Document daQqOOboNbKaaBO967qZOj6jR

'lit At-Ia.I 14q2)n;--, 65 L-IevietSciencePuhli%hersB.V. All tght%rcscrvcd(KXIS-27()()/Y2/SO5.(Xl 7- s BBALIP i4Oll The mechanism underlyingthe hypolipemic 'effect of perfluorooctanoiaccid(PFOA), perfluorooctanesulphonic acid(PFO.SA) and clofibriaccid Bente Haughom and Oystein Spydevold lnstititoife.VfedicaBliochemistry,Unit;ersiro%f.Oslo,Oslo (Norwa@%-) (Received 12 June 1992) Key words: Lipidmetabolism; TriacylglycerolC;holesterol,Hepatocyte; (Rat liver) ne influence of the peroxisomal prolif-.ratorpserfluorooctanoic acid (PFOA), perfluorooctane sulphonic acid (PFOSA) and cloribricacid on lipidmetabolism in ratswas studied. Dietary treatment of male Wistar ratswith these three compounds resultrd in rapid and t)ronounced retinctionin both cholester@l and triacylitlycerolisn serum. T-ne concentration of livertriacvlglycerols was increased by about 300% by PFOSA.Free cholesterol was increased by botfi--Pe mpounds. CholestMI ester was reduced to 50To by PFOSA as well by clofib.-ateI.n hepatocytes from fed rats.allthe compounds resulted in reduced cholesterol svnthesis from aceta!Le.pyruvate and hydroxymethy=glutarate, but there was no reduction of synthesis from mevalonic acid. :Me qxidaii@-oo-fn r)aimitaat@elwsaos increased in all group . The perfluoro compounds, but not cloiribratec,aused some reduction in fattvacid svnthesis.The activit-o.f, liverHMG-COA reductase was reduced to 50% or less in alltreatment groups and allthree compounds led to lower activitvof acyl-CoA: cholesterol acyltransfefa-se(AC-AT5. Changes in other enzymes related to lipid metabolism were iii-consisten'tn.e present data suggest that the hypolip-emi effect of these compounds may, at least partly,be mediated via a common mechanism. impaired production of lipoprotein particlesdue to reduced synthesis and esteriricationof cholesterol together with enhanced oxidation of fatty acids in the liver. Introduction Many hypolipcmic drugs cause p,-oliferationof per- oxisomes and i!Lcreasc the actii-,v of the peroxisomal 0-oxidation in rats [1-3]. Chemically, these drugs con- stitute a heterogeneous group including clofibrate,tib- ric acid, niadenate and long-chain acvlthioacetic acids, tiadenol, long-chain thia acids and MEDICA 16 (1-6]. It has been suggested that the increase in the fatty acyl-CoA oxidizing system contributes to the hy- polipemic effect of these drugs [7,S].The dominating mechanism underlying reduction of serum triacyl- glycerols and cholesterol by these drugs is,however, uncertain. Other mechanisms x-hich may be important for the hypolipemic effect include; reduced hepatic synthesis of fatty acids and cholesterol [9-12]; reduced Correspondence to:0. Spydevold. Instituteof Medical Biochemistry, P.O. l@ax I112. Blindern, N-031' Oslo. Noruay. Abbreviations:PFOA. perfluoroociint,icac:d:PFOSA. perfluorooclanc sulphonic acid: ACAT, ;ic-fl-CoAc:huictiteroalcyltransferase; HMG. hydroxy'mcthylglutliriacc'ic!: te-.-@i-decylthioaceiaiccid; VLDL very-low-dcnsitylipaprotcins. triacylglycerol release by the liver (13,141; increased rate of VLDL degradation [6]; increased uptake or reduced release of fatty acids by adipose tissue [15,16) and increased excretion of cholesterol into bile and feces (17]. Ikeda et al. (18,19] observed that perfluorooctanoic acid (PFOA) and perfluorooctane sulphonic acid (PFOSA) efficientlyinduced the peroxisomal 0-oxida- tion in rats that were fed these compounds (0.02% in the diet).Just et al.[20] reported that the perfluorocar- boxylic acids alter hepatic lipid metabolism and reduce serum lipid levels showing that these compounds also belong to the group of pcroxisomal proliferators with hypolipemic effect.The perfluorinated compounds are particularly interesting since they obviously are not subject to ordinary metabolic modifications. The ef- fects of these compounds must, therefore, be due to effects of the compounds per se. Peroxisomal inducers may bring about the hy- polipemic effect by affecting different steps in lipid metabolism. However. it seems likelythat,some impor- tant steps in lipid metabolism are common targets for these compounds. The unphysiological character of the perfluoro compounds makes itpossible that their effect 66 on lipidmciabolism follows a patternthat would makc it easier to understand possiblemechanism for the lipidreductioneffectof pcroxisomal prolifcrators. In the search for a possiblecommon mechanism underlying the hypolipemic effectof peroxisomal inducers, we have compared hepatic fairyacid metabolism,cholesterolsynthesisand the activitieosf enzymes relatedto these metabolic processdsin the liverof rats fed perfluorooctanesulphonic acid(PFOSA), perfluorooctanoicacid(PFOA) and clofibrate. Methods and Materials Animals Male Wistar ratswere used. They were divided into fivegroups. One group was allowed normal food ad libitum(controlrats).Three other groups were given food which contained 0.3% clofibriaccid,0.02% perfluorooctanoic acid or 0.02% perfluorooctane sulphonic acid,respectivelyn.e dietswere prepared by soaking standard food (in pelletform) in diethylether in which the compounds had been dissolved.The fifth group was restrictedin food intake to that consumed by the PFOSA group. The stock diet was used for these paired feeding experiments. Tle average body weight when the experiments startedwas 269 g and the average dailyfood consumption per rat in the different groups were; control:23.7 g, clofibrate2:3.1 g, PFOA: 22.7 g, PFOSA: 20.0 g. Materials (2-"'C]Mevalonic acid, [1-14 C)pyruvicacid and (2- 14 Clpyruvic acid were from New England Nuclear. [1-"C)aceticacid was obtained from Amersham (UK). Clofibratewas from Fluka (Buchs, Switzerland).Perfluorooctanoicacid was purchased from AldrichChemie (Steinheim, Germany) and Perfluoro-octane sulphonic acid was from Fluorochem (Old Glossop, UK). Other chemicals were from Sigma (St. Louis, MO, USA). Preparatiopotf heparocytcs The effectsof dietarytreatmcnt were studied in hcpatocylesisolatedfrom ratsfed the dictsfor I week and for the study of directeffectsof the compounds, hepatocyteswere isolatedfrom rats fed the standard diet.Isolationof hepatocyteswas performed by pcrfu. sion with coliagenase,according to Berry and Friend (21],with the modificationsdescribed by Segien [22). Oxidationof palmitateand conl-ersioonf labelledsubstrateintolipids Fatry acid oxidationwas measured according to Christiansenet al.[23)with 0.5 mM palmitate as substrate.Fatty acids and cholesterol*,nthesized from radioactiveprecursorswere extracted from the cell suspension after90 min incubation.The reaction was stopped by the additionof 5% saturated KOH in ethanol and the mixture was heated at 90*C for I h. The nonsaponiflable lipids were extracted with petroleum ether.The suspension was then acidified with HCI. 'ne extractswere evaporated to dryness and the radioactivelipidresidue was dissolvedin 100 pLI hexane. The lipidextractswere chromatographed with hexane/diethylether/aceticacid (80:20:1) on silica gel thin-layerplates.The spots corresponding to cholesteroland fattyacidswere identifiedby standards and were then isolatedfor measurement of radioactivity. Enzyme assaysand measurements of DNA, protein and lipids Pyruvate dehydrogenase was estimated by measur- ing the "C02 liberatedwhen hepatocyto-wsere incu- MM [1_14 bated with 5 C]pyruvate for 30 min at 37'C. Acetate thiokinasewas measured as described by Jones and Lipman (24).Liver microsomes were prepared accordingto Easom and Zammit [25),and HMG-COA reductase was measured according to Drevon et al. (26). CDPcholine: 1,2-diacylglyceroclholinephospho- transferase,EC 2.7.8.2and lysolecithinacyltransferase TABLE I Body and lil-ewreightsand lit-eUrpid contentin ratsfed differendtietsfor 7 days Cloribratcwas givenas 0.3'7(cw,,w) and PFOA and PFOSA as 0.02% in the diet.Values are givenas means S.E. There were four observations in each group. Fisher'sP-values are given;* P < 0.05;* *P < 0.01vs.controlgroup. Control Coribrate Body weight (g) Liverweight (g) Uver triacylrlyccrt(),lusmol/g li%-cr) Uvcr non-esicririecdholusicrol(Amol/'& liver) Livercholcstert)eisicr(Amol,lr,livcr) 305 2 11.3 0.4 4.1 0-5 4.4 0.4 0.38 0.08. 298 3 16.1 0.8 4.0 0.7 4.5 0.3 0.24 0.02 PFOA :lq 18.8 0.7 .1.9= 0.3 ().S 0.4 PFOSA 275 :t 6 15.9=0,7 13.8 18 7.6 0.3 0.27:t 0.03 Stock.diet limitedfcd 282 7 4.1 0.6 (EC 2.3.1.23w)as mcasurcd as dcscribcdbv Parthasar- athy ct al.[27].Acyl-CoA:cholesicrol acyltransfcrasc (ACAT) was measurcd according to Rustan et al.[281 and the synthcsis of phosphatidviscrinc, phosphatidylethanolamine and phosph atidylcholine was measured as described by Vance [29].Other enzymes were measured as described earlier[30]. DNA was measured by the mithod of Ubarca and Paigen [311 and protein .%-asestimated by the biuret method or by the method of Lowry et al.[32]. Liver lipids were extracted with chloroform/ methanol (2: 1, v/v). Triacylglycerolswas determined directlyon the dried extractwith a kitmethod (Nyco, Oslo, NorwaN,). Free and esterifiedcholesterol was determined by Nycotest kit method for cholesterol (Nyco) after separation of the extract on thin-layer chromatography. Serum cholesteroland triacylglycerols was measured directlyby the kit methods. Results Table I shows that 0.02% PFOA or PFOSA in the diet resulted in a lower body weight after 7 days of feeding as compared with the control group. ne clofibrate diet (0.3%) did not affectthe rat weight. The group with restrictedfood intake to that of the PFOSA group had about the same weight as the PFOSA group. All the compounds resulted in a 40-60% increased liverweight. Similar effecton the liverweight has been observed earlierin rats fed clof-ibrat[e9,10,33]. Changes in serum and lil:elripidsby clofibrate,PFOA and PFOSA Fig. IA shows that allthree dietssignificantlyreduced serum cholesterol.In all treatment groups, cholesterolwas significantlryeduced (to 50-70% of control) after 24 h. Dietary treatment for 2 weeks resulted in further cholesterolreductions by 70% or more. In agreement with other observations,fastingfor 2 days did not bring about significantcholesterol changes. 20 E 0 L 10 67 A x E B a3 0 2 05 u 00 D.7 ,. 1 L 9a 7 14 Days Fig. 1.Effectof cloribrateP,FOA and PFOSA on the serum lipids. (A) Cholesterol(:B) triacylglycero&l.s;fasted-0.. 0.3% cloribrate; 0.02% PFOA. a. 0.02% PFOSA. The datarepresentmean S.E. of four observations. Fig. 1B shows the effectsof clofibrate,PFOA and PFOSA on triacylglycerolisn the rat serum. None of the compounds resulted in significantchanges after 1 day of treatment. After 7 days, clofibrateand PFOA resulted in reduction to about 60% of control value and no further reduction was obtained with further treatment. PFOSA reduced triacylglyccroltso about 50% and 30% of controlvalue after 1 and 2 weeks of treatment. The significantreduction obtained by fasting for 2 days was as expected. In the rats fed stock diet but restrictedto that consumed by the PFOSA group, the serum triacylglycerolwsas 2.43 0.13 mM after 1 week of treatment (not shown) which is not significantllyower than observed in the control group (fed stock diet ad libitum). Table I shows that PFOSA increased the livertriacylglycerolscontent to more than 3-times the control value. This effect of PFOSA on triacylglycerolwsas clearlyin contrast to the effectof clofibrateand PFOA which did not affectthe content of triacylglycerolasnd clearlyindicates that PFOSA inhibitsthe excretion of triacylglycerolfsrom the liver.Both PFOA and in particularPFOSA increased the livercontent of nonesterifiedcholesterol.In contrast to this,a significant TABLEII Corr-iersioonflabellesdubstrateisnto" C-lobellcehdolesterboylhepaiocyteosfratsfeddifferednitetfsor 7days Clofibratweas givenas0.3c,(cw/lw)and PFOA and PFOSA as0.02% inthedietn.e concentratioonf[2-"Clmcvalonataend (3-"C]hydroxymethylglutaratewas 0.5mM. Incubationswith 11-14C]acetateand 12-1"Clpyruvatewere conducted with 5 mM of the labelledsubstratesand includedalsouniabelledglucose(10 m.M). Values are means S.E. There were fourobservationsin each group. F-tsher"Ps values are given p < 0.05. p < 0.01. Substrate I-" C]Acclate 12-1"C]Pyruvate 1.11,4_C]Iiydrcixy-mcihylglutaratc [2-1"C]Mcvalonatc Formation of '4C-labclledcholesterol(nmol substratecarbon -mg DNA- control 147.6 21.4 7N.2 14.6 I-,.Y 3.3 2:80 IiO cloribrate 59 13 *1 46 9.5 3.6 0.9 2160 280 PFOA 3'.6 16.5 35.2t 9.7 3.6t 1.3 1730 1-40 1-h-1) PFOSA 21.2 .-7.4 26.5 5.4 5.0 1.1 2030 310 TABLEIII Ctuff@ersiCt),fplnabellesdithstraiirssit(i fairyacidcand oxidatic(iAtfJiU-'"ClpalmttaItne.hepattw.reiiefsratvfeddifferendtiets .f(i7rdo@A-.v The concentrationsof compounds inthedietand the concentrations of [1-"Cldcetateand 12-"Clpyru---iwteire as describedinTable 11. The concentrationof [L;-" ClpaImitatewas 0.5 mM. Values are means= S.E. There were 4 obsc-%-ationisn each group. Fisher'sP valuesare given:*. P < 0.05; P < 0.01vs.controlgroup. Substrate "C-Labelledfatiyacidformation (nmol substratecarbon-MIDNA - h - 1) control cloribratePFOA PFOSA 11-14C@Acetate 25228 180@28 [2-"C]Pyruvate 22121 322=23* 90@54* 182@-55 104 28** 128 21 [U-1"C]Palmitateoxidation (nmol substrateoxidized-M&DNA control clof-ibratePFOA 502:t50 r-6 36 0 564 63 1-h - 1) PFOSA 861102 0 reduction(toapprox.50%) in esterifiecdholesterolin the treatedrats(PFOA-treated ratsnot measured) was found. Conve?-sioonf labellesdubstrateisntocholesterboyl hepatocyteosf ratsfed differendtiets Table 11shows the incorporationof labelledcarbon atoms from differentllyabelledacetate,pyruvate, mevalonate and hydroxymethylglutariacid (HMG). The major pointemerging from thistableisthat the rate of cholesterolsynthesiswas significantlryeduced from allsubstrateswhich are proximalto the HMGCoA dehydrogenase*step in alltreatmentgroups. In contrast,cholesterolsynthesisfrom mevalonate was not reduced in any of the groups. Table III shows the s%-nthesiosf fattyacids from pyruvateand acetate.Both PFOA and PFOSA treat- TABLEIV Ctitit-erovfio/nI-"'laretatiento '*'r-laheflaltetd.a%r.itiavnd icroktv hqmtor@-reisn thePFcvcnceOfkl'polipr"cl,i(cm?ilxpiindv The concentratioonf []-"Clacciatwcas i mM and the incubalitins alsoincluded 10 mM glucose.Each incubationflaskctinlained4 Mr. cellproteinand 1.5% albun3inin 3 ml Krcbs-Henselcitbicarbonate buffer.The concentration@f hypolipcmiccompounds was I mM. Values aremeans S.E.There were fourobservationsin each group. Fisher'sP valuesare given.*' P < 0.01v.s.controlgroup..- Addition None TTA Cloribriaccid PFOA PFOSA Formation of 14C-labelled cholesterol Formation of '4C-labcllce fattyacid (nmol substratcearbonmg DNA- '-h- 1) 114 1.4 57 63 65 8 0.07 0.7 11 7 313 14 7-5 1.3 243 8 611 t:20 1908 29 ment reduced lipidsynthesis(notsignificanftrom pyruvate in the PFOA group).No reductionin fattyacid synthesiswas found in hepatocytesfrom clofibratetreatedanimals.The tablefurthershows that the peroxisomal inducers,as expected,resultedin an increasedrate of paimitateoxidation(althoughPFOA did not resultin a significanitncrease)T.he experiments do not excludethe possibilityhat intracellular content of the testedcompounds might have a direct (reversiblee)ffecton fattyacidand cholesteroslynthesisinvivo.Such effectscould have escaped our detectionsincethe compounds might have been washed out during the cellisolationprocedure.To testpossible directinhibitoryeffectsof thesedrugs,th@esynthesisof fattyacidsand cholesterolin hepatocyteswith additionsof the compounds to the incubationmedium were performed(Table IV).The tableshows thatI mM of clofibriaccid,PFOA and FPOSA inhibitedthe choles- TABI F V ActiL;iotfycn--ymerselatedtothesynthesiosfcholesteraonld fairyacidsinliverofratsfed differendtietfsor 7 days The concentrationosf compounds in thedietwere as describedin Table 1.Values are givenas means S.E.There were fourobservationisn each group.* P < 0.05: P < 0.05-, P < 0.01vs.controlgroup. Enzyme Pyruvat dehydrogenase Citratesynthase ATP-citratelyase Acetate thiokinase Malate dehydrorenase(NADP) (decarhoxylating) Malate dehydrogenase Gluco-,c-&phosph2le dehydrotena.%c lsocitratdeehydrogen;L-;c HMG-COA reductusc Activit(yAmol -(mg DNA min 1) control clofibrate 0.59 0.05 0.97:t0.12 1.540.34 8.7 0.64 0.360.05 27.1 3.2 0.50 0.10 2.44 0.35 0.580.04 0.94 0.06 0.78 0.38 5.7 0.85 0.83 0.05 18.4 2.4 0.170.03 2.71:t0.56 (nmol -,gg protein -'-min -') 0.31 0.0-1 0.16 0.04 PFOA 0.42 0.08 0.80 0.05 0.75 17 * 3.5 0.5 1..44 0.05 20.7 2.6 0.21 O.D4 2.7@ 0.19 0.15 0.05 PFOSA 0.36 0.04* 1.10 29 0.24 0.06 * 5.7 0.28 0.21 0.01 19.6 3.0 0. 11 0.01 0.13 0.11:tO.ol 69 tcroI synthesisto the same extent (approx.50%,).Surprisinglyt.ctradccylthioacctaicid (TTA). another induccr of pcroxisomal 0-oxidationwith hypolipemic effcct in rats [3.41.almost completel%.inhibitedcholcsterols@,nthesisT.he pronounced inhibitoryeffectof T-TA on the fattyacid synthesisobserved by Skrede et al.[34]was also confirmed.Clofibricacid reduced the fattyacid production by about 20%. However, a direct inhibitoryeffecton the fattyacid synthesisis not a property shared by all the tested compounds. Both perfluorinatedcompounds unexpectedly'stimulatedthe rate of fattyacid synthesisstronglv.It is unlikelythat thiswas due to an inhibitionof Krebs cvcie with a concomitant increase in lipogenic precursors,since these compounds inhibitedthe cholesterolsynthesis.It seems more likelythat the compounds stimulatea rate-limitinegnzyme, e.g.,acetyl-CoA carboxylase.At lower concentrations(0.5and 0.1 mM), there were very small inhibitoryeffectsof clofibricacid, PFOA and PFOSA (data not shown). En--yme,srelatedto the synthesisof cholesteroalnd fatty acidsfrom pyrut-ateand acetate Table V shows thatpyruvate dehydrogenase, citrate synthase and acetatethiokinasewere only slightlayffected by treatment with the three compounds. ATPcitratelyase activitywas reduced to about 15% of controlvalues by PFOSA. PFOA reduced the activity of thisenzyme significantltyo 50%. The effectof the compounds on three NADPH-generating enzymes shows a remarkable pattern.All three compounds significantlyreduced the activitoyf glucose-6-phosphate dehydrogenase. PFOSA reduced the activityto 20% of cdntrol values. In contrast,isocitratedehydrogenase was unaffected by allthree compounds. The activitoyf malic enzyme was increased 2- and 3.5-foldby cloftbrate and PFOA, respectively\.Ialatedehydrogenase, which isnot specificalliynvolved in lipidsynthesis,was virtuallyunchanged by any of the compounds. The activitoyf HMG-COA reductasc.the ratc-limitinagnd regulatedstep in cholesterolsynthesis,was reduced to 50% or lessin allthree treatment groups. En--yntesrelatedto the synthesisof cliolestere-svtlerand phospholipidsiiiratsfed differendtiets Table VI shows thatacyl-CoA: cholesterolacyltransferase(ACAT), was significantldyownregulated by all three compounds. PFOSA, which had the strongest effect,reduced the activittyo one thirdof control.The downregulation of thisenzyme isin keeping with the reductionof livercholesterolester(Table I).The table furthershows that the activitieosf two enzymes important in the phophoiipid turnover,acyl-CoA:I-acylglycero-3-phosphocholineacyltransferaseand CDPcholine: 1.2-diacylglycerolcholinephosphotransferase, were not significantlaylteredin the ratsfed any of the hypolipemic drugs.The activityof phosphatidyiscrinc synthase, phosphatidylethanomaline synthase and phosphatidylcholinesynthase.were alsounaffected by any of the dietaryregimes utilized(not shown). ACtiL'itioefSen.-ymesrelatedto phospholipidsvnfhesisin thePresenceof hipolipemicdrugs Table VII shows thatat 0.5 mM clofibriaccid had littleffecton activitieosf the enzymes listedin the table.The five enzymes were all inhibitedby both perfluorinated.compounds. PFOA had a particular inhibitoryeffecton phosphatidylserinesynthase activity which was reduced to 18% of normal with 0-5 mM PFOA. PFOSA had strongestinhibitoryeffecton the activitoyf CDP-choline: 1,2-diacylglyceroclholinephosphotransferase,phosphatidyiserinesynthase and phosphatidylethanolaminesynthase,which were reduced to 14%, 13% and 28% of control,respectivelywith 0.5 mM PFOSA. At lower concentration(0.1 mM) the inhibitoryeffectsof the perfluorinatedcompounds were very moderate (data not shown). Clofibricacid had no significanetffectat 1 mM concentration. TABLE Vi Aciii-iorf%e.n--)-mreeslatetdoj)-nihesoi.s(cholesteersytlerasnd phospholipiidnslit-eorfsratsfed hypolipemidcrugsfor 7 days The concentrationisnthe dietwere as desc:-ibiendTable1.Valuesare givenas means S.E.There were four observationisn each group. P < 0.01vs.controlgroup. Enzv.me Acyl-CoA Cholesterol acyltransfera(sAeCAT) (nmol/mg proteinpermin) Ac-yl-CoA:I-acylrlyccrc-v-." phosphocht)linaecyltr;jnsferuse (nnit)nli.g.p,rt)tcipnermin) CDP-chc)lineI:.:!-diacvlglvcertil cht)line,nht)spht@transfcr;&%$. (nmt)lmig prt)tcipnermin) Control 660 51 i,.o- 1.3 lo.4--1.3 Cloribrate 427 18 12.4- 1.0 9.1 1.2 PFOA 323 39 12.7= 1.@ 7.6- 0.7 PFOSA Z.17 41 11.8 1.8 S.h 1.0 70 TABLE Vil Artit-itioefsctt--vmcrselatedio ph(;sphidipi-dn-tirhesiinsthepresmice (If ltilrA-Jlipc-ir d-gs Concentratedst)lutit()In(s)mM) of clilfibarciicd.PFOA and PFUSA in DMSO were dilutedto 0-5 m.%i finalconcentrationin a.%.%aystre.m (The controlswere added the same amount of pure DMSO.) The effectof the drugs an each of the enz@mes were testedfour times and in each of the experiments.the activitiweesre calculatedas per cent of the controlvalue.Values are givenas means @-S.E- Control CloribricPFOA acid Acyl-CoA l@-sophosphatidyl transferase 100 CDP-choline: 1.2-diacylglycerol cholinephosphotransferase 100 Phosphatidyiscrinnemthase 100 Phosphatidylethanolamine synthase 100 Phosphatidylcholinesynthase 100 90@-6 56 7 95 5 101=9 71 t 2 18 4 1013 89 -@3 475 475 PFOSA 615 140.6 131.4 282 483 The effect of the compounds on the activityof ACAT was tested only at 100 AM and with thisconcentration the ACAT activitywas unaffected by the compounds (not shown). Discussion Reduction of the steady-statelevelsof serum lipids may be visualized as the resultof downregulation of lipidsynthesisor increased clearance from plasma. The aim of this study was to evaluate the effectsof three different peroxisomal proliferatorson a series of en- z@,,mesinvolved in hepatic lipidsynthesis. The most important observations with allthree compounds was downregulation of H.MG-COA reductase, the ratelimitingenzyme of cholesterolsvnthesisand of choles- terol esterificationenzyme (ACAT). The agreement between the observed alter-ationisn the enzyme activi- ties and production of cholesteroland lipidsin intact livercells provides some support for the assumption that such enzyme measurements do reflectreal changes in metabolic activityin the intactorgan and shed light on the mechanism of the h@-polipemic effectsof the agents investigated in this study.These compounds do not show a correlated effect on the lipogenic and cholesterogenic pathways in the in vivo experiments. The in vitroexperiments indicatethat there are direct, presumably reversible,effectson these pathways. These effectsare differentthan those observed after dietary manipulation. The latterprobably represent changes in enzyme concentrations since, at least,the changes in the cholesterogenic path%%-ayare correlated with the changes in HMG-COA reductase. The pcrfluorinated compounds utilizedin this study are strong local irritantsT.he observed actions should not be regarded as mcrc unspecifictoxic effectssince. in additionto reductionof theactivitoyf some enzyme systems,thesecompounds increase the activity of other enzymes, e.g.. cnzvmes of the pcroxisomal fattv acid 6-oxidation system [18,191.However, local irritation mav explain the reduction of food intake and slower weight increase obser-@ed after feeding perfluorinatcd compounds. However, this probably contributes little to the reduction in serum triacylglyccrolss,ince no significantreduction was observed in rats with restrictedfood intake. Increased liverweight Correlates well with earlierstudies on p-eroxiSOmal prolifc-rators [35,361. Cholesterol and fairy acid sinthesis The reduction of cholesterolsvnthesis from differ- ent labelledsubstrates fitswell with the reduced activ- ity of HMG-COA reductase observed in this study. Three substrates,proximal to the reductase step,were incorporated into cholesterolat a reduced rate whereas no reduction from mevalonate was observed in any of the treatment groups. Lowering of HMG-COA reduc- tase levelsby clofibrateisin accord with other observa- tions[9,37,381.Even though the three compounds tested in thisstudy reduce the HMG-COA reductase activity. thiseffect may not be a characteristicof allperoxiso- mal inducers.MEDICA 16,another compound in this group, does not act via an effect on this reductase (12,39],but rather inhibitsthe synthesisof cholesterol at a step distalto HMG-COA reductase. Our observa- tions with clofibrateare at variance with the data obtained by Azarnoff et al.(10)who found thatcholesterol synthesis from mevalonic acid was reduced in Iiversof ratsfed clofibrate. In contrastto the inhibitoryeffectof MEDICA 16 on ATP-citrate lyase (39],clofibricacid, PFOA and PFOSA had essentiallyno directeffecton ATP-citrate lyase at 1 mM (data not shown), but the drugs downregulated the enzyme afterdietaryadministration.This effectmay contribute to reduced synthesis of choles- terolin vivo,since the enzyme isimportant in the main pathyway for cholesterolprecursor synthesis.Itisinter- esting that all three compounds reduced one of the NADPH generating enzymes (glucose-6-phosphate de- hydrogcnase), while isocitratedehydrogenase (NADP) was unchanged. Reduced capacity for NADPH generation can therefore hardly contribute to reduction in serum lipid levels. The synthesisof fattyacids from acetate or pyruvate was not reduced in hepatocytes from ratsfed clofibrate (Table 111),but was significantlryeduced in the hepato- cytes from rats fed PFOSA. The reduced rate of f-,ittv acid synthesis in the PFOA and PFOSA groups is probably not related to reduction in serum triicNI- glyccrols.In rats fed PFOSA. there'was accumulation of livertritcylglycerolTsh.e lack of reduction of fitin acid s@-nthcsisby clofibratcsuggcsis that other prt)- ccsscsin lipidmetabolism must be more essentialfor the hypolipemia.Tomarelli et al.[40]found increased synthesisof lipidsfrom acetatein ratsfed cloribriaccid and also in ratsfcd the .,cr-pyotcni hypolipcmic compound Nk*N'-14.643. S@,iiihesoifscliolestereyslicrand phospholipids Liver exportslipidsto other organs mainly as VLDL particles.In addition to apolipoproteins.triacylglycerolsand free choltstcrol,these particlesconsist mainly of cstcrifiedcholesteroland phospholipids. Hence. reduced hepatic synthesisof these components might lead to reduced transportof lipidsfrom the liver. The hepatic contentof cholcstcrylesterwas reduced both in clofibrateand PFOSA fed rats(PFOA fed rats were not tested)even where free cholesterolwas not reduced (Table 1).Similareffectwas obtained by Avignan et al.[411 with the hypolipemic drug MER-29. Reduced cholesterylester production is most likelya resultof downregulation of the ACAT activitywhich was observed in all treatment groups. A direct inhibitor.%e,ffect on the enz@-mc by the compounds is probably of lessimportance since 100 AM of the compounds did not affectthe enzyme activir(vhigherconcentrationswere not tested).Vance [2@] argued that synthesisof phospholipids may be limitingfor lipoprotein synthesis.In thisstudy, we have not observed any downregulation of enzymes involvedin the synthesisor metabolism of phospholipids.We have observed only small directeffectsof cloribricacid on these enzymes. It seems likely,however, that reduced phospholipid synthesisplaysa rolein the lipid-reducingeffectof the perfluorinatedcompounds, since these had directinhi@itory effect on several important enzymes. Parthasarathy et al.[271 suggested from theirstudies that inhibitionof phosphatidyi-cholinesynthesis,particularlyby the lysolecithinacN.t1ransferase pathway, may be related to a drug's effectivenessin decreasing serum lipids.We found that the transferasewas moderatelyinhibitedby the perfluorinatedcompounds. We also observed thatthisenzyme was virtualluynaffected by 0.5 mM (Table VII) and also by 1 mM of clofibric acid. This agrees with the data reported by Parthasarathyet al.[27).Our data do not support the hypothesis that reduced actin-itoyf lysolecothineacyltransferaseplays a central role in the hypolipemic effectof the compounds we have tested.Inhibitionof other enzymes of phospholipid svnthesismay play a role. Rediiced releaseof lipid.fTrot?iiliclit-er In earlierstudies,ithas been reported that clofi- braic reduces the releaseof lipidsfrom the livcr[13.141. A directcffccton the excretionproccssbv clofibratc was not confirmed by accumulation of livertriacylglycerolsin clufibrate-fcdn.or in PFOA-fed animals.A 71 reduced releaseof lipidsfrom liverwas probably an effectof PFOSA. however, since the compound increased livertriacylglvccroblys about 200% in spiteof reduced rate of fattyacid synthesis(Table IV). Since interferencewith t@c synthesisof cholesteryl estermay cause reduced hepatic lipidoutput [42],it may be concluded thatreductionincholesterolsynthcsis and @stcrificatiodnue to downregulation of HMGCoA reductase and ACAT together with enhanced fattyacid oxidationin the liver,are effectscaused by clofibricacid as well as by the perfluorinatedcompounds. This may reduce VLDL productionby liver which plays a centralrole after the postprandialchylomicroncmic period during which the liveristhe dominatingorgan for deliveryof lipidsto serum. In addition,the differenthypolipemic drug may act by inhibitingthe synthesisof other lipoproteincomponents such as phosphatidylcholine[27]. Acknowledgements Financial support was received from Norwegian Council on Cardiovascular Diseases; Anders Jahres Foundation for promotion of Science, Norway; and The InsulinFond, Copenhagen, Denmark. We thank Dr. Robert Horn forhelpfuldiscussionsand Mrs Mctte Ursin for skillfultechnicalassistance. References I Lazarow,P.B.(1977)Science197,580-581. 2 Reddy, J.K. Warren, J.R.,Reddy, M.K. and Lahvani,N.D. (1982)Ann. N.Y. Acad. Sci.386,81-110. 3 Berge,R.Y- and Bakke, O.M. (1981)Biochem. Pharmacol.30, 2251-2256. 4 Spydevold,0. and Bremer,J.(1989)Biochim.Biophys.Acta 1003,72-79. 5 Lundgren.B., Meijer.J. and DePierre,J.W. (1987)Eur.J. Biochem. 163,423-431. 6 Bar-Tana.J.,Rose-Kahn.G.,Frenkel,B.,Shafer,Z- and Fainaru, M. (1988)J.Upid Res.29,431-441. 7 lde.T..Oku. H. and Sugano,M. (1982)Metabolism31 (No. 10), 1065-1072. 8 SinoriC,R. (1981)Lancetii,1362. 9 Avoy. D.R.. S%%"d, E.A.and Gould,R.G. (1965)J.Upid Res.6. 369-376. 10 Azarnoff.D.L. Tucker,D.R. and Barr,G.A. (1965)Metabolism 14,959-965. II Maragoudakis.M.E. (1969)J.Biol.Chem. 244,5005-5013. 12 Bar-Tana,J..Rose-Kahn.G. and Srcbnik,M. (1985)J. Biol. Chcm. 260.8404-9410. 13 Gould. R.G., S-al7ydE,.A..Coan. BJ. and Avoy, D.R. (1966)J. AtherosclerR.eL 6.555-564. 14 Duncan. C.H.. Rest.M.M. and Despopoulos.A. (1964)Circula- tion30.7 (Abstr.). 15 Nc%tclP.J.andAustinW.. (1968J).Athero.-.cRleesr.8..827-833. 16 Barret.A.M. (1%6) Brit.J.Pharmacoi.26.363-.17). 17 Grundv. S.Ni..Ahrens.E.H. Jr..Salen G..Schreibman.P.H. and Nc.%telP'.J.(IY72) J.Lipid Res. 13.5@1-55 1. IS Ikeda.T..Aiba. K..Fukuda.K. and Tanaka.M. (I9b5)J.Biochum. 98.475-482. 72 19 Ikeda.T..Fukuda. K- Mori. I..Enomoto. M- Komai. T. and Suga. T. (1987)inPeroxL%omes inBioloo and Medicine(Fahimi, H.D. and Sics.H- ecL-@)p.p 3n4--IW. Springer Vcriar_New York. 20 Just.W.W.. Gorgas. K-.Hartl.F.-U..Heincmann. P..Salzer.M. and Schimassek.)4.(IM) Hepatology 9.S70-581. 21 Berry.M.N. and Friend.D.S. (1969)J.CellBiol.43.506-520. 2'-)Segien,P.O. (1973) F-xpt.Ccil.Res.82-391-398. 23 Christiansen,R..Borrcbaelc@B. and Bremer. J.(1976)FF-BS Utt. 62. 313-'317. 24 Jones.M.E and lipmann, F. (195S)Methods Enrymol.1. S85- 591. 25 Easom RA. and Zimmit, V-k (1984)Siochem. J.220.739-745. 26 Drevon. C.A..Weinstein.D.B. and Steinberg,D. (1980)J. Biol. Cheni- 2S5.9128-9137. 27 ParLhasarathy,S. Kritchevsk@-D.. and Baumann, WJ. (1982) Pror-Nati.Acad. Sci.USA 79. 6890-6893. 29 Rustan, A.C. Nosscn. J.0- Osmundsen, H. and Drevon, C-A. (1988)J.Biol.Chem. 263,91264132. 29 Vance. J.E (1991)J.Biol.Chcm. 266,89-97. 30 Spvdcvold,S.0, Grccnbaunx. A.L, Baquer, N.7- and McLean, P. (1978) Eur. J.Biochem. 89.3-19-339. 31 Labarca. C and Pai&en, Y- (1980) AnaL BiochenL IOZ 344-352. 32 lawny. O.H.. Rosebrough. NJ, Farr, A.L and Randall. Ri. (19:51J).Biol.Chcm. 193,265-275. 33 Adams. LL. Webb, W.W. and Fallon.HJ. (1971)J.ain. Invcsl. 50. 2339-ZM6. 34 Skrcdc.S.Narec. M., Bergmth. S.and Bremer. J.(1989)Biachim. Biophys.Acta 1005.296-302. 35 Pagct,G.E. (1963)J.Atheroscier.RCL 3,7-)9-7.46. 36 Kolde, G, Roessner,A. and Thernann, H. (1976)Virchows Archiv.B. CellPathoL(Berlin)2Z 73-97. 37 Cohen, B.], Raicht.R.F.,Shefer.S. and Mosbach, E.H. (1974). Biochim. Biophys.Acta 369, 79-95. 38 Berndt,J.,(jaumert,P- and StillJ,. (1978).Atherosclerosi3s0. 147-152. 39 Rose-Kahn, G. and Bar-Tana, J. (1985) J. Biol.Chcm. 260. 8411-8415. 40 Tomarelli,R.M, Bauman, LM. and SavinL S. (1978)Ather(>. sclerosis30, 301-311. 41 Avigm J.,SteinbergD., Vroman. H.F-,'nompson, MJ. and Mosettig,I- (1960)J.BioL Chem. 235, 3123-3126. 42 Khan, B.,Wilcox,H.G. and Heimber& K (1989)Biochem. J. 259. W7-816.