European Journal of Pharmacology, 218 (1992)369-372

369

© 1992 Elsevier Science Publishers B.V. All rights reserved 0014-2999/92/$05.00

EJP 21100 Short communication

13-Hydroxyoctadecadienoic acid attenuates oedema formation induced by leukotriene B 4 in vivo in rabbit skin T h e r e s a L. Buckley, M a r c o J. V a n de V e l d e , P a u l A.J. H e n r i c k s , F e r d i E n g e l s and F r a n s P. N i j k a m p Department of Pharmacology, Faculty of Pharmacy, Universityof Utrecht, P.O. Box 80.082, 3508 TB Utrecht, Netherlands

Received 16 April 1992,revised MS received 12 June 1992,accepted 16 June 1992

Intradermal 13-hydroxyoctadecadienoic acid (13-HODE; 10-11-10 -9 mol/site) inhibited oedema formation induced by the neutrophil-dependent mediator leukotriene B 4 (LTB 4) in the presence of calcitonin gene-related peptide (CGRP; 10 -11 tool/site) in rabbit skin. In contrast, the responses to the direct acting mediators bradykinin and histamine were unaffected by 13-HODE. 13-HODE failed to induce oedema formation in rabbit skin when injected alone or in the presence of the potent vasodilator CGRP. These results present a novel interaction between 13-HODE and LTB4 that could have important implications in the pathogenesis of inflammation. 13-HODE (13-hydroxyoctadecadienoic acid); Oedema formation; Leukotriene B4; Skin (rabbit)

1. Introduction

Studies investigating the role of fatty acids in inflammation have focussed predominantly on arachidonic acid and its metabolites. However, recently there is increasing awareness of the importance of linoleic acid (and its metabolites) which is also present in plasma membranes. Linoleic acid is metabolised into various products including 13-hydroxyoctadecadienoic acid (13-HODE) by endothelial cells (Buchanan et al., 1985b), epithelial cells (Oosthuizen et al., 1990) and leukocytes (Claeys et al., 1982; Engels et al., 1991). 13-HODE, which is continuously synthesised by endothelial cells under basal conditions, has been reported to inhibit the adhesion of platelets to endothelial cells in vitro (Buchanan et al., 1985a). However, 13-HODE induced chemotaxis of bovine and human leukocytes in vitro (Henricks et al., 1991). Oedema formation and neutrophil accumulation are prominent features of inflammation. Mediators such as the arachidonate metabolite leukotriene B 4 (LTB 4) can induce neutrophil-dependent oedema formation in rabbit skin whereas histamine and bradykinin induce o e d e m a by acting directly on the endothelium (Wedmore and Williams, 1981). In contrast vasodilators, including calcitonin gene-related peptide (CGRP)

Correspondence to: T.L. Buckley, Department of Pharmacology, Faculty of Pharmacy, University of Utrecht, P.O. Box 80.082, 3508 TB Utrecht, Netherlands. Tel. 31.30.537 352, fax 31.30.537 420.

and prostaglandin 12 (PGI2), do not induce oedema formation but as a consequence of their vasodilator activities they can synergistically potentiate oedema induced by mediators of increased microvascular permeability (Brain and Williams, 1985; Buckley et al., 1991). The possible role of 13-HODE in inflammation has been the subject of limited studies most of which have investigated the 'chemorepellent' effects of 13-HODE against platelet adhesion. In this study, we investigated whether 13-HODE could induce or modulate oedema formation in rabbit skin.

2. Materials and methods 2.1. Preparation o f 13-HODE

13-HODE was prepared as described in Engels et al. (1991). Briefly, linoleic acid was incubated with lipoxidase-1 enzyme (type IV) at 4°C. The formation of the linoleic acid intermediate 13-hydroperoxyoctadecadienoic acid (13-HPODE) was followed by measuring samples of the incubation mixture at appropriate intervals in a spectrophotometer (235 nm). After the linoleic acid had been metabolised the incubation was stopped by adding HCI to obtain a p H of 4. 13-HPODE was separated from non-metabolised linoleic acid using thin layer chromatography and reduced with sodium borohydrate. After reduction the purity of 13-HODE was measured by reverse phase

370 F~(4)~

H P L C and was found to be > 95% (Engels et al., 1991). The formed 13-HODE was then stored in methanol at - 80°C.

2.2. Measurement of oedema formation

F--(4)~

8O

5o !

o

40

Rabbits (New Zealand White; 2.5-3.0 kg) were anaesthetised with sodium pent0barbitone (30 mg k g - 1) after the dorsal fur was shaved. O e d e m a formation was measured over a 30 min period by the local extravascular accumulation of i.v. injected t25I-albumin (0.2 MBq per rabbit; mixed in a 2.5% Evans blue solution) as previously described (Williams, 1977). The agents under investigation were injected intradermally in 0.1 ml volumes, each treatment having six replicates with injections given according to a balanced site plan. At the end of the accumulation period a 10 ml cardiac blood sample was taken and the animal was killed with an overdose of sodium pentobarbitone. The dorsal skin was subsequently removed and the injected sites were punched out with a 17 mm punch. The injected sites, together with plasma samples, were counted in a gamma counter. Exudate volumes, expressed in /zl of plasma, were calculated by dividing 125I counts per skin site by counts per 1/~1 plasma and then subtracting/zl leakage at saline-injected sites (saline 4.2 + 0.I ~ l / s i t e , mean + S.E.M., n = 7 rabbits).

2.3. Chemicals Bradykinin, linoleic acid, lipoxidase-1 enzyme (type IV) and histamine were purchased from Sigma Chemical Company Ltd., St. Louis, Missouri, USA. H u m a n o~CGRP was purchased from Bachem Feinchemikalien AG, Bubendorf, Switzerland. LTB 4 was a gift from Dr. A.W. Ford-Hutchinson, Merck Frosst Canada Inc., Quebec, Canada. 125I-albumin was purchased from Amersham Nederland BV, 's-Hertogenbosch, The Netherlands.

2.4. Statistical analysis The data, expressed as mean _+ S.E.M., were analysed using two-way analysis of variance (ANOVA). Differences were determined using the N e w m a n Keuls procedure and P < 0.05 was considered to be significant.

~ 30 c

F--(4)~

(7)

20

o

tool/site Fig

1

LTB4

hLstarnine

13 I-~C)E

i0 -I°

3.10 -9

i0 -I°

Skin p l a s m a v o l u m e

13- HOOE

10 -9

CGIqP

10 -11

(/zl/site) induced by histamine,

LTB 4

and 13-HODE. Responses induced by these mediators are shown in the presence (filled columns) and absence (open columns) of C G R P

(10-u mol/site). All test agents were co-injected in the rabbit skin. Results are expressed as mean, with vertical lines showing S.E.M., for (n) rabbits.

(10 -1° and 10 -9 mol/site) did not induce oedema formation when injected alone and co-injection with C G R P failed to potentiate this response (fig. 1). In addition, 13-HODE (10 - l ° mol/site) did not potentiate oedema formation induced by histamine (histamine 31.1 ___2.7 ~l, 13-HODE 3.0 _+ 0.7/zl, histamine + 13H O D E 28.9 _+ 2.7 /~l, mean _+ S.E.M., n = 6 sites) or the small oedema response induced by LTB 4 (LTB 4 11.1 _+ 1.7/zl, LTB 4 + 13-HODE 12.1 _+ 0.7/xl, mean _+ S.E.M., n = 6 sites).

20 10

S o

o

E -~o o ~o ~ c 2_ U

-20 -30 -40

-50 -60

-ii 10 13-HODE

-i0 10 (mol/s~te)

Fig. 2. Percentage (%) change in oe de ma responses induced by

3. Results Histamine ( 3 X 10 -9 mol/site) and LTB 4 (10 -1° mol/site) induced comparable oedema formation in the presence of the potent vasodilator C G R P (10 -11 mol/site); C G R P alone did not induce a leakage response (fig. 1). The linoleic acid metabolite 13-HODE

inflammatory mediators in the presence of 13-HODE. Responses induced by histamine (3 x 10 -9 mol/site; hatched bar), LTB 4 (10-I0 mol/site; filled bar) and bradykinin (10-10 mol/site; cross hatched bar) in the presence of CGRP (10 -11 mol/site)were 68.4+ 11.5, 70.15:4.4 and 123.9+ 11.5 /zl (means: S.E.M.), respectively. Results are expressed as mean % change from control, with vertical lines showing S.E.M. for n = 4 rabbits in each group except # where n = 3 rabbits. 13-HODE, at doses of 10-11 mol/site and 10-l° tool/site, significantly inhibited (P < 0.05; ANOVA) oedema formation induced by LTB4 + CGRP.

371 In contrast, 13-HODE significantly attenuated the marked oedema formation induced by LTB 4 + C G R P in the rabbit skin whilst having no effect on responses induced by histamine + C G R P or bradykinin + C G R P (fig. 2). 13-HODE, at doses of 10 -I1 m o l / s i t e and 10-~° mol/site, inhibited responses induced by LTB 4 (10 -l° m o l / s i t e ) + C G R P (10 -11 mol/site) by 31.6 _+ 6.4 and 39.6 + 5.8% respectively. In studies where a higher dose of 13-HODE (10 -9 mol/site) was used, 13-HODE inhibited responses induced by LTB 4 (10-H~ mol/site) + C G R P (10-Jl mol/site) by 35.5 + 8.5% (mean _+ S.E.M., n = 6 sites). Although this inhibition was comparable to the inhibition effected by 13-HODE, at a dose of 10 -l° mol/site, responses induced by bradykinin (10-10 mol/site) + C G R P (10-11 mol/site) were also partially inhibited by the higher dose of 13-HODE (14.0 _+ 8.6% inhibition; mean _+ S.E.M., n = 6 sites). It would appear that higher doses of 13H O D E exhibit a non-specific action on oedema formation induced by direct acting and neutrophil-dependent inflammatory mediators.

4. Discussion This present study showed that 13-HODE significantly inhibited oedema formation induced by the neutrophil-dependent mediator LTB 4 in the presence of C G R P (fig. 2). In addition 13-HODE did not induce a leakage response alone or in the presence of the potent vasodilator C G R P (fig. 1). In vitro studies have shown that LTB 4 is a potent chemotactic agent for polymorphonuclear leukocytes (PMNs; Ford-Hutchinson et al., 1980). In addition, LTB 4 and other chemotactic agents such as the complement fragment C5a and the bacterial peptide n-formyl-methionyl-leucyl-phenylalanine (FMLP), are mediators of increased microvascular permeability in rabbit skin. Oedema formation induced by these mediators was found to be dependent on the presence of circulating PMNs (Wedmore and Williams, 1981). Interestingly, 13-HODE also has chemotactic activity (albeit weaker) for bovine and human PMNs (Henricks et al., 1991), however, in this present study, 13-HODE failed to induce oedema formation in rabbit skin. Moreover, 13-HODE also exhibited anti-inflammatory activity against responses mediated by LTB 4 in the presence of CGRP. There are various possibilities to explain this surprising and novel finding. Firstly, 13-HODE could decrease local blood flow. Decreasing microvascular blood flow has been shown to attenuate oedema formation induced by neutrophildependent and direct acting mediators in rabbit skin (Brain et al., 1989). However, the responses to histamine (which were comparable to the LTB4-induced responses) and bradykinin were unaffected by co-injec-

tion of 13-HODE; this explanation is therefore unlikely. Secondly, 13-HODE has been shown to release prostacyclin (PGI 2) from the endothelium. Rampart and Williams (1986) demonstrated that an i.v. infusion of PGI 2 (50 ng kg- ~ m i n - ~) suppressed neutrophil-dependent oedema formation while leakage induced by direct acting mediators was unaffected. It is therefore possible that endogenous PGI 2 is responsible for the in vivo modulation of LTB4-induced oedema formation as shown in fig. 2. However, PGI 2, like CGRP, is a vasodilator and intradermal PGI 2 would be expected to enhance leakage responses. In this study intradermal 13-HODE did not potentiate oedema formation induced by histamine (data in text). Finally, 13-HODE could be acting on the PMN directly. Linoleic acid metabolites including 13-HODE, which are lipophilic mediators, could be incorporated into the plasma membrane. This incorporation could alter the membrane fluidity hence influencing the PMNs response to neutrophil-dependent mediators. Alternatively the anti-inflammatory activity of 13H O D E could be a receptor-mediated event and we are currently investigating the presence of a specific receptor for 13-HODE on the PMN plasma membrane. If either membrane incorporation occurs or a functional receptor is present, 13-HODE could be an important modulator of PMN function. In physiological situations 13-HODE, which is continuously released from unstimulated endothelial cells, could act to inhibit PMN adhesion (Buchanan et al., 1985a). This action would prevent unnecessary sticking of PMNs to the healthy endothelium. Therefore the balance of linoleic acid and arachidonic acid and the release of their respective metabolites could be critical for the normal functioning of the endothelium. Disturbance of this balance could have disastrous pattiophysiological consequences. In future studies we will investigate further the actions of linoleic acid metabolites on neutrophil function in vitro and oedema formation in vivo.

Acknowledgement T.L.B. is in receipt of a Royal Society European Exchange fellowship.

References Brain, S.D. and T.J. Williams, 1985, Inflammatory oedema induced by synergism between calcitonin gene-related peptide (CGRP) and mediators of increased vascular permeability, Br. J. Pharmacol. 86, 855. Brain, S.D., D.C. Crossman, T.L. Buckley and T.J. Williams, 1989, Endothelin-l: Demonstration of potent effects on the microcirculation of humans and other species, J. Cardiovasc. Pharmacol. 13, S147.

372 Buchanan, M.R., R.W. Butt, Z. Magas, J. Van Ryn, J. Hirsch and D.J. Nazir, 1985a, Endothelial cells produce a lipoxygenase derived chemorepellant which influences platelet/endothelial cell interactions: Effects of aspirin and salicylate, Thromb. Haemost. 53, 306. Buchanan, M.R., T.A. Haas, M. Lagarde and M. Guichardant, 1985b, 13-Hydroxyoctadecadienoic acid is the vessel wall chemorepellant, LOX, J. Biol. Chem. 260, 16056. Buckley, T.L., S.D. Brain, M. Rampart and T.J. Williams, 1991, Time-dependent synergistic interactions between the vasodilator calcitonin gene-related peptide (CGRP) and mediators of inflammation, Br. J. Pharmacol. 103, 1515. Claeys, M., M.-C. Coene, A.G. Herman, G.H. Jouvenaz and D.H. Nugteren, 1982, Characterisation of monohydroxylated lipoxygenase metabolites of arachidonic and linoleic acid in rabbit peritoneal tissue, Biochim. Biophys. Acta 713, 160. Engels, F., G.C.R. Kessels, A.J.M. Schreurs and F.P. Nijkamp, 1991, Production of arachidonic acid and linoleic acid metabolites by human bronchoalveolar lavage cells, Prostaglandins 42, 441. Ford-Hutchinson, A.W., M.A. Bray, M.V. Doig, M.E. Shipley and

M.J.H. Smith, 1980, Leukotriene B, a potent chemokinetic and aggregating substance released from polymorphonuclear leukocytes, Nature 286, 264. Henricks, P.A.J., F. Engels, H. Van der Vliet and F.P. Nijkamp, 1991, 9- and 13-hydroxy-linoleic acid possess chemotactic activity for bovine and human polymorphonuclear leukocytes, Prostaglandins 41, 21. Oosthuizen, M.J., F. Engels, B. Van Esch, P.A.J. Henricks and F.P. Nijkamp, 1990, Production of arachidonic and linoleic acid metabolites by guinea-pig tracheal epithelial cells, Inflammation 14, 401. Rampart, M. and T.J. Williams, 1986, Polymorphonuclear leukocyte-dependent plasma leakage in the rabbit skin is enhanced or inhibited by prostacyclin, depending on the route of administration, Am. J. Pathol. 124, 66. Wedmore, C.V. and T.J. Williams, 1981, Control of vascular permeability by polymorphonuclear leukocytes in inflammation, Nature 289, 646. Williams, T.J., 1977, Chemical mediators of vascular responses in inflammation: two mediator hypothesis, Br. J. Pharmacol. 62, 447.