SREBP2 Activation Induces Hepatic Long-chain Acyl-CoA Synthetase 1 (ACSL1) Expression in Vivo and in Vitro through a Sterol Regulatory Element (SRE) Motif of the ACSL1 C-promoter*

Long-chain acyl-CoA synthetase 1 (ACSL1) plays a key role in fatty acid metabolism. To identify novel transcriptional modulators of ACSL1, we examined ACSL1 expression in liver tissues of hamsters fed a normal diet, a high fat diet, or a high cholesterol and high fat diet (HCHFD). Feeding hamsters HCHFD markedly reduced hepatic Acsl1 mRNA and protein levels as well as acyl-CoA synthetase activity. Decreases in Acsl1 expression strongly correlated with reductions in hepatic Srebp2 mRNA level and mature Srebp2 protein abundance. Conversely, administration of rosuvastatin (RSV) to hamsters increased hepatic Acsl1 expression. These new findings were reproduced in mice treated with RSV or fed the HCHFD. Furthermore, the RSV induction of acyl-CoA activity in mouse liver resulted in increases in plasma and hepatic cholesterol ester concentrations and reductions in free cholesterol amounts. Investigations on different ACSL1 transcript variants in HepG2 cells revealed that the mRNA expression of C-ACSL1 was specifically regulated by the sterol regulatory element (SRE)-binding protein (SREBP) pathway, and RSV treatment increased the C-ACSL1 abundance from a minor mRNA species to an abundant transcript. We analyzed 5′-flanking sequence of exon 1C of the human ACSL1 gene and identified one putative SRE site. By performing a promoter activity assay and DNA binding assays, we firmly demonstrated the key role of this SRE motif in SREBP2-mediated activation of C-ACSL1 gene transcription. Finally, we demonstrated that knockdown of endogenous SREBP2 in HepG2 cells lowered ACSL1 mRNA and protein levels. Altogether, this work discovered an unprecedented link between ACSL1 and SREBP2 via the specific regulation of the C-ACSL1 transcript.

In mammals, the predominant long-chain fatty acids (LCFAs) 3 have 16 and 18 carbons with different degrees of sat-uration. Prior to entering various metabolic pathways, free LCFAs must be esterified with coenzyme A by members of the long-chain acyl-CoA synthetase (ACSL) family (1,2). This activation process requires ATP, CoA, and LCFAs. Activated fatty acids can then undergo degradation through ␤-oxidation, be utilized for cellular lipid synthesis, or serve as lipid anchors for protein modifications (3)(4)(5)(6)(7)(8).
The mammalian ACSL family consists of five distinct members, including ACSL1, ACSL3, ACSL4, ACSL5, and ACSL6. These isoforms differ substantially in their substrate preference, tissue distribution, and subcellular compartmentation as well as in their responses to nutritional status (9), all of which contributes to the unique functions of each isoform in a particular tissue or a cell type.
ACSL1 has been the most extensively studied isoform of this family since it was cloned from rat liver in 1990 (10) and has broad substrate specificity for saturated FAs of 16 and 18 chain lengths and unsaturated FAs of 16 -20 carbon atoms. ACSL1 is the abundant isoform in major metabolic tissues, including liver, heart, adipose, and muscle. Whereas in vitro studies carried out in different cell lines have reported similar as well as conflicting functions of ACSL1 in mediating FA oxidation, FA uptake, and triglyceride (TG) synthesis (9,11,13), recent animal studies with tissue-specific knock-out of Acsl1 revealed some definitive functions of this enzyme. In heart and adipose tissues, Acsl1 deficiency generated a profound decrease in FA oxidation without an impact on [ 14 C]oleate incorporation into TG or phospholipid (14,15). In skeletal muscle, Acsl1 deficiency caused a 91% loss of ACSL specific activity and 60 -85% decreases in muscle FA oxidation. Furthermore, systemic glucose homeostasis became severely compromised in mice lacking Acsl1 in skeletal muscle (16).
Despite the recent advance in our understanding of the tissue-specific functions of ACSL1, to date, only limited information is available regarding the regulation of ACSL1 at transcriptional levels. It is reported that rat Acsl1 is regulated by the PPAR family members through a PPAR-responsive element (PPRE) embedded in the C-promoter region of the Acsl1 gene (17,18). In particular, treating rats with fenofibrate, an activator of PPAR␣, increased Acsl1 expression in liver and kidney but not in heart or muscle.
In mammals, both FA and cholesterol biosynthetic processes are controlled by a common family of transcription factors designated sterol regulatory element-binding proteins (SREBPs). The family consists of three different SREBP proteins: SREBP1a, SREBP1c, and SREBP2. Together, SREBPs transcriptionally activate a cascade of enzymes required for endogenous cholesterol, FA, TG, and phospholipid synthesis (19,20). The activity of SREBPs is regulated by cellular sterol levels. SREBPs are synthesized as inactive precursors localized to the membrane of the endoplasmic reticulum. In the inactive state within the endoplasmic reticulum, the C-terminal domains of the SREBPs interact with another membrane protein, SREBP cleavage-activating protein, which functions as a sterol sensor. Under low sterol conditions, SREBP cleavage-activating protein escorts the precursor forms of SREBPs from the endoplasmic reticulum to the Golgi, where they are processed by two membrane-associated proteases, the site 1 and site 2 proteases, which release the NH 2 -terminal transcriptionally active mature form of the SREBPs from the precursor proteins (21,22). The mature SREBPs translocate to the nucleus, where they bind to the SRE motifs and E-boxes within promoters of SREBP target genes (23). Interestingly, despite the key roles of ACSL1 in FA metabolism, an intrinsic relationship between ACSL1 and SREBPs has not been firmly established.
In this current study, we utilized normolipidemic and hypercholesterolemic hamsters and mice as in vivo models to examine ACSL1 expression levels in liver tissues under conditions of suppression of SREBP pathways by dietary cholesterol and activation of SREBP2 by treating animals with a statin. These in vivo studies revealed a strong correlation of ACSL1 expression and activity with SREBP2. By using human hepatoma-derived HepG2 as a model system, we identified ACSL1 transcript variant C-ACSL1 as a novel molecular target of SREBP2 and further mapped the SREBP2 binding site to a novel SRE motif located in the proximal region of the C-ACSL1 promoter. Finally, by using a specific siRNA targeting SREBP2, we demonstrated that knockdown of endogenous SREBP2 in HepG2 cells directly reduced ACSL1 mRNA and protein levels.

Experimental Procedures
Cells and Reagents-HepG2 cells were obtained from ATCC. Cholesterol and 25-hydroxycholestrol were purchased from Sigma, and rosuvastatin (RSV) was purchased from AK Scientific Inc. (Mountain View, CA).
Animals, Diets, and RSV Treatment-All animal experiments were performed according to procedures approved by the Veterans Affairs Palo Alto Health Care System Animal Care and Use Committee. Ten-week-old male golden Syrian hamsters were purchased from Harlan. Hamsters were housed (2 animals/cage) under controlled temperature (72°F) and lighting (12-h light/dark cycle). Animals had free access to autoclaved water and food. For the diet study, 18 hamsters were divided into three diet groups of a normal diet (ND), a high fat diet (HFD) (TD.88137 from Harlan; ϳ42% of total calories from fat, 0.15% cholesterol), and a high cholesterol and high fat diet (HCHFD) (D12336, 1.25% cholesterol; Research Diets, Inc., New Brunswick, NJ) for 2 weeks. The main ingredients of these diets are summarized in Table 1. At the experimental termina-tion, after a 16-h fast, hamsters were anesthetized, and terminal blood samples and tissue samples were collected.
For the RSV treatment, 18 hamsters were fed a high fructose diet (60% fructose; Diets, Inc., Bethlehem, PA) for 3 weeks. While continuously on the fructose diet, hamsters were randomly divided into two groups (n ϭ 9/group) and were given a daily dose of 20 mg/kg RSV or the vehicle (0.5 ml of 10% 2-hydroxypropyl-␤-cyclodextrin in autoclaved water) by oral gavage. The drug treatment lasted 7 days. Twenty-four h after the last dosing, all animals were sacrificed.
For the mouse study, 18 adult male C57BL/6J mice were divided into three groups; group one (n ϭ 6) were fed the HCHFD for 2 weeks, group two (n ϭ 6) were fed with an ND and orally administered with vehicle, and group three (n ϭ 6) were fed an ND and were orally administered with RSV at a daily dose of 20 mg/kg for 2 weeks. At the end of the experiment, mice were fasted for 4 h, and blood samples were collected for the analysis of serum total cholesterol (TC) and free cholesterol (FC). Liver tissue samples were collected and stored at Ϫ80°C for the analysis of mRNA and protein expression and lipid contents.
Measurement of Serum and Hepatic Lipid Levels-Serum was isolated at room temperature and stored at Ϫ80°C. Extraction of lipids from liver tissues was performed as described (24). Standard enzymatic methods were used to determine TC, TG, and FC levels of serum and hepatic lipids using commercially available kits purchased from Stanbio Laboratory (Boerne, TX). Each sample was assayed in duplicate. Cholesterol ester (CE) and ester ratio were calculated by using the following formulas provided in the kit manual.

RNA Isolation and Quantitative RT-PCR (qRT-PCR)-Total
RNA isolation, generation of cDNA, and qRT-PCR were conducted as reported previously (24). Each cDNA sample was assayed in duplicate. The correct size of the PCR product and the specificity of each primer pair were validated by examination of PCR products on an agarose gel. Species-specific primer sequences used in qRT-PCR are listed in Table 2. Target mRNA expression in each sample was normalized to the housekeeping gene GAPDH. The 2 Ϫ⌬⌬Ct method was used to calculate relative mRNA expression levels.
Western Blot Analysis-Approximately 50 mg of frozen tissue was homogenized in radioimmune precipitation assay buffer containing 1 mM PMSF and protease inhibitor mixture (Roche Applied Science). After protein quantitation using BCA protein assay reagent (Pierce), 50 g of homogenate proteins from individual tissue samples were resolved by SDS-PAGE and transferred to nitrocellulose membranes. The anti-human ACSL1 antibody (catalog no. ab76702) was obtained from Abcam (Cambridge, MA). The rabbit anti-SREBP2 polyclonal antibodies recognizing both the precursor form and the nuclear mature form of SREBP2 have been described in our previous studies (26). The membranes were reprobed with an anti-␤actin (catalog no. A1978, Sigma) antibody or with anti-GAPDH (catalog no. AM4300, Thermo Fisher). Immunoreactive bands of predicted molecular mass were visualized using a Super-Signal West Femto Maximum Sensitivity Substrate ECL kit (Thermo Scientific) and quantified with Alpha View software with normalization by signals of ␤-actin.
Measurement of Tissue Acyl-CoA Synthetase Activity-Approximately 50 mg of frozen tissue were homogenized in a buffer containing 20 mM HEPES, 1 mM EDTA, and 250 mM sucrose, pH 7.4. After a centrifugation at 16,000 rpm, cell lysates were collected, and protein concentrations of cell lysates were determined by the BCA method (Pierce) prior to the ACSL activity assay. The activity assay incubation mixture contained 175 mM Tris-HCl, pH 7.4, 8 mM MgCl 2 , 5 mM dithiothre-itol, 1 mM ATP, 0.2 mM CoASH, 0.5 mM Triton X-100, 10 M EDTA, and 50 M palmitate mixed with 0.1 Ci of [ 3 H]palmitic acid (25). The reaction was initiated by adding the homogenized sample and terminated by adding 1 ml of Dole's reagent as described previously (24). Generated [ 3 H]acyl-CoA was extracted, and the radioactivity was determined in a scintillation counter. The radioactivity in the reaction that contained all components but omitted homogenate was included as a negative control.
ACSL1 Promoter Cloning and Reporter Construction-A 2.5-kb DNA fragment covering the 5Ј-flanking region from Ϫ2215 to ϩ300, relative to the 5Ј end of exon 1C of the human  Table 2. The PCR was performed using the Invitrogen High Fidelity PCR system. PCR products were initially cloned into the TOPO-TA PCR 2.1 cloning vector (Invitrogen). After digestion with SacI and XhoI, the insert was subcloned into pGL3 basic vector at the SacI and XhoI sites, thereby generating the reporter vector ACSL1 promoter C-Luc.
Luciferase Reporter Activity-HepG2 cells seeded in 96-well culture plate (3 ϫ 10 4 cells/well) were transfected with 0.1 g of ACSL1 promoter C-Luc, pLDLR234, or pGL3-basic vector and cotransfected with 10 ng of pCMV-␤-Gal as an internal control by using Polyjet transfection reagent (SignaGen Laboratories, Gaithersburg, MD). Twenty-four h after transfection, cells were cultured in medium containing 0.5% FBS for 12 h and then treated with DMSO, cholesterol (10 g/ml), or RSV (5 M) for 24 h. Luciferase activities were assayed using the Single-Luciferase Reporter Assay System (Promega, Madison, WI). For ␤-galactosidase activity, 50 l of cell lysates were added to 50 l of ␤-galactosidase 2ϫ assay buffer, and the mixture was incubated for 30 min at 37°C. The absorbance at 420 nm was recorded. Promoter luciferase activity was normalized by ␤-galactosidase activity.
Electrophoretic Mobility Shift Assay-EMSA was performed using a LightShift chemiluminescent EMSA kit (Pierce). Briefly, 200 fmol of 5Ј-biotin end-labeled double-stranded oligonucleotide probes were incubated with 0.2 g of purified recombinant nSREBP2 protein in 1ϫ binding buffer containing 80 mM Hepes, pH 7.5, 500 mM KCl, 5 mM MgCl 2 , 10 mM DTT, 0.8 mM EDTA, 2.5% (v/v) glycerol, 50 ng/l poly(dI-dC), and 0.5% Non-idet P-40 in a total of 20 l of reaction mixture for 20 min at room temperature. DNA-protein complexes were subjected to electrophoresis on a 5% TBE gel retardation gel (Bio-Rad) in 0.5ϫ TBE buffer, transferred onto the positively charged nylon membrane (GE Healthcare), UV-cross-linked, and visualized by a chemiluminescent nucleic acid detection module (Pierce). For the specific oligonucleotide and nonspecific competition assay, the competing and non-competing unlabeled doublestranded oligonucleotide were added to the reaction with the biotin-labeled probe at 100-fold molar excesses. The sense sequences of biotinylated probes and competitors used in the EMSA assays are listed in Table 2. Purified His-tagged recombinant nSREBP2 protein was used in the EMSA. To purify the SREBP2 protein, pCMV-His-nSREBP2 plasmid was transfected into HEK-293A cells, and the protein was purified using the Probond purification system (Invitrogen), following the manufacturer's instructions.
Chromatin Immunoprecipitation (ChIP) Assay-The ChIP assay was performed using the ZymoSpin ChIP kit (Zymo Research), following the manufacturer's guidelines. Briefly, HepG2 cells were seeded in 100-mm dishes and transfected with 5 g of plasmid of pCMV-His-nSREBP2 or pCMV-His empty vector as a control. After 48 h of transfection, cells were trypsinized, resuspended in PBS at a dilution of 6 ϫ 10 6 cells/ ml, and fixed with 1% formaldehyde (Sigma) for 10 min. To obtain nuclear lysates, fixed cells were resuspended in 500 l of chromatin shearing buffer and sonicated at 4°C in a Bioruptor 300 instrument (Diagenode, Inc.) for 12 cycles of 30 s on/30 s off at a high setting with intermittent vortex mixing. Chromatin containing nuclear lysates (100 l) was incubated overnight

Mouse primers
with 3 g of either mouse anti-His (Ab18184, Abcam) or isotype control IgG antibodies (Santa Cruz Biotechnologies, Inc.), and complexes were immunoprecipitated using protein A magnetic beads. ChIP DNA was eluted, cross-linking was reversed, and protein-free DNA was purified before PCR amplification with site-specific primers covering the ACSL1 promoter C SRE region. Primer sequences for the ChIP assay are listed in Table 2. SREBP2 siRNA Transfection-siRNA duplexes (SREBP2 siRNA (sc-36559) and siRNA control (sc-44231)) were obtained from Santa Cruz Biotechnology. HepG2 cells were seeded in 6-well plates (3 ϫ 10 5 cells/well) and transfected with SREBP2 siRNA or control siRNA duplexes at concentrations of 20 nM using siPORT TM NeoFX TM transfection agent (Invitrogen), according to the manufacturer's instructions. Transfection complexes were performed in serum-free OptiMEM (Invitrogen). Cell medium was changed 24 h after transfection. After 72 h of transfection, cells were harvested for protein and RNA expression analysis.
Statistical Analysis-Values are presented as the mean Ϯ S.E. Significant differences between diet groups were assessed by two-tailed Student's t test (nonparametric Mann-Whitney test) or one-way analysis of variance with Dunnett' post-test. Statistical significance is displayed as p Ͻ 0.05 (*), p Ͻ 0.01 (**), or p Ͻ 0.001 (***).

Repression of ACSL1 Expression in Hamsters by Dietary
Cholesterol-To determine whether ACSL1 expression could be affected by dietary cholesterol, we fed hamsters three different diets (ND, HFD, or HCHFD) for 2 weeks. Measurement of serum TC and TG as well as hepatic TC and TG levels confirmed the hyperlipidemic status of hamsters in the HCHFD group and HFD group (Table 3). Previous studies have shown that ACSL1 expression was induced by a specific ligand of PPAR␣. To determine whether the hyperlipidemic diets could affect Acsl1 gene expression in liver tissue through PPAR␣, we performed real-time qRT-PCR to measure the mRNA levels of Acsl1, Ppar␣, Srebp1, and Srepb2 in liver tissues of all diet groups. The results showed that Acsl1 mRNA levels were slightly decreased by HFD but were lowered by ϳ40% by HCHFD, whereas Ppar␣ mRNA levels were the same among the three diet groups (Fig. 1A). The mRNA levels of Srebp2 were reduced in HFD and HCHFD groups, whereas the levels of Srebp1 were elevated by both hyperlipidemic diets.
To further examine the association between ACSL1 and SREBP2, we examined protein levels of ACSL1 and SREBP2 in the liver homogenates of the three diet groups by Western blot analysis (Fig. 1B). Compared with the ND group, hepatic ACSL1 and mature SREBP2 protein abundances were reduced to 31% (p Ͻ 0.01) of control and 37% (p Ͻ 0.001) of control, respectively, by HCHFD feeding but not by HFD feeding. Altogether, these results provided the first evidence that ACSL1 expression in vivo was down-regulated by dietary cholesterol along with the active form of SREBP2.  We next measured ACSL activity of individual homogenates of liver tissue from ND, HFD, and HCHFD groups. ACSL activity assays using liver homogenate and [ 3 H]palmitic acid (16:0) demonstrated a 16% (p Ͻ 0.05) reduction of palmitoyl-CoA synthetase activity in the HCHFD group (Fig. 1C). Collectively, these results demonstrated that dietary cholesterol exerted inhibitory effects on the expression levels of Acsl1 mRNA and protein, thereby leading to a reduced ACSL enzymatic activity.
Induction of Hepatic ACSL1 Expression in Hamsters by RSV Treatment-The above dietary studies in hamsters revealed a positive correlative relationship between ACSL1 and SREBP2. To examine the direct effect of SREBP2 activation on Acsl1 gene transcription in vivo, we examined hepatic Acsl1 mRNA and protein expressions in hamsters that were on a fructose diet for 3 weeks and were then treated with 20 mg/kg RSV or vehicle for 7 days (26). Fig. 2A shows that ACSL1 protein levels were 71% (p Ͻ 0.01) higher in RSV-treated livers as compared with the vehicle group. qRT-PCR measurement of hepatic Acsl1 mRNA levels corroborated the results of Western blot analysis with a 75% increase over the vehicle control group (Fig. 2B).
ACSL1 Expression Is Down-regulated in Livers of Mice Fed HCHFD and Up-regulated by RSV Treatment-We further examined the regulation of ACSL1 expression and activity in livers of mice that were fed the HCHFD or treated with RSV for 2 weeks. Fig. 3, A and B, shows that the HCHFD significantly suppressed ACSL1 protein expression by more than 50% (p Ͻ 0.001) and ACSL specific activity by 16% (p Ͻ 0.05), respectively. SREBP2 protein expression was also significantly suppressed by the HCHFD diet (Fig. 3A). In contrast, the amount of ACSL1 protein and specific activity were up-regulated significantly by 160% (p Ͻ 0.001) and 21% (p Ͻ 0.001) (Fig. 3, A and B) in livers of mice treated with RSV compared with vehicle treatment. The changes in Acsl1 and Srebp2 mRNA levels correspond to their protein expression; HCHFD diet dramatically suppressed Acsl1 and Srebp2 mRNA levels, whereas RSV treat- Hamsters on a fructose diet were orally administered vehicle (n ϭ 9) or 20 mg/kg RSV (n ϭ 9) once per day for 7 days. Twenty-four h after the last treatment, all animals were euthanized, and liver tissues were harvested. In A, total protein extracts were individually prepared from 6 randomly chosen liver samples/group. Equal amounts of homogenate proteins (50 g) were resolved by SDS-PAGE, and ACSL1 proteins were detected by immunoblotting using anti-ACSL1 antibody. The membrane was reprobed with an anti-␤-actin antibody. In B, total RNA was isolated from each liver sample, and relative mRNA abundances of Acsl1 and Srebp2 were determined by conducting qRT-PCR and normalized to Gapdh. Values are the mean Ϯ S.E. (error bars) of 9 animals/group. **, p Ͻ 0.01 as compared with the vehicle group.

FIGURE 3. Opposite regulations of ACSL1 expression and activity in liver tissues of mice by HCHFD and RSV.
Male C57BL/6J mice were fed an HCHFD for 2 weeks (n ϭ 6), or fed an ND and orally administered vehicle (n ϭ 6) or administered RSV at 20 mg/kg (n ϭ 6) for 2 weeks. Mice were sacrificed, and serum and livers were isolated at the termination of the experiment. A, individual liver homogenates were prepared, and protein concentrations were determined. 50 g of homogenate proteins of 5 randomly chosen liver samples/diet group were resolved by SDS-PAGE. ACSL1 and SREBP2 proteins were detected by immunoblotting using anti-ACSL1 and anti-SREBP2 antibodies. The membrane was reprobed with anti-␤-actin antibody. The protein abundance of ACSL1 or SREBP2 was quantified using Alpha View software. Values are the mean Ϯ S.E. of 5 samples/group. B, total ACSL enzymatic activity in tissue homogenate was measured using 4 g of homogenate proteins at 37°C in the presence of 3 H-labeled PA. Values are mean Ϯ S.E. (error bars) of 6 mice/group. C, total RNA was isolated from individual livers, and relative mRNA abundances of the indicated genes were determined by conducting qRT-PCR and normalized to Gapdh. Values are mean Ϯ S.E. of 6 mice/group. *, p Ͻ 0.05; **, p Ͻ 0.01; ***, p Ͻ 0.001 as compared with the ND group. ment significantly up-regulated mRNA expression of Acsl1 and Srebp2 (Fig. 3C).
We next examined the hepatic cholesterol distributions in all groups of mice (Fig. 4, A-D). Feeding mice the HCHFD for 2 weeks resulted in massive accumulations of liver TC, CE, and FC contents without changing the hepatic CE ratio as compared with the liver of the ND group. Interestingly, under normal diet conditions, 2-week RSV treatment significantly increased the hepatic content of CE, reduced FC, and increased liver CE ratio by 57% as compared with the control mice. Next, we examined the TC and FC in serum samples of all groups and observed changes similar to hepatic cholesterol contents (Fig. 4,  E-H). These new findings suggest that increased ACSL1 activity by RSV treatment led to enhanced CE synthesis.
In mammals, endogenous CEs are synthesized by acyl-CoA cholesterol acyltransferase (ACAT) that catalyzes the formation of cholesteryl esters from cholesterol and long-chain fatty acids. There are two known genes (Acat1 and Acat2) that encode the two ACAT enzymes, and both enzymes are present in the liver tissue of adult mice. Utilizing qRT-PCR, we measured the mRNA levels of hepatic Acat1 and Acat2 in all three groups. The results showed that the mRNA expressions of Acat1 and Acat2 were coordinately regulated with Acsl1 mRNA by RSV and HCHFD (Fig. 3C). Considering the functional roles of ACAT1 and ACAT2 in CE formation and the function of ACSL1 to provide the substrate acyl-CoAs for CE synthesis, the coordinated up-regulation of ACSL1 and ACAT might provide an underlying mechanism for the observed elevations of CE levels in liver and plasma of RSV-treated animals.
Identification of ACSL1 Transcript Variant 3 (C-ACSL1) as the Molecular Target of SREBP2-Previous studies of rat Acsl1 gene structure revealed three ACSL1 transcript variants derived from the Acsl1 gene through alterative splicing. These transcripts are under the control of its own promoter, and the C-promoter contained a PPRE sequence that was responsive to fibrate stimulation (17,18). Similar to the rat ACSL1 gene, the human ACSL1 gene gives rise to three transcript variants that all encode the full-length ACSL1 protein with 698 amino acids. The three transcripts (v1, v2, and v3) contain the same coding exon 2 but have different 5Ј-UTRs (Fig. 5A). To be consistent with the rat ACSL1 mRNA nomenclature, we designated human ACSL1 v1 as A-ACSL1, v2 as B-ACSL1, and v3 as C-ACSL1. Using HepG2 as a model system, we sought to identify the ACSL1 mRNA species that is regulated by the SREBP pathway.
HepG2 in triplicate dishes were treated with RSV or cholesterol for 24 h. Fig. 5B shows that cholesterol repressed and RSV stimulated the formation of mature SREBP2, resulting in the suppression and up-regulation of the protein levels of ACSL1 and LDLR, a well known SREBP2 target gene. We performed qRT-PCR using a common ACSL1 primer set that amplifies the same coding region of all ACSL1 mRNA species. The qRT-PCR results showed a 25% reduction by cholesterol treatment and a 2.1-fold increase by RSV treatment of ACSL1 mRNA levels in HepG2 cells (Fig. 5C), which were consistent with the results obtained from liver tissues of hamsters and mice. Next, we designed three sets of PCR primers targeting the 5Ј-UTR sequence of individual ACSL1 mRNA species and performed qRT-PCR analysis. In untreated HepG2 cells, A-ACSL1 mRNA was highly expressed; B-ACSL1 mRNA could not be detected, and C-ACSL1 mRNA was expressed at a low level of ϳ5% of A-ACSL1 abundance. However, the low abundance C-ACSL1 mRNA levels were dramatically changed in cells treated by cholesterol and RSV. Its mRNA levels declined more than 90% in sterol-treated cells and were increased 11.2-fold by RSV treat-ment (Fig. 5D). In contrast to C-ACSL1, A-ACSL1 mRNA expression was slightly reduced by cholesterol treatment, and it was not up-regulated by RSV. We repeated the qRT-PCR assays using another set of exon-specific primers to individually amplify each transcript variant and obtained identical results. These results provided the first evidence to demonstrate the specific regulation of the C-ACSL1 transcript by the SREBP pathway in hepatic cells.
Transactivation of ACSL1 Promoter C Activity by SREBP2-To further demonstrate the regulation of C-ACSL1 mRNA levels by SREBP pathway at the transcriptional level, we analyzed the nucleotide sequence within the 2.5-kb 5Ј-flanking sequence of exon 1C of the human ACSL1 gene and found one consensus SRE site along with several Sp1, E-box, and PPRE motifs immediately upstream of the 5Ј-end of exon 1C. The SRE site, 5Ј-GTGGGGTGAC-3Ј (5Ј-GTCACCCCAC-3Ј in the complementary strand), contains sequences identical to those of the SRE-1 motif found in the LDLR promoter and is located at positions Ϫ7 to ϩ1, relative to the 5Ј-end of exon 1C (Fig. 6A). Using HepG2 genomic DNA, we cloned the 2.5-kb DNA fragment covering the human C-ACSL1 proximal promoter region from Ϫ2215 to ϩ300 relative to the 5Ј-end of exon 1C to pGL3basic luciferase (Luc) vector to generate Promoter C-Luc reporters (promoter C) with SRE-1 intact (WT) or mutated. The ACSL1 promoter C constructs along with a LDLR promoter reporter (pLDLR234) were transfected into HepG2 cells and treated with RSV or cholesterol. Fig. 6B shows that the wild-type ACSL1 promoter C activity was strongly induced by RSV and was repressed by cholesterol to extents similar to the changes in LDLR promoter activity. Mutation of the SRE site in promoter C luciferase constructs rendered the promoter C totally unresponsive to either RSV or cholesterol treatment. These results strongly suggest that the SRE site mediates the cholesterol and RSV-dependent regulation of the ACSL1 promoter C activity.
Next, we co-transfected promoter C wild-type and SRE mutated constructs with plasmid pCMV-His-nSREBP2, encoding the His-tagged active nuclear form of SREBP2 (Fig. 6C). In this set of experiments, again we included pGL3-basic and LDLR promoter reporters as negative and positive controls. Compared with the mock transfection control (pCMV-His), luciferase activities of promoter C WT plasmid and pLDLR234 were increased 42-and 31-fold in HepG2 cells overexpressing nSREBP2. In contrast, promoter C mutant constructs did not respond to SREBP2 overexpression at all.
To determine whether SREBP2 could bind directly to the SRE site in the ACSL1 promoter C, first we used purified recombinant nSREBP2 protein to perform EMSA. Fig. 6D shows the result of EMSA that detected a single shifted band (Fig. 6D, lane 2) compared with the reaction mixture containing only the biotinylated probe (Fig. 6D, lane 1). Competition experiments showed that a 100-fold molar excess of unlabeled wild-type probe blocked the formation of the probe-nSREBP2 complex (Fig. 6D, lane 3), whereas 100-fold molar amounts of unlabeled mutant probe failed to compete (Fig. 6D, lane 4), thereby demonstrating the binding specificity.
To further examine the in vivo interaction of SREBP2 with the ACSL1 promoter C, we performed a ChIP assay in HepG2 cells. Because the anti-SREBP2 antibody does not work in an FIGURE 5. Regulations of ACSL1 expression in HepG2 cells by cholesterol and RSV. HepG2 cells in triplicate wells were cultured in medium containing lipoprotein-deficient bovine serum overnight and then treated with cholesterol (10 g/ml cholesterol and 1 g/ml 25-hydroxycholestetol or 5 M RSV for 24 h. A, diagram of human ACSL1 gene structure. The diagram is not drawn to scale. In B, total cell lysates were isolated, and the protein abundance of ACSL1 or LDLR was examined by Western blotting, and specific signals were quantified. In C, total RNA was isolated, and mRNA levels of authentic SREBP target genes and total ACSL1 mRNA levels were measured and presented relative to control without treatment. Data are mean Ϯ S.E. (error bars) of triplicate RNA samples with duplicate measurements of each cDNA sample. In D, expression levels of individual ACSL1 transcript variants were separately measured by qRT-PCR. The normalized mRNA level of A-ACSL1 in untreated cells is expressed as 100%, and C-ACSL1 mRNA levels are expressed relative to that. **, p Ͻ 0.01; ***, p Ͻ 0.001 as compared with untreated control.
immunoprecipitation assay, we expressed the His-nSREBP2 in HepG2 cells and used anti-His antibody to immunoprecipitate the SREBP2-bound chromatins. The empty vector pCMV-Histransfected HepG2 cells were included in this experiment as the mock control. Isolated DNA fragments were used as templates to perform the PCR with one set of primers that amplify a 181-bp fragment of the human ACSL1 promoter C region from Ϫ116 to ϩ65, encompassing the SRE site, and another set of primers to amplify a 147-bp fragment within the promoter A region of the human ACSL1 gene as a negative control. Fig. 6E shows that a strong binding signal of His-nSREBP2 to ACSL1 promoter C was detected using immunoprecipitation materials obtained from pCMV-His-nSREBP2-transfected cells, but very little background binding signal was detected using the ChIP materials of the empty vector-transfected HepG2 cells (pCMV-His). Furthermore, this strong binding was not observed in the promoter A region of the ACSL1 gene. We also did not observe specific signals using a control antibody (anti-IgG). These results demonstrated the direct binding of SREBP2 to the ACSL1 promoter C under in vivo conditions. Altogether, these results provided a molecular basis for increased mRNA levels of C-ACSL1 transcript by RSV treatment and further demonstrate that C-ACSL1 gene transcription in liver cells is activated by SREBP2 through the SRE site of promoter C.
Finally, we examined the functional role of SREBP2 in ACSL1 expression by transfecting HepG2 cells with a SREBP2 siRNA or a control nonspecific siRNA. Transfection of SREBP2 siRNA resulted in a marked decrease in SREBP2 mRNA levels (Ϫ67%) and also lowered total ACSL1 mRNA levels by 30% in HepG2 cells (Fig. 7A). Consistent with the quantitative PCR results, the protein levels of SREBP2 and ACSL1 were all reduced in SREBP2 siRNA-transfected cells as compared with the cells transfected with the scrambled control siRNA (Fig. 7B).

Discussion
Abnormality in LCFA metabolism is increasingly recognized as an underlying causal factor for cardiovascular diseases, type II diabetes mellitus, insulin resistance, obesity, nonalcoholic fatty liver disease, and metabolic syndrome (27,28). Because ACSL1 is the dominant isoform of the ACSL family in a majority of metabolic tissues, variations in ACSL1 levels of expression and activity will inevitably impact the LCFA metabolism at cellular levels and at the whole body level. In particular, changes in the upstream signaling pathway and transcriptional networks that govern ACSL1 expression might influence the entry of fatty acyl-CoAs into different metabolic pathways.
Previous studies have demonstrated a primary role of PPAR␣ in FA ␤-oxidation (29,30). Thus, it is conceivable that activation of PPAR␣ led to the increased expression of ACSL1 to provide more conjugated acyl-CoAs to be utilized as fuels through the FA oxidative pathway. On the other hand, several in vitro studies have demonstrated the activity of ACSL1 in promoting TG and phospholipid synthesis (11). Until this current study, it is unknown whether the activity of ACSL1 in promoting the entry of FA into these anabolic pathways is regulated by different upstream signals that do not involve the transcriptional activation of PPAR␣.
We set out in this study to identify transcriptional networks that might regulate ACSL1 expression under hyperlipidemic conditions with excessive amounts of LCFAs and dietary cholesterols. By profiling Acsl1 mRNA and protein expressions in liver tissues of hamsters fed a cholesterol-enriched HCHFD, an HFD, and a normal diet, our studies revealed a strong inhibitory effect of dietary cholesterol on hepatic Acsl1 mRNA and protein expressions; however, there still remains the possibility that different fat compositions between these diets could also be an influence on Acsl1 expression. A previous study using adenovirusmediated overexpression of SREBP1 has shown a positive effect of SREBP1c on ACSL5 expression in livers of diabetic mice (11). In our study, we did not observe a correlation of SREBP1 with ACSL1 at the mRNA level in vivo or in hepatic cells. Instead, we found a strong correlation of down-regulation of ACSL1 with reduced SREBP2 in the liver of hamsters.
We further confirm the correlation of ACSL1 regulation by SREBP2 in C57BL/6J mice in which HCHFD dramatically suppressed ACSL1 and SREBP2 expression. In addition, ACSL specific activity was also reduced significantly in livers of hamsters and mice. To seek additional evidence to support the role of SREBP2 in ACSL1 expression, we examined ACSL1 expression in liver tissues of hamsters and mice that were treated with RSV, the activator of the SREBP pathway. Compared with the vehicle group, Acsl1 mRNA and protein were both elevated in RSV-treated liver samples. These results from two different rodent models strongly suggest a positive role of SREBP2 in ACSL1 gene transcription.
The above new findings brought up an important question as to which ACSL1 transcript is the target of SREBP pathway. In rats, it has been demonstrated that the Acsl1 gene generates three transcripts (A-ACSL1, B-ACSL1, and C-ACSL1) by alterative transcription from three different promoters. The C-promoter was shown to be highly inducible by the PPAR␣ agonist fenofibrate in rat liver. To identify the ACSL1 transcript variants that are responsive to SREBP activation, we decided to use HepG2 cells as a model system because in HepG2 cells, ACSL1 mRNA and protein levels were repressed by sterols and induced by RSV to degrees similar to what we have observed in liver tissues. Our analysis of human ACSL1 gene structure revealed three transcripts (v1, v2, and v3) that contain the same exon 2 but different first exons that are preceded by specific 5Ј-upstream regulatory sequences. We named these three FIGURE 7. siRNA-mediated knockdown of SREBP2 decreases ACSL1 mRNA and protein levels in HepG2 cells. SREBP2 siRNA and scrambled siRNA were transfected into HepG2 cells in 6-well plates with duplicate wells per treatment condition. Two days after transfection, whole protein lysates and total RNA and were isolated, and cDNA was synthesized from total RNA. A, hepatic mRNA levels of SREBP2, ACSL1, and GAPDH were assessed by qRT-PCR using specific primers with triplicate measurement of each cDNA sample. After normalization with GAPDH mRNA levels, the relative levels are presented. The data shown are the summarized results of duplicate wells. Error bars, S.E. B, Western blotting was performed using the indicated antibodies after SREBP2 siRNA or control siRNA transfection. Results are representative of two independent transfections with similar results.
variants A-ACSL1, B-ACSL1, and C-ACSL1. By conducting qRT-PCR with variant-specific primers (Fig. 5D), we demonstrated that A-ACSL1 is highly abundant in HepG2 cells and attributes to ϳ95% of total ACSL1 mRNA species in HepG2 cells under uninduced conditions. Transcript B-ACSL1 could not be detected in HepG2 cells, and transcript C-ACSL1 was expressed at a low level, ϳ5% of the A-ACSL1 mRNA amount. However, in RSV-treated HepG2 cells, mRNA levels of A-ACSL1 did not change, but C-ACSL1 mRNA abundance increased Ͼ10-fold, which probably accounted for the ϳ2-fold increase in total ACSL1 mRNA levels when measured using the primers that recognize all ACSL1 transcripts.
Our sequence analysis of the 5Ј-flanking region of human C-ACSL1 promoter revealed a single SRE site adjacent to the 5Ј-end of exon 1C. The role of the SRE site in the sterol-dependent regulation of the ACSL1 promoter C was supported by the observation that in HepG2 cells, treatment with RSV and overexpression of the nuclear forms of SREBP-2 significantly activated the ACSL1 promoter C activities, whereas mutation in SRE resulted in a complete loss of activation by SREBP2 and statin treatment. By performing EMSA and ChIP assays, we provided direct evidence for the interaction of SREBP2 with the SRE site in the ACSL1 promoter C specifically. Finally, by utilizing siRNA targeting SREBP2, we provided functional evidence that depletion of endogenous SREBP2 led to direct inhibition of ACSL1 expression.
Although more comprehensive lipidomic analysis is needed in future studies in order to thoroughly determine the outcomes of ACSL1-produced fatty acyl-CoAs under SREBP-activated (RSV treatment) and repressed (HCHFD feeding) conditions, in the current mice study, we have detected a strong association of higher ACSL enzymatic activity with increased CE and decreased free cholesterol in liver as well as in plasma of RSV-treated animals (Fig. 4). Because SREBP2 is a master transcription activator for many genes encoding enzymes in the cholesterol de novo biosynthetic pathway, the transcriptional activation of ACSL1 by SREBP2 through a SRE site embedded in the C-ACSL1 gene promoter could provide a molecular mechanism to explain the function of ACSL1 in channeling acyl-CoAs into an anabolic pathway including cholesterol ester synthesis.
Statins have been the corner stone to treat hypercholesterolemia to reduce the risk of heart disease. The identification of ACSL1 as a new molecular target of SREBP2 may unravel a new aspect of statin efficacy, a potential benefit of statin therapy beyond the LDL cholesterol lowering.