Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 131–139
Transcriptional regulation of the human type 8 17�-hydroxysteroiddehydrogenase gene by C/EBP�
Joaquın Villar a,1, Jon Celay a, Marta M. Alonso a,2, Mirja Rotinen a,Carlos de Miguel b, Marco Migliaccio a, Ignacio Encıo a,∗
a Departamento de Ciencias de la Salud, Universidad Publica de Navarra, Avda. Baranain s/n, 31008 Pamplona, Spainb Departamento de Bioquımica y Biologıa Molecular, Universidad de Navarra, Apartado Postal 177, 31080 Pamplona, Spain
Received 11 August 2006; accepted 7 December 2006
bstract
17�-Hydroxysteroid dehydrogenases (17�-HSD) regulate the intracellular concentration of active sex steroid hormones in target tissues. Toate, at least 14 different isozymes have been identified. The type 8 17�-hydroxysteroid dehydrogenase (17�-HSD8) selectively catalyzes theonversion of estradiol (E2) to estrone (E1). To map the promoter region and to investigate its regulation, we cloned and fused a 1600 bp DNAragment upstream of the 17�-HSD8 transcriptional start site to a luciferase reporter gene. After transient transfection in HepG2 cells, thisragment was shown to possess promoter activity. Deletion constructs of the 5′ flanking region of the 17�-HSD8 gene led to the identificationf the minimal promoter region within the first 75 bp upstream of the transcriptional start site. This region included two CCAAT boxesnd sequences closely resembling the consensus Sp1 and NF-�B motifs. Site directed mutagenesis revealed that the CCAAT boxes were
ssential for transcription in HepG2. EMSA, supershift and chromatin immunoprecipitation reflected that these sequences were binding sitesor C/EBP�. Furthermore, promoter activity was increased by the co-transfection of a C/EBP� expression vector, and this transactivation washrough both CCAAT boxes. Our studies indicate that C/EBP� is essential for the transcription of the 17�-HSD8 gene in the liver.
The 17�-hydroxysteroid dehydrogenases (17�-HSD) arenzymes responsible for reversible interconversion of 17-ydroxy and 17-keto steroids. Multiple different mammalian7�-HSD have been cloned and characterized to date. Human7�-HSDs differ in nucleotide cofactor and substrate speci-cities, subcellular compartmentalization and tissue-specific
xpression patterns [1–6]. One of these enzymes, the type17�-HSD (HSD17B8, also known as ke6) was primarily
haracterized as a protein whose abnormal gene regulation
1 Present address: Laboratory of Experimental Carcinogenesis, Depart-ent of Radiation Medicine, Georgetown University Medical Center,ashington, DC, USA.2 Present address: Department of Neuro-Oncology, The University ofexas MD Anderson Cancer Center, Houston, TX, USA.
s linked to the development of recessive polycystic kid-ey disease (PKD) in the mouse [7]. The mouse 17�-HSD8atalyzes the in vitro oxidation of E2, testosterone and dihy-rotestosterone and the reduction of E1 [8]. In addition, recentodelling studies predicted it to be most likely involved in
he regulation of fatty acid metabolism [9].Expression of 17�-HSD8 gene has been detected in sev-
ral somatic tissues in the mouse. 17�-HSD8 was found toe particularly abundant in the kidney and liver, and it wasetected at lower levels in ovary, testes, heart and brain [8].he 17�-HSD8 mRNA has been detected in uterus, oviduct,agina, mammary gland, prostate, pituitary gland, adrenalland, dorsal skin and lung by in situ hybridization [10].y immunofluorescence, the presence of the 17�-HSD8 pro-
ein has been detected in the cumulus cells surrounding theocyte [8], in the endothelial cells of the mouse ovary [11]nd in normal male reproductive tissues [12]. In the human,orthern blot analysis showed that the 17�-HSD8 mRNA is
redominantly expressed in the liver and pancreas andoderately present in the skeletal muscle and kidney
13].C/EBP� is a member of the family of basic region leucine
ipper (bZIP) transcription factors that also includes C/EBP�,/EBP�, C/EBP�, C/EBP� and C/EBP� [14]. C/EBP� and/EBP� are the predominant C/EBP isoforms expressed bydult hepatocytes in healthy liver [15]. In this organ, tran-cription of C/EBP� results in a single mRNA that cane translated into the full-length transcriptional activator/EBP�-LAP (35 kDa) and the amino-terminally truncated
soform C/EBP�-LIP (20 kDa), that behaves as a transcrip-ion antagonist [16]. These proteins dimerize with otherZIP family members, as well as other transcription fac-ors, and as homo- or heterodimers affect transcription ofifferent genes. Through this mechanism C/EBP� playsrole in a wide range of important cellular processes,
uch as adipocyte differentiation, carbohydrate metabolism,nflammation and cellular proliferation and survival17–19].
Here, we report the molecular characterization and pro-oter analysis of the 5′ flanking-region of the human
7�-HSD8. This gene that maps to the human leukocytentigen region at 6p21.3 and spans a 2.2-Kb stretch ofNA is composed by 8 exons and the corresponding introns
nd is physically linked to the HKE4 gene [13,20]. Theseenes are separated by only 219 bp and are transcribednto the same orientation. Therefore, the promoter-enhanceregion of the 17�-HSD8 gene should overlap with the′-untranslated region of the HKE4 gene. To determinehether this region has functional promoter activity, several7�-HSD8 5′-flanking sequence-luciferase reporter plasmidsere constructed and transiently transfected into HepG2
ells. Transfection experiments revealed that two CCAAToxes are sufficient for efficient transcription of the gene inhis cell line and this result was confirmed by site directed
utagenesis. EMSA, supershift and chromatin immunopre-ipitation studies showed specific binding of C/EBP� to theselements. Our data demonstrate that C/EBP� is involvedn the transcriptional regulation of the human 17�-HSD8ene.
. Materials and methods
.1. Reagents
17�-Estradiol was purchased from Sigma–Aldrich (St.ouis, MO). The C/EBP� expression vector (pMSV-/EBP�) [19] was a gift from Dr. Steven L. Mcknight (UTouthwestern Medical Center, Dallas, TX, USA).
Restriction enzymes were obtained from Promega (Madi-on, WI). Antibodies against C/EBP�, C/EBP� and/EBP� (C/EBP�(N-19), C/EBP�(�198) and C/EBP�(M-7), respectively) were purchased from Santa Cruziotechnology (Santa Cruz, CA).
53Cup
Molecular Biology 105 (2007) 131–139
.2. Construction of human 17β-HSD8 5′-flankingegion/luciferase genes
The 5′ untranslated region of the human 17�-SD8 gene was isolated by PCR with Pfu-polymerase
Stratagene, La Jolla, CA) using the oligonucleotides′-GATGGTCTGGCCATTGGGGCTTCCT-3′ (nucleotides323–1347 of GenBank accession number NM 006979) and′-CCAAGGCCAGTGCGGAGCGGAG-3′ (nucleotides3–52 of GenBank accession number NM 014234) andloned into pGEM-T vector (Promega) to create pAVKE2.his plasmid was digested with Bam HI and Nco I and the
esulting fragment was cloned in the Nco I and Bgl II sitesf the pGL3-Basic luciferase reporter vector (Promega) toreate pJV1670. Constructs pJV1538, pJV1185 and pJV635ere generated from pJV1670 by 5′-end deletion of its insertsing the Erase-a-Base® System (Promega) following manu-acture’s protocol. Construct pJV260 was generated via PCRith the primers 5′-GCTAGCCACTGTGTTTGAATCGA-′ and 5′-CTTTATGTTTTTGGCGTCTTCCA-3′ and theloning of the PCR product, after digestion with Nco I, inhe Nco I and Sma I sites of pGL3-Basic. Construct pJV150as generated by digestion of pJV1670 with Sma I and
ircularization of the largest fragment. Construct pJV75as generated by digestion of pJV1670 with EcoICR I andsu36 I and circularization of the largest fragment afterlling in the cut ends. The structure and identity of eachonstruct was confirmed by DNA sequencing.
The putative transcription factor binding sites on the′-flanking region of the human 17�-HSD8 gene weredentified by computer analysis using the Alibaba 2.1http://www.gene-regulation.com/pub/programs/alibaba2)nd Transcription Factor sites scan (http://www.motif.enome.jp) based on TRANSFAC databases [21].
.3. Site-directed mutagenesis
Site-directed mutagenesis was performed using theuick Change Kit (Stratagene) to alter the DNA
equence of two putative CCAAT boxes and one Sp1ite within the −260 construct. Primers used to mutatehe CCAAT box situated at −5 were: sense strand 5′-CG-CCCTTGCGCTGGGAGGAGCTTAGTGCTGGGATTC-′ and antisense strand 5′-GAATCCCAGCACTAAGC-CCTCCCAGCGCAAGGCG-3′. Primers used toutate the CCAAT box situated at −46 were: sense
trand 5′-TGGCGGGCCTTTGAGGAGCCCCGGCTT-GC-3′ and antisense strand 5′-GCAAAGCCGGG-CTCCTCAAAGGCCCGCCA-3′. Primers used to mutate
he putative Sp1 situated at −239 were: sense strand-GTGTTTGAATCGAGGGGAAAAGGTGGTAACCGG-
′ and antisense strand 5′-CCGGTTACCACCTTTTC-CCTCGATTGAAACAC-3′ (mutated sites are shown asnderlined letters). These primers were annealed to theJV260 vector. Pfu DNA polymerase was used to synthesize
CsItaC5′-TGTGGGTGGGTGGGAAGC-3′ (antisense, +8/+25).Amplification products were purified using the QIAquick
J. Villar et al. / Journal of Steroid Biochem
he mutated promoter, followed by digestion of the parentallasmid by Dpn I according to the manufacturer’s instruc-ions. The mutated plasmid was transformed into XL1-Blueompetent cells, and the resulting plasmid was isolated andequenced to confirm the mutations.
.4. Cell culture and transfection
Human HepG2 were obtained from American Type Cul-ure Collection (Manassas, VA). The cells were cultured at7 ◦C in 5% CO2 in Dulbecco’s modified Eagle’s mediumInvitrogen, Carlsbad, CA) supplemented with 10% foetalovine serum and penicillin (Invitrogen, Carlsbad, CA).edia were renewed every 2 days and cells were subcultured
t a ratio of 1:3.Cells were transfected using the calcium phosphate pre-
ipitation method [22]. Cells were seeded in 6-well plates at× 105 cells per well. After 24 h, cells were transfected withg of 17�-HSD8 promoter construction, 2 g of pUC18
carrier DNA) and 3 ng of pRL-SV40 as a control for trans-ection efficiency. After 18 h, transfected cells were washednd fresh medium was added. Cells were harvested 24 hater, and then reporter gene activity was measured in cellsxtracts.
Cells were lysed with passive lysis buffer (Promega) for5 min. Lysates were spun at 10,000 × g for 10 min to pelletellular debris. The supernatant was collected and assayedor total protein using the BioRad Protein Assay (BioRad,ercules, CA). The level of reporter gene expression fromstandardized amount of cell extract was quantified byeasuring luciferase activity using a luminometer (Bertholdechnologies, Oak Ridge, TN) and the Dual-luciferaseeporter assay system (Promega). Firefly luciferase activ-ty reflects 17�-HSD8 promoter activity. Renilla luciferasectivity was used to normalize data. pGL3 basic andGL3 control vectors (Promega) were used as transfectionontrols.
.5. Mapping of transcription start site
To identify the transcription start site, primer extensionnalysis was performed. A 17�-HSD8-specific primer (5′-AGCGGAGTCGGTTCTGG-3′) labelled with fluoresceinas hybridized to HepG2 total RNA, and extended with
he MMLV reverse transcriptase (Invitrogen) under the con-itions described by the manufacturer. The products ofxtension were separated in a 6% polyacrylamide sequenc-ng gel and sized by comparing them with the products ofsequencing reaction performed with the same primer thatas run parallelly.
.6. Nuclear extraction
Nuclear extraction was prepared by a modification of theethod of Dignam et al. [23]. All buffers contained protease
nhibitor PMSF 0.2 mM (Sigma–Aldrich). After washing
PauM
Molecular Biology 105 (2007) 131–139 133
ith PBS, cells were collected with a cell scraper, recov-red in a microtube and pelleted by centrifugation. Cellsere then suspended with five more times of their cellu-
ar volume in Buffer A [HEPES 20 Mm pH 7.9, MgCl2.5 mM, KCl 10 mM, DTT 0.5 mM (Sigma) and PMSF.2 mM], vortexed, mixed and centrifuged at 3000 rpm formin at 4 ◦C. The supernatant was removed, and the pelletas resuspended in one volume of buffer B [HEPES 20 mMH 7.9, glycerol 25%, MgCl2 1.5 mM, KCl 20 mM, EDTA.2 mM, DTT 0.5 mM and PMSF 0.2 mM], followed by incu-ation for 30 min at 4 ◦C with vortex-mixing continuously.fter centrifugation at 15,000 rpm for 15 min at 4 ◦C, the
upernatant was recovered and then used for EMSA. Pro-ein concentration was determined using the BioRad Proteinssay (BioRad).
.7. Electrophoretic mobility shift assays (EMSA)
EMSAs were carried out using LightshiftTM Chemilu-inescent EMSA kit (Pierce Biotechnology, Rockford, IL).he reactions were conducted in buffer containing 10 mMRIS pH 7.5, 50 mM KCl, 1 mM DTT and 0.1 mM EDTA.probe with the first 90 bp of the promoter region, which
as previously labelled with biotin using Biotin 3′ End DNAabelling Kit (Pierce), was used. Five micrograms of nuclearxtracts were incubated with 1 g of poly(dI-dC) and 50 fmolf probe. To determinate the specificity of the binding tohe probe, competition experiments were conducted by coin-ubation with 100-fold excesses of unlabeled competitorligonucleotides. Competitor oligonucleotides carrying con-ensus binding sites for NF-1, NF-Y, NF-�B, Sp1 or C/EBPamily members were obtained from Santa Cruz Biotechnol-gy. Reactions were incubated for 20 min at 20 ◦C beforeeing subjected to electrophoresis under native conditionshrough a 4.5% (w/v) polyacrylamide (BioRad) gel in 0.5BE buffer. Gels were transferred to a Nylon membrane
Millipore, Billerica, MA) and developed following the man-facture’s protocol.
.8. Chromatin immunoprecipitation (ChIP) assay
ChIP assay was carried out essentially using the ChIP-IThromatin immunoprecipitation Kit (Active Motif, Rixen-
art, Belgium). Anti-C/EBP� antibodies or normal mousegG (negative control) were used for immunoprecipita-ion of protein-DNA complexes and the precipitated DNAnalyzed by PCR using as primers the oligonucleotides 5′-CTAAGCAGCAGTGTCGG-3′ (sense, −142/−125) and
CR purification kit (Qiagen, Venlo, The Netherlands)ccording to the manufacturer’s instructions, and sequencedsing an ABI Prism 310 system (Perkin-Elmer, Wellesley,A).
The data was analyzed with one factor ANOVAs. Sig-ificance was set at p < 0.05. When statistically significantifferences were found data was subjected to analysis withanhane post-tests in which the value necessary for sig-ificance (p < 0.05) is lowered dividing by the numberf comparisons that are made. Analysis was done usingPSS 12.0 for Windows. The results are expressed asean ± S.E.M.
. Results
.1. The 5′ untranslated region of the 17β-HSD8 gene isfunctional promoter
To date, regions suspected to contain human17�-HSD8romoter activity have not been analyzed. Interestingly, theequence upstream of the known 17�-HSD8 cDNA is highlyC-rich, with a 70% G + C content within the first 300
ucleotides. Primer extension analysis revealed several tran-cription start points, with strong signals at −30 and −154 ofhe previously published +1 (data not shown). Furthermore,omputer analysis of about 2 kb of DNA sequence upstream
hmr1
ig. 1. Molecular analysis of the 17�-HSD8 5′-flanking region. (A) Putative transcrihe 17�-HSD8 5′-flanking region and its relative position to the first nucleotide of thranscription start sites detected by primer extension are indicated by bold faced leatches of the human sequence to regulatory elements in the TRANSFAC databas
Molecular Biology 105 (2007) 131–139
he initiation codon with the MOTIF software (Bioinformat-cs Center, Kyoto University), showed that this region lacks
consensus TATA box but contains two putative CCAAToxes, as well as a number of other putative transcriptionactor-binding sites (Fig. 1A). Among them, we could notnd an oestrogen response element (ERE) nor a typical cAMPesponse element, which are the sites most frequently asso-iated with oestrogen and luteinizing hormone signalling,espectively. However, other motifs frequently found in genesnvolved in steroidogenesis or lipid metabolism [24], as theinding sites for Sp1, NF-�B, SREBP-1, C/EBP� and YY-, were present in the sequence. An ERE half site, whichay allow for oestrogen signalling in association with other
ranscription factors such as Sp1, was also detected. Addi-ionally, comparison of these 2 kb of DNA sequence withhe 5′-flanking region of the mouse Ke6 [25] gene showed
60% of identity over a continuous fragment of 420 bp inength starting at +1 (Fig. 1B), with transcription factor coreinding motifs well conserved between both species. Sincedentity among promoter regions of homologous genes in
igher eukaryotes is unusual, the conservation of this frag-ent suggests that it is important for the transcriptional
egulation of the gene. To study such a possibility, the human7�-HSD8 5′-flanking region was cloned into pGL3 basic
ptional regulatory sequences of the 17�-HSD8 gene. The sequence found ine cDNA is shown. Putative transcription factor binding sites are underlined.tters. (B) Sequence alignment of the 17�-HSD8 human and mouse genes.
J. Villar et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 131–139 135
Fig. 2. Transcriptional activity of the 17�-HSD8 promoter-luciferase chimeras. A schematic representation of the promoter constructs used in this study iss ection ei ean ± Sp (p < 0.0p
vHsa
3a
otaimaspmpirftmsta
Ceiob
fbdtbtde1tatuH
3w
Cb1itacFHF
hown on the left. Luciferase activity was normalised to the internal transft is expressed as % activity of the largest construct (pJV1680). Values (merformed in triplicates. Significant differences to the reference constructGL3-basic was always included in each experiment as a negative control.
ector to form pJV1680 and was transiently transfected inepG2 cells. Importantly, pJV1680 elicited a robust tran-
criptional activity (Fig. 2), thus confirming the existence offunctional promoter in the 17�-HSD8 5′-flanking region.
.2. CCAAT boxes are essential for 17β-HSD8 promoterctivity
To further analyze the function of the 5′ flanking regionf the human 17�-HSD8 gene, a series of 5′ deletions con-aining sequences from −1538, −1185, −635, −260, −150nd −75 to +36 were constructed and transiently expressedn HepG2 cells. As shown in Fig. 2, shortening of the frag-
ent from −1680 to −1538, −1185 or −635 did not result inny substantial change in luciferase activity, suggesting thatequences upstream −635 most likely do not contribute toromoter activity in HepG2 cells. However, different resultsight be obtained in other cell-lines or if other stimuli are
resent. Deletion of the fragment up to −260, −150 or −75,ncreased the promoter activity by 2.5-, 2.1- and 1.7-foldespectively. These results clearly showed that the fragmentrom −75/+36 is enough to efficiently drive transcription ofhe reporter gene, and thus, can be considered as the mini-
al promoter unit. Despite this fact, results also indicate thatequences from −260 to −75 are required for full transcrip-ional activity and that the region −635/−260 might containnegative regulatory element(s).
Previous sequence analysis revealed the presence of twoCAAT boxes and several Sp1 binding sites as the most rel-
vant motifs within the 296 bp promoter fragment includedn pJV260 (Fig. 1). The role of these elements in the functionf the human 17�-HSD8 promoter was further investigatedy reporter gene analysis of intact and mutated −260/+36
wcac
fficiency control (pRL-SV40), that was cotransfected simultaneously, and.E.M.) were calculated from at least eight independent experiments, each5) are denoted by asterisks (*). Transfection of the empty parental vector
ragments. In HepG2 cells, the mutation of the CCAATox located at −5 (pMC1) and −46 (pMC2) resulted in aecrease in the luciferase activity of 80% and 99%, respec-ively (Fig. 3). This result clearly indicates that the CCAAToxes are required for promoter activity. However, muta-ion of the GC box at −239 (pMS1) only led to a modestecrease on the luciferase activity, thus suggesting that thislement has a minor role in the transcription of the human7�-HSD8 gene. Curiously, simultaneous mutation of thewo CCAAT boxes caused a 12-fold induction in promoterctivity (pMC1C2, pMC1C2S1), which suggests that simul-aneous deletion of the two CCAAT boxes allows a newnidentified transcription factor to bind and activate the 17�-SD8 promoter.
.3. Interaction of the 17β-HSD8 minimal promoterith the transcription factor C/EBPβ
There are several transcription factors able to bind toCAAT motif. These factors include the CCAAT/enhancerinding protein (C/EBP), nuclear factor Y and nuclear factorfamily factors [14,26,27]. To determine which of them is
nvolved in the regulation of the transcriptional activity ofhe human 17�-HSD8, we performed EMSA analysis using3′-end labelled double stranded DNA fragment (−75/+36)alled sJV75 and HepG2 nuclear extracts. As shown inig. 4A–C, sJV75 formed 3 DNA-protein complexes with theepG2 nuclear extract, two of them almost comigrating inig. 4A. The formation of the middle and bottom complexes
ere prevented by an unlabeled oligonucleotide carrying the
onsensus binding site for C/EBP family members (Fig. 4And C) but not by NF-1, NF-Y, NF-�B or Sp1 unlabeledonsensus oligonucleotides (Fig. 4A), thus suggesting that
136 J. Villar et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 131–139
Fig. 3. Functional analysis of transcription factor binding sites in the 17�-HSD8 proximal promoter. Luciferase activities were determined in cell extracts ofHepG2 cells transfected with derivatives of pJV260 in which the CCAAT boxes located at −5 and −46 and a putative Sp1 binding site located at −239 werem g sites a− t are indV experimc
acCdtpiniCataD1t1
omwcfCNpti
3
t
tscptHopsairvs
4
cocpS
fbsa
utated, either alone or in combination. Putative transcription factor bindin46) and 3 (Sp1 binding site at −239); mutations within the binding elemenalues (mean ± S.E.M.) were calculated from a total of eight independentonstruct are denoted by asterisks (*p < 0.05; **p < 0.01).
C/EBP related protein binds to sJV75. However, whenompetitor oligonucleotides HSD-C1 (−29/+7) and HSD-2 (−67/−38) were used (Fig. 4B) only the middle band wasisplaced indicating that this is the specific complex and alsohat the two CCAAT boxes are able to bind the C/EBP relatedrotein. DNA-protein interactions were further investigatedn supershift experiments performed with sJV75, HepG2uclear extracts and C/EBP�, C/EBP� and C/EBP� antibod-es. As shown in Fig. 4B, no supershift was detected with the/EBP� or C/EBP� antibodies. However, when the C/EBP�ntibody was included in the binding reaction (Fig. 4C),he middle (specific) complex vanished as expected for anntibody whose epitope (amino acids 199–345) maps to theNA binding domain of the transcription factor (amino acids93–225). Taken together, these results clearly demonstratehat the CCAAT boxes within the proximal promoter of the7�-HSD8 gene bind to C/EBP� in a specific manner.
These results prompted us to verify the in vivo occupancyf the C/EBP� to the CCAAT sites. With this purpose, chro-atin immunoprecipitation (ChIP) assays were performedith HepG2 cells after formaldehyde crosslinking. The pre-
ipitated DNA was subjected to PCR with specific primersor the 17�-HSD8 promoter region. Analysis revealed that/EBP� binds strongly to the 17�-HSD8 promoter (Fig. 4D).o amplification product was detected with DNA immuno-recipitated with IgG. This result provided the direct evidencehat C/EBP� forms the complex with 17�-HSD8 promotern vivo.
.4. C/EBPβ transactivates the 17-βHSD8 promoter
The above results indicate that the binding of C/EBP� tohe CCAAT boxes is required for the minimal transcription of
dr
f
re shown as ovals and referred as 1 (CCAAT box at −5), 2 (CCAAT box aticated by a cross. Luciferase activity is referred to as the wild type pJV260.ents, each performed in triplicates. Significant difference to the reference
he human17�-HSD8 gene. To examine the effects of expres-ion of C/EBP� on 17�-HSD8 promoter activity, HepG2ells were cotransfected with the reporter plasmids pJV260,MC1, pMC2 or pMC1C2, and a C/EBP� expression vec-or. As shown in Fig. 5, the overexpression of C/EBP� inepG2 resulted in an increase in the transcriptional activityf the wild type pJV260 and mutated constructs pMC1 andMC2 cells by 3.5-, 4.5- and 7-fold, respectively. However,imultaneous mutation of the two CCAAT boxes in pMC1C2bolished the enhancement of the 17�-HSD8 promoter activ-ty induced by C/EBP� overexpression. Taken together, theseesults confirm the involvement of C/EBP� in the transacti-ation of 17�-HSD8 gene via binding to the CCAAT boxequences.
. Discussion
To define the promoter of the human 17�-HSD8 gene, weharacterized a fragment (1.6 kb) of the 5′-upstream sequencef the gene. This region lacks a consensus TATA box butontains two putative CCAAT boxes. It also includes severalutative binding sites for different transcription factors likep-1 and NF-�B among others (Fig. 1).
In this study, we demonstrate that the −96/+8 promoterragment contains two CCAAT boxes that are essential forasal transcription of the gene (Fig. 3). There are several tran-cription factors able to bind to CCAAT motifs; among themre NF-Y, NF-1 and C/EBP family [14,26,27]. Our results
emonstrate that transcription factor C/EBP� binds to thisegion of the human 17�-HSD8 gene.
C/EBP� is a member of the C/EBP family of transcriptionactors (C/EBP�, C/EBP�, C/EBP�, C/EBP�, C/EBP� and
J. Villar et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 131–139 137
Fig. 4. Interaction of C/EBP� with the 17�-HSD8 proximal promoter. (A) EMSA showing specific binding of HepG2 nuclear extracts to sJV75. When indicated,a 100-fold molar excess of unlabelled oligonucleotide competitors carrying the consensus binding site for NF-1, NF-Y, C/EBP family members, NF-�B or Sp1were included in the binding reactions. Competed DNA-nuclear protein complexes are identified by arrows. (B) EMSA and supershift analysis of sJV75 usingHepG2 nuclear extracts and a 100-fold molar excess of unlabelled competitor oligonucleotides HSD-C1 and HSD-C2 (−29/+7 and −67/−38 of the human17�-HSD8 gene, respectively) or 2 g of the C/EBP�(N-19) or C/EBP�(M-17) antibody as indicated. All EMSA were performed at least three times. Arrowidentifies the specific DNA-nuclear protein complex. (C) EMSA and supershift analysis of sJV75 using HepG2 nuclear extracts and a 100-fold molar excess ofu f the Cc d IgG anC hat ampa
Cailalt[scp
trctair
t
ttbCriEtpdhElmend
nlabelled competitor oligonucleotide for C/EBP family members or 2 g oomplex. (D) ChIP assay performed with HepG2 cells using the C/EBP� anCTAAGCAGCAGTGTCGG-3′ and 5′-TGTGGGTGGGTGGGAAGC-3′ tminimum of two times.
/EBP�) that bind to consensus DNA sequences as homo-nd heterodimers and affect the transcription of various genesnvolved in proliferation and differentiation, especially in theiver and the immune system [15]. Mice deficient for C/EBP�re viable but display immune defects including lymphopro-iferative disorders, imbalanced T-helper responses, impairedumor cytotoxicity and increased susceptibility to infections28]. Interestingly, C/EBP�-null mice are sterile [29], andhow high levels of steroid and prolactin receptor expressiononcomitant with a decrease in proliferation in response toregnancy hormones [30].
The mouse 17�-HSD8 protein catalyzes in vitro oxida-ion of estradiol, testosterone and dihydrotestosterone andeduction of estrone [8]. Thus, this protein participates in theontrol of the intracellular levels of active steroids. The pro-ein expressed by the human 17�-HSD8 has not been isolatednd its activity remains in question. However, recent model-
ng studies have predicted 17�-HSD8 to be involved in theegulation of fatty acid metabolism [9].
Based on the activity of the mouse 17�-HSD8 protein,he involvement of C/EBP� in the hepatic regulation of
nda
/EBP�(�198) antibody. Arrow identifies the specific DNA-nuclear proteintibodies. The results were evaluated by PCR using the oligonucleotides 5′-lify the region −142/+25 of 17�-HSD8 gene. ChIP assays were performed
he 17�-HSD8 gene suggests that C/EBP� indirectly con-rols the levels of active intracellular estrogens in the livery governing the transcription of hepatic 17�-HSD8 gene./EBP� is known to be a critical mediator of steroid hormone
esponsiveness in the uterus. C/EBP� expression is rapidlynduced in response to E2. This induction is mediated byR, because treatment with ICI 182,780, an ER antagonist
hat blocks the transcriptional activity of the receptor, sup-ressed the E2-induced C/EBP� expression [31]. Recently, airect interaction between ER� and C/EBP� bound to DNAas also been described in breast cancer cells activated with2 [32]. Since 17�-HSD8 protein catalyze the reactions that
ead to the inactivation of androgen and estrogens, C/EBP�-ediated regulation of 17�-HSD8 should lead to less active
strogens and androgens available and may help to maintainormal levels of these hormones in human tissues, especiallyuring foetal development.
The 17�-HSD8 gene was first discovered as Ke6 in con-ection with PKD in the mouse where it was shown to beown-regulated in different models of PKD mice (cpk, jcknd pcy) and in the liver of the congenital model of the
138 J. Villar et al. / Journal of Steroid Biochemistry &
Fig. 5. 17�-HSD8 promoter transactivation by C/EBP�. HepG2 cells weretransiently transfected with 2 g of the C/EBP� expression vector and 2 gof pJV260 or one of the three mutated constructs as indicated. Luciferaseactivity was normalised to the internal transfection efficiency control (pRL-SV40), that was cotransfected simultaneously and it is referred to the activityof the constructs in the absence of the pC/EBP� expression vector. Values(p(
mtooFweaaasmtiHscC[f
ziiiii
A
o
wptS
R
[
[
[
[
[
[
[
[17] C.F. Calkhoven, C. Muller, C.F. LeutzCalkhoven, C. Muller, A. Leutz,
mean ± S.E.M.) were calculated from four independent experiments, eacherformed in triplicates. Significant differences to the reference constructp < 0.05) are denoted by asterisks (*).
urine PKD (cpk homozygote) [7,33]. It has also been shownhat inhibition of the Ke6 gene expression in organ culturesf embryonic kidneys leads to the development of numer-us cysts against a background of normal nephrogenesis [8].or these reasons, the abnormal regulation of the Ke6 geneas linked to the development of recessive renal cystic dis-
ase, a developmental disorder of the kidney. Moreover, ovarynd testes of cpk/cpk mice are underdeveloped and arrestedt an early stage and direct measurements of 17�-HSD8ctivity in these organs showed a marked reduction in sexteroid metabolism, thus suggesting that estrogen/androgenetabolism play an important role in the development of
he urogenital system [11,12]. Another gene down-regulatedn PKD is 11�-Hydroxysteroid dehydrogenase type 1 (11�-SD1) [34]. This enzyme plays a very important role in
teroid metabolism, since it inactivates glucocorticoids byonverting them to their 11-keto derivatives. Interestingly,/EBP regulates hepatic transcription of 11�-HSD1 gene
35]. Our results suggest that the C/EBP transcription factoramily could be involved in the PKD.
In conclusion, this study contributes to the characteri-ation of the transcriptional regulation of 17�-HSD8 genen HepG2 cells. Understanding the regulatory mechanismnvolved in the expression of the human 17�-HSD8 gene ismportant, because this gene participates in the control of thentracellular levels of active steroids and could be involvedn the regulation of fatty acid metabolism.
cknowledgments
This work was supported by a grant from the Departmentf Health of the Government of Navarra, Spain. J. Villar
[
Molecular Biology 105 (2007) 131–139
as the recipient of a “Formacion y perfeccionamiento deersonal investigador navarro” fellowship from the Depar-amento de Educacion y Cultura del Gobierno de Navarra,pain.
eferences
[1] S. Ishikura, K. Matsumoto, M. Sanai, K. Horie, T. Matsunaga, K.Tajima, O. El-Kabbani, A. Hara, Molecular cloning of a novel type ofrat cytoplasmic 17{beta}-hydroxysteroid dehydrogenase distinct fromthe type 5 isozyme, J. Biochem. 139 (2006) 1053–1063.
[2] L.Y. Zhou, D.S. Wang, B. Senthilkumaran, M. Yoshikuni, Y. Shibata, T.Kobayashi, C.C. Sudhakumari, Y. Nagahama, Cloning, expression andcharacterization of three types of 17beta-hydroxysteroid dehydroge-nases from the Nile tilapia, Oreochromis niloticus, J. Mol. Endocrinol.35 (2005) 103–116.
[3] R. Mindnich, G. Moller, J. Adamski, The role of 17 beta-hydroxysteroiddehydrogenases, Mol. Cell. Endocrinol. 218 (2004) 7–20.
[4] V. Luu-The, Analysis and characteristics of multiple types of human17beta-hydroxysteroid dehydrogenase, J. Steroid Biochem. Mol. Biol.76 (2001) 143–151.
[5] F. Labrie, V. Luu-The, S.X. Lin, J. Simard, C. Labrie, Role of 17 beta-hydroxysteroid dehydrogenases in sex steroid formation in peripheralintracrine tissues, Trends Endocrinol. Metab. 11 (2000) 421–427.
[6] H. Peltoketo, V. Luu-The, J. Simard, J. Adamski, 17beta-hydroxysteroid dehydrogenase/17-ketosteroid reductase family;nomenclature and main characteristics of the 17HSD/KSR enzymes,J. Mol. Endocrinol. 23 (1999) 1–11.
[7] N. Aziz, M.M. Maxwel, B. St-Jacques, B.M. Brenner, Down-regulationof Ke 6, a novel gene encoded within the major histocompatibility com-plex, in murine polycystic kidney disease, Mol. Cell. Biol. 13 (1993)1847–1853.
[8] J. Fomitcheva, M.E. Baker, E. Anderson, G.Y. Lee, N. Aziz, Charac-terization of Ke 6, a new 17betahydroxysteroid dehydrogenase, and itsexpression in gonadal tissues, J. Biol. Chem. 273 (1998) 22664–22671.
[9] V.Z. Pletnev, W.L. Duax, Rational proteomics IV: modeling the primaryfunction of the mammalian 17beta-hydroxysteroid dehydrogenase type8, J. Steroid Biochem. Mol. Biol. 94 (2005) 327–335.
10] G. Pelletier, V. Luu-The, S. Li, F. Labrie, Localization of type 8 17beta-hydroxysteroid dehydrogenase mRNA in mouse tissues as studied byin situ hybridization, J. Histochem. Cytochem. 53 (2005) 1257–1271.
11] D. Woo, G.Y. Lee, E. Anderson, N. Aziz, Immature ovaries and poly-cystic kidneys in the congenital polycystic kidney mouse may be dueto abnormal sex steroid metabolism, Mol. Cell. Endocrinol. 176 (2001)155–162.
12] N. Aziz, E. Anderson, G.Y. Lee, D.D. Woo, Arrested testis developmentin the cpk mouse may be the result of abnormal steroid metabolism,Mol. Cell. Endocrinol. 171 (2001) 83–88.
13] A. Ando, Y.Y. Kikuti, A. Shigenari, H. Kawata, N. Okamoto, T. Shiina,L. Chen, T. Ikemura, K. Abe, M. Kimura, H. Inoko, cDNA cloning of thehuman homologues of the mouse Ke4 and Ke6 genes at the centromericend of the human MHC region, Genomics 35 (1996) 600–602.
14] D.P. Ramji, P. Foka, CCAAT/enhancer-binding proteins: structure,function and regulation, Biochem. J. 365 (2002) 561–575.
15] A.M. Diehl, Roles of CCAAT/enhancer-binding proteins in regulationof liver regenerative growth, J. Biol. Chem. 273 (1998) 30843–30846.
16] P. Descombes, U. Schibler, A liver-enriched transcriptional activatorprotein, LAP, and a transcriptional inhibitory protein, LIP, are translatedfrom the same mRNA, Cell 67 (1991) 569–579.
Translational control of C/EBP� and C/EBP� isoform expression,Genes Dev. 14 (2000) 1920–1932.
18] M. Buck, M. Chojkier, Signal transduction in the liver: C/EBP� mod-ulates cell proliferation and survival, Hepatology 37 (2003) 731–738.
19] Z. Cao, R.M. Umek, S.L. McKnight, Regulated expression of threeC/EBP isoforms during adipose conversion of 3T3-L1 cells, GenesDev. 5 (1991) 1538–1552.
20] Y.Y. Kikuti, G. Tamiya, A. Ando, L. Chen, M. Kimura, E. Ferreira, K.Tsuji, J. Trowsdale, H. Inoko, Physical mapping 220 kb centromericof the human MHC and DNA sequence analysis of the 43-kb segmentincluding the RING1, HKE6, and HKE4 genes, Genomics 42 (1997)422–435.
21] E. Wingender, X. Chen, E. Fricke, R. Geffers, R. Hehl, I. Liebich, M.Krull, V. Matys, H. Michael, R. Ohnhauser, M. Pruss, F. Schacherer,S. Thiele, S. Urbach, The TRANSFAC system on gene expressionregulation, Nucl. Acids Res. 9 (2001) 281–283.
22] F.L. Graham, A.J. van der Eb, Transformation of rat cells by DNA ofhuman adenovirus 5, Virology 54 (1973) 536–539.
23] J.D. Dignam, R.M. Lebovitz, R.G. Roeder, Accurate transcription initi-ation by RNA polymerase II a soluble extract from isolated mammaliannuclei, Nucl. Acids Res. 11 (1983) 1475–1489.
24] N.A. Kolchanov, E.V. Ignatieva, E.A. Ananko, O.A. Podkolodnaya, I.L.Stepanenko, T.I. Merkulova, M.A. Pozdnyakov, N.L. Podkolodny, A.N.Naumochkin, A.G. Romashchenko, Transcription regulatory regionsdatabase (TRRD): its status in 2002, Nucl. Acids Res. 30 (2002)312–317.
25] S. Ramirez, I. Fomitcheva, N. Aziz, Abnormal regulation of the Ke 6gene, a new 17beta-hydroxysteroid dehydrogenase in the cpk mousekidney, Mol. Cell. Endocrinol. 143 (1998) 9–22.
26] C. Romier, F. Cocchiarella, R. Mantovani, D. Moras, The NF-
YB/NF-YC structure gives insight into DNA binding and transcriptionregulation by CCAAT factor NF-Y, J. Biol. Chem. 278 (2003)1336–1345.
27] R.M. Gronostajski, Roles of the NFI/CTF gene family in transcriptionand development, Gene 249 (2000) 31–45.
[
Molecular Biology 105 (2007) 131–139 139
28] I. Screpanti, L. Romani, P. Musiani, A. Modesti, E. Fattori, D. Laz-zaro, C. Sellitto, S. Scarpa, D. Bellavia, G. Lattanzio, F. Bistoni, L.Frati, R. Cortese, A. Gulino, G. Ciliberto, F. Costantini, V. Poli, Lym-phoproliferative disorder and imbalanced T-helper response in C/EBPbeta-deficient mice, EMBO J. 14 (1995) 1932–1941.
29] E. Sterneck, L. Tessarollo, P.F. Johnson, An essential role forC/EBPbeta in female reproduction, Genes Dev. 11 (1997) 2153–2162.
30] S.L. Grimm, A. Contreras, M.H. Barcellos-Hoff, J.M. Rosen, Cell cycledefects contribute to a block in hormone-induced mammary gland pro-liferation in CCAAT/enhancer-binding protein (C/EBPbeta)-null mice,J. Biol. Chem. 280 (2005) 36301–36309.
31] S.R. Mantena, A. Kannan, Y.P. Cheon, Q. Li, P.F. Johnson, I.C. Bagchi,M.K. Bagchi, C/EBPbeta is a critical mediator of steroid hormone-regulated cell proliferation and differentiation in the uterine epitheliumand stroma, Proc. Natl. Acad. Sci. USA 103 (2006) 1870–1875.
32] J. Dong, C.H. Tsai-Morris, M.L. Dufau, A novel estradiol/estrogenreceptor alpha-dependent transcriptional mechanism controls expres-sion of the human prolactin receptor, J. Biol. Chem. 281 (2006)18825–18836.
33] M. Maxwell, J. Nearing, N. Aziz, Ke 6 gene. Sequence and organizationand aberrant regulation in murine polycystic kidney disease, J. Biol.Chem. 270 (1995) 25213–25219.
34] N. Aziz, D. Brown, W. Lee, A. Naray-Fejes-Toth, Aberrant 11b-hydroxysteroid dehydrogenase-1 activity in cpk mouse: implicationsfor regulation by the Ke 6 gene, Endocrinology 137 (1996) 5581–5588.
35] L.J. Williams, V. Lyons, I. MacLeod, V. Rajan, G.J. Darlington, V. Poli,J.R. Seckl, K.E. Chapman, C/EBP regulates hepatic transcription of11beta-hydroxysteroid dehydrogenase type 1, A novel mechanism forcross-talk between the C/EBP and glucocorticoid signalling pathways,J. Biol. Chem. 275 (2000) 30232–30239.
