Characterization of the human gene encoding the scavenger receptor expressed by endothelial cell and its regulation by a novel transcription factor, endothelial zinc finger protein-2.

The scavenger receptor expressed by endothelial cell (SREC), mediates the selective uptake of modified low density lipoprotein (LDL), such as acetylated LDL and oxidized LDL, into endothelial cells. The SREC gene spans 12 kilobase pairs and contains 11 exons. Analysis of full-length cDNA clones of SREC from a peripheral blood leukocyte cDNA library revealed that at least five alternatively spliced cDNAs were present, and two of them encoded soluble forms of SREC. The transcription start site of the SREC gene was mapped, and DNA sequence analysis revealed an Sp1 binding site in its proximal region. Deletion analysis of the 5'-flanking sequence revealed that sequence between base pairs -108 and -98 was critical for the promoter activity. This region contained half of an inverted repeat (IR) sequence with a triple nucleotide spacer (IR-3). A protected sequence between base pairs -268 and +17 was defined by in vitro DNase I footprinting analysis using human umbilical vein endothelial cell (HUVEC) nuclear extract. A novel transcription factor, endothelial zinc finger protein-2 (EZF-2), that binds to the 5'-flanking critical region of the SREC promoter activity was cloned from a HUVEC cDNA library employing a one-hybrid system. Whereas purified recombinant Sp1 alone produced similar protection in in vitro DNase I footprinting analysis, EZF-2 also bound to the 5'-flanking region SREC promoter. Co-transfection of SREC promoter and Sp1 or EZF-2 expression plasmids in HUVEC revealed that EZF-2 but not Sp1 increased SREC promoter activity. On the other hand, the mutation of either the Sp1 motif or IR-3 motif resulted in a decrease in the promoter activity. These results suggest that whereas Sp1 is the major nuclear protein bound to the regulatory region of the promoter, both EZF-2 and Sp1 are responsible for its regulation.

Scavenger receptors mediate the endocytosis of chemically modified lipoproteins, such as acetylated low density lipoprotein (LDL) 1 and oxidized LDL, and have been implicated in the pathogenesis of atherosclerosis. The scavenger receptor gene family is composed of a series of unlinked genes that encode membrane proteins that bind multiple ligands (1)(2)(3)(4)(5), and we have cloned a subgroup of this family, the class F receptor, scavenger receptor expressed by endothelial cell (SREC), from a human umbilical vein endothelial cell (HUVEC) cDNA library (6,7). SREC mediates the binding and degradation of acetylated LDL and oxidized LDL in endothelial cells. The primary structure of the molecule has no significant homology to other types of scavenger receptors. SREC has several characteristic domain structures, including an N-terminal extracellular domain with seven epidermal growth factor (EGF)-like cysteine pattern signatures and an unusually long C-terminal cytoplasmic domain (391 amino acids) composed of a Ser/Prorich region followed by a Gly-rich region. SREC is expressed at high levels in endothelial cells but not in smooth muscle cells.
Regulation of gene expression at the transcription level is mediated by the interaction of trans-acting factors with cisacting DNA sequences in genes (8). However, the molecular mechanism underlying the transcriptional regulation of the SREC gene has not been clarified. Therefore, identification of basal and regulatory DNA elements in the 5Ј-flanking region of the human SREC gene will provide important insight into the molecular mechanisms underlying the regulation of expression of this gene. In this paper, we report the isolation and functional analysis of the promoter of the human SREC gene. To examine the regulation of the human SREC gene promoter, chimeric constructs containing serial deletions of the 5Ј-flanking region of the human SREC gene ligated to the luciferase reporter gene were prepared and transfected into HUVEC. We have identified the region that plays an important role in determining the basal promoter activity of the gene. This region contained an inverted repeat (IR) sequence with a triple nucleotide spacer (IR-3) and Sp1 motif. Employing a one-hybrid system, we cloned a novel transcription factor, endothelial zinc finger protein-2 (EZF-2), that bound to the 5Ј-flanking region, IR-3 motif of the SREC gene from a HUVEC cDNA library. Co-transfection of the SREC promoter and Sp1 or EZF-2 expression plasmids in HUVEC revealed that EZF-2 but not Sp1 increased SREC promoter activity. On the other hand, the mutation of either the Sp1 motif or IR-3 motif resulted in a decrease in promoter activity. It was also found that inflammatory cytokines such as IL-1␣, IL-1␤, and TNF-␣ inhibited SREC promoter activity. Our results suggest that whereas Sp1 is the major nuclear protein bound to the regulatory region of the SREC promoter and can regulate the gene expression, EZF-2, a novel transcription factor, is also a major regulator of the promoter activity.

EXPERIMENTAL PROCEDURES
Isolation of Human SREC Gene-A human genomic DNA library (CLONTECH) was screened by a combination of PCR and hybridization (9) using two oppositely oriented oligonucleotides from the 5Ј region of the SREC cDNA (5ЈCCCGCTGCTGCTGCTCTG-3Ј and 5Ј-TGGCTG-GCTCGCACTG-3Ј) that corresponded to nucleotides 47-64 and 354 -370 of SREC cDNA, respectively (7), causing the amplification of 324 bp. Once a human genomic DNA library was plated at a density of 5000 plaque-forming units/plate on LB agar plates using Escherichia coli XL1-Blue MRA as a host. The crude phage solution of each plate was recovered with SM buffer. Crude phage solutions were screened by PCR. The positive phage solution was then plated on an LB agar plate. Hybridization was conducted using a fluorescein-12-dUTP-labeled PCR product as a probe. Signals were detected using anti-fluorescein horseradish peroxidase conjugate (PerkinElmer Life Sciences) and ECL plus (Amersham Biosciences). The clone obtained, SREC-1, had an insert of 15 kb and contained sequences at the 5Ј-end of the SREC cDNA. For cloning of the 3Ј-end of the SREC gene, the same human library was screened by PCR using two oppositely oriented oligonucleotides from the 3Ј region of the SREC cDNA (7) (5Ј-GCCAGCGAGAGATGGAGC-3Ј and 5Ј-TCTTGGGGTGGCACACA-3Ј) that corresponded to nucleotides 1391-1408 and 1641-1657 of the SREC cDNA, respectively, causing the amplification of 267 bp. Two clones were identified. One of them, SREC-2, had an insert of 16 kb and contained the 3Ј-end of the SREC cDNA.
Restriction Map and Exon/Intron Junctions of SREC-The DNAs from clones were digested with restriction enzyme XbaI or XhoI, and the restriction fragments were subcloned into pUC19 vector. The exon/ intron junctions of exons 1-11 were determined by sequencing of 15 kb of overlapping DNA fragments from SREC-1 and SREC-2. Sequencing was performed using an automated sequencer (Applied Biosystems model 377-18 DNA Sequencer; PerkinElmer Life Sciences).
Cloning of Full-length SREC cDNAs from Human Peripheral Blood Leukocyte cDNA Library-A cDNA library was prepared employing poly(A) ϩ RNA human peripheral blood leukocyte as described (7) and then screened by PCR essentially as described above.
RNA Ligase-mediated Rapid Amplification of cDNA Ends (RLM-RACE) to Define the Transcription Start Sites of SREC Gene-RLM-RACE (10) was performed using the First Choice TM RLM-RACE Kit (Ambion, Austin, TX) according to the manufacturer's instructions. In brief, 0.5 g of poly(A) ϩ RNA from HUVEC was treated with calf intestine phosphatase to remove 5Ј-phosphate from degraded mRNA, rRNA, and DNA and then treated with tobacco acid pyrophosphatase (TAP) to remove the cap from full-length, intact mRNA. RNA adaptors were ligated to decapped mRNA with T4 RNA ligase and then reverse transcribed with Moloney murine leukemia virus reverse transcriptase and random decamers. Nested PCR was then conducted to amplify the 5Ј-end of a specific transcript using two nested primers.
Samples were amplified for 25 cycles under the following conditions: denaturation for 1 min at 94°C, annealing for 1 min at 55°C, and extension for 1 min at 72°C employing AmpliTaq Gold (Applied Biosystems, Inc.). A PCR product was then digested with restriction enzymes KpnI and XhoI and then cloned into pGL3-Basic promoterless plasmid containing the firefly luciferase gene and sequenced using a DNA sequencer (Applied Biosystems, Inc.).
Cell Culture and Transfection-HUVEC and CASM were purchased from Clonetics and maintained in EGM-1 or SmGM medium according to the manufacturer's instructions. For transfection experiments, the cells were grown on a 24-well plate at a density of 3.76 ϫ 10 4 cells/well. After 16 h of plating, cells were washed once with serum-free OPTI-MEM. Either 1.6 g of pGL3-Promoter plasmid or an equivalent molar amount of test plasmid was co-transfected into HUVEC along with 0.032 g of pRL-CMV plasmid using the synthetic cationic lipid component, Lipofectin reagent, according to the manufacturer's instructions (Invitrogen). The pRL-CMV vector containing the Renilla luciferase gene under control of the herpes simplex virus thymidine kinase promoter (Promega) was used as an internal control of differences in transfection efficiency and cell number. For functional analysis of the basal promoter region of the SREC gene, the transfected cells were maintained for 48 h in serum-supplemented medium before harvesting. For analysis of regulation of the SREC promoter by various stimulants, the transfected cells were maintained for 24 h in serum-supplemented medium and incubated for 24 h in serum-supplemented medium supplemented with or without stimulants before harvesting. At the end of the culture period, the transfectants were lysed, and the luciferase activity in the cell lysates was measured by a dual luciferase reporter assay system (Promega).
Preparation of Nuclear Extracts-All steps were carried out at 4°C or on ice. The cells were grown in 10-cm dishes to ϳ80% confluence, washed twice with ice-cold phosphate-buffered saline, and harvested by scraping into 1 ml of ice-cold phosphate-buffered saline containing 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 2 g/ml leupeptin, and 2 g/ml aprotinin. The cells were pelleted by centrifugation at 500 ϫ g for 5 min, gently resuspended in 1 ml of hypotonic buffer (10 mM HEPES-NaOH, pH 7.9, 10 mM KCl, 1.5 mM MgCl 2, 0.1 mM EGTA, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 2 g/ml leupeptin, and 2 g/ml aprotinin), and incubated on ice for 10 min. The cells were then lysed by homogenization, and the nuclei were collected by centrifugation for 5 min at 1000 ϫ g. Nuclear extracts were prepared by a modification of the method of Dignam et al. (11). The nuclei were resuspended in 150 ml of ice-cold extraction buffer (10 mM HEPES-NaOH, pH 7.9, 420 mM NaCl, 1.5 mM MgCl 2, 0.1 mM EGTA, 0.5 mM dithiothreitol, 5% glycerol, and 1 mM phenylmethylsulfonyl fluoride). After slowly mixing for 30 min at 4°C, the suspensions were centrifuged at 13,000 ϫ g for 15 min. The supernatant was dialyzed against dialysis buffer (20 mM HEPES-NaOH, pH 7.9, 75 mM NaCl, 0.1 mM EDTA, 0.5 mM dithiothreitol, 5% glycerol, and 1 mM phenylmethylsulfonyl fluoride for 2 h. The dialysate was then centrifuged at 13,000 ϫ g for 15 min, divided into aliquots, and stored at Ϫ80°C. Protein concentration was determined by the method of Bradford (12) with a kit from Bio-Rad using bovine serum albumin as a standard.
In Vitro DNase I Footprinting-In vitro DNase I footprinting was performed as described by Sandaltzopoulos and Becker (13), employing a Sure Track footprinting kit (Amersham Biosciences). DNA fragments containing bp Ϫ268 to ϩ17 (for noncoding strand analysis) of the human SREC gene were labeled with biotin using the biotin-labeled primer, 5Ј biotin-GCGCTCGGGTTCGTCTGG-3Ј, and 5Ј-ATGGAAAGC-CGTGGCTCCCA by PCR. Samples were amplified for 15 cycles under the following conditions: denaturation for 1 min at 94°C, annealing for 1 min at 55°C, and extension for 1 min at 72°C employing AmpliTaq Gold. Samples were then purified using a QIAquick PCR Purification Kit (Qiagen). About 50 pg of the labeled DNA fragment was incubated with 25 g of nuclear extract or bovine serum albumin and 1 g of poly(dI-dC)⅐poly(dI-dC) for 25 min on ice and then 2 min at room temperature. In some experiments, recombinant human Sp1 (1 footprinting unit; Promega) or EZF-2 (as described below) was used in addition to bovine serum albumin. Samples were treated with increasing doses of DNase I (0.25-1 units with bovine serum albumin and 0.5-2 Kunitz units with nuclear extract) at room temperature for 2 min. Samples were analyzed as described (13,14) using a 6% denaturing polyacrylamide/urea gel and an SEQ-Light TM DNA sequencing system. For comparison, DNA-sequence reactions of the genomic DNA around the transcription-initiation site of the SREC gene were run in parallel (G, A, C, and T) on the gel using a primer (5Ј biotin-ATGG-AAAGCCGTGGCTCCCA-3Ј).
Mutagenesis-The "megaprimer" method (15) was used to mutagenize bp Ϫ128 to ϩ17 of the 5Ј-flanking sequence of SREC promoter cloned in pGL3-Basic. In the first step, PCRs were carried out using an internal mutagenic primer with an external vector-targeted primer (RVprimer3, 5Ј-CTAGCAAAATAGGCTGTCC-3Ј). The internal reverse primers (mutated bases underlined) were IR3-mut (5Ј-GAGAGGGAAT-AGGTGGCTGG-3Ј) and Sp1-mut (5Ј-CCGGAGGATAGGGCTGCCAC-3Ј). The PCR product from each set of first step PCRs (e.g. RVprimer3/ IR3-mut and RVprimer3/Sp1-mut) were purified and used in the second step PCR with an external vector-targeted primer (GLprimer2, 5Ј-CTTTATGTTTTTGGCGTCTTCC-3Ј) primer. Products from the secondstep PCRs were digested with KpnI and XhoI and cloned into pGL3-Basic and sequenced to confirm the presence of the desired mutations. To construct the Sp1/IR3 double mutant, the IR3 site mutant cloned in pGL3-Basic was used as template. The mutagenic primer was Sp1-mut. The second step PCR product was cloned and sequenced as described above.
Cloning of Novel Transcription Factor from HUVEC cDNA Library-A target reporter construct and a reporter yeast strain were prepared by using a matchmaker one-hybrid system (CLONTECH). Three tandem copies of the nucleotide sequence between Ϫ110 and Ϫ89 of the 5Ј-flanking sequence of the SREC gene (5Ј-CACCCCCTCCCTCT-CAGGGAGG-3Ј) were prepared by annealing two complementary strands of synthetic oligonucleotides, which include the cohesive ends of EcoRI and XbaI, and inserted upstream of the reporter gene HIS3 in pHISi. The resulting target-reporter construct was integrated at the HIS3 locus of yeast strain YM4271, yielding a reporter yeast strain. This reporter yeast strain was used as a host strain for the library screening. HUVEC cDNA libraries fused to the GAL4 activation domain, which were constructed in yeast expression vectors pGAD424 (CLONTECH), were screened according to the manufacturer's instructions. Selection media included 10 mM 3-amino-1,2,4-triazole to inhibit background HIS3 expression. After screening, cDNAs from ␤-galactosidase assay positive clones were sequenced. Among them, half of the cDNAs encoded a novel protein, EZF-2, and its activity was further confirmed by in vitro footprinting assay. EZF-2 was expressed in E. coli W3110 as an RGS-His 4 epitope-tagged form as described (16) and purified employing chelating Sepharose and DEAE-Sepharose column chromatographies.

Cloning of SREC Gene and Alternatively Spliced cDNAs-A
human genomic DNA library (CLONTECH) was screened by PCR using two oppositely oriented oligonucleotides from the 5Ј region and then the 3Ј region of the human SREC cDNA (7). Two overlapping clones that covered the entire coding region of SREC were identified. A restriction map of the insert of these clones and the location of SREC exons and introns is shown in Fig. 1. The SREC gene spans 12 kb and contains 11 exons and 10 introns. The exon-intron boundaries of the SREC gene were determined by sequencing the SREC gene entirely, and it was found that all donor and acceptor splicing sequences contained consensus GT and AG dinucleotides, respectively. Most of the exons were small, with the exception of exon 11 (ϳ2 kb), which contained a 3Ј-untranslated sequence.
As shown in Fig. 2, various alternatively spliced forms of SREC cDNAs were cloned from the peripheral blood leukocyte cDNA library. The nucleotide sequences of these clones contained exon-intron boundaries on the SREC gene, indicating that they were alternatively spliced transcripts for SREC. Among them, two transcripts, SREC-2 and SREC-4, encoded soluble forms of SREC. By alternative splicing, a stop codon was introduced before the transmembrane region of the SRECs. Whereas SREC-2 encoded 342 amino acids, SREC-4 encoded 337 amino acids and was five amino acids shorter than SREC-2 in domain 8 of the EGF-like repeat. Three transcripts, including the previously reported SREC-1 (7), encoded membrane-bound forms of SREC. SREC-3 encoded a different intracellular domain from SREC-1 by skipping the last 37 bp of exon 9. SREC-5 encoded a shorter extracellular domain by skipping the last 34 bp of exon 5, the entire exon 6, and the first 47 bp of exon 7. When these three membrane-bound forms were expressed transiently in COS-1 cells, they all showed receptor activity toward DiI-acetylated LDL (data not shown).
Mapping of the Transcriptional Start Site-As the first step to characterize the human SREC gene promoter, the transcriptional start site was mapped by RLM-RACE employing poly(A) ϩ RNA from HUVEC. In the first amplification, no specific PCR product was detectable using "outer RNA adaptor primer" as the sense primer complementary to "RNA adaptor" and SREC outer primer as the gene-specific antisense primer. The resulting cDNA extending to the cap site was then amplified by nested PCR in the presence of an "inner RNA adaptor primer" as the sense primer complementary to the "RNA adaptor" and an SREC inner primer as the gene-specific antisense primer. Products of ϳ120 bp were then cloned and sequenced.
The results of sequencing analysis demonstrated that all selected clones terminated at the same nucleotide (designated as ϩ1) located 35 bp upstream of the translation site (Fig. 3), indicating that this nucleotide is the major transcriptional initiation site of the human SREC gene. The 5Ј-CA-3Ј nucleotide pair at this position is the most common transcription initiation site (17). The 5Ј-flanking sequence of the human SREC gene lacks TATA and CCAAT boxes near the transcriptional start site as in the case of endothelial cell-specific KDR/ flk-1 and intercellular adhesion molecule 2 genes (18,19). Comparison of the 5Ј-flanking sequence with sequences in the transcription factor data base revealed several putative DNAbinding elements that may play roles in basal and regulated SREC transcriptional activity (Fig. 3). Putative binding sites of MZF1, GATA-1, GATA-2, AML-1a, Lyf-1, and Sp1 are shown in Fig. 3.
Basal Promoter Region of the Human SREC Gene-To determine whether the 5Ј-flanking sequence of the human SREC gene possesses functional promoter activity, we directionally subcloned an Aor51HI (Eco47III) DNA fragment (ϳ2.4 kb) at the SmaI site of the pGL3-basic promoterless luciferase plasmid. The Ϫ2358/ϩ17 region of the human SREC gene ligated to the luciferase reporter gene (Ϫ2358/ϩ17 Luc) showed substantial promoter activity in HUVEC (Fig. 4). To identify the region regulating the basal promoter activity of the human SREC gene, a series of 5Ј-deletion constructions were transiently transfected into HUVEC and assayed for luciferase activity. Since the 5Ј-deletion construct deleted to Ϫ108, but not that deleted to Ϫ98, retained the promoter activity (Fig. 4), the region between nucleotides Ϫ108 and Ϫ98 was considered to contribute to the basal promoter activity of the human SREC gene in HUVEC.
To define DNA-protein interactions in this region and determine the proteins binding to the SREC promoter, we first performed in vitro DNase I footprinting assays using a biotinlabeled fragment from Ϫ268 to ϩ17 of the SREC promoter as a probe to disclose interactions protecting the coding strand. A densely protected nucleoprotein complex was formed in the presence of HUVEC nuclear extract but not of control and CASM nuclear extract on nucleotide sequence from Ϫ108 to Ϫ65 (Fig. 5).
Cloning of a Transcription Factor Bound to the 5Ј-Flanking Region of the Human SREC Gene-Since the nucleotide sequence between Ϫ108 to Ϫ98 was important for the regulation of SREC gene expression, we next attempted to clone a transcription factor bound to the 5Ј-flanking region of the SREC gene by employing a one-hybrid system. After screening of the HUVEC cDNA library, one candidate cDNA that interacted with the nucleotide sequence from Ϫ110 to Ϫ89 of the SREC promoter was cloned, and the nucleotide sequence was determined. The data base search revealed a match with an unknown human placental cDNA (AK001999); sequences from the human genome sequencing projects, unknown cDNAs (AF074985 and BC021282), and chromosome 19 (AC024580); and highly similar sequence from unknown cDNA (XM_063584) and chromosome 15 at 15q21.3 (AC090515 and AC025918). The deduced amino acid sequence and the schematic structure of the cDNA are shown in Fig. 6. The deduced amino acid sequence contained leucine-rich region/SCAN (SRE-ZBP, Ctfin51, AW-1, and number 18 cDNA) domain (20) and the C2H2-type zinc finger domain. We temporary termed it as EZF-2, since another Krü ppel-like zinc finger protein termed EZF was already cloned from the HUVEC cDNA library (21). Recombinant EZF-2 was expressed in E. coli W3110 as an RGS-His 4 epitope-tagged form and purified by chelating Sepharose and DEAE-Sepharose column chromatographies.
Sp1 and EZF-2 Bind to the SREC Promoter-To identify the nuclear protein(s) present in HUVEC nuclear extract that contribute to the nucleoprotein complex, we compared the binding activity of HUVEC nuclear extract with that of recombinant Sp1 in in vitro footprinting assays using the human SREC fragment from Ϫ268 to ϩ17 as a probe. To our surprise, recombinant Sp1 had essentially the same binding activity as HU-VEC nuclear extract (Fig. 5), thereby presenting the possibility that Sp1 was attributable mainly to the binding activity of HUVEC nuclear extract to the fragment. As expected, EZF-2 also bound to the region between Ϫ122 and Ϫ88 bp, immediately upstream of the Sp1 binding site (Fig. 5).
EZF-2 Modulates the SREC Promoter Activity-To examine if elevated Sp1 or EZF-2 expression was sufficient to increase SREC promoter activity, HUVEC were transiently transfected with the Ϫ2358/ϩ17 Luc, human SREC promoter luciferase reporter construct, and expression plasmids for either Sp1 or EZF-2 (Fig. 7A). Overexpression of EZF-2 caused a nearly 2.5-fold increase in SREC promoter activity in HUVEC. On the other hand, expression of Sp1 showed little effect on the basal SREC promoter activity in HUVEC. Transfection of reporter construct together with Sp1 and EZF-2 expression plasmids showed nearly the same promoter activity as that of reporter construct with EZF-2 expression plasmid.
To confirm the significance of these transcription factors further, we constructed mutant plasmids having mutated Sp1 and/or IR-3 motifs. As shown in Fig. 7B, mutation of either Sp1 motif or IR-3 motif caused a decrease in promoter activity of nearly 50% in HUVEC. Moreover, mutant plasmid having a mutation in both motifs had little promoter-enhancing activity. These results suggest that both Sp1 and EZF-2 play roles in the induction of SREC gene expression.
Regulation of the SREC Promoter by Cytokines-To determine whether or not the expression of the SREC gene was physiologically and/or pathologically regulated, HUVEC transfected with Ϫ2358/ϩ17 Luc construct were treated with various biologically active substances. Whereas treatment with phorbol 12-myristate 13-acetate or platelet-derived growth factor-AB slightly increased luciferase activity in HUVEC transfected with the Ϫ2358/ϩ17 Luc construct, treatment with inflammatory cytokines such as IL-1␣, IL-1␤, or TNF-␣ inhibited the expression of luciferase activity (Fig. 8). These results suggested that the expression of the SREC gene could be regulated by these inflammatory cytokines. DISCUSSION In this study, we have characterized the human gene encoding SREC. The gene is ϳ12 kb in length and contains 11 exons and 10 introns. Only a single SREC transcription start site was Ϫ268 to ϩ17 of the SREC promoter in pGL3-Basic plasmid (1.6 g) and 0.06 g of either Sp1 or EZF-2 expression plasmid in pEAK8 vector (Edge Biosystems) were co-transfected into HUVEC along with 0.032 g of pRL-CMV plasmid using Lipofectin reagent. The pRL-CMV vector containing the Renilla luciferase gene under control of the herpes simplex virus thymidine kinase promoter was used as an internal control for differences in transfection efficiency and cell number. B, identification of important regulatory elements in the SREC promoter by site-directed mutagenesis. Luciferase constructs containing the wild type Ϫ128 to ϩ17 or derivatives thereof, mutated as indicated by stars, were transfected with pRL-CMV vector into HUVEC. Corrected liciferase activity was calculated, and promoter activity was expressed as a percentage of that of wild type promoter. identified in cultured HUVEC using RLM-RACE. Alternatively, spliced SREC variants were identified by cloning of full-length cDNA clones from a human peripheral blood leukocyte cDNA library. The sizes of the cDNAs were all ϳ3.5 kb and could not be distinguished by Northern blot analysis. Among them, two encode soluble forms of SREC, and we have observed that SREC-2 was indeed expressed in HUVEC (data not shown). Although the expression of these alternatively spliced forms of SREC in other tissues is not known, it is conceivable that such soluble receptors may modulate the pathophysiological function of membrane-bound receptor through interaction with its ligand (22)(23)(24)(25). Further studies are required to elucidate the pathophysiological function of these soluble forms.
Transient expression of the 5Ј-flanking region fused to the luciferase gene in HUVEC showed substantial promoter activity. We have defined the basal promoter activity within the 5Ј-flanking region of the human SREC gene, a fragment extending from Ϫ108 to Ϫ98, that, when deleted, reduced the activity of a luciferase reporter construct to the background level in transient transfection assays. A consensus binding site for the transcription factor Sp1 was present 78 bp upstream of the transcription start site. Sp1, which is ubiquitously expressed in mammalian tissues, is believed to regulate cell typespecific expression of genes such as KDR/flk-1 in HUVEC (18). These results suggest that elements within this region, which have proximal potential binding sites for Sp1, are critical for endothelial cell-specific expression of SREC and that these elements are likely to serve as the core promoter of this TATAless gene.
To define DNA-protein interactions in this region and determine the proteins binding to the SREC promoter, we performed in vitro DNase I footprinting assays. A densely protected nucleoprotein complex was formed on the nucleotide sequence from Ϫ108 to Ϫ65 bp in the presence of HUVEC nuclear extract but not of control and CASM (Fig. 5), suggesting that this region is critical for the transcriptional regulation of the SREC gene. Furthermore, we have cloned a novel transcription factor, EZF-2, that bound to this critical region, by employing a onehybrid system. The domain structure and the amino acid sequence of EZF-2 are similar to those of several zinc finger transcription factors such as Krü ppel-like ZNF191 (26) or ZNF24. The leucine-rich region/SCAN domain has been shown to be able to mediate homo-and hetero-oligomerization (27). The C2H2 zinc finger is composed of two short ␤-strands followed by an ␣-helix and the conserved cysteines and histidines that coordinate with a zinc ion. Since many members of the C2H2 zinc finger gene family regulate differentiation processes and genetic mutation in some zinc finger genes has been associated with specific human diseases (28,29), it is interesting to elucidate the role of this novel transcription factor. Further work is required.
Recombinant Sp1 had essentially the same binding activity as HUVEC nuclear extract (Fig. 5), suggesting that the major binding protein in HUVEC nuclear extract was attributable to Sp1. On the other hand, recombinant EZF-2 was found to bind to the sequence between Ϫ122 and Ϫ88, immediately upstream of the Sp1 binding site. Overexpression of EZF-2 increased SREC promoter activity by nearly 2.5-fold in HUVEC, while expression of Sp1 alone showed no significant effect on basal SREC promoter activity in HUVEC. On the other hand, mutation analysis of reporter constructs revealed that the intact Sp1 and IR-3 motifs were required for the maximal expression of the SREC gene. These results suggested that Sp1 and EZF-2 can induce SREC gene expression independently. It is conceivable that in the overexpression experiments, Sp1 had little effect because of the high and saturation level of endogenous Sp1 expression in HUVEC. The precise mechanism of the endothelial cell-specific expression of the SREC gene remains to be elucidated.
Employing the 5Ј-flanking region of SREC, the regulation of SREC expression in HUVEC in vitro by various stimuli was determined at the transcriptional level. Among them, inflammatory cytokines such as IL-1␣, IL-1␤, and TNF-␣ inhibited promoter activity. A parallel decrease in DiI-acetylated LDL uptake was observed after treatment with these stimuli (data not shown). These results suggested the possibility that cytokines regulated the expression level of SREC and thus modified the inflammatory process. Pathophysiological implications of these regulations should be elucidated by analyzing the biological function of SREC.
In summary, we have characterized the human SREC gene promoter in detail and found that a novel transcription factor, EZF-2, regulated the gene expression. The gene expression was regulated by several cytokines, such as TNF-␣, IL-1␣, and IL-1␤, suggesting that SREC plays a role in the course of inflammation. Since several investigators have reported that the stimulation of scavenger receptors in endothelial cells induces the expression of proteins related to vascular functions, the elucidation of the underlying mechanism for SREC gene regulation will provide new insight into the pathological function of the SREC.