Sterol Regulatory Element Binding Protein-1 Activates the Cholesteryl Ester Transfer Protein Gene in Vivobut Is Not Required for Sterol Up-regulation of Gene Expression*

The plasma cholesteryl ester transfer protein (CETP) plays a central role in high density lipoprotein metabolism and reverse cholesterol transport. Plasma CETP levels are increased in response to dietary or endogenous hypercholesterolemia as a result of increased gene transcription in liver and periphery. Deletional analysis in human CETP transgenic mice localized this response to a region of the proximal promoter which contains a tandem repeat of the sterol regulatory element (SRE) of the 3-hydroxy-3-methylglutaryl-CoA reductase gene. The purpose of the present study was to evaluate the role of the SRE-like element in CETP promoter activity. Gel shift assays using CETP promoter fragments containing these elements showed binding of the transcription factors, sterol regulatory element-binding protein-1 (SREBP-1) and Yin Yang-1 (YY-1). Point mutations in the SRE-like element, designated MUT1 and MUT2, resulted in decreased binding of SREBP-1 (MUT1) or SREBP-1 and YY-1 (MUT2). To determine the in vivo significance of this binding activity, CETP transgenic mice were prepared containing these promoter point mutations. MUT1 and MUT2 transgenic mice expressed CETP activity and mass in plasma. In response to high fat, high cholesterol diets, both MUT1-CETP and MUT2-CETP transgenic mice displayed induction of plasma CETP activity similar to that observed in natural flanking region (NFR) CETP transgenic mice. Moreover, in stably transfected adipocyte cell lines, MUT1 and MUT2 CETP promoter-reporter genes showed significant induction of reporter activity in response to sterols. To evaluate transactivation by SREBP-1, NFR- and MUT1-CETP transgenic mice were crossed with SREBP-1 transgenic mice. Induction of the SREBP transgene in the liver with a low carbohydrate diet resulted in a 3-fold increase in plasma CETP activity in NFR-CETP/SREBP transgenic mice, but there was no significant change in activity in MUT1-CETP/SREBP transgenic mice. Thus, SREBP-1 transactivates the NFR-CETP transgene in vivo, as a result of interaction with the CETP promoter SREs. However, this interaction is not required for positive sterol induction of CETP gene transcription. The results suggest independent regulation of the CETP gene by SREBP-1 and a distinct positive sterol response factor.

The plasma cholesteryl ester transfer protein (CETP) 1 plays a central role in the catabolism of high density lipoprotein (HDL) cholesteryl esters and in reverse cholesterol transport i.e. the transfer of cholesterol from cells in peripheral tissues to the liver via the plasma compartment (1). In the initial steps of reverse cholesterol transport, cellular cholesterol is taken up by HDL and esterified by lecithin:cholesterol acyltransferase. CETP transfers cholesteryl esters from HDL to triglyceriderich lipoproteins with subsequent removal by the liver. The activity of CETP results in formation of smaller HDL species that may be efficient mediators of cellular cholesterol efflux and optimal substrates for the plasma lecithin:cholesterol acyltransferase reaction (2,3). In human genetic CETP deficiency, HDL levels are increased but there appears to be an excess of coronary heart disease (4). Conversely, in hypertriglyceridemic CETP transgenic mice, where formation of small, pre-␤-HDL is increased, CETP expression decreases atherosclerosis even while lowering overall HDL levels (5). These findings could indicate that reverse cholesterol transport stimulated by CETP has an anti-atherogenic role.
Plasma CETP levels and activity are increased in response to dietary or endogenous hypercholesterolemia, presumably reflecting the role of CETP in reverse cholesterol transport. This response has been observed in a variety of species including humans (6 -9). Mice do not normally have plasma CETP activity. However, natural flanking region CETP transgenic mice express the human gene in an authentic tissue pattern and show 2-3-fold induction of plasma CETP levels and hepatic CETP mRNA when placed on a high cholesterol diet (10). Moreover, after breeding onto LDL receptor and apoE gene knock-out backgrounds, a marked 4 -10-fold induction of plasma CETP levels and hepatic CETP mRNA is observed, and there is a close relationship between CETP and plasma cholesterol levels (11). These changes are due to increased CETP gene transcription in liver and peripheral tissues (10).
In an attempt to define the regulatory elements responsible for increased CETP gene expression in response to hypercholesterolemia, we recently prepared CETP transgenic mice with different lengths of natural flanking sequences (12). A transgenesis approach was necessitated by our inability to obtain robust sterol responses of CETP transgenes or CETP promoterreporter constructs in cell culture (13). The in vivo analysis suggested that a region of the CETP promoter between 138 and 370 bp upstream of the transcription start site was required for the up-regulation of CETP gene expression in response to a high cholesterol diet (12). In the middle of this region we identified a tandem repeat of an element highly homologous to the sterol regulatory element (SRE) of the HMG-CoA reductase promoter; also, the presence of an upstream NF-1 site and the promoter position of the SREs were similar in the two genes (14). Because the reductase SRE mediates increased gene expression under sterol depletion conditions (15), this suggested either 1) the existence of separate positive and negative sterol response elements in the CETP promoter or 2) the use of a common element (the SRE) for both positive and negative sterol responses.
To differentiate between these two possibilities we have prepared CETP transgenic mice with point mutations in the tandem SRE-like elements of the CETP promoter and characterized their response to a high fat, high cholesterol diet. Parallel experiments with wild-type and mutant reporter genes were carried out in cell culture. Mutation of the SRE element did not abolish the positive response to sterols, supporting the first possibility above. Also, mutant and natural flanking region CETP transgenic mice were crossed with mice overexpressing SREBP-1, a sterol-regulated transcription factor that binds and activates the SRE of the HMG-CoA reductase promoter (15,16). These experiments demonstrate in vivo transactivation of the CETP promoter by SREBP-1 via the SRE element. The findings suggest independent regulation of the CETP promoter by SREBP-1 and a distinct transcription factor with a positive response to sterols.

EXPERIMENTAL PROCEDURES
Transgene Constructs and Development of Transgenic Mice-NFR-CETP mice used in this study were line 5203 (10). Additional founders containing the NFR-CETP transgene were also prepared. CETP transgenic mice containing promoter point mutations were obtained as follows. The NFR-CETP plasmid containing ϳ3400 bp of 5Ј flank was constructed as described previously (10). The MUT1-CETP plasmid and the MUT2-CETP plasmid containing point mutations in the CETP promoter region (Fig. 1) were created using the megaprimer method (17). The resulting PCR product was subcloned into the TA cloning vector (Invitrogen, Carlsbad, CA). Mutant constructs were screened by restriction enzyme mapping of a novel RsaI site introduced into MUT1-CETP and a novel AflIII site introduced into MUT2-CETP. Mutagenesis was confirmed by automated sequence analysis of positive transformants using the M13 forward and reverse primers on the TA vectors containing the mutated inserts. A 270-bp NsiI/XbaI fragment from each mutant construct was subcloned into the corresponding sites of the NFR-CETP plasmid, replacing the wild-type fragment with the mutated fragment. Mutagenesis was again confirmed by restriction enzyme mapping as described above. The NFR-, MUT1-, and MUT2-CETP plasmids were purified by CsCl gradient centrifugation. The resulting plasmids were digested with NotI and SalI, the ϳ10-kb transgene fragments were isolated by agarose gel electrophoresis and subjected to another round of CsCl gradient fractionation. The transgenes were dialyzed against a solution containing 5 mM Tris (pH 7.5), 10 M NaCl and 0.1 mM EDTA. The purified transgenes were injected into the pronuclei of fertilized mouse eggs taken from superovulated F1 females as described by Walsh et al. (18). Screening of CETP transgenic mice and subsequent breeding was carried out according to Agellon et al. (19).
Protein Preparations-The amino-terminal 490-amino acid fragment of SREBP-1a was prepared as described (20). A plasmid vector containing the YY-1 cDNA fused to glutathione S-transferase was a kind gift of Dr. Kathryn Calame. The YY-1/GST fusion protein was expressed in DH5␣ competent cells and purified by S-linked glutathione-agarose beads using standard techniques (21).
Electophoretic Mobility Shift Assays-Standard gel shift assays were performed as described previously (22). Oligonucleotides were synthesized (Life Technologies, Inc.) corresponding to a region from Ϫ180 to Ϫ220 of the CETP promoters for NFR-CETP (top strand 5Ј-AGGGAT-GGCAAAAATGGTGCAGATGGTGGAGGGGAGACAA-3Ј), MUT1-C-ETP (top strand 5Ј-AGGGATGGCAAAAATGGTACAGATGGTAGAGG-GGAGACAA-3Ј), and MUT2-CETP (top strand 5Ј-AGGGATGGCAAA-AATGTTGCAGATGTTGGAGGGGAGACAA-3Ј). The complementary strands were annealed and end-labeled with [␥-32 P]dATP (3000 Ci/ mmol) and T 4 polynucleotide kinase (Life Technologies, Inc.). Briefly, 50 ng of either the purified amino-terminal of SREBP-1a or the YY-1-GST fusion protein was incubated with 0, 2, 20, or 200 ng of unlabeled probe fragment for 15 min at room temperature. Then, 1 ng of labeled probe was added and allowed to incubate for an additional 30 min on wet ice. Following the incubation, 2 l of loading dye (0.1% bromphenol blue, 0.1% xylene cyanol) was added to each sample. The samples were run out on a 4.5% nondenaturing polyacrylamide gel in 0.22ϫ TBE at 450 V at 4°C. The gels were then dried and subjected to autoradiography on Kodak X-OMAT AR film (Sigma) at Ϫ70°C with an intensifying screen.
SREBP-1a Overexpressing Mice-Mice expressing the active nuclear form of SREBP-1a under the control of the phosphoenolpyruvate carboxykinase promoter as described previously (16) were kindly provided by Drs. Brown and Goldstein. Male SREBP-1a transgenic mice (Tg1a) were mated to female NFR-CETP and MUT1-CETP mice. Progeny were screened for the presence of the SREBP-1a transgene by Southern blot using an ϳ800-bp SacI fragment from the SREBP-1 cDNA labeled with [␣-32 P]dCTP (3000 Ci/mmol) using the Multiprime DNA labeling system (Amersham Pharmacia Biotech).
Dietary Studies-Heterozygous transgenic mice of both sexes from F1 and F2 generations were used in all experiments, except for experiments involving MUT2-CETP mice and simultaneously developed NFR-CETP transgenic mice in which founder mice were studied. Fourto eight-month-old animals were maintained on a chow diet (Purina Chow 5001, Ralston Purina Co., St. Louis, MO). Blood was collected from the tail veins of experimental animals (n ϭ 3-6 animals/group), and the mice were switched to either a high cholesterol diet (20% hydrogenated coconut oil, 0.15% cholesterol) (Research Diets C11070, New Brunswick, NJ) for 2 weeks or to a very high cholesterol/bile salt diet (7.5% cocoa butter, 1.25% cholesterol, 0.5% sodium cholic acid) (Research Diets C13002, New Brunswick, NJ) for 4 days (see figure legends) at which time blood was again collected from the tail veins. NFR-CETP/SREBP-1 and MUT1-CETP/SREBP-1 transgenic mice were treated similarly, except that the experimental diet was a low carbohydrate diet (65% protein, 10% carbohydrate) (Purina Mills Test Diet 5789C-3, St. Louis, MO) used to induce the expression of the SREBP-1a transgene.
Plasma CETP Activity and Mass-CETP activity was determined in diluted plasma using 3 H-labeled CE-HDL as a CE donor and a mixture of human very low density lipoprotein and LDL as a CE acceptor as described previously (23). CETP mass was determined by enzymelinked immunosorbent assay using the Cholesteryl Ester Transfer Protein Quantitative Test Kit (REAADS Medical Products, Inc., Westminster, CO) RNA Analysis-Total RNA was isolated from mouse livers by the RNAzol B method (Tel-Test, Inc., Friendswood, TX), and poly(A) ϩ RNA was isolated from total RNA using a Mini-Oligo(dT) Cellulose Spin Column Kit (5 Prime-3 Prime, Inc., Boulder, CO) and analyzed by Northern blot using as probes an ϳ1-kb EcoRV fragment of the CETP cDNA and an ϳ300-bp EcoRI/XbaI fragment of the ␤-actin cDNA. Signal strengths were quantified by PhosphorImager analysis.
Reporter Constructs-Reporter constructs to test the function of the wild-type, MUT1 and MUT2 promoters were made using the Great EscAPe SEAP Reporter System 2 (CLONTECH Laboratories, Inc.). The plasmid pSEAP Basic was digested with EcoRI, and the ends were filled in with the large (Klenow) fragment of DNA polymerase I. The vector was then cut with KpnI and isolated by agarose gel electrophoresis. A 1670-bp KpnI/EcoRV fragment containing 570 bp of 5Ј flank, exon I, intron I, and 57 bp of exon 2 was isolated from NFR-CETP, MUT1-CETP, and MUT2-CETP and subcloned in-frame into the digested pSEAP Basic vector.
Cell Culture and Transfection Experiments-The 3T3-L1 preadipocytes (American Type Culture Collection, Rockville, MD) were maintained in Dulbecco's modified Eagle's medium with 10% fetal bovine serum at 37°C. Cells were cotransfected with 20 g of CETP-SEAP, MUT1-SEAP, or MUT2-SEAP and 1 g of pWL-neo (Stratagene, La Jolla, CA) using the Stratagene mammalian transfection kit (Stratagene). Cells were selected in medium containing 0.5 mg/ml G418 (Life Technologies, Inc., Gaithersburg, MD) for 2-3 weeks. Resistant colonies were pooled and propagated in mass culture in the presence of 0.5 mg/ml G418. When cells were grown to 80 -90% confluence, the culture medium was replaced with phenol red-free Dulbecco's modified Eagle's medium containing 10% lipoprotein deficient serum (LPDS). The medium was supplemented with a mixture of human LDL, cholesterol, and 25-hydroxycholesterol or ethanol as a control. After a 48-h incubation, medium was collected for alkaline phosphatase activity assays, and the cells were harvested to prepare cell lysates which were used to quantify protein by Lowry's method.
Assays-Alkaline phosphatase assays were carried out using SIGMA FAST pNPP (p-nitrophenyl phosphate) tablets (Sigma Chemical Co., N-9389). Fig. 1 shows the SRE-like elements found in the proximal promoter regions of the HMG-CoA reductase and CETP genes. The CETP promoter contains a tandemly repeated SRE element identical to that in the reductase promoter, except for the last nucleotide of the second repeat. This nucleotide is not crucial for sterol regulation of the reductase promoter (14). Based on point mutations that were shown to markedly reduce sterol regulation of the reductase promoter (14), we made two sets of point mutations in each of the tandem repeats of the CETP promoter designated MUT1 and MUT2 (Fig. 1) and evaluated the effects of these mutations on the binding of transcription factors and on promoter function.

Characterization of Binding of SREBP-1 and YY-1 to CETP Promoter in Gel Shift Assays-
The SRE element of the reductase promoter has been shown to bind two different transcription factors: SREBP-1 and Yin Yang-1 (YY-1) (originally designated as Red25 and Red60) 2 (15,24) in a partially overlapping fashion (14). To determine if the CETP promoter also binds these transcription factors, gel shift assays were carried out using purified transcription factors incubated with wild-type, MUT1-, and MUT2-CETP promoter fragments. Direct binding assays of SREBP-1 with MUT1 and MUT2 probes were complicated by a high nonspecific background, as reported previously (15). However, in competition assays, the wild-type probe bound SREBP-1, and this was partially competed (ϳ75% by laser densitometry) by an excess of cold wild-type probe; under the same conditions, MUT1 and MUT2 probes did not compete with the wild-type probe (Fig. 2). Binding assays using YY-1 showed similar specific binding for wild-type and MUT1 probes but markedly impaired binding of YY-1 by the MUT2 probes (Fig. 3). The broadness of the signal reflects multiple shift complexes, apparent at lower exposures (e.g. as shown in Fig. 4). Similarly, competition assays showed that the binding of YY-1 to the wild-type probe was competed by wild-type and MUT1 fragments but not by MUT2 (Fig. 4). These results show that the CETP promoter binds both SREBP-1 and YY-1. The binding of SREBP-1 is impaired in MUT1 while the binding of both SREBP-1 and YY-1 is defective in the MUT2 promoter.

Response of NFR-, MUT1-, and MUT2-CETP Transgenic
Mice to High Cholesterol Diets-Mutant promoter fragments (MUT1 and MUT2) were then used to replace the corresponding wild-type fragments in the promoter of the NFR-CETP transgene, generating MUT1-and MUT2-CETP transgenes. The MUT1-CETP and MUT2-CETP transgenes were used to prepare transgenic mice which had plasma CETP activity and mass at levels 20 -50% of NFR-CETP transgenic mice (line 5203). Line 5203 was originally selected for its high activity (10), so the absolute differences in activities between lines, from 80% to 350% of human plasma CETP activity, are of uncertain significance. The MUT1-CETP and MUT2-CETP mice with the highest activity were further characterized. We first evaluated the response of MUT1-CETP and MUT2-CETP transgenic mice to high cholesterol diets. After 4 days on the very high cholesterol/bile salt diet, there was a marked induction of plasma CETP activity in NFR-CETP, MUT1-CETP, and MUT2-CETP transgenic mice (Fig. 5). An increase in plasma CETP levels was also seen by immunoassay, which showed a 2.9-fold induction in plasma CETP levels in NFR-CETP mice and a 2.7-fold increase in CETP levels in MUT1-CETP mice in response to a high cholesterol diet. Moreover, a positive response (i.e. a 1.5-to 3.5-fold increase in plasma CETP activity) of the MUT1 transgene to the high cholesterol diet was also confirmed in each of four different founders containing MUT1 transgenes (not shown). To verify that the MUT2 transgene was responsive to a high cholesterol diet, several new founders containing NFR-CETP and MUT2 transgenes were developed. These studies revealed a robust response to the very high cholesterol/bile salt diet for both NFR and two different MUT2 founder animals (Table I). An equivalent positive response to the high cholesterol diet in NFR-CETP and MUT2-CETP founders was observed on several different occasions. To evaluate a potential difference in response to a cholesterol depletion diet, NFR and MUT1 mice were placed on a diet containing 2% colestipol hydrochloride and 0.25% Fluvastatin. However, no statistically significant change in activity was observed in the plasma of either group of animals after 1 week on the diet compared with the chow diet.
NFR-and MUT1-CETP Transgenic Mice Crossed with SREBP-1a Transgenic Mice-The lack of effect of the promoter mutants on up-regulation of CETP expression by dietary cholesterol could indicate that the gel shift activity (Figs. 2-4) is not functionally significant. To evaluate the in vivo significance of the binding of SREBP-1 to the SRE element of the CETP promoter, we crossed SREBP-1a transgenic mice with NFR-CETP transgenic mice or with CETP transgenic mice contain-2 T. Osborne, unpublished data. The probe is a 40-mer oligonucleotide representing the wild-type CETP promoter fragment containing the SREBP-1 binding site (Ϫ180 to Ϫ220, see Fig. 1). An aliquot of the active, nuclear form of SREBP-1 (amino acids 1-490) (50 ng) was incubated for 15 min at room temperature with either no competitor or 200 ng of an unlabeled 24-mer oligonucleotide containing the wild-type, MUT1-, or MUT2-CETP SREBP-1 binding sequence as indicated. 1 ng of 32 P-labeled probe was then added for an additional 30 min, and the reaction was run out on a 4.5% nondenaturing polyacrylamide gel. After electrophoresis, the gel was exposed to Kodak XAR film for 16 h at Ϫ70°C with an intensifying screen.
ing point mutations (MUT1) in the upstream flanking sequence. NFR-CETP transgenic mice containing the SREBP-1a transgene showed increased plasma CETP activity compared with NFR-CETP mice, whereas MUT1-CETP/SREBP-1a mice had the same activity as MUT1-CETP mice (Fig. 6). The SREBP-1a transgene is induced by a low carbohydrate diet (16). In response to this diet, there was a significant 3-fold induction of plasma CETP activity in NFR-CETP/SREBP-1a transgenic mice (Fig. 6). In marked contrast, there was no induction of plasma CETP activity by the low carbohydrate diet in MUT1-CETP/SREBP-1a transgenic mice. In Fig. 6, plasma CETP activity was measured under conditions where activity is proportional to mass; similar results were obtained when hu-man CETP was measured by immunoassay in mouse plasma (not shown). An analysis of poly(A) ϩ mRNA taken from the livers of these animals showed a significant induction of hepatic CETP mRNA in NFR-CETP/SREBP-1a mice which paralleled the increase in plasma mass and activity, but no such induction in the MUT1/SREBP-1a mice (Fig. 7).These results indicate in vivo activation of the wild-type CETP promoter by SREBP-1a but no similar activation of the MUT1-CETP promoter.
Sterol Up-regulation of CETP Promoter-Reporter Constructs in Cell Culture-To further evaluate the transcriptional response of wild-type and mutant CETP promoters, we used an alkaline phosphatase reporter gene stably transfected into preadipocytes. This construct includes the signal peptide of the   FIG. 3. EMSA of YY-1 binding to wild-type, MUT1, and MUT2 probes. EMSAs were carried out using as probes 32 P-labeled 40-mer oligonucleotides corresponding to the NFR-, MUT1-, or MUT2-SREBP-1 binding sequences as indicated (Ϫ180 to Ϫ220). 50 ng of the YY-1-GST fusion protein was incubated for 15 min on wet ice with either no competitor or 2, 20, or 200 ng of each unlabeled self-competitor as indicated. 1 ng of labeled probe was then added to each reaction and allowed to incubate for an additional 30 min on ice. The reactions were run out on a 4.5% nondenaturing polyacrylamide gel. After electrophoresis, the gel was exposed to Kodak XAR film for 16 h at Ϫ70 0 C with an intensifying screen.

TABLE I Increase in plasma CETP mass in NFR-and MUT2-CETP founder
mice in response to a very high cholesterol diet Plasma CETP mass was determined by enzyme-linked immunosorbent assay as described under "Experimental Procedures." The values are expressed as micrograms of CETP/ml of plasma and represent an analysis in duplicate (with a variation of Ͻ10% between duplicates) for each founder before and after 4 days on a very high cholesterol/bile salt diet. NFR-CETP 2 represents a new founder generated with the NFR-CETP transgene simultaneously to the MUT2 founders. CETP gene and results in secretion of alkaline phosphatase activity into media. For each construct multiple clones were obtained and pooled. In response to treatment with sterols, MUT1 and MUT2 reporter constructs displayed highly reproducible, graded increases in reporter gene activity with maximum 1.5-2.0-fold increases in alkaline phosphatase activity above basal expression (Fig. 8). In contrast to this finding, the wild-type promoter showed no significant response to sterol loading. Also, a viral promoter-reporter (pSV-alkaline phosphatase) showed no response to sterols (Fig. 8). Thus, mutation of the SREBP-1 binding site of the CETP promoter appears to unmask positive sterol responsiveness in stable cell lines containing promoter-reporter constructs.

DISCUSSION
By cross-breeding NFR-CETP transgenic mice with SREBP-1a transgenic mice, we have demonstrated in vivo activation of the wild-type human CETP promoter by SREBP-1. In contrast, in another CETP transgenic mouse line containing point mutations in the tandem repeat SRE-like promoter elements (MUT1), there was readily detectable plasma CETP activity, but when the MUT1-CETP transgenic mice were crossed with SREBP-1a transgenic mice, there was no increase in plasma CETP activity. This shows that the effect of SREBP-1 on the activity of the wild-type CETP transgene is due to transactivation of the SRE-like elements of the CETP promoter. Because SREBP-1 is expressed in the liver of chowfed animals (25), these results could indicate that SREBP-1 contributes to CETP gene expression under basal conditions.
Although not transactivated by SREBP-1, MUT1-CETP transgenic mice showed induction of plasma CETP activity equivalent to NFR-CETP transgenic mice in response to a high fat, high cholesterol diet. A positive response to the high cholesterol diet was also obtained in mice containing a second mutant CETP transgene, MUT2, which displayed impaired binding of both SREBP-1 and YY-1. Moreover, in contrast to wild-type promoter constructs, both MUT1 and MUT2 reporter genes displayed positive responses to sterol loading in cell culture. These results provide strong evidence that the positive sterol response does not require binding of SREBP-1 and YY-1 to the CETP proximal promoter. SREBP-1, also known as ADD-1, was described as a basic helix-loop-helix transcription factor expressed during adipose differentiation (26). SREBP-1, and the related molecule SREBP-2, were independently identified as transcription factors activating LDL receptor gene expression under sterol depletion conditions in cell culture (27)(28)(29). SREBP-1 and -2 exist as larger precursor proteins in the cellular endoplasmic reticulum (30,31). Under sterol depletion conditions, an aminoterminal fragment containing the transcriptional activation domain is liberated by protease activity and enters the nucleus (30,31). SREBP-1 also activates SRE-like elements of the ratelimiting enzymes of cholesterol and fatty acid biosynthesis (32)(33)(34)(35). Transgenic mice with hepatic overexpression of SREBP-1 have massive fatty livers due to increased cholesterol and triglyceride synthesis and also display ectopic hepatic expression of lipoprotein lipase (16). The SREBP-1a transgenic mice used in our study also had massive fatty livers after induction with the low carbohydrate diet (Figs. 6 and 7). While these changes (e.g. increased cholesterol biosynthesis) could have contributed indirectly to increased CETP activity, the lack of response of the MUT1-CETP transgene strongly suggests a direct effect of SREBP-1 on the CETP promoter.
The in vivo transactivation of CETP expression by SREBP-1 suggests that SREBP-1 contributes to expression of CETP in the liver and adipose tissue. Although originally discovered as an activity induced in cells by sterol depletion, active SREBP-1 was found in the liver in chow-fed hamsters, but it disappeared under cholesterol depletion conditions (36). In contrast, active SREBP-2 appeared only with cholesterol depletion. Thus, SREBP-1 may contribute to CETP gene expression in chow-fed animals. Recently, SREBP-1 has been shown to be an insulinresponsive factor and to mediate transactivation of certain genes by insulin (37). Although it is not known if the CETP gene is insulin-responsive, this could theoretically be another way that SREBP-1 could influence CETP gene expression.
The SRE-like elements of the HMG-CoA reductase and CETP genes contain a half-site identical with that in the LDL receptor gene, as well as five additional 5Ј-nucleotides (14). This may explain why, in addition to SREBP-1, the SRE-like elements of the reductase and CETP promoters also bind the transcription factor Yin Yang-1. As a result of binding to enhancers in different genes, YY-1 can either activate or repress gene expression, and in some circumstances YY-1 can act as a transcriptional initiator (38). Although our studies do not elucidate the role of YY-1 in the CETP promoter, they clearly indicate that binding of YY-1 to the SRE is not essential for CETP gene expression and is not responsible for sterol upregulation of CETP activity.
Interestingly, the MUT1 and MUT2 CETP reporter genes gave a highly reproducible, although modest, positive response to sterols in cell culture, whereas the NFR-CETP reporter gene gave no response. While we cannot exclude a possible role of clonal variation in these different responses, the more robust induction of the mutant promoters could reflect the decreased contribution of SREBP-1 to gene expression. The predominant isoform of SREBP-1 in cell culture (SREBP-1a) is much more transcriptionally active than the major isoform expressed in tissues in intact animals (SREBP-1c) (39). This high activity of SREBP1a in tissue culture experiments could somehow mask the contribution of the putative positive sterol-responsive factor. Although we have excluded a role of SREBP-1 and YY-1 at the SRE-like element, the unknown positive factor could be acting at this site independent of the point mutations we made ( Fig. 1) or at a distinct site in the CETP promoter. In support of the latter interpretation, the SRE-like element is poorly conserved in the related phospholipid transfer protein gene, which is also induced by hypercholesterolemia (40).
The identity of the putative positive sterol response factor remains unknown. Recently, LXR, a nuclear hormone transcription factor activated by hydroxysterols, was shown to transactivate a multimerized element of the CYP7A promoter that in some species is induced by a high cholesterol diet (41). However, the CETP promoter does not contain the consensus LXR binding site, and preliminary experiments indicate that CETP promoter fragments do not bind LXR. 3 Steroidogenic factor-1 may also activate transcription in response to oxysterols, but this factor is not highly expressed in the liver (42). The orphan nuclear receptor LRH-1, which is homologous to steroidogenic factor-1, could be another possible candidate (43). Alternatively, a novel activity could be involved. Our findings suggest that the CETP gene promoter responds both to SREBP-1 and to an independent positive sterol response factor. This may provide the advantage of complex, graded responses to conditions of sterol deprivation or excess or may allow integrated responses to sterol changes and non-sterol stimuli acting through SREBP-1.