Pathophysiology 11 (2005) 201–208

ReducedLPAexpression after peroxisome proliferator-activated receptoralpha (PPAR�) activation inLPA-YAC transgenic mice

Paivi A. Teivainen-Laedrea,b,∗, Knut A. Eliassenc, Marit Slettena,b,Adrian J. Smithd, Kare Berga,b

a Institute of Medical Genetics, University of Oslo, P.O. Box 1036, Blindern, NO-0315 Oslo, Norwayb Department of Medical Genetics, Ullev˚al University Hospital, Oslo, Norway

c Department of Basic Sciences and Aquatic Medicine, Norwegian School of Veterinary Science, Oslo, Norwayd Laboratory Animal Unit, Norwegian School of Veterinary Science, Oslo, Norway

Received 12 October 2004; received in revised form 19 December 2004; accepted 20 December 2004

Abstract

Background: Apolipoprotein(a) (apo(a)), which is part of the atherogenic lipoprotein Lp(a), shares structural homology with plasminogen( rm ofc by whichL eb transgenicft PARTt ,3-x sampledp llected ands atfi levelsi sm rt tion topa©

K

1

ttg

s-

of

on-ismalvivo.r the

0d

plg). Genes coding for plasminogen (PLG) and apo(a) (LPA) are linked and situated 40 kb apart in the telomeric region of the long ahromosome 6.LPA is naturally expressed only in primates and hedgehogs. Thus, access to knowledge regarding the mechanismPAexpression is regulated is limited due to shortage of appropriate animal models. However, mice transgenic for the humanLPAgene haveen produced. Lp(a) levels in man are genetically determined and not altered significantly by dietary changes. In contrast, mice

or LPA-yeast artificial chromosome (LPA-YAC) have markedly reduced apo(a) levels after maintenance on a high-fat diet.LPA-YAC carrieshe 40 kbLPA–PLG intergenic region, which includes a putative binding site for peroxisome proliferator-activated receptor alpha (P�).herefore, we examined if fibrates, which exert their effect via PPAR�, could alterLPAexpression in transgenic mice.Methods: Two LPA

ransgenic mouse lines with or without theLPA–PLG intergenic region we fed either PPAR� agonist fenofibrate (FF) or 4-chloro-6-(2ylidino)-2-pyrimidinylthioacetic acid (WY 14643) containing diets for 3 weeks. For the study of serum apo(a) levels, blood wererior the experiment and when the animals were sacrificed. For the study of gene expression pattern pieces of livers were coubmerged in RNAlater buffer and stored at−70◦C until analysis by quantitative PCR.Results and conclusions:The results showed thbrates reduceLPA expression inLPA-YAC transgenic mice, but have no impact on hepatic apo(a) mRNA or serum apo(a) proteinn LPA-cDNA transgenic mice, which lack theLPA–PLG intergenic region. This suggests that the effect of fibrates onLPA expression i

ediated upstream of theLPA gene. However, on the basis of current data it is not possible to conclude that PPAR� is the primary factohat repressesLPA expression inLPA-YAC transgenic mice. Negative correlation between FXR and apo(a) mRNA levels, in addiutative FXR DNA binding sequence inLPA–PLG intergenic region, suggest that it is equally likely that reduced expression ofLPAcould besecondary consequence of PPAR� activation on other genes, such as FXR.2004 Elsevier Ireland Ltd. All rights reserved.

eywords:LPA; Apo(a); Transgenic mice; Fibrates

. Introduction

Lipoprotein(a) (Lp(a)) consists of a low-density lipopro-ein (LDL) particle covalently linked through a disulfide bondo apolipoprotein(a) (apo(a))[1]. Apo(a) is an atherogeniclycoprotein with several kringle (K) domains that are ho-

∗ Corresponding author. Tel.: +47 22 11 98 60; fax: +47 22 11 98 99.E-mail address:p.a.teivainen@medisin.uio.no (P.A. Teivainen-Laedre).

mologous to plasminogen (plg) KIV and KV[2]. The genecoding for plg (PLG) and apo(a) (LPA) are linked and situated 40 kb apart in the telomeric region of the long armchromosome 6 (band 6q26–27)[3].LPA is naturally expressed only in man, Old World m

keys, great apes and hedgehogs[4,5] and its expressionmainly confined to the liver. The lack of appropriate animodels has been a barrier to the study of apo(a)/Lp(a) inHowever, the recent development of mice transgenic fo

928-4680/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.pathophys.2004.12.002

202 P.A. Teivainen-Laedre et al. / Pathophysiology 11 (2005) 201–208

humanLPAgene[6,7]has made it possible to examine effectsof xenobiotics onLPAexpression and may lead to increasingknowledge of regulation ofLPA transcription in vivo.

Lp(a) levels in man do not appear to be affected signifi-cantly by dietary changes[8]. In comparison, high-fat feedingreducesLPAexpression inLPA-yeast artificial chromosome(LPA-YAC) transgenic mice, which carry the 40 kbLPA–PLGintergenic region[9–12].

In rodents, unlike in humans, bile acids synthesizedfrom fat are closely regulated by cholesterol 7�-hydroxylase(CYP7A1), which prevents overaccumulation of bile acidsand cholesterol[13,14]. Studies with receptor-selective ag-onists have shown that liver X receptor� (LXR�) and far-nesoid X receptor (FXR) as retinod X receptor (RXR) het-erodimeric partners regulate CYP7A1 levels. Whereas di-etary cholesterol stimulates CYP7A1 transcription via activa-tion of LXR� [15], bile acids repress its transcription throughFXR-mediated induction of short heterodimer partner (SHP)protein [16]. In this way, rodents are capable of excretingexcess dietary cholesterol as bile acids. Therefore, our pre-vious observation that reducedLPAexpression is associatedwith significantly increased cholesterol CYP7A1 expressionin high-cholesterol fedLPA-YAC transgenic mice, but not inhigh-cholesterol/high-fat fed mice[17], most likely reflect al-terations in the volume and/or composition of bile acid pool.

Fibrates are a group of hypolipidemic agents, which exertt r al-p erw crip-t lism[ ithhe e site2 Rb nte

ed-iaw na -o wee d4 Y1 sw

2

2

in-t ea nalL -

senting the humanLPAgene (LPA-cDNA) fused to the mousetransferrin promoter in a hybrid genetic background of strainC57BL/6 crossed with strain SJL were a gift from Dr. RichardLawn, CV Therapeutics, Inc., Palo Alto, CA, USA[7]. Non-transgenic FVB mice were purchased from Harlan (Bicester,England). For experiments in Oslo, mice were bred in theLaboratory Animal Unit at Norwegian School for VeterinaryScience, which is accredited by the Association for Assess-ment and Accreditation of Laboratory Animal Care and UseInternational (Brussels, Belgium). The transgenicity of theanimals was tested by PCR and RIA analyses, as describedpreviously[27].

Mice were housed in a room with 12-h light:12-h darkcycle and 55% relative humidity at 21◦C, and kept onRMI(E)SQC mouse diet (Special Diet Services, Witham, Es-sex, England) until the start of the experiment. Sentinel an-imals were used to run a FELASA-style health-monitoringscheme and mice were clinically healthy.

The experiments were approved by the Norwegian AnimalResearch Authority and were performed in accordance withthe Norwegian Gene Technology Act of 1994.

2.2. Feeding experiment

Twelve wild type (WT) female mice, 27LPA-YAC trans-g ew var-i oftY o(a)[ e inp , wec k-oldL ea .T -76,I iets 5%( maA ce-t withw wast WY1

iceh wered hreew dmin-i m( nep flu-a s twop ax-i vicald

heir effect via peroxisome proliferator-activated receptoha (PPAR�) [18]. Activated PPAR� binds as a heterodimith RXR to its response element and modulates trans

ion of genes that regulate most aspects of lipid catabo19]. Furthermore, PPAR� is highly expressed in tissues wigh fatty acid oxidation, such as liver tissue[19]. A 180 bpnhancer region, located at the Dnase I hypersensitiv8 kb upstream of theLPAgene, contains a putative PPA�inding site[20]. However, fibrates do not exert significaffect on plasma Lp(a) levels in primates[21–24].

SinceLPA expression is reduced in response to fat feng in LPA-YAC transgenic mice and PPAR� regulates bilecid synthesis and transport, at least in rodents[25,26], itas of interest to study if PPAR� activation could result ilteredLPAexpression inLPA transgenic mice with or withut theLPA–PLG intergenic region. In the present study,xamined the effect of PPAR� ligands fenofibrate (FF) an-chloro-6-(2,3-xylidino)-2-pyrimidinylthioacetic acid (W4643) onLPAexpression in twoLPAtransgenic mouse lineith or without theLPA–PLG intergenic region.

. Materials and methods

.1. Mice

FVB mice transgenic for a 270 kb YAC carrying theactLPAgene and the 40 kbLPA–PLGintergenic region wer

gift from Dr. Edward Rubin, Lawrence Berkeley Natioaboratory, CA, USA[6]. Mice transgenic for cDNA repre

enic female mice and 10LPA-cDNA transgenic female micere included in the experiment. The number of mice

ed from four to six per experimental group. The agehe WT mice varied from 20 to 22 weeks. FemaleLPA-AC transgenic mice exhibit high plasma levels of ap6]. In such mice, we have observed a slight increaslasma apo(a) level after 10 weeks of age. Thereforeompared the effect of fibrates between 10- and 36-weePA-YAC transgenic mice.LPA-cDNA transgenic mice werge-matched to the 36-week-oldLPA-YAC transgenic micehe animals were fed a semi-synthetic mouse diet (AIN

CN, Asse-Relegem, Belgium, control diet) or AIN-76 dupplemented with either 0.1% (w/w) fenofibrate or 0.0w/w) WY 14643. Both chemicals were obtained from Sigldrich (St. Louis, USA). Diets were made by soaking a

one or ether containing FF or WY 14643, respectively,eighted amounts of AIN-76 diet. The acetone or ether

hen evaporated under continuous mixing, leaving FF or4643 containing diet.

During the 3 weeks of the feeding experiment, the mad free access to tapwater and food. Blood samplesrawn from the saphenous vein prior to the experiment. Teeks later, the animals were anaesthetized by s.c. a

stration of 7.5 ml/kg of a mixture of one part Dormicu5 mg/ml midazolam) (Alpharma, Oslo, Norway) plus oart Hypnorm (0.315 mg/ml fentanyl and 10 mg/mlnisone) (Janssen Pharmaceutica, Beerse, Belgium) pluarts sterile water. Blood was drawn from a cut in the

lla and the mice were killed under anesthesia by cerislocation.

P.A. Teivainen-Laedre et al. / Pathophysiology 11 (2005) 201–208 203

2.3. Tissue preparation

The liver from each mouse was quickly excised and imme-diately washed with ice-cold saline before weighing. Piecesof liver, weighing approximately 500 mg, were submergedin 2 ml of RNAlaterTM buffer (Ambion Diagnostic, Cam-bridgeshire, England) and stored at−70◦C until RNA isola-tion took place.

2.4. Serum apo(a), lipids, aspartate aminotransferase(asat) and alanine aminotransferase (alat) levels

Human serum apo(a) levels in transgenic mice were mea-sured using a commercially available apo(a) RIA kit (Mer-codia AB, Uppsala, Sweden) as described previously[27].According to the manufacturer’s recommendations, the back-ground level was set to 17 U/l, which is approximately equalto 1.2 mg/dl.

Serum total cholesterol, triglycerides, asat and alat lev-els were measured at the Central Laboratory at NorwegianSchool of Veterinary Science by enzymatic, colorimetricmethods (Advia 1650, Bayer HealthCare LLC, Germany).

2.5. Quantitative PCR (qPCR)

the6 sterC It forqo eK

e de-t A).Sp dF ed0 m),aM ity,C ycleaf e-t No.2 izedil forc re-s

2.6. Statistical analysis

The non-parametric Mann–Whitney test was performedto search for statistically significant differences betweengroups. Data are presented as mean± S.E.M.

3. Results

3.1. Serum apo(a) levels in LPA-YAC and LPA-cDNAtransgenic mice

Feeding FF supplemented diet for 3 weeks resulted in re-duced serum apo(a) level in both 10- and 36-week-oldLPA-YAC transgenic mice (P< 0.05 andP< 0.005, respectively)(Fig. 1), compared to their respective controls. A similar re-duction in serum apo(a) levels was observed in WY 14643-treated 10-week-oldLPA-YAC transgenic mice (P< 0.05).

Overall, LPA-cDNA transgenic mice exhibited muchlower serum apo(a) levels thanLPA-YAC transgenic mice(P< 0.005). Feeding the FF supplemented diet toLPA-cDNAtransgenic mice had no effect on serum apo(a) levels.

3.2. Effect of fibrates on hepatic apo(a) mRNA levels inLPA-YAC and LPA-cDNA transgenic mice

up-p velst Fht s nott

3W

l asL ntf icee fedt mtY rols,d

onsY -oldL to-tT

TS ssion n

G Revers

C CAGA ATF CGTC

Total RNA was isolated from 10 mg mouse liver using100 Nucleic Acid PrepStation (Applied Biosystems, Foity, CA, USA) and eluted in 150�l elution buffer. DNase

reatment for 15 min was included in the procedure. cDNAPCR was prepared from 2.5�g total RNA in a final volumef 50�l of ddH2O using the High Capacity cDNA Archivit (Applied Biosystems, Foster City, CA, USA).qPCR was performed in the ABI Prism 7700 sequenc

ection system (Applied Biosystems, Foster City, CA, USequences for housekeeping genes andLPA primers androbes are published elsewhere[12] and for CYP7A1 anXR are presented inTable 1. The reaction mixture contain.3�M primers and probe (Eurogentech, Seraing, Belgiund 2�l of 1:10 diluted cDNA template in 25�l of 1× Taq-an Universal Master Mix (Applied Biosystems, Foster CA, USA). PCR cycle parameters were as follows: one ct 50◦C for 2 min and at 95◦C for 10 min; 40 cycles at 95◦C

or 15 s and at 60◦C for 1 min. Relative mRNA level was dermined by the Standard Curve Method (User Bulletin, Applied Biosystems). Expression level was normal

n relation to the means of cyclophilin and�-actin mRNAevel. In this method, inter-experiment variation (CV%)yclophilin and�-actin mRNA levels was 1.2 and 1.4%,pectively.

able 1equences for CYP7A1 (accession number: NM007824) and FXR (acce

ene Forward primer

YP7A1 CTTTGATGACATGGAGAAGGCTAAG TXR GGGTTTAGAAAATCCAATTCAGATTAGT C

LPA-YAC transgenic mice fed the FF or WY 14643 slemented diet had significantly lower apo(a) mRNA le

han mice fed the control diet (Fig. 2). In comparison, Fad no effect on hepatic apo(a) mRNA level inLPA-cDNA

ransgenic mice (the WY 14643 supplemented diet waested on these animals).

.3. Serum triglyceride and total cholesterol levels inT and LPA-YAC transgenic mice

Serum triglyceride levels in FF-treated WT as welPA-YAC transgenic mice were not significantly differe

rom their respective controls. WY 14643-treated WT mxhibited lower serum triglyceride level than WT micehe control diet (P< 0.05,Table 2). The lower mean seruriglyceride level in WY 14643-treated 10-week-oldLPA-AC transgenic mice, compared to their respective contid not reach statistically significance.

FF or WY 14643 had no statistically significant effectserum total cholesterol levels in WT or 10-week-oldLPA-AC transgenic mice. However, the FF treated 36-weekPA-YAC transgenic mice had significantly higher serum

al cholesterol levels than their respective controls (P< 0.01,able 3).

umber: AF293370) primers and probes used in qPCR

e primer Probe

GGCTGCTTTCATTGCT CGCACCTTGTGATCCTCTGGGCCGGCACAAATCC TTCACCACAGCCACCGGCTGTC

204 P.A. Teivainen-Laedre et al. / Pathophysiology 11 (2005) 201–208

Fig. 1. Effect of 3 weeks feeding of fenofibrate (FF) or WY 14643 (WY)supplemented diet on serum apo(a) levels inLPA transgenic mice with orwithout theLPA–PLG intergenic region. Apo(a) levels in individual mouseserum were measured using a commercially available RIA kit. Data arepresented as mean± S.E.M. (mg/dl). Number of mice is presented in paren-theses. Black bars represent serum apo(a) levels at the start and open barsrepresent levels after 3 weeks feeding of control, FF or WY 14643 sup-plemented diet. The asterisk (*P< 0.05) and double asterisk (** P< 0.005)indicate statistically significant differences compared to respective mice fedthe control diet. FF: 0.1% (w/w) fenofibrate supplemented diet; WY 14643:0.05% (w/w) WY 14643 supplemented diet.

3.4. Hepatic FXR, LXR, CYP7A1 and APO A5 mRNAlevels in WT and LPA-YAC transgenic mice

The difference in CYP7A1 mRNA levels between the FF-treated WT, 10- or 36-week-oldLPA-YAC transgenic miceand respective mice fed the control diet did not reach statisti-

Fig. 2. Effect of 3 weeks feeding of fenofibrate (FF) or WY 14643 (WY)supplemented diet on hepatic apo(a) mRNA levels inLPA transgenic micewith or without theLPA–PLG intergenic region. For normalization of ex-pression data means of cyclophilin and�-actin mRNA level were used. Dataare presented as the mean± S.E.M. of normalized expression levels. Num-ber of mice is presented in parentheses. Open bars represent mice on thecontrol diet. Grey or black bars represent levels in mice fed the FF or WY1 ea aredt pple-m

Table 2Serum triglyceride (TG) and total cholesterol (TC) levels in WT andLPA-YAC transgenic mice fed a control diet or diets supplemented with fenofibrate(FF) or WY 14643 (WY) for 3 weeks

Mice genotype Age (weeks)n Diet TG (mmol/l) TC (mmol/l)

Wild type 20–22 4 Control 1.9± 0.2 4.1± 0.4Wild type 20–22 4 FF 1.5± 0.1 4.7± 0.5Wild type 20–22 4 WY 0.7± 0.0* 3.9 ± 0.4

LPA-YAC 10 5 Control 2.1± 0.2 3.6± 0.6LPA-YAC 10 5 FF 1.6± 0.2 4.6± 0.2LPA-YAC 10 5 WY 1.1± 0.3 4.7± 0.3

LPA-YAC 36 6 Control 1.5± 0.0 3.1± 0.1LPA-YAC 36 6 FF 1.5± 0.1 4.9± 0.5**

Hypnorm–Dormicum anesthetized mice were bled from a cut in the axilla atthe end of the feeding experiment. Serum triglyceride and serum total choles-terol levels were measured using the Advia 1650 automatic analyzes system.Data are mean± S.E.M. (mmol/l). The asterisk (*P< 0.05) and double aster-isk (** P< 0.01) indicate statistically significant differences in comparisonto mice fed the control diet. FF: 0.1% (w/w) fenofibrate supplemented diet;WY: 0.05% (w/w) WY 14643 supplemented diet.

cally significant levels. However, WT mice fed the WY 14643supplemented diet had significantly lower, and 10-week-oldLPA-YAC transgenic mice fed the WY 14643 supplementeddiet significantly higher, CYP7A1 mRNA levels than theirrespective controls (Fig. 3).

Hepatic FXR mRNA levels (Fig. 4) were significantlyhigher in all groups that were given the FF supplementeddiet compared to their respective control groups. The WY14643 supplemented diet had a similar effect.

Hepatic LXR mRNA levels were significantly increasedonly in WT mice fed the WY 14643 supplemented diet(Fig. 5), compared to their respective controls.

Hepatic apolipoprotein A5 (APO A5) mRNA levels didnot show significant variation between the different groups(Fig. 6).

Table 3Serum aspartate aminotransferase (asat) and alanine aminotransferase (alat)levels in mice fed control diet or diets supplemented with fenofibrate (FF)or WY 14643 (WY) for 3 weeks

Mice genotype Age (weeks) n Diet Asat (U/l) Alat (U/l)

Wild type 20–22 4 Control 162± 20 39± 2Wild type 20–22 4 FF 208± 55 68± 5*

Wild type 20–22 4 WY 358± 92* 161 ± 26*

LPA-YAC 10 5 Control 456± 58 81± 18LL

LL

H illa att e Advia1a aredt diet;W

4643 supplemented diet, respectively. The asterisk (*P< 0.05) and doublsterisk (** P< 0.005) indicate statistically significant differences comp

o respective mice fed the control diet. FF: 0.1% (w/w) fenofibrate suented diet; WY 14643: 0.05% (w/w) WY 14643 supplemented diet.

PA-YAC 10 5 FF 548± 222 282± 57PA-YAC 10 5 WY 258± 40 116± 12

PA-YAC 36 6 Control 315± 12 55± 7PA-YAC 36 6 FF 257± 50 109± 10*

ypnorm–Dormicum anesthetized mice were bled from a cut in the axhe end of the study. Serum alat and asat levels were measured using th650 automatic analyzes system. Data are presented as mean± S.E.M. Thesterisk (*P< 0.05) indicates statistically significant differences comp

o mice fed the control diet. FF: 0.1% (w/w) fenofibrate supplementedY: 0.05% (w/w) WY 14643 supplemented diet.

P.A. Teivainen-Laedre et al. / Pathophysiology 11 (2005) 201–208 205

Fig. 3. Hepatic CY7A1 mRNA levels in WT orLPA-YAC transgenic micefed control diet or diets supplemented with fenofibrate (FF) or WY 14643(WY) for 3 weeks. For normalization of expression data means of cyclophilinand�-actin mRNA level were used. Data is presented as mean± S.E.M. ofnormalized expression levels. Number of mice is presented in parentheses.Open bar represents median levels in mice fed the control diet. Grey andblack bars represent median levels in mice fed the FF or the WY 14643supplemented diet, respectively. The asterisk (*P< 0.05) indicates a statis-tically significant difference compared to the respective control group. WT:wild type; FF: 0.1% (w/w) fenofibrate supplemented diet; WY 14643: 0.05%(w/w) WY 14643 supplemented diet.

3.5. Serum asat and alat level in WT and LPA-YACtransgenic mice

Serum asat levels were significantly increased only inWY 14643-treated WT mice (P< 0.05,Table 3). Serum alatshowed some increase in FF-treated WT and 36-week-oldLPA-YAC transgenic mice compared to their respective mice

Fig. 4. Hepatic FXR mRNA level in WT andLPA-YAC transgenic mice fed acontrol diet or diets supplemented with fenofibrate (FF) or WY 14643 (WY)f n and�

m ice fedt mice,rd ontrolg 05%(

Fig. 5. Hepatic LXR mRNA level in WT andLPA-YAC transgenic micefed a control diet or diets supplemented with fenofibrate (FF) or WY 14643(WY) for 3 weeks. For normalization of expression data means of cyclophilinand�-actin mRNA level were used. Data is presented as mean± S.E.M. ofnormalized expression levels. Open bars represent median values in micefed the control diet. Grey and black bars represent FF or WY 14643 treatedmice, respectively. The asterisk (*P< 0.05) indicates statistically significantdifference compared to the respective control group. FF: 0.1% (w/w) supple-mented fenofibrate diet; WY 14643: 0.05% (w/w) WY 14643 supplementeddiet.

controls, as well as in WY 14643-treated WT mice. How-ever, all values appear to be within the limits for normal liverfunction in mice[28].

3.6. Relative liver weight (% of body weight)

All mice fed the FF or the WY 14643 supplemented dietshowed a two- to threefold increase in relative liver weight(RLW) compared to mice on the control diet (Table 4).

F ef 4643( hilinan ice fedt mice,r 643:0

or 3 weeks. For normalization of expression data means of cyclophili-actin mRNA level were used. Data are presented as mean± S.E.M. of nor-alized expression levels. Open bars represent median values in m

he control diet. Grey and black bars represent FF or WY 14643 treatedespectively. The asterisk (*P< 0.05) and double asterisk (** P< 0.005) in-icate statistically significant differences compared to the respective croup. FF: 0.1% (w/w) supplemented fenofibrate diet; WY 14643: 0.w/w) WY 14643 supplemented diet.

ig. 6. Hepatic APO A5 mRNA level in WT andLPA-YAC transgenic miced a control diet or diets supplemented with fenofibrate (FF) or WY 1WY) for 3 weeks. For normalization of expression data means of cyclopnd�-actin mRNA level were used. Data is presented as mean± S.E.M. oformalized expression levels. Open bars represent median values in m

he control diet. Grey and black bars represent FF or WY 14643 treatedespectively. FF: 0.1% (w/w) supplemented fenofibrate diet; WY 14.05% (w/w) WY 14643 supplemented diet.

206 P.A. Teivainen-Laedre et al. / Pathophysiology 11 (2005) 201–208

4. Discussion

To examine if the fibrate doses used in present experimentshad hepatotoxic effects, we studied serum transaminase lev-els, which are widely used as markers of liver injury. If thereis liver damage, transaminase values would be expected toincrease many fold. Thus, the slight increase in transami-nase levels observed (Table 3) after maintenance on a fibrate-supplemented diet most likely reflects increased hepatic geneexpression rather than significant liver damage[29,30].

It is known that fibrates induce peroxisomal prolifera-tion and growth and are associated with hypertrophy of theliver, through PPAR� activation [31]. Therefore, it seemslikely that the observation of increased relative liver weight(Table 4) in mice in response to FF or WY 14643 treatmentsreflects PPAR� activation.

To the best of our knowledge, we have demonstrated forthe first time that fibrates reduce serum apo(a) and hepaticLPAexpression levels inLPA-YAC transgenic mice. The ef-fect of fibrates inLPA-YAC transgenic mice is most prob-ably mediated upstream of theLPA gene since fibrates hadno effect on serum apo(a) or hepatic apo(a) mRNA levelsin mice transgenic for cDNA representing the humanLPAgene. Reduction inLPA transcription has also been observedin high-fat fedLPA-YAC transgenic mice[9–12]. In humans,Lp(a) levels do not appear to change significantly in responset st at in-fl

t isc co-t at att bes othergi igh-

TR ietss

M

WWW

LLL

LL

A ly ex-c bodyw s-t /w)f ntedd

fat diet is not associated with significant changes in hepaticPPAR� expression[12].

It is known that intake of fibrates increases the relative lev-els of primary bile acid cholic acid[25], which is a high affin-ity FXR ligand [33,34]. Thus, the observation of increasedhepatic FXR mRNA level in response to fibrate-treatmentboth in WT andLPA-YAC transgenic mice most likely re-flects alterations in composition of the bile acid pool.

Previously, it has been shown that FXR negatively regu-lates bile acid production by inhibiting transcription of theCYP7A1 gene[35]. This is in agreement with our observa-tion that both WT andLPA-YAC transgenic mice tend to havelower CYP7A1 levels after maintenance on FF-containingdiet. However, the present experiment does not explain whyWY 14643 increased CYP7A1 mRNA level inLPA-YACtransgenic mice. Whether or not this finding is related toLPAtransgenicity must be studied further.

Activated FXR binds after heterodimerization with RXRto its response element. The optimal DNA binding sequencefor the FXR/RXR heterodimer is an inverted repeat com-posed of two AGGTCA halfsites spaced by one nucleotide[36]. Alignment analysis of the 40 kb nucleotide sequence ofLPA–PLGintergenic region (accession number: NT007422)and binding sequence for the FXR/RXR heterodimer revealedthat about 25 kb upstream of theLPA start codon the het-erodimer sequence TTTATGAAGGTCAAATGACCTT ex-i acingb emsp ucha .T con-tg ate inh p(a)l etet

at nt ytesa hy-p ,w A5m iths -oldL erumt gestt micet sento e tofi hichs glyc-e ia-t ts.

o dietary changes or fibrate intake[8,21–24]. This suggesthe existence of species-specific transcription factors thuenceLPAexpression.

The decreasedLPA expression after fibrate treatmenontrary to findings of Puckey et al. They showed inransfection assays that PPAR� enhancesLPAexpression viheLPA–PLG intergenic region[32]. This may suggest thhe reduction inLPAexpression we have observed couldecondary to an effect of fibrates on the expression ofenes than PPAR�. Accordingly, reducedLPA expression

n LPA-YAC transgenic mice after maintenance on a h

able 4elative liver weights (% of bodyweight) in mice fed control diet or dupplemented with fenofibrate (FF) or WY 14643 (WY) for 3 weeks

ice genotype Age (weeks) n Diet RLW

ild type 20–22 4 Control 4.7± 0.2ild type 20–22 4 FF 8.9± 0.4*

ild type 20–22 4 WY 12.3± 0.3*

PA-YAC 10 5 Control 4.6± 0.1PA-YAC 10 5 FF 11.4± 0.3*

PA-YAC 10 5 WY 12.1± 0.3*

PA-YAC 36 6 Control 4.4± 0.1PA-YAC 36 5a FF 9.3± 0.5*

t the end of the study, the mice were sacrificed and livers were quickised and weighed. Relative liver weight (RLW) is calculated as % ofeight. Data are mean± S.E.M. The asterisk (*P< 0.05) indicates stati

ically significant differences compared to control mice. FF: 0.1% (wenofibrate supplemented diet; WY: 0.05% (w/w) WY 14643 supplemeiet.a Value from one mouse is missing.

sts. Although this sequence contained 2-base pair spetween the putative FXR/RXR binding halfsites, it seossible that the FXR/RXR heterodimer could bind to ssequence and, thus, be able to regulateLPA expression

herefore, we suggest that bile acids through FXR mayribute to suppressedLPA transcription inLPA-YAC trans-enic mice. This putative mechanism seems not to operumans since fibrate intake has no significant effect on L

evels in humans[21,23,24]. However, it could still underlihe transcriptional mechanism, which results in reducedLPAxpression upon fibrate-treatment or fat-feeding, inLPA-YACransgenic mice.

APO A5, which is important in determining plasmriglyceride levels in mouse and man[37], was recently showo be highly up-regulated by fibrates in human hepatocnd it was suggested that it is a major mediator of theotriglyceridemic effect of fibrates in humans[38]. Howevere found no significant effect of fibrate-treatment on APORNA levels in mice, and this appears to be in conflict w

uch a notion. On the other hand, our WT and 10-weekPA-YAC transgenic mice appeared to have reduced s

riglyceride levels after fibrate treatment. This may sughat the molecular mechanism for the hypotriglycerideffect of fibrates observed in WT and 10-week-oldLPA-YAC

ransgenic mice may be independent of APO A5. The prebservation of increased FXR mRNA levels in responsbrate treatment together with recent studies in rats, whowed that synthetic FXR ligand decreases plasma trirides levels[39], suggest that FXR is a potential med

or of the hypotriglyceridemic effect of fibrates in roden

P.A. Teivainen-Laedre et al. / Pathophysiology 11 (2005) 201–208 207

This is also supported by the findings that the activationof the FXR-SHP-SREBP-1c regulatory cascade leads to hy-potriglyceridemia in mice[40]. Furthermore, it is plausiblethat the lack of a hypotriglyceridemic effect and the pres-ence of hypercholesterolemia in fibrate-treated 36-week-oldLPA-YAC transgenic mice reflect reduced hepatic capacity tometabolize lipids in old animals, as suggested previously byothers[41].

Today, there is no prescription drug available with sig-nificant effects on plasma Lp(a) levels. Accordingly, newknowledge concerning regulation ofLPA expression is es-sential for the development of pharmaceuticals. The presentdata clearly demonstrates that PPAR� activation is followedby reduced serum apo(a) levels and hepaticLPAexpression,which is associated with increased hepatic FXR expression,in fibrate-fedLPA-YAC transgenic mice. However, on the ba-sis of current data it is not possible to conclude that PPAR�is the primary factor repressingLPAexpression inLPA-YACtransgenic mice. It seems equally likely that reduced expres-sion ofLPA could be a secondary consequence of PPAR�’sactivation on the expression of other genes.

5. Conclusions

The results showed that fibrates reduceLPA expres-s onam g-g fectv oft it isn -tm o(a)m e-q ise -o s,sm

A

ianR triciaE

R

ofNatl.

hen,man

apolipoprotein(a) is homologous to plasminogen, Nature 330 (1987)132–137.

[3] P. Magnaghi, E. Citterio, N. Malgaretti, F. Acquati, S. Ottolenghi,R. Taramelli, Molecular characterisation of the human apo(a)-plasminogen gene family clustered on the telomeric region of chro-mosome 6 (6q26–27), Hum. Mol. Genet. 3 (1994) 437–442.

[4] K. Makino, A. Abe, S. Maeda, A. Noma, M. Kawade, O. Takenaka,Lipoprotein(a) in nonhuman primates. Presence and characteristicsof Lp(a) immunoreactive materials using anti-human Lp(a) serum,Atherosclerosis 78 (1989) 81–85.

[5] P.M. Laplaud, L. Beaubatie, S.C. Rall Jr., G. Luc, M. Saboureau,Lipoprotein[a] is the major apoB-containing lipoprotein in theplasma of a hibernator, the hedgehog (Erinaceus europaeus), J. LipidRes. 29 (1988) 1157–1170.

[6] K.A. Frazer, G. Narla, J.L. Zhang, E.M. Rubin, The apolipoprotein(a)gene is regulated by sex hormones and acute-phase inducers in YACtransgenic mice, Nat. Genet. 9 (1995) 424–431.

[7] G. Chiesa, H.H. Hobbs, M.L. Koschinsky, R.M. Lawn, S.D. Maika,R.E. Hammer, Reconstitution of lipoprotein(a) by infusion of hu-man low density lipoprotein into transgenic mice expressing humanapolipoprotein(a), J. Biol. Chem. 267 (1992) 24369–24374.

[8] B.A. Clevidence, J.T. Judd, E.J. Schaefer, J.L. Jenner, A.H. Lichten-stein, R.A. Muesing, J. Wittes, M.E. Sunkin, Plasma lipoprotein (a)levels in men and women consuming diets enriched in saturated,cis-,or trans-monounsaturated fatty acids, Arterioscler. Thromb. Vasc.Biol. 17 (1997) 1657–1661.

[9] T. Huby, V. Afzal, C. Doucet, R.M. Lawn, E.L. Gong, M.J. Chap-man, J. Thillet, E.M. Rubin, Regulation of the expression of theapolipoprotein(a) gene: evidence for a regulatory role of the 5′ dis-tal apolipoprotein(a) transcription control region enhancer in yeastartificial chromosome transgenic mice, Arterioscler. Thromb. Vasc.

[ nd,e hu-

[ , A.ns-Res.

[ rg,el in

[ istry

[ i. 3

[ nantom-

[ ee,th,

ofacid

[ les-LPA

[ ters-and

[ tors:8.

[ R.m of

0399.[ t of

ion in LPA-YAC transgenic mice, but have no impactpo(a) mRNA or protein levels inLPA-cDNA transgenicice, which lack theLPA–PLG intergenic region. This suests that the effect of fibrates, which exert their efia PPAR�, on LPA expression is mediated upstreamhe LPA gene. However, on the basis of current dataot possible to conclude that PPAR� is the primary fac

or that repressesLPA expression inLPA-YAC transgenicice. Indeed, negative correlation between FXR and apRNA levels, in addition to putative FXR DNA binding suence inLPA–PLG intergenic region, suggests that itqually likely that reduced expression ofLPAcould be a secndary consequence of PPAR� activation on other geneuch as FXR, after fibrate-treatment ofLPA-YAC transgenicice.

cknowledgements

This work was supported by a grant from the Norwegesearch Council (grant no. 134426/140). We thank Pangen for her skilful technical assistance.

eferences

[1] C. Brunner, H.G. Kraft, G. Utermann, H.J. Muller, Cys4057apolipoprotein(a) is essential for lipoprotein(a) assembly, Proc.Acad. Sci. U.S.A. 90 (1993) 11643–11647.

[2] J.W. McLean, J.E. Tomlinson, W.J. Kuang, D.L. Eaton, E.Y. CG.M. Fless, A.M. Scanu, R.M. Lawn, cDNA sequence of hu

Biol. 23 (2003) 1633–1639.10] P.A. Teivainen, K.A. Eliassen, K. Berg, K. Torsdalen, A. Svindla

Atherogenesis and vascular calcification in mice expressing thman LPA gene, Pathophysiology 11 (2004) 113–120.

11] F. Acquati, R. Hammer, B. Ercoli, V. Mooser, R. Tao, V. RonickeMichalich, G. Chiesa, R. Taramelli, H.H. Hobbs, H.J. Muller, Tragenic mice expressing a human apolipoprotein[a] allele, J. Lipid40 (1999) 994–1006.

12] P.A. Teivainen, K.A. Eliassen, E.M. Rubin, S. Djurovic, K. BeHuman apoB contributes to increased serum total apo(a) levLPA transgenic mice, Lipids Health Dis. 3 (2004) 8.

13] D.W. Russell, K.D. Setchell, Bile acid biosynthesis, Biochem31 (1992) 4737–4749.

14] J.Y.L. Chiang, Regulation of bile acid synthesis, Front. Biosc(1998) D176–D193.

15] S. Gupta, W.M. Pandak, P.B. Hylemon, LXR alpha is the domiregulator of CYP7A1 transcription, Biochem. Biophys. Res. Cmun. 293 (2002) 338–343.

16] B. Goodwin, S.A. Jones, R.R. Price, M.A. Watson, D.D. McKL.B. Moore, C. Galardi, J.G. Wilson, M.C. Lewis, M.E. RoP.R. Maloney, T.M. Willson, S.A. Kliewer, A regulatory cascadethe nuclear receptors FXR, SHP-1, and LRH-1 represses bilebiosynthesis, Mol. Cells 6 (2000) 517–526.

17] P. Teivainen, K. Eliassen, S. Djurovic, K. Berg, Effect of choterol feeding on apo(a), CYP7A1 and LXR alpha expression intransgenic mice, Atheroscler. Suppl. 4 (2003) 296.

18] B. Staels, J. Dallongeville, J. Auwerx, K. Schoonjans, E. Leidorf, J.C. Fruchart, Mechanism of action of fibrates on lipidlipoprotein metabolism, Circulation 98 (1998) 2088–2093.

19] B. Desvergne, W. Wahli, Peroxisome proliferator-activated recepnuclear control of metabolism, Endocr. Rev. 20 (1999) 649–68

20] D.P. Wade, L.H. Puckey, B.L. Knight, F. Acquati, A. Mihalich,Taramelli, Characterization of multiple enhancer regions upstreathe apolipoprotein(a) gene, J. Biol. Chem. 272 (1997) 30387–3

21] G. Dahlen, T. Gillnas, A.L. Borresen, K. Berg, C. Ericson, Effecgemfibrozil on serum lipid levels, Artery 7 (1980) 224–231.

208 P.A. Teivainen-Laedre et al. / Pathophysiology 11 (2005) 201–208

[22] D.M. Neele, A. Kaptein, H. Huisman, E.C. de Wit, H.M. Princen, Noeffect of fibrates on synthesis of apolipoprotein(a) in primary culturesof cynomolgus monkey and human hepatocytes: apolipoprotein A-Isynthesis increased, Biochem. Biophys. Res. Commun. 244 (1998)374–378.

[23] P.H. Jones, H.J. Pownall, W. Patsch, J.A. Herd, J.A. Farmer, C.Payton-Ross, K.T. Kimball, A.M. Gotto, J.D. Morrisett, Effect ofgemfibrozil on levels of lipoprotein[a] in type II hyperlipoproteine-mic subjects, J. Lipid Res. 37 (1996) 1298–1308.

[24] S.J. Bredie, H.T. Westerveld, H.C. Knipscheer, T.W. de Bruin, J.J.Kastelein, A.F. Stalenhoef, Effects of gemfibrozil or simvastatinon apolipoprotein-B-containing lipoproteins, apolipoprotein-CIII andlipoprotein(a) in familial combined hyperlipidaemia, Neth. J. Med.49 (1996) 59–67.

[25] M.C. Hunt, Y.Z. Yang, G. Eggertsen, C.M. Carneheim, M. Gafvels,C. Einarsson, S.E. Alexson, The peroxisome proliferator-activatedreceptor alpha (PPARalpha) regulates bile acid biosynthesis, J. Biol.Chem. 275 (2000) 28947–28953.

[26] S.M. Post, H. Duez, P.P. Gervois, B. Staels, F. Kuipers, H.M. Princen,Fibrates suppress bile acid synthesis via peroxisome proliferator-activated receptor-alpha-mediated downregulation of cholesterol7alpha-hydroxylase and sterol 27-hydroxylase expression, Arte-rioscler. Thromb. Vasc. Biol. 21 (2001) 1840–1845.

[27] K. Berg, A. Svindland, A.J. Smith, R.M. Lawn, S. Djurovic, A. Ale-strom, P. Alestrom, K. Eliassen, Spontaneous atherosclerosis in theproximal aorta of LPA transgenic mice on a normal diet, Atheroscle-rosis 163 (2002) 99–104.

[28] Reference values for clinical chemical analytes, in: W.F. Loeb, F.W.Quimby (Eds.), The Clinical Chemistry of Laboratory Animals, Perg-amon Press, New York, 1989, pp. 415–495.

[29] C. Tomkiewicz, F. Muzeau, A.D. Edgar, R. Barouki, M. Agger-artate(2004)

[ eck,nofi-e pro-Lett.

[ oetz,d dis-atedc ef-

fects of peroxisome proliferators, Mol. Cell. Biol. 15 (1995) 3012–3022.

[32] L.H. Puckey, B.L. Knight, Interaction of oestrogen and peroxi-some proliferator-activated receptors with apolipoprotein(a) gene en-hancers, Biochem. J. 366 (2002) 157–163.

[33] M. Makishima, A.Y. Okamoto, J.J. Repa, H. Tu, R.M. Learned, A.Luk, M.V. Hull, K.D. Lustig, D.J. Mangelsdorf, B. Shan, Identi-fication of a nuclear receptor for bile acids, Science 284 (1999)1362–1365.

[34] D.J. Parks, S.G. Blanchard, R.K. Bledsoe, G. Chandra, T.G. Con-sler, S.A. Kliewer, J.B. Stimmel, T.M. Willson, A.M. Zavacki, D.D.Moore, J.M. Lehmann, Bile acids: natural ligands for an orphannuclear receptor, Science 284 (1999) 1365–1368.

[35] J.Y. Chiang, R. Kimmel, C. Weinberger, D. Stroup, Farnesoid Xreceptor responds to bile acids and represses cholesterol 7alpha-hydroxylase gene (CYP7A1) transcription, J. Biol. Chem. 275 (2000)10918–10924.

[36] B.A. Laffitte, H.R. Kast, C.M. Nguyen, A.M. Zavacki, D.D. Moore,P.A. Edwards, Identification of the DNA binding specificity and po-tential target genes for the farnesoid X-activated receptor, J. Biol.Chem. 275 (2000) 10638–10647.

[37] L.A. Pennacchio, M. Olivier, J.A. Hubacek, J.C. Cohen, D.R. Cox,J.C. Fruchart, R.M. Krauss, E.M. Rubin, An apolipoprotein influ-encing triglycerides in humans and mice revealed by comparativesequencing, Science 294 (2001) 169–173.

[38] N. Vu-Dac, P. Gervois, H. Jakel, M. Nowak, E. Bauge, H. De-hondt, B. Staels, L.A. Pennacchio, E.M. Rubin, J. Fruchart-Najib,J.C. Fruchart, Apolipoprotein A5, a crucial determinant of plasmatriglyceride levels, is highly responsive to peroxisome proliferator-activated receptor alpha activators, J. Biol. Chem. 278 (2003) 17982–17985.

[ dra,is,the2974.

[ els-yc-c, J.

[ M.ceri-anges66.

beck, Opposite regulation of the rat and human cytosolic aspaminotransferase genes by fibrates, Biochem. Pharmacol. 67213–225.

30] A.D. Edgar, C. Tomkiewicz, P. Costet, C. Legendre, M. AggerbJ. Bouguet, B. Staels, C. Guyomard, T. Pineau, R. Barouki, Febrate modifies transaminase gene expression via a peroxisomliferator activated receptor alpha-dependent pathway, Toxicol.98 (1998) 13–23.

31] S.S. Lee, T. Pineau, J. Drago, E.J. Lee, J.W. Owens, D.L. KrP.M. Fernandez-Salguero, H. Westphal, F.J. Gonzalez, Targeteruption of the alpha isoform of the peroxisome proliferator-activreceptor gene in mice results in abolishment of the pleiotropi

39] P.R. Maloney, D.J. Parks, C.D. Haffner, A.M. Fivush, G. ChanK.D. Plunket, K.L. Creech, L.B. Moore, J.G. Wilson, M.C. LewS.A. Jones, T.M. Willson, Identification of a chemical tool fororphan nuclear receptor FXR, J. Med. Chem. 43 (2000) 2971–

40] M. Watanabe, S.M. Houten, L. Wang, A. Moschetta, D.J. Mangdorf, R.A. Heyman, D.D. Moore, J. Auwerx, Bile acids lower trigleride levels via a pathway involving FXR, SHP, and SREBP-1Clin. Invest. 113 (2004) 1408–1418.

41] E. Sanguino, M. Ramon, L. Michalik, W. Wahli, M. Alegret, R.Sanchez, M. Vazquez-Carrera, J.C. Laguna, Lack of hypotriglydemic effect of gemfibrozil as a consequence of age-related chin rat liver PPARalpha, Biochem. Pharmacol. 67 (2004) 157–1