Portal de Periódicos da CAPES

Cloning and Characterization of the Human Selenoprotein P Promoter

1997; Elsevier BV; Volume: 272; Issue: 46 Linguagem: Inglês

10.1074/jbc.272.46.29364

1083-351X

Ingeborg Dreher, Tatjana Jakobs, Josef Köhrle,

Sperm and Testicular Function

We isolated an 18-kilobase (kb) genomic selenoprotein P clone from a human placenta library and cloned, sequenced, and characterized the 5′-flanking region of the human selenoprotein P gene. Sequence analysis revealed an intron between base pairs (bp) −13 and −14 upstream of the ATG codon and another one between bp 534 and 535 of the coding region. The major transcription start site of selenoprotein P in human HepG2 hepatocarcinoma cells was mapped to bp −70 by 5′-rapid amplification of cDNA ends and by primer extension. 1.8 kb of the 5′-flanking sequence were fused to a luciferase reporter gene. They exhibited functional promoter activity in HepG2 hepatocarcinoma and Caco2 colon carcinoma cells in transient transfection experiments. Treatment of transfected HepG2 cells with the cytokines interleukin 1β, tumor necrosis factor α, and interferon γ repressed promoter activity. Nuclear extracts of interferon γ-treated cells bound to a signal transducer and activator of transcription response element of the promoter in gel retardation experiments. By transfection of promoter-deletion constructs, a TATA box and a putative SP1 site were identified to be necessary for selenoprotein P transcription. These data indicate that the human selenoprotein P gene contains a strong promoter that is cytokine responsive. Furthermore, selenoprotein P, secreted by the liver, might react as a negative acute phase protein. We isolated an 18-kilobase (kb) genomic selenoprotein P clone from a human placenta library and cloned, sequenced, and characterized the 5′-flanking region of the human selenoprotein P gene. Sequence analysis revealed an intron between base pairs (bp) −13 and −14 upstream of the ATG codon and another one between bp 534 and 535 of the coding region. The major transcription start site of selenoprotein P in human HepG2 hepatocarcinoma cells was mapped to bp −70 by 5′-rapid amplification of cDNA ends and by primer extension. 1.8 kb of the 5′-flanking sequence were fused to a luciferase reporter gene. They exhibited functional promoter activity in HepG2 hepatocarcinoma and Caco2 colon carcinoma cells in transient transfection experiments. Treatment of transfected HepG2 cells with the cytokines interleukin 1β, tumor necrosis factor α, and interferon γ repressed promoter activity. Nuclear extracts of interferon γ-treated cells bound to a signal transducer and activator of transcription response element of the promoter in gel retardation experiments. By transfection of promoter-deletion constructs, a TATA box and a putative SP1 site were identified to be necessary for selenoprotein P transcription. These data indicate that the human selenoprotein P gene contains a strong promoter that is cytokine responsive. Furthermore, selenoprotein P, secreted by the liver, might react as a negative acute phase protein. Selenoprotein P (SeP) 1The abbreviations used are: SeP, selenoprotein P; kb, kilobase pair(s); bp, base pair(s); GAS, interferon-γ activation site; IFNγ, interferon-gamma; PCR, polymerase chain reaction; RACE, rapid amplification of cDNA ends; IL, interleukin; TNFα, tumor necrosis factor α; FCS, fetal calf serum; SRE, serum response element; LTR, long terminal repeat; CAT, chloramphenicol acetyltransferase; 5′DI, type I iodothyronine 5′-deiodinase; dATPαS, adenosine 5′-O-(thiotriphosphate).is a glycosylated plasma protein whose physiological function still remains unknown. It originally was purified as a 57-kDa species showing heparin binding capacity (1Herrman J.L. Biochim. Biophys. Acta. 1977; 500: 61-70Crossref PubMed Scopus (56) Google Scholar, 2Yang J.G. Morrison-Plummer J. Burk R.F. J. Biol. Chem. 1987; 262: 13372-13375Abstract Full Text PDF PubMed Google Scholar, 3Motchnik P.A. Tappel A.L. Biochim. Biophys. Acta. 1989; 993: 27-35Crossref PubMed Scopus (21) Google Scholar, 4Eberle B. Haas H.J. J. Trace Elem. Electrolytes Health Dis. 1993; 7: 217-221PubMed Google Scholar). However, recently at least five different forms with identical amino-terminal sequences have been isolated from rat plasma based on different SDS-polyacrylamide gel electrophoresis migration and heparin Sepharose affinity (5Himeno S. Chittum H.S. Burk R.F. J. Biol. Chem. 1996; 271: 15769-15775Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar). cDNAs encoding selenoprotein P have been cloned and sequenced for man, rat, and mouse homologues. In all species, the open reading frame of the cDNA contains 10 TGA codons, each of which might encode a selenocysteine residue (6Hill K.E. Lloyd R.S. Yang J.-G. Read R. Burk R.F. J. Biol. Chem. 1991; 266: 10050-10053Abstract Full Text PDF PubMed Google Scholar, 7Hill K.E. Lloyd R.S. Burk R.F. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 537-541Crossref PubMed Scopus (161) Google Scholar, 8Steinert P. Ahrens M. Gross G. Flohé L. BioFactors. 1997; 6: 311-319Crossref PubMed Scopus (20) Google Scholar). The 3′-untranslated region of SeP mRNA contains two selenocysteine incorporation stem-loop structures (7Hill K.E. Lloyd R.S. Burk R.F. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 537-541Crossref PubMed Scopus (161) Google Scholar, 9Martin III, G.W. Harney J.W. Berry M.J. RNA. 1996; 2: 171-182Crossref PubMed Scopus (5) Google Scholar). The amino acid composition of SeP is almost identical to that suggested by the cDNA sequence, although only 7–8 of the 10 predicted selenocysteine residues were detected by amino acid analysis (10Read R. Bellew T. Yang J.-G. Hill K.E. Palmer I.S. Burk R.F. J. Biol. Chem. 1990; 265: 17899-17905Abstract Full Text PDF PubMed Google Scholar). Association of the protein with cell membranes has been described (11Wilson D.S. Tappel A.L. J. Inorg. Biochem. 1993; 51: 707-714Crossref PubMed Scopus (34) Google Scholar). In bovine brain, the expression of a selenoprotein P-like protein with 12 selenocysteine residues has been demonstrated recently at the mRNA and protein levels (12Saijoh K. Saito N. Lee M.J. Fujii M. Kobayashi T. Sumino K. Mol. Brain Res. 1995; 30: 301-311Crossref PubMed Scopus (50) Google Scholar). Although SeP transcripts are found in nearly all tissues examined (13Burk R.F. Hill K.E. J. Nutr. 1994; 124: 1891-1897Crossref PubMed Scopus (157) Google Scholar, 14Dreher I. Schmutzler C. Jakob F. Köhrle J. J. Trace Elem. Med. Biol. 1997; 11: 83-91Crossref PubMed Scopus (61) Google Scholar, 15Adams M.D. Kerlavage A.R. Fleischmann R.D. et al.Nature. 1995; 377: 3-174PubMed Google Scholar), the liver is the main source for circulating SeP in the plasma (13Burk R.F. Hill K.E. J. Nutr. 1994; 124: 1891-1897Crossref PubMed Scopus (157) Google Scholar). In the rat, about two-thirds of the plasma selenium are contained in SeP (13Burk R.F. Hill K.E. J. Nutr. 1994; 124: 1891-1897Crossref PubMed Scopus (157) Google Scholar); therefore, a role in selenium storage and distribution has been discussed (16Motsenbocker M.A. Tappel A.L. Biochim. Biophys. Acta. 1982; 709: 160-165Crossref PubMed Scopus (30) Google Scholar). However, former in vivo studies revealed a preferential synthesis of SeP compared with the cytosolic glutathione peroxidase under selenium-deficient conditions (17Hill K.E. Lyons P.R. Burk R.F. Biochem. Biophys. Res. Commun. 1992; 185: 260-263Crossref PubMed Scopus (134) Google Scholar). Some observations suggest a selenium storage function for cytosolic glutathione peroxidase rather than for SeP (17Hill K.E. Lyons P.R. Burk R.F. Biochem. Biophys. Res. Commun. 1992; 185: 260-263Crossref PubMed Scopus (134) Google Scholar, 18Gross M. Oertel M. Köhrle J. Biochem. J. 1995; 306: 851-856Crossref PubMed Scopus (79) Google Scholar, 19Sunde R.A. Burk R.F. Selenium in Biology and Human Health. Springer Verlag, New York1994: 45-47Crossref Google Scholar). Studies in the rat model revealed a protective effect of SeP against liver injury caused by diquat or paraquat. Therefore, SeP might serve as an antioxidant in the extracellular space (20Burk R.F. Lawrence R.A. Lane J.M. J. Clin. Invest. 1980; 65: 1024-1031Crossref PubMed Scopus (220) Google Scholar, 21Burk R.F. Hill K.E. Awad J.A. Morrow J.D. Kato T. Cockell K.A. Lyons P.R. Hepatology. 1995; 21: 561-569PubMed Google Scholar). To get more information about the biological function and regulation of SeP gene expression, we isolated a genomic clone from a human placenta library and cloned and sequenced a 1.8-kb fragment of the 5′-flanking region of the human SeP gene. In transient transfection experiments of cultured hepatocarcinoma and colon carcinoma cells, the activity of this proximal promoter region and several deletion constructs was analyzed using chimeric luciferase reporters. We also identified the major start site of transcription at position −70 and an interferon γ (IFNγ)-responsive element at nucleotides −742 to −732. As secretion of liver proteins is affected by acute phase reaction (22Baumann H. Gauldie J. Immunol. Today. 1994; 15: 74-88Abstract Full Text PDF PubMed Scopus (468) Google Scholar), the effect of different cytokines on SeP promoter activity was investigated; IFNγ as well as tumor necrosis factor α (TNFα) and interleukin 1β (IL1β) decreased promoter activity in transfected HepG2 cells, suggesting a repression of SeP expression during acute phase reaction. A human placenta genomic library in the Lambda FixII-vector (Stratagene) was screened with a 1200-bp cDNA probe encoding the 5′ part of the rat 16C1 SeP clone (kindly provided by Dr. K. Hill, Nashville, TN) according to a protocol provided by the manufacturer. A total of 4 × 106 plaques were screened, yielding one positive clone. Lambda Fix phage DNA was isolated and digested with EcoRI, XbaI, andPstI (Fig. 1). One 3.8-kbEcoRI, four XbaI (3, 2.1, 1.7, and 1.4 kb), and two PstI fragments (5.5 and 4.2 kb) were subcloned into the pBluescript SKII+ vector (Stratagene). Partial sequencing of theXbaI clones revealed that the 3-kb and the 2.1-kb fragments contained sequences of the 5′ part of the selenoprotein P coding region. The EcoRI and the 5.5-kb PstI clone were mapped to the 5′-untranslated region by Southern blot with oligonucleotide ID4 (5′-ACAACCACTCCAACGGGCC-3′, comprising the base pairs −33 to −14 with respect to the ATG codon). Oligonucleotide primers for DNA sequencing reactions were purchased from Pharmacia (Uppsala, Sweden). Sequencing reactions were performed with T7 DNA polymerase (Pharmacia) according to the manufacturer's instructions. The reaction products were separated on a 6% polyacrylamide gel or, for long range sequence reading, after separation on a 3.5% polyacrylamide gel directly blotted onto a nylon membrane by using the GATC 1500-System Long Run DNA Sequencer (MWG-Biotech, Munich, Germany) according to the supplier's protocol. The nucleotide sequences of both strands were determined. DNA sequence analysis was performed using DNASIS for Windows software. For sequence comparison, we used the Genetics Computer Group (GCG) sequence analysis software package. The sequence was scanned through the non-redundant data bases GenBank, EMBL, DDBJ, and PDB. The 5′-end of the SeP mRNA was cloned after rapid amplification of cDNA ends (RACE) as modified by Edwards et al. (23Edwards J.B.D.M. Delort J. Mallet J. Nucleic Acids Res. 1991; 11: 1475-1489Google Scholar) using the 5′-Amplifinder RACE kit from CLONTECH. Poly(A)+ RNA for 5′-RACE was prepared from HepG2 cells with the Dynabead mRNA direct kit (Dynal, Oslo, Norway). For cDNA preparation, 2 μg of HepG2 poly(A)+ RNA were incubated with primer ID3 (5′-CCTAGGAGCCAACTCTGAAT-3′), annealing at nucleotides 894 to 875 of the SeP open reading frame, and avian myeloblastosis virus reverse transcriptase for 30 min at 52 °C. After ligation of an "anchor" to the cDNA according to the manufacturer's protocol, PCR was carried out with the primer ID9 (5′-GGCCACATCTATCATATATGAGGA-3′), corresponding to nucleotides 423 to 399 of the open reading frame of the human SeP. Amplification products were polished with the Klenow fragment of DNA polymerase I, phosphorylated, and cloned into the SmaI restriction site of pBluescript SKII for nucleotide sequence determination. 15 ng of 32P-labeled primer ID10 (5′-ACAGGTATCAGCTGGCTTGAAGAAG-3′, corresponding to base pairs 184 to 160 of the SeP cDNA) were annealed to 2 μg of HepG2 poly(A)+ RNA for 10 min at 70 °C. The primer was extended with avian myeloblastosis virus reverse transcriptase (Life Technologies, Inc., Eggenstein, Germany) in a total volume of 20 μl containing 4 μl of 5-fold reaction buffer, 1 mmdithiothreitol, 0.5 mm each of dATP, dCTP, dGTP, and dTTP, and 0.5 units/ml RNasin (Promega) for 60 min at 42 °C. The products were subjected to ribonuclease I digestion (0.8 μg/ml, Sigma, Deisenhofen, Germany) for 20 min at 37 °C, extracted with phenol/CHCl3, ethanol precipitated, and resuspended in 4 μl of H2O. 2 μl of formamide containing stop buffer were added before separation on a 6% sequencing gel. The pGL2-basic vector, which contains a luciferase reporter gene for analysis of promoter activity of cloned fragments located immediately downstream of a polylinker, was obtained from Promega. The 1.8-kb promoter fragment of the 5.5-kb PstI subclone was amplified by PCR using the T7 promoter primer (annealing to the pBSSKII+ plasmid) and a modified ID4 primer with an additionalBglII restriction site (5′-TGCTGCAGATCTGACAACCACTCCAACGGGCC-3′). The PCR fragments were treated with the Klenow fragment of DNA polymerase I (Life Technologies, Inc.) and purified from a 1% agarose gel using the Nucleotrap kit from Macherey & Nagel (Düren, Germany). Subsequently, the fragments were cleaved with KpnI andBglII and cloned into pGL2. Identity of isolated clones were confirmed by nucleic acid sequencing. One clone (pBK15), which contained the entire 1.8-kb promoter fragment and 50 bp of the pBluescript vector, was chosen for further studies. The reporter gene plasmids pGL2control and pCATcontrol (both obtained from Promega), used as controls, contain a luciferase or chloramphenicol acetyltransferase (CAT) gene downstream of a SV40 promoter. In the plasmid pGLh5′DIpr the promoter and enhancer regions of the human type I iodothyronine 5′-deiodinase (5′DI) are fused to the luciferase reporter gene (24Jakobs T. Schmutzler C. Meissner J. Köhrle J. Eur. J. Biochem. 1997; 247: 288-297Crossref PubMed Scopus (76) Google Scholar). Based on the nucleotide sequence of clone pBK15, several restriction enzyme sites were found suitable for the construction of a set of deletion clones. Fig. 2 shows the deletion constructs. pBK15 was partially digested for 1 h with StuI, purified from a 1% agarose gel as described above, and religated, resulting in the plasmid pStu3. For the construction of ΔStuI, pStu3 was recut with StuI and withSmaI. Deletion of a 517-bp fragment of BK15 resulted in the plasmid pΔNsi, which by restriction with SmaI andNsiI was modified to ΔNsi/Sma. By restriction of BK15 withSmaI and PvuII (ΔSma/Pvu), the 50-bp pBluescript fragment and 47 bp of the very 5′-end of the promoter sequence were eliminated (not shown). For the construction of Nsi/Stu, the Nsi fragment of pBK15 was cloned into the PstI site of pBluescript SKII. A fragment containing a 123-bp Nsi/Stu SeP promoter fragment and 20 bp of pBluescript was isolated byStuI/HindIII digestion. The fragment was cloned into the SmaI and HindIII sites of pGL2. To eliminate the retroviral LTR at the 5′-end of the cloned SeP promoter, the entire insert of BK15 was isolated bySmaI/BglII digestion. The insert was recut withBstEII for 2 h at 60 °C. After purification, the insert and the BglII/SmaI cut vector were ligated. Afterward, the BstEII site of the insert was filled in by Klenow fragment treatment, and the resulting construct (ΔBstE/Sma) was ligated again. All constructs were confirmed by sequencing. HepG2 and Caco2 cells were obtained from the American Type Culture Collection. The human hepatocarcinoma line HepG2 (ATCC, HB 8065) was cultured in Dulbecco's modified Eagle's medium-F12 medium with 10% fetal calf serum and 1 mg/ml l-glutamine; the human colon carcinoma cell line Caco2 (ATCC, HTB37) was maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 1% non-essential amino acids, 0.4 mm Hepes, 1 mg/mll-glutamine. All cells were cultured in 75-mm2plastic flasks and incubated at 37 °C in a humidified atmosphere consisting of 5% CO2 and 95% air. For transfection, the cells were seeded in 6-well plates and transfected at 50–70% confluence. Plasmids for transfection were prepared and purified using a Midi prep kit purchased from Qiagen (Hilden, Germany) according to the supplier's protocol. For control of transfection efficiency, a plasmid containing the gene for bacterial β-galactosidase under the control of the constitutive SV40 promoter (pSV-β-galactosidase, Promega) was cotransfected in each experiment. Liposome-mediated transfection was performed using LipofectAMINE reagent (Life Technologies, Inc.) according to the method of Felgneret al. (25Felgner P.L. Gadek T.R. Holm M. Roman R. Chan H.W. Wenz M. Northrop J.P. Ringold G.M. Danielsen M. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 7413-7417Crossref PubMed Scopus (4388) Google Scholar). 10 μl of LipofectAMINE and 1 μg of BK15 plasmid plus 1 μg of pSV-β-galactosidase were diluted in 100 μl of serum-free medium, respectively. The solutions were combined and incubated for 30 min at ambient temperature, 800 μl of serum-free medium were added, and the solution was then carefully dripped onto the cells, which had been washed twice with serum-free medium. Caco2 cells were incubated for 6 h before the transfection solution was removed and replaced by serum-containing medium. After an incubation period of 20 h, the cells were harvested. HepG2 cells were incubated for 20 h with the transfection solution. Then the transfection solution was exchanged for serum-free medium. Cells were incubated for 20 h with this medium alone or with medium containing TNFα (50 ng/ml), IL1β (60 ng/ml), interleukin 6 (IL-6, 100 units/ml), or IFNγ (100 ng/ml) (obtained from PBH, Hannover, Germany). Cells were harvested in 180 μl of 1 × reporter lysis buffer (Promega). 20 μl of the cell lysate were used for luciferase measurement, and 50 μl (Caco2) or 100 μl (HepG2) were applied for determination of β-galactosidase activity. Background luciferase expression from the vector was determined by transfection of the empty pGL2-basic plasmid. Analysis of luciferase activity was performed using the reporter gene assay provided by Promega in a microplate luminometer (EG&G Berthold, Bad Wildbad, Germany). β-Galactosidase activity was measured in a Uvicord III photometer (Pharmacia) according to a protocol provided by Promega. For control of transfection efficiency the obtained luciferase values were divided by the β-galactosidase activities. When appropriate, data were analyzed by analysis of variance. When the main effect was significant, theU test of Mann and Whitney was applied as a post hoc test to determine individual differences between means. The interaction of HepG2 nuclear proteins with an oligonucleotide corresponding to the nucleotides −745 to −726, comprising a putative interferon-γ activation (GAS) site of the consensus sequence TTCNNNGAA (26Schindler C. Darnell J.E. Annu. Rev. Biochem. 1995; 64: 621-631Crossref PubMed Scopus (1657) Google Scholar) and flanking bases, was analyzed by electrophoretic mobility shift assays. HepG2 cells were seeded in 6-well plates in a culture medium with 10% FCS. At approximately 80% confluence, the culture medium was replaced by serum-free medium containing 1, 10, and 100 ng/ml IFNγ for 10 min, 25 ng/ml for 5 and 15 min, or 50 ng/ml for 20 to 60 min, respectively. Afterward, the medium was removed and the cells were washed twice with phosphate-buffered saline. Nuclear extracts were prepared according to a modified protocol of Grandison et al. (27Grandison L. Nolan G.P. Pfaff D.W. Mol. Cell. Endocrinol. 1994; 106: 9-15Crossref PubMed Scopus (22) Google Scholar; see also Ref. 28Schreiber E. Mattias P. Mueller M.M. Schaffner W. Nucleic Acids Res. 1989; 17: 6419Crossref PubMed Scopus (3918) Google Scholar). Nuclear cell extracts (15 μg of protein) were incubated in a reaction buffer of 10 mm Tris-HCl (pH 7.5), 100 mm NaCl, 1 mm EDTA, 1 mm dithiothreitol, 0.2% Nonidet P-40, 5% glycerol containing 1 μg of poly(dI·dC), and approximately 50 fmol of 32P-labeled double-stranded oligonucleotide (15,000–20,000 cpm) in a volume of 15 μl for 30 min at ambient temperature. The oligonucleotides GASID1 (5′-GGTCTTCCAGGAAGTACGAC-3′) and its complementary GASID2 were annealed and end labeled using T4 polynucleotide kinase (Life Technologies, Inc.) and [γ-32P]ATP (Amersham, Braunschweig, Germany). For competition experiments, a 10-, 50-, and 100-fold molar excess of unlabeled GASID1/2 was added to the reaction mixture before the addition of reaction buffer. Additional competition experiments were carried out with a related oligonucleotide from the human aromatase promoter (29Zhao Y. Nichols J.E. Bulun S.E. Mendelson C.R. Simpson E.R. J. Biol. Chem. 1995; 270: 16449-16457Abstract Full Text Full Text PDF PubMed Scopus (216) Google Scholar) (5′-GGTGTTTCCTGTGAAAGTTC-3′) as described above. Reaction mixtures were analyzed by non-denaturing polyacrylamide gel electrophoresis and autoradiography. To map the transcription initiation site of SeP, the 5′-end of the mRNA from HepG2 cells was cloned using a modified RACE protocol. Poly(A)+ mRNA was isolated from the human hepatocarcinoma cell line HepG2, which showed high SeP expression in Northern blot experiments (14Dreher I. Schmutzler C. Jakob F. Köhrle J. J. Trace Elem. Med. Biol. 1997; 11: 83-91Crossref PubMed Scopus (61) Google Scholar). Specifically primed cDNA was prepared with an oligonucleotide annealing to nucleotides 894 to 875 of the open reading frame of the SeP mRNA. PCR amplification of the 5′-mRNA end was carried out with a nested oligonucleotide corresponding to the nucleotides 423 to 399 of the coding region. A single clear band of about 500 bp in length was demonstrated by gel electrophoretic analysis as shown in Fig.3 A. The band was eluted from the gel and subcloned into pBluescript vector (Fig. 3 B), and the nucleotide sequence of three clones was determined, which was the same for all clones. 459 bp were identical to the known SeP sequence. The remaining DNA stretch started with a G corresponding to position −70 with respect to the ATG translation start codon. The position of the transcription start site was also determined by primer extension analysis using primer ID10 (184 to 160) and HepG2 mRNA. To determine whether the length of the primer extension product (Fig.3 C, lane 1) fits to the transcriptional start site mapped by RACE, a sequencing reaction with the same primer and one of the RACE clones was run in parallel (Fig. 3 C, lanes 2–5). To achieve appropriate exposure times, the sequencing reaction was carried out with [α-32P]dCTP. As sequencing with this isotope results in poorer resolution, the autoradiogram was compared with a [35S]dATPαS-sequencing reaction. Apart from a strong band supporting the finding that transcription initiation occurs at G (−70), an additional faint band was visible, which corresponds to C (−60). Screening of a human genomic library with a rat SeP clone resulted in the isolation of a single clone containing about 18 kb of inserted DNA. Partial sequencing of two XbaI subclones showed nucleotide sequence identity with the 5′ part of the published human SeP cDNA sequence and at least two exon-intron boundaries, one between bp −13 and −14 upstream of the ATG codon and the other one between bp 534 and 535 of the coding region. Sequence information about known exon-intron boundaries is given in Fig. 1 and Fig.4. The exon-intron boundaries mapped to date agree with the GT-AG rule for splice sites. Nucleotide sequence analysis of a 3.8-kb EcoRI subclone and a 5.5-kb PstI clone demonstrated that the clones contained 620 bp and 1800 bp, respectively, of the 5′-flanking region of the SeP gene. Location of the fragments with respect to the cloned 5′-end of SeP is shown in Fig. 1. Based on sequence inspection, putative binding sites for several transcription factors were found as indicated in Fig.4 A. A TATA box was present at position −101 to −95, and a CAAT box was present at position −196 to −193. GC boxes (SP1 sites) characteristic for the promoters of housekeeping genes were found between −338 and −320, and two putative AP1 binding sites were identified at −408 to −402 and at −290 to −284. At nucleotides −742 to −732, the sequence CTTCCAGGAAG was found, which corresponds to elements (GAS elements) known for DNA binding of IFNγ-responsive transcription factors of the signal transducer and activator of transcription family (25Felgner P.L. Gadek T.R. Holm M. Roman R. Chan H.W. Wenz M. Northrop J.P. Ringold G.M. Danielsen M. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 7413-7417Crossref PubMed Scopus (4388) Google Scholar). Comparison of the nucleic acid sequence with entries in non-redundant data bases revealed sequence identity in 620 bp of human SeP promoter sequence submitted by Yasui et al. (30Yasui Y. Hasada K. Yang J.G. Koiwai O. Gene (Amst.). 1996; 175: 269-270Crossref PubMed Scopus (9) Google Scholar). At bp −429 to −445, we found a stretch of 17 T nucleotides where Yasui had reported 16 Ts. A DNA stretch of the very 5′-end (bp −1808 to −1367) of the DNA fragment revealed sequence homology (84%) to an LTR of human endogenous viral DNA (31Emi M. Horii A. Tomita N. Nishide T. Ogawa M. Mori T. Matsubara K. Gene (Amst.). 1988; 62: 229-235Crossref PubMed Scopus (38) Google Scholar). To assess basal promoter activity of the 5′-flanking region of the SeP gene, the full-length fragment (BK15) was transfected into HepG2 and Caco2 cells. In both cell lines, the luciferase activity of the promoter construct was 100–1000 times higher than background activity, defined as luciferase activity of the promoterless pGL2 plasmid. To identify sequences within the 5′-flanking region of the SeP gene, important for maximal basal promoter activity, several deletion constructs (Fig. 2) were transfected into HepG2 and Caco2 cells. Luciferase activities of the deletion constructs, expressed as percent of the values of the full-length construct BK15, are shown in Fig.2 C. The Stu3 construct (Δ −1008 to −353) revealed no change in luciferase activity compared with the full-length construct in Caco2 and a slight, but not significant increase in HepG2 cells, whereas ΔStu (Δ −1800 to −353) reached only 45% in HepG2 and 84% in Caco2 cells, respectively. Elimination of the sequence between −747 and −230 (ΔNsi) resulted in luciferase activity of only 20% of BK15 in HepG2 cells and 52% in Caco2 cells, whereas removal of the complete 5′ region upstream of nucleotide −230 (ΔNsi/Sma) led to a slight but significant increase in promoter activity to 45% (HepG2,p < 0.001) and 75% (Caco2, p < 0.005) relative to control BK15 expression. By cloning a 123-bp Nsi/Stu fragment into the pGL2 plasmid (Nsi/Stu), 3% (HepG2) and 9% (Caco2) of the full-length construct's activity was achieved. However, compared with the promoterless pGL2, luciferase activity of Nsi/Stu was increased up to 40 times. In contrast, transfection of HepG2 cells with a 105-bp fragment derived from the human 5′DI promoter (24Jakobs T. Schmutzler C. Meissner J. Köhrle J. Eur. J. Biochem. 1997; 247: 288-297Crossref PubMed Scopus (76) Google Scholar), which constitutes the basal promoter of 5′DI, led only to an increase of luciferase activity by a factor of 3 to 4 (data not shown). This ensures that the Nsi/Stu region of the SeP promoter containing the SP1 and one AP1 site confers marked activation to the luciferase reporter construct. Elimination of the remaining stretch of pBluescript in BK15 by SmaI/PvuII digestion had no effect on promoter efficiency (data not shown). Removal of the retroviral LTR (ΔBstE/Sma) did not reduce SeP promoter activity. In contrast, a significant (p < 0.005) increase to 140% was obtained in HepG2 cells. HepG2 cells were treated with increasing concentrations of IFNγ before nuclear extracts were prepared and analyzed by electrophoretic mobility shift assays with an oligonucleotide comprising the GAS element described in Fig.4 A. The oligonucleotide used for the electrophoretic mobility shifts also contains the flanking bases from the SeP promoter. In the IFNγ-treated probes, a specific band appeared whose intensity correlated with increasing IFNγ concentrations (Fig.5 A, lanes 2–4). Specificity was ensured as incubation with an excess of unlabeled probe led to the dose-dependent disappearance of the slower migrating band (lanes 7–9). In contrast, a nonspecific band present in every lane was not competed by excess of unlabeled probe. Additional competition experiments with a related oligonucleotide derived from the aromatase promoter (29Zhao Y. Nichols J.E. Bulun S.E. Mendelson C.R. Simpson E.R. J. Biol. Chem. 1995; 270: 16449-16457Abstract Full Text Full Text PDF PubMed Scopus (216) Google Scholar) did not influence IFNγ-dependent protein binding to the GASID1/2 oligonucleotide (lanes 10–12). The time course of protein-DNA interaction is shown in lanes 5 and 6and Fig. 5 B. While after 5 min no retarded band appeared, it was fully present after 15 min of IFNγ treatment, persisting for at least 60 min. HepG2 cells were treated with cytokines for 20 h after transfection of the 1.8-kb SeP promoter construct, and luciferase activities were determined as shown in TableI. IL-6 had no significant effect on promoter activity, whereas IL1β, IFNγ, and TNFα revealed a significant repression of luciferase activity (46, 40, and 55% of control, respectively). Expression of β-galactosidase was also affected by cytokine treatment (Table I). Therefore, division of the luciferase activities by β-galactosidase values altered the relative levels of r