorporation of radiolabel from the specific polyunsaturated fatty acid source. When the concentration of LTBa eliciting maximum chemotactic response of human polymorphonuclear leukocytes, 50 ng/ml (1.5 x lo-' M), and that eliciting a maximum aggregation response, 20 ng/ml(5.9 X lo-' M), were compared with the interpolated values of LTB4 eliciting comparable effects, the potency of LTBS relative to LTB4 was approximately 1:s as a chemotactic agent and about 1:20 as an aggregating agent. The double lipoxygenase products and the resolved 6-trans-diastereoisomers of the pentaene and tetraene series were about 2 logs less active as chemotactic factors than LTB4 and only *This research was supported in part by Grants AI-07722, AI-10356, AI-20081, HL-13262, and HL-17382 from the National Institutes of Health and in part by grants from the Lillia Babbitt Hyde Foundation, the New England Peabody Home for Crippled Children, and the National Science Foundation. 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. Arachidonic acid released from membrane phospholipids during cell activation may be reacylated or oxidatively metabolized by the cyclooxygenase or lipoxygenase pathway (1, 2). 5-Lipoxygenase generates 5S-hydroperoxy-6-trans-8,11,14cis-eicosatetraenoic acid from arachidonic acid and this unstable intermediate is hydrolyzed to 5s-hydroxy-6-trans-8,11,14-cis-eicosatetraenoic acid (3) or is enzymatically converted to an unstable epoxide, 5,6-oxido-7,9-trans-11,14-ciseicosatetraenoic acid (leukotriene A,) (4-6). LTA,' is converted by an epoxide hydrolase to 5S,12R-dihydroxy-6,14-cis-8,lO-trans-eicosatetraenoic acid (leukotriene B4) ( 7 ) or by a glutathione-S-transferase to 5s-hydroxy-6R-S-glutathionyl-7,9-trans-11,14-cis-eicosatetraenoic acid (leukotriene C,) (8).
Comparative analysis has revealed natural and synthetic LTB4 to be 100 times more potent than the double lipoxygenase product (5S,12S)-6-trans-8-cis-LTB4 and the 6-trans-LTB, 12-S/R diastereoisomers in eliciting a comparable chemotactic response of human PMN (11). LTBI has addi-Leukotriene B5 and Analogs tional biologic effects on human leukocytes which include aggregation of PMN (12), chemokinesis and chemotaxis for eosinophils and monocytes (13,14), enhanced expression of C3b receptors on PMN and eosinophils (13), and release of lysosomal enzymes and stimulation of superoxide anion generation by PMN (15). The injection of 1.6 nmol of LTB, into human skin elicited an indurated lesion which, upon biopsy, revealed a predominant PMN infiltration (16). Because EPA is incorporated into membrane phospholipids along with arachidonic acid and other polyunsaturated fatty acids in humans and animals on a fish-enriched diet, certain oxidative products of EPA generated via the cyclooxygenase pathway have been characterized (17)(18)(19)(20). The only analysis of the dihydroxy products of EPA generated via the 5-lipoxygenase pathway has utilized ionophore-induced activation in vitro of mouse mastocytoma cells obtained from tumor-bearing mice maintained on a diet enriched in EPA (21). Authentic LTB5 and the diastereoisomers of 6-trulzs-LTB5 were resolved by RP-HPLC and characterized by UV absorption spectra and by mass spectrometry (21). No functional comparisons were made among the 5,12-dihydroxy derivatives within the pentaene series or relative to the tetranes. Accordingly, each of the natural 5,12-dihydroxydiastereoisomeric pentaene derivatives, including the double lipoxygenase product of EPA, has been prepared by biosynthesis, isolated, and analyzed for comparative chemotactic and aggregating effects on human PMN.
Preparation of PMN-derived 5,12-Dihydroxy Derivatives of Arachidonic Acid and EPA--1.6 to 9.2 X 10' PMN, obtained from 150-500 ml of citrate-anticoagulated blood from 6 normal donors, were purified to greater than 95% by dextran sedimentation, centrifugation through Ficoll-Hypaque and hypotonic lysis of erythrocytes (26), and were washed three times in HBSS. The PMN were suspended in HBSS at a concentration of 1 X lo' cells/ml, warmed to 37 "C for 5 min, mixed with 20 pg/ml of arachidonic acid or EPA, and immediately activated with 10 p~ calcium ionophore A23187. In selected experiments, ["Clarachidonic acid or ["CIEPA was added with the unlabeled arachidonic acid or EPA at 0.1 pCi/107 cells. After incubation of the mixture at 37 "C for 2 min, the reaction was stopped by the addition of 4 volumes of ethanol and rapid cooling to 4 "C. The mixture was allowed to stand for 1 h at 4 "C and the precipitate was removed by sedimentation at 1000 X g for 15 min at room temperature. The supernatant was evaporated to dryness under reduced pressure and the residue was suspended in a mixture of 65% methanol, 34.9% water, 0.1% acetic acid, v/v, pH 5.6 (Solvent I) for RP-HPLC.
Preparation of Tetraene and Pentaene Double Lipoxygenase Products-In order to generate the tetraene andpentaene products derived from the sequential action of 5-and 12-lipoxygenase on arachidonic acid (10) and on EPA, 5-HETE and rac-5-HEPE were separately incubated with platelet sonicates. Venous blood of 3 normal donors was collected into separate plastic tubes containing 0.8% citric acid, 2.2% trisodium citrate, and 2.45% dextrose (w/v) (ACD) (1 volume of ACD:9 volumes of blood) and centrifuged at 100 X g for 15 rnin at 20 'C. The platelet-rich plasma was collected, acidified to pH 6.4 with 0.15 N citric acid, and centrifuged at 900 x g for 20 min at 20 'C. The supernatant was discarded. The platelet pellets were suspended in 6 ml of calcium-free Tyrode's/Hepes/ACD buffer, pH 6.4 (9 volumes of Tyrode's solution with 5 mM Hepes:l volume of ACD), washed by repeated centrifugation in the same buffer, incubated with 0.1 mM aspirin for 15 min at 20 'C, washed again by centrifugation, and resuspended in calcium-free Tyrode's/Hepes/ACD at a concentration of 1 to 7.5 X 109/ml. Platelet purity was virtually 100% with less than 1 erythrocyte/lO' platelets (27). Cell-free sonicates were produced at 4 "C with a Branson sonifier (setting 3, 40% pulse cycle, 20 pulses).
Replicate samples of 2.75 pg of 5-HETE or rac-5-HEPE lithium salts were evaporated to dryness under nitrogen; the residues were suspended in 900 pi of Tyrode's buffer, pH 7.4, with 1.3 mM calcium chloride and prewarmed to 37 "C for 5 min. One hundred pl of the platelet sonicate was added to each sample of substrate, incubation was continued for 5 min at 37 "C, and the reaction was stopped by the addition of 1 ml of methanol and cooling to 4 "C. After centrifugation at 1000 X g for 15 min at 4 "C, the supernatants were collected, pooled by substrate, and evaporated to dryness under reduced pressure; the residue of each was resuspended in Solvent I for RP-HPLC.
Purification and Characterization of the 5,12-Dihydroxy Derivatives of Arachidonic Acid and EPA-Products from each reaction were initially resolved by RP-HPLC on a 10-pm Cls Ultrasil-ODS column In the experiments carried out in the presence of ["CC]arachidonic acid and ["CIEPA, the column was eluted at the same flow rate in Solvent I for 30 min, then with a gradient to 100% methanol over 5 min, and isocratically in 100% methanol for 10 min more. In this solvent program, synthetic 5-HETE and arachidonic acid eluted at 37 and 39 min, respectively, and rac-5-HEPE and EPA eluted at 36.5 and 38.5 min, respectively. One hundred-pl samples of each fraction were added to 10 ml of scintillation fluid (Hydrofluor, National Diagnostics, Somerville, MA) and radioactivity was measured in a liquid 0scintillation counter (Mark 111, Tracor Analytic, Elk Grove, IL). In selected experiments, 5-pl samples of each fraction were evaporated to dryness under reduced pressure. The residues were resuspended in 500 pl of 10 mM Tris-HC1 buffer, pH 7.4, containing 0.15 M NaCl and 0.1% gelatin (Isogel Tris buffer), and a 100-pl sample of each was assessed for reactivity in a radioimmunoassay for LTB, (28).
Fractions from several RP-HPLC procedures which eluted in peaks of AZe9 with the same retention time were pooled and the solvent was evaporated under reduced pressure. Each residue was suspended in 1 ml of a mixture of 59% methanol, 40.9% water, and 0.1% acetic acid (v/v, pH 5.6) (Solvent 11) and applied to a separate CIS Ultrasil-ODS column equilibrated in the same solvent. The column was eluted at a flow rate of 1 ml/min, UV absorbance was continuously monitored at 269 nm, and 1-ml fractions were collected. The fractions corresponding to a single product were combined, analyzed for UV absorption spectra and concentrations, and stored in Solvent 11 under argon at -20 "C until assessment of immunoreactivity and biological function.
Immunoreactivity of the 512-Dihydroxy Derivatives of Arachidonic Acid and EPA-As determined by UV absorbance, defined amounts of the resolved natural products stored in Solvent I1 were dried under reduced pressure and resuspended in 250 p1 of Isogel Tris buffer. The per cent inhibition of the binding of [14,15-3H]LTB4 (5100-7525 cpm) to 15 p1 of rabbit immune plasma by 0.03-5 ng of each of the natural products was assessed in duplicate in at least 3 separate by guest on March 21, 2020 http://www.jbc.org/ Downloaded from experiments with a previously described competition assay (28). Serial dilution of the unlabeled synthetic or natural 5J2-dihydroxy derivatives of arachidonic acid and EPA, [14,15-3H]LTB4, and rabbit imbuffer for 60 min at 37 "C before the addition of 100 p1 of goat anti-mune plasma were incubated in a final volume of 250 p1 of Isogel Tris rabbit IgG plasma (previously titrated to antibody excess with respect to the rabbit IgG in the immune plasma and in the control nonimmune plasma). The mixtures were shaken and the immune precipitates were formed overnight at 4 "C and sedimented at 1500 X g for 60 min at 4 "C. The supernatants were discarded and the immunoprecipitates were dissolved in 200 p1 of 0.1 N NaOH and mixed with 2.5 ml of scintillation fluid for measurement of radioactivity. Synthetic LTB, was detectable on the linear portion of its radioligand binding inhibition curve over a dose range from 0.05-1.0 ng. Biological Properties of the 5,12-Dihydroxy Derivatives of Arachidonic Acid and EPA-The chemotactic activity of each of the natural 5,12-dihydroxy derivatives was compared with synthetic LTB, over a concentration range from 0.01-50 or 100 ng/ml using PMN from 3 normal donors in a modified Boyden assay. In two further experiments with different donors, the chemotactic activity of synthetic LTB, was compared to synthetic LTB,. As determined by UV absorbance, known amounts of the synthetic or natural products stored in Solvent I1 were dried under reduced pressure and resuspended in HBSS containing 30 mM Hepes, pH 7.35 (HBSS/Hepes), at defined concentrations. Eight hundred-pl samples were placed into the lower wells of chemotactic chambers. One ml of 2 X lo6 PMN in HBSS/ Hepes buffer containing 0.4% ovalbumin (w/v) was added to the upper well of each chemotactic chamber; upper and lower wells were separated by an 8-pm pore size nitrocellulose filter (Sartorius, Gottingen, Germany). After incubation for 90 min at 37 "C, the filters were removed, fixed in ethanol, stained with hematoxylin/lithium carbonate and mounted bottom-side uppermost on glass slides. The cells which had migrated through the entire thickness of the filter were counted and the chemotactic activity was expressed as the number of PMN/10 high power fields after subtraction of the background migration in the chambers without a chemotactic stimulus (29). The experiments were performed in duplicate and the intraassay coefficient of variation for duplicate filters was 19.5%.
Human PMN were aggregated in a Model 300 aggregometer (Payton Associates, Buffalo, NY) (30). One-half ml of a PMN suspension of 1 X lo' cells/ml of HBSS was added to a cuvette containing a stirring bar and the suspension was warmed for 2 min at 37 "C. Doseaggregation response relationships were assessed by the addition of each natural or synthetic product in 1-40 p1 of HBSS to a final concentration of 0.1-40 ng/ml. Addition of the agonist produced a rapid initial increase in light transmission, followed by a second more gradual increment. Since only the second phase increment in light transmission was associated with microscopic evidence of PMN aggregation, as also reported by ultrastructural analysis (30), this second phase response, recorded as millimeters deflection from base-line, was used to compare the response to each product with that to synthetic LTB,.

RESULTS
Generation and Initial Characterization of 5,12-Dihydroxy Derivatives of Arachidonic Acid and EPA from PMN-The products of ionophore activation of PMN in the presence of arachidonic acid from two separate experiments were resolved by RP-HPLC in Solvent I as depicted for one experiment in  (-) and immunoreactivity in the LTB, RIA (0) were determined on 1ml eluate fractions. The column was calibrated for the retention times of synthetic LTC,, LTB5, 6-trans-LTB4, and LTB, as indicated. 11.4% in the solvent front, 28.2% with 5-HETE, and 45.8% with arachidonic acid.

RP-HPLC in Solvent I of the products generated by the ionophore activation of PMN in the presence of either arachidonic acid (A) or EPA (B). Absorbance at 269 nm
The products of ionophore activation of PMN in the presence of EPA from four separate experiments were resolved by RP-HPLC in Solvent I, as depicted for one experiment in Fig. 1B. As assessed by A269 and incorporation of radiolabel from EPA, there were two peaks of products eluting before the diastereoisomers of 6-trans-LTB, and LTB4. The most polar products eluting as a doublet with average retention times of 11.5 k 0.2 min and 12.5 f 0.4 min (mean k S.D., n = 4) were presumed to be the diastereoisomers of 6-trans-LTB, based on the incorporation of radiolabel from EPA and the RP-HPLC retention times reported for these products when eluted in similar solvent systems (21). The major peak eluting at 14.6 f 0.3 min (mean f S.D., n = 4) was presumed to be LTB5 based on prominent immunoreactivity, incorporation of radiolabel from EPA and co-elution with synthetic LTB,. In the four experiments, the average quantities of the putative diastereoisomers of 6-trans-LTB,, LTB,, 6-trans-LTB,, and LTB, were 2.8 f 1.2 pg (mean f S.D., n = 4), 1.1 f 0.6 pg, 0.8 f 0.3 pg, and 0.3 f 0.5 pg per 1 X 10' cells, respectively. In the presence of radiolabeled EPA, 4.8% of the recovered counts eluted in the position of the putative diastereoisomers of 6-trans-LTB5, 2.0% with LTB6, 8.5% in the solvent front, 37.5% with 5-HEPE, and 39.8% with EPA.
Generation and Initial Characterization of Tetraene and Pentaene Double Lipoxygenase Products from Platelet Sonicates-Borgeat et al. (10,31) previously demonstrated that platelets converted 5-HETE to a single product which was characterized as (5S,12S)-6,10-trans-8,14-cis-eicosatetraenoic acid. In each of two similar experiments, 5s-HETE and rac-5-HEPE were separately incubated with sonicates of 1 X 10' by guest on March 21, 2020 http://www.jbc.org/ Downloaded from platelets and the products were resolved on RP-HPLC in Solvent I (Fig. 2, A and B). As assessed by AZe9 as well as by immunoreactivity, a single major product from 5-HETE eluted at a retention time of 22.7 min and a single major product from rac-5-HEPE eluted at 15.1 min. Additional experiments in which 2.75 pg of 5-HETE and 2.75 pg of rac-5-HEPE were separately incubated with sonicates of 22, 64, and 75 X los platelets from three separate donors yielded separate major products from each substrate which eluted from RP-HPLC as a single peak at 22. Additional Purification of Each Pentaene and Tetraene Product with Analysis of Its UV Spectrum-In order to obtain adequate quantities of homogeneous products for further characterization, comparable fractions from samples resolved by replicate RP-HPLC in Solvent I were chromatographed in Solvent 11. In this isocratic solvent system, synthetic LTC, and LTB, had average retention times of 27.5 and 59 min, respectively, and the natural 5,12-dihydroxy derivatives of arachidonic acid and EPA eluted as discrete peaks with the retention times noted in Table I. LTB, and LTB5, which  HEPE (B) with aspirin-treated platelet sonicates. Inset in B shows RP-HPLC in Solvent I1 of the double lipoxygenase products generated from rac-5-HEPE. Absorbance was monitored at 269 nm (-) and immunoreactivity in the LTB, RIA (0) was determined on 1-ml eluate fractions. The columns were calibrated for the retention times of synthetic LTCr, LTB,, and LTB6 as indicated.  eluted as single peaks in Solvent I, also eluted as single peaks when chromatographed in Solvent 11. By applying only the ascending and descending portions of the doublet representing the diastereoisomers of 6-trans-LTB4 and analogous portions of the doublet representing the diastereoisomers of 6-trans-LTBs, it was possible to resolve each diastereoisomeric pair into homogeneous peaks whose mean retention times were separated by 6.2 and 2.5 min, respectively. In the case of the 6-trans-LTBs diastereoisomers, the initial component of the doublet was presumed to be the 5S,12R-isomer by analogy to the relative retention times of the 6-trans-LTB4 diastereoisomers. (5S,12S)-6-trans-8-cis-LTB4 resolved as a distinct peak in Solvent I and again in Solvent 11. The product formed by the interaction of the rac-5-HEPE and platelet sonicates was resolved as 2 unequal peaks in Solvent I1 with retention times separated by 2 min (Fig. 2B, inset); the fractions from the descending portion of the major and less polar peak were pooled and presumed to represent (5S,12S)-6-trans-8-cis-LTB5. Both authentic LTB, and LTBS gave UV absorption maxima (Xmax) greater than 269 nm, whereas the X , , , of the other products was less than 268 nm (Table I); for each product the UV spectrum exhibited shoulders at X, , , f 10 nm.
Biological Activities of the 5,12-Dihydroxy Derivatives of Arachidonic Acid and EPA-The chemotactic activities of each of the natural 5,12-dihydroxy derivatives of arachidonic acid and EPA were compared to each other and to synthetic LTB, with PMN from 3 different donors. Natural LTB, elicited a dose-dependent chemotactic response curve which was superimposable on that of synthetic LTB,. The response to natural LTB, reached 365 f 80 PMN/10 hpf (mean f S.E.) at 50 ng/ml as compared to 497 +_ 48 PMN/10 hpf at 50 ng/ ml of natural LTB,. By interpolation, 6 ng of LTB, would attract the same number of cells as 50 ng of LTB5, thereby indicating that LTB, was approximately 1 h as potent (Fig.  4A). When the dose-related chemotactic activity of synthetic LTB5 was compared to that of synthetic LTB4 using 2 additional PMN donors, LTB, and LTB5 each reached a plateau response at 50 ng/ml even though the maximum dose tested was 100 ng/ml; the maximum response to LTB5 was only ' % that of LTB,. By interpolation, 8 ng of LTB, would attract the same number of cells as 50 ng of LTB, indicating that synthetic LTB5 was approximately ' / 6 as potent. Each of the double lipoxygenase products, (5S,12S)-6-truns-8-cis-LTB4 and (5S,12S)-6-truns-8-cis-LTB5, was about '114 as potent as LTB,. Each of the diastereoisomers of 6-truns-LTB4 was somewhat more chemotactically active than the 6-truns-LTB5 diastereoisomers and, for each series, the 6-trans-diastereoisomers were similar or slightly less active than the double lipoxygenase products (Fig. 4, A and B).
The aggregating response of human PMN to natural LTB, was analyzed in 2 experiments, to synthetic LTB, in 3 experiments, and to the other natural compounds in 6 experiments, with synthetic LTB, used as the standard in each. Natural LTB, elicited a dose-dependent second phase aggregation curve which was virtually superimposable on that elicited by synthetic LTBd (Fig. 5). The PMN aggregation in response to natural LTB, reached a plateau at 20 ng/ml as compared to the response plateau reached at 5 ng/ml of synthetic or natural LTB,. Further, the maximal aggregating response elicited by LTB5 at 20 ng/ml was Y 3 that obtained with LTB, at 5 ng/ml. By interpolation, 1 ng/ml of LTB4 would elicit the same response as 20 ng/ml of LTB,, indicating that LTBs was approximately l/zo as potent. When the dose-related PMN aggregating activity of synthetic LTB5 was compared to that of synthetic LTB,, LTB, was about %3 as potent as LTB,. Of the other natural 5,12-dihydroxy derivatives of arachidonic acid and EPA, only (5S,12S)-6-truns-8-cis-LTB4 had PMN aggregating activity and that was maximal at 20 ng/ml, but achieved only '16 the maximal response evoked by 5 ng/ml of LTB4. Thus, both LTB, and (5S,12S)-6-truns-8-cis-LTB4 performed as weak and partial agonists.

DISCUSSION
EPA, which is incorporated from dietary sources into membrane phospholipids along with arachidonic acid (17)(18)(19)(20)(21)32) and other polyunsaturated fatty acids (33), has been documented to yield pentaene-derived products via the cyclooxygenase pathway. Two of these, thromboxane A3 and prostaglandin 13, are markedly less potent than the analogous arach- idonic acid derivatives (19,34). The limited information available on the sulfidopeptide pentaene products of the lipoxygenase pathway reveals little or no difference in potency of their spasmogenic activities on ileal smooth muscle (35). Because of the likely fundamental importance of leukocyte chemotaxis and aggregation in response to LTB, in inflammatory processes and the complete lack of comparative biologic studies with the pentaene analogues, the natural 5,12dihydroxy derivatives of both arachidonic acid and EPA were compared within each series and between the two, with the same source of human cells being used to generate the eight natural products under study.
Human PMN were interacted with ',C-labeled and unlabeled arachidonic acid to generate authentic LTB, and the 6truns-LTB, diastereoisomers and with I4C-labeled and unlabeled EPA to produce the additional products LTB5 and the 6-truns-LTB5 diastereoisomers. After resolution by RP-HPLC (Fig. 1, A and B ) , the products were tentatively identified by their elution times relative to standard markers, the incorporation of radiolabel which defined the polyunsaturated fatty acid source of the product, and their competition in a RIA for LTB,, which revealed whether a product expressed immunochemically detectable determinants. Synthetic 5S-HETE, which corresponded to a natural 5-lipoxygenase pathway product, and ruc-5-HEPE were interacted with sonicates of human platelets to yield the double lipoxygenase products (5S,12S)-6-truns-8-cis-LTB4 and (5S,12S)-6-truns-8-cis-LTB5, respectively. These products were tentatively identified by their elution as single reaction products approximately 2 min after synthetic LTB, and LTB5 standards and by their immunoreactivity (Fig. 2).
The eight natural products were resolved by RP-HPLC in a second solvent and each was further characterized by its immunoreactivity, UV absorption spectrum, and biologic activity. Authentic LTB, and LTB5 resolved as single peaks in both Solvents I and I1 at the elution times of the appropriate synthetic markers and yielded characteristic UV absorption spectra with X , , , at 269.6 nm and 269.2 nm, respectively, and shoulders at X , , , f 10 nm. The resolution of the diastereoiso-mers of 6-trans-LTB4 and of 6-truns-LTB5 required that each doublet obtained in Solvent I be divided into fractions composed only of the ascending and descending limbs. These fractions were then fully resolved in Solvent I1 with peak retention times that differed by 6.2 min for (5S,12R)-and (5S,12S)-6-truns-LTB4 and by 2.5 min for (5S,12R)-and (5S,12S)-6-truns-LTB5. The order of elution for the diastereoisomers of 6-trans-LTB4 had been previously established with synthetic standards (36,37) and was assumed also to represent the order of elution for the 6-truns-LTB5 diastereoisomers. The UV absorption spectra of the 6-trans-diastereoisomers of the tetraene and pentaene series yielded a X , .
of 267.8 nm (Table I), with shoulders at X , , k 10 nm. The 1.8 nm difference for X , , , between LTB, and each'of its isomers is consistent with the observations of Borgeat and Samuelsson (7) in which the difference in the Xmax was 2 nm. In contrast to the present study, the only previous UV absorption data on 5,12-diols of the pentaene series failed to recognize a different Xmax for authentic LTB5 and its 6-trans-LTB5 diastereoisomers (21). Whereas (5S,12S)-6-trans-8-cis-LTB, resolved as a single peak in both Solvent I and Solvent 11, the products formed from the ruc-5-HEPE resolved in Solvent I1 as two unequal peaks whose mean retention times were separated by 2 min. The descending portion of the less polar peak (Fig. 2B, inset) contained only the predominant product and exhibited a X , , of 267.8 nm with shoulders at Xmax f 10 nm, identical with that of (5S,12S)-6-trans-8-cis-LTB, (Table I) (10).
Competitive inhibition for binding of [3H]LTB4 to immune rabbit plasma by the natural products served as an additional parameter for identifying the degree of structural relationship between pairs of pentaene and tetraene products. The rank order of cross-reactivity was similar for both the pentaene and tetraene series: LTB5/LTB4 > (5S,12S)-6-trans-8-cistrans-LTB5/LTB4 (Fig. 3). That each of the compounds in the pentaene series was more immunoreactive on a weight basis than the analogous compounds in the tetraene series may have resulted from preparing the immunogen by coupling

Leukotriene B5
the carrier to LTBl through the 12-oxo function. The resulting proximity of the C-17 double bone to the coupling site may contribute to active site specificity by the greater rigidity of the pentaene. The definition of the degree of cross-reactivity for each moiety provides the possibility of measurement in complex biologic fluids following their resolution. The chemotactic and aggregating activities of LTB5 for human PMN were approximately 1 h and %3 to l/zo as potent, respectively, as those of LTB, (Figs. 4 and 5 ) . The impaired chemotactic activity of LTB5 has been reported in a preliminary statement (38). The assessment of potency was made by comparing the concentrations of natural and synthetic LTB5 giving a maximal effect with the interpolated values of natural and synthetic LTB, giving a comparable response. Further, the maximum effect achieved by LTB, was less than that of LTB4, especially in the aggregation assay, indicating that LTB, functions as a weak partial agonist as compared to LTB4. The double lipoxygenase products, (5S,12S)-6-trans-8cis-LTB4 and (5S,12S)-6-truns-8-cis-LTB5, were about 14-fold less potent than LTB, as chemotactic factors and (5S,12S)-6-trans-8-cis-LTB5 failed to elicit aggregation at the maximum dose tested whereas (5S,12S)-6-trans-8-cis-LTB4 was a weakly active partial agonist. The diastereoisomers of 6-trans-LTB5, as well as those of 6-trans-LTB4, did not elicit aggregation at the maximal doses studied and each set of these nonenzymatic degradation products of their respective epoxide (LTA) leukotriene were minimally active as a chemotactic factor. Thus, LTB5, the pentaene analog of LTB4, and its naturally occurring isomers are markedly functionally impaired in their agonist actions on human PMN, a finding which may have important implications for the use of diets rich in EPA to attenuate the functions of the cyclooxygenase pathway (17-21, 32, 34, 39).