Leukotriene C promotes prostacyclin synthesis by human

Proc. Nati Acad. Sci. USA Vol. 80, pp. 4109-4113, July 1983

Immunology

Leukotriene C promotes prostacyclin synthesis by human endothelial cells (slow reacting substance/prostaglandins)

EVA B. CRAMER*, LAURA POLOGEt, NICHOLAs A. PANWLOWSKIt, ZANVIL A. COHNt, AND WILLIAM A. SCOTTt tThe Rockefeller University, New York, New York 10021; and *Department of Anatomy and Cell Biology, Downstate Medical Center, Brooklyn, New York 11203

Contributed by Zanvil A. Cohn, April 4, 1983

ABSTRACT Cultured endothelial cells from human umbilical vein were labeled with [3H]arachidonic acid for 16 hr. The radiolabel was localized primarily in phospholipids (93%) and 73% was distributed equally between phosphatidylcholine and phosphatidylethanolamine. Leukotriene C (10-1,000 nM) promoted a dose-dependent release of radiolabel into the culture medium. This response was 3.3 times control values at 100 nM. The major arachidonic acid metabolite synthesized was prostacyclin, which was 33% of the total released radiolabel. Endothelial cells also released small amounts of prostaglandin F2a (6.1%), unidentified lipoxygenase products (14.8%), and unreacted arachidonic acid (33%). The 30-min time course of release was independent of the leukotriene C concentration used. Leukotriene D at similar concentrations also promoted endothelial cells to release primarily prostacyclin and unreacted arachidonic acid. The release of prostacyclin, a potent vasodilatory agent, may be an important mediator in slow reacting substance effects on the vasculature.

Slow reacting substances (SRS) consist of the cysteine-containing leukotrienes (LT) C, D, and E and are derived from the lipoxygenase pathway of arachidonic acid (20:4) metabolism (1). Although the effects of LTC, LTD, and LTE show both quantitative and species differences (2, 3), they are strong smooth muscle stimulants causing bronchial (4), ileal, and uterine (5) contraction. In addition, LTC and LTD are vasoactive. Both compounds produced a transient constriction of arterioles that was followed by exudation from postcapillary venules in the hamster cheek pouch (6). Likewise in monkeys, LTC promoted a rapid, transient pulmonary and systemic hypertension followed by a more prolonged fall in blood pressure with decreased cardiac output (7). This may be due to vasodilation and increased vascular permeability (6, 8), a direct or secondary effect of SRS. To explore possible mechanisms for the vasoactive effect of SRS, human endothelial cells prelabeled with [3H]20:4 were exposed to LTC. The results of this study demonstrate that 10-8 to 10-6 M LTC promotes the release of prostacyclin, a potent vasodilatory agent. MATERIALS AND METHODS Materials. Radiolabeled prostaglandins (PG) were purchased from New England Nuclear and included 6-keto(120-180 Ci/mmol; 1 Ci = 3.7 [5,8,9,11,12,14,15-3H(N)]PGF1,, X 10'° Bq), [5,6,8,9,11,12,14,15-3H(N)]thromboxane B2 (100150 Ci/mnmol), [5,6,8,11,12,14,15_3H(N)]PGF2,,, (150-180 Ci/

mmol), and [5,6,8,11,12,14,15-3H(N)]PGE2 (100-200 Ci/mmol). Radiolabeled lipoxygenase products were prepared in this laboratory from various sources as described (9, 10). Nonradiolabeled LTC, used to promote 20:4 metabolism by

endothelial cells, was purified by reverse-phase high-performance liquid chromatography (HPLC) from culture medium of resident mouse peritoneal macrophages stimulated with unopsonized zymosan (10). LTD was prepared from LTC by treatment with y-glutamyltranspeptidase (11) and isolated by reverse-phase HPLC. Human thrombin was a gift of J. W. Fenton (Division of Laboratories and Research, Department of Health, Albany, NY). Isolation and Culture of Endothelial Cells. Endothelial cells were isolated by collagenase (type II; Worthington) treatment of human umbilical veins (12). The cells were collected by centrifugation, resuspended in medium 199 plus 20% heat-inactivated human serum, and plated in 35-mm tissue culture dishes. Studies were performed on first- or second-passage cells. All cells examined were positive for factor VIII as determined by indirect immunofluorescence staining (13). 20:4 Release and Synthesis of 20:4 Oxygenated Products. Endothelial cell cultures were washed with phosphate-buffered saline and overlaid with 1 ml of medium 199 containing 10% heat-inactivated human serum and 1.0 liCi of [5,6,8,9, 11,12,14,15-3H]20:4 ([3H]20:4) (91.2 Ci/mmol; New England Nuclear). At the end of the labeling period (16 hr), the cells were washed three times with phosphate-buffered saline, placed in medium 199 without serum, and incubated at 370C. Duplicate 50-1l aliquots of medium were removed at the indicated times for radioactivity measurements. At the end of the experiment, the medium was removed, and after washing with phosphate-buffered saline, the cells were scraped into 1.0 ml of 0.5% Triton X-100. Duplicate aliquots of the cell ysates were assayed for protein (14) and for 3H. Metabolites of 20:4 were extracted from the culture medium by the method of Unger et al. (15) and were separated by reverse-phase HPLC. System I was used to separate total 20:4 metabolites. A column of 5-,um Ultrasphere ODS, 4.6 mm X 25 cm (Altex, Rainin Instruments, Woburn, MA) was eluted at 1 ml/min with (i) 60 ml of methanol/water/acetic acid (65:34.9:0.01, vol/vol) adjusted to pH 5.4 with ammonium hvdroxide followed by (ii) 40 ml of methanol/water/acetic acid (75:25:0.01, vol/vol) and (iii) 40 ml of methanol/acetic acid (100:0.01, vol/vol). One-milliliter fractions were collected. The radiolabel contents of whole fractions or aliquots of fractions were determined by liquid- scintillation counting in Hydrofluor (National Diagnoistics, Somerville, NJ). Corrections were made for counting efficiency. The cyclooxygenase products in fractions 4-16 were pooled and the solvent was evaporated under Abbreviations: 20:4, arachidonic acid; LTC, leukotriene C, (5S)-hydroxy-(6R)-S-glutathionyl-7,9-trans-11,14-cis-icosatetraenoic acid; LTD, leukotriene D, (5S)-hydroxy-(6R)-cysteinylglycyl-7,9-trans-11, 14-cisicosatetraenoic acid; LTE, leukotriene E, (5S)-hydroxy-(6R)-S-cysteinyl-7,9-trans-11, 14-cis-icosatetraenoic acid; PGE2, prostaglandin E2; PGF2,, prostaglandin F2a; 6-keto-PGFia, 6-ketoprostaglandin Filt; SRS, slow reacting substance.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 4109

4110

Proc. Natl. Acad. Sci. USA 80 (1983)

Immunology: Cramer et al.

nitrogen. Separation of individual cyclooxygenase products was accomplished by using a 3.6 mm X 30 cm Waters fatty acid analysis column eluted at 2 ml/min with 50 ml of water/acetonitrile/benzene/acetic acid (76.7:23.0:0.2:0.1, vol/vol) followed by 20 ml of methanol/acetic acid (100:0.01, vol/vol) (16) (system II). As described above, fractions were collected at 1min intervals, and the radiolabel contents of fractions were determined. The elution times of prostaglandins, leukotrienes, monohydroxyicosatetraenoic acids (mono-HETEs), and 20:4 were determined by using radiolabeled standards. Lipid Extraction, Separation, and Analysis. For fatty acid analyses of endothelial cell lipids, cultures maintained in serumcontaining media were washed thoroughly with phosphatebuffered saline and scraped into 0.9% saline. The lipids were extracted by the method of Bligh and Dyer (17). Concentrated lipid extracts were applied to columns containing 0.4 g of activated silicic acid (Silicar CC-4; Mallinckrodt). Neutral lipids and phospholipids were sequentially eluted with 10 ml of chloroform and then 10 ml of methanol (18). Fatty acid methyl esters were prepared by transesterification in methanolic HCl and analyzed by gas/liquid chromatography on 1/8 inch x 10 foot (3 mm X 305 cm) columns of 10% SP-2330 on 100/200 Chromosorb WA/W (Supelco, Bellefonte, PA) at 180'C with a carrier gas flow of 30 ml/min (18). To determine the localization of [3H]20:4 in endothelial cell lipids, lipid extracts of radiolabeled cultures were subjected to one-dimensional thin-layer chromatography on plates of Redi Coat G (Supelco) prepared as described (19) and developed in diethyl ether. The distribution of 3H among individual phospholipids was determined by two-dimensional thin-layer chromatography (20). In both systems, unlabeled lipids (0.1 ,umol) extracted from J774 cells (a macrophage-like cell line) or serum lipids (Supelco) were added to samples to facilitate visualization of lipids on the plates. The lipid-containing areas were detected by exposing the plate to iodine vapor, and the silica gel in these regions was scraped from the plates into scintillation vials for

radioactivity determinations. Bioassay of LTC. The SRS activity of LTC was measured by using an isolated guinea pig ileum as described by Chakravarty (21).

RESULTS Localization of [3H]20:4. Fatty acid analyses indicated that the phospholipids of human endothelial cells contain 22 mol % 20:4, a level comparable to that previously reported (22) for these cells and similar to the 20:4 content of macrophages (23), which are potent sources of 20:4 metabolites. The distribution of the [3H]20:4 in endothelial cell lipids was measured. Endothelial cell cultures took up 40-65% of the radiolabeled fatty acid supplied in the culture medium. Greater than 99% of the cell-associated 20:4 was incorporated into cell lipids and less than 1% was recovered as free fatty acid (Table 1). Of the [3H]20:4 esterified in cell lipids, 93% was recovered in phospholipid and 5% in neutral lipid. Approximately 73% of the radiolabeled fatty acid in cell phospholipids was equally distributed between phosphatidylcholine and phosphatidylethanolamine (Table 1). Significant amounts of label were also recovered in phosphatidylinositol (12.4%) and phosphatidylserine (7.1%). 20:4 Release. Fig. 1 shows the time course of radiolabel release by endothelial cells exposed to various concentrations of LTC. The radiolabel content of the medium increased for 30 min and leveled off thereafter. LTC at 10-9 M caused little or no release of radiolabel above control levels. Maximal release occurred at 10-8 to 10-6 M LTC. Although the quantity of ra-

Table 1. Distribution of [3H]20:4 in endothelial cell lipids % of radiolabel Lipid 7.1 Phosphatidylserine 12.4 Phosphatidylinositol 0.2 Sphingomyelin 36.5 Phosphatidylcholine 36.0 Phosphatidylethanolamine 1.1 Cardiolipin 5.0 Neutral lipid 0.7 Free fatty acid 1.0 Other

Cultures were labeled overnight with 1.0 ,uCi of [3H]20:4 and washed with phosphate-buffered saline. The cells were scraped into isotonic saline and the lipids were extracted. The lipid extracts were subjected to two-dimensional thin-layer chromatography. Areas of the chromatogram containing lipid were scraped, and the radioactivity of each was determined. Data are the average values of two determinations and are expressed as the percent of recovered radiolabel.

diolabel released was dependent on the concentration of LTC, the time course of release was independent of the stimulus level. At 10-7 M LTC, 4.3 ± 1.2% of the total cell radiolabel was recovered in the culture medium, whereas 1.3 ± 0.8% was recovered from control cells incubated in medium 199 alone. LTD also promoted radiolabel release by endothelial cells prelabeled- with [3H]20:4; LTD was effective at concentrations similar to those for LTC and the kinetics were also similar. Release of radiolabel induced by thrombin (1 unit/ml), a protease known to stimulate 20:4 release and metabolism by endothelial cells, was 1.2 to 2.9 times greater than with LTC (10-7 to 10-6 M) (Fig. 2). In addition, the kinetics of the thrombin-stimulated response were more rapid than with LTC-stimulated endothelial cells. Maximal levels were reached in approximately 10 min. Identification of 20:4 Metabolites. The 20:4 metabolites released by endothelial cells in response to a LTC challenge were determined by using HPLC system I. Two major radiolabeled peaks together with a number of smaller peaks were obtained (six experiments; results are presented as mean ± SEM). A representative experiment is shown in Fig. 3A. The initial peak

1046

7

~~~~~~~~10-8

6

6

-1

4

C

2

0

1

10

20

I

Control

30

40

50

60

Minutes FIG. 1. Time course of 20:4 release by endothelial cells as a function of LTC concentration (M). Control cultures were incubated in medium 199 without additions. Data are reported as duplicate determinations of the 3H content of media from duplicate cultures (mean +

range).

Immunology:

Cramer et al.


3A). It could be estimated from six experiments that 67.7 + 11.4% of the [3H]20:4 released in response to a LTC stimulus was converted to oxygenated metabolites. In control cultures, 49.5 19. 1% of the released [3H]20:4 was metabolized. Individual cyclooxygenase metabolites were separated by HPLC system II (Fig. 3B). In five experiments, the predominant compound was 6-keto-PGFia (62.2 7.6%), the breakdown product of prostacyclin. A smaller peak of PGF2,. contained 11.6 4.6% of the radiolabeled cyclooxygenase products. Occasionally a small peak of PGE2 (2-4% of the radiolabel) was seen in some preparations. Therefore 33% of the total released radiolabel was converted to prostacyclin. This constituted the major 20:4 metabolite synthesized by endothelial cells stimu-

7

I

6

Thrombin

4111

0

±

5i -

1j

T

i

tLTC

a)

±

U,

0

Q)

4

0~~~~~~~~~~~

±

0)

i.)

3

a)

a_

2

lated with LTC. Thrombin (Fig. 4A) also caused the release of cyclooxvgenase (51 8.7%) and lipoxygenase (16.7 7.0%) products in addition to free 20:4 (32 5.3%). As seen in this study (Fig. 4B) and observed by others (24), prostacyclin was the major 20:4 metabolite produced by endothelial cells exposed to thrombin. The notable difference from cultures exposed to LTC was that a higher percentage of the released 20:4 was converted to PGE2 and PGF2.G. In three experiments, one of which is shown in Fig. 4B, 58.7 14.7% of the cyclooxygenase products were 6-ketoPGF1a, 20.3 2.9% were PGF2G, and 12.0 1.7% were PGE2. Effect of FPL 55712. The drug FPL 55712 (Fisons, Leicestershire, England) blocks LTC- and LTD-mediated contraction of smooth muscle (5, 25). It was of interest to determine whether this compound has similar antagonistic effects on prostacyclin synthesis by LTC-treated endothelial cells. Endothelial cells were incubated for 20 min in medium containing FPL 55712 at 0.5, 1, 5, or 10 ,ug/ml and then challenged with 50 nM LTC

Control

±

I

IY 0

10

20

30

40

50

60

Minutes

FIG. 2. Time course of [3H]20:4 release in response to 100 nM LTC at 1 unit/ml.

or thrombin

±

±

±

(4-16 min) represented 52.8 9.3% of the recovered radiolabel and contained primarily cyclooxygenase products. The second major peak (111-150 min) with an average of 32.7 11.1% of the radiolabel was unreacted 20:4. Lipoxygenase products (17-110 min) made up a smaller proportion (14.8 + 4.6%) of the radiolabeled material. The basal release from control cultures (no stimulus) produced a similar HPLC profile (Fig. ±

±

A

18-

B

16-

I

LTC

14-

LTC

1210&0

x

6-

E 0.

'a

4-

n U.ll I v

I I

I

I

I

--,

+ I t 6 keto

-1 A 1

1

PGF.

PGF

Control

I

Pg

Control

0

20

4k

d

do

lo

iio 4o Minutes

--Ii6o

. o

20

I

40

60

8o

FIG. 3. HPLC chromatograms of 3H-labeled products released by endothelial cells in to 60-min incubation in the of LTC in medium 199 (no stimulus). (A) Total 20:4 metabolites released by cells. Media extracted and separated Ultrasphere ODS C18 verse-phase column eluted with system I. Elution times: cyclooxygenase products, 4-16 min; lipoxygenase products, 17-110 min; and 20:4, 111150 min. (B) Cyclooxygenase products eluting at 4-16 min from the chromatograms in A rechromatographed in system II. Elution times: 6keto-PGF1,, 11 min; thromboxane, 24 min; PGF2a, 29 min; and PGE2, 34 min. The peaks eluted with methanol (55-80 min) represent unidentified radiolabeled material. response

or

were

a

presence

on an

were

re-

Immunology: Cramer et al.

4112

Proc. NatL Acad. Sci. USA 80 (1983)

181614-

12-

q

A

I

*T

I

a)

l

LTC

T

I

() 0

Thrombin

a1)

a1)

10-

c

8-

0Q

a)

.

4-

x

0

0

2x

E

-o0.

B

12108-

PGF20 4-

4,APvGE2

2-

20

40

60

80

100

120

_-

-

T

-0

iO

Control ~~~~ --

o Iot

10

20

30

40 50 Minutes

60

70

80

FIG. 5. Effect of a 20-min preincubation with FPL 55712 (FPL) at 10 jig/ml on 3H released by endothelial cells labeled with [3H]20:4 and exposed (arrow) to 50 nM LTC. Control denotes no additions.

0

14- 6-keto PGF,

~

T//

Ilo FP

6-

140

Minutes FIG. 4. HPLC chromatograms of 3H-labeled products released by endothelial cells in response to a stimulus of thrombin at 1 unit/ml for 10 min. (A) Total 20:4 metabolites. Elution times are those given in Fig. 3. (B) Cyclooxygenase products eluting at 4-16 min from HPLC system I (A) were rechromatographed in system Elution times: 6-keto-PGF10, 10 min; thromboxane, 18 min; PGF2a, 25 min; and PGE2, 32 min. fl.

in the presence of the drug. As shown for FPL 55712 at 10 ,tg/ ml (Fig. 5), all concentrations of the drug failed to block LTCmediated prostacyclin synthesis. In separate experiments, the same concentrations of drug inhibited the contractile effects of

2 nM LTC on guinea pig ileum. However, with higher doses of LTC (50 nM) comparable to those used to stimulate endothelial cells, an FPL 55712 concentration of 10 Ag/ml was needed to block ileal contraction. After washing, the ileum was responsive to LTC (2 nM), indicating that the effects of drug and high LTC levels were reversible. DISCUSSION LTC and LTD initiate 20:4 release and the synthesis of the vasodilatory agent prostacyclin by cultured human endothelial cells. Concentrations of 10- to 10-6 M LTC were required to stimulate 20:4 release and metabolism above basal levels. At 10' M LTC, this corresponded to a 4.3 ± 1.2% release of the total cell [3H]20:4 and compared favorably with the release promoted by calcium ionophore A23187 (26). Unlike thrombin (Fig.

2; ref. 24), histamine (27), and bradykinin (28), which produce maximal levels of prostacyclin release within 10 min, LTC in-

duced prostacyclin synthesis by endothelial cells over approximately a 30-min period. The radiolabel assay used to monitor 20:4 release and metabolism indicated that prostacyclin, unreacted 20:4, and PGF1,t accounted for approximately 33%, 33%, and 6%, respectively, of the total radiolabel released into the culture medium by cells prelabeled with [3H]20:4 and exposed to LTC. Occasionally PGE2 (2-4%) was observed in some experiments. These products were similar to those released by thrombin-treated cells. However, thrombin promoted the synthesis of higher amounts of PGF2a, and PGE2. Although this assay does not provide a quantitative measure of prostacyclin synthesis, we can say that the amounts produced in response to a maximal LTC stimulus ranged from 0.3 to 0.8 of the amount promoted by thrombin at 1 unit/ml. A number of structurally unrelated vasoactive compounds are known to stimulate prostacyclin synthesis by endothelial cells. These include angiotensin 11 (29), bradykinin (28), and histamine (27). It would appear that the vasodilatory effects of prostacyclin modulate the actions of these vasoactive substances either by counteracting their vasoconstrictor actions or by amplifying or mediating their vasodilatory actions. From the results of this study, LTC and LTD can now be included in this category of mediators. SRS (6), histamine (6), and bradykinin (28) are also known to promote transudation of plasma from the microvasculature. However, prostacyclin produces little plasma transudation (30). It has been proposed that a second mediator or mechanism regulates vessel wall permeability and that it is a combination of a vasodilatory agent such as prostacyclin together with a permeability-increasing factor that influences the amount of plasma exudate (30). While SRS, histamine, and bradykinin may all promote vasodilation through a common mediator, it is unknown if a common mechanism also controls vascular permeability. SRS has also been shown to have spasmogenic effects. LTCand LTD-mediated contraction of smooth muscle can both be inhibited by FPL 55712 (5, 25). In contrast, FPL 55712 does not block prostacyclin synthesis by endothelial cells challenged with LTC. Although the mechanism of action of this drug is unknown, these differences may provide a distinction between the action of LTC on endothelial cells and its effects on the contraction of smooth muscle. Leukotrienes play roles in the complex sequelae of immediate hypersensitivity and other acute inflammatory reactions

Immunology:

Cramer et al.

(31). It seems possible, therefore, that in response to inflammatory stimuli their generation by macrophages (10, 32) and monocytes (unpublished data), which are either in or in close approximation to the vasculature, would lead to the synthesis of prostacyclin. In addition, granulocytes (33, 34) and mast cells (35) are sources of similar compounds. Taking these findings together, one can speculate on a metabolite cascade in which cells of the immune system influence the vasculature via products of the cyclooxygenase (23) and lipoxygenase pathways. The authors thank E. Abraham, M. Andreach, and A. Hamill for their excellent technical assistance. This work was supported in part by National Institutes of Health Grants HL-27186, AI-16480, and AI-07012. 1. Samuelsson, B. (1982) in Leukotrienes and Other Lipoxygenase Products, eds. Samuelsson, B. & Paoletti, R. (Raven, New York), pp. 1-17. 2. Peck, M. J., Piper, P. J. & Williams, T. J. (1981) Prostaglandins 21, 315-321. 3. Ueno, A., Janaka, K., Katori, M., Hayashi, M. & Arai, Y. (1981) Prostaglandins 21, 637-648. 4. Dahlen, S.-E., Hedqvist, P., Hammarstr6m, S. & Samuelsson, B. (1980) Nature (London) 288, 484-486. 5. Hedqvist, P., Dahlen, S.-E., Gustafsson, L., Hammarstrom, S. & Samuelsson, B. (1980) Acta Physiol. Scand. 10, 331-333. 6. Dahlen, S.-E., Bjork, J., Hedqvist, P., Arfors, K.-E., Hammarstr6m, S., Lindgren, J.-A. & Samuelsson, B. (1981) Proc. Natt Acad. Sci. USA 78, 3887-3891. 7. Smedegaard, G., Hedqvist, P., Dahlen, S.-E., Revenas, B., Hammarstrom, S. & Samuelsson, B. (1982) Nature (London) 295, 327-329. 8. Lewis, R. A. & Austen, K. F. (1981) Nature (London) 293, 103108. 9. Pawlowski, N. A., Scott, W. A., Andreach, M. & Cohn, Z. A. (1982) J. Exp. Med. 155, 1653-1664. 10. Rouzer, C. A., Scott, W. A., Hamill, A. L. & Cohn, Z. A. (1982) J. Exp. Med. 155, 720-733. 11. Sok, D.-E., Pai, J.-K., Atrachi, V. & Sih, C. J. (1980) Proc. Nati. Acad. Sci. USA 77, 6481-6485. 12. Jaffe, E. A., Nachman, R. L., Becker, C. G. & Minick, C. R. (1973) J. Clin. Invest. 52, 2745-2756. 13. Jaffe, E. A., Hoyer, L. W. & Nachman, R. L. (1973)J. Clin. Invest. 52, 2757-2764.


4113

14. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 15. Unger, W. G., Stanford, I. F. & Bennett, A. (1981) Nature (London) 233, 336-337. 16. Alam, I., Ohuchi, K. & Levine, L. (1979) Anal. Biochem. 93, 339345. 17. Bligh, E. G. & Dyer, W. J. (1959) Can. J. Biochem. Physiol. 37, 911-917. 18. Mahoney, E. M., Hamill, A. L., Scott, W. A. & Cohn, Z. A. (1977) Proc. Natl. Acad. Sci. USA 74, 4895-4899. 19. Keith, A. D., Waggoner, A. S. & Griffith, 0. H. (1968) Proc. Nati Acad. Sci. USA 61, 819-826. 20. Rouser, G., Fleischer, S. & Yamamoto, A. (1970) Lipids 5, 494496. 21. Chakravarty, N. (1959) Acta Physiol Scand. 46, 298-313. 22. Rastogi, B. K. & Nordoy, A. (1980) Thromb. Res. 18, 629-641. 23. Scott, W. A., Zrike, J. M., Hamill, A. L., Kempe, J. & Cohn, Z. A. (1980) J. Exp. Med. 152, 324-335. 24. Weksler, B. B., Ley, C. W. & Jaffe, E. A. (1978) J. Clin. Invest. 62, 923-930. 25. Piper, P. J., Samhoun, M. N., Tippens, J. R., Morris, H. R., Jones, C. M. & Taylor, G. W. (1981) nt. Arch. Allergy Appl. Immunol 66, Suppl. 1, 107-112. 26. Thomas, J. M. F., Chap, H. & Douste-Blazy, L. (1981) Biochem. Biophys. Res. Commun. 103, 819-824. 27. Baenziger, N. L., Fogerty, F. J., Mertz, L. F. & Chernuta, L. F. (1981) Cell 24, 915-923. 28. Hong, S. L. & Deykin, D. (1982)J. Biol. Chem. 257, 7151-7154. 29. Swies, J., Radomski, M. & Gryglewski, R. J. (1979) Pharmacol. Res. Commun. 11, 649-655. 30. Williams, T. J. (1979) Br. J. Pharamacol. 65, 517-524. 31. Samuelsson, B. (1981) lnt. Arch. Allergy Appi Immunol 66, Suppl. 1, 98-106. 32. Rouzer, C. A., Scott, W. A., Hamill, A. L. & Cohn, Z. A. (1980) J. Exp. Med. 152, 1236-1247. 33. Borgeat, P. & Samuelsson, B. (1979) Proc. Natl Acad. Sci. USA 76, 2148-2152. 34. Jorg, A., Henderson, W. R., Murphy, R. C. & Klebanoff, S. J. (1982)J. Exp. Med. 155, 390-402. 35. MacGlashan, D. W., Jr., Schleimer, R. P., Peters, S. P., Schulman, E. S., Adams, G. K., III, Newball, H. H. & Lichenstein, L. M. (1982) J. Clin. Invest. 70, 747-751.