205

Biochimica et Biophysica Attn. 1 I28 ( 1 9 9 2 ) 205-21t} © 1992 Elsevier Science Publishers B.V. ALl rights reserved t~)05-2760/92/505.00

BBALIP 54017

Hypoxia increases stimulus-induced PAF production and release from human umbilical vein endothelial cells M i c h a e l S. C a o l a n % L u b a A d l e r ", A n n e Kelly ~' a n d W e t H s u e h h a Department of Pediatrics, Evan.~ton I-!osvital. Evanrton. !1. (USA) and ~ Department of Pathol,:~.'. Children ~ Memori,,t! Ho,,pital, Northwestern Universi~" Medical School. CT~igago. IL tUSA)

(Recetved 2(1 February 1992)

(Revised manuscript received 21 May 1992)

Key words: PAF; Hypoxia: Endothelial cell: Pho.~pholipaseA,: Acetylhydrolasc Hypoxia alters endothelial cell function and metabolism. S;nce PAF is syitthcsize,! ~y ,--aothc!ial colts a.nd c.:~p:,bleel modulating endothelial cell responses, we investigated the effect ~f hypoxia on synthesis and release of PAF t'row, endothelial cells. We found: (1) Approx. 90% of the radylPAF derivative in stimulated endothelial cells is acylPAF. (2) Acute hylmxie (15 min-I h) priming increased ionophore- and thrombin-induced radylPAF accumulation. (3) Long-term h.vgoxic exrmsure increased radylPAF accumulation at 24 and 48 h in the presence of ionophore. ~4) Bioactive PAF was released into media and hypoxia :.,nd ionophore synergistically increased PAF release. (5) Hypoxia and ionophore stimulation increased phospholipase A: activity and decreased acetylhydrolase activity in end~theiial cells. We conclude that hypoxia and ionophore increase PAF synthesis and release from endothelial cells.

Introduction

PAF (platelet-activating factor, PAF-acether) is a potept pho~pholipid mediator wit~ diverse biological effects. Exogenous administration of PAF into animals produces systemic hypotension, capillary leakage, pulmonary hypertension, bronchocons~riction, thrombocytopenia, neutropenia and intestinal injury [1-3]. We have previously shown that hypoxia increa?,es circulating plasma PAF concentrations in experimental animals [4]. Furthermore. it has been shown that hypoxia increases PAF in bronchoalveolar Image [5], perhaps by stirrulating phospholipase A : [6]. Although in rive data suggest an association between hypoxia and elevated PAF levels, the effect of h~ooxia on PAF metabolisnt in vitro is unclear. Endothelial cells are important in the regulation of vas"ular tone, angiogenesis, thrombogenesis, inflammation and the immune zesponse [7]. Hypoxia alters membrane permeability of endothelial cells [8], induces changes in vescmar tone and may modulate other

Correspondence to: M,S. Caplaa. Department ot Fediatrics, Evanston l-iospilal, .2650 Ridge Av~, Evanston. IL (~0201, USA. Abbreviations: PAF. platclt.[-a~.dvatinr, factor; alk-~IPAF, l-alkyl-2acetyl-sa-glycero-3-phosphoeholine; acylPAF, l-aeyl-2-acetyl-seglyccro-3-phosphocholinc; HUVEC, h u m a n umbilical velr,,-,Oothelial cells; A23187. calcium ionophore.

endothelial cei; f, mctions. Human umbilical win cndotllelial cell (HUVEC) monolayers have been well studied and are known to syntitc.~;ze PAF in response to several agonists including thrombim bi-adykinin, histamine, ATP and leukotrienes [9-11]. In addition~ recent reports have shown that the predomina~'_ derivative of rady!PAF in stimulatet~ endothelial cells is acylPAF (l-acyl-2-acetyl-sn-glycero-3-phosphocholine), a compound with 10t~fold less biologic activity than ... , i,,~alky,-,.-acetyi-sn-glycero-3-phosphochol alkyipA ~ line) t,,--,','.r"~, ,~ r.,,u,,h,,rm,,r,.," - ~ - ~.'may ,~tud:cs suggest that all newly synthesized radylPAF remains cell-associated [9-11]. In this report, we document that hypoxia primes agonist-induced radylPAF synthesis in endothelial cells; that the predominant derivative of hypoxia primed radylPAF accumulation is acyiPAF; and that HUVEC's release bioactive PAF into the surrounding medk:. "

~,'b

Materials and Methods HUI.'EC i.~olatitm

Endothelial ceils were harvested from fresh (24-48 h) full-tr.t,n human umbilical wins, as previously described [15]. Vein~ were washed with Hanks' balanced saI~. solution (HBSS) and endothelial cells separated by collagenase incubation for 15 rain at 37°C. The effluent was centrifuged and the cell pellet was resuspended in compLic medium M-199 witla 20% rcml bovlnc st;rum,

206 L-glutamine (2 raM). penicillin (t00 units/m!) and strcptomycin (100 ~g/mI). Cells were incubated in 37~C, 5% CO_,, 21% O, and batar.ced nitrogen until mon~Iaycrs reached c-nflu~nce. The presence of endothcfial ceIIs was confirmed by morphology, using phase contrast microscopy and the demonstration of factor VIII antigen by indirect immunofluorescence. Ceil viability was tested using the Trypan blue exclusion technique.

Experimental protocol To study the acute effect of hypoxia and agonist stimulation on radyIPAF synthesis, HUVEC monolayers were either exposed to hypoxia (1% O~, 5% CO z) or normoxia (21% O_~, 5% CO.,) for various time intervals, with or without caIeium ionophore (A23187, Sigma, St. Louis, MO) or thrombin (l U / m i . Sigma). Two doses of A23187 were used: l p.M throughout the incubation or 10 tzM added at the end of incubation (to avoid cytotoxicity). For investigation of long-term effect of hypoxia, 10 p.M A23187 (added at the end) was used. RadylPAF synthesis was determined at different time intervals using [3H]acetate incorporation method [15]. Lastl~. HUVEC's were studied tor the ability to ~ecrete PAF into media. Cells were incubated for 30 min or 2 h in hypoxic or normoxic environment with or v, ithout 1 /.tM A23187. Media was collected, phospholipids extracted and assayed for PAF activity using the rabbit platelet serotonin release method [16].

P,4F quantitation using ['~tt]acetate incorporation method [ 15] Following hypoxia or normoxia, media was removed from monolayers and HUVEC's were washed with buffer A (HEPES 2.38 g/I, glucose 1 g/I, NaCI 8 g/I, KCI 0.4 g/i, Na~HPO4-7H.,O 0.09 g/I, KH.,PO4 0.116 g/I) aqd incubated 10 rain at room temperature in i mi buffer A, containing calcium (2 raM) and 25 p-Ci ['~H]acetate. In some experiments, the incubation included 10 p-M A23187. The reaction was stopped with acidified methanol, monolayers were scraped, and PAF was extracted using chIoroform/methanol and separated by thin-layer chromatgraphy (silica-gel G, Analtech, Ncwark, Delaware). Zones comigrating with PAF standard were quantified using scintillation spec.troscopy (Beckman, Fullerton, CA).

Separation of nulyiPAF species into l-acyl or l-aikyl derit'atit'es RadylPAF, recovered from the [~H]acetate incorporation method, was extracted [17], evaporated and reconstituted in HCI-Tris buffer (100 mM (pH 7.4)) as prcviousiy described [12,13]. The sample and PAF standards were then incubated in ethyl ether containing phospholipasc C (50 units/sample), lipids extracted using Bligh and Dyer method [18] and organic phase

recovered. These derivatives were acetylated with acetic anhydride (0.5 ml) and pyridine (0.I mi) for 16 h at 37°C, and the organic phase recovered and redeveloped or. a TLC system containing hexane/ethyi ether/formic acid (90:60:6). Plates were redeveloped twice to increase separation between the 1-alkyl and 1-acyl derivatives, and scraped in 0.25 or 0.5 cm blocks. The portions comigrating with alkylPAF standards subjected to phospholipasc C digestion and reacetylation, were counted and considered alkylPAF; the large peak migrating behind this fraction was !-acyl-2-acetyl-GPC, as described previously [12,13].

PAF bioassay: rabbit platelet serotonhz release method i161 At the end of exposure, HUVEC's were separated from media, total lipids from media were extracted by Folch's method [17], and PAF was separated by thinlayer chromatography using chloroform/methanol/ water (65:35:6, v/v). For some samples, PAF was further separated using high-performance liquid chromatography. Varian HPLC was equipped with an UItrasphere-Si, 5 p-m column. The solvent starts with 96% isopopanol/hexane (1:1) and 4% water with a linear gradient to 8% water over a 15 min period. The flow rate was 2 ml/min [19]. Fractions recovered corresponding to standard PAF were collected, dried under nitrogen and resuspended in PBS/albumin fer bioassay. Piatelet-rich plasma was recovered from rabbit blood (New Zealand white rabbit, Charles River, NY) and labelled with [3H]serotonin. A standard PAF (l-alkyl-2-acetyi-sn-glycero-3-phosphocholine, Sigma) curve, ranging from 100 pg/tube to 1 rig/tube, was compared to amounts of [3Hlserotonm released from each sampie in order to calculate PAF concentrations. PAF bioaetivity was confirmed by comparing samples to incubations pretreated with PAF receptor antagonist SRI 63-441 (2 mg/ml, Sandoz Research Institute, Hanover, New Jersey) or by sample treatment with porcine pancreas phospholipase A 2 prior to assay incubation. [~H]PAF was added to each sample to calculate PAF recovery.

Phospholipase A 2 assay To investigate the importance of PAF synthesis in this model, phospholipase A , activity of HUVEC monolayers was measured. The radiolabelled PAF specific substrate l-alkyl-2-[~H]arachidonyI-GPC (0.01 # C i / t u b e , 0.013 p-M/mCi) was used in an incubation assay with ltCI-Tris buffer (100 mM (pH 7.2)), BSA (5 mg/ml), deoxycholate (0.1 mg/ml), CaC! 2 (2 mM), cold l-alkyl-2-arachidonoyl-GPC (10 p.l, 30 mM) and sample (50 p-I sample). The substrate was first sonicated vigorously for 5 min and then sample added and incubated for 30 min at 37°C. The reaction was stopped with chloroform/methanol (2:1, 8 vol.), and the or-

207 ganic phase developed on TLC system containing chloroform/methanol/water (65:35:6). Areas comigrating with phosphatidylchotine and arachidon,c acid standards were scraped and quantified using scintillation spectroscopy. Phospholipase A ~ activity was expressed as pmol arachidonic acid formed/mg protein/rain (protein measured using,, spectrophotometric a~:~y (Bio-Rad, Richmonc~, CA).

Acetylhydrolase assay

Statistical analysis All data are expressed as the mean _+ S.E. and significance was assigned where P was less than 0.05. Differences between group means were tested using analysis of variance, and individual group differences were compared by the Bonferroni method. Unless specified, each data point represents the mean of four experiments. Results

Using the [3H]acetate incorporation method (Fig. 1 and 2), we found that almost all radylPAF formation was cell-associated. Low dose of A23187 (l p.M) stimulated radyiPAF accumulation (P < 0.05, compared with

~,,

"E

.,0-

~. 40./ .'

,t.

/

30-

207

.

" .

~ .... ~

"

• ..Q.

\ ~. < j. ~ " .l ~

.

.

1

o.1

i

~ .~

. - L"

IO

1oo

TIME ( h )

Fig. 2. Effect of hypoxia en endothelial cell radylPAF pr~;duction ovt:r time. RadylPAF synthesis in cell monolayers was quantifiecl usin.~: the [3}l]acetate incorporation method with lO /.tM A23187 give,; at end of exposure for 10 rain, Open squares with dashed line repre,:en~ normoxmc monolayers: crossed squares with solid line rep.~-~sen, hypoxic incubations. All points represent the mean of four or five e~perimcnts. * P < 005 and * * P < 0,01 compared to normoxia.

controls, Fig. 1A). This ionophore-induced respon~ was enhanced by hypoxia (Fig. 1A, 15 and 30 rain, P < 0.05, compared to A23187 alone), whereas hypoxia alone had litde effect. Fig. IB shows that higher dose A23187 (10/zM, added at the end of the incubation) increased radyIPAF accumulation. Although hypoxia alone did not alter radylPAF production, at 15 and 30 min hypoxia markedly enhanced A23187-induced radyiPAF synthesis ( P < 0 . 0 5 compared to A23187 alone). In the presence of A23187, we found a biphasie increase in radylPAF accumulation in response to hypoxia (Fig. 2). Hypoxia enhanced cell-associated radylPAF accumulation acutely at 15 and 30 min, and this response subsided by I h. However, at 6 h of hypoxic exposure, there was a steady increase in radylPAF synthesis which persisted at 48 h. There were no differ-

• HY'POXIA

o NOI~t~I~XIA

m/~

I

U

NORMOXIA

20000

~" l ~ ~_

....

;:

ii.-HYPOXIA

II

~-, I0000

cam/million calls} ...........................

~oi

To assess the effect of hypoxia on HUVEC PAF degrading ability, acetylhydrolase activity was measured, as previously described [20]. Briefly, samp!e (100 #1) was incubated with [3H]alkylPA'. and HEPES buffer, and products recovered, sepa-ated on TLC, and quantified using scintillation spec,roscopy. To ensure that PAF degradation was due to acetylhydrolase, in similar experiments the eft ec, of diisopropyl flourophosphate (DFP, selective aretylhydrolase inhibitor, 10 mM, Sigma, St. Louis, Me L~was studied. Acetylhydrolase activity was estimate,, as nmol lysoPAF formed/rag protein per rain.

A

(x 1 0 0 0

PAF 6,

•~

• HYI~OXIA

\

oNoNXtA.,a,~,sr

~

tl I~fP0XIA÷ ild¢/b?.3187

15000

~.

+ 10$14 A 2 3 1 8 7

"

•

Q.

~

10000

v

5000

~

.

•~

.ITY?OXIA + iOpN A23107

5000

......

o .g~e'" 0 30

~

0 ~ 0

L '7---:-,

90

60 TIME

(mn)

120

,

3O

,

....

t

60 lIME (h)

Fig. [. (A) Effect of hypoxia and A23187 ( 1 ~.M) on endothelial cell radylPAF accumulation over time. (B) Effect of hypoxia and A23187 (10 # M. given at end of incubation for 10 min) on endothelial cell radylPAF formation over time. RadyIPAF synthesis was mea,.ured in the cell monolayers using the ['~HJacetate incorporation method. Data represent tile mean :tS.E. * identifies P < (l.05 compared to normoxia controls. • * represents P < 0.05 compared to A2.3187 alone.

208 TABLE I

TABLE II

Characterization of radylPAt" .wech's bt endothelial cells erposed to hypoxia and ionophore at 2 h

Bioactirc PAI-' com'entrotions hz media fi'om emlothelial cells exposed to h vpoxia attd ionophore

Phospholipids were extracted and separated by TLC. and the radylPAF fraction furlher separaled by phospholipase C digestion, acetytation, and TLC as described in M~,terials and Methods. "the data represent the meat, _+S.E. of four experiments.

Phospholipids were extracted, separated by TLC and HPLC. and PAF quantified using the rabbit platelet serotonin release bioassay as described in Materials and Methods. The data represent the mean 4- S.E. for five different experiments.

Control alkyiPAF (cpm) acylPAF (cpm) acyl / a l kyl ratio

llypoxia

Ionophore

30+5

30+ 6

60+t(I

75 -+5

Nt_+ 14

52(I_+45*

-(at

-(a)

8.6:1

tlyp + h)n 9

85+

89.5+_ 135 ** 10.5: !

(a): Ratio unobtainable becatJ.,,e baseline control cpm = 30. * represcntx P < (1.05 compared to control and hypoxia. ** represents P < 0.05 compared (o ionophore alone.

e n c e s in cell viability s t u d i e s b e t w e e n n o r m o x i c a n d hypoxic H U V E C ' s ( > 9 0 % ) until 48 h o f h y p o x i a w h e n cells d e m o n s t r a t e d d e c r e a s e d viability ( 7 5 - 8 5 % ) . Similar to t h e a c u t e r e s p o n s e , c h r o n i c h y p o x i a o r n o r m o x i a w i t h o u t A 2 3 1 8 7 did not s i g n i f i c a n t l y i n c r e a s e r a d y l P A F a c c u m u l a t i o n ( d a t a n~,t s h o w n ) . T h e 1-acyl d e r i v a t i v e w a s t h e m a j o r f o r m o f radyiP A F i d e n t i f i e d in H U V E C ' s e x p o s e d to h y p o x i a a n d i o n o p h o r c ; t h e a c y l / a l k y l r a t i o w a s a p p r o x . 9 - 1 0 : 1 in s t i m u l a t e d cells ( T a b l e !, Fig. 3). A l t h o u g h a c y l P A F was the p r e d o m i n a n t derivative, hypoxia and i o n o p h o r e synergistically increased both alky!PAF and acylPAF a c c u m u l a t i o n . T r i p l i n g t h e d o s e o f [ 3 H ] a c e t a t e to 75 # C i clarified t h e s e p a r a t i o n o f l-alkyl a n d l - a c y l P A F d e r i v a t i v e s (Fig. 3). U s i n g t h e P A F b i o a s s a y , w e f o u n d significant q u a n tities o f P A F r e l e a s e d into t h e m e d i a ( T a b l e !11. F u r -

°'~--~

4 00[

Control

Hypoxia

Ionophore

ltyp+ Ion

5.64-3.1

1.8+_1.8

6.8+-6.8

22.7+_7.3*

9.(1_+6.1

12.7_+8.6

2.8_+2.11 5.2_+4.8

• represents P < 0.05 compared to other groups at 3(I rain.

t h e r m o r e , 30 m i n o f h y p o x i a a n d A 2 3 1 8 7 (! p . M ) t o g e t h e r significantly i n c r e a s e d P A F r e l e a s e ( P < 0.05). By 2 h, this r e s p o n s e to h y p o x i a a n d i o n o p h o r e w a s n o longer detected. Since hypoxia augmented the ionophore-induced r a d y l P A F a c c u m u l a t i o n in H U V E C ' s , w e s t u d i e d t h e effect of hypoxia and ionophore on PAF enzyme regulation. A l t h o u g h h y p o x i a a l o n e a n d i o n o p h o r e a l o n e h a d n o statistically s i g n i f i c a n t e f f e c t o n H U V E C p h o s p h o l i p a s e A , a n d a c e t y l h y d r o l a s e activity, h y p o x i a a n d i o n o p h o r e t o g e t h e r s y n m , istically i n c r e a s e d p h o s p h o -

TABLE II! The effect of hypoxia and iom~phore on emtothelial celt-associated iduzq~ito[ipase A 2 (PLA ,) actirity a,td acetylhydrokise (AH) actirity at 30 min

Cell monotayers were scraped :rod enzyme activity assessed as described in Materials and Methods. Data represent the mean_+S.E. for five experiments. Control PLA (pmol/mg/min) 0.l) AH (nmol/mg/min) 2.3+0.1

5001 >.:

PAF at 30 min (ng/ml) PAF at 2 h (ng/ml)

Hypoxia

lonophore Hyp+ion

11 ±0.4 6 _+11.3 2.2_+0.2 1.7_+0,1

1(18 +42 * t.4+ 11.3 *

• Represents P < 0.05 using ANOVA.

,-~eL"~jOI ,--d

TABLE IV

200 L ;00

t

0

g

o-o-o-ooooooo~ 5

,

o

'

+o_ot

~ ooo-o

10 Distance (cm)

15

The effect of h)'poxht and th,mtbin on emiothelml ccll.assocmted rcldylPAF acctmmlation at different #wttbation times

Cell monotayer., were exposed to normoxic or hypoxic condilions and subsequently treated with thrombin tl U / m l ) a n d assessed fi)r radylPAF accumulation as described in Materials and Methods. Dala represent mean + S.E. for four experimenls.

Fig. 3. Identification of i-acyl.2.[~ll]acelyi-GPC and i-O-alkyl-2[3lt]acclyl-GPC in human umbilical v,:in endothelial cells. RadyIPAF was further separated into i-acyl :md l-atkyl dcriwttives using phospholipase C digestion, reacelylation, and thin-layer chromatography as described in Materials and Methods. Zone 2 corresponds to porlitms migrating with atky|PAF standards treated identically to the samples. Zone 1 corresponds m !h'~ 1-:leyl-2-acetyi-GPC derivative.

Control radylPAF at 3t1' (cpm) rudyiPAF at 1211' (cptr:

|lypoxia Thrombin

.145+7,-, 47t~±79

89311___857*

124_+I11 132_+35 5113_+508 *

Hyp+Throm 11613+_ 12711 * 9486_+ 11195 **

• Signifies P < 0.01. ** P < 0.1t5 compared to thrombin alone.

209 lipase A~ and decreased acctyihydrolase activity after 30 min incubations (Table !I|). DFP completcIy blocked lysoPAF formation in the acctylhydrolase assay, therefore PAF degradation was all secondary to acetylhydrolase activity. Finally, thrombin-induced radylPAF accumulation was augmented by hypoxic priming similar to ionophorc stimulation (Table IV). At 30 min, thrombin significantly increased radylPAF accumulation, and the major derivative was l-acylPAF (data not shown). At 120 min hypoxia augmented the thrombin-induccd increase in radyiPAF accumulation (9486± 1095 cpm vs. 5113 + 508 cpm, P < 0.05). Discussion

We have clearly shown that hypoxic priming enhances HUVEC radyiPAF accumulation in the presence of ionophore or thrombin. Although greater than 90% of radylPAF identified is the acylPAF derivative, we have shown that bioactive PAF is secreted from endothelial cells into the surrounding media. The rapid increase in radylPAF accumulation to agonist challenge in this study is consistent with the acute radyiPAF response to agonist stimulation of endothelial cells published previously [9-11]. Acute alterations of endothelial cell radylPAF production may have multiple pathophysiological effects and recent studies suggest that increased ceil-associated radylPAF alters the inflammatory response by increasing neutrophil adhesion to HUVEC monolayers [21]. The role of hypoxiainduced PAF synthesis in HUVEC's on angiogenesis, thrombogenesis and vascular reactivity remains unknown. Calcium ionophore increased radylPAF accumulation in this study similar to others [22]. Furthermore, hypoxia-induced radylPAF formation required the presence of calcium ionophore or thrombin; HUVEC monolayers with or without hypoxia had negligible radylPAF production without A23187 or thrombin. Since monolayers of endothelial cells are an artificial system, it is plausible that low levels of circulating agonists in vivo, similar to ~hrombin, stimulate cell metabolism and allow i,cre:,,sed PAF accumulation under stressful conditions. In this manner, acute hypoxic stress can increase PAF synthesis, and modulate endothelial c~ti physiologic responses. Although many cell types synthesize and release PAF [1,2], most reports suggest that endothelial cells retain all newly formed PAF [9,10]. However, Camussi et al. has previously shown using the platelet aggregation bioassay that PAF is rele:~sed into media at 31) min from HUVEC's stimulated b~ ionophore [221. In addition, Bussolino and co-workers have shown that human endothelial cells release significant quantities of PAF

aftc~ 6 h of exposure to interleukin-I [23]. In our study. we confirmed that PAF i'; released from HUVEC's into ~hc surrounding media. We have characterized the released compound as authentic PAF based on (a) digcstio0 by phospholipase A 2, (b) separation on HPLC identical to that of known PAF standard, and (c) the activation of rabbit platelct serotonin release and inhibition by PAF rcceptor antagonists. Although the mechanism of PAF release from cells is poorly understood, re:cnt evidence suggests that the rugulation depends on localization of PAF to the plasma membrane, transbilascr movement, and PAF removal from the membrane by extracellular and intracellular accep. tor molecules [24]. Enhanced PAF release following hypoxic stress may impact on several endothelial cell functions including regulation of vascular tone, the inflammatory and hnmune response and "~",,,,O,,,t,,.,~,.,t,.-~k ...... :~is. and warrants further investigation. Several recent sludi".s have evaluated endothelial cell radylPAF accumulation, and have found a predominance oflhe I-acvi derivative in :;t=1a " ...... " ' - d cells [ 12, i2]. AcylPAF has approx, l(~l-toid less biological activity than alkylPAF [!4], and the significance o1 this derivative remains unclear. Our findings coincide with previous reports that acylPAF is the predominant PAF derivative formed in stimulated endothelial cells 112,13] and in addition, suggest that hypoxia primes the formation of this phospholipid compound. Further studics are necessary m delineate the physiologic importance of this phenomenon. Hypoxic priming of agonist-induced radylPAF accumulation in HUVEC's was a~sociated with increased phospholipase A 2 activity and decreased acetylhydrolase activity io this study. Therefore, both activated PAF synthesis and inhibited PAl- degradation may contribute to increased radylPAF accumulation in endothelial cells. Increased phospholipase A 2 activity has been described in endothelial cells [25], but our data are novel in that they suggest an effect of hypoxia on phovpholipase activation in HUVEC's. Although the major derivative of radylPAF is acylPAF in endothelial cells, recent repores have shown that phospholipase A : [26] and acetylhydrolase [27] can effectively utilize the acyl derivative as a subst,'ate. These important enzymes of PAF regulation play a critic:d role in endothelial cell response to hypoxia. In summary, to our knowledge this report is the first to describe increased radylPAF synthesis in stimulated HUVEC's toilowing hypoxic exposure. Furthermore, HUVEC's release PAF into the surrounding circulation which may then act locally on smooth muscle cells. inflammatory cells, and perhaps other neighboring endothelial cells. Further studies are required to investigate the pathophysiotogic effect of hypoxia on endothelial cell function, to define the mechanisms of hypoxiainduced alterations in PAF synthesis, and to clarify the

210 b i o l o g i c i m p o r t a n c e o f a c y l P A F in s t i m u l a t e d e H d o t h e Iial cells. Acknowledgements T h i s w o r k w a s s u p p o r t e d in p a r t b y U S P u b l i c Health Service, NIH grant RR-05370 and Dee and Moody grant, Evanston Hospital Corporation,

References I Hanahan, D.J. (t986) Annu. Rev. Biochem. 55. 483-509. 2 Snyder, F. (1990) Am. J. Physiol. 259, C697-708. 3 Benveniste, J. (1988) in Biological Membranes: Aberrations in Membrane Structure and Function (Karnovsky, M.L., Leaf, A., 139lis, L.C., eds.), pp. 73-85, Alan R. Liss, New York. 4 Caplan, M.S., Sun, X.M. and Hsueh, W. (1990) Gastroenterology 99, 979-986. 5 Prevost, M.C., Cariven, C., Simon, M.F., Chap, H. and Douste,'~ O A ' h ~ ,'j, Biazy, L. ~Jgo,,j Biochcm. Biophys. Res. Commun. ! !9, .8-6... 6 Prevost, M.C., Cariven, C. and Chap, H. (1988) Biochim. Biophys. Acta 962, 354-361. 7 Fajardo, L.F. (1989) Am. J. Clin. Pathol. 92, 241-250. 8 0 g a w a , S., Geriach, H., Esposit~, C., Pasagian-Macaulay, A., Brett, J, and Stern, D. (1990) J. Clin. Invest. 85. 1090-1098. 9 Prescott, S.M., Zimmerman, G.A. and Mclntyre, T.M. (1984) Proc. Natl. Acad. Sci. USA 81, 3534-3538. 10 Mcintyre, T.M., Zimmerman, G.A. and Prescott, S.A. (1986) Proc. Natl. Acad. Sci. USA 83. 2204-2208. 11 Mclntyrc, T.M., Zimmerman, G.A., Satoh, K, and Prescott, S.M. (1985) J. Clin, Invest. 76, 271-280.

12 Mueller, H.W., Nollert, M.U. and Eskin, S.G. (1991) Biochem. Biophys. Res. Commun. 176, I557-1564. 13 Clay, K.L., Johnson. C. and Worthen, G.S. (1991) Biochim. Binphys. Acta 11194,43-5"1, 14 Triggiani. M., Goldman, D.W. and Chilton, F.H. (1991) Biochim. Biophys. Acta 1084, 41-47. 15 Zimmerman, G.A., Whatley, R.E., Mclntyre, T.M., Benson, D.M. and Prescott, S.M. (1990) Methods Enzymol. 187, 521-538. 16 Gonzalez-Crussi, F., Hsueh, W. and Arroyave, J.L. (1987) FASEB J. 1,403-405. 17 Folch, J., Leesa M. and Sloane Sianley, G.H. (1957) J. Biol. Chem. 226, 497-509. 18 Bligh, E. and Dyer, W. (1959) Can. J. Biochem. Physiol. 37. 911-915. 19 Blank, M.L. and Snyder, F. (1983) J. Chromatogr. 273. 415-420. 20 Caplan, M., Hsueh, W.. Kelly, A. and Donovan, M. (1990) Prostaglandins 39, 705-714. 21 Zimmerman, G.A., Mclntyre, T.M. and Prescott, S.M. (1985) J. Clin. Invest. 76, 2235-2246. 22 Camussi, G., Agliena, M., Malavasi. F., Tetra, C., Piacibello, W., Sanavio, F. and Bussolino, F. (I983) J. hnmunol. 131, 2397-2403. 23 Bus.~eqno, F., Breviario, F., Tetta, C., Aglietta, M., Mantovani, A. and Dejana, E. (1986) J. Clin. Invest. 77, 2027-2033. 24 Bratton, D.L., Kailey, J.M., Clay, K.L. and Henson, P.M. (1991) Biochim, Biophys. Acta 1062, 24-34. 25 Whatl,~y, R.E., Nelson, P., Zimmerman, G.A., Stevens, D.I... Parker, C.J., Mclnlyre, T.M. and Prescott, S.M. (1989) J. Biol. Chem. 264, ~,325-6333. 26 Angle, M.J., Paltauf, F. and Johnston, J.M. (1988) Biochim. Biophys. Acta 962, 234-240. 27 Stafforini, D.M., Prescott, S.M. and Mclntyre, T.M. (1987) J. Biol. Chem. 262, 4223-4230.