Induction of Peroxisomal Fatty Acid ,&Oxidation and Liver Fatty Acid ...

Download PDF
3 downloads 124 Views 7MB Size Report
Comment
Raja M. Kaikaus$$, William K. Chanlf, Nina LysenkoS, Rupsa Ray$, Paul R. Ortiz de ..... perature for 1 h, wells were washed again with PBS/Tween buffer,.
Vol. 268, No. 13, Issue of May 5, pp. 9593-9603,1933 Printed in U.S.A.

OF BIOLOGICAL CHEMISTRY THEJOURNAL

6 1993 by The American Society for Biochemistry and Molecular Biology, Inc

Induction of Peroxisomal Fatty Acid ,&Oxidation and Liver Fatty Acid-binding Protein by Peroxisome Proliferators MEDIATION VIA T H E CYTOCHROME P-4501VA1 w-HYDROXYLASE PATHWAY* (Received for publication, August 13, 1992,and in revised form, January 11, 1993)

Raja M. Kaikaus$$, William K. Chanlf, Nina LysenkoS, Rupsa Ray$,Paul R. Ortiz deMontellanoll$, and NathanM. Bass$$1 I From the Departmentsof SMedicine and lfPhrmaceutica1 Chemistry and the §Liver Center, University of California, S a n Francis& California 94143-0538

Both the enzymes of peroxisomal fatty acid @-oxida- acid metabolism have been recognized relatively recently (2, tion and the liver fatty acid-binding protein (L-FABP) 3). Thecontribution of peroxisomes to overall fatty acid are induced in the liver by peroxisome proliferators, oxidation in hepatocytes in the basal state remains a subject such as clofibrate (CF), as well as high fat diets. One of debate (4, 5 ) , although under conditionsof impairment (6) proposed mechanism for this induction is that it rep- or overload of mitochondrial @-oxidation (7) increased fatty resents an adaptive response to altered intracellular acid oxidation via the microsomal and peroxisomal pathways fatty acid fluxes, mediated by dicarboxylicfatty acids occurs. Diets containing very long-chain fatty acids induce formed via the cytochrome P-450IVA1 w-oxidation the enzymes of both peroxisomal and microsomal oxidation pathway. The studies presented in this paper were (8), while diets rich in long-chain fatty acids have been shown designed to investigate the role of the products of P- to induce both peroxisomal @-oxidation(9) and the 14-kDa 450IVA1 w-oxidation in the regulationof peroxisomal &oxidation and L-FABP. In primary hepatocyte cul- cytosolic liverfatty acid-binding protein (L-FABP)’ (10). This tures exposed to CF,the increase in P-450IVA1 activ- abundant protein, which constitutes 2-5%of total hepatic for long-chain fatty ity preceded the induction of peroxisomal &oxidation cytosolic protein, exhibits marked affinity and L-FABP. The CF-mediated increases in peroxiso- acids, and has been considered to play a significant role in mal@-oxidation and L-FABP,butnot P-450IVA1, fatty acid transportand metabolism (11). The most procould be significantly inhibited pretranslationally by nounced induction of the enzymes of peroxisomal and microconcurrent exposureof cultured hepatocytes to inacti- somal oxidation (la), aswell as of L-FABP (13, 14), occurs in vators of cytochromes P-450, such as 1-aminobenzo- response to a group of diverse chemicals collectively termed peroxisome proliferators(12), whichincludehypolipidemic triazoleand10-undecynoic acid. Hexadecanedioic acid, a 16-carbon dicarboxylicfatty acid, that is poorly drugs such as clofibrate (CF),phthalate esterplasticizers, and @metabolized inhepatocytes,inducedperoxisomal herbicides. oxidation and L-FABP, but not P-450IVA1, via a preThe mechanism(s) by which these chemicallydissimilar translational mechanism that was not inhibited by 1- compounds effect similar biological responses ispoorly underaminobenzotriazole. Long-chain monocarboxylic acids stood. A solublereceptor, termed theperoxisome proliferatorwere without such inducing effect. In further studies,activated receptor (PPAR),belonging to the nuclear hormone non-&oxidizable dicarboxylic acidanalogs were found receptor superfamily, has recently been isolated, and has been to display greater potency as inducers of peroxisomal shown to confer peroxisome proliferator responsiveness to &oxidation when compared to hexadecanedioic acid. COS cells intransactivationassays (15, 16).Therecent The inducing effects of the dicarboxylic acid analogs identification of response elements in the 5”upstream region were also independent of the P-450 w-oxidation pathreg- of the peroxisomal fatty acyl-CoA oxidase (FACO) gene furway. The results of these studies suggest that the regulation of the enzymes ulation of peroxisomal&oxidationenzymes and L- ther confirmsa role forPPAR in the FABP is mediated, to a significant extent, by poorly of peroxisomal @-oxidation (17, 18). However, none of the metabolized long-chain dicarboxylic acids formed via peroxisome proliferators tested thus far have been shown to bind to PPAR, and the“endogenous” ligands of this receptor the P-450IVA1 pathway. remain unknown (16). The cellular response to peroxisome proliferators is characterized by an increase in the transcriptionof several genes Although mitochondrialfatty acid @-oxidationhas been that are involved in fatty acid metabolism (16). It has been extensively studied and its regulation well characterized (I), suggested that the peroxisome proliferator response reprefor a fundamental the presence of peroxisomal @-oxidation andmicrosomal cy- sents an adaptation initiated to compensate tochrome P-450-mediated w-hydroxylation pathways of fatty disruption of normal fatty acid metabolism effected by these diverse agents (19). Indeed, peroxisome proliferators and their * This work was supported by Research Grants DK-02056, DK- CoA derivatives have been shown to inhibit carnitine palmi-

32926, DK-13328-21, GM-25515, and Liver Core Center Grant DK26743 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1) To whom correspondence should be addressed Box 0538, HSW 1120,U.C.S.F., 3rd. & Parnassus Aves., San Francisco, CA94143. Tel.: 415-476-6424;Fax: 415-476-0659.

’

The abbreviations used are: L-FABP, liver fatty acid-binding protein; ABT, 1-aminobenzotriazole; BCMTO, l,8-bis(carboxymethy1thio)octane; BSA, bovineserum albumin; CF, clofibrate; EHS, Engelbreth-Holm-Swarm; FACO, fatty acyl-CoA oxidase; Me2S0, dimethyl sulfoxide; P-450IVA1, cytochrome P-450 4A1 (CYP4A1); PPAR, peroxisome proliferator-activated receptor; TMDD, 2,2,ll,ll,tetramethyldodecanedioicacid; UDYA, 10-undecynoicacid.

9593

9594

Regulation of Peroxisomal 0-Oxidation and

L-FABP

(ii) Lactic dehydrogenase measurements were performed by the toyltransferase I activity in mitochondria (20). It has been hypothesized that this inhibition of carnitine palmitoyltrans- method of Cabaud and Wr6blewski (32) using a kitprovided by Sigma ferase I results in fatty acyl accumulation (substrate overload) (No. 500) with pyruvate as asubstrate. To compensate for the differences in density of plating of hepatocytes on the EHS-coated and thesubsequent formation, via the microsomal P-450IVA1 plates, lactic dehydrogenase concentration in the medium was exw-oxidation pathway, of long-chain dicarboxylic fatty acids pressed as nanomole of pyruvate consumed per mg of cellular protein/ which then serve as the proximate inducers of peroxisomal min. (iii) The ability of hepatocytes to synthesize new protein was also enzymes (19, 21), possibly by acting as ligands for PPAR. However, direct evidence in support of such a mechanism is used as a measure of cell toxicity (33). At the end of a treatment period, 7 pCi of [3H]leucine (5 Ci/mmol) was added to the medium. lacking. After 30 min, the medium was discarded, plates washed twice with The studies presented in thispaper were designed to inves- ice-cold phosphate-buffered saline, pH 7.4 (PBS), and the cellular tigate the role of the P-450IVA1 w-hydroxylase pathway in pellet harvested. Hepatocytes were first extracted with 10% trichlomediating the induction of peroxisomal @-oxidationand L- roacetic acid for 2 min, and the supernatant(precursor pool) fraction FABP. It has recently been shown that primary cultures of saved. The pellet was re-extracted, twice with 10% trichloroacetic rat hepatocytes maintained on a laminin- and collagen type acid and once with 1:l diethyl ether/ethanol at 4 "C, and dissolved in 2% Na2C03, 0.1 N NaOH (Lowry reagent). Incorporation of [3H] IV-rich substratum, prepared by salt extraction of the Engel- leucine into newly synthesized protein was expressed as the ratio of breth-Holm-Swarm (EHS) mouse sarcoma, maintaintheir radioactivity obtained inthe trichloroacetic acid-precipitable fraction differentiated state for prolonged periods (22) and are an to theradioactivity in the precursor pool fraction. excellent model system for investigating the regulation of Treatment of Cell Cultures-Treatments were started 18 h after peroxisomal @-oxidation andL-FABP (23), as well as of hepatocytes were plated onEHS-coated dishes. Medium waschanged cytochromes P-450 (24). The present studies were, therefore, daily, except when otherwise specified. CF, ABT, and UDYA were in Me2S0, the final concentration of which was 1% (v/v) conducted utilizing the EHS-supported primary hepatocyte dissolved in the medium. Control cultures were exposed to 1%Me2S0 only. culture model. Hepatocytes were exposed to various treatments for periods ranging from 6 h to 5 days. A final CF concentration of 0.3 mM, previously shown in cultured hepatocytes to produce maximal induction of peroxisomal 0-oxidation and L-FABP (23), was routinely employed Materials in these studies. Fatty acids, including dicarboxylic fatty acids and EHS matrix was prepared as described by Bissell et al. (22). 1- the fatty acid analogs TMDD and BCMTO, were dissolved in medium Aminobenzotriazole (ABT) was synthesized according to the method containing 1%BSA, with control cells being exposed only to 1%BSA. P-45OIVAl Activity Assay-P-45OIVAl activity was measured as of Campbell and Rees (25). 10-Undecynoic acid (UDYA) wassynthesized as described (26). 2,2,11,1l-Tetramethyldodecanedioic acid the rate of w-hydroxylation of [l-'4C]lauric acid (34). Hepatocytes (TMDD) and 1,8-bis(carboxymethylthio)octane(BCMTO) were syn- were pooled from two 10-cm dishes, and after sonication of the cell thesized as described by Chan (27).Clofibric acid, dimethyl sulfoxide suspension, the assay was carried out at 37 "C in a reaction volume (Me2SO), bovine serum albumin (BSA), palmitic acid, oleic acid, of 1 ml containing 2 mg of total protein, 600 pM [l-"C]lauric acid, dodecanedioic acid, hexadecanedioic acid, and palmitoyl-CoA were and 1 mM NADPH in 50 mM potassium phosphate buffer, pH 7.4. from Sigma. Flo-Scint I1 scintillator was from Packard Instrument The reaction was stopped after 20 min with 0.4 ml of 10% (v/v) Co. Dispase was purchased from Collaborative Research, Inc. [3H] sulfuric acid. Unlabeled standards of lauric acid (10 pmol) and 12Leucine (5 Ci/mmol) and [1-"Clpalmitoyl-CoA (59.9 mCi/mmol) hydroxylauric acid (10 pmol) were added, followed byfour extractions were obtained from Du Pont-New England Nuclear. [cx-~*P]~CTPwith diethyl ether. The combined ether extracts, containing the 12(3000 Ci/mol), [Y-~'P]ATP(6000 Ci/mmol), and [1-"Cllauric acid hydroxylauric acid product, were methylated in2 ml of diazomethane in diethyl ether at room temperature for 45 min. After overnight (50 mCi/mmol) were from Amersham Corp. Plasmid pMJ125, which is aplasmid pBR322 construct containing evaporation of the solvent at room temperature, the precipitate was dissolved in 0.5 ml of 80% methanol/water. High performance liquid a partial cDNA of rat FACO (28) was kindly provided by Dr. T. Osumi (Shinshu University, Shinshu, Japan). Plasmid pJG418 (a chromatographic analysis was performed using a Du Pont Zorbax pBR322 construct containing the full-length cDNA for rat L-FABP ODS column (4.6 X 250 mm) eluted with 80% methanol/water at 0.5 Ref. 29) wasprovided by Dr. Jeffrey Gordon (Washington University, ml/min. The effluent was monitored using a Beckman 166 UV St. Louis, MO). Dr. T. Enoch (Harvard University, Cambridge, MA) detector set at215 nm and a Flo-One Beta radioactive Flow Detector I1 at 1.5 ml/min, to provided the rat 7-actincDNA, and ratalbumin cDNA was obtained with a 2.5-ml flow cell running with Flo-Scint quantify the amount of w-hydroxylated [l-'4C]lauric acid formed. from Dr. John Taylor (Gladstone Foundation, University of CaliforPeroxisomal Fatty Acid &Oxidation Assay-This was measured as nia, San Francisco). Toquantitate P-4501VA1 mRNA, a 39-mer cyanide-insensitive palmitoyl-CoA oxidation in hepatocyte sonicates, oligonucleotide, complementary to the P-45OIVA1 mRNA and spanas described by Hertz etal. (35). Acid-solubleoxidation products were ning bases 611-649 (5'- ATTTCCATCCACCTGAACACTGCCA- determined as described by Mannaerts et al. (4). TTGTGGCTGAAGGC) of the P-450IVA1 cDNA sequence (301, was Liver Fatty Acid-binding Protein Assay-L-FABP was determined synthesized at theBiomolecular Resource Center, University of Calin the 105,000 X g cytosolic fraction of sonicated hepatocytes by a ifornia, San Francisco. sensitive noncompetitive, direct enzyme-linked immunosorbent assay developed according to the method of Paulussen et al. (36). Briefly, Methods enzyme-linked immunosorbent assay plates were coated with diluPrimary Hepatocyte Cultures-Hepatocytes were isolated by col- tions of 1-8 ng of pure L-FABP (standards) or hepatocyte cytosolic lagenase perfusion of the livers of 250-g male Sprague-Dawley rats, proteinin 0.1 ml of 0.1 M Na2C03,pH 8.2. After an overnight and a highly purified hepatocyte population obtained by centrifugal incubation at 4 "C, wellswere washed with PBS, 0.1% Tween 20 elutriation (22). Cells were plated on either 3.5- or 10-cm plastic buffer, and 0.1 ml of 1:3000 dilution of a specific polyclonal rabbit dishes layered with EHS gel in modified serum-supplemented medium anti-rat L-FABP antibody (23) was added to each well. After 1 h, 199 (22, 23). The cell density was 1.5 X lo6 for the 3.5-cm plates. wells were washed, and 0.1 ml of goat anti-rabbit IgG horseradish Hepatocytes were harvested by digesting the matrix with Dispase, a peroxidase conjugate (Bio-Rad No. 170-6505), diluted 1 to 2000 in mixed protease solution, followed by centrifugation at 800 X g to PBS/Tween, was added to each well. After incubation at room temobtain the cellular pellet. Hepatocytes were then resuspended and perature for 1 h, wells were washed again with PBS/Tween buffer, sonicated as previously described (23). Proteinwas determined by the and 0.1 ml of peroxidase substrate (Bio-Rad No. 172-1064) wasadded. Oxalic acid (2%) was added to stop the reaction after an hour, and method of Lowry et al. (31). Assessment of Hepatocyte Viability and Cytotoxicity-(i) An aliquot absorbance was read at 415 nm using a DynatechRminireader (Model of hepatocytes was diluted with an equal volume of 0.4% trypan blue MR 300). Northern Hybridization of mRNA-Hepatocyteswere harvested solution immediately after harvest, and thepercentage of cells stained with trypan blue was determined by counting on a hemocytometer. from a 10-cm culture plate and total cellular RNA was extracted by The ability of hepatocytes to exclude trypan blue was taken as a the method of Chomczynski and Sacchi (37). Total cellular RNA (1030 Kg) was fractionated on a1%agarose, 2.2 M formaldehyde gel, and measure of viability. EXPERIMENTALPROCEDURES

Regulation of Peroxisomal @-Oxidation L-FABP and blotted on nylon membranes (HybondTMN, Amersham) by capillary transfer. Membranes were baked at 80 "C for 2 h, and prehybridized a t 50"C for 4h in 50% formamide, 0.1% SDS,5 X Denhardt's solution, 750 mM NaCI, 5 mM EDTA, 50 mM NaH2P04,pH 7.7, and 200 pg/ml denatured salmon sperm DNA. The linearized cDNAs for FACO, L-FABP, and albumin were labeled with [w3'P]dCTP by the random primer method (38),and hybridizations with the RNA blots were carried out overnight at 50 "C. The hybridization solution was similar to the prehybridization solution except that it contained 1 X Denhardt's solution. Membranes were washed for 10 min X 2 in 2% SSC, 0.1% SDS at room temperature, and for 30 min each in 0.2% SSC, 0.1% SDS at 50 "C, and 0.1% SSC (1 X SSC: 0.15 M Nacl, 0.15 M sodium citrate, pH 7.0), 0.1% SDS at 55 "C. In the case of the [oc-32P]CTP-labeledy-actin cRNA probe, hybridizations were conducted at 50 "C, and the final wash was at 60 "C. The P-450IVA1 oligonucleotide probe was end-labeled with [y-32P]ATP(6000 Ci/ mmol) (39), and hybridization was performed at 42 "C in 6 X SSC, 5 X Denhardt's solution, 100 pg/ml denatured salmon sperm DNA, 0.05% SDS, and 0.05% sodium pyrophosphate. Membranes were washed once for 5 min at room temperature in 10 X SSC and then for 30 min at 65 "C in 6 X SSC, 0.05% sodium pyrophosphate. Signal densities for each mRNA were quantified by densitometric analysis of the autoradiograms. Normalization for input RNA were done by rehybridizing the blots with either the albumin cDNA or the y-actin cRNA probe (23). Prior to reprobing, membranes were stripped by washing for 1-2 h at 65-75 "C in 0.005 M Tris-HC1, pH 8.0, 0.002 M Na,EDTA, and 0.1 X Denhardt's solution. Complete removal of probes was confirmed autoradiographically. Metabolism of Dicarboxylic Fatty Acids by Isolated HepatocytesHepatocytes were isolated by collagenase perfusion of rat liver, and 2 x lo5 cells suspended in Hank's glucosemedium with 2% BSA. Hexadecanedioic, or palmitic, acid was added to a concentration of 0.3 mM (containing 0.1 mCi ofeither 1,16-"C-labeled hexadecanedioic or [l-"C]palmitic acid), and incubated for 5, 10, and 15 min. Total lipids from 1ml of the incubate were then extracted (40).Radioactivities in the fatty acid and in the esterified fractions were determined by thin layer chromatography, and water-soluble oxidation products were measured in the upper phase of the Folch extraction. The dimethyl ester of [1,16-"C]hexadecanedioic acid was synthesized by boiling the free acid with BFs/methanol and subsequently purifying the diester by thin layer chromatography (41). Successful synthesis of radiolabeled dimethyl hexadecanedioic acid was confirmed by mass spectrometry. Metabolism by isolated hepatocytes of the radiolabeled dimethyl ester of hexadecanedioic acid was then determined as described above. Statistical Analysis-Data arepresented as means _t S.D. The significance of the differences between the means of treatment groups and controls was determined by Student's t test.

9595

n

6

12

2 41 2 0

72

HOURS

FIG. 1. Time course of CF-mediated induction of P450IVA1 activity, peroxisomal &oxidation, and L-FABP in primary hepatocyte cultures. Hepatocytes cultured on EHS gel were exposed to 0.3 mM clofibrate, and harvested at the designated time points. P-450IVA1 activity, peroxisomal @-oxidation,and LFABP were determined as described under "Methods." Values represent three pooled culture dishes. Results are presented relative to P-450IVA1 control (untreated) hepatocyte values (controls = 1).0, activity; a,peroxisomal @-oxidation;a,L-FABP.

T o investigatethe role of P-450IVA1 intheinduction of peroxisomal @-oxidation and L-FABP, we next determined the effect of inhibition of cytochrome P-450 on CF induction of peroxisomal @-oxidation and L-FABP. 1-ABT, a suicide inhibitor of cytochrome P-450 (42, 43), has been previously shown in vivo to profoundly inactivate lauricacid w-hydroxylase activity in control and CF-treated rat livers (44), and to be relatively nontoxic after in vivo administration (45). In initial experiments, the effect of different concentrations of ABT on the CF-mediated induction of P-450IVA1 activity, peroxisomal @-oxidation, and L-FABP was determined. Hepatocytes were exposed for 5 days to 0.3 mM CF and ABT in concentrationsranging from 0.2 to 10 mM. At 5days, P 4501VA1 activity in control cells (1%MezSO only) was undetectable, even after pooling hepatocytes from three 10-cm culture dishes. CF markedlyinduced P-450IVA1 activity (Fig. 2 4 ) . Profound inhibition of this induction was obtained with ABT at concentrations aslow as 0.5 mM, with >95% inhibition of P-450IVA1 activity at ABT concentrations of 5 and 10 mM (Fig. 2 A ) . The induction of peroxisomal @-oxidation was also significantly suppressed,with -55% inhibition by 5 RESULTS mM ABT and-75% inhibition by 10 mM ABT (Fig. 2B). The T i m e Course of Induction of P-45OIVAl Activity, Peroxiso- CF-mediated increase in L-FABPwas completely suppressed malp-Oxidation,and L-FABP-In initial experiments, the at 5 days by ABTconcentrations of 1 mM and higher time course of induction of P-450IVA1 activity, peroxisomal (Fig. 2C). p-oxidation, and L-FABP abundance was determined in cell Since in these initial experiments, P-450IVA1 activity in culture (Fig. 1).After 6 h exposure to 0.3 mM CF, no change control hepatocytes diminished to undetectable levels by the in P-450IVA1 activity, peroxisomal p-oxidation, or L-FABP fifth day, all further experiments were conducted using hewas observed. At 12h, P-450IVA1 activitywas increased -1.7 patocytes harvested after 3 days of treatment. The effect of X control values, while peroxisomal @-oxidationwas only 1.1 ABT (10 mM) on the inductionby CF of P-450IVA1 activity, X control values. L-FABP abundance was unchanged at 12 h peroxisomal @-oxidation, and L-FABP abundance is shown (Fig. 1).By 24 h, P-450IVA1 activity was increased 4 times in Fig. 3. By 3 days, P-450IVA1 activity was induced 14- to above control values, and rose further to32 times the control 20-fold by CF compared to controls, and this induction was values by 72 h. Peroxisomal@-oxidation wasmodestly in- completely inhibited by 10 mM ABT ( p < 0.05; Fig. 3 A ) . creased at 24 h (-1.7 x control), but was induced 11.5- and While theinduction of peroxisomal @-oxidation was also 15-fold above controls at 72 and 120 h,respectively. L-FABP significantly inhibited ( p < 0.005; Fig. 3 B ) , this inhibition abundance, in keepingwith its slow rate of turnover (14),was was not as complete as that of P-450IVA1 activity. Although unchanged after the first 24 h of exposure to CF. By 72 and L-FABP induction by CF was consistently inhibited by ABT 120 h,L-FABPhadincreasedto 1.5 and 2.2 X control, after 3 days of exposure, this reduction did not achieve statisrespectively. Thus,theincreasein P-450IVA1 activityin tical significance. response to CF occurred earlier and toa greater degree than In order to determine whether the effect of ABT in supthe increase in peroxisomal p-oxidation or L-FABP abunpressing the CF induction of peroxisomal @-oxidationwas due dance. to a direct inhibition of the enzymes of the peroxisomal pEffect of 1-Aminobenzotriazoleon CF Induction of P- oxidation pathway, ABT (1-20 mM) was added to a homoge45OIVAl Activity, Peroxisomal @-Oxidation, and L-FABPnate prepared from the liver of a rat fed a diet containing

Regulation of Peroxisomal ,&Oxidation and L-FABP

9596 ~

A.

1.0

ABT [mM]

FIG. 2. Effect of ABT concentration on induction by CF of P-45OIVA1 activity, peroxisomal &oxidation, and L-FABP. CONT ABT CF CF+ABT Hepatocytes were exposed to 0.3 mM CF and the indicated concentrations of ABT for 5 days, and P-450IVA1 activity( A ) ,peroxisomal FIG. 3. Effect of ABT on induction by CF of P-450IVA1 @-oxidation( B ) ,and L-FABP (C) abundance was determined.Values activity, peroxisomal &oxidation, and L-FABP. Cultured herepresent three pooled dishes. No P-450IVA1 activitywas detectable patocytes were exposed to 0.3 mM clofibrate, t o 10 mM ABT, or to in control hepatocytes at 5 days. Shaded areas in panels B and C both. Cells were harvested after 3 days, and P-450IVA1 activity( A ) , represent the respective control(untreated cells) means S.D. 0,P- peroxisomal@-oxidation ( B ) , andL-FABP (C) weremeasuredas 450IVA1 activity;O, peroxisomal @-oxidation; B, L-FABP abundance. described under “Methods.” 0, P-450IVA1 activity; B, peroxisomal 0-oxidation; a, L-FABP. Values represent mean S.D. forfive on different days with primary hepatocyte 0.5% CF for 9 days. In concentrations up to 10 mM, ABT had separate experiments done no effect upon peroxisomal @-oxidation(Fig. 4), whereas P - cultures from different animals. *, p < 0.05; **, p < 0.005, compared with controls. Cont, control. 450IVA1 activity was rapidly andprofoundly inhibited under these conditions (data not shown). Peroxisomal @-oxidation 10, 1 was slightly decreased at 20 mM ABT. In another series of experiments, hepatocytes in culture were exposed to 0.3 mM CF for 3 days, after which 10 mM ABT was added. After 4 h of exposure to ABT, there was marked suppression of P 450IVA1 activity, but no change in peroxisomal @-oxidation (Fig. 5). These results excluded a direct inhibitory effect of ABT on peroxisomal @-oxidation enzymes as an explanation for the inhibition of their activity. Effect of ABT on CF Inductionof P-45OZVA1, FACO, and 0 10 20 L-FABP rnRNA-Total cellularRNA was extractedfrom ABT [mM] cultured hepatocytes treated for 3 days with CF, ABT, or both FIG. 4. Effect of ABT on peroxisomal @-oxidationin liver CF and ABT. After Northern blotting, membranes were sehomogenates from a CF-treated rat. A 60-day-old Sprague-Dawquentially hybridized with the probesfor P-450IVA1, FACO, ley rat was fed a standard laboratory diet containing0.5% (w/w) CF and L-FABP, asdescribed under “Methods.”As shown inFig. for 9 days. Peroxisomal 8-oxidation determined inthe liver homogeto the values 6, CF treatment led to a marked induction of P-450IVA1, nate from this animalwasinduced8-foldcompared previously obtained in homogenates from untreated rats fed only the FACO, and L-FABP mRNA expression. ABT had minimal effectontheCF-mediatedincreasein P-450IVA1 mRNA standard diet (data not shown). ABT in concentration ranging from 0.5 to 20 mM was added to the liver homogenate from the CF-treated abundance, but markedly inhibited the induction of FACO rat, and after a 15-minincubation,peroxisomal@-oxidation was and L-FABP mRNA. This inhibitionwas observed at both 5 determined as described under “Methods.” and 10mM ABT. Fig. 7 shows the resultsof the densitometric analyses of signal densities on autoradiograms from three reduced by ABT to-25% of the values seen with CFalone. separateexperiments.Inprimaryhepatocytecultures,no Toxicity of ABT inHepatocyte Cultures-In order toexclude signal for P-450IVA1, FACO, or L-FABP mRNAwas detect- toxicity to hepatocytesas a cause for the inhibitory effects of able in control hepatocytes. Hence, changes in the respective ABT, hepatocyte viability and function were systematically mRNAs could not be represented relative to controls, and areevaluated. There were no differences in viability of cultured shown relative to the effect noted with CF treatment alone. hepatocytes as determinedby their ability to exclude trypan The increases in P-450IVA1 mRNA abundance by CF were blue. At the end of 3 days of treatment, 75-85% of control, essentiallyunchanged by ABT,whereastheCF-mediated CF-, or CF and ABT-treated hepatocyteswere viable by this increases in FACO and L-FABP mRNA abundances were criterion. However, in view of the lack of sensitivity of the

*

*

Regulation ofL-FABP Peroxisomal and@-Oxidation

CF

CF+ABT

FIG. 5. Effect of short-term exposure to ABT on P-450IVA1 activity and peroxisomal &oxidation preinduced by clofibrate. Hepatocytes were exposed to 0.3 mM CF (10 culture dishes) or to 1%Me2S0 (controls). After 3 days, 10 mM ABT was added to five of the CF-containing culturedishes. Hepatocytes were harvested 4 h after ABT was added, and P-450IVA1 activity and peroxisomal @-oxidationquantified as described under "Methods." Results for P450IVA1 were obtained from three pooled dishes, whereas values for peroxisomal @-oxidationrepresent means derived from two separate dishes. Results are presented relative to controls where controls = 1. 0, P-450IVA1 activity; M,peroxisomal @-oxidation.

A.

f

9597

"."" P-4501VA1

FACO

L-FABP

FIG.7. Effect of ABT on the CF induction of P-450IVA1, FACO, and L-FABP mRNAs, determined by densitometric evaluation of autoradiograms. Signals on autoradiograms from three different experiments were analyzed by an optical densitometer. Due to the absence of detectable signals in controls, mRNA abundances after exposure to both CF and ABT are presented relative to CF. The mean induction by CF of P-450IVA1, FACO, and L-FABP mRNAs is represented by the value of 1 (dashed line), and the bars represent the mRNA abundances after treatment with both CF and ABT relative to CF alone. Values represent means f S.D. of three peroxisomal @-oxiseparate experiments. 0, P-450IVA1 activity; dation; B,L-FABP.

.,

P-4501VA1

B. 2mM

FIG. 6. Analysis of P-450IVA1, FACO, and L-FABP mRNA by Northern hybridization. Total cellular RNA was extracted from hepatocytes after 3 days exposure to CF(0.3 mM), ABT (5 mM), to both CF and ABT, or to 1%Me2S0. After Northern blotting, a 32P-end-labeled39-mer synthetic oligonucleotide was used to probe for P-450IVA1 mRNA. After stripping the P-450IVA1 oligonucleotide, the blot was hybridized with a 32P-labeledcDNA for FACO. Subsequently, the FACOcDNA probe was stripped, and mRNA abundances for L-FABP and albumin were quantified using 32Plabeled full-length cDNAs for L-FABP and albumin. Albumin mRNA expression, which is stably maintainedin hepatocytes grown on EHS matrix and is unaffected by CF treatment (data notshown), was used to control for inequalities of input RNA. CONT, control.

5mM l O m M

lOmM

FIG. 8. Lactic dehydrogenase release into the culture medium after 3 days exposure to ABT. Hepatocytes were exposed to (i) ABT at concentrations ranging from 2 to 10 mM, (ii) CF (0.3 mM) alone, and (iii) CF (0.3 mM) with ABT (10 mM). Control (CONT) cells were exposed to vehicle (Me2SO) only. After 3 days, lactic dehydrogenase ( L D H ) release into the medium was determined as described under "Methods." Values are means & S.D., n = 3.

as well as the ratioof radioactivity incorporated into trichloroacetic acid-precipitable cellular fraction to theradioactivity in the trichloroacetic acid-soluble (precursor pool) fraction. ABT in concentrations from2 to 10mM did not significantly affect incorporation of [3H]leucine into newly synthesized protein (trichloroacetic acid-precipitable fraction). When results were normalized to the precursor[3H]leucine pool, ABT at 5 mM significantly increased [3H]leucine incorporation into newly synthesized protein (28% above controls, p < 0.01). Effect of IO-Undecynoic Acid on Peroxisomal /3-Oxiclation trypan blueexclusion method,toxicity of ABT was also and L-FABP-Since ABT isa global inhibitor of cytochromes assessed by (i) measuring lactic dehydrogenase activity reP-450, it is conceivable that a product of a P-450 enzyme leased by hepatocytes into the culture medium; and (ii) deter- other than P-450IVA1 is the mediator of peroxisome prolifmining the effect of ABT on [3H]leucine incorporation into erator effects, or that a metabolite of CF formed via the Pnewly synthesized protein by cultured hepatocytes (33). As 450 pathway is the active form of CF. Further experiments shown in Fig. 8, exposure of hepatocytes to ABT in concen- were therefore conducted using UDYA, a n 11-carbon fatty trations up to 10 mM did not result in any significant increase acid with a terminal acetylenic bond, which specifically and in release of lactic dehydrogenase into the culture medium. irreversibly inactivates P-450IVA1 by alkylating the protein Similarly, CF alone or with 10 mM ABT was nontoxic to moiety of the enzyme (26), in contrast to ABT which forms hepatocytes as determinedby lactic dehydrogenaserelease. an adduct with the heme group of P-450IVA1 (43). UDYA The resultsof [3H]leucine incorporation intonewly synthe- has previously been shown to inactivate P-450IVA1 in vivo, sized protein are shown in Table I. Data are presented both in both CF-treated and control animals (44). as the total radioactivity incorporated into cellular protein, In initial experiments, hepatocyteswere exposed for 3 days

Regulation of Peroxisomal ,&Oxidation and

9598

TABLEI Effect of ABT on protein synthesisin cultured hepatocytes Hepatocytes were exposed to different concentrations of ABT (2, 5, and 10 mM) in 1%Me2S0 for 3 days. At the end of this period, 7 pCi of [3H]leucinewereadded to the medium. Hepatocytes were harvested after 30 min and lysed with 10%trichloroacetic acid. After centrifugation of the lysate, the trichloroacetic acid-soluble radioactivity in the supernatant was taken to represent the precursor pool; t3H]leucine incorporationinto protein was determined in the pellet as describedunder “Experimental Procedures.”Valuesrepresent mean f S.D. for determinations in three culture dishes. [3H]Leucine Ratio of [3H]leucine incorporated into incorporated into newly newly synthesized synthesized protein to protein radioact’vityprecursor pool radioactivity

dpm X lO-’/dish

*

Control 800 f 90 11 0.6 661 f 50 11 f 4 ABT (2 mM) 934 +- 96” 10.8 f 0.6 ABT (5 mM) ABT (10 mM) 1100 f 380 13.8 f 5.4 ‘ p < 0.01 (compared with controls).

71.8 f 4.1

64.6 f 22.4 92.1 f 5.1“ 89.6 f 12.7

L-FABP

F

y-ACTIN

1

-

P

-FACO

c

L-FABP

-

FIG. 10. Effect of UDYA on induction by CF of FACO and L-FABP mRNA abundance. Total cellular RNA was extracted from hepatocytes after 3 days exposure to CF (0.3 mM), UDYA (1 mM), toboth CF and UDYA,or to 1% Me2S0. AfterNorthern a 32P-labeledcDNA for FACO. blotting, the blots were hybridized with Subsequently, the FACO cDNA probe was stripped, and mRNA abundances for L-FABP andy-actin were quantified using 3ZP-labeled full-length cRNAs for L-FABPand y-actin. The y-actin RNA abundance was used to control for variations in input RNA. CONT, control.

8

T 6

*

* T

T

t-

2

8

4

X

2

0

FIG. 9. Effect of UDYA on induction by CF of peroxisomal &oxidation and L-FABP. Cultured hepatocytes were exposed to 0.3 mM CF, 1 mM UDYA, or to both. Controls were exposed to 1% Me2S0 only. Cells were harvested after 3 days, and peroxisomal 0oxidation and L-FABP were measured as described under “Methods.” W, peroxisomal @-oxidation;B,L-FABP. Values represent mean f S.D. for four separate experimentsdoneondifferentdayswith primary hepatocyte cultures from different animals.*, p < 0.005; **, FIG. 11. Effect of mono- and dicarboxylic fatty acids on p < 0.05, compared with controls. peroxisomal @-oxidationand L-FABP in primary hepatocyte cultures. Cultured hepatocytes were exposed to palmitic acid, oleic to CF(0.3 mM), UDYA, or both CF and UDYA. In a concen- acid,dodecanedioicacid (C-12 DCA), hexadecanedioicacid (C-16 tration of 1 mM, UDYA was found to be nontoxic to hepato- DCA), or to hexadecanedioic acid and ABT (10 mM). All fatty acids in a concentration of 0.8 mM. Medium also contained 1%BSA. cytes, and completely blocked the CF induction of P-450IVA1 were Values represent mean f S.D. for nine separate experiments (except activity (data not shown). Similar to the results with ABT, for C-16 DCA+ABT and ABT experiments where n = 3) done on UDYA also markedly (by -75%) inhibited the CF induction different days with primary hepatocyte cultures from different aniof peroxisomal @-oxidation, and completely suppressed the mals. Values are given as means f S.D. relative to controls, where increaseinL-FABPabundance (Fig. 9). Theseeffects of control = 1. M, peroxisomal @oxidation;B,L-FABP. *, significantly UDYA were mediated via a pretranslational mechanism, as different from controls ( p < 0.005). demonstrated by the marked inhibition of the CF-mediated increases in FACO and L-FABP mRNA(Fig. 10). acid (C-16 dicarboxylic acid) and dodecanedioic acid ((2-12 Thus, inactivation of P-450IVA1 by two inhibitors with dicarboxylic acid), as well as to palmitic andoleic acids bound different chemical structures, mechanisms of action, as well to 1% BSA. As shown in Fig. 11, hexadecanedioic acid inas specificities, demonstrated that an intact P-450IVA1 path- creased peroxisomal @-oxidation by 2-fold ( p < 0.005), but way is necessaryfor the peroxisome proliferator-mediated didnotaffectL-FABPabundance (Fig. 11). By contrast, induction of peroxisomal FACO and L-FABP. palmitic, oleic, and dodecanedioic acids did not significantly Effect of Dicarboxylic Fatty Acids on Peroxisomal @-Oxida- affect eitherperoxisomal @-oxidationor L-FABP (Fig. 11). tion and L-FABP-In order to further test the hypothesis In Northern hybridization experiments,20 pg of total celthat long-chain dicarboxylic fatty acids, formed via the P- lular RNA was probed with a 32P-labeled oligonucleotide for 450IVA1 pathway, are the proximate mediators of the effects P-450IVA1, and cDNAs forFACO and L-FABP (Fig. 12). As of CF, cultured hepatocytes were exposed to medium- and shown in Fig. 12A, nomRNA signalfor P-450IVA1 was long-chain dicarboxylic fatty acids, as well as to long-chain detected either in control or fatty acid-treated cells. Faint monocarboxylic fatty acids, and their effects onperoxisomal signalsfor FACO were detectableincontrols, oleic acid@-oxidation and L-FABP abundancewere determined. Cells treated, or dodecanedioic acid-treated cells, whereas hepatowere exposed for 3 days to 0.8 mM each of hexadecanedioic cytes exposed to hexadecanedioic acid consistently showed

Regulation of Peroxisomal P-Oxidation A. Oleic Cont

C-12

and L-FABP

9599

s"

C-16

DCADCA acid

P

n n n m cP-450

IVA1

.. . cFACO

-'

u

-L-FABP

+Albumin

FIG.12. Effect of fatty acids onP-450IVA1, FACO, and LFABP mRNAs. Hepatocytes were exposed for 3 days to oleic acid, dodecanedioicacid ( C - I 2 DCA), or to hexadecanedioic acid (C-16 DCA). Controls (CONT) wereexposed to 1%BSA only. A, after extraction, total cellular RNA (IO pg) was fractionated and blotted on nylon membranes. Northern hybridizations were performed with a 32P-labeled P-450IVA1 oligonucleotide probe, as described under "Methods." After stripping, the blotswere reprobed with 32P-labeled cDNA probe for FACO. Individual lanes represent RNA derived from separate experiments. Equal input RNA in thegels for these experiments was confirmed by ethidium bromide staining. B, total cellular RNA (20 pg) was fractionated, transferred to nylon membranes, and probed with 32P-labeled cDNAs for L-FABP (top panel, autoradiogram exposed for 1 day) and albumin(bottom panel,autoradiogram exposed for 3 days).

increased FACO mRNA abundance. The increase in FACO mRNA, however, was not aspronounced as thatproduced by CF. As shown in Fig. 12B, L-FABP mRNA was increased by both dodecanedioic acid and hexadecanedioic acid. The albumin mRNA signal was also greater in the lanes containing RNA from dodecanedioic acid- and hexadecanedioic acidtreated cells, pointing to a greater input RNA in these lanes. The increase in L-FABP mRNA compared to controls, determined densitometrically, after normalizing for albumin mRNA was only 2-fold in the dodecanedioic acid-treated and 10-fold in the hexadecanedioic acid-treated cells. No increase in L-FABP mRNA was found in oleic acid-treated cells. LFABP mRNA therefore showed an absolute increase only in cells exposed to hexadecanedioic acid. If long-chain dicarboxylic fatty acids are indeed the proximate effectors of the induction of peroxisomal @-oxidation and L-FABP, then theirinducing effects should be independent of the P-450IVA1 pathway and not be inhibited by ABT. In a separateseries of experiments, hepatocytes were exposed for 3 days to eitherhexadecanedioic acid (0.8 mM) alone or to both hexadecanedioic acid (0.8 mM) and ABT (10 mM), and both peroxisomal @-oxidationas well as FACO mRNA abundance were determined (Figs. 11 and 13). Hexadecanedioic acid increased peroxisomal @-oxidation-2-fold and L-FABP -1.5-fold,while the presence of ABT did not affect these inductions. Indeed, concurrent treatment with ABT resulted in a consistent, although nonsignificant, increase in the hexadecanedioic acid-mediated increase in FACO mRNA, when normalized for y-actin mRNA (Fig. 14). Metabolism of HexadecanedioicAcid by Isolated Hepatocytes-Although hexadecanedioic acid showed consistent induction of peroxisomal @-oxidation, and ofFACO and LFABP mRNAs, the magnitude of the respective increases was modest compared with the effects of peroxisome proliferators.

I

.

-

FACO

FIG.13. Effect of ABT on hexadecanedioic acid induction of FACO mRNA. Hepatocytes were exposed for 3 days to ABT (10 mM), hexadecanedioic acid (0.8 mM) (C-16 DCA), or to both hexadecanedioic acid and ABT. Controls (CONT)were exposed to 1%BSA only. After extraction, total cellular RNA (20 pg) was fractionated and transferred to nylon membranes. Northern hybridizations were then performed with a 32P-labeled cDNA probefor FACO (top panel, autoradiogram exposed for 2 days). The blot was then stripped and reprobed with 32P-laheled cRNA for y-actin (bottom panel,autoradiof three separate ogram exposed for1 day). This blot is representative experiments.

0