Statin Induction of Liver Fatty Acid-Binding Protein (L-FABP) Gene Expression Is Peroxisome Proliferator-activated Receptor- (cid:1) -dependent*

Statins are drugs widely used in humans to treat hypercholesterolemia. Statins act by inhibiting cholesterol synthesis resulting in the activation of the transcription factor sterol-responsive element-binding protein-2 that controls the expression of genes involved in cholesterol homeostasis. Statin therapy also decreases plasma triglyceride and non-esterified fatty acid levels, but the mechanism behind this effect remains more elusive. Liver fatty acid-binding protein (L-FABP) plays a role in the influx of long-chain fatty acids into hepatocytes. Here we show that L-FABP is a

Statins are competitive inhibitors of 3-hydroxy-3-methylglutaryl CoA reductase, the rate-limiting enzyme in cholesterol biosynthesis. The resulting lower intracellular cholesterol concentration after statin treatment leads to a proteolytic activa-tion of the transcription factor sterol responsive element-binding protein-2 (SREBP-2), 1 which up-regulates several genes controlling cholesterol homeostasis, including the LDL receptor (LDLr) (1). Induction of LDLr in the liver enhances clearance of circulating LDL resulting in decreased plasma LDL-cholesterol levels. Statins also increase, at least in part via a PPAR␣-dependent mechanism (2), the level of high density lipoproteins (3,4). As a consequence, statins improve the blood cholesterol profile and markedly reduce cardiovascular mortality and morbidity in dyslipidemic patients (5)(6)(7). Statins also influence plasma TG and NEFA levels in rats and humans (4, 8 -11) through mechanisms not yet fully elucidated. Sustained hepatic clearance of TG-rich very low density lipoproteins by the LDLr (12), statin-dependent up-regulation of the lipoprotein lipase (LPL) gene, and down-regulation of the LPL inhibitor apolipoprotein C-III (13) may all contribute to the TG-lowering effect of statins. By contrast, it is not known if statins control the cellular uptake and the metabolic fate of lipids. Moreover, the NEFA lowering effect of statins is not yet explained.
Liver fatty acid-binding protein (L-FABP) is a cytoplasmic protein exhibiting a strong affinity for long-chain fatty acids (LCFA). It is highly expressed in liver, where it represents more than 5% of cytosolic proteins (14). The transcription rate of the L-FABP gene is tightly regulated and induced by both fibrate hypolipidemic drugs and LCFA through a peroxisome proliferator-activated receptor (PPAR)-responsive element located in the proximal part of the promoter. Several lines of evidence show that L-FABP participates in the cellular uptake and trafficking of LCFA. Using stably transfected L-cell fibroblasts it was shown that L-FABP overexpression leads to an increased uptake of LCFA (15). Conversely, antisense L-FABP knock-down reduced cellular LCFA influx in a dose-dependent manner in HepG2 cells (16). L-FABP expression can also affect the metabolic fate of LCFA. Indeed, overexpression of L-FABP was reported to be associated with an induced ␤-oxidation activity in McA-RH7777 hepatoma cells (17). In the mouse, targeted deletion of the L-FABP gene resulted in a reduced rate of LCFA uptake by the liver under conditions of high lipid supply (e.g. intravenous LCFA bolus or prolonged fasting) (18,19). Fasted L-FABP-null mice were characterized by lower hepatic TG synthesis and fatty acid oxidation than wild-type controls (19,20). This phenotype is likely caused by the limitation of cellular LCFA availability for further metabolic use (i.e. esterification and oxidation) under high fatty acid load.
Altogether, these findings highlight the pivotal role exerted by L-FABP in hepatic lipid disposal and metabolic utilization. In the present study, we evaluated whether the hypolipidemic effect of statins is associated with an up-regulation of L-FABP expression in liver.

MATERIALS AND METHODS
Materials-Simvastatin was kindly provided by Merck Research Laboratories (Rahway, NJ). Fenofibrate and Wy14,643 were purchased from Sigma. All products for cell culture were purchased from Invitrogen.
Triglyceride Measurement-Plasma triglyceride concentrations were determined by an enzymatic assay adapted to microtiter plates using commercially available reagents (Roche Applied Science).
Animals and Experimental Treatments-French guidelines for the use and care of laboratory animals were followed. Wild-type or PPAR␣-null mice in the SV129 background (a kind gift of F. Gonzalez, Laboratory of Metabolism, National Institutes of Health, Bethesda, MD) were used. Animals were housed individually in a controlled environment (constant temperature and humidity, darkness from 8 p.m. to 8 a.m.) and fed ad libitum a standard chow (UAR A04, Usine d'Alimentation Rationnelle). PPAR␣(ϩ/ϩ) and PPAR␣(Ϫ/Ϫ) mice were force-fed for 5 days with simvastatin (100 mg/kg) and/or fenofibrate (40 mg/kg). Controls received the vehicle alone (1% carboxymethyl cellulose) by the same route. After sacrifice, livers were removed, snap frozen in liquid nitrogen, and stored at Ϫ80°C until RNA extraction.
Cell Culture and Treatments-Rat hepatocytes were isolated by collagenase perfusion according to the modified procedure of Seglen (21). Hepatocytes were plated in 60-mm dishes in Williams E medium (Invitrogen) containing 100 units/ml penicillin, 100 g/ml streptomycin, and 10% fetal calf serum (FCS). After an attachment period of 6 h, the culture medium was replaced by serum-free medium, and cells were incubated for an additional 24 h with 2.4 -24 M simvastatin. Control cultures received the vehicle alone (2 l/ml ethanol).
Rat Fao hepatoma cells (passages 20 -30) were cultured in a controlled environment (37°C, 5% CO 2 ) in Ham's F-12 medium, 100 units/ml penicillin, 100 g/ml streptomycin supplemented with 10% FCS. On the first day of confluency, cells were incubated for 24 h in the culture medium in the presence of Wy14,643 (10 M) or simvastatin (12 M) alone or in combination. Control cultures received the vehicle alone (2 l/ml ethanol).
Northern Blotting-Total RNAs were extracted with TRIzol reagent (Invitrogen) and electrophoresed on a 1% agarose gel and then transferred to GeneScreen membranes (PerkinElmer Life Sciences) using previously published procedures (22). Rat L-FABP and rat acyl-CoA oxidase (ACO) cDNA probes were labeled with [␣-32 P]dCTP (3000 Ci/ mmol, ICN) using the Prime-It RmT Random Primer Labeling kit (Stratagene). A 24-residue oligonucleotide specific for rat 18 S rRNA was used as probe to ensure that equivalent amounts of RNAs were loaded and transferred. This oligonucleotide was 5Ј-end-labeled using T 4 polynucleotide kinase and [␥-32 P]ATP (3000 Ci/mmol, ICN).
Real Time Quantitative Reverse Transcriptase-PCR-PPAR␣ and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA or 28 S rRNA levels were determined by real time reverse transcriptase-PCR. RNA integrity was checked by ethidium bromide staining after electrophoresis. The cDNA was produced from total RNA (1 g) by reverse transcription using Superscript reverse transcriptase (200 units, Invitrogen) in a 20-l reaction volume containing 1ϫ Superscript buffer (Invitrogen), 1 mM dNTP, 20 ng of random hexamers, 10 mM dithiothreitol, and 20 units of RNase inhibitors. After 50 min at 42°C incubation, the reaction was stopped (10 min at 70°C). PCR amplification of PPAR␣ and GAPDH or 28 S used Syber PCR master mix and the following primer sequences: PPAR␣ (forward, 5Ј-CCTCTTC-CCAAAGCTCCTTCA-3Ј; reverse, 5Ј-GTACGAGCTGCGCATGCTC-3Ј), GAPDH (forward, 5Ј-TTCACCACCATGGAGAAGGC-3Ј; reverse, 5Ј-GGCATGGACTGTGGTCATGA-3Ј), 28 S (23). Optimized PCR consisted of 40 cycles of amplification at 95°C for 15 s followed by amplification at 60°C for 1 min using the Abi Prism 7700 Sequence detection system (Qiagen). Significant PCR fluorescent signals were normalized for each sample to a PCR fluorescent signal obtained using GAPDH or 28 S rRNA as control. The comparative ⌬⌬C T method was used for the relative mRNA quantification.
Relative quantification of the target normalized to an endogenous reference gene (GAPDH or 28 S) and to a relevant control was calculated as follows: relative quantification ϭ 2 Ϫ⌬⌬CT , with ⌬⌬C T defined as the difference between the mean ⌬C T (sample) and the mean ⌬C T (control) , and ⌬C T as the difference between the mean C T (PPAR␣) and the C T (GAPDH or 28 S) as the endogenous control (C T ϭ threshold cycles).
Relative L-FABP Protein Quantification-Liver homogenates were prepared in ice-cold 0.154 M KCl, 0.01 M phosphate buffer, pH 7.4, using a glass Teflon potter. Proteins were quantified using the BCA Assay kit according the manufacturer's protocol (Uptima, Interchim). 100 g of sample was used for L-FABP dosage using the Rat L-FABP enzymelinked immunosorbent assay test kit (Hycult Biotechnology).
Transfection Assays-Caco-2 cells were plated in 6-well plates in Dulbecco's modified Eagle's medium supplemented with 10% FCS at 40 -50% confluency. Transfection mixes contained 4 g of L-FABPchloramphenicol acetyltransferase (CAT) or mPPAR␣-CAT reporter plasmid and 500 ng of ␤-galactosidase expression vector. Cells were transfected overnight by the calcium phosphate precipitation method. The medium was changed, and Dulbecco's modified Eagle's medium was supplemented with 1% of lipoprotein-depleted serum alone or with various simvastatin concentrations (from 2.4 -24 M) was added to the cells for an additional 24-h period. Cell extracts were prepared and assayed for CAT and ␤-galactosidase activities.
Statistical Analysis-The results are expressed as means Ϯ S.E. Statistically, significant differences between groups were determined by Student's t test.

Simvastatin Induces L-FABP mRNA Levels in Primary Rat
Hepatocytes-To determine whether the expression of L-FABP gene is regulated by statins, rat hepatocytes were cultured in the presence of various concentrations of simvastatin. As shown in Fig. 1, L-FABP mRNA levels were increased in a dose-dependent manner by simvastatin. Indeed, a 1.9-and 2.9-fold rise of L-FABP mRNA was found after incubation with 12 and 24 M simvastatin, respectively.
Simvastatin-responsiveness of the L-FABP Promoter Is Mediated by Its PPRE-Statins act on gene transcription by activating SREBP-2, which specifically up-regulates genes containing sterol regulatory elements (SRE) and/or E-boxes (1,25). Three putative SREBP-binding sites were localized in the rat L-FABP promoter by in silico analysis: two SRE-like sequences located at positions Ϫ801/Ϫ790 and Ϫ245/Ϫ233, respectively, and an E-box at position Ϫ306/Ϫ301 (Fig. 2A). To explore their respective functionality, the effect of progressive 5Ј-deletions of the L-FABP promoter on the capacity of simvastatin to transactivate the CAT reporter gene was determined by transfection analysis. Surprisingly, activation by simvastatin was observed even when constructs that did not contain any SRE or E-box (Fig. 2B). This unexpected result demonstrates that these putative sequences are not SREBP-binding sites. The fact that the L-FABP 75 wt sequence contains a PPRE known to be crucial for the regulation of the L-FABP gene by fibrates led us to evaluate the possible involvement of this PPAR-binding site in this regulation. Interestingly, mutations in the PPRE (L-FABP 275 mut and L-FABP 275 del ) fully abolished the activation of the L-FABP promoter by simvastatin (Fig. 2C). This unresponsiveness was not caused by functional alterations of the promoter constructs used because the L-FABP gene regulator hepatocyte nuclear factor 1 (26) was still able to transactivate a reporter gene driven by the mutated or deleted promoters (data not shown). Taken together these data suggest that the PPRE is implicated in the up-regulation of L-FABP by simvastatin.
L-FABP Gene Expression Is Synergistically Induced by Simvastatin and the PPAR␣ Agonist Wy14,643-To assess the involvement of PPAR␣ in the statin-mediated induction of L-FABP gene expression, Fao cells were cultured with simvastatin and/or the specific PPAR␣ agonist Wy14,643. This cell line was chosen because it was previously used to study PPAR-mediated regulation of the L-FABP gene by LCFA (27). As observed in primary rat hepatocyte cultures, simvastatin robustly induced L-FABP mRNA levels to an extent similar as Wy14,643. Co-treatment with simvastatin and the PPAR␣ ag- onist triggered a synergistic rise of L-FABP mRNA levels, which were induced by more than 60-fold compared with controls (Fig. 3). Interestingly, similar effects were also found for ACO, the rate-limiting enzyme of peroxisomal ␤-oxidation, a prototypical PPAR␣ target gene (28). These observations demonstrate that this regulatory pathway is not restricted to the L-FABP gene (Fig. 3).
Statin-mediated Induction of L-FABP Gene Expression Is Disrupted in PPAR␣-null Mice-To evaluate the physiological relevance of these in vitro data, wild-type and PPAR␣-null mice were force-fed for 5 days with simvastatin and/or fenofibrate. In wild-type mice, in line with the results obtained in Fao cells, simvastatin and fenofibrate alone up-regulated L-FABP mRNA levels, and co-treatment exerted a drastic synergistic effect (Fig. 4A). These changes were associated with a rise in L-FABP protein amounts, although combined treatment did not produce an additive induction (Fig. 4B). These treatments also induced ACO mRNA levels (Fig. 4C). Interestingly, whereas plasma triglycerides did not change when simvastatin or fenofibrate were given alone, combined treatment resulted in a significant drop in plasma TG levels (81 Ϯ 5 in simvastatin/ fenofibrate treated animals versus 117 Ϯ 12 mg/dl in controls, Fig. 5). By contrast, deletion of the PPAR␣ gene fully abolished the statin-mediated up-regulation of L-FABP expression (mRNA and protein levels, Fig. 4, A and B), ACO mRNA levels (Fig. 4C) and the decrease of plasma TG (Fig. 5), demonstrating the crucial role of PPAR␣ in the induction of L-FABP gene expression by statins.
Simvastatin Induces PPAR␣ mRNA Levels and PPAR␣ Promoter Activity-The effect of simvastatin on PPAR␣ mRNA levels was evaluated both in vitro and in vivo. As shown in Fig.  6A, expression of this nuclear receptor was significantly induced by simvastatin in Fao cells as well as in wild-type mice. To further investigate the mechanism by which this regulation takes place, transfection studies were performed using a construct containing the mouse PPAR␣ promoter cloned upstream of the CAT reporter gene. Increasing concentrations of simvastatin produced a dose-dependent transactivation of the reporter gene (Fig. 6B), which strongly supports the hypothesis of a transcriptional control of PPAR␣ gene expression by statins. DISCUSSION 3-Hydroxy-3-methylglutaryl CoA reductase inhibitors (statins) are widely used to correct hypercholesterolemia in humans. Statins act through the activation of SREBP-2 that controls genes involved in cholesterol homeostasis (1). Statin treatment also normalizes blood TG and NEFA levels, but the mechanism is not yet fully understood. Although the expression of the LDLr, known to be involved in the LDL clearance, is induced by statins (29), this change cannot fully explain the hypotriglyceridemic action of these drugs. Indeed, TG levels are still decreased by statins in patients suffering from familial hypercholesterolemia in which the LDLr is not functional (30,31). This observation suggests the existence of additional statin target genes that specifically control TG metabolism. In line with this hypothesis, statins are reported to induce the expression of LPL, an endothelial enzyme responsible for the hydrolysis of the TG-rich lipoproteins, chylomicrons, and very low density lipoproteins (13,32,33). This effect is further enhanced by a statin-mediated decreased in expression of apolipoprotein CIII (apo CIII), a potent inhibitor of LPL activity (13). Interestingly, both LPL (positive) and apoC-III (negative) are PPAR␣ target genes (34 -37). In hepatocytes, statins also inhibit the microsomal triglyceride transfer protein (MTP) gene that encodes a lipid-binding protein implicated in very low density lipoproteins assembly (38,39). Moreover, statins stimulate the ␤-oxidative pathways through mechanisms that re-main to be determined (40,41). In the present study, we demonstrate for the first time that L-FABP gene expression is also up-regulated by statins, an effect that might contribute to the hypolipidemic action of these drugs by facilitating LCFA influx and trafficking into hepatocytes. Collectively, these data strongly suggest that the hypotriglyceridemic effect of statins results from of an integrative regulatory pathway affecting a set of genes encoding for enzymes and lipid-binding proteins responsible for the hydrolysis of TG-rich lipoproteins and the cellular uptake and metabolic fate of LCFA.
In hepatocytes, the cholesterol depletion induced by statins triggers the proteolytic activation of the SREBP transcription factors, which after translocation into the nucleus modulate the transcription rate of sterol target genes (1). Surprisingly, although three putative SREBP-binding sites were identified by in silico analysis of the L-FABP gene, they were apparently non-functional in the context of the natural rat promoter. By contrast, statin regulation required the DR1 Ϫ67/Ϫ55 sequence, which confers PPAR-responsiveness to the L-FABP gene (42). The physiological relevance of this finding was confirmed using PPAR␣-null mice. Indeed, deletion of the PPAR␣ gene fully abolished the statin-mediated up-regulation of L-FABP expression (mRNA and protein levels) demonstrating a crucial role for PPAR␣.
PPAR␣ activity is regulated at various levels including at the transcriptional level by ligand activation, cofactor recruitment, and by post-transcriptional modification, such as ubiquitination and phosphorylation (43). The inhibition of PPAR␣ phosphorylation by cerivastatin was recently reported to be respon-sible for the induction of apolipoprotein A-I gene expression (2). This regulation was accounted for by a statin-dependent inhibition of geranylgeranylation of Rho-A protein (2), known to play a role in the c-jun NH 2 -terminal kinase and p38 mitogen- Controls received the vehicle alone (1% carboxymethyl cellulose). A, L-FABP mRNA levels were evaluated by Northern blotting using 20 g of total RNAs from liver. The bar graph represents L-FABP data normalized to 18 S rRNA. B, L-FABP protein expression was determined by enzyme-linked immunosorbent assay. C, ACO mRNA levels were evaluated by Northern blot analysis using 20 g of total RNAs from liver. Data are normalized to 18 S rRNA. Means Ϯ S.E.; *, p Ͻ 0.05; ***, p Ͻ 0.001. activated protein kinase cascades (44,45). It is likely that such a regulation might also occur for other genes including L-FABP. However, the induction of PPAR␣ gene expression by statins might constitute an alternative complementary mechanism. Indeed, simvastatin induced a significant rise in PPAR␣ mRNA levels both in vitro and in vivo (Fig. 6). These data are in line with previously published results obtained in endothelial cells (46) and in hypertriglyceridemic rats treated with statins (9). A transcriptional mechanism for this regulation is suggested by the fact that simvastatin is able to transactivate the PPAR␣ promoter (Fig. 6). Further experiments are required to determine whether PPAR␣ is regulated by a direct SREBP-dependent mechanism, as shown for PPAR␥ (47), or whether an indirect regulation occurs.
A synergistic action on L-FABP mRNA levels was observed upon combination treatment (statins plus PPAR␣ agonists) in both hepatoma Fao cells and in livers from wild-type mice (Fig.  4). Such an effect, which has already been pointed out by others (2,48), is consistent with the existence of a cross-talk between the PPAR␣ and statin signaling pathways. This observation A, PPAR␣ mRNA levels were evaluated in Fao hepatoma cells and in liver from wild-type mice by real time quantitative reverse transcriptase-PCR. Results were normalized to GAPDH (Fao cells) and to 28 S (for mice). Means Ϯ S.E., n ϭ 6; *, p Ͻ 0.05. B, Caco-2 cells were transiently transfected with the mouse PPAR␣ promoter. Cells were transfected in 10% FCS medium. 12 h after transfection, the cells were cultured for an additional 24 h in a medium containing 1% lipoprotein-deprived serum in presence of 2.4, 12, or 24 M simvastatin. CAT activity is normalized to ␤-galactosidase activity. Results are represented as arbitrary units of normalized CAT activity. Means Ϯ S.E., n ϭ 3. could provide a molecular basis and a scientific rationale for the therapeutic association of these drugs.
Because PPAR␣ controls cellular fatty acid ␤-oxidation, we speculate that the unexplained rise in LCFA oxidation activity by statins (40,49,50) is at least in part mediated via PPAR␣ activation. This hypothesis is strongly supported by the fact that the induction of ACO mRNA levels by simvastatin is fully suppressed in PPAR␣-null mice (Fig. 4C). Statin-mediated induction of L-FABP might contribute to the metabolic utilization of NEFAs because L-FABP is required for NEFA uptake and its subsequent ␤-oxidation (20). Moreover, hepatic TG synthesis is positively correlated with hepatic LCFA concentration (51). L-FABP up-regulation by statins might also contribute to a decrease of TG synthesis and thereby to the reduction of plasma very low density lipoproteins concentrations. Therefore, the regulation of L-FABP by statins could also contribute to the effects of statins on blood NEFA concentrations, which could be beneficial in patients suffering from diabetes or the metabolic syndrome, pathologies that are characterized by dyslipidemia and high NEFA levels (52,53).
In conclusion, L-FABP induction by statin treatment might provide a mechanism contributing to the hypolipidemic effect of these drugs. Our data strongly suggest that PPAR␣ is involved in this phenomenon, which could be caused at least in part by a transcriptional induction of the PPAR␣ gene by statins.