Activation of TM7SF2 promoter by SREBP-2 depends on a new sterol regulatory element, a GC-box, and an inverted CCAAT-box

Biochimica et Biophysica Acta 1801 (2010) 587–592

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r.com/ locate /bba l ip

Activation of TM7SF2 promoter by SREBP-2 depends on a new sterol regulatoryelement, a GC-box, and an inverted CCAAT-box

Gianluca Schiavoni a, Anna Maria Bennati a, Marilena Castelli b, Maria Agnese Della Fazia b,Tommaso Beccari c, Giuseppe Servillo b, Rita Roberti a,⁎a Department of Internal Medicine, Laboratory of Biochemistry, University of Perugia, via del Giochetto, 06122 Perugia, Italyb Department of Clinical and Experimental Medicine, University of Perugia, Italyc Department of SEEA, University of Perugia, Italy

Abbreviations: SRE, sterol regulatory element; SRENF-Y, nuclear factor Y; LDLR, low density lipoprotein receLov, lovastatin; 25-OH chol, 25-hydroxycholesterol; LPDEMSA, electrophoretic mobility shift assay; ChIP, chromat⁎ Corresponding author. Tel.: +0039 075 585 7426;

E-mail address: [email protected] (R. Roberti).

1388-1981/$ – see front matter © 2010 Elsevier B.V. Adoi:10.1016/j.bbalip.2010.01.013

a b s t r a c t

Article history:Received 10 July 2009Received in revised form 20 January 2010Accepted 27 January 2010Available online 4 February 2010

Keywords:CholesterolTM7SF23beta-hydoxysterol delta14-reductaseSREBP-2Promoter

TM7SF2 gene encodes 3β-hydroxysterol Δ14-reductase, responsible for the reduction of C14-unsaturatedsterols in cholesterol biosynthesis. TM7SF2 gene expression is controlled by cell sterol levels through theSREBP-2. The motifs of TM7SF2 promoter responsible for activation by SREBP-2 have not been characterized.Using electrophoretic mobility shift assays and mutation analysis, we identified a new SRE motif, 60%identical to an inverted SRE-3, able to bind SREBP-2 in vitro and in vivo. Co-transfection of promoter–luciferase reporter constructs in HepG2 cells showed that the binding of SREBP-2 to SRE producedapproximately 26-fold promoter activation, whereas mutation of the SRE motif caused a dramatic decreaseof transactivation by SREBP-2. The function of additional motifs that bind transcription factors cooperatingwith SREBP-2 was investigated. An inverted CCAAT-box, that binds nuclear factor Y (NF-Y), cooperateswith SREBP-2 in TM7SF2 promoter activation. Deletion of this motif resulted in the loss of promoterinduction by sterol starvation in HepG2 cells, as well as a decrease in fold activation by SREBP-2 in co-transfection experiments. Moreover, co-transfection of the promoter with a plasmid expressing dominantnegative NF-YA did not permit full activation by SREBP-2. Three GC-boxes (1, 2, 3), known to bind Sp1transcription factor, were also investigated. The mutagenesis of each of them produced a decrease in SREBP-2-dependent activation, the most powerful being GC-box2. A triple mutagenized promoter construct did nothave an additive effect. We conclude that, besides the SRE motif, both the inverted CCAAT-box and GC-box2are essential for full promoter activation by SREBP-2.

BP-2, SRE-binding protein 2;ptor; FBS, foetal bovine serum;S, lipoprotein deficient serum;in immunoprecipitationfax: +0039 075 585 7428.

ll rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

TM7SF2 gene encodes 3β-hydroxysterol Δ14-reductase, a proteinof the endoplasmic reticulum catalyzing the reduction of C14-unsaturated sterol intermediates during the conversion of lanosterolto cholesterol [1]. Disruption of Tm7sf2 gene in mice did not result inimpairment of cholesterol biosynthesis, due to recovery of theenzymatic activity by lamin B receptor, a protein of the inner nuclearmembrane [2,3]. Despite this evidence, unlike lamin B receptor gene,TM7SF2 gene appears to play a crucial role in cholesterol biosynthesis,its expression being controlled by cell sterol levels [4]. Microarrayanalysis of transgenic mice indicated that this control is exertedthrough Sterol Regulatory Element-Binding Protein-2 (SREBP-2) [5].

Our previous studies demonstrated SREBP-2-dependent transactiva-tion of TM7SF2 promoter and defined the minimal region fortransactivation to occur [4]. However, the Sterol Regulatory Element(SRE) sequence retrieved by TRANSFAC analysis of this region wascharacterized by low score and exhibited very low homology withdescribed SREs.

Ever since SRE sequences were described in the promotersof low density lipoprotein receptor (LDLR), HMG-CoA synthase,and HMG-CoA reductase [6–8], the number of genes involved incholesterol homeostasis regulated through SREs present in theirpromoters has increased continuously. Concomitantly, it becameevident that SREs are characterized by quite variable sequences.Although most of them exhibit significant homology with the wellcharacterized SRE-1 of LDLR promoter, SRE-2 and SRE-3 sequenceshave also been reported [7,9].

SREBPs are weak activators and require additional transcriptionfactors to achieve optimal regulation of sterol sensitive genes. Sp1 andthe trimeric nuclear factor Y (NF-Y) are transcription factors actingas SREBP co-activators in the promoter of several genes involved infatty acids and cholesterol metabolism [10–14]. The region−200/−1

588 G. Schiavoni et al. / Biochimica et Biophysica Acta 1801 (2010) 587–592

upstream the transcription start site of TM7SF2 gene exhibits pro-moter activity [4] and contains, besides a potential SRE, Sp1 and NF-Ybinding motifs. Although SREBP-2-dependent transactivation ofTM7SF2 promoter is known [4], it has not been demonstratedwhetherthe potential SRE is able to bind SREBP-2. Due to the variable se-quences of SRE motifs, other not yet identified SREs could fulfilthis function. In this paper, we investigate the binding of SREBP-2 toTM7SF2 promoter. Using electrophoretic mobility shift assays (EMSA)and co-transfection of promoter–luciferase reporter constructs, wedemonstrated through various mutation analyses that the previouslyretrieved SRE is not able to bind SREBP-2, whereas a new SRE motifexhibits this feature, leading to promoter activation. In addition, weestablished that a GC-box and an inverted CCAAT-box are essentialfor full promoter transactivation. Chromatin immunoprecipitation(ChIP) assays confirmed the ability of TM7SF2 promoter to bindSREBP-2, NF-Y, and Sp1 in vivo.

2. Materials and methods

2.1. Materials

Minimum essential medium (MEM), Dulbecco's Modified Eagle'smedium (DMEM), foetal bovine serum (FBS), and other culturereagents were from GIBCO (Invitrogen, Milan, Italy). Lipoproteindeficient serum (LPDS) was prepared as described previously [4].Lovastatin (Lov) and 25-hydroxycholesterol (25-OH chol) werepurchased from Sigma (Milan, Italy). Complete protease inhibitorcocktail tablets, T4 DNA ligase, T4 polynucleotide kinase, andGenopure Plasmid maxi-kit were from Roche Diagnostics (Milan,Italy). Hartmann Analytic GmbH (Germany) provided [γ-32P]-dATP.MicrospinTM G-50 columns and poly(dI–dC)/poly(dI–dC) werefrom Amersham Biosciences. Dual Luciferase Reporter Assay System,pGL2-basic plasmid, and pRL-SV40 plasmid were from Promega(Madison, WI). QuikChange® Site-Directed Mutagenesis Kit and pfuTurbo DNA polymerase were from Stratagene. Lipofectamine 2000and custom oligonucleotides were purchased from Invitrogen. MBIFermentas (Lithuania) provided all other reagents.

2.2. Plasmids

The −200/−1 promoter region upstream the transcription ini-tiation site of human TM7SF2 gene was amplified from p326 plasmid[4] using Pfu Turbo DNA polymerase and the following primers:forward, 5′-ATGGTACCAAACATGTGTGGCCTCCTCCTC-3′; reverse, 5′-AAGCTTGGACACGGAACGCAGACAAGG-3′. The amplified region corre-sponds to −326/−127 of p326, which was numbered starting fromthe ATG [4]. The fragment was subcloned upstream the luciferasegene in pGL2-basic expression vector, using the KpnI and HindIIIrestriction sites (in italics) present in forward and reverse primers,respectively. The obtained p200 construct was used as template forbase substitution, insertion, or deletion mutations using the Quik-Change® site-directed mutagenesis kit (Stratagene), according to themanufacturer's protocol. Oligonucleotides designed to mutate eachelement are indicated in Table 1. The pGC-boxes1/2/3 mut, in which

Table 1Forward primers used for mutagenesis of p200. Modified bases (indicated by numbers in b

SRE (−71/−68)SRE (−113/−109)Δ Inv CAATTinv CAATT mut (−91/−89)Ins-5GC-box1 (−168/−164)GC-box2 (−152/−148)GC-box3 (−58/−54)GC-boxes1/2

all the three GC-boxes were mutated, was obtained by using first theGC-box2 primer and the pGC-box3mut as template to obtain the pGC-box2/3 mut, and then the GC-boxes1/2 primer and the pGC-box2/3mut as template. All the constructs were verified by DNA sequencing.The plasmid encoding aminoacids 1–481 of human SREBP-2 (pCS2-SREBP2) was a generous gift of Dr. T. F. Osborne (University of Cali-fornia, Irvine). The cDNA encoding SREBP-2 was amplified fromlinearized pCS2 plasmid using the forward primer 5′-GCCTCGAGATG-GACGACAGCGGCGAGCT-3′ and the reverse primer 5′-GCGGATCCT-CACCGTGAGCGGTCTACCATGC-3′ by pfu Turbo DNA polymerase. Thefragment was subcloned in bacterial expression vector pET-15b(Novagen), using the XhoI and BamHI restriction sites (in italics)present in forward and reverse primers, respectively. The plasmidΔ4NFYA13m29, expressing the dominant negative of NF-YA and thecontrol plasmid Δ4NFYA13 were a kind gift of Dr. Roberto Mantovani(Milan, Italy). The CMV-Sp1 plasmid was Addgene plasmid 12097(thanks to Dr. Robert Tjian).

2.3. Electrophoretic mobility shift assays

Human SREBP-2 was overexpressed in E. coli BL21(DE3) fromplasmid pET-15b by induction with 0.8 mM isopropyl β-D-1-thioga-lactopyranoside. After 3 h at 37 °C cells were lysed and supernatantpurified from debris by centrifugation. The supernatant was used inEMSA experiments. As a control, a supernatant from cells transformedwith empty pET-15b plasmid was used. Protein concentration ofsupernatants was determined by Bradford method [15]. Probes weredesigned to cover the −143/−1 region upstream the transcriptioninitiation site (Fig. 1A), responsive to SREBP-2 [4]. LDL receptoroligonucleotide 5′-AAAATCACCCCACTGCAAACTCCTCCCCCTGC-3′,containing the SRE-1 element, was used as positive control (LDLR-SRE-1). Annealed probes were end-labelled with [γ-32P]-dATP and T4polynucleotide kinase and then purified with MicrospinTM G-50columns. For the binding reaction, E. coli cleared lysates (18 μg pro-tein) were preincubated for 15 min at room temperature with 3 μgpoly(dI–dC)/poly(dI–dC) in a solution containing 12.5 mM Hepes(pH 7.9), 6 mM MgCl2, 50 mM KCl, 5.5 mM EDTA, 0.5 mM DTT,0.25 mg/ml BSA, 5% glycerol, and protease inhibitor cocktail, finalvolume 20 μl. Equal amounts of probes (40,000 cpm, 0.1 pmol, 1 μl)were then added and incubation continued for 30 min at roomtemperature. For competition assays, the indicated molar excess ofunlabeled LDLR-SRE-1 was added to the preincubated bindingmixture and incubated for additional 30 min at room temperatureprior to adding labelled probes. Protein–DNA complexes were thenresolved by electrophoresis through 4.5% polyacrylamide gels for 3 h,and the dried gels were subjected to autoradiography.

2.4. Cell culture and transfections

HepG2 human hepatoma cells were grown in a 5% CO2 incubatorat 37 °C in MEM supplemented with 10% FBS, 2 mM L-glutamine,100 U/ml penicillin, and 0.1 mg/ml streptomycin. HEK293 cells weremaintained in DMEM as described above. Cells grown in 24-wellplates at 60–70% confluence were transfected in duplicate with 0.8 μg

rackets) are in italics bold.

−89/GGCTGGTGTGCTGGGCCCTTGTTGGGCAGGGGGCGG/−54−128/CTGAGGCGCCCGCGTTCACGCCGCCCCCGAGCGCC/−94−104/CCCCGAGCGCCG - - - - - CTGGTGTGCTGGGCCCAGC/−69−108/CCGCCCCCGAGCGCCGAAGTGCTGGTGTGCTGGGCCC/−72−109/CCCGAGCGCCGATTGGAATAACTGGTGTGCTGGGCCC/−72−189/GCCTCCTCCTCTCGCTTGCTGAATTCGCCTTTCCGGGGGCGGGG/−146−172/GCTGGGCGGGCCTTTCCGGGAATTCGGTTTTGAAGCTGAGGC/−131−77/GGGCCCAGCATGGGCAGGGAATTCTTCCACTTAAAAACCCTGGG/−34−189/GCCTCCTCCTCTCGCTTGCTGAATTCGCCTTTCCGGGAATTCGG/−146

Fig. 1.DNA binding of SREBP-2 on TM7SF2 promoter. A) Sterol responsive sequence of human TM7SF2 gene promoter (Ref. [4]) upstream the transcription initiation site. The putativeSRE-116 motif, the new SRE motif identified in this study, an inverted CCAAT-box, and three CG-boxes are underlined. Numbers with arrows indicate the start of oligonucleotideprobes used in gel mobility shift assays. B) Gel mobility shift assays were performed with end-labelled, double stranded oligonucleotides spanning the −143/−1 region. Equalaliquots (40,000 cpm, 0.1 pmol) were incubated with extracts (18 μg protein) of E. coli transformed with pET-15b/SREBP-2 plasmid. Labelled LDLR-SRE1 was used as a control(lane a, no bacterial extract; lane b, extract of E. coli transformed with pET-15b empty vector). The indicated molar excess of unlabelled oligonucleotide LDL-SRE1 was used incompetition assays.

589G. Schiavoni et al. / Biochimica et Biophysica Acta 1801 (2010) 587–592

reporter gene constructs plus 10 ng of Renilla luciferase expres-sion vector pRL-SV40 as an internal control, using lipofectamine2000 transfection reagent. pCS2-SREBP2 plasmid (15 ng) was co-transfected in SREBP-2 transactivation experiments. In the experi-ments using Δ4NFYA13 and Δ4NFYA13m29 plasmids, the totalamount of DNA in each transfection was adjusted to 1.6 μg by addingpSG5 plasmid. Cells were collected 48 h after transfection and extractsof lysed cells were assayed for Firefly and Renilla luciferase activitiesusing the Dual Luciferase Reporter Assay System. Relative luciferaseactivities were determined as the ratio of Firefly luciferase activityof each sample to Renilla luciferase activity.When indicated, 16 h aftertransfection, cells were switched to medium supplemented with5% LPDS, or LPDS plus 10 μM lovastatin (sterol starvation condition),or LPDS plus 1.5 μg/ml 25-hydroxycholesterol (sterol rich condition),incubated for additional 36 h, and assayed as described above.

2.5. Chromatin immunoprecipitation

ChIP assays were performed on HepG2 cells grown on sterol star-vation condition (5% LPDS plus 10 μM lovastatin), using the EZ-Chipkit in accordance to the protocol provided by the manufacturer(Millipore-Upstate, USA). Briefly, after cross-linking with 1% formal-dehyde and quenching with 125 mM glycine, cells were recovered,resuspended in SDS-lysis buffer in the presence of protease inhibitors,and sonicated to obtain 200–1000 bp chromatin fragments. Aliquotsof the fragmented chromatin corresponding to about 106 cells wereprecleared with protein G-agarose. The precleared chromatin (1%)was kept as “input” and the residual was incubated overnight with thefollowing antibodies: anti-NF-YB (Genespin, Milan, Italy), anti-Sp1,mouse IgG, and rabbit IgG (Millipore-Upstate, USA), anti-SREBP-2

(generous gift of T. Osborne). Five μg of each antibody except anti-SREBP-2 andmouse IgG (10 μg) were used. Antibody–DNA complexeswere captured by incubation with protein G-agarose, eluted, andsubjected to cross-linking reversal. The TM7SF2 sequence of interestwas amplified by PCR using AmpliTaq Gold (Applied Biosystems-Roche, Milan, Italy) and primers forward (−201/−181) 5′-TAAA-CATGTGTGGCCTCCTCC-3′, and reverse (−10/−29) 5′-AACGCAGA-CAAGGACCGCTC-3.′ PCR cycling conditions were: 5 min denaturationat 95 °C followed by 35 cycles of 30 s at 95 °C, 30 s at 62 °C, and 30 s at72 °C.

3. Results and discussion

3.1. Identification of potential SREs in the TM7SF2 gene promoter

Our previous studies on human TM7SF2 gene promoter identified aregion responsible for promoter regulation by sterols and by co-expressed SREBP-2 [4]. This region spans 200 bp upstream thetranscription start site and contains 3 GC-boxes and an invertedCCAAT-box (Fig. 1A), sequences that bind transcription factors knownto cooperate with SREBPs [10,11]. In addition, a potential SRE,characterized by low score (63.4), was predicted by TFSEARCHat −116/−106 [4] (Fig. 1A). To investigate the ability of this and/orother not yet identified SREs to bind SREBP-2 in vitro, EMSAwas performed. The probes used covered the −143/−1 promoterregion, which was activated by SREBP-2 in co-transfection experi-ments [4]. No significant shift was observed when −143/−97 and−48/−1 labelled probes were incubated with bacterial lysatescontaining SREBP-2, indicating that the previously described −116/−106 SRE is not able to bind SREBP-2 in vitro. In contrast, incubation

Fig. 3.Mutational analysis of the putative SRE motifs on TM7SF2 promoter. The putativeSRE-116 and SRE-73 present in the−200/−1 promoter region (p200) of TM7SF2 genewere subjected to mutagenesis (see Table 1). Control and mutated promoter constructsfused to luciferase gene in pGL2 vector were transfected in HepG2 and HEK293 asdescribed in Materials and methods, in the absence (open columns) or in the presenceof pCS2-SREBP2 (filled columns). Relative luciferase activities were normalized to 1 forp200 alone. Data are the mean±SD of at least three experiments performed induplicate.

590 G. Schiavoni et al. / Biochimica et Biophysica Acta 1801 (2010) 587–592

of the −96/−49 labelled probe resulted in the formation of a shiftedcomplex that was competed by molar excess of unlabelled LDLR-SRE-1 probe (Fig. 1B). The ability to form a shifted and competedcomplex was maintained also by the−87/−60 28-bp probe, locatedbetween the inverted CCAAT-box and the GC-box 3 (Fig. 1A). Toidentify a potential SRE within the 28-bp sequence, a set ofoligonucleotides bearing 4-bp successive transverse mutations wasused. Mutants 4 to 6 lost the ability to form a shifted complex (Fig. 2),indicating that the −75/−64 region is able to bind SREBP-2. Thecorresponding sequence, GCCCAGCATGGG, shares very low identitywith the SRE-1 present in the promoters of genes involved in variouslipid biosynthetic pathways, whereas bases 3–10 are 60% identical tothe inverted SRE-3 element (CTAGTGTGAG) found in the humansqualene synthase promoter [14]. On this basis, we postulate thatTM7SF2 gene promoter contains a new SRE-like sequence, locatedat −73/−64, able to bind SREBP-2.

3.2. Mutational analysis of the potential SREs within TM7SF2 promoter

To verify the ability of the SRE-like elements to participatein TM7SF2 gene transactivation by SREBP-2, promoter–luciferasereporter constructs containing the −200/−1 promoter sequence(Fig. 1A)were transfected in HepG2 and HEK293 cells. Co-transfectionof HepG2 cells with p200 and pCS2-SREBP2 resulted in about 26-foldpromoter activation (Fig. 3). When a 4-bp mutation (Table 1) wasintroduced within the −73/−64 SRE-like element, a strong reduc-tion of transactivation by SREBP-2 was observed. Contrarily, no effectwas produced by a 5-bp mutation (Table 1) introduced within the

Fig. 2. Mutational analysis of the 28-bp region (nucleotides −87/−60) of TM7SF2promoter responsible for SREBP-2 binding. Upper panel. Sequence of the wild type andmutant oligonucleotides used for electrophoretic mobility shift assays. Four mutatednucleotides in each probe are boxed. Numbers with arrows indicate the sequenceinvolved in the binding of SREBP-2. Lower panel. Electrophoretic mobility shift assayswere performed as described in the legend to Fig. 1. The SREBP-2 is indicated by thearrow. SREBP-2 dimer is indicated by the asterisk (Ref. [16]).

−116/−106 putative SRE sequence (Fig. 3). A similar behaviour,although the transactivation level was much lower (about 3-fold),was observed in transfection experiments of HEK293 cells (Fig. 3).These results confirm those obtained in EMSA experiments and de-monstrate that the ability of the new −73/−64 SRE-like element tobind SREBP-2 results in promoter transactivation, indicating a func-tional role for this element, whereas the −116/−106 putative SREcannot be considered functional.

3.3. Role of GC-boxes and invCCAAT-box in transactivation of TM7SF2promoter by SREBP-2

SREBPs alone are weak transcriptional activators and require theintervention of co-regulatory transcription factors to enhance theiractivation potency [11,16,17]. The ubiquitously expressed Sp1transcription factor, able to bind GC-box sequences, acts as SREBPco-activator in several promoters [6,10,12,18]. Transfection of HepG2cells with p200 and the Sp1 expression vector CMV-Sp1 did not resultin increased transactivation by SREBP-2 (data not shown), probablydue to endogenous availability of GC-box binding factors. To inves-tigate whether the three GC-boxes present in the p200 promoterconstruct cooperate with the −73/−64 SRE, each of them wassubjected to mutagenesis and co-transfected in HepG2 cells togetherwith pCS2-SREBP2. Mutation of GC-box 2, as well as mutation ofall the three GC-boxes, resulted in significant reduction of relativeluciferase activity in the presence of SREBP-2, compared to thecontrol. When fold activation was calculated as the ratio of relativeluciferase activities in the presence and in the absence of SREBP-2,promoter transactivation by SREBP-2 appeared reduced with anymutated construct. The lowest transactivation (about 47% of control)was obtained with pGC-box2 mut (Fig. 4). Mutation in the threeelements together did not result in additional reduction of transacti-vation, suggesting that GC-box2 is the main element involved incooperation with −73/−64 SRE. The reduction of transactivation bySREBP-2 observed with mutated GC-box1 and GC-box3 could beascribed to indirect effects on the adjacent GC-box2 and −73/−64SRE motif, respectively, although their direct role cannot be excluded.Contrarily to what observed in HepG2 cells, all the mutated GC-boxconstructs showed the same activation by SREBP-2, compared tocontrol, when transfected in HEK293 cells (data not shown). Thiscould be explained in terms of different network of transcriptionfactors in these cells, compared to HepG2.

The human TM7SF2 gene promoter contains a 5′-ATTGG-3′ se-quence (an inverted CCAAT-box) 14 bp upstream the−73/−64 SRE-

Fig. 4. Contribution of GC-boxes to TM7SF2 promoter activation by SREBP-2. The TM7SF2 promoter–luciferase chimeric construct p200 is reported with potential regulatory elementsmarked as boxes. Base substitutions were introduced in the GC-boxes (dashed) (see Table 1). The reporter constructs were transfected in HepG2 cells alone or in co-transfection withpCS2-SREBP2. Relative luciferase activities were normalized to 1 for p200 alone. For each construct, fold activation was calculated as the ratio of relative luciferase activity in thepresence (filled columns) and in the absence of SREBP-2 (open columns). Data are the mean±SD of at least three experiments performed in duplicate.

591G. Schiavoni et al. / Biochimica et Biophysica Acta 1801 (2010) 587–592

like motif (Fig. 1A). The inverted CCAAT-box was previously de-monstrated to bind the transcription factor NF-Y in both farnesyldiphosphate synthase and HMG-CoA synthase promoters, thusparticipating in sterol-mediated regulation of gene expression [11].Following these observations, an increasing number of promotersof genes involved in cholesterol metabolism that are controlledin a coordinated manner by SREBP and NF-Y has been described[14,18–20]. To study the role of NF-Y in TM7SF2 promoter activation,control and mutated p200 promoter constructs were transfected in

Fig. 5. Activation of TM7SF2 promoter by SREBP-2 requires the inverted CCAAT-box. The inmutagenesis by 3-bp substitution (pInvCCAAT mut) and by complete deletion (pΔInvCCAATtransfected with p200 wild type construct (open columns) or with pΔInvCCAAT (filled columfeeding (25-OH chol) conditions (see Materials and methods). Relative luciferase activitiesmean±SD of at least three experiments performed in duplicate. (B) HepG2 cells were transthe presence of pCS2-SREBP2 (filled columns). Relative luciferase activities were normalizthe mean±SD of at least three experiments performed in duplicate. (C) HepG2 cells were tra0–800 ng of pΔ4NFYA13m29 (NF-YA dominant negative). Data are expressed as percent actdominant. Data are the mean±SD of at least three experiments performed in duplicate.

HepG2 cells. Promoter activity increased about 4-fold when trans-fected cells were grown in medium containing delipidated serumand lovastatin (Fig. 5A). The activation by sterol starvation, mea-sured as the ratio of lovastatin to 25-hydroxycholesterol treatment,was 3.4±0.3, confirming previous results [4]. Deletion of the invertedCCAAT-box (pΔInvCCAAT) reduced the basal promoter activity toabout 25% and completely abolished promoter induction by sterolstarvation (Fig. 5A), suggesting a role of NF-Y in SREBP-2-mediatedpromoter activation. Indeed, strong reduction of promoter activation

vCCAAT-box in TM7SF2 promoter–luciferase chimeric construct p200 was subjected to), or spaced from SRE-73 by 5-bp insertion (pIns5) (see Table 1). (A) HepG2 cells werens). Transfected cells were grown in FBS, or in sterol starvation (LPDS, Lov), or in sterolwere normalized to 1 for p200 wild type construct in cells grown in FBS. Data are thefected with control or mutated reporter constructs in the absence (open columns) or ined to 1 for p200 alone. Fold activation was calculated as in legend to Fig. 4. Data arensfected with p200 wild type construct, or co-transfected with p200, pCS2-SREBP2, andivation of p200 by SREBP-2, with 100% referred to activation in the absence of negative

Fig. 6. Binding of SREBP-2, NF-Y, and Sp1 to TM7SF2 promoter in vivo. Chromatinimmunoprecipitation assays were performed on chromatin derived from HepG2 cellsgrown in the presence of 5% LPDS plus 10 μM lovastatin. Primers specific for −201/−10 sequence of TM7SF2 promoter were used to amplify DNA from complexesimmunoprecipitated with α-SREBP-2, α-NF-YB subunit, and α-Sp1 antibodies. Input,fragmented DNA before immunoprecipitation. Negative controls, mIgG for SREBP-2 andrIgG for NF-YB/Sp1.

592 G. Schiavoni et al. / Biochimica et Biophysica Acta 1801 (2010) 587–592

by co-transfected SREBP-2 occurred when InvCCAAT was deleted(pΔInvCCAAT), or a 3-bp mutation was introduced (pInvCCAATmut),or 5 bp were inserted between the −73/−64 SRE and the ATTGG(pIns5) (Fig. 5B). Similar reduction was observed when the sameconstructs were co-transfected with SREBP-2 in HEK293 cells (datanot shown). The involvement of NF-Y in promoter transactivation bySREBP-2 was further demonstrated by co-transfection of HepG-2 cellswith the p200 promoter construct and Δ4NFYA13m29 plasmid. Thelatter expresses a dominant negative form of NF-YA that rendersthe trimeric NF-Y complex unable to bind the CCAAT-box [21].This dominant negative produced a dose-dependent reduction ofpromoter transactivation by SREBP-2, reaching 63% of control(Fig. 5C). No effects were observed by co-transfecting Δ4NFYA13control plasmid (not shown). These results indicate that NF-Y andSREBP-2 cooperate in TM7SF2 promoter, contributing to sterol-dependent regulation of gene expression. In accordance with studiesin other promoters [13,17], the relative position of the NF-Y andSREBP binding elements is critical for cooperation, as demonstrated bythe decrease of transactivation by SREBP-2 with pIns5, in whichadditional 5 bp are interposed between the two elements. Indeed, thisrepresents a half helical turn variation, responsible for negative effectson transcription [22].

3.4. SREBP-2, NF-Y, and Sp1 binding to TM7SF2 promoter in vivo

To investigate whether the transcription factors transactivatingTMSF2 promoter are able to bind their consensus sequences in vivo,ChIP assays were performed on HepG2 cells grown in sterol starvationcondition. Antibodies against SREBP-2, NF-YB subunit, and Sp1 wereused. PCR amplification of the −201/−10 TMSF2 promoter region,containing the elements under investigation, was performed in theimmunoprecipitated DNA. The expected 192-bp fragment of theproximal TM7SF2 promoter was amplified when ChIP was performedwith anti-NF-YB, or anti-Sp1, or anti-SREBP-2 antibodies (Fig. 6). Noamplification product was observedwith normal rabbit IgG (rIgG) andmouse IgG (mIgG), as negative controls for NF-YB/Sp1 and SREBP-2,respectively. SON, DHFR, and LDLR were amplified as positive controlgenes for NF-YB, Sp1, and SREBP-2, respectively (data not shown).

ChIP experiments demonstrate the ability of SREBP-2, NF-Y, andSp1 to bind the proximal TM7SF2 promoter in vivo, thus strengtheningthe results of mutagenesis and co-transfection experiments. In HepG2cells grown in the presence of insulin and glucose, TM7SF2 promoterbinds SREBP-1 and NF-Y, but not Sp1 [23]. The results obtained in ourexperimental conditions could reflect differences in the responsive-ness of TM7SF2 gene to different stimuli.

In this study we identified a new SRE motif in TM7SF2 promoter,demonstrating its ability to bind SREBP-2 in vitro and in vivo and toparticipate in promoter activation. Two additional upstream motifs,the inverted CCAAT-box and GC-box2 are involved. Indeed, muta-genesis of each of them resulted in decreased promoter activity,compared to control, in co-transfection with SREBP-2, suggesting thatboth are essential to achieve maximal promoter transactivation.

Acknowledgements

The financial support of Telethon-Italy (grant no. GGP030102) andFondazione Cassa di Risparmio di Perugia (Project 2009.020.0046) isgratefully acknowledged. M. C. is supported by an AUCC fellowship.

References

[1] R. Roberti, A.M. Bennati, G. Galli, D. Caruso, B. Maras, C. Aisa, T. Beccari, M.A. DellaFazia, G. Servillo, Cloning and expression of sterol delta14-reductase from bovineliver, Eur. J. Biochem. 269 (2002) 283–290.

[2] C.A.Wassif, K.E. Brownson,A.L. Sterner,A. Forlino, P.M.Zerfas,W.K.Wilson,M.F. Starost,F.D. Porter, HEM dysplasia and ichthyosis are likely laminopathies and not due to 3β-hydroxysterol Δ14-reductase deficiency, Hum. Mol. Genet. 16 (2007) 1176–1187.

[3] A.M. Bennati, G. Schiavoni, S. Franken, D. Piobbico, M.A. Della Fazia, D. Caruso,E. De Fabiani, L. Benedetti, M.G. Cusella de Angelis, V. Gieselmann, G. Servillo,T. Beccari, R. Roberti, Disruption of the gene encoding 3β-hydroxysterol Δ14-reductase in mice does not impair cholesterol biosynthesis, FEBS J. 275 (2008)5034–5047.

[4] A.M. Bennati, M. Castelli, M.A. Della Fazia, T. Beccari, D. Caruso, G. Servillo, R. Roberti,Sterol dependent regulation of human TM7SF2 gene expression: role of the encoded3β-hydroxysterol Δ14-reductase in human cholesterol biosynthesis, Biochim.Biophys. Acta 1761 (2006) 677–685.

[5] J.D. Horton, N.A. Shah, J.A. Warrington, N.N. Anderson, S.W. Park, M.S. Brown,J.L. Goldstein, Combined analysis of oligonucleotide microarray data from trans-genic and knockout mice identifies direct SREBP target genes, PNAS 100 (2003)12027–12032.

[6] P.A. Dawson, S.L. Hofmann, D.R. van der Westhuyzen, T.C. Südhof, M.S. Brown,J.L. Goldstein, Sterol-dependent repression of low density lipoprotein receptorpromoter mediated by 16-base pair sequence adjacent to binding site fortranscription factor Spl, J. Biol. Chem. 263 (1988) 3372–3379.

[7] J.R. Smith, T.F. Osborne, M.S. Brown, J.L. Goldstein, G. Gil, Multiple sterol regu-latory elements in promoter for hamster 3-hydroxy-3-methylglutaryl-coenzymeA synthase, J. Biol. Chem. 263 (1988) 18480–18487.

[8] T.F. Osborne, G. Gil, J.L. Goldstein, M.S. Brown, Operator constitutive mutation of3-hydroxy-3-methylglutaryl coenzyme A reductase promoter abolishes proteinbinding to sterol regulatory element, J. Biol. Chem. 263 (1988) 3380–3387.

[9] J. Ericsson, S.M. Jackson, B.C. Lee, P.A. Edwards, Sterol regulatory element bindingprotein binds to a cis element in the promoter of the farnesyl diphosphatesynthase gene, Proc. Natl. Acad. Sci. U.S.A. 93 (1996) 945–950.

[10] H.B. Sanchez, L. Yieh, T.F. Osborne, Cooperation by sterol regulatory elements-binding protein and Sp1 in sterol regulation of low density lipoprotein receptorgene, J. Biol. Chem. 270 (1995) 1161–1169.

[11] S.M. Jackson, J. Ericsson, T.F. Osborne, P.A. Edwards, NF-Y has a novel role in sterol-dependent transcription of two cholesterogenic genes, J. Biol. Chem. 270 (1995)21445–21448.

[12] J.M. Lopez, M.K. Bennett, H.B. Sanchez, J.M. Rosenfeld, T.F. Osborne, Sterolregulation of acetyl coenzyme A carboxylase: a mechanism for coordinate controlof cellular lipid, Proc. Natl. Acad. Sci. U.S.A. 93 (1996) 1049–1053.

[13] J. Ericsson, S.M. Jackson, P.A. Edwards, Synergistic binding of sterol regulatoryelement-binding protein and NF-Y to the farnesyl diphosphate synthase promoteris critical for sterol-regulated expression of the gene, J. Biol. Chem. 271 (1996)24359–24364.

[14] G. Guan, P.-H. Dai, T.F. Osborne, J.B. Kim, I. Shechter, Multiple sequence elementsare involved in the transcriptional regulation of the human squalene synthasegene, J. Biol. Chem. 272 (1997) 10295–10302.

[15] M.M. Bradford, A rapid and sensitive method for the quantitation of microgramquantities of protein utilizing the principle of protein-dye binding, Anal. Biochem.72 (1976) 248–254.

[16] S.M. Jackson, J. Ericsson, R. Mantovani, P.A. Edwards, Synergistic activation oftranscription by nuclear factor Y and sterol regulatory element binding protein,J. Lipid Res. 39 (1998) 767–776.

[17] J. Inoue, R. Sato, M. Maeda, Multiple DNA elements for sterol regulatory element-binding protein and NF-Y are responsible for sterol-regulated transcription of thegenes for human 3-hydroxy-3-methylglutaryl coenzyme A synthase and squalenesynthase, J. Biochem. 123 (1998) 1191–1198.

[18] J.-H.Kim, J.N. Lee, Y.-K.Paik, Cholesterolbiosynthesis from lanosterol.A concerted rolefor Sp1 and NF-Y-binding sites for sterol-mediated regulation of rat 7-dehydrocho-lesterol reductase gene expression, J. Biol. Chem. 276 (2001) 18153–18160.

[19] M.K. Bennett, T.F. Osborne, Nutrient regulation of gene expression by the sterolregulatory element binding proteins: increased recruitment of gene-specificcoregulatory factors and selective hyperacetylation of histone H3 in vivo, Proc.Natl. Acad. Sci. U.S.A. 97 (2000) 6340–6344.

[20] M. Nagai, J. Sakakibara, Y. Nakamura, F. Gejyo, T. Ono, SREBP-2 and NF-Yare involved in the transcriptional regulation of squalene epoxidase, Biochem.Biophys. Res. Commun. 295 (2002) 74–80.

[21] R. Mantovani, X.-Y. Li, U. Pessara, R.H. van Huisjduijnen, C. Benoist, D. Mathis,Dominant negative analogs of NF-YA, J. Biol. Chem. 269 (1994) 20340–20346.

[22] R. Mantovani, The molecular biology of the CCAAT-binding factor NF-Y, Gene 239(1999) 15–27.

[23] B.D. Reed, A.E. Charos, A.M. Szekely, S.M. Weissman, M. Snyder, Genome-wideoccupancy of SREBP-1 and its partners NF-Y and Sp1 reveals novel functional rolesand combinatorial regulation of distinct classes of genes, PLoS Genetics 4 (7)(2008) e1000133.