Stimulation of Adipose DifferentiationRelated Protein (ADRP) Expression in

Adipocyte Precursors byLong-Chain Fatty Acids

JUN GAO, HONG YE, AND GINETTE SERRERO*Department of Pharmaceutical Sciences, University of Maryland School of Pharmacy

and Program of Oncology, Marlene and Stewart Greenebaum Cancer Center,University of Maryland

Adipose differentiation related protein (ADRP) is a 50-kDa protein expressed inadipocytes and transcriptionally activated when adipocyte precursors differenti-ate into mature adipocytes. Recent experiments have demonstrated that ADRP isa fatty acid binding protein that specifically facilitates the uptake of long-chainfatty acids. The present investigation provides evidence that ADRP mRNA andprotein expression in preadipocytes is stimulated by fatty acids in a time- anddose-dependent fashion. ADRP mRNA expression was maximally stimulated atfatty acid concentrations of or above 10–5 M. Stimulation of ADRP expression wasobserved with the nonmetabolizable fatty acid 2-bromopalmitate and with nat-ural fatty acids. Stimulation of ADRP mRNA expression by fatty acids peakedbetween 5 and 8 hr and decreased by 24 hr. Stimulation of ADRP expression byfatty acids was completely inhibited by treatment with actinomycin D, suggestingthat fatty acid stimulates ADRP gene expression at the transcriptional level.Comparison of the effect of several fatty acids with varying carbon chain lengthsindicated that long-chain fatty acids were active in stimulating ADRP, whereasshort-chain fatty acids such as caproate and 2-bromooctanoate had no effect. Thedegree of saturation of fatty acids did not influence their ability to stimulate ADRPexpression. These studies provide new information on the regulation of ADRP andidentify a new target regulated by fatty acids during adipose differenti-ation. J. Cell. Physiol. 182:297–302, 2000. © 2000 Wiley-Liss, Inc.

The regulation of the expression of genes involved inlipid metabolism by nutrients, and particularly by di-etary fat is now well established in adipose tissue andin adipogenic cell lines (Murphy et al., 1993; Baillei etal., 1996; Bernlohr et al., 1997). Recent studies haveshown in particular that fatty acid (FA) treatment ofpreadipocytes induces expression of several genes en-coding proteins implicated in FA metabolism. Theseinclude the adipocyte lipid-binding protein ALBP, alsoknown as aP2 (Distel et al., 1992), lipoprotein lipase(Amri et al., 1996), phosphoenolpyruvate carboxyki-nase (Antras-Ferry et al., 1994), and angiotensinogen(Safonova et al., 1997). In addition, polyunsaturatedfatty acids (PUFA) have been shown to decrease theexpression of fatty acid synthase (Balke and Clarke,1990), stearoyl-CoA desaturase (Waters et al., 1997),and glucose transporter Glut4 (Tebbey et al., 1994).

Adipocyte differentiation is characterized by a coor-dinate increase in gene expression. In most cases, thisincrease can be accounted for by activation of genetranscription (Cornelius et al., 1994). Studies of theadipose differentiation program have been facilitatedby the availability of established adipogenic cell linesthat can be induced to differentiate on treatment with

a mixture of stimulants. In particular, these cell lineshave provided useful systems for the identification ofnovel gene products induced during the onset of theadipocyte differentiation program (Cornelius et al.,1994).

Differential hybridization screening of a mouse adi-pocyte cDNA library has led to the cloning of a cDNA

Contract grant sponsor: National Institutes of Health; Contractgrant number: RO1 DK 51463; Contract grant sponsor: JuvenileDiabetes Foundation International; Contract grant number:194174; Contract grant sponsor: American Heart AssociationMid-Atlantic Affiliate; Contract grant number: 9951222U.

Jun Gao is currently at the Department of Central NervousSystem and Cardiovascular Research, Schering Plough ResearchInstitute, Kenilworth, NJ 07033.

Hong Ye is currently at the Neurosciences Department, Chil-dren’s Hospital, Boston, MA 02115.

*Correspondence to: Dr. Ginette Serrero, Department of Pharma-ceutical Sciences, University of Maryland School of Pharmacy, 20N. Pine Street, Baltimore MD 21201-1180.E-mail: gserrero@rx.umaryland.edu

Received 1 July 1999; Accepted 21 September 1999

JOURNAL OF CELLULAR PHYSIOLOGY 182:297–302 (2000)

© 2000 WILEY-LISS, INC.

encoding a novel 50-kDa protein named adipose differ-entiation related protein (ADRP; Jiang et al., 1992). Itwas shown that the 1.7-kb ADRP mRNA was induced50- to 100-fold, a few hours after the initiation of theadipose differentiation program, making it an earlymarker of the process (Jiang and Serrero, 1992). Sim-ilarly, the 50-kDa ADRP protein was undetectable inundifferentiated cells and its expression increased asthe cells differentiated into adipocytes (Jiang and Ser-rero, 1992). Factors known to inhibit differentiationwould inhibit the expression of ADRP (Serrero et al.,1992; Xia and Serrero, 1999). Tissue distribution ofADRP mRNA expression in mice indicated that ADRPmRNA was expressed at a high level in adipose tissue(Jiang and Serrero, 1992). In addition, recent experi-ments have shown that ADRP was also found in manydifferent types of cells and tissues that accumulated orsynthesized lipids, although at a lower level than inadipose tissue (Brasaemle et al., 1997a). ADRP expres-sion has been found in the livers of mice treated withthe carnitine palmitoyltransferase I inhibitor etomoxir,which caused neutral lipid accumulation in the organ(Steiner et al., 1996).

These studies point out to the possible role of ADRPin the formation or stabilization of lipid droplets inadipocytes and other lipid-accumulating cells. Recentexperiments where ADRP was transiently expressed inCos-7 cells demonstrated that ADRP localized to theplasma membrane and facilitated the uptake of long-chain fatty acids by increasing initial uptake velocity(Gao and Serrero, 1999). Moreover, purified recombi-nant mouse ADRP protein was shown to bind fluores-cently labeled fatty acid in a stoichiometric manner(Serrero, Frolov, Schroeder and Gelhaar, submitted forpublication). These data demonstrated that ADRP wasa novel fatty acid binding protein distinct from the 14-to 17-kDa cytosolic fatty acid binding proteins and fromthe recently characterized transporters, fatty acidtransport protein (FATP) and scavenger protein FAT(Abumrad et al., 1992; Schaffer and Lodish, 1994).Thus, it is important to identify the factors stimulatingADRP expression. Since ADRP is induced during theonset of the differentiation program, this question wasinvestigated in adipocyte precursors. We have previ-ously shown that ADRP was transcriptionally acti-vated by cyclooxygenase inhibitors in undifferentiatedcells (Ye and Serrero, 1998). In the present study, wehave examined the effect of fatty acids on the expres-sion of ADRP mRNA and protein in 1246 adipocyteprecursors.

MATERIALS AND METHODSMaterial

Dulbecco’s modified Eagle’s medium, Ham’s F12, fe-tal bovine serum, and Trizol were purchased from LifeTechnologies (Gaithersburg, MD). Fatty acids, bovinepancreatic insulin, human plasma transferrin, bovinefetuin, and fatty acid–free bovine serum albumin werefrom Sigma Chemical Co. (St. Louis, MO). 2-Bromo-palmitate and 2-bromooctanoate were from Aldrich(Milwaukee, WI). The enhanced chemiluminescence kitwas from Amersham (Piscataway, NJ). All chemicalsused for polyacrylamide gel electrophoresis and forWestern blotting analysis as well as Zeta probe mem-brane were from Bio-Rad (Hercules, CA). Immobilon P

membrane was from Millipore (Bedford, MA). Tissueculture plasticware were from Beckton Dickinson(Franklin Lakes, NJ).

Cell culture and adipose differentiationFor the experiments, we used the mouse C3H tera-

toma-derived adipogenic cell line 1246 (Serrero andKhoo, 1982); 1246 cells were maintained in DME/F12medium (1:1 mixture of Dulbecco’s modified Eagle me-dium and Ham’s nutrient F12 supplemented with 10%fetal bovine serum [FBS]). For experiments, 1246 cellswere cultivated in defined medium corresponding toDME/F12 medium supplemented with bovine insulin(10 mg/ml), human transferrin (10 mg/ml), and bovinefetuin (250 mg/ml) as previously described (Jiang andSerrero, 1992). At day 4, when the cells reached con-fluency, increasing concentrations of fatty acids wereadded to the culture medium for different lengths oftime. Fatty acids were dissolved in ethanol at a con-centration of 100 mM and then diluted to 1 or 10 mM ina solution of 5 mg/ml fatty acid–free bovine serumalbumin (BSA) in phosphate buffer saline (PBS), pH7.0. Aliquots of the fatty acid–BSA solutions were in-cubated for 10 min in a sonicating water bath prior tobeing added to the culture medium in order to obtainthe final fatty acid concentrations indicated in the Re-sults section. Total RNA and proteins were collectedafter the treatments as described later.

RNA isolation and Northern blot analysisTotal cellular RNA was isolated from duplicate

60-mm plates of cells with Trizol. Ten micrograms ofRNA were electrophoresed on a denaturing 1.2% aga-rose gel containing 0.22 M formaldehyde in 13 MOPS(0.02 M MOPS, 5 mM NaAc, 0.5 mM EDTA). RNA wasblotted to Zeta probe nylon membrane (Bio-Rad) byovernight capillary transfer in 103 SSC (203 SSC: 3 MNaCl; 0.3 M NaCitrate, pH 7.0). The filters were bakedat 80°C under vacuum for 2 hr. Prehybridization, hy-bridization, and washing conditions of the filters havebeen described elsewhere (Ye and Serrero, 1998). Hy-bridization was performed at 42°C with approximately106 cpm of randomly primed 32P-labeled cDNA probesper milliliter of fresh hybridization solution. Afterwashing, the filters were exposed to Kodak XAR-5 filmat 270°C with intensifying screens. Ribosomal proteinL32 mRNA was detected as internal control for normal-izing RNA loading (Bowman, 1987). Quantitation ofADRP mRNA level was determined by densitometricscanning of the films after autoradiography and nor-malized to the expression of RPL32 mRNA.

Preparation of 1246 cell lysates and Westernblot analysis of ADRP expression

Conditions for cell culture and fatty acid treatment ofthe 1246 cells were carried out as described earlier. Foreach experimental condition, cells from duplicate60-mm culture dishes were washed once with PBS andlysed by adding 500 ml per dish of 13 SDS samplebuffer (62.5 mM Tris–HCl, pH 6.8; 2% SDS; 10% glyc-erol) without b-mercaptoethanol (2-Me) and bromophe-nol blue. Cell lysate was then sonicated at 40% 20 Woutput for 10 sec using a Vibra Cell sonicator (Sonics &Materials Inc. Danbury, CT), and centrifuged at10,000g for 10 min at 4°C, after which the supernatant

298 GAO ET AL.

was collected. The protein concentration of cell lysatewas measured by using a micro BCA protein assayreagent kit (Pierce, Rockford, IL). After adding 1/10 volof 103 loading dye (50% 2-Me, 1% bromophenol blue),equal amounts of protein (20 mg) were used for SDS–PAGE and for Western blot analysis, using anti-ADRPantibody to measure ADRP protein expression as pre-viously described (Ye and Serrero, 1998). Immunoreac-tive proteins on the membranes were visualized by theenhanced chemiluminescence (ECL) detection system(Amersham). Quantitation of ADRP levels was deter-mined by densitometric scanning.

All experiments were repeated at least three times.Values are expressed as mean 6 SD.

RESULTSTime-dependent stimulation of ADRP mRNA

by treatment of adipocyte precursorswith 2-bromopalmitate

Investigation of the effect of fatty acids on gene ex-pression has often used the nonmetabolizable fatty acid2-bromopalmitate (Grimaldi et al., 1992; Safonova etal., 1997). Thus, experiments described here were ini-tially carried out using this fatty acid derivative. The1246 cells were cultivated in defined medium as de-scribed in the Materials and Methods section. At day 4,the medium was replaced by fresh defined medium andthe cells were treated with 10 mM 2-bromopalmitate forincreasing lengths of time up to 24 hr. RNA was ex-tracted at the indicated time points to examine ADRPmRNA expression by Northern blot analysis (Fig. 1).RPL32 mRNA expression was measured as internalcontrol for RNA loading. A time-dependent increase ofADRP mRNA expression was observed on addition of2-bromopalmitate. The stimulation of ADRP expres-sion by 2-bromopalmitate reached a maximum at 8 hr(10.8 6 0.8–fold above control) and slightly decreasedby 24 hr. No change in ADRP mRNA expression wasobserved in control cells maintained in defined mediumwithin the same period of time (data not shown).

Dose-dependent stimulation of ADRP mRNAexpression by 2-bromopalmitate

The 1246 cells were then treated for 6 hr with in-creasing concentrations of 2-bromopalmitate (C16) be-fore extracting RNA to measure ADRP mRNA expres-sion. As shown in Figure 2A, a dose-dependent increaseof ADRP mRNA expression was observed at all doses. Amaximum increase of 12.2 6 2.1–fold above control wasobserved with a concentration of 3 3 10–5 M.

In contrast, corresponding short-chain fatty acid

2-bromooctanoate (C8) did not have any effect on theADRP mRNA expression, even at a concentration of10–4 M (Fig. 2B).

Transcriptional activation of ADRP expressionby 2-bromopalmitate

The 1246 cells were treated with 2-bromopalmiticacid (10–5 M) in the presence or absence of 5 mg/mlactinomycin D or 10 mg/ml cycloheximide. As shown inFigure 3, the increase of ADRP mRNA expression ob-served in cells treated for 6 hr with 2-bromopalmitatewas abolished by actinomycin D. In this case, ADRPexpression returned to the level observed in controlcells that were not treated with 2-bromopalmiate. Ad-dition of cycloheximide to cells treated with 2-bromo-palmitate resulted in only a slight decrease in ADRPmRNA expression These results suggest that fatty acidregulates the expression of ADRP gene mainly bymeans of transcriptional activation.

Time- and dose-dependent stimulation of ADRPmRNA and protein expression by treatment of

adipocyte precursors with oleic acidWe then investigated the effect of natural fatty acids

on ADRP mRNA expression. The cells cultivated in

Fig. 1. Time course of stimulation of ADRP mRNA expression by2-bromopalmitate. At day 4, 1246 cells cultivated in defined mediumwere treated with 10–5 M 2-bromopalmitate (2-BP) for up to 24 hr.RNA was extracted at the indicated times to measure ADRP andRPL32 mRNA expression. Fig. 2. Dose-dependent increase of ADRP expression in 1246 cells

treated with increasing concentrations of 2-bromopalmitate. A: At day4, 1246 cells cultivated in defined medium were treated for 6 hr withincreasing concentrations of 2-bromopalmitate. RNA was extracted tomeasure ADRP mRNA expression by Northern blot analysis. B: Com-parison of 2-bromopalmitate (2-BP) and 2-bromooctanoate (2-BO) onthe stimulation of ADRP mRNA expression. Cells cultivated in de-fined medium were treated for 6 hr with 10–4 M 2-bromopalmitate or2-bromooctanoate before collecting RNA for measuring ADRP andRPL32 mRNA expression.

Fig. 3. Stimulation of ADRP mRNA expression is inhibited bytreatment with actinomycin D. The 1246 cells cultivated for 4 days indefined medium were treated for 6 hr with 3 3 10–5 M 2-bromopalmi-tate (2-BP) alone or in the presence of either 5 mg/ml actinomycin D(ActD) or 10 mg/ml cycloheximide (CX). Control cells (Cont.) werecultivated in the absence of fatty acid. RNA was collected to examinethe level of ADRP mRNA expression.

299ADRP EXPRESSION BY LONG-CHAIN FATTY ACIDS

defined medium were treated at day 4 with 30 mM ofoleic acid for increasing lengths of time up to 24 hr. Atime-dependent increase of ADRP mRNA expressionwas observed on addition of oleic acid (Fig. 4A). Stim-ulation of ADRP mRNA was maximal at 5 hr (3.0 60.4–fold above control) and decreased thereafter to re-turn to the basal level of control untreated cells by 24hr. No change in ADRP mRNA expression was ob-served in control cells maintained in defined medium,containing just BSA vehicle, over the same time period.

The 1246 cells were treated for 5 hr with increasingconcentrations of oleate from 10–6 to 10–4 M followedby RNA extraction to measure ADRP mRNA expres-sion (Fig. 4B). The stimulation of ADRP mRNA expres-sion by oleate was dose dependent, starting at a con-centration of 3 3 10–6 M and reaching a 9 6 0.7–foldmaximum stimulation above control with 10–4 Moleate.

The effect of oleic acid on ADRP protein expression in1246 cells was also examined by Western blot analysiswith anti-ADRP antibody (Fig. 5). Treatment of the

cells with oleate (3 3 10–5 M) stimulated ADRP proteinexpression in a time-dependent manner. Stimulation ofADRP protein expression reached a maximum by 5 hr(15.3 6 1.9–fold above control) and remained elevatedafter that.

Influence of carbon-chain length and degree ofsaturation of fatty acids on their ability to

stimulate ADRP expressionThe experiments carried out with the 2-bromo fatty

acid derivatives indicated that long-chain 2-bromo-palmitate (C16:0) was effective in stimulating ADRPmRNA expression in contrast to the short-chain deriv-ative, 2-bromooctanoate (C8). Moreover, the experi-ments described in Figure 4 had shown that the long-chain monounsaturated fatty acid oleic acid (C18:1)also stimulated ADRP mRNA expression. In thepresent experiment, the effect of other natural fattyacids varying in the length and degree of saturation ofthe carbon chain was investigated. They were caproicacid (C6:0), palmitic acid (C16:0), linoleic acid (C18:2),and arachidonic acid (C20:4).

Time course studies were carried out. For all fattyacids examined, stimulation of ADRP mRNA expres-sion peaked at 5 hr similarly to oleic acid (data notshown). The data corresponding to the fold-stimulationof ADRP mRNA expression at 5 hr of treatment withfatty acids are summarized in Table 1. The resultsshow that long-chain fatty acids, either unsaturated orsaturated, were effective in stimulating ADRP mRNAexpression. In contrast, short-chain fatty acid caproate(C6: 0) had no effect. This confirmed the result obtainedwith 2-bromooctanoate (Fig. 2B). Northern blot analy-sis of dose-response studies is shown for palmitic acid

Fig. 4. Effect of oleate on ADRP mRNA expression. A: Time courseof oleic acid effect on ADRP mRNA expression. The 1246 cells werecultivated in defined medium as described in the Materials and Meth-ods section. At day 4, cells were treated with 3 3 10–5 M oleate for upto 24 hr. Control cells were treated with BSA vehicle only. Total RNAwas collected at the indicated times to examine mRNA expression ofADRP and of RPL32 (used as an internal control for RNA loading) byNorthern blot analysis. B: Effect of increasing concentrations of oleateon ADRP mRNA expression. The sample of 1246 cells was cultivatedin defined medium. On day 4, the cells were treated for 5 hr withincreasing concentrations of oleate. Only BSA was added in the con-trol cells. Total RNA was extracted to measure ADRP and RPL32mRNA expression by Northern blot analysis.

Fig. 5. Time-dependent stimulation of ADRP protein expression byoleate. The 1246 cells were cultivated in defined medium. On day 4,cells were exposed to 3 3 10–5 M oleate for increasing amounts of timeup to 24 hr. Whole cell lysates were collected as described in theMaterials and Methods section and 20 mg of protein were analyzed byWestern blot using anti-ADRP antibody. Signals were visualized byenhanced chemiluminescence.

TABLE 1. Effect of fatty acids with various carbon chain lengthson ADRP mRNA expression

Fatty acid1ADRP mRNA

fold-stimulation2

Caproate (C6:0) 1Palmitate (C16:0) 2.3 6 0.5Oleate (C18:1) 3.25 6 1.15Linoleate (C18:2) 4.25 6 1.2Arachidonate (C20:4) 3.96 6 0.45

1Fatty acids were added to the 1246 cells cultivated in defined medium for 5 h ata concentration of 3 3 1025 M except for caproate, which was added at 1024 M.Cells cultivated in defined medium in the absence of fatty acid were used ascontrols.2RNA was extracted to measure ADRP mRNA and RPL32 mRNA expression.Quantitation of ADRP mRNA level was determined by densitometric scanning ofthe films after autoradiography and normalized to the expression of RPL32mRNA. Values were expressed as fold-stimulation above the level of ADRPmRNA in control untreated cells. Data are expressed as mean 6 SD.

300 GAO ET AL.

and for arachidonic acid (Fig. 6). Palmitic acid andarachidonic acid stimulated ADRP mRNA expressionwith about the same dose-response pattern as oleicacid.

These experiments indicate that long-chain fatty ac-ids, whether they are saturated, mono-, or polyunsat-urated, but not short-chain fatty acids, are effective instimulating ADRP expression.

DISCUSSIONAdipose differentiation related protein (ADRP) is a

novel 50-kDa protein induced very rapidly and at avery high level when adipocyte precursors differentiateinto adipocytes (Jiang and Serrero, 1992). The datapresented here show that the expression of ADRPmRNA and protein in adipocyte precursors is stimu-lated by fatty acids in a dose- and time-dependentmanner. The stimulatory effect of fatty acids on ADRPexpression was observed with natural as well as non-metabolizable fatty acids such as 2-bromopalmitate.Dose-dependent studies with various fatty acidsshowed that ADRP mRNA expression was stimulatedwithin a range of concentrations from 10–5 to 10–4 M, inagreement with what was observed for other adipose-specific genes also activated by fatty acids in adipocyteprecursors (Amri et al., 1991, 1996; Grimaldi et al.,1992). The fold-stimulation observed for ADRP mRNAexpression by long-chain fatty acids was similar to theones reported for ALBP mRNA expression, which var-ied from threefold in Ob17 cells with 5 3 10–5 M palmi-tate (Amri et al., 1994) to 20-fold in 3T3-F442A with10–4 M oleate (Distel et al., 1992). The time coursestudy presented here showed that the stimulation ofADRP mRNA expression by fatty acid was transient,peaked at 5 to 6 hr and returned to the basal level by 24hr. Treatment of the cells with the nonmetabolizablefatty acid 2-bromopalmitate did not result in such atransient pattern of stimulation of ADRP mRNA ex-pression. This would suggest that fatty acids are mosteffective to stimulate ADRP mRNA expression beforethey are metabolized. This property has also been re-ported for several adipose-specific genes such as aP2phosphenolpyruvate carboxykinase and angiotensino-gen, where nonmetabolizable fatty acids were found tobe more effective than their metabolizable counter-parts (Grimaldi et al., 1992; Antras-Ferry et al., 1994;Safonova et al., 1997).

It is interesting to note that the stimulation of ADRP

protein by fatty acids was more pronounced and sus-tained than that of mRNA expression, since the stim-ulation by fatty acid above basal level was maintainedup to 24 hr. The difference could be due to the fact thatADRP protein may be stabilized by fatty acids. Such astabilization of proteins by lipid has been reported forperilipin, although in this case, stabilization requiredthe fatty acid to be metabolized into neutral lipids to beeffective, since 2-bromopalmitate had no effect (Bra-saemle et al., 1997b). Whether fatty acids also regulateADRP at the posttranslational level remains to be elu-cidated.

The comparison of the effect of saturated and unsat-urated (mono- or polyunsaturated) fatty acids with var-ious carbon lengths was carried out. Palmitic, oleic,linoleic, and arachidonic acids stimulated ADRPmRNA expression in a time- and dose-dependent man-ner. Stimulation of ADRP expression was also observedwith lauric acid (C12:0; data not shown). In contrast,caproate acid (C6:0) and 2-bromooctanoate (C8:0) wereunable to stimulate ADRP mRNA expression, evenwhen added at 10–4 M. These data indicate that thesimulation of ADRP expression was observed withlong-chain fatty acids, regardless of their degree ofsaturation.

The function of ADRP as a fatty acid binding proteinhas recently been elucidated using cells expressingADRP (Gao and Serrero, 1999) and purified recombi-nant ADRP protein (Serrero et al., submitted for pub-lication). It was shown by transient transfection intoCos-7 cells that ADRP increased uptake of fatty acidsby increasing the initial uptake velocity (Gao and Ser-rero, 1999). Comparison of the uptake of fatty acidswith various carbon-chain lengths and degree of satu-ration indicated that ADRP specifically mediated theuptake of long-chain fatty acids, regardless of theirdegree of saturation (Gao and Serrero, 1999). In con-trast, uptake of short-chain fatty acids did not takeplace. These results and the ones obtained here areinteresting, as they suggest that the fatty acids thatinduce ADRP expression are the ones whose uptake isstimulated by the protein once it is induced.

Among the various fatty acids examined, it is note-worthy that free arachidonic acid stimulated ADRPexpression. However, we have shown previously thatonce metabolized to cyclooxygenase products such asPGF2a, inhibition of adipose differentiation and ofADRP expression was observed (Serrero et al.,1992a,b). This would suggest that arachidonic acid andits metabolites may have multiple and distinct func-tions related to differentiation and expression of earlymarkers such as ADRP. This is possible by assumingthat arachidonic acid may be acting on different effec-tor systems. In support of this possibility, reports haveshown that arachidonic acid and other polyunsatu-rated fatty acids (PUFA) inhibited the expression oflipogenic enzymes by a prostanoid-mediated pathway(Mater et al., 1998), whereas they inhibited Glut4 ex-pression in adipocytes by affecting transcription andprotein turnover (Tebbey et al., 1994).

Fatty acids have been shown to activate nuclear per-oxisome proliferator activated receptors PPARs (Ton-tonoz et al., 1994; Kliewer et al., 1997; Krey et al.,1997). PPAR response elements (PPRE) have beenidentified in a number of adipocyte genes that, in turn,

Fig. 6. Effect of various saturated and unsaturated fatty acids onADRP mRNA expression. At day 4, the 1246 cells were treated for 5 hrwith increasing concentrations of either palmitate (C16:0) or arachid-onate (C20:4). Cells cultivated in defined medium in the absence offatty acid were used as controls (C). Cells cultivated in defined me-dium and treated with BSA only (BSA) were also used as controls.RNA was extracted to measure ADRP mRNA expression as describedin the Materials and Methods section.

301ADRP EXPRESSION BY LONG-CHAIN FATTY ACIDS

are transcriptionally activated by fatty acids on activa-tion of PPARg (Tontonoz et al., 1994). So far, analysisof the ADRP gene (Eisinger and Serrero, 1993) hasfailed to identify typical PPRE on ADRP proximal pro-moter region, thereby raising the question of whetherthe effect of fatty acids on ADRP is mediated withPPAR activation. A similar observation has been maderecently for angiotensinogen gene that is stimulated byfatty acids during adipocyte differentiation of Ob1771cells, but does not present PPRE in its promoter region(Safonova et al., 1997). It has been postulated in thiscase that activation of angiotensinogen gene by fattyacids could be indirect by activation of PPARg, which inturn would modulate the activity of proteins able tobind to other cis-acting elements to be characterized inthe promoter. Recent report in the literature has indi-cated that polyunsaturated fatty acids and PPAR mayregulate lipogenesis and peroxisomal gene expressionby distinct mechanisms (Clarke and Jump, 1997).Whether such a possibility exists for ADRP remains tobe examined.

The results presented in this study provide new in-formation about ADRP regulation in adipocyte precur-sors and also identify a new member of the gene prod-ucts regulated by fatty acids in adipocyte precursors.These data also suggest that modulation of ADRP byfatty acids may also occur in vivo during changes ofnutritional status of the animals. These possibilitiesare presently explored in our laboratory.

ACKNOWLEDGMENTSThe authors thank Drs. Jun Hayashi and Koichiro

Tanaka for helpful discussion.

LITERATURE CITEDAmri E, Bertrand B, Ailhaud G, Grimaldi P. 1991. Regulation of

adipose cell differentiation: fatty acids are inducers of aP2 geneexpression. J Lipid Res 32:1449–1456.

Amri EZ, Teboul L, Vannier C, Grimaldi PA, Ailhaud G. 1996. Fattyacids regulate the expression of lipoprotein lipase gene and activityin preadipose and adipose cells. Biochem J 314:541–546.

Antras-Ferry J, Le Bigot G, Robin GP, Robin PD, Forest C. 1994.Stimulation of phosphoenolpyruvate carboxykinase gene expres-sion by fatty acids. Biochem Biophys Res Commun 203:385–391.

Baillei RA, Jump DB, Clarke SD. 1996. Specific effects of polyunsat-urated fatty acids on gene expression. Curr Opin Lipidol 7:53–55.

Bernlohr DA, Coe N R, Simpson MA, Hertzel AV. 1997. Regulation ofgene expression in adipose cells by polyunsaturated fatty acids. AdvExp Med Biol 422:145–56.

Blake WL, Clarke SD. 1990. Suppression of rat hepatic fatty acidsynthase and S14 gene transcription by dietary polyunsaturatedfat. J Nutr 120:1727–1729.

Bowman LH. 1987. The synthesis of ribosomal protein S16 and L32 isnot autogenously regulated during mouse myoblast differentiation.Mol Cell Biol 7: 4464–4471.

Brasaemle DL, Barber T, Wolins NE, Serrero G, Blanchette-MackieEJ, Londos C. 1997a. The adipose differentiation related protein isa widely distributed lipid droplet-associated protein: adipocyteADRP expression is temporally regulated. J Lipid Res 38:2249–2263.

Brasaemle DL, Barber T, Kimmel AR, Londos C. 1997b. Post-trans-lational regulation of perilipin expression: stabilization by storedintracellular neutral lipids. J Biol Chem 272:9378–9387.

Clarke SD, Jump D. 1997. Polyunsaturated fatty acids regulate lipo-

genesis and peroxisomal gene expression by independent mecha-nisms. Prostaglandins Leukot Essent Fatty Acids 57:65–69.

Cornelius P, McDougald OA, Lane MD. 1994. Regulation of adipocytedevelopment. Annu Rev Nutr 14:99–129.

Distel R, Robinson GS, Spiegelman BM. 1992. Fatty acid regulation ofgene expression: transcriptional and post-transcriptional mecha-nisms. J Biol Chem 267:5937–5941.

Eisinger DP, Serrero G. 1993. Structure of adipose differentiationrelated protein (ADRP) gene. Genomics 16:638–644.

Gao J, Serrero G. 1999. Adipose differentiation related protein(ADRP) expressed in transfected Cos-7 cells selectively stimulateslong chain fatty acid uptake. J Biol Chem 274:16825–16830.

Grimaldi PA, Knobel SM, Whitesell RR, Abumrad NA. 1992. Induc-tion of aP2 gene expression by non-metabolizable long-chain fattyacids. Proc Natl Acad Sci USA 89:10930–10934.

Jiang H, Serrero G. 1992. Isolation and characterization of a fulllength cDNA coding for a novel adipose differentiation related pro-tein (ADRP). Proc Natl Acad Sci USA 89:7856–7860.

Jiang H, Harris SE, Serrero G. 1992. Molecular cloning of a differen-tiation-related mRNA in the adipogenic cell line 1246. Cell GrowthDiffer 3:21–30.

Kliewer S,. Sundeseth SS, Jones SA, Brown PJ, Wisely GB, Koble CS,Devchand P, Wahli W, Wilson TM, Lenhard JM, Lehmann JM.1997. Fatty acids and eicosanoids regulate gene expression throughdirect interactions with peroxisome proliferator activated receptorsalpha and gamma. Proc Natl Acad Sci USA 94:4318–4323.

Krey G, Braissant O, L’Horset F, Kalkhoven F, Perroud M, ParkerMG, Wahli W. 1997. Fatty acids, eicosanoids and hypolipidemicagents identified as ligands of peroxisome proliferator activatedreceptors by coactivator-dependent receptor ligand assay. Mol En-docrinol 11:779–791.

Murphy MC, Zampelas AS, Puddicombe, M, Furlonger, NP MorganLM, Williams CM. 1993. Adipose tissue site specificity of lipoproteinlipase mRNA expression in rats fed diets containing different fattyacid compositions. Biochem Soc Trans 21:145S.

Mater MK, Pan D, Bergen WG, Jump DB. 1998. Arachidonic acidinhibits lipogenic gene expression in 3T3–L1 adipocytes through aprostanoid pathway. J Lipid Res 39:1327–1334.

Safonova, I, Aubert J, Negrel R, Ailhaud G. 1997. Regulation by fattyacids of angiotensinogen gene expression in preadipose cells. Bio-chem J 322:235–239.

Serrero G, Khoo, JC. 1982. An in vitro model to study adipose differ-entiation in serum-free medium. Anal Biochem 120:351–359.

Serrero G, Lepak NM, Goodrich SP.1992a. Prostaglandin F2a inhibitsthe differentiation of adipocyte precursors in primary culture. Bio-chem Biophys Res Commun 183:438–442.

Serrero G, Lepak NM. and Goodrich SP. 1992b. Paracrine regulationof adipose differentiation by arachidonate metabolites: PGF2a in-hibits early and late markers of differentiation in the adipogeniccell line 1246. Endocrinology 131:2545–2551.

Sessler AM, Ntambi JM. 1998. Polyunsaturated fatty acid regulationof gene expression. J Nutr 128:923–926.

Steiner S, Wahl D, Mangold BLK, Robison R, Raymackers L, MeheusL, Anderson NL, Cordier A. 1996. Induction of the adipose differ-entiation-related protein in liver of etomoxir-treated rats. BiochemBiophys Res Commun 218:777–782.

Tebbey PW, McGowan KM, Stephens JM, Buttke TM, Pekala PH.1994. Arachidonic acid down-regulates the insulin-dependent glu-cose transporter gene (GLUT4) in 3T3–L1 adipocytes by inhibitingtranscription and enhancing mRNA turnover. J Biol Chem 269:639–644.

Tontonoz P, Hu E, Spiegelman BM. 1994. Stimulation of adipogenesisin fibroblasts by PPAR gamma 2, a lipid- activated transcriptionfactor. Cell 79:1147–1156.

Waters KM, Miller CW, Ntambi JM. 1997. Localization of a polyun-saturated fatty acid response region in stearoyl-CoA desaturasegene 1. Biochim Biophys Acta 1349:33–42.

Xia X, Serrero G. 1999. Inhibition of adipose differentiation by PI-3-kinase inhibitors. J Cell Physiol 178:9–16.

Ye H, Serrero G. 1998. Stimulation of adipose differentiation relatedprotein (ADRP) expression by ibuprofen and indomethacin in adi-pocyte precursors and in adipocytes. Biochem J 330:803–809.

302 GAO ET AL.