Linoleic and dihomo-γ-linolenic acids modulate granuloma growth and granuloma macrophage eicosanoid release

European Journal of Pharmacology, 124 (1986) 325-329

325

Elsevier

LINOLEIC AND DIHOMO-~-LINOLENIC ACIDS MODULATE GRANULOMA GROWTH AND G R A N U L O M A M A C R O P H A G E E I C O S A N O I D RELEASE GRAHAM R. ELLIOTT *. MARTINUS J.P. ADOLFS, MARJAN VAN BATENBURG and IVAN L. BONTA Pharmacology Department. Erasmus Universi(v Rotterdam. Postbox 1738, 3000 DR Rotterdam. The Netherlands

Received 22 May 1985, revised MS received 23 January 1986, accepted 4 March 1986

G.R. ELLIOTT, M.J.P. ADOLFS, M. VAN BATENBURG and I.L. BONTA, Linoleic and dihomo-y-linolenic acids modulate granuloma growth and granuloma macrophage eicosanoid release, European J. Pharmacol. 124 (1986) 325-329. We have already demonstrated that arachidonic acid (AA) inhibites carrageenin-induced granuloma growth in vivo and that this effect is related to an increase in prostaglandin E 2 formation. As prostaglandin E1 has been shown to be more effective in inhibiting granuloma growth than prostaglandin E 2 we investigated the effect of linoleic acid (LA) (18 : 2 ~6) and dihomo-7-1inolenic acid (DHGLA) (20 : 3 ~6), potential precursors of prostaglandin E 1, in this model. LA and DHGLA inhibited the development of carrageenin-induced granulomas in the rat when injected locally. Both fatty acids (FA) stimulated the release of prostaglandin (PG) E 1 from granuloma macrophages (M~) in vitro, DHGLA being most effective. LA had little effect on the release of PGE 2, 6ketoPGFl,, the stable product of prostacyclin (PGI 2) or thromboxane (Tx) B2, the stable metabolite of TxA 2. DHGLA had no effect on the release of 6ketoPGF~,, but inhibited PGE 2 and, to a lesser extent, TxB 2 synthesis. Dihomo-y-linolenic acid PGE 1

Granuloma macrophage

I. Introduction The essential fatty acid (EFA), linoleic acid (LA), is metabolised by desaturation and chain elongation to dihomo-7-1inolenic acid ( D H G L A ) , the precursor of monoenoic cyclooxygenase metabolites such as prostaglandin E 1 (PGE1), and arachidonic acid (AA), the precursor of the dienoic P G E 2, (Lagarde et al., 1981). P G E 1 and P G E 2 have been shown to inhibite the macrophage ( M ~ ) dependent, tissue proliferative, stage of carrageenin sponge granuloma growth in the rat ( P a r n h a m et al., 1979; Bonta and Parnham, 1979), p r o b a b l y by stimulating M ~ c A M P synthesis so depressing expression of their pro-inflammatory functions (Bonta et al., 1981). We have recently demonstrated that A A also stimulates M ~ c A M P synthesis and inhibits granuloma growth in vivo (Elliott et al., 1983; 1985). As P G E 1 was more * To whom all correspondence should be addressed. 0014-2999/86/$03.50 ¢3 1986 Elsevier Science Publishers B.V.

Granuloma growth

Linoleic acid

PGE 2

potent in inhibiting granuloma growth than P G E 2 we have now investigated the effect of its fatty acid (FA) precursor, D H G L A , on the growth of carrageenin sponge granulomas in the rat. We have also assessed the anti-inflammatory potential of LA in this model as it has been reported to be effective in the treatment of adjuvant induced polyarthritis in the rat (Stuyvesant and Jolley, 1976). Changes in granuloma formation in vivo were related to changes in granuloma Mq~ eicosanoid release induced by D H G L A and LA in vitro.

2. Materials and methods Ethanolic solutions of D H G L A and LA (purity greater than 95%) were gifts of Unilever, Vlaardingen, The Netherlands. Just before use the FAs were diluted to the required concentration in Geys Balanced Salt Solution (GBSS). Antisera against

326

PGE1, PGE 2, thromboxane Be (TxB2) and 6ketoPGF~, were bought from the Institute Pasteur. Radiolabelled eicosanoids were obtained from Amersham International, England. Standard compounds were ordered from Sigma Fine Chemicals Ltd.

TABLE 1 The effect of linoleic and dihomo-T-linolenic acids on carrageenin sponge induced g r a n u l o m a s in the rat. Results are the means_+S.E.M, of 10 granulomas. Significance values were c a l c u l a t e d using the M a n n - W h i t n e y U-test. Treated vs. control. ~' P < 0.05: b p < 0.025; " P < 0.001: a 5 /*g vs. 10 p,g dihomo~'-Iinolenic acid, P < 0.025. * / , g fatty a c i d / s p o n g e . G r a n u l o m a dry weight (mg)

2.1. The effect of LA and DHGLA on carrageenin sponge granuloma formation The methods used have already been described in detail elsewhere. Briefly, male Wistar rats (175200 g) were placed in groups of 5, matched for body weight, and cannulated, carrageenin-impreghated sponges were implanted subdorsally (2 per animal) as described by Bonta et al. (1979). On days 4, 5, 6 and 7 after implantation 250 ktl of the FA solution, or solvent control, was injected into each granuloma via the cannula. On day 8 the granulomas were excised, the sponges removed and the granuloma tissue weighed. Granuloma dry weight was measured after heating overnight at 80°C.

2.2. The effect of LA and DHGLA granuloma M ~ eicosanoid release

on 4 day

The FAs, 10 >l of the ethanolic solutions, were added to 1 ml Plastibrand polypropylene reaction vials and the ethanol evaporated under a stream of nitrogen. Granuloma Mq)s were isolated from 4 day granulomas by pronase digestion according to the method of Bonta et al. (1981). The Mq) preparation was greater than 98% viable and greater than 85% pure. The cells (2 × 106 ml GBSS) were incubated with the FAs for 1 h at 37°C in a water bath, the cells centrifuged and analysed for c A M P (Gilman, 1970) and the supernatants stored at - 7 0 ° C for measurement of eicosanoids by radioimmunoassay (RIA).

3. Results

3.1. The effect of LA and DHGLA on granuloma growth LA inhibited granuloma growth (expressed as dry weight) with a maximal effect at 5 txg/sponge.

Linoleic acid Dihomo-7linolenic acid

0 big *

1 p.g

5 #g

10 big

220+15

184_+12 "

175+14 h

184_+13

260_+18

220+11

190+

230+18d

8~

D H G L A also had an optimal inhibitory action at 5 , t g / s p o n g e but at 10 /*g/sponge there was a clear reduction in its effectiveness (table 1).

3.2. The effect of LA and DHGLA granuloma McI) eicosanoid release

on 4 day

Cross reactivity between the RIA antisera used and LA (10/*g/ml) and D H G L A (10/xg/ml) was below the level of detection of the assays. LA had no effect on the release of PGE 2, TxB 2 or 6ketoPGF1, ~from granuloma Mq~s over a 1 h incubation period (fig. 1). D H G L A however inhibited P G E x synthesis in a dose-dependent manner but had little effect on TxB 2 and failed to influence the levels of 6ketoPGF1, (fig. 2). Both FAs stimulated PGE1 formation, D H G L A being the more effective substrate (fig. 3). Basal cAMP levels were below the limit of detection of the protein binding assay used (1.0 pmol) and were not stimulated to detectable levels after incubation with the FAs. Cross reactivity between the PGE 1 antisera and P G E 2 was 15% (at 50% binding between PGE I and its antiserum). However, in view of the fact that PGE z levels decreased in the presence of D H G L A , it is unlikely that this cross reactivity contributed significantly to the levels of PGE~ assayed. This proposal is strengthened by the fact that PGE~ could not be detected in the supernatant of control Mq)s, although P G E : was present.

327

ng/ml 1.9-

ng/ml 6 Keto PGFIa

1.9 •

PGE

1.5

2

t 6 Keto PGF1a

1.5

o

1.0

1.0 eO

0.5

~

TXB2

PGE 2 0,5 ~

IJg linoleic acid Fig. 1. The effect of LA on 4 day granuloma Mq~ PGE2, TxB 2 and 6ketoPGF1,~ synthesis. M q~s (2× I06/mi) were incubated with LA in GBSS for I h at 37°C. The cells were spun down and the supernatant stored at -70°C for later analysis of eicosanoids and the cell pellet analysed for cAMP. Results are the mean-l-S.D, of triplicate incubations for each LA concentration. Macrophages were isolated from 4 day granulomas and pooled. The effects of LA and DHGLA were investigated using the same macrophage preparation.

0

T

X

; 1

B

2

's

,'0

1Jg dihomo- ~ -linolenic acid Fig. 2. The effect of D H G L A on 4 day granuloma Mq~ PGE2, TxB 2 and 6ketoPGF1, synthesis. Conditions are as for fig. 1. Significance values, vs. controls, were obtained using Student's t-test (* P < 0.05, ** P < 0.01, *** P < 0.005).

PGE 1 (ng/rnl) 1.5

dihomo r linolenic acid

4. Discussion

We reported earlier that AA (1-10 /tg/sponge) inhibited the development of carrageenin sponge granulomas in the rat (Elliott et al., 1985). G r a n u l o m a inhibition was related to a preferential increase in P G E 2 synthesis both in vivo (granuloma exudate) and in vitro (granuloma M ~ eicosanoid release). We have now shown that LA and D H G L A can also inhibit granuloma growth in vivo although the effect was maximal at 5 t~g/sponge and, in the case of D H G L A , was reversed at 10/~g/sponge. Further we could detect no relationship between the effect of these FAs on P G E (E 1 + E2) synthesis by granuloma Mq~ in vitro and the pattern of inhibition in vivo. It is

1.0 linoleic acid

0.5

0 0

1 Fatty

5 a c i d (~Jg/ml)

10

Fig. 3. The effect of LA and D H G L A on 4 day granuloma M ~ PGE 1 synthesis. Conditions are as for fig. 1.

328

possible to argue however, in the light of our previous findings, that at least part of the inhibitory effect of these FAs was mediated via PGE~. In that case, it could be reasoned, LA was less inhibitory as it stimulated PGE~ synthesis to a lesser degree. To our knowledge this is the first report demonstrating that Mq~s can metabolise exogenous LA to PGE~. As LA is released by activated cells, for example alveolar Mq~s of rabbits injected with complete Freunds adjuvant (Freeman and Lynn, 1980), it is possible that PGE~ could play a role in regulating inflammatory reactions in vivo. This proposal is supported by the report that antisera against PGE~ inhibited experimental allergic encephalomyelitis in rats. At low concentrations PGE~ and LA initially reversed this effect but, as the concentration of these mediators increased, an inhibition was again observed (Mertin and Stackpoole, 1981; Gurr, 1982). That is, PGE~ and LA had bell shaped or biphasic dose-response curves similar to that observed by us for inhibition of granuloma growth by D H G L A . The reason for the lack of inhibition observed at 10 /~g/sponge D H G L A is not clear, D H G L A cannot be metabolised to leukotrienes as it lacks desaturation at the A5 position (Willis, 1981), but can give rise to polyhydroxy FAs (Lagarde et al., 1981), which could have a pro-inflammatory action. However AA, which has been shown to be metabolised to hydroxy FAs by granuloma tissue (Bragt et al., 1980), would be expected to show a similar dose-response pattern if these compounds played an important role in granuloma development. This is not the case. Similarly it would not appear likely that granuloma growth was inhibited by some non-specific, membrane, effect of D H G L A . It is clear however that the anti-proliferative actions of LA and D H G L A are complex, possibly due to interactions between the various FAs and endogenous AA or its metabolites. LA has been reported by a number of authors to inhibit the release of AA metabolites from cells in culture (Spector et al., 1981; Kaduce et al., 1982). However we could not detect any change in the release of dienoic cyclooxygenase metabolites during a 1 h incubation period. PGEa formation was stimulated however indicating that this FA metabolite is preferentially formed from LA.

D H G L A , besides stimulating PGE 1 synthesis, partly inhibited P G E 2 release and induced a smaller, but definite, reduction in TxB 2 production. No alteration in 6ketoPGF~,, synthesis was detected. D H G L A and AA are metabolised to PGs by the same cyclooxygenase enzymes. However, as for LTs, because D H G L A lacks desaturation at the A5 position, it cannot be converted to prostacyclin. Further, because monoenoic endoperoxides are poor substrates for thromboxane synthestase, at least in platelets, little TxB~ is formed (Lagarde et al., 1981). Thus, PGE~ is one of the major cyclooxygenase metabolites synthetized from D H G L A . It is not surprising therefore that PGE 2 formation was inhibited by D H G L A , due to competition between the mono and dienoic endoperoxides for the PGE isomerase. It is not clear however why 6ketoPGF~, levels should remain unaffected. It could be argued that if dienoic endoperoxides were prevented from being converted to PGE 2 and, to a lesser extent, TxB 2 then they would be shunted via the PGI 2 synthetase. This was not the case. Interestingly rat Mq)s synthetized more PGE 2 from AA (8.0 n g / m l from 10 p,g/ml AA under identical conditions), than PGE~ from either LA (0.5 n g / m i ) or D H G L A (1.4 n g / m l ) (Elliott et al., 1985; this study). Platelets however metabolise D H G L A in preference to AA in vitro (Lagarde et al., 1981). A detailed study of LA and D H G L A metabolisms by rat Mq~s, using cells preloaded with the different FAs, is needed to see if endogenous AA is also preferentially metabolised by rat M~s. We have shown that pharmacological doses of AA can inhibit granuloma growth in the rat and other workers have shown that LA inhibits adjuvant-induced polyarthritis (Stuyvesant and Jolley, 1976) and experimental allergic encephalomyelitis in the rat {Mead and Mertin, 1978). Similarly orally administered D H G L A or evening primrose oil (containing both LA and ¥-iinolenic acid, the immediate precursor of D H G L A ) have proved useful in treating a number of immunoinflammatory conditions in animals and humans (Horrobin et al., 1981; Kunkel et al., 1981). However, our results, and those of others (Gurr, 1982), indicate that levels of these FAs, above a critical con-

329 centration, could have an effect opposite to that desired. Further experiments into the mechanism by which DHGLA modified granuloma growth could b e u s e f u l in a s s e s s i n g t h e p o t e n t i a l advantages and problems associated with EFA treatment of inflammatory conditions.

Acknowledgement We wish to thank Unilever, Vlaardingen, The Netherlands, for providing linoleic and dihomo-y-linolenic acids.

References Bonta, 1.L., M,J.P, Adolfs and M.J. Parnham, 1979, Cannulated sponge implants for the study of time dependent pharmacological influences on inflammatory granuloma, J. Pharmacol. Meth. 2, 1. Bonta, I.L., M.J.P. Adolfs and M.J. Parnham, 1981, Prostaglandin E 2 elevation of cyclic-AMP in granuloma macrophages at various stages of inflammation: Relevance to anti-inflammatory and immunomodulatory functions, Prostaglandins 22, 95. Bonta, I.L. and M.J. Parnham, 1979, Time-dependent stimulatory and inhibitory effects of prostaglandin E 1 on exudative and tissue components of granulomatous inflammation in rats, Br, J. Pharmacol. 65, 465. Bragt, P.C., I.L. Bonta and M.J.P. Adolfs, 1980, Cannulated teflon chamber implant in the rat: A new model for continuous studies on granulornatous inflammation, J. Pharmacol. Meth. 3, 51. Elliott, G.R., M.J.P. Adolfs and I.L. Bonta, 1983, The modulation of macrophage activity by exogenous arachidonic acid, in: lmmunotoxicology, eds. Gibson, Hubbard and Parke (Academic Press, London) p. 463. Elliott, G.R., M.J.P. Adolfs and I.L. Bonta, 1985, Arachidonic acid, an in vivo inhibitor of carrageenin induced granulomas in the rat, European J. Pharmacol. 110, 367.

Feeman, B.A. and W.S. Lynn, 1980, Fatty acid secretion and metabolism in 'activated' rabbit alveolar macrophages, Biochim. Biophys. Acta 620, 528. Gilman, A.G., 1970, Protein binding assay for adenosine 3'5'cyclic monophopshate, Proc. Natl. Acad. Sci. U.S.A. 67, 305. Gurr, M.I., 1982, The role of lipids in the regulation of the immune system, in: Progress in Lipid Research, ed. Holman (Pergamon Press) p. 257. Horrobin, D.F., A. Campbell and C.G. McEwen, 1981, Treatment of the sicca syndrome and sjogrens syndrome with E.F.A., pyridoxine and vitamin C, in: Progress in Lipid Research, ed. Holman (Pergamon Press) p. 253. Kaduce, I.L,, A.A. Spector and R.S, Bar, 1982, Linoleic acid metabolism and prostaglandin production by cultured bovine pulmonary artery endothelial cells, Arteriosclerosis 2, 380. Kunkel, S i . , H. Ogawa, P.A. Ward and R.B. Zurier, 1981, Supression of chronic inflammation by evening primrose oil, in: Progress in Lipid Research, ed. Holman (Pergamon Press) p. 885. Lagarde, M., M. Guichardant and M. Dechavanne, 1981, Human platelet PGE~ and dihomogammalinolenic acid, Comparison to PGE 2 and arachidonic acid, in: Progress in Lipid Research, ed. Holman (Pergamon Press) p. 439, Mead, C.J. and J. Mertin. 1978, Fatty acids and immunity, in: Advances in Lipid Research, eds, Paoletti and Kritschevsky (Pergamon Press) p. 127. Mertin, J. and A. Stackpoole, 1981, Anti-PGE antibodies inhibit in vivo development of cell-mediated immunity, Nature 294, 456. Parnham, M.J., I.L. Bonta and M.J.P. Adolfs, 1979, Distinction between prostaglandin E 2 and prostacyclin as inhibitors of granulomatous inflammation, J. Pharm. Pharmacol. 31,565. Spector, A.A., T.L. Kaduce, J.C. Hoak and G.L. Fry, 1981, Utilisation of arachidonic and linoleic acids by cultured human endothelial cells, J. Clin. Invest. 68, 1003. Stuyvesant, V.J. and W.B. Jolley, 1976, Anti-inflammatory activity of a-tocopherol (vitamin E) and linoleic acid, Nature 216. 585. Willis, A.J., 1981, Unanswered questions in EFA and PG research, in: Progress in Lipid Research, ed. Holman (Pergamon Press) p. 839.