JOURNAL OF CELLULAR PHYSIOLOGY 151571-578 (1992)

Fibroblast Growth Factor Upregulates PCG/H Synthase in Rabbit Microvascular Endothelial Cells by a Glucocorticoid Independent Mechanism TARIQ MOATTER AND MARY E. CERRITSEN * Department of Phy,iology, New York Medical College, Valhalla, N e w York 10595 (J.M.); institute of Arthritis dnd Autoimmunity, Miles Inc., West Haven, Connecticut 065 16 (.M.E. C. ) The present study was undertaken to determine the effects of acidic fibroblast growth factor (aFGF) on eicosanoid synthesis in microvcssel endothelial cells derived from rabbit left ventricular muscle (RCME cells). We observed that aFGF increased AA conversion to PCE, in a time- and dose-dependent manner, and the stimulatory effect was abolished by actinomycin D and cycloheximide. Acidic FGF increased the recovery of PGGiH synthase activity following aspirin treatment, suggesting an action on de novo PGGiH synthase synthesis. Acidic FGF increased the incorporation of \35S] methionine into a 70 kD immunoreactive PGGiH synthase band. PGGiH syrithase synthesis following aspirin treatment was also increased by transforming growth factor b, while epidermal growth factor basic FGF and platelet derived growth factor were without effect. In addition, the actions of aFGF on de novo PGGiH synthase were compared in several endothelial preparations. Acidic FGF treatment of aspirin treated endothelial cells from rabbit lung microvessels and small pulmonary artery and from human lung microvessels all showed an increase in PGGiH synthase recovery. In contrast, similar treatment of human umbilical vein endothelial cells was without effect. Pretreatment of RCME cells with dexamethasone (1 pM) did not alter the aFGF induction of PGGiH synthase activity. We conclude that aFGF stimulates PGE, production by a mechanism that includes the de novo synthesis of PGGiH synthase. This mechanism appears to be distinct from previously described glucocorticoid sensitive translational controls of PC synthase synthesis by epidermal growth factor i n smooth muscle and me5angial cells. o 1992 Wiiey-Liss, Inc.

Fibroblast growth factors (FGF) constitute a family of structurally related polypeptides, of which the best characterized are acidic FGF (aFGF) and basic FGF (bFGF). Initially identified by virtue of its ability to stimulate the proliferation and phenotypic transformation of 3T3 fibroblasts, bFGF shares 55%total sequence homology with aFGF (Gospodarowicz et al., 1987). Both aFGF and bFGF modulate many properties of vascular endothelial cells, including growth rate and morphology. Culture supplements from bovine brain or retina enriched in FGFs are routinely used in the isolation and continuous culture of large and small vesselderived endothelial cells. It has been reported that when large vessel derived human endothelial cells are cultured in the presence of aFGF and heparin, a marked decrease in PGI, production occurs as a consequence of a dose- and time-dependent decrease in cyclooxygenase and PGI, synthase activity (Weksler, 1990). In contrast to the published observations with human umbilical vein endothelial cells, preliminary observations in this laboratory had suggested that microvascular endothelial cells grown in media 0 1992 WILEY-LISS, INC.

containing crude FGF preparations exhibited marked increases in basal and stimulated eicosanoid production. Recent reports (Bailey et al., 1988, 1989; Harris and Badr, 1990) have demonstrated that epidermal growth factor (EGF) elicits an increase in de novo synthesis of PGGiH synthase in smooth muscle and mesangial cells. The following study was undertaken to test the hypothesis that aFGF might modulate PG biosynthesis in microvessel endothelial cells in a n analogous manner.

MATERIALS AND METHODS Materials RPMI, fetal bovine serum, and other cell culture material were from GIBCO (Grand Island, NY). Pipets and sterile plasticware were from Falcon (VWR Scientific, Piscataway, NJ). Highly purified natural aFGF (derived from bovine brain), bFGF (derived from bovine

Received November 20, 1991; accepted January 28,1992.

* To whom reprint requestsicorrespondence should be addressed.

572

MOATTER AND GERRITSEN

brain), transforming growth factor p l (TGFP, isolated from porcine platelets), and platelet derived growth factor (PDGF, derived from human platelets) were from R and D Systems (Minneapolis, MN). Recombinant human epidermal growth factor (EGF) was purchased from Genzyme (Boston, MA). Aspirin, actinomycin D, NP40, Triton-X-100, Protein A, Protein A agarose, and aprotinin were from Sigma (St. Louis, MO). [35Sl methionine was from Dupont-New England Nuclear (Boston, MA); 13HJPGE2, r3Hl 6-keto PGF,, and ll"CI arachidonic acid were from Amersham (Clearbrook, IL). PGE, and 6-keto PGE',, standards were from Upjohn (Kalamazoo, MI), and prostaglandin antisera were from Advanced Magnetics (Cambridge, MA). Monoclonal and polyclonal antisera against PGH, synthase were from Oxford Biomedical (Oxford, MI). Heparin was from Fisher Scientific (Pittsburgh, PA). Arachidonic acid was from Nuchek Prep (Elysian, MN). Nuserum was from Collaborative Research (Bedford, MA).

Cell culture Microvessel endothelial cells (RCME) were isolated and cultured from rabbit ventricular cardiac muscle, identified as previously described (Gerritsen and Cheli, 1983; Gerritsen et al., 1988) and were routinely grown in RPMI/lO% FBS. Rabbit pulmonary artery (RPA), rabbit lung microvessel (RLB), and human lung microvessel endothelial (HLE) cells were kindly provided by Dr. William Carley (Miles, Inc.). Human umbilical vein endothelial cells (HUVE) were purchased from Clonetics (San Diego, CA). All cells used were virtually pure populations of endothelial cells (> 99%) based on three established criteria for the identification of endothelial cells: uptake of DiI-acetylated LDL, expression of Factor VIII related antigen, and angiotensin converting enzyme as described previously (Carley et al., 1990a,b; Gerritsen and Cheli, 1983; Gerritsen et al., 1988). Radioimmunoassay Radioimmunoassays for PGE, and 6-keto PGF,, were performed a s detailed in earlier publications (Gerritsen and Cheli, 1983; Gerritsen e t al., 1987). All reagents, growth factors, etc., used in the cell incubations were evaluated for interference in the radioimmunoassays and none was detected a t the concentrations used in this study. All washes and incubations with arachidonic acid were performed using Dulbecco's Modified Phosphate Buffered Saline (PBS; GIBCO) with 1 mM C a + + and 1 mM Mg'+. Most experiments were performed on 24-well culture plates unless otherwise indicated. Due to variability from one culture plate to another and from one cell preparation to another, each multiwell dish contained the appropriate controls to which comparisons were made. Immunoprecipitation RCME cells were grown to confluence in 60 mm petri dishes, washed twice with 2 ml of PBS, and incubated with methionine free medium containing 3% dialyzed FCS for 2 hr. The media were then replaced with 1ml of methionine free medium containing 3% dialyzed FCS, 100 pCi/ml [35Slmethionine with or without 200 ngiml aFGF. The cells were incubated 5 h r a t 37°C. At the end

of the incubation the cells were washed three times with PBS, then lysed with 200 p1 lysing buffer (diethyldithiocarbamic acid, 1mM; EDTA, 10mM; 1%NP40 in 50 mM Tris, pH 8.0). Aliquots of the lysates equivalent to 1 x lo7 cpm were transferred to 1.5 ml microcentrifuge tubes, precleared by incubation with 10 pl of nonimmune rabbit serum diluted to 100 pl in extraction buffer (190 mM NaC1, 6 mM EDTA, 2.5% Triton X-100,lO Uiml aprotonin, and 50 mM Tris, pH 7.4) for 2 h r a t 4"C, followed by addition of 40 p1of a Stapholococcus aureus suspension, and incubation a n additional 45 min. The tubes were then centrifuged at 15,OOOg for 5 min and supernatants transferred to new tubes. The supernatants were incubated with a n additional aliquot of S . aureus and centrifuged as above. The precleared supernatants were then incubated with 5 pl polyclonal anti PGGiH synthase antiserum (PG 20) in a total volume of 100 pl and immune complexes recovered with protein A sepharose. The pellets were washed three times with 1 ml buffer A (5mM EDTA, 150 mM NaC1,0.5% Triton X-lOO,O.l% SDS, 10 Uiml aprotonin in 50 mM Tris, pH 8.0) followed by three washes with 1 ml buffer B (5mM EDTA, 150mM NaC1,lO Uiml aprotonin, in 50 mM Tris, pH 8.0). Following the last wash the pellets were solubilized in 50 pl of SDS sample buffer (100 mM dithiothreitol, 2% SDS, 62 mM Tris, pH 6.8,0.001% bromophenol blue), and boiled 3 min. Samples were analyzed by SDS-PAGE using 10% Novex Tris-Tricine minigels.

Aspirin protocol Endothelial cells were grown to confluence in RPMIi 10% FBS. The culture media were removed and the cells incubated 16 h r in RPMI/5% serum, then washed with PBS ( 3 x 1ml). Cells were incubated with 400 pM aspirin in PBS for 45 min. Preliminary experiments were performed to establish that this dose and time interval inhibited the conversion of 5 FM arachidonic acid (AA)to PGE? greater than 95%. After removal of the aspirin containing buffer, the cells were washed twice with 1ml of PBS, then incubated in 3 ml of RPMI with or without growth factor for various periods of time. Following the indicated recovery periods, the cells were washed twice with PBS and incubated with 5 p M AA in PBS for 15 min. All incubations, washes, and treatments were carried out a t 37°C. Prostaglandin production in non-extracted supernatants was measured by radioimmunoassay . [14C] Arachidonic acid incubations with cells and phospholipid analyses [l*C] Arachidonic acid was evaporated under N, on ice and resuspended in 50 mM Na,C03. After a 15 min incubation on ice the AA was diluted in RPMI/3% dialyzed FBS to a final concentration of 0.5 pCiiml. Confluent RCME cells were incubated in the aforementioned labelling media for 24 h r a t 37°C. The cells were washed three times with 3 ml PBSi35 mm dish and incubated with or without aFGF for 5 h r in RPMI/O.5% FBS. After incubation with the cells the media were removed and the cells washed twice with PBS, removed by scraping in cold PBS, and pelleted by centrifugation (800g, 5 min) and lipids extracted by the method of Sligh and Dyer (1959). Phospholipids were separated

573

FGF INCREASES PGE, BIOSYNTHESIS TABLE 1. Effects of aFGF on RCME cell number, total protein, and membrane arachidonic acid1 Parameter Cell number (cells/35 mm well) Total protein ( p g / F mm well) Arachidonic acid (pg/mg protein)

Control

TABLE 2. Effects of aFGF on the redistribution of [14C] arachidonic acid in RCME cells’ Control aFGF (% of total cpm/well)

aFGF treated 0.1x 105

7.6 i 0.1x lo5

7.8

410 f 20

420 5 15

0.46 f 0.05

0.76 f 0.04*

‘RCME cells were incubated with or wthout (control) 200 ng/ml aFGF in RPMI/5% FBS (see Matenals and Methods) for 5 hr Cell number was determined using a hemocytometer, protein using the Pierce BCA assay, and arachidonic acid content by gas chromatography Data are expressed as the mean = S E M (n = 3) *Significantly different from control, P < 0 05

by thin layer chromatography on Silica gel H thin layer chromatography (TLC) plates following the method of Touchstone e t al. (19811, and the distribution of [l4C1 AA in the different phospholipids determined by scraping the appropriate areas of the TLC plate, transfer of the silica gel to scintillation vials, addition of 4 ml Liquiscint, and radioactivity quantitated in a Nuclear Chicago @-scintillationcounter. In other experiments, cells were prelabelled with [l4C1AA a s above, washed with PBS, then incubated with or without 200 ngiml aFGF for 0, 30, 90, 180, and 300 min. At the indicated time points, media were removed from the cells and lipids extracted as described previously (Medow et a]., 1989). Arachidonic acid and various metabolites were analyzed by TLC in silica gel G plates a s described previously (Gerritsen and Cheli, 1983; Medow et al., 1989). Extraction efficiency exceeded 90% for all samples. G a s c h r o m a t o g r a p h y and determination of arachidonic content of RCME cells Total membrane arachidonic acid content in control and aFGF treated (5 hr) RCME cells was determined by gas liquid chromatography according to the method of Kaduce e t al. (1982). The mass of fatty acids was determined using C17:O as internal standard.

Statistical analyses Data are expressed as the mean k standard error of the mean of n observations. For each experiment, unless otherwise indicated, the n was equal to or greater than three, and every protocol was reproduced at least twice. In experiments where multiple comparisons were made against a single control group, data were first analyzed by one way analysis of variance, followed by the Bonferroni modification of Student’s t test (Wallenstein et al., 1980). In other experiments, groups were compared using a non-paired Student’s t-test. Significance was accepted at P < 0.05. RESULTS aFGF effects on RCME membrane composition Treatment of RCME cells with aFGF (200 ngiml) for 5 h r did not alter total proteiniwell nor cell number/ well (Table 11, but did significantly increase total cellular arachidonic acid content. Cells prelabelled with 11-14ClAA for 24 hr, followed by a 6 h r incubation with or without aFGF (200 ngiml) exhibited minor differences in the distribution of labelled AA between the

Phospholipid class Lysophosphatidylcholine Sphingomyelin Phosphatidylcholine Phosphatidylserine Phosphatidylethanolamine Phosphatidylinositol Neutral liaids

0.6 i 0.1 0.6 0.1 32.0 f 3.7 2.0 i 0.2

* *

7.2 0.9 10.4 f 0.3 46.4 f 4.0

** *

0.5 0.1 0.6 0.1 35.0 1.4 3.2 i 0.1 16.8 f 1.4 12.9 f 1.1 32.5 4.0

*

‘RCME cells were prelabelled with 0.5 aCi/rnl [“C] arachidonic acid for 24 hr as detailedin Materials andlllethods. Cells were washedand incubated with orwithout aFGF(200 ng/ml) for5hrinRPMI/0.5RFBS,afterwhichlipidswereextractedand analyzed by thin layer chromatography as described in Materials and Methods. Uata are expressed as the percent of total radioactivityrecovered from the cells (total radioactivity was not different in the two groups) and are the mean of three experiments, standard error of the mean.

*

T i m e (min) Fig. 1. Effect of aFGF on the release of radiolabelled PGE, from RCME cells. RCME cells were prelabelled 24 hr with IL4ClAA and washed a s described in Materials and Methods. Medium with ( 0 ) or without (0) 200 ng/ml aFGF was added to cells and followed by incubation for the indicated time periods. The radiolabelled media were extracted and analyzed as described in Materials and Methods. Data are the mean of duplicate determinations. PGE,, solid lines; AA, dashed lines.

two grou s, with aFGF treated cells exhibiting slightly more [l BCl AA in phosphatidylethanolamine and slightly less in the neutral lipid fraction (Table 2). Of the membrane phospholipids, the major pools of [l4C1 AA were phosphatidylcholine, phosphatidylinositol, and phosphatidylethanolamine in both treatment groups. Addition of aFGF to [l4C1AA prelabelled cells resulted in a time-dependent increase in the production of radiolabelled PGE, compared to control cells such t h a t a t 5 hr the aFGF treated cells synthesized 3.4-fold more 1l4C1 PGE, than the controls (Fig. 1).This increase was not a consequence of a n increase in basal [14C] AA release since the amounts of unreacted [I4C] AA in the media were similar (Fig. 1). To assess possible effects of aFGF on the release and metabolism of arachidonic acid from endogenous stores, the release of radiolabelled AA and PGE, following a 10 min incubation with 5 pg’ml calcium ionophore A23187 was evaluated. Under these conditions the majority of radiolabel released co-migrated with authentic

574

MOATTER AND GERRITSEN

7 .- 40001

.-'

t

-

-e

15

T i m e (hrs) 001

Fig. 2. Time dependence of the effect of aFGF on synthesis of PGGiH synthase de novo. Cells were treated with aspirin as described in Materials and Methods and allowed t o recover in the absence (0) or presence (a) of 200 ngiml aFGF. At the indicated time points, media were removed from the cells, thc cells washed with PBS, and conversion of 5 pM AA to PGE, determined. Data are expressed as the mean ? S.E.M. (n = 3); for many of the time points the error was smaller than the symbol size.* indicates significantly different from control group at the same time point (P< 0.05).

PGE, (60-65%) with the remainder as unreacted AA. The amount of v4Cl co-migrating with 6-keto PGF,, the stable PGI, metabolite, was no greater than background. Acidic FGF treated cells s nthesized more [l4C1 PGE, (3,165 L 44 cpmi7 x 10 cells) than the controls (1,366 k 79 cpm/7 x lo6 cells) (n = 3). Additionally, aFGF treated cells released more [14C] AA (1,646 2 124 c p d 7 x lo6 cells) compared t o controls (918 4417 x lo6 cells) upon stimulation with 5 pgiml A23187. Aspirin irreversibly inhibits prostaglandin GiH synthase (cyclooxygenase) by acetylating a serine residue a t the active site. Recovery of cyclooxygenase activity following aspirin treatment requires synthesis of new enzyme, and thus provides a measure of de novo synthesis. Following aspirin treatment of RCME cells, a slow recovery of arachidonic acid conversion to prostaglandins (primarily PGE,, the principal arachidonic acid metabolite of RCME cells LGerritsen and Cheli, 19831) occurred, attaining pre-aspirin treatment levels after 8-12 hr. If cells were treated with actinomycin D (2 pg1ml) or cycloheximide (2 pgiml) following the aspirin treatment, there was no recovery of cyclooxygenase activity (data not shown). Inclusion of 200 ngiml aFGF to the culture media during the recovery period resulted in a n increased conversion of AA to PGE, at all time points tested (Fig. 2). The effects of aFGF were dose-dependent, with a near maximal effect at 200 ngiml (Fig. 3).The response to 400 ng/ml aFGF was not statistically different from the response to 200 ngiml. Heparin is known to potentiate the mitogenic activity of both crude or purified FGFs, a n action mediated at least in part by reducing inactivation (Damon et al., 1989). Heparin (25 pg) alone inhibited the recovery of PGE, synthesis following aspirin, yet potentiated the activity of a low dose of aFGF (Fig. 3, inset). Heparin did not modify the response to 200 pgiml aFGF (data not shown). The effects of several other growth factors

1

I

I

1 0

10

100

1000

[aFGF] n g / m l Fig. 3. Dose dependence of aFGF effect on the recovery of PGGiH synthase activity 5 hr following aspirin treatment (n = 3). Inset. Effects of heparin on control and aFGF stimulated recovery of PGGiH synthase 5 hr following aspirin treatment. Controls heparin (25 pg/ml) ( Q ) ,aFGF (10 ngiml) (ail,and heparin in combination with '*Significantlydifferent from control; significantly differaFGF (a). ent from aFGF (10 ng/ml) alone.

(n),

+

OCONTROL

EGF

EG3PDGF

B

*

I

0 1

E4bFGF 100 50

Fig. 4. Comparison of the effects of different growth factors (all tested at 200 ngiml) on the recovery of PGE, biosynthesis in RCME cells 5 hr after aspirin treatment. Control, no growth factor added. Data are expressed as the mean i S.E.M. (n = 4)." Significantly different from untreated control cells,p < 0.05.

were compared with that of aFGF. As shown in Figure 4, EGF, bFGF, and PDGF (200 ngiml) did not significantly affect recovery of cyclooxygenase activity, whereas TGFp (200 ngiml) increased PGE, production from AA although to a reduced extent compared to aFGF. Since dose-response relationships with other growth factors were not evaluated, this observation does not imply that aFGF is the most potent growth factor eliciting this effect. The effects of aFGF were compared across several other endothelial cell preparations (Fig. 5). Acidic FGF inclusion in the media following aspirin treatment also increased prostaglandin production in rabbit pulmonary artery, rabbit lung microvessel, and human lung microvessel endothelial cells, but had no significant effect on prostaglandin synthesis in human umbilical vein endothelial cells.

PGG/H synthase biosynthesis The foregoing observations were consistent with the possibility that aFGF treatment of RCME cells resulted in a n increased activity of PGGiH synthase. To directly assess the effects of aFGF on this enzyme, control and

FGF INCREASES PGE, BIOSYNTHESIS

6 LUU r

o 150

:13c

c

; 50

L

0

Fig. 5. Comparison of the effects:of aFGF on the recovery of PGGiH synthase activity in different endothelial cells. HUVEC, human umbilical vein endothelial cells; RLB, rabbit lung microvessel endothelial cells; RPA, rabbit pulmonary artery endothelial cells; HLE, human lung endothelial microvessel endathelial cells. Data are expressed as the percent of control PGE, synthesis observed upon incubation with 5 pM AA (n = 4).PGI, synthesis in control and aFGF treated groups yielded similar observations (data not shown).*Significantly different from untreated control cells of the same type, p < 0.05.

aFGF treated RCME cells were labelled with [35Slmethionine a s detailed in Materials and Methods, and immunoreactive [35S]methionine labelled PGG/H synthase determined by immunoprecipitation and SDSPAGE analyses. Cells treated with aFGF exhibited a twofold increase in the labelling of the 70-kD band when compared with controls (Fig. 6). We attempted to resolve the temporal sequence of aFGF stimulation of PGE, synthesis into transcriptional and translational phases. RCME cells pretreated with cycloheximide (2 pgirnl) or actinomycin D (2 pg/ ml), followed by incubation for 5 h r in RPMI showed no change in AA conversion to PGE, compared to untreated controls (Table 3). Pretreatment with actinomycin D or cycloheximide completely blocked the stimulatory effects of aFGF on AA conversion to PGs (Table 3). In Figure 7, RCME cells were incubated with aFGF for various time periods, after which a small aliquot of actinomycin D (2 pgiml final concentration) was added to the culture media. Actinomycin D addition prevented the induction of PGE, synthesis up to 45 min after the addition of aFGF. However, the effects of aFGF were independent of new RNA synthesis after 90 min. Addition of cycloheximide a t 90 min blocked the increase in PGE, production (not shown).

Effects of dexamethasone EGF induction of cyclooxygenase in smooth muscle cells and fibroblasts is blocked by pretreatment with dexamethasone (Bailey et al., 1988, 1989). Similarly, dexamethasone blocks interleukin-1 and TNF induction of PGG/H synthase in various cell types (Raz et al., 1989; Burch and Tiffany, 1989). However, PMA induced PGGiH synthase in Madin Darby canine kidney cells (Coyne et al., 1990) is not blocked by corticosteroids. To determine the actions of corticosteroids on aFGF effects on RCME PGGiH synthase, cells were treated with and without 1OP6Mdexamethasone for 24 h r prior t o the addition of aFGF. Previous studies by this laboratory have shown that RCME cells similarly treated with dexamethasone exhibit a 60-80% reduction in agonist stimulated eicosanoid release (Rosenbaum and Gerritsen, 1976). As shown in Table 4, pre-

575

treatment with dexamethasone did not block the increase in AA conversion to PGE, in response to aFGF.

DISCUSSION Although the availability of arachidonic acid is generally accepted as the major rate limiting factor in prostaglandin production, the conversion of AA to PGGiH is a second major control point. Work by many groups has established that PGGM synthase can be auto-inactivated by free radicals formed during the synthesis of oxygenated products and that endothelial and many other nucleated cells rapidly and continuously produce PGG/H synthase. Rapid inactivation of the cyclooxgenase pathway in endothelial cells was first observed by Brotherton and Hoak (1983). In the present study we took advantage of the actions of aspirin which irreversibly inhibits PGGiH synthase by acetylating the active site (Roth et al., 1975, 1983). Thus, recovery of PGE, production provided a relatively simple readout of de novo synthesis of PGGiH synthase. This was verified by the demonstration that the recovery of PGE, biosynthesis following aspirin treatment was completely blocked by cycloheximide. Our previous studies (Rosenbaum and Gerritsen, 1976) had demonstrated that the doses of cycloheximide (2 pgiml) and actinomycin D (2 p/ml) used in this study inhibit RCME cell protein and RNA synthesis, respectively, by greater than 95%. Direct inhibitory effects of cycloheximide on PG biosynthesis were ruled out by the demonstration that inhibition of protein synthesis with this agent had no effect on AA conversion to PGs in cells not treated with aspirin. In the present study we found that aFGF and TGFp increased the conversion of AA to PGE, in cells treated with aspirin and allowed to recover for 5 hr. The actions of aFGF were further characterized and demonstrated to be dose dependent, and at low concentrations potentiated by heparin. Heparin alone decreased AA conversion t o PGE,, a n observation similar with the previously described actions of heparin on human umbilical vein PGI, synthesis (Weksler, 1990).The mechanism of heparin’s inhibitory actions is unknown at this time and was not further evaluated in this study. The enhancement of low-dose aFGF actions on AA conversion to PGE, by heparin is probably due to the ability of heparin to bind (Maciag et al., 19841 and stabilize FGF bioactivity (Damon et al., 1989). aFGF treatment also increased PGE, biosynthesis in cells not treated with aspirin. Several lines of evidence indicate that this effect was mediated, at least in part, by a n increased de novo synthesis of PGGiH synthase: 1)aFGF stimulated the incorporation of [35S]methionine into a 70-kD band immunoprecipitated by antisera to PGGiH synthase and 2) the effects of aFGF were blocked by pretreatment of cells with inhibitors of protein (cycloheximide) and mRNA synthesis (actinomycin D). However, aFGF also increased the potential for microvessel endothelial cells to produce eicosanoids at several additional steps in the eicosanoid cascade. Fatty acid analyses in this study indicated that aFGF increased the membrane content of arachidonic acid roughly twofold in just 5 hr. Additionally, aFGF exerted effects on the distribution of arachidonic acid in

576

MOATTER AND GERRITSEN

Control Control

FGF

FGF

Mr 68kD Fig. 6 . Representative autoradiogram of immunoprecipitation of PGGIH synthase with the polyclonal antibody PG20. Acidic FGF treated cells exhibited a twofold increase in L”S] methionine into a 70-kD band. Standard: 68 kD Mr. TABLE 3. Actions of aFGF on RCME PGE:! biosynthesis require ncw mRNA and protein synthesis1 None Group Control aFGF

763 2.328

Pretreatment Cycloheximide Actinomycin D PGEz synthesis (ng/well/lS min)

*+ 15186*

504 k 27 795 f 198

665 k 45 527 55

+

‘Cunfluent RCME cells were pretreated 15 min with indicated agent in RPMI/5%

FCS, thebufferremovedandreplacedwithRPMI/6%FCS withorwithout 200ng/ml aFGF as well as the original pretreatmentagent. Aftera 5 minincubation,themedia wcrr removed, the cells washed with PRS, and the conversion of 3 pM A A to PGEa determined. Data are the mean of 4 determinations. *Significantly different from untreated control cells, p < 0.05.

Time (rnins) Fig. 7. Effects of actinomycin D on aFGF induced increase in PGE, synthesis. aFGF (200 nglml) was added a t time 0. At the indicated lime points, aliquots of actinomycin D ( 2 p,g/ml final concentration) were added to various wells. All cells were incubated a total of 5 hr; after which conversion o f 3 p M arachidonic acid to PGE, was determined as described in Materials and Methods. Solid circles indicate, at time 0, the synthesis of PGE, in cells not treated with aFGF and, at 5 hr, the synthesis o f PGE, by cells treated with aFGF and no actinomyein. The triangles correspond to the levels ofPGE, synthesized by cells treated with aetinomycin D at the arrowheads. Data are the mean i S.E.M. (n = 4).

membrane lipid pools, a n action which could result in a greater release of arachidonic acid upon stimulation of various acylhydrolases. Evidence in support of this latter possibility was provided by our observation that aFGF treated cells exhibited greater basal and A23187 stimulated release of [l4C1PGE, and arachidonic acid. Many growth factors and cytokines have been reported to modulate prostaglandin synthesis at the level of either (or both) phospholipase A, and PGGiH synthase. For example, EGF stimulates the recovery of PGGiH synthase following aspirin treatment in vascular smooth muscle (Bailey et al., 1988, 1989) and rat

TABLE 4. Effects of aFGF on RCME PGEz biosynthesis are not blocked by dexamethasonel None Group Control aFGF

Pretreatment M dexamethasone (PGE2 synthesis pg/well/l5 min)

445 f 83 727 f 108’

279

* 55

821 i 59’

M dexamethasone. The ‘RCME cells were incubated 24 hr with or without preincubation media were removed and replaced with RPMI/l% dialyzed FBS with or without (control) 200 ng/ml aFGF. After 5 hrincuhation, conversion of 3 pM AA to PGEI was determined as described above. Data are expressed as the mean standard error of the mean (n = 4). %Significantlydifferent fmm untreated control.

*

mesangial cells (Harris and Badr, 1990) by a mechanism requiring new RNA and protein synthesis and involving a n increased expression of the message for PGG/H synthase. Interleukin-1 has been shown to increase de novo synthesis of PGG/H synthase in a number of cell types, including endothelial cells, and, a t least in some cell types, may exert its actions by increasing the message levels for PGGiH synthase (Maier et al., 1990). Although the mechanism(s) whereby different growth factors and cytokines alter PGG/H synthase in activity are unclear, a n observation made by many laboratories has been the ability of glucocorticoids to block the induced increase in PGG/H synthase activity. In contrast, this does not occur for aFGF induction of PGG/H synthase activity in RCME cells, since pretreatment of the cells with M dexamethasone had no effect on the actions of aFGF. The molecular mechanisms of aFGF effects on PGGiH synthase in RCME cells were not evaluated in this study, although our observations t h a t the effects of aFGF required new RNA and protein synthesis suggest two major possibilities: 1)aFGF increases PGGiH synthase mRNA transcription and/or stabilizes the message or 2) aFGF affects the rate of PGGiH synthase mRNA translation. These questions will be addressed in future studies. The actions of growth factors on endothelial cell eicosanoid biosynthesis are controversial. Weksler (1990) and others (Hasegawa et al., 1988: Boutherin-Falson and Blaes, 1989) have reported that FGF and heparin decrease prostaglandin synthesis, while other groups have reported opposing findings (Kuwasima et al., 1988). The different sources of endothelium used in these disparate studies offer a possible explanation for the apparent discrepancy. In this study we evaluated

FGF INCREASES PGE, BIOSYNTHESIS

the effects of aFGF in detail on one endothelial preparation, the rabbit cardiac muscle microvessel endothelial cells, and compared some of these observations to several other endothelial cell preparations that were available t o us. We found that aFGF increased de novo synthesis of PGGiH synthase in small vessel and microvessel endothelial cells both rabbit and human in origin, but did not appear to have this activity in endothelial cells derived from larger blood vessels (HUVE). The inhibitory actions of FGF on HUVE reported by Weksler (1990)were observed a t time points of 24 hr or more. Similar t o our observations, Weksler (1990) found no significant effect of FGF on HUVE prostaglandin synthesis with incubations as short as 5 hr. The divergence in the actions of FGF on different endothelial cells may suggest yet another fundamental aspect of heterogeneity in the properties of large vs. small vessel endothelial cells, and emphasizes further the caution required in extrapolating observations made with one source of endothelial cells to all endothelial cells in general. The ability of growth factors to regulate eicosanoid synthesis at the level of PGGM synthase may play an important role in inflammation and wound healing. It is well known that PGGiH synthase is irreversibly autoinactivated by its metabolic products (Egan et al., (1976). Thus following the initial phase of inflammation, endothelial cell eicosanoids would be expected to be decreased rather than increased unless de novo synthesis of the enzyme were to occur. The increase in prostaglandin production in microvascular endothelial cells evoked by angiogenic factors such as aFGF TGF, may play roles in maintaining new vessel patency by inhibiting formation of intraluminal platelet aggregates and by eliciting prolong local vasodilation. Basic FGE’ is produced by endothelial cells (Hannan et al., 1988;Vlodavsky et al., 19871, is present in the extracellular matrix (Vlodavsky et al., 1990), and may be liberated by proteases generated by extravasating leukocytes or activated endothelial cells (Vlodavsky et al., 1990). Several recent studies now suggest that aFGF may be localized to various sites outside of the nervous system, including the cardiac myocyte (Casscells et al., 1990) and cells of the vessel wall (Spier et al., 1991; Weich et al., 1990). Therefore, following injury or inflammation aFGF and/or bFGF could play important roles in regulating the local endothelial cell response.

LITERATURE CITED Bailey, J.M., Makheja, N., Pash, J., and Verma, M. (1988) Corticosteroids suppress cyclooxygenase messenger RNA levels and prostanoid synthesis in cultivated vascular cells. Biochem. Biophys Res. Commun., 157:1159-1163. Bailey, J.M., Makheja, A.N., Pash, J., and Verma, M. (1989) Regulation of cyclooxygenase messenger RNA levels by EGF and corticosteroids in cultured vascular smooth muscle cells. New Trends in Lipid Mediator Res., 3:%16. Bligh, E.G., and Dyer, W.J. (1959)A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol., 37911-917. Boutherin-Falson, O . , and Blaes, N. (1989) Decreased prostacyclin production by human umbilical vein cells cultured with endothelial cell growth factor and heparin. Thromb. Res., 54:487492. Brotherton, A., and Hoak, J.C. (1983). Prostacyclin biosynthesis in cultured vascular endothelium is limited by deactivation of cyclooxygenase. J. Clin. Invest., 7231255-1261. Burch, R.M., and Tiffany, C.W. (1989) Tumor necrosis factor causes

577

amplification of arachidonic acid metabolism in response to interleukin 1,bradykinin and other agonists. J . Cell. Physiol., 141:8589. Carley, W.W., Niedbala, M.J., and Gerritsen, M.E. i1990a) Isolation and characterization of human lung microvessel endothelial cells. J. Cell Biol., 111:185a. Carley, W.W., Tanoue, L., Merker, M., and Gillis, C.N. (1990b) Isolation of rabbit pulmonary microvascular endothelial cells and characterization of their angiotensin converting enzyme activity. Pulmonary Pharmacology, 3:3540. Casscells, W., Speir, E., Sasse, J., Klagsburn, M., Allen, P., Lee, M., Calvo, B., Chiba, M., Haggroth, L., Folkman, J., and Epstein, S.E. (1990) Isolation, characterization and localization of heparin binding growth factors in the heart. J. Clin. Invest., 85:433441. Coyne, D.W., Mordhorst, M., and Morrison A.R. (1990) Regulation of eicosanoid biosynthesis by phorbol ester in Madin Darby canine kidney cells. Am. J . Physiol., 259:F69%F703. Damon, D.H., Lobb, R.R., D’Amore, P.A., and Wagner, J.A. (1989) Heparin potentiates the action of acidic fibroblast growth factor by prolonging its biological half life. J . Cell. Physiol., 138:221-226. Egan, R.W., Pazton, J., and Kuehl, F.A. (1976) Mechanism for irreversible self-deactivation of prostaglandin synthetase. J. Biol. Chem., 251:732%7335. Gerritsen, M.E., Nganele, D.M., and Kodrigues, A.M. (1987) Calcium ionophore (A231871and arachidonic acid stimulated prostaglandin release from microvascular endothelial cells: Effects of calcium antagonists and calmodulin inhibitors. J. Pharmacol. Exp. Ther., 240:837-846. Gerritsen, M.E., and Cheli, C.D. (1983) Arachidonic acid and prostaglandin endoperoxide metabolism in isolated rabbit coronary microvessels and isolated cultivated coronary microvessel endothelial cells. J . Clin. Invest., 721658-1671. Gerritsen, M.E., Carley, W., and Milici, A.J. (1988) Microvascular endothelial cells: Isolation, identification and cultivation. Adv. Cell Culture, 6:3547. Gospodarowicz, D., Ferrara, N., Schweigerer, L., and Neufeld, G. (1987)Structural characterization and biological functions of fibroblast growth factor. Endocrinol. Rev., 8:95-114. Hannan, R.L., Kourembanas, S., Flanders, K.C., Rogelj, S.J., Roberts, A.B., Faller, D.V., and Klagsburn, M. (1988) Endothelial cells synthesize basic fibroblast growth factor, and transforming growth factor. Growth Factors, 1:7-17. Harris, R.C., and Badr, K.F. (1990) Recovery of prostaglandin synthesis in rat glomerular mesangial cells after aspirin inhibition: Induction of cyclooxygenase activity by serum and epidermal growth factor. Prostaglandins, 39:2 13-222. Hasegawa, H., Yamaomoto, M., and Yamamoto, K. (1988)Stimulation of cell growth and inhibition of prostacyclin production by heparin in human umbilical vein cells. J. Cell. Physiol., 137603-607. Kaduce, T.L., Spector, A.A., and Bar, R.S. (1982) Linoleic acid metabolism and prostaglandin production by cultured bovine pulmonary artery endothelial cells. Arteriosclerosis, 2:38&389. Kuwashima, L., Graeber, J., and Glaser, B.M. (1988) Stimulation of endothelial cell prostacyclin release by retinal derived growth factors. Invest. Ophthalmol. Vis. Sci., 29:1213-1210. Maciag, T., Mehlman, T., Friesel, R., and Schreiber, A.B. (1984).Heparin binds endothelial cell growth factor, the principal endothelial cell mitogen in bovine brain. Science 225:932-935. Maier, J.A.M., Hla, T., and Maciag, T. (1990) Cyclooxygenase is a n immediate early gene induced by interleukin-1 in human endothelial cells. J . Biol. Chem., 26.5:10805-10808, Medow, M.S., Intrieri, L.I., Moatter, T., and Gerritsen, M.E. (1989) Dexamethasone effects on microvascular endothelial cell lipid composition. Am. J. Physiol., 257C512-C519. Rosenbaum, R., and Gerritsen, M.E. (1976) Effects of dexamethasone on rabbit coronary microvessel endothelial (RCME) cell prostaglandin synthesis. Am. J . Physiol., 250:970-977. Roth, G.J., Stanford, N., and Majerus, P.W. (1975)Acetylation ofprostaglandin synthase by aspirin. Roc. Natl Acad. Sci. U.S.A., 72:3073-3076. Roth, G.J., Machuga, E.T., and Ozols, J. (1983) Isolation and covalent structure of the aspirin modified active-site region of prostaglandin synthetase. Biochemistry, 222:46724675. Spier, E., Sasse, J., Shrivastav, S., and Casscells, W. (1991) Cultureinduced increase in acidic and basic fibroblast growth factor activities and their association with the nuclei of vascular endothelial and smooth muscle cells. J. Cell. Physiol., 147:362-373. Touchstone, J.C., Levin, S.S., Dobins, M.F., and Carter, P.J. (1981)

578

MOATTER AND GERRITSEN

Differentiation of saturated and unsaturated phospholipids on column chromatography. High Resolution Chromatogr., 4:432-424. Wallenstein, D., Zucker, C.L., and Fleiss, J.L. (1980) Some statistical methods useful in circulation research. Circ. Res., 471-9. Weich, HA. Iberg, N., Klagsburn, M., Folkman, J. (1990) Expression of acidic and basic fibroblast growth factors in human and bovine vascular smooth muscle cells. Growth Factors 2 3 1 3 3 2 0 . Weksler, R.B. (19901 Heparin and acidic fibroblast growth factor interact to decrease prostacyclin synthesis in human endothelial cells by affecting both prostaglandin H synthase and prostacyclin synthase. J. Cell. Physiol., I42:51.1-522. Wyche A,, and Needleman P. (1989) Temporal and pharmacological

division of fibrobl ast cyclooxygenase expression into transcriptional and translational phases. Proc. Natl. Acad. Sci. U.S.A.,86:1657lti61. Vlodavsky, I., Frieman, R., Sullivan, R., Sasse, J., and Klagsburn, M. (1987) Aortic endothelial cells synthesize basic fibroblast growth factor which remains cell associated and platelet derived growth factor which is secreted. J. Cell. Physiol., 131:402408. Vlodavsky, I., Korner, G., Ishai-Michaeli, R., Bashkin, P., Bar-Shavit, R., and Fuks, Z. (1990)Extracellular matrix-resident growth factors and enzymes: Possible involvement in tumor metastasis and angiogenesis. Cancer Metastasis Rev:, 9:203-226.

H synthase in rabbit microvascular endothelial cells by a glucocorticoid independent mechanism.

The present study was undertaken to determine the effects of acidic fibroblast growth factor (aFGF) on eicosanoid synthesis in microvessel endothelial...
867KB Sizes 0 Downloads 0 Views