Cholesterol Biosynthesis from Lanosterol

The 7-dehydrocholesterol reductase (Dhcr7) is the terminal enzyme in the pathway of cholesterol biosynthesis. We have previously reported that sterol depletion in vivo caused a significant induction of both liver mRNA and enzyme activity of Dhcr7 (Bae, S.-H., Lee, J. N., Fitzky, B. U., Seong, J., and Paik, Y.-K. (1999) J. Biol. Chem. 274, 14624–14631). In this paper, we also observed liver cell-specific sterol-mediated Dhcr7 gene induction in vitro by sterol depletion in rat hepatoma cells, suggesting the presence of sterol-mediated regulatory elements in the Dhcr7 gene. To understand the mechanisms responsible for regulating Dhcr7 expression, we have isolated the 5′-flanking region of the gene encoding rat Dhcr7 and have characterized the potential regulatory elements of the gene that are responsible for sterol-mediated regulation. The Dhcr7 promoter contains binding sites for Sp1 (at −177, −172, −125, and −20), NF-Y (at −88 and −51), and SREBP-1 or ADD1 (at −33). Deletion analysis of the Dhcr7 gene promoter (−1053/+31), employing a nested series of Dhcr7-luciferase constructs, demonstrated that the −179 upstream region of the gene is necessary and sufficient for optimal efficient sterol-regulated transcription. DNase I footprinting and electrophoretic mobility shift assay showed that the SRE1/E box (−33/−22) involved in sterol response of many sterol-related enzyme genes was protected specifically by the overexpressed recombinant ADD1. Mutational analysis for the functional relationship between the identified cis-elements in this region indicate that one of the binding sites for Sp1 (GC box at −125) and NF-Y (CCAAT box at −88) plays a cooperative role in the sterol-mediated activation, in which the latter site also acts as a co-regulator for SREBP-activated Dhcr7 promoter activity. We believe that Dhcr7 is the first enzyme characterized with a sterol-regulatory function in the post-lanosterol pathway. This may be important for understanding the coordinated control of cholesterol biosynthesis as well as the molecular mechanism of Smith-Lemli-Opitz syndrome-related protein in mammals.

The 7-dehydrocholesterol reductase (Dhcr7, 1 EC 1.3.1. 21) catalyzes the reduction of the ⌬ 7 -double bond of sterol intermediates, which is the terminal reaction, in the pathway of cholesterol biosynthesis from lanosterol (1)(2)(3). Studies of this enzyme have recently drawn our attention due to its implication in both cholesterol metabolism and mammalian developmental biology (4 -6). In particular, mutations of Dhcr7 has been involved in genetic disease such as Smith-Lemi-Opitz syndrome (SLOS) (7)(8)(9), which is an autosomal recessive disorder with various developmental abnormalities. SLOS is characterized by elevated 7-dehydrocholesterol in serum (10 -18). This connection between cholesterol and the developmental program in mammals was established based on the critical role of cholesterol in autoprocessing of sonic hedgehog (shh) protein (5,6,19), a morphogen, which binds to, patched protein in developing the central nervous system and limbs. For example, the mouse mutated in shh showed SLOS-like phenotype (20,21).
Recently, cDNAs encoding Dhcr7 from mouse (22), rat (4), human (7), and Arabidopsis thaliana (23) have been isolated. Dhcr7 mRNA was expressed mainly in liver and its regulation appeared to be controlled by tissue sterol contents as well as some cholesterogenic inhibitors (2,4). However, molecular cloning and the nature of sterol-mediated transcriptional control of the Dhcr7 gene in any species have not been reported previously. Furthermore, the relative importance of multiple motifs at the 5Ј-flanking region in transcriptional regulation of the Dhcr7 gene is also not known. In order to characterize the 5Ј-flanking region and the response of multiple regulatory elements to sterols, we isolated the 5Ј-flanking region of the rat Dhcr7 gene and the complete genomic sequence (GenBank TM accession number AF279892). 2 In this paper, we describe the promoter activity and the sterol-mediated regulation of this gene by various mutation analyses, with emphasis on the proximal promoter (Ϫ179 to ϩ1), which controls transcriptional induction of the gene in response to starvation of cellular sterols.
RNA Preparation and Northern Blot Analysis-RNA was isolated by phenol-chloroform extraction using TRI reagent Kit (Molecular Research Center Inc.). Total RNA (20 g) was electrophoresed in a 1% formaldehyde-agarose gel and vacuum-transferred to Hybond N ϩ membrane (Amersham Pharmacia Biotech) with 20 ϫ SSC (150 mM NaCl, 15 mM sodium citrate, pH 7.0). After prehybridization of the membranes at 42°C for 5 h in 5 ϫ SSC, 10 ϫ Denhardt's solution, 100 g/ml denatured herring sperm DNA, the membranes were hybridized to 32 P-labeled rat Dhcr7 cDNA probe for 18 h. Then the membranes were washed with 0.1% SSC, 0.1% SDS at 65°C. Hybridization signals were visualized by autoradiography using Hyper x-ray film with an intensifying screen at Ϫ70°C. The signals that appeared in the membranes were quantitatively analyzed with a BAS 2500 system (Fuji Photo Film Co.) and normalized to the quantity of ␤-actin mRNA expression.
Cloning of the 5Ј-Flanking Region of the Rat Dhcr7 Gene-At the initial stage of this work, rat Dhcr7 promoter was amplified from DNA provided in the rat Promoter Finder TM DNA walking kit (CLONTECH) according to the manufacturer's instructions. The primer r7R-9 (ϩ49 ϳ ϩ28) 5Ј-GCGTGCGGGATCCGGGCGGTTG-3Ј, corresponding to the cDNA for rat Dhcr7, and anchor primer AP1 of the promoter finder kit were used in the primary PCR. The primary PCR products were diluted and used for the secondary PCR with nested primer r7R-48 (ϩ31 ϳ ϩ7) 5Ј-GTTGATTCCAAGCTCCAGCAGCGCC-3Ј in the 5Ј region of Dhcr7 and anchor primer AP2. Secondary PCR products were cloned into pT7Blue(R)-T vector and sequenced with an ABI automated DNA sequencing system. The cloned Dhcr7 promoter region was also confirmed by data obtained from genomic clone sequencing.
Cell Culture and Transfection of Cells-HepG2, CHO-K1, and H4IIE cells were grown in RPMI medium 1640 (Life Technologies, Inc.) supplemented with 5% (v/v) FBS, 1 mM glutamine, and 10 g/ml gentamy-cin in a 5% CO 2 incubator at 37°C. Cells plated onto 6-well plates were grown to 50 -70% confluence before transfection. Two g of test constructs were co-transfected into HepG2 cells with 0.2 g of Renilla luciferase control vector, pRL-SV40 (Promega). Each plasmid containing Dhcr7 promoter-luciferase fusion gene and 2.0 l of Lipofectin reagent (Life Technologies) was diluted into 80 l of Opti-MEM I (Life Technologies). Plasmid DNA and Lipofectin reagent were then combined and added to each plate according to the Lipofectin manufacturer recommendations. The cells were transfected on serum-free media for 18 h and then were switched to a medium with 5% LPDS plus 10 M Lovastatin (designated as sterol starvation condition) or 1.5 g/ml 25-hydroxycholesterol (designated as sterol rich condition). After incubation for 48 h, with the appropriate medium, cells were harvested, and extracts were assayed for luciferase activity (see below for details).
Preparation of Cell Extracts and Luciferase Enzyme Assays-Following transfection and drug treatment, the cells were washed with phosphate-buffered saline and lysed in 0.5 ml of 1 ϫ passive lysis buffer (Promega). Cell extracts were assayed for Firefly and Renilla luciferase activities using the Dual-luciferase Reporter Assay System TM according to the manufacturer's recommendations (Promega). Amounts of lysates employed for the Firefly luciferase activity assays of test constructs were normalized to the Renilla luciferase activities.
DNase I Footprinting Assays-32 P-Labeled DNA probes (Ϫ258/ϩ41) were prepared from Dhcr7 gene and incubated with 0.2 or 2 g of purified ADD1 protein in Buffer A (10 mM Tris (pH 7.6), 100 mM KCl, 10 mM MgCl 2 , 1 mM EDTA, 1 mM dithiothreitol, and 4 g of poly(dI-dC)⅐poly(dI-dC)) at room temperature for 30 min in a total volume of 20 l. At the end of the reaction, 80 l of a solution containing 3 mM CaCl 2 and 6 mM MgCl 2 was added the mixture. After digestion with DNase I, the DNA samples were extracted by a phenol/chloroform solution and electrophoresed on a 6% sequencing gel with a same end-labeled sequencing ladder.

Induction of Dhcr7 Gene by Sterol Starvation-
We have previously reported (4) that when rats were maintained on a sterol starvation condition by feeding CL-diet (5% (w/w) cholestyramine plus 0.1% (w/w) Lovastatin in chow) for 14 days, both hepatic mRNA expression and the specific enzyme activity of Dhcr7 were increased in vivo about 3-and 5-fold, respectively. To examine if this type of sterol-mediated Dhcr7 induction can also be seen in cultured cells in vitro, Northern blot analysis was performed using Dhcr7 mRNA obtained from H4IIE cells and CHO-K1 cells that had been maintained in either FBS or LPDS medium. As shown in Fig. 1A, mRNA from the Dhcr7 gene was specifically induced more than 3-fold in H4IIE cells (hepatic cells). However, this induction by sterol depletion was seen only slightly in CHO-K1 cells (non-hepatic cells). Dhcr7 mRNA was highly inducible in a dose-dependent manner when cells were grown under sterol starvation conditions, where cells were treated with the two cholesterogenic enzyme inhibitors, AY-9944 (Dhcr7 inhibitor) and Lovastatin (HMG-CoA reductase inhibitor) ( Fig. 1, B and C). In contrast to this observation, the Dhcr7 gene was suppressed by treatment of cells with 25-hydroxycholesterol (Fig. 1D). These results predicted sterolmediated regulatory sequences in the Dhcr7 gene. To test this possibility, we isolated and analyzed the 5Ј-flanking region of this gene in detail.
Deletion Analysis of Dhcr7-Luciferase Recombinant DNA for Sterol-response Regulation-To identify the transcriptional regulatory regions in rat Dhcr7 gene and to determine whether this region is regulated by exogenous sterols, a nested series of fragments containing various lengths of the 5Ј-flanking region, together with a part of the first exon (ϩ31 out of 52 bp) were generated by either PCR or restriction endonuclease digestion. The parental plasmid used for this study was designated as p7R1084, spanning nucleotides Ϫ1053 to ϩ31. Each deletion construct was prepared by inserting DNA fragments of various lengths into the luciferase gene and then analyzed for promoter activity using a transient assay system, as described under "Experimental Procedures." These deletion constructs were transfected into HepG2 cells and their luciferase activities measured following treatment with sterols (Fig. 3). To normalize activity in different transfected cells, we co-transfected the cells with pRL-SV40, a non sterol-regulated Renilla luciferase expression vector. Results were expressed as "-fold" activation by taking the ratio of the luciferase activities after sterol starvation (Ϫsterol) or sterol supplemented (ϩsterol) conditions. For the analysis of the regulation of the Dhcr7 promoter, HepG2 cells were chosen due to their relatively high transfection efficiency (Ͼ30-fold) as compared with that of H4IIE cells (data not shown). The promoter activity of the longest fragment, beginning at Ϫ1053 (p7R1084), was set at 100% for the results obtained from HepG2 cells in 5% LPDS (plus 10 M Lovastatin; sterol starvation condition). The basal promoter FIG. 1. Effects of LPDS medium, sterol and cholesterollowering drugs on expression of Dhcr7 mRNA. A, after H4IIE and CHO-K1 were incubated for 24 h in either FBS or LPDS medium at 70% confluence, total RNA was isolated from the cells. Twenty g of the total RNAs were electrophoresed on a 1% agarose gel and transferred to a Hybond N ϩ membrane. The filter was hybridized with 32 P-labeled rat Dhcr7 cDNA probe and exposed to film. The same filter was subsequently hybridized with ␤-actin probe. activity was set at the luciferase activity of p7R72 (Ϫ41/ϩ31). As shown in Fig. 3, the promoter activity of the parent construct (p7R1084) was decreased more than 11-fold when sterol (1.5 g/ml 25-hydroxycholesterol) was added into the LPDS media, confirming the presence of sterol-mediated regulatory elements such as the SRE1/E box located at Ϫ33/Ϫ22 in the 5Ј-flanking region (Fig. 2). Deletion of regions spanning nucleotides Ϫ1053 and Ϫ133 (p7R537, p7R318, p7R210, and p7R164) caused no significant change in either promoter activity or sterol-mediated activation of Dhcr7 promoter activity. However, deletion of nucleotides between Ϫ133 and Ϫ117 caused almost a 3.5-fold reduction in both the promoter activity relative to the Ϫ133 region and the sterol-mediated activation. That is, there was a change in the magnitude of sterol-mediated activation between these two constructs (Ϫ133/ϩ31 (13.6fold) versus Ϫ117/ϩ31 (4.0-fold)). This result suggests the presence of positive cis-acting sequences in this region (Ϫ133 and Ϫ117), which seems sensitive to the absence of sterol. This is further confirmed by the fact that the deletion constructs, p7R148 (Ϫ117/ϩ31) and p7R72 (Ϫ41/ϩ31), almost abolished a sterol-mediated activation (ϳ4.0 to ϳ2.9-fold). Therefore, it appears that the two tandem repeat GC boxes at Ϫ177 (GC1) and Ϫ172 (GC2) are less important than other proximal GC boxes (e.g. GC3) for Dhcr7 promoter activity in response to sterols. Addition of the fragment containing sterol-regulated regions which enhanced activity (Ϫ287/Ϫ117) of the reporter gene itself (pGL3-Basic vector; e.g. p7R287-117), exhibited much lower luciferase activity in the absence of sterol than that of basal promoter construct, suggesting the absence of promoter activity in this sequence. A similar pattern of expression was also observed for these constructs when they were transfected into either CHO-K1 cells or H4IIE cells (Fig. 4). Therefore, we have focused on nucleotides spanning Ϫ287 to ϩ31 (p7R318 showed the highest activity on sterol depletion conditions) in the 5Ј-flanking region of the rat Dhcr7 gene for further analysis of sterol response.
To examine which oxysterol may be most effective in suppression of the Dhcr7 promoter, luciferase activities were measured for p7R318 (Ϫ287/ϩ31) in the presence of different oxysterols. As shown in Fig. 5 creases of 150 and 180% in HepG2 cells that had been transfected with p7R318. The concentration of Lovastatin or AY-9944 was empirically determined prior to this experiment. These results are consistent with those of endogenous mRNA expression under similar conditions (Fig. 1, B and C). As expected, addition of 25-hydroxycholesterol to LPDS medium abolished the AY-9944-or Lovastatin-induced activation of Dhcr7 promoter activity (Fig. 5), suggesting a dominant suppressive role for 25-hydroxycholesterol. In FBS medium (a sterol-supplemented condition), Lovastatin (HMG-CoA reductase inhibitor) produced a 3-fold increase of Dhcr7 promoter activity; AY-9944 (Dhcr7 inhibitor) only resulted in 1.3-fold induction, implying a different mode of action for these inhibitors in FBS media.
Interaction of Recombinant ADD1 Transcription Factor with Promoter Sequence of Dhcr7-To characterize transcription factors binding to the Dhcr7 promoter, ADD1, a rat homolog of human SREBP-1, was overexpressed. DNase I footprinting and EMSA were performed using the DNA fragment covering Ϫ258/ϩ41 and a synthetic 29-mer oligonucleotide binding to ADD1. As shown in Fig. 6A, a recombinant ADD1 binds specifically to the proximal promoter specifically in a dose-response fashion, and its binding to the probe was abolished by the addition of 100-fold molar excess of a double-stranded 29mer oligonucleotide competitor. As shown in Fig. 6B, the sequence elements such as SRE1/E box involved in sterol response of Dhcr7 was protected by overexpressed recombinant ADD1. A footprint covering nucleotides Ϫ40 to Ϫ22 was clearly visible; this protected region includes the consensus SRE1/E box (Ϫ33/Ϫ22). The binding of overexpressed ADD1 to the Dhcr7 promoter was also observed by EMSA (Fig. 6C). The annealed double-stranded oligonucleotides probe ADD (Ϫ46 to Ϫ18), containing SRE1/E box, binds efficiently to ADD1 in EMSA. Its specific binding intensity was reduced in the presence of a 100-fold molar excess of specific competitors such as ABS (contained ADD1-binding site: E box) and SRE-1 consensus sequence (Fig. 6C). However, binding competition was not observed for any of these mutant oligonucleotide competitors (lanes 4 -5 in Fig. 6C). These results suggest that the required sequences for binding to the overexpressed ADD1 are located within the nucleotides between Ϫ40 and Ϫ22, which are overlapped with E box. To examine any binding specificity for Sp1 and NF-Y to the proximal region of Dhcr7 promoter, EMSA was performed using the appropriate probes covering this region in the presence or absence of competitor. Specifically, GC3 at Ϫ125 binds to Sp1 and CCAAT1 at Ϫ88 binds to NF-Y (Fig. 7,  A and B). These results provide direct evidence that sequences proximal to nucleotide Ϫ133 serve as binding sites for both Sp1 and NF-Y.
Mutational Analysis of Sterol-mediated Regulatory Sequences within Dhcr7 Promoter- Fig. 3 demonstrates that the Ϫ287 upstream region of the Dhcr7 promoter is both necessary and sufficient for optimal sterol-regulated transcription. To further investigate functional relationships between the identified cis-elements in the Dhcr7 promoter, which are shown to be involved in the sterol-mediated regulation, we introduced a series of multiple point mutations into this region and fused the resulting mutant promoters to a luciferase reporter gene. Each construct was then transiently transfected into HepG2 cells and assayed for sterol-regulated transcription by maintaining the transfected cells either under induced (sterol starvation) or suppressed (supplementation of 25-hydroxycholesterol) conditions. Constructs containing mutations within the FIG. 6. SREBP-1/ADD1 binds to sequences in the proximal rat Dhcr7 promoter. A, a probe corresponding to nucleotides Ϫ137 to ϩ41 and containing the SRE-1 was generated by PCR from the Dhcr7 promoter. This probe was end-labeled with [␣-32 P]dCTP and incubated with ADD1 (lanes 2 and 3) and competitor double strand-ADD (lane 4). B, the end labeled probe (Ϫ258 to ϩ41) was incubated in the absence or presence of recombinant ADD1. DNase I digestion was performed and the resulting products were separated by denaturing polyacrylamide gel electrophoresis with the same end-labeled sequencing ladder. The presence of a DNase I protected region and its corresponding nucleotide sequence are indicated. The proteins added are 0.2 g (lane 1) and 2 g (lane 2) of ADD1, and 2 g of bovine serum albumin (lane 3), respectively. C, EMSA of double strand-ADD by binding to ADD1. The doublestranded oligomer (ds-ADD) containing the SRE-1 sequence was used as the probe. The labeled probe was added into the reaction mixture that contained the recombinant ADD1 and 100-fold molar excess amount of competitor oligonucleotides as described under "Experimental Procedures." Lane 1, unbound ds-ADD probe; lanes 2 and 3 probe incubated with 20 and 100 ng of recombinant ADD1; lanes 4 -7, show competition assays. Separation was done by electrophoresis on a 5% native polyacrylamide gel. region Ϫ287 to ϩ1, relative to the transcription start site are shown (Fig. 8), and the ratio of reporter gene activity obtained from either induced or suppressed conditions is reported as -fold activation (Ϫsterol/ϩsterol). Mutations of the two distal Sp1-binding sites (⌬GC1/GC2; M1), second NF-Y-binding site (⌬CCAAT2; M7), and SREBP/ADD1-binding site (⌬SRE1/Ebox; M9-M11) caused no change in sterol-mediated response (Fig. 8). However, mutation of either the third Sp1-binding site (⌬GC3; M2, M3) or the first NF-Y-binding site (⌬CCAAT1; M5, M6) caused a reduction of sterol regulatory response. Mutations of the proximal Sp1-binding site (⌬GC4; M4, M12) and the whole region of SRE1/E box or the SRE1 part of SREBP/ ADD1-binding sites (⌬SRE1/E box; M8, ⌬SRE; M13, M14) caused moderate change in sterol regulatory response. Taken together, these results suggest that GC3 and CCAAT1 are essential to the sterol-mediated regulation of the Dhcr7 promoter. The data shown in Fig. 8 confirmed that the sequences located between nucleotides Ϫ33 and Ϫ25 are functional SRE-1 of the rat Dhcr7 promoter and behave as the direct repeat element 5Ј-PyCAPy-3Ј, like other sterol-response genes (35,36).
Activation of Dhcr7 Promoter by SREBP-1-Activation of Dhcr7 gene transcription by sterol starvation required GC3 and CCAAT1, in addition to the SRE1/E box (Fig. 8). Thus, co-regulatory transcription factors, such as Sp1 and NF-Y, may function in concert with SREBP-1 (ADD1) to mediate the sterol-regulatory response of the Dhcr7 gene (32,37,38). To test this, Dhcr7 transcription was measured by SREBP-1 co-transfection (24). As shown in Fig. 9A, the activity of Dhcr7-directed luciferase activity increased in a dose-dependent manner up to 50 ng when cells were co-transfected with pSREBP1. The activity of the Dhcr7 promoter by pSREBP1 was abolished when the sequence containing GC3 and CCAAT1 was deleted (Fig.  9B), suggesting that either Sp1 binding to GC3 or NF-Y binding to CCAAT1 is necessary for SREBP-1-mediated activation of the promoter. To further analyze the specificity and potential mechanism for this activation, we employed mutational analysis to find which regulatory element (i.e. GC3 or CCAAT1) is more critical for activation by SREBP-1 overexpression. These various point mutants in cis-acting elements of the Dhcr7 promoter were compared with wild type promoter for activation where the SREBP-1 was supplied exogenously from an expression plasmid. Interestingly, the mutation of CCAAT1 (M5, 6) and SRE1/E box (M8; Refs. 13 and 14) resulted in significant reduction (i.e. 35 and 58%) of SREBP activated Dhcr7 promoter activity as compared with the wild type (Fig. 10). However, mutation of GC3 (M2, 3) resulted in a less significant reduction of the activation rate when pSREBP1 plasmid was added. This result is also consistent with the data shown in Fig. 9B. A mutant construct of p7R148, in which GC3 was deleted, sustained 5.7-fold activation by co-transfection with pSREBP1. These results clearly demonstrate that SREBP-activated Dhcr7 promoter preferentially required the CCAAT1 element in HepG2 cells. However, GC3 site for Sp1 may contribute to the basal Dhcr7 promoter activity in response to sterol. DISCUSSION The purpose of this study was to determine how multiple motifs in the proximal promoter region are involved in sterolmediated regulation of the Dhcr7 gene. Using combinations of deletion and site-specific mutations in the Dhcr7 promoter, we have identified and characterized cis-acting elements involved in the sterol-mediated regulation, which requires differential interaction of Sp1 (to GC3), NF-Y (to CCAAT1), and SREBP-1 (to SRE1/E box). (Fig. 3, Fig. 8 and Fig. 10). Mapping of the Dhcr7 promoter with respect to sterol-mediated gene expression was important for two reasons. First, Dhcr7 is involved in the final reaction of cholesterol biosynthesis, so control of gene expression by the end product (cholesterol) was intriguing. Second, since mutations of Dhcr7 genes already were found to be associated with SLOS, regulation of its gene expression should be critical for normal development. The results presented in this paper provided at least three interesting points with regard to Dhcr7 gene expression. First, rat Dhcr7 gene expression can be regulated by exogenous sterols including the end product, cholesterol (Fig. 1). In particular, 25-hydroxycholesterol effects seem dominant since induction of Dhcr7 promoter activity, mediated by sterol starvation condition, was abolished by this sterol regardless of the presence of cholesterol biosynthesis inhibitors (Fig. 5). Second, sterol-mediated Dhcr7 gene regulation appears to require participation of specific GC boxes (i.e. GC3 at Ϫ125/Ϫ120) and a CCAAT box (CCAAT1 at Ϫ88/Ϫ81) in addition to the SRE1/E box sequence (Ϫ33/Ϫ22). In addition, the most critical nucleotide sequence(s) in the SRE/E box was found to be Segment II of the SRE/E box ("5Ј-TCAC"-3Ј, Table I) because mutation of this portion (i.e. M13 in Fig. 8) caused more change in sterol responsiveness of SRE/E box as compared with those in other segments (Segments I or III) of SRE/E box sequences, for example, M9, M10, and M11.
Sequence analysis also revealed that the rat Dhcr7 promoter region contains high GC content and lacks a canonical TATA box, suggesting that it has overall similarity to promoters of HMG-CoA reductase and human squalene synthase in the cholesterol biosynthesis pathway and a number of housekeeping genes (29,33,39). Third, the fact that activation of Dhcr7 by addition of AY-9944 or Lovastatin led to a notion that the endogenous sterol can down-regulate the sterol-sensitive Dhcr7 gene, as evidenced by data in Fig. 1. Taken together, the maximal activation of Dhcr7 promoter by sterol starvation requires multiple cis-acting elements that are known to utilize the general transcription factors, NF-Y, Sp1, and ADD1, the rat homolog of SREBP-1, which may have some roles in recruiting SREBP-1, as well as activating it in these promoters. Although both NF-Y and Sp1 appear to work in a concerted manner in sterol-mediated regulation, the former participates in SREBP-1-mediated activation, while the latter acts in general transcription activation. A similar case may be the promoters of fatty acid synthase and the LDL receptor gene which require the SREBP-1 and generic co-regulator Sp1 for synergistic activation (34,40). Recently, the synergistic activation by SREBP-1 and Sp1 has also been reviewed on an artificially assembled chromatin template in vitro and Sp1 has been shown to interact with NF-Ya in vitro in the absence of DNA (41,42). Hence, it is believed that an interaction between NF-Y and Sp1 plays a crucial role in the sterol regulation of rat Dhcr7 promoter. Direct evidence should come from studies of promoter-factor interaction.
In conclusion, SLOS-related Dhcr7 is the first enzyme in the post-lanosterol pathway displaying a typical sterol-mediated gene regulation, thereby providing concerted, coordinated con- FIG. 9. Differential SREBP-activated expression of rat Dhcr7 promoter. A, the p7R318 transfected HepG2 cells (50 -70% confluence) were co-transfected with different amounts of pSREBP1 plasmid for 24 h and incubated in RPMI 1640 and 5% LPDS in the presence of 1.5 g/ml 25-hydroxycholesterol to suppress endogenous SREBPs, and analyzed by a standard transient transfection assay for luciferase expression from a Dhcr7 promoter construct, p7R318. B, HepG2 cells were co-transfected with 50 ng of pSREBP1 and the indicated Dhcr7 promoter deletion constructs along with the reference plasmid pRL-SV40. After transfection, the cells were incubated in RPMI 1640 with 5% LPDS supplemented with 1.  trol of cholesterol biosynthesis between enzymes in the earlier pathway (HMG-CoA reductase and squalene synthase) and those in the distal pathway.