[CANCER RESEARCH 44,120-125, January 1984]

Inhibition of Stimulated Prostaglandin Biosynthesis by Retinoic Acidin Smooth Muscle Cells1

Sanford M. Goldstein,2 Michael A. Moskowitz,3 and Lawrence Levine4

Department of Biochemistry, BrandéisUniversity, Waltham, Massachusetts 02254 [S. M. G., L L], and Department of Neurology and Neurosurgery MassachusettsGeneral Hospital, Harvard Medical School, Boston, Massachusetts 02114 [M. A. M.]

ABSTRACT

Bovine aorta smooth muscle cells (SMC) incubated with atumor-promoting phorbol ester, 12-O-tetradecanoylphorbol-IS-

acetate (TPA), released increased levels of prostaglandin I2[measured as its stable hydrolytic product, 6-keto-prostaglandin Fi„(6-keto-PGF,,,)], and this response was inhibited by all-frans-

retinoic acid (RA) at concentrations as low as 17 nw. Retinol andretinyl acetate, at concentrations as high as 1.7 and 1.5 U.M,respectively, did not inhibit the TPA-stimulated 6-keto-PGF, „production. RA was not cytotoxic at 1.7 U.M,as assayed byexclusion of trypan blue dye. Inhibition by RA was increasedafter preincubation of the SMC with RA prior to TPA stimulation.The inhibition of arachidonic acid (AA) metabolism by RA wasnot specific for TPA stimulation; RA inhibited prostaglandin production after SMCs were stimulated by serotonin; melittin; theCa2+ ionophore, A23187; and fetal calf serum. RA had no effect

on phorbol ester binding to SMC, nor did it inhibit increased 6-keto-PGF,,, production in SMC treated with exogenous AA. WhileRA inhibited TPA-stimulated production of 14C-labeled 6-keto-PGF,,, from [1*C]AA-labeled cells, it did not inhibit the accumulation of [14C]AA in the culture medium. The data suggest that

RA inhibits stimulated, rather than basal, levels of prostaglandinproduction. RA does not seem to act by inhibiting the deacylationof AA from cellular phospholipid pools, insofar as this is reflectedin the accumulation of AA in the media, but may inhibit reactionsat, or after, the generation of endoperoxides by cyclooxygenase.

INTRODUCTION

Retinoids, synthetic and natural derivatives of vitamin A, causestriking and diverse changes in the growth and differentiation ofmany tissues and cultured cells. One approach to the study ofretinoid action has been to investigate their biochemical effectsor their modulation of other effectors of cellular development (2,16). Not surprisingly, they seem to affect many biochemicalpathways. These alterations have varied from cell type to celltype, and opposite effects have occurred in different cell lines.This is especially true of the recently documented effects ofretinoids on AA5 metabolism. Several retinoids stimulated PGE2

1This work was supported by Grants GM 27256, CA 17309, and NS 19038-01

from the NIH. Publication No. 1461 of the Department of Biochemistry BrandéisUniversity, Waltham, Mass. 02254.

2Supported by NIH National Research Service Award AM06999-01.3Established Investigator of the American Heart Association.'American Cancer Society Research Professor of Biochemistry. To whom

requests for reprints should be addressed.sThe abbreviations used are: AA, arachidonic acid; PGE2, prostaglandin E2;

PGFj. prostaglandin Fj.; PGIj, prostacyclin; 6-keto-PGF,., 6-keto-prostaglandinF,„;TPA, 12-O-tetradecanoylphortiol-13-acetate; SMC, smooth muscle cell; RA,all-frans-retinoic acid; MEM-, Eagle's minimal essential medium; MEM*. 10% fetal

calf serum; BSA, bovine serum albumin; PDBU, phorbol-12,13-dibutyrate; HPLC,high-performance liquid chromatography; RIA, radioimmunoassay

Received May 16,1983; accepted September 28,1983.

and PGF2„,and PGI2 (measured as its nonenzymatic hydrolyticproduct, 6-keto-PGF,,,) in MDCK dog kidney cells (15), butinhibited TPA-stimulated 6-keto-PGF,,, production in bovine aortaSMCs (13) and TPA-stimulated PGE2 synthesis in chick embryo

fibroblasts (17). Retinoids also inhibited PGE2 formation by rheumatoid synovial cells in culture (3), and one retinoid, 4-hdyroxy-phenyl retinamide, inhibited prostaglandin production in severalcell lines (12).

AA, the precursor of the prostaglandins in mammalian cells, ismetabolized by a complex series of reactions. Inhibitors that actat different steps of the overall reaction have been described.Nonsteroidal antiinflammatory drugs, such as indomethacin andaspirin, inhibit the enzyme cyclooxygenase and prevent formation of endoperoxide, a substrate for prostaglandin production;the antiinflammatory steroids, such as dexamethasone, inhibitexpression of acylhydrolase activity (7), possibly by inducing thesynthesis of an inhibitor (1, 6); while antioxidants also blockproduction of an endoperoxide, a substrate for the prostaglandins (14). However, the mechanisms by which retinoids inhibitAA metabolism are not understood. Therefore, we chose a cellline in which AA metabolism was inhibited by RA in order tomore fully characterize this inhibition and obtain informationconcerning its possible mechanism or site of action in vivo.

MATERIALS AND METHODS

Cell Culture. Bovine aorta SMCs were prepared from primary cultureof freshly obtained calf aorta as described by Ross (20). Cultures weremaintained in MEM*. Cells were used from passages 2 to 6 and within

6 weeks from first passage. For assays of the effects of agonists ofprostaglandin production and retinoids, cells were treated with trypsinand plated at a concentration of 10s cells/35-mm dish (SMC) and incubated overnight at 37°in MEM* containing penicillin and streptomycin.

The medium was then aspirated, and the cells were washed twice andwere then incubated for specified periods (usually overnight) at MEM~

with penicillin and streptomycin containing various concentrations of TPAor other agonists, with or without retinoids. Media were then removedand assayed for AA metabolites.

Binding of Phorbol Ester to SMC. Phorbol ester binding to SMC wasquantitated by a modification of the assay described by Jaken eíal. (9).Trypsin-treated SMCs were plated in MEM* in 24-well dishes at 1.75 to2.5 x 105 cells/well and were incubated overnight. Medium was then

aspirated, and the cells were washed twice and then treated for 30 minwith MEM- supplemented with BSA, at 4 mg/ml, or MEM+, and contain

ing various concentrations of tritiated PDBU with or without unlabeledPDBU (20 MM). Some wells were also treated with RA. The cells werethen placed on ice and washed twice with ice-cold BSA-containingmedium. Cell lysates were collected after cells were treated for 30 minat 37°with 0.1 N NaOH, 200 /.I/well, followed by one washing of the

wells with 200 Ml 1% Triton X-100. The NaOH and Triton X-100 fractionswere combined and counted in a Beckman LS7500 scintillation counter.

Labeling of Cellular Lipids with [14C]AA and Analysis of AA Metabolites. SMC (30 ml) at 105 cells/ml in MEM- were mixed with 50 /il (0.005

120CANCER RESEARCH VOL. 44

on May 21, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

•-r:w

Inhibition of Prostaglandin Production by Retinole Acid

mCi) of [14C]AA (New England Nuclear; specific activity, 52.9 mCi/mmol)and plated at 10s cells/35-mm dish. Cells were incubated overnight,

during which time AA is incorporated into cellular lipids (7). The cellswere then washed 5 times with MEM" containing BSA (4 mg/ml) to

remove free, unincorporated AA [media with BSA was used throughout,since previous experience (8) suggested that higher levels of releasedAA are found in the media after stimulation of cells, possibly due tobinding of AA to BSA in media]. Preliminary experiments (data not shown)showed that no significant differences in radioactivity released from TPAor TPA plus RA-treated cells were found using MEM" without BSA

(although overall levels were lower). The cells were then preincubatedwith RA (1.7 JIM) or medium for 2 hr before the medium was aspiratedand replaced with MEM"/BSA with TPA (10 ng/ml), with or without RA

(1.7 UM). Since preliminary experiments showed that released radioactivity was highest after an overnight treatment with TPA, media werecollected the next morning for analysis by HPLC and were pooled into 2groups: TPA alone; or TPA with RA. Part of each pool was retained forRIA analysis. Lipids in the remaining media were extracted with ethylacetate and analyzed by reverse-phase HPLC on a ¿tBondapakcolumn

as described previously; recoveries of the prostaglandins (PGE2, PGF^,and 6-keto-PGF,„)ranged from 64 to 89% (11). The results presented

are not corrected for this recovery. A constant amount of tritiated PGF2„also was added prior to HPLC as a retention time control. All fractions(1 ml) were lyophilized and then counted as above; cpm per fractionwere normalized to cpm per ml of original culture fluid or percentage oftotal cpm eluted from the column, which gave similar results.

RIAs for the Prostaglandins. RIAs were carried out as describedpreviously (14). PGI2 was measured as its nonenzymatic hydrolyticproduct 6-keto-PGF,,,.

Trypan Blue Assay. Viability of cells treated with RA (1.7 MM)for 20hr was assessed by exclusion of try pan blue. Cells (10s) were treatedwith trypsin and taken up in 1.0 ml MEM+ and then centrifuged at 800

x g for 8 min. They were then resuspended in 200 n\ of 0.02% trypanblue dye and counted in a hemocytometer.

Materials. RA, retinol, and retinyl acetate were products of SigmaChemical Co. (St. Louis, Mo.). Stocks (1 mg/ml) were made up in dimethylsulfoxide and stored under argon at -80°. Working stocks were kept infoil at -20° for short periods. Purity was assessed by thin-layer chro-

matography using silica gel 60 thin-layer plate (layer thickness, 0.25 mm;

E. Merck, Germany) developed by hexane/ether, 1/1 (v/v). Serotonin,melittin, and the calcium ionophore A23187 were obtained from Sigma.

RESULTS

Inhibition of AA Metabolism by Retinole Acid. RA inhibited6-keto-PGF,,, production in SMC stimulated by TPA (10 ng/ml)

(Chart 1). Inhibition was seen with doses of RA as low as 16.6HM; RA at 6.6 IIM did not cause statistically significant inhibition(p > 0.5). The dose-response curve was steep and showedsaturation behavior, with maximal inhibition of 70%. Half-maximal

inhibition was caused by RA at 27 nM. Retinol and retinyl acetatein doses as high as 1.7 and 1.5 MM, respectively, caused nosignificant inhibition (p > 0.25 and p > 0.5, respectively). SMCsproduce small amounts of PGE2 and PGF2„.As seen in Chart 2,RA at 1.7 ft»caused 60% inhibition of 6-keto-PGFi„,but only

moderate inhibition of stimulated levels of PGF2<<(42%) and lessinhibition of PGE2 (21%). Inhibition by RA at 1.7 ^M was not dueto cytotoxicity as assessed by trypan blue exclusion; 98.9% ofRA-treated SMC excluded trypan blue, compared to 98.5% of

control cells treated with dimethyl sulfoxide carrier.Effect of Preincubation of Cells with RA. Assay of 6-keto-

PGF,,. in culture media from SMC at various times after additionof TPA showed an accumulation of the metabolite (Chart 3, fop).RA inhibition was not immediate; no difference between cells

g 60o.

if 50o.

èu 40

ID

o

Z

9!IO

IO-» IO'8 IO-7 IO'6CONCENTRATION OF RETINOID, M

IO-5

Chart 1. Inhibition of 6-keto-PGF,,, production in SMC stimulated by TPA andtreated with either RA (A), retinol (D), or retinyl acetate (O). Inhibition was determined by comparison of RIAs of individual TPA/retinoid-treated dishes (n = 4) to

mean of dishes treated with TPA alone (0% inhibition) minus mean of dishes withoutTPA (no stimulation). Values of inhibition for individual dishes were averaged andare expressed as mean ±S.E. Significance was determined by comparison ofmeans by f test. Curve was fitted by a nonlinear computer plotting program.

0.3

_ 0.2E

0.1 •

PGr2.—PGDIEZ—

I6

1.4

1.2

1.0

0.8

0.6

04

0.2

6-K»to-PGF,.

n-I < < <5 " fr

5 Z fr

»z

Chart 2. Inhibition of PGFj„,PGE2, and 6-keto-PGF,. by RA (1.7 pM). Mediafrom dishes in each group (n = 5/bar) were pooled and assayed in triplicate. Allresults were within 10% of each other and are expressed as the mean 6-keto-PGF,. (ng/ml) in culture media.

stimulated by TPA and those treated with TPA and RA could beseen for the first 90 min. Thereafter, TPA-stimulated SMC exposed to RA ceased producing 6-keto-PGFi„,while TPA-stimulated cells without RA continued to produce 6-keto-PGF!,,, al

though at a slower rate. Subsequent experiments (data notshown) demonstrated that preincubation of cells with RA (1.7UM) prior to TPA stimulation increased the inhibition caused by

JANUARY 1984 121

on May 21, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

S. M. Goldstein et al.

2 3

Tim* (Hours)

Charts. Top, time course of release of 6-keto-PGF,„into media after TPAstimulation. Bars, mean ±S.E. of cells treated with TPA (O), n = 4; cells treatedwith TPA and RA (A), n = 4; and MEM" control cells (D). «= 3. Bottom, effect of

a 3-hr preincubation of RA on 6-keto-PGF1a production. O, cells preincubated withMEM" prior to TPA. A, cells preincubated with RA prior to TPA plus RA as in a; D,

control cells preincubated in MEM .

RA, and that the additional inhibition was proportional to thelength of time of precinbuation with RA. This preincubation withRA was still effective in inhibiting TPA-stimulated 6-keto-PGF,„

production even when cells were washed several times afterpreincubation, and before TPA alone was added to cultures. Theeffect of a 3-hr preincubation of SMC with RA prior to TPA

stimulation is shown in Chart 3 (Chart 3, bottom). It is clear thatpreincubation inhibited the early production of 6-keto-PGF1a bySMC after TPA treatment. Thus, there is a certain lag before theeffects of RA are seen in stimulated SMC, and this can bedecreased by exposure of cells to RA prior to TPA stimulation.

Lack of Effect of RA on Phorbol Binding to SMC. Since thebinding of phorbol esters to cell membranes can be quantitated,we chose to investigate the effect of RA on binding of [3H]PDBU

in SMC. As shown in Chart 4, RA at 1.7 MM had no effect onbinding of PDBU to SMC. Similar data (not shown) were alsoobtained for SMC treated with RA (1.7 MM)for 3 or 24 hr priorto the binding assay. Thus, we were unable to demonstrate anyeffect of RA on binding of PDBU, even when used in dosessufficient to cause maximal inhibition of TPA-stimulated 6-keto-

PGF1oproduction.RA Inhibition of 6-Keto-PGF,,, Synthesis Stimulated by a

Variety of Different Agonists: Lack of Specificity of Inhibition.The data cited above, as well as those in the literature (9),suggest that RA does not act by interfering with the initialinteraction of TPA with the cell. Therefore, we chose to investi-

gate the effect of RA on prostaglandin production in SMC afterstimulation by a number of unrelated agonists of prostaglandinsynthesis. As shown in Chart 5, RA inhibition was not specificfor SMC treated with TPA. RA at 1.7 /¿Minhibited 6-keto-PGF,,,

production stimulated by serotonin, melittin, A23187, and 10%fetal calf serum. RA seemed to inhibit stimulated cells and, asnoted in all experiments, had no statistically significant effect onbase-line levels in unstimulated cells (P > 0.5).

Lack of Effect of RA on 6-Keto-PGF,, Production Stimu

lated by Addition of Excess Exogenous Substrate (AA). SMCsupplied with exogenous substrate in the form of AA producedincreased levels of 6-keto-PGFia. Other cells have been shown

to incorporate exogenous AA into cellular lipids, but also produceprostaglandins from exogenous AA, although at a much lowerrate (19). Therefore, exogenous AA was used, since the effectof RA on cells so stimulated might shed some light in vivo onthat part of the pathway of AA metabolism subsequent to releaseof AA from cellular lipids. The stimulated PGI2 production byexogenous AA (16 UM) was not inhibited by RA (Chart 6a).However, SMCs were capable of yet further stimulation by TPA¡nthe presence of exogenous AA, and RA (at 3.3 MM)did inhibitthis additional 6-keto-PGF1a production (Chart 6b). The exoge

nous AA (16 MM)did not interfere with RIA (data not shown).Lack of Effect of RA on Deacylation of Phospholipids.

SMCs whose cellular lipids were labeled with [14C]AA were

treated with RA (1.7 MM)for 2 hr prior to TPA stimulation in orderto maximize the RA inhibition. After an overnight incubation ofthe radiolabeled cells with TPA, the media from dishes (n = 6) inthe TPA or TPA plus RA (n = 7} group were combined, and a

small aliquot was counted for released radioactivity, which represented labeled AA and its metabolites. The cpm per ml ofmedia were similar between the 2 groups: 24,070 cpm/ml (TPA)versus 21,570 cpm/ml (TPA plus RA). However, duplicate assaysof 6-keto-PGF1o in each pool by RIA showed a large inhibition:

6.0 ng/ml (TPA) versus 2.0 ng/ml (TPA plus RA). The media werethen extracted (see "Materials and Methods") and analyzed by

HPLC (Chart 7). Total counts in the 6-keto-PGF1n and AA peaks

70

6010O

z 40

Io 30u.

a3> 20

°-io

I.O 10.0FREE PDBU ADDED (nM>

IOO

Chart 4. Effect of RA on binding of [3H]PDBU to SMC. Bars, mean ±S.E. of(O) cells treated with [3H]PDBU (n = 3), and mean of (A) cells treated with RA aswell as [3H]PDBU (n = 2) (Values for cells treated with RA as well as [3H]PDBU

were within 5% of the mean plotted.)

122 CANCER RESEARCH VOL. 44

on May 21, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Inhibition of Prostaglandin Production by Retinole Acid

were calculated and compared in the 2 groups. The AA peaksaccounted for most of the released radioactivity, and the 2 peaksdiffered by only 4%, as did the total radioactivity eluting from theHPLC column. However, the 6-keto-PGF1o peak in the chroma-

togram from TPA plus RA SMC media was only 47% as largeas that for TPA-stimulated cells. Control marker [3H]PGF2„dif

fered by only 1.2% between the HPLC runs.Additive Effects of Corticosteroids and RA. Addition of

dexamethasone to cultures of SMC treated with a maximallyinhibiting dose of RA (1.7 UM) resulted in additive inhibition. Asshown in Chart 8, dexamethasone at 0.1 »g/mlcaused a 71%inhibition of TPA-stimulated 6-keto-PGF,„production, and RA

caused a 68% reduction. However, when combined, they inhibited TPA stimulation by 87%, this further inhibition being statistically significant when compared to dexamethasone or RA usedsingly with TPA (p < 0.025).

Effect of Protein Synthesis Inhibitors of RA Inhibition. It ispostulated that RA may act through binding to a specific cyto-

plasmic receptor which has a subsequent effect on transcriptionand, ultimately, protein synthesis (18). However, attempts todeduce the effects of protein synthesis inhibitors failed to elucidate the necessity of protein synthesis for RA inhibition. Cyclo-

heximide, at 0.1 ng/m\, inhibited protein synthesis 46%, but didnot inhibit prostaglandin production (it stimulated production of6-keto-PGFi„slightly). Cycloheximide, at 1.0 fig/m\, inhibitedprotein synthesis 90%, but it inhibited base-line 6-keto-PGF1osignificantly. Therefore, although RA inhibited TPA-stimulated 6-keto-PGF1n more than did cycloheximide, the dependence of this

inhibition on protein synthesis could not be definitively determined.

6-Keto-PGFu (ng/ml)

roö

Oi

000b

MEM

MEM

)}

A23I87

A23I87 + RA

Melittin

Meiittm + RA

Serotonin

Serotonin + RA

MEM*

MEM+ + RA

Chart 5. Inhibition by RA ot stimulation caused by various agents. Bars, mean±S.E. except for serotonin (0.1 ng/mi) with identical values in 3 dishes; n = 3,except A23187 (0.4 fig/ml; (n = 6). Melittin was used at 1 fig/ml. Data are all fromovernight cultures. All differences were statistically significant with p values between < 0.05 (melittin)and < 0.005) (serotonin),except for base-lineMEM versusMEM- plus RA (see text).

28 r

20

I6

(9a

I2

UJ

4OL

2uiZ

uiS

UI

UIZ

Charte, a. Lack of effect of RA on AA-stimulated 6-keto-PGF,„production.Media was assayedafter overnight incubationwith agents as shown. Bars, pool of5 dishes assayed in duplicate (values were within 10% of each other). RA wasused at 3.3 ¡IM.b, RA inhibition of 6-keto-PGF,,,production stimulated by TPA inthe presence of exogenous AA. Identical dishes as used in a were also treatedwith TPA in addition to AA with or without RA (3.3 *IM).Note inhibition does notextend below baselineof MEM plus AA (a).

3300

2IOO

900

° 4OO

IDO

20 30 40 90Fraction

60 70 80

Chart?. HPLC chromatogram of extracted [14C]AAand metabolites. Arrows,where standards elute on this column. 12-HETE, 12-hydroxyeicosatetraenoicacid;PE, phosphatidyl ethanolamine;PC, phosphatidylcholine; LTB4,leukotriene B4.O,SMC treated with TPA. ».TPA plus RA (1.7

JANUARY 1984 123

on May 21, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

S. M. Goldsteinetal.

s

2 <CL+

2t-+su2

xUJO

ChartS Additive effect of dexamethasone and RA. Values are mean ±S.E. ofdishes (n = 3, except MEM" plus TPA, with n = 4); 6-keto-PGF,a production in thepresence of dexamethasone (DEX) plus RA (last bar) was significan«), lowercompared to that in the presence of each agent used separately (p < 0.025). Media

were assayed after overnight treatment.

DISCUSSION

These studies have described the inhibition of AA metabolismby RA in cultured SMCs. This inhibition was seen at doses muchlower than reported previously (3, 13). It is also interesting thatthe onset of inhibition, while rapid, was by no means immediate.Indeed, there was a significant lag seen with relatively high (1.7UM)doses of RA, and this could be eliminated by preincubationof SMC with RA prior to stimulation by TPA. In addition, nosignificant inhibition was noted with retinol or retinyl acetate.

Our finding that RA does not affect phorbol binding is consistent with previous studies which found no effect of RA on bindingto isolated cell membranes (5) and in whole cells (9). We havefound that RA induces no detectable change in binding evenafter prolonged periods in intact cells, at least at doses of RAsufficient to inhibit the stimulatory effects of TPA. We did notanalyze our binding data by Scatchard plots to compare thenumber of binding sites per cell, in light of criteria recentlyreviewed by Klotz (10). We were interested in correlating theeffect of RA on binding of PDBU, with RA inhibition of AAmetabolism in identically treated cells. We were not able to makemeasurements at PDBU concentrations at or above the inflectionpoint in our binding curve (Chart 4). Therefore, any estimates ofnumber of receptor sites would likely be in error. However, ourdata are sufficient to support the conclusions we made onbinding over the range of PDBU we used. In addition, our dataon the lack of specificity of RA inhibition, with regard to otherstimulators of AA metabolism, make it unlikely that phorbol esterbinding plays an important part in the inhibition of AA metabolismby RA. It is important to note that similar binding of PDBU hasrecently been observed using phospholipid liposomes presumably free of receptor sites (4).

There are several potential points in the pathways of AAmetabolism where RA might exert its inhibition directly, orthrough some metabolite or induced effector molecule. Since RAinhibited prostaglandin production stimulated by a variety ofagents, RA probably acts at a step or process in stimulating thecell shared by all of these agents, or at some common early

point in a pathway subsequent to an initial interaction betweenan agonist and the cell. An initial common point, which couldreflect changes in the earliest steps during and after stimulation,would be the deacylation of AA by cellular acylhydrolases. However, we found no inhibition in levels of [14C]AA released con

comitant with the very significant decreases in labeled andunlabeled 6-keto-PGF1<Tcaused by RA. If changes in prostaglandin production are to be accounted for by similar changes in thesteady-state concentration of free AA available to prostaglandinsynthetic enzymes (reflected by AA in media), then thesechanges should be proportional, as discussed in detail by Robinson ef al. (19). Clearly, this is not the case with RA inhibition inSMC. A similar argument is true of alterations in steady-state

levels of AA caused by increasing the rate of reacylation of AAback into phospholipids. Therefore, RA does not seem to act atthe level of release or reacylation of AA from or into cellularphospholipids, respectively. This interpretation rests on assumptions that: (a) AA released into media reflects rates of deacylationand reacylation of AA in the cell; (b) AA released into the mediais an accurate reflection of intracellular AA concentration; and (c)the [14C]AA accurately represents the kinetically important pool

of AA inside the cell. If there are small, individually regulatedpools of phospholipids which are used differentially by variousprostaglandin-synthetic enzymes, or compartmentalized poolsof AA, then small, significant changes might not be detectable inoverall accumulation of AA in the media.

Our data on AA deacylation are in contradiction with an earlierreport by Mufson ef al. (17), who described the inhibition oflabeled AA release by RA in chick embryo fibroblasts. Our systemdiffered from theirs in 2 respects; (a) we used a different cell line;and (D) these authors documented a peak in PGE2 and AArelease at 3 hr after TPA treatment, with a return to base-line ornear base-line levels at 24 hr. Our system seems to respond

very differently, with steady accumulation of these compoundsin media without observed declines.

The next well-described reaction in AA metabolism is the

generation of endoperoxides from free AA by cyclooxygenase.Inhibition of cyclooxygenase by RA has been reported by Zibohef al. (23). Their conclusion was based on inhibition of prostaglandin synthesis in microsomal fractions of guinea pig skin atRA concentrations at and above 160 MM and on inhibition ofoxygen uptake in preparations of sheep vesicular gland by RAat concentrations of 80 MM or greater when preincubated withRA for 1 min prior to initiation of the reaction. Inhibition ofcyclooxygenase alone is not sufficient to explain all of our datawithout taking several points into consideration. If RA at concentrations effective in vivo in our cells also inhibits cyclooxygenaseactivity in vivo, then one would have expected inhibition of AA-stimulated 6-keto-PGF,,, just as was observed with anothercyclooxygenase inhibitor, indomethacin.6 However, we found

that RA did not inhibit the increased production of 6-keto-PGF1<tcaused by exogenous AA, but did inhibit 6-keto-PGFia produc

tion stimulated by TPA, even in the presence of exogenous AA.These data might be explained in one of 2 ways, (a) TPA mightstimulate those oxidative reactions which are required for prostaglandin synthesis, and RA may block this stimulation. Thiswould account for RA inhibition of TPA-stimulated 6-keto-PGFi„

production, and the lack of effect of RA after AA treatment,

* S. M. Goldstein, M. A. Moskowitz. and L. Levine, unpublished results.

124 CANCER RESEARCH VOL. 44

on May 21, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

'•*."•'V.-V•-"- *

which presumably causes higher PGI2 production by increasingconcentrations of substrate available to the basal levels of synthetic enzymes within the cell. However, there are no data toindicate that TPA acts in this way, although theoretically it is alikely possibility (19). (b) Alternatively, our data could be explainedif there were several cyclooxygenase pools or pathways of AAmetabolism which are differentially sensitive to RA. The existenceof such separate pathways is suggested by other investigators.Schwartzman ef al. (21 ) described 2 pools of esterified AA in theperfused rabbit kidney. They proposed a model in which PGE2released from renal cells originate from 2 different enzymaticsystems: a base-line hormone-insensitive and a stimulated hormone-sensitive, coupled lipase-cyclooxygenase system. Theirmodel would account for the minimal inhibition by RA of basal 6-keto-PGF1(, production in SMC versus the large inhibition by RAof TPA-stimulated production, In addition, their model postulates

different cyclooxygenases in each pathway; our data are consistent with this model. More recently, Willems ef al. (22) proposed the existence of at least 2 cyclooxygenase pools inendothelial cells treated with acetylsalicylic acid. Prostaglandinproduction was measured using exogenous and endogenousAA substrate. The metabolism of exogenous AA showed a shortturnover time after aspirin was withdrawn, and a more rapidrestoration of activity than for the cyclooxygenase using endogenous AA. Such separate pools could explain our results, especially if the enzyme metabolizing exogenous AA is an exoen-

zyme, as they postulated, and not accessible to inhibitors synthesized within the cells, as our time course studies suggest. Inaddition, while the need for 4-log-higher concentrations of RA,

by Ziboh ef al. (23), might be explained by their use of an in vitrosystem, they measured inhibition of cyclooxygenase by RA usingexogenous AA.

Finally, it is possible that RA inhibits the conversion of endo-

peroxides to PGI2 by acting at the level of the terminal syntheticpathways which generate PGI2. However, our data and that ofother workers (3) demonstrated that RA can inhibit the production of prostaglandins as well as PGI2. This suggests that RAwould have to act at 2 or more separate points in these terminalpathways, if it were to inhibit prostaglandin and PGI2 productiondistal to the conversion of AA to endoperoxides, rather than atone point which is common to both pathways (i.e., cyclooxygenase).

REFERENCES

1. Blackwell, G. J., Camuccio, R., Di Rosa, M., Flower, R. J., Parente, L., andPersico, P. Macrocortin: a polypeptide causing the anti phospholipase effectof glucocorticoids. Nature (Lond.), 287:147-149,1980.

Inhibition of Prostaglandin Production by Retinoic Acid

2. Blalock, J. E., and Gifford, G. E. Retinoic acid (vitamin A acid) inducedtranscriptional control of interferon production. Proc. Nati. Acad. Sei. U. S. A.74: 5382-5386.1977.

3. Brinckerhoff, C. E., McMillan, R. M., Dayer. J.-M., and Harris, E. D., Jr. Inhibitionby retinóteacid of collagenase production in rheumatoid synovial cells. N. Engl.J. Med., 303:432-436,1980.

4. DeLeers, M., and Malaisse, W. J. Binding of tumor-promoting and biologicallyinactive phorbol esters to artificial membranes. Cancer Lett., 77: 135-140,1982.

5. Driedger, P. E., and Blumberg, P. M. Specific binding of phorbol ester tumorpromoters. Proc. Nati. Acad. Sei. U. S. A., 77: 567-571,1980.

6. Mirata, R., Schiffmann, E., Venkatasubramanian, K„Salomon, D. and Axelrod,J. A phospholipase Az inhibitory protein in rabbit neutrophils induced byglucocorticoids. Proc. Nati. Acad. Sci. U. S. A., 77: 2533-2536,1980.

7. Hong, S. L., and Levine, L. Inhibition of arachidonic acid release from cells asthe biochemical action of anti-inflammatory corticosteroids. Proc. Nati. AcadSci.U. S. A., 73:1730-1734,1976.

8. Hong, S. L., and Levine, L. Stimulation of prostaglandin synthesis by bradykininand thrombin and their mechanisms of action of MC5-5 fibroblasts. J. BtolChem., 251: 5814-5816,1976.

9. Jaken, D., Tashjian, A. H., Jr., and Blumberg, P. M. Characterization of phorbolester receptors and their down-modulation in GH.C, rat pituitary cells. CancerRes., 47:2175-2181,1981.

10. Klotz, I. M. Numbers of receptor sites from Scatchard graphs: facts andfantasies. Science (Wash. D. C.), 2Õ7: 1247-1249,1982.

11. Kobayashi, T., and Levine, L. Arachidonic acid metabolism by erythrocytes. J.Btol. Chem., 258: 9116-9121,1983.

12. Levine, L. N-(4-HydroxyphenylXetinamide: a synthetic analog of vitamin A thatis a potent inhibitor of prostaglandin biosynthesis. Prostaglandins Med., 4:285-296,1980.

13. Levine, L. Stimulation of cellular prostaglandin production by phorbol estersand growth factors and inhibition by cancer chemopreventive agents. In:Prostaglandins and Cancer, First International Conference: Prostaglandins andRelated Upids, Vol. 2, pp. 189-204. New York: Alan R. Uss, Inc., 1982.

14. Levine, L. Inhibition of the A-23187-stimulated teukotriene and prostaglandinbiosynthesis of rat basophil leukemia (RBL-1) cells by non-steroidal anti-inflammatory drugs, anti-oxidants and calcium channel blockers. Stochern.Pharmacol., 32: 3023-3026,1983.

15. Levine, L., and Ohuchi, K. Retinoids as well as tumour promoters enhancedeacylation of cellular iipids and prostaglandin production in MOCK cells.Nature (Lond.,) 276: 274-275,1978.

16. Lotan, R., Neumann, G., and Lotan, D. Characterization of retinoic acid inducedalterations in the proliferation and differentiation of a murine and a humanmelanoma cell line in culture. Ann. N. Y. Acad. Sci., 359:150-170,1981.

17. Mufson, R. A., DeFoe, D., and Weinstein, I. B. Effects of phorbol ester tumorpromoters on arachidonic acid metabolism in chick embryo fibroblasts. Mol.Pharmacol., 76: 569-578,1979.

18. Ong, 0. E., and Chitil, F. Cellular retinóteacid binding protein from rat testis.J. Btol. Chem., 253: 4551-4554,1978.

19. Robinson, D. R., Bastían,D., Hamer, P. J., Pfchiarallo, D. M., and Stephenson,M.L. Mechanism of stimulation of prostaglandin synthesis by a factor fromrheumatoid synovial tissue. Proc. Na«.Acad. Sci. U.S.A., 78: 5160-5164,1981.

20. Ross, R. The smooth muscle cell. II. Growth of smooth muscle in culture andformation of elastic fibers. J. Cell Bid., 50:172-186,1971.

21. Schwartzman, M., LJberman, E., and Raz, A. Bradykinin and angiotensin IIactivation of arachidonic acid deacylation and prostaglandin E2 formation inrabbit kidney. J. Btol. Chem., 256: 2329-2333,1981.

22. Willems, C., DeGroot, P. G., Pool, G. A., Gonsalvez, M. S., Van Aken, W. G.,and Van Mourik, J. A. Arachidonate metabolism in cultured human vascularendothelial cells. Evidence for two prostaglandin synthetic pathways sensitiveto acetylsalicylic acid. Biochim. Biophys. Acta, 773: 581-588,1982.

23. Ziboh, V. A., Price, B., and Fulton, J. Effects of retinoic acid on prostaglandinbiosynthesis in guinea pig skin. J. Invest. Dermatol., 65:370-374,1975.

JANUARY 1984 125

on May 21, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1984;44:120-125. Cancer Res   Sanford M. Goldstein, Michael A. Moskowitz and Lawrence Levine  Acid in Smooth Muscle CellsInhibition of Stimulated Prostaglandin Biosynthesis by Retinoic

  Updated version

  http://cancerres.aacrjournals.org/content/44/1/120

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/44/1/120To request permission to re-use all or part of this article, use this link

on May 21, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from