(Received for publication, August 19, 1996, and in revised form, February 7, 1997)
From the Department of Biochemistry, F. Edward
Hébert School of Medicine, Uniformed Services University of the
Health Sciences, Bethesda, Maryland 20814, * Department of Molecular
Biology and Biochemistry, University of California, Irvine, California
92717, and the § Dana-Farber Cancer Institute and Department
of Cell Biology, Harvard Medical School,
Boston, Massachusetts 02115
The expression of human squalene synthase (HSS)
gene is transcriptionally regulated in HepG-2 cells, up to 10-fold, by
variations in cellular cholesterol homeostasis. An earlier deletion
analysis of the 5-flanking region of the HSS gene demonstrated that
most of the HSS promoter activity is detected within a 69-base pair sequence located between nucleotides
131 and
200. ADD1/SREBP-1c, a
rat homologue of sterol regulatory element-binding protein (SREBP)-1c binds to sterol regulatory element (SRE)-1-like sequence (HSS-SRE-1) present in this region (Guan, G., Jiang, G., Koch, R. L. and Shechter, I. (1995) J. Biol. Chem. 270, 21958-21965). In our
present study, we demonstrate that mutation of this HSS-SRE-1 element
significantly reduced, but did not abolish, the response of HSS
promoter to change in sterol concentration. Mutation scanning indicates
that two additional DNA promoter sequences are involved in
sterol-mediated regulation. The first sequence contains an inverted
SRE-3 element (Inv-SRE-3) and the second contains an inverted Y-box
(Inv-Y-box) sequence. A single mutation in any of these sequences
reduced, but did not completely remove, the response to sterols.
Combination mutation studies showed that the HSS promoter activity was
abolished only when all three elements were mutated simultaneously.
Co-expression of SRE-1- or SRE-2-binding proteins (SREBP-1 or SREBP-2)
with HSS promoter-luciferase reporter resulted in a dramatic
increase of HSS promoter activity. Gel mobility shift studies
indicate differential binding of the SREBPs to regulatory sequences in the HSS promoter. These results indicate that the transcription of the
HSS gene is regulated by multiple regulatory elements in the
promoter.
Squalene synthase (farnesyl diphosphate-farnesyl diphosphate farnesyltransferase, EC 2.5.1.21) catalyzes the reductive head-to-head condensation of two molecules of farnesyl diphosphate (FPP)1 to form squalene, the first specific intermediate in the cholesterol biosynthesis pathway. The expression of squalene synthase, as that of several other key enzymes in the pathway, such as 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase, HMG-CoA synthase, and FPP synthase, is highly regulated by the cholesterol homeostasis in cells (1-4). It has previously been demonstrated that the expression of squalene synthase is regulated transcriptionally (5), but the mechanism for this regulation is unknown.
Low density lipoprotein receptor (LDLR) and HMG-CoA synthase are regulated by interaction between recently described transcriptional factors, sterol regulatory element-binding-proteins (SREBPs). These interact with sequences called sterol regulatory elements (SRE-1 and SRE-2) that exist in the promoters of the two genes (6-8). Two functionally related SREBPs, SREBP-1 and SREBP-2, have been purified from human cells and hamster cells, and the mechanism by which they regulate the expression of LDLR and HMG-CoA synthase has been studied extensively (9-11). Human SREBP-1 and SREBP-2 are 47% identical. At the NH2-terminal region of each protein, there is a basic-helix-loop-helix leucine zipper (bHLH-Zip) structure that serves as a transcriptionally active domain. Next to the bHLH-Zip domain there are two membrane attachment domains. Nascent SREBP-1 and SREBP-2 are localized in the ER by these domains, and they are inactive in stimulating transcription. At lower concentrations of sterol in cells, an ER-associated, sterol-sensitive protease is activated and proteolytically activates the SREBPs by a cleavage at a site between the leucine zipper and the membrane attachment domains to release the bHLH-Zip domain (12). The active bHLH-Zip segment of SREBP-1 was shown to localize at the nucleus. This form of SREBP binds to SRE-1 in promoters of LDLR and HMG-CoA synthase, which results in the transcriptional activation of the two genes (13, 14). The sterol-regulated release of the active form of SREBP-2 from cell membranes was shown to require two sequential proteolytic cleavages of the transmembrane segment of the nascent protein (12). Human SREBP-1 and SREBP-2 were chromosomally mapped to 17p11.2 and 22q13, respectively (15). It was recently demonstrated that the expression of fatty acid synthase (16-19), an essential enzyme in fatty acid biosynthesis, and acetyl-coenzyme A carboxylase (20) are also transcriptionally activated by the binding of SREBP-1 to the sterol regulatory elements in their promoters. This observation provides evidence that directly links the metabolism of fatty acids and cholesterol and the importance of the SREBP family of transcription factors in the regulation of lipid metabolism. More recently, another sterol regulatory element, SRE-3, was defined in FPP synthase promoter (21). This SRE-3, with a limited identity to SRE-1, also interacts with SREBP-1 and thereby initiates the transcription of the gene. The same sequence exists in the promoter of HSS as well (22).
The difference in cellular function of two different SREBPs is presently unknown, since they share similar structure. The two proteins bind in vitro to SRE-1 with the same specificity. In cultured cells they stimulate SRE-1-containing promoters in an additive fashion (10). Both undergo proteolytic activation processing and presume to stimulate the expression of the sterol-regulated enzymes by the same mechanism.
Similar sterol-related transcription regulation may exist in rodents as well. The rat homologue of SREBP-1c has been independently cloned as adipocyte determination- and differentiation-dependent factor 1 (ADD1) (18). ADD1/SREBP-1c is expressed predominantly in white adipose tissue, brown adipose tissue, and liver, and its expression is induced at a very early stage of adipogenesis (18, 19). Unlike other bHLH-Zip transcription factors, ADD1/SREBP-1c has a unique dual DNA binding specificity to the E-box (CANNTG) and SRE-1 elements (23). This dual DNA binding specificity of ADD1/SREBP-1c homodimer is conferred by the presence of atypical tyrosine residue in the basic domain instead of an arginine residue, which is present in all other bHLH proteins (23). Recently, it has been demonstrated that ADD1/SREBP-1c has an important role in adipocyte differentiation as well as fatty acid metabolism. Retroviral expression of ADD1/SREBP-1c stimulates adipocyte differentiation and induces the expression of adipocyte-specific genes under strong differentiation conditions (19). Furthermore, in several cell lines, ADD1/SREBP-1c can induce expression of fatty acid synthase and lipoprotein lipase, two key enzymes involved in fatty acid metabolism (19). However, its detailed involvement in the regulation of cholesterol homeostasis is presently unclear.
In an earlier report, we demonstrated the importance of a 69-bp
sequence in the promoter region of human squalene synthase (HSS) gene
in its sterol-mediated transcriptional regulation. Gel retardation and
DNase I footprinting verified that ADD1/SREBP-1c binds to the modified
SRE-1 (HSS-SRE-1) element present within the 69-bp region. An 8-bp
sequence identical to 8 out of the 10 bp of the SRE-1 element
(SRE-1(8/10)) at nucleotide 101 did not show protection in the
footprinting assay (22). These findings suggest that the transcription
of HSS may be regulated similarly to the LDLR and HMG-CoA synthase,
namely by the interaction between SRE-1 and SREBPs.
The present study focuses on identifying sequences of the cis-acting DNA elements in the HSS promoter involved in the sterol-mediated regulation of the HSS gene, and their regulatory relationship. We also describe the differential interaction between these elements and the trans-acting SREBPs.
Standard molecular biology
methods were used (24). DNA sequencing was performed using the
Sequenase 7-diaza-dGTP DNA sequencing kit (Amersham Corp.). Restriction
enzymes and modification enzymes were purchased from New England
Biolabs. [-32P]dCTP was obtained from DuPont NEN.
Lipofectin reagent was purchased from Life Technologies, Inc.
Anti-NF-YA (a antibody against the A-subunit of human NF-Y protein) was
purchased from Biodesign.
pHSS1kb-Luc, the fully functional HSS promoter-luciferase reporter gene, was constructed as described previously (22). ADD1-403 is an expression vector containing a cDNA fragment of the active transcriptional factor ADD1/SREBP-1c (amino acids 1-403) controlled by an SV40 promoter (pSVSPORT1). ADD1-DN, an inactive protein, is the same construct as ADD1-403 with an alanine point mutation replacing tyrosine in the bHLH motif of ADD1/SREBP-1c at residue 320 (19). pCMV-CSA10 contains the coding sequence from human SREBP-1a from amino acids 1-490 inserted into the pCMV5 expression vector as described previously (25). pCMV-CS2 is a similar expression plasmid that produces amino acids 1-481 of the human SREBP-2 protein. RSABB (SREBP-1 amino acids 234-490) and C2BB (SREBP-2 amino acids 236-401) are bacterial fusion proteins for SREBP-1 and SREBP-2, respectively, and are designed to express the indicated protein fragments as fusion proteins in the pRSET vector (invitrogen). The recombinant proteins were purified from Escherichia coli by nickel chelation chromatography.
Cell Culture and Transient TransfectionHuman hepatoma
HepG-2 cells were maintained in 35-mm plates in a minimum essential
medium (Life Technologies, Inc.) supplemented with 10% fetal bovine
serum, 1 mM glutamine, 1 mM pyruvate, 100 units/ml penicillin, and 100 µg/ml streptomycin at 37 °C and 5% CO2. Transient transfection of HepG-2 cells was conducted
using the Lipofectin reagent transfection procedure according to the manufacturer's manual and previously described (22). Each transfection included 5 µg of HSS-promoter-luciferase reporter DNA and 4 µg of
DNA of pCMV--galactosidase (pCMV-
GAL) expression vector for calculation of transfection efficiency. When indicated, various amounts
of one of the following DNAs was also included in the transfection
mixture: pCMV-CSA10, pCMV-CS2, ADD1-403, or ADD1-DN. On the day
following transfection, the cells were treated for 24 h with
either sterol(+) (1 µg/ml 25-OH cholesterol plus 10 µg/ml
cholesterol in 10% lipid-depleted serum-containing medium) or
sterol(
) (5 µg/ml lovastatin substituting the sterols in the same
medium). The procedures for cell harvesting and the assays for
luciferase activity and
-galactosidase activity have been described
previously (22). Relative luciferase activity is expressed as the ratio
of luciferase activity in relative light units to
-galactosidase
activity (in A574).
The construct used for
the generation of the replacement mutations is a pBluescript phagemid
containing the 1-kb 5-flanking region of HSS gene (pHSS1kb-BS, +73 to
934) (22). First, single-strand DNA was prepared from pHSS1kb-BS
following the procedure recommended in the Sculptor in vitro
mutagenesis kit manual (Amersham Corp.). Briefly, TG1 host cells were
transformed with pHSS1kb-BS to obtain fresh transformants. On the
following day, one colony was selected and grown in 2 × YT medium
for 3 h. VCS-M13 helper phage (Stratagene) was then added to the
culture and the incubation continued for an additional 4 h. The
cells were then lysed and the single-strand DNA was purified by
polyethylene glycol 8000 precipitation, phenol extraction, and finally
by ethanol precipitation. Oligonucleotide-directed mutagenesis was
performed as described previously (24). To overcome the high GC content
in the mutation site, annealing and extension of mutation primers were
performed at a relatively higher temperature. Therefore, the following
procedure was employed. Single-strand DNA and mutation primer
containing a unique restriction enzyme site (see Table I) were
denatured at 94 °C for 5 min and slowly cooled to 25 °C to allow
annealing. The annealed primer was then extended in a buffer containing
2 mM DTT, 6 mM MgCl2, 0.2 mM each of dNTP, and 5 units of VentTM DNA polymerase in a
total volume of 100 µl. The reaction was kept at 0 °C for 5 min,
incubated at room temperature for 10 min, 55 °C for 30 min, and
finally, 70 °C for 2 h. Then, 400 units of T4 DNA ligase was
added and the reaction incubated at 37 °C for 1 h to allow for
extension and ligation. The resulting DNA was used to transform
XL-1Blue competent cells. The mutated colonies were selected by colony hybridization and restriction enzyme digestion using the unique site
introduced into mutation primers (Table I) and verified by DNA
sequencing. The mutated plasmids were used for the subsequent subcloning of the mutated 1-kb HSS insert into HindIII and
BamHI sites of a pXP1 luciferase reporter gene vector as
described previously for the native promoter (22). The resulting
mutations in pXP1 constructs were confirmed again by DNA sequencing,
and the plasmid DNAs were prepared for transient transfections. This
procedure was utilized for the introduction of a single transversal
replacement mutation in the pHSS1kb-Luc promoter. The various single
replacement mutations that were introduced in the region
48 to
190
relative to the transcription starting site are shown in Fig.
1.
|
The primers described in Table I were used in the procedure described above to generate single replacement mutations. To prepare replacement mutations in multiple regions, oligonucleotides containing several of the mutated sequences were employed. Thus, the doubly mutated pHSSm12-Luc is a combination mutation of both the HSS-SRE-1 and the Inv-SRE-3 sequences (containing mutations present in pHSSm10-Luc and pHSSm1-Luc); the doubly mutated pHSSm13-Luc is a combination mutation of both the HSS-SRE-1 and the Inv-Y-box sequences (containing mutations present in pHSSm1-Luc and pHSSm6-Luc); the doubly mutated pHSSm23-Luc is a combination mutation of both the Inv-Y-box and the Inv-SRE-3 sequences (containing mutations present in pHSSm6-Luc and pHSSm10-Luc); and finally, the triply mutated pHSSm123-Luc is the combination mutation of all three sequence elements (containing mutations present in pHSSm6-Luc and pHSSm12-Luc).
Electrophoretic Mobility Shift AssaysNuclear extracts were prepared from HepG-2 cells treated with either 5 µg/ml 25-OH cholesterol or 5 µg/ml lovastatin based on a protocol described originally by Dignam et al. (26). The probes used in the shift assays are synthetic oligonucleotides. The sequences of these oligonucleotides are shown below. Binding condition and labeling of probes have been described previously (22). In each assay, 0.2 ng of DNA from a 32P-end-labeled probe (8 × 104 cpm) was incubated on ice for 30 min with one of the following proteins: 1.5 µg of RSABB, 0.1 µg of C2BB, or 4 µg of nuclear extract. In the competition assay, excess amounts (as indicated for different experiments) of non-labeled competing DNAs were added to the reaction just before the addition of the probe. Mobility immunosupershift assays were carried out under similar conditions except the nuclear extracts were incubated with anti-NF-YA (2 µg) on ice for 1 h prior to addition of the probe. After the binding reaction, the mixture was electrophoresed on a 4% polyacrylamide gel and the binding signal was detected by autoradiography. The following DNA probes were used in the above assays (sequence name indicated by underlining, mutations indicated in bold).
![]() |
![]() |
We have previously reported that the transcriptional regulation of
the HSS gene is localized within a 69-bp sequence in HSS promoter. In
addition, we have shown that ADD1/SREBP-1c binds to an HSS-SRE-1
element existing in this region (22). To verify the regulatory effect
of this interaction between ADD1/SREBP-1c and HSS-SRE-1 in cultured
cells, the effect of ADD1/SREBP-1c on HSS promoter activity was tested
by expressing different ADD1/SREBP-1c constructs together with
pHSS1kb-Luc in HepG-2 cells grown in the presence of sterol. A 3.6-fold
increase in HSS promoter activity was observed in the presence of a
transcriptionally active portion of ADD1/SREBP-1c protein (ADD1-403).
However, its effect was much weaker than that observed in cells grown
in the absence of sterols (Fig. 2). In addition, when a
point mutation is introduced into the bHLH motif of ADD1/SREBP-1c
(ADD1-DN), the protein loses most of its activation effect. These
results indicate that the HSS-SRE-1 element in the HSS promoter may be
functional.
To further confirm that HSS-SRE-1 is involved in
sterol-dependent regulation of HSS transcription, a
mutation replacing the HSS-SRE-1 element was introduced into
pHSS1kb-Luc. The resulting effect of the mutant (pHSSm1-Luc) on HSS
promoter activity was tested in a transient transfection system using
HepG-2 cells both in the absence (sterol-depleted, sterol()) or
presence (sterol (+)) of sterol. Unexpectedly, mutation of the
HSS-SRE-1 sequence element still allowed 70% regulation of HSS
promoter activity in response to sterol depletion (Fig.
3A). This result suggests that in addition to
the HSS-SRE-1 element, other sequences in the promoter may also
contribute to the sterol-related regulation of HSS transcription.
Therefore, to identify DNA sequence elements that are involved in this
regulation, a series of transversion mutations were made in which
sequences within
48 to
190 promoter region were successively
replaced. The effect of these mutations on the HSS promoter activity in
both the presence and absence of sterols was investigated. As expected,
in addition to the sequence mutated in pHSSm1-Luc, a DNA region
that includes the sequences mutated in pHSSm4-Luc,
pHSSm5-Luc, and pHSSm6-Luc (nucleotides
113 to
153) also showed
significant involvement in the regulation of HSS promoter activity
(Fig. 3A). The largest decrease in sterol-mediated regulation was observed for pHSSm6-Luc. The relative luciferase activity in cells transfected with pHSSm6-Luc was reduced to about 30%
of that observed for pHSS1kb-Luc transfected cells. The mutations in
pHSSm2-Luc, pHSSm3-Luc, pHSSm7-Luc, pHSSm8-Luc, and pHSSm9-Luc did not
significantly affect the promoter activity or the sterol-mediated response.
Sequence analysis reveals that an SRE-3 element, previously reported in
FPP synthase promoter (21), also exists in the HSS promoter
(nucleotides 138 to
147) and overlaps the region covered by pHSSm4
and pHSSm5 (Fig. 1). This SRE-3 element has an inverted orientation in
comparison to the SRE-3 in FPP synthase promoter. Therefore, it is
named Inv-SRE-3. In addition, an A (at position
145) in the HSS
element substitutes a C present in the FPP gene. An inverted perfect
Y-box sequence CCAATCAG (Inv-Y-box,
121 to
132) is present 3
to
the latter sequence. Y-box element is reported to be involved in major
histocompatibility complex class II gene regulation (27). The mutation
in both pHSSm5-Luc and pHSSm6-Luc modified this Inv-Y-box. Mobility
shift assays indicate that the Inv-Y-box sequence found in the HSS
promoter does, in fact, bind the NF-Y transcription factor present in
nuclear extracts of HepG-2 cells. When labeled Inv-Y-box probe is used,
a clear mobility shift signal is observed (Fig. 4,
lanes 1 and 2). This signal diminishes in the
presence of a competing non-labeled Inv-Y-box DNA probe (lanes
5-8) but not with a competing HSS-SRE-1 probe (lanes
9-12). The final verification that the interacting transcription factor is indeed NF-Y comes from an immunological assay in which a
specific anti-NF-YA interaction with the nuclear extract is shown to
supershift the signal to a more retarded position (lanes 3 and 4).
The observation that none of these single region replacement mutations resulted in complete loss of HSS promoter activity made us consider the possibility that multiple elements in HSS promoter, and a few transcription proteins factors that bind to these elements, may act in concert for maximal sterol-mediated regulation of HSS transcription. To verify this possibility, five more mutants were prepared and their effects on HSS promoter activity were investigated in the transiently transfected HepG-2 cells. First, a mutation that exactly replaced the Inv-SRE-3 element was introduced into pHSS1kb-Luc (pHSSm10-Luc). Transient expression of this construct in cells demonstrated that this replacement mutation indeed decreases the response of the HSS promoter to sterol depletion (Fig. 3A). Mutations in multiple regions were then generated. Thus, a mutation of both the HSS-SRE-1 and the Inv-SRE-3 sequences (pHSSm12-Luc), a mutation of both the HSS-SRE-1 and the Inv-Y-box sequences (pHSSm13-Luc), a mutation of both the Inv-Y-box and the Inv-SRE-3 sequences (pHSSm23-Luc), and finally, a mutation of the above three elements (pHSSm123-Luc) were tested.
The effects of these combination mutations on HSS promoter activity are depicted in Fig. 3B. pHSSm12-Luc and pHSSm23-Luc reduced luciferase activity in sterol-depleted cells by more than 70%. pHSSm13-Luc reduced activity in these cells by about 90%. Finally, pHSSm123-Luc transfected into the cells failed almost completely to show sterol-mediated transcriptional regulation. These results indicate an additive regulatory effect of the three transcription elements in the HSS promoter.
The transcriptional protein factors that bind to SRE-1 in the LDLR and
the HMG-CoA synthase promoters and SRE-3 in the FPP synthase promoter
are SREBPs (8, 9, 21). To determine whether the known SREBPs play a
similar role in the regulation of HSS gene transcription, we expressed
two vector constructs that encode transcriptionally active forms of
SREBP-1 and SREBP-2. A dose response in the activation of the
pHSS1kb-Luc of the plasmids of three transcription factors encoding
ADD1/SREBP-1c, SREBP-1, and SREBP-2 shows maximal activation at 0.5 µg of DNA/plate for SREBP-2 and approximately 1 µg of DNA/plate for
the other two. However, the level of activation observed for the
rat-derived ADD1/SREBP-1c is only about 40% of that observed for the
human-derived SREBPs at 1 µg of DNA/plate (Fig.
5).
We then investigated the binding of the two human-derived SREBP protein
factors to the regulatory cis-DNA sequence elements present
in the HSS promoter. Fig. 6 shows the result of
electrophoretic mobility shift assay using synthetic oligonucleotides
with various sequences of the regulatory elements. SREBP-1 is shown to
bind strongly to oligonucleotides containing both the LDL-SRE-1 and the
HSS-SRE-1 sequences. It also shows a weak, yet significant, binding to
the FPP-SRE-3 sequence of the FPP synthase promoter as well as a weak
binding to an oligonucleotide containing the SRE-1(8/10) element. It
does not bind to the oligonucleotides containing either the Inv-SRE-3
or the Inv-Y-box sequences, which are present in the HSS promoter.
In comparison, SREBP-2 shows similar binding to SRE-1, HSS-SRE-1, and a much stronger binding signal to SRE-3 and the Inv-SRE-3 sequences. In addition, it also displays a strong binding signal to the oligonucleotide containing the SRE-1(8/10) sequence.
Previous studies identified a 69-bp sequence within the HSS
promoter region that is important for the transcriptional regulation of
the gene. We have previously hypothesized that an SRE-1-like element
(HSS-SRE-1) at 180 to
189 is responsible for this regulation (22).
In the present report, we demonstrate that this HSS-SRE-1 is not the
only element regulating HSS promoter activity and gene transcription. A
second sterol response element, Inv-SRE-3, located at nucleotides
138
to
147, originally described in an inverted orientation for FPP
synthase promoter, is also shown to contribute to the activation of HSS
promoter. Mutation of this Inv-SRE-3 sequence element reduced the
activation of the HSS promoter in response to sterol depletion. This
reduction in activation is similar to the effect caused by mutation of
the HSS-SRE-1 element, suggesting that both elements are involved in
the sterol-mediated regulation of HSS transcription. Mutations in both
the HSS-SRE-1 and Inv-SRE-3 sequences decreased the sterol depletion
response by more than 70%, indicating an additive effect of the two
elements. Although these two elements are known to exist and function
in other genes involved in cholesterol production, this study of HSS is
the first report to show the presence of both transcriptional elements
in a gene in this pathway.
The functional relationship of the regulatory sequences HSS-SRE-1 and
Inv-SRE-3 in HSS promoter is unknown. The observation that 70% of the
regulation remains, when either of these sequences remains unmodified,
may indicate that they may also function independently and not
necessarily regulate HSS transcription in concert. Our multiple
sequences mutation studies demonstrate the importance of the sequence
modified in pHSSm6-Luc for HSS promoter activity. We hypothesize that
the Inv-Y-box located within this region may function with both the
HSS-SRE-1 and Inv-SRE-3 elements and, thereby, enhance their
transcriptional regulatory activity. It has been demonstrated that the
transcriptional factor NF-Y, which binds to the Y-box sequence, is
important in the regulation of FPP synthase and HMG-CoA synthase: two
additional genes involved in the production of sterols (28). Evidence
also shows that the activation of transcription of the LDLR gene
requires a synergistic cooperation of SREBP and the common
transcriptional factor Sp1 (25, 29). Therefore, the emerging theme
suggests that there is a synergistic interaction between SREBPs and one
of the more constitutively expressed transcriptional factors, such as
Sp1 or NF-Y, for the regulation. Since no regulatory transcriptional
sequence is identified immediately 3 to the HSS-SRE-1, and since
mutation of the Sp1 sequence at nucleotides
52 to
57 does not
affect the regulation, the possibility is raised that both the
HSS-SRE-1 and the Inv-SRE-3 elements function coordinatively with
the Inv-Y-box located 3
to both. Alternatively, we cannot eliminate
the possibility of yet another, hitherto unknown, regulatory element
that may exist within the DNA region
113 to
153, be
sterol-responsive, and function in the regulation of HSS
transcription.
It is of interest that the SRE-3 and the Y-box in the HSS promoter are
both in inverted orientation compared with those found in FPP synthase.
Their relative position in the promoter is reversed as well. In FPP
synthase, the Y-box is 5 to the SRE-3 sequence and they are separated
by a 21-bp spacer. In HSS, the Inv-SRE-3 is 5
to the Inv-Y-box with a
9-bp separation. In both genes, the two elements are arranged
sequentially to allow orientation in the same direction. It is possible
that this arrangement is required to enable the interaction between
SREBP and NF-Y.
There are two adjacent perfect Sp1 binding sites (GC-box) within the
promoter region of HSS. The observations that pHSSm9-Luc did not show
repressed activity and that pHSSm123-Luc loses activity completely
suggest that the most 5 Sp1 sequence element may not be important in
the regulation of HSS. The maximal activation of pHSS1kb-Luc by the two
SREBPs as well as by the rat-derived ADD1/SREBP-1c was obtained at
approximately 1 µg plasmid DNA/plate. However, this maximal effect is
not the same for all three. A much lower activation was obtained by
ADD1/SREBP-1c at this concentration. This activation could not be
increased by addition of plasmid DNA. While the actual concentrations
of activating factors within the cells is unknown, these results
indicate that HSS may be relatively unresponsive to ADD1/SREBP-1c.
The Inv-Y-box sequence is shown to bind to the NF-Y transcription factor, which is present in nuclear extracts of HepG-2 cells (Fig. 4). No apparent difference was observed in the amount of this transcription factor present in cells treated with either lovastatin or 25-OH cholesterol (Fig. 4). Since the NF-Y factor is not tissue-specific, and the Y-box is reported to exist in nonsterol-related genes (27, 30, 31), it raises the possibility that NF-Y may interact with one of the SREBPs for the sterol-mediated regulation of HSS transcription similar to the interaction reported for the FPP gene (28).
The work described here, which provides evidence for differences in interaction with regulatory sequences between SREBP-1 and SREBP-2 (Fig. 6), may indicate distinct functional specificity for the two proteins. This corresponds with the report that the expression of SREBP-1 and SREBP-2 in the hamster liver is not coordinated during variations in sterol concentrations (32). Accordingly, it was proposed that SREBP-1 is responsible for basal transcription of LDLR and HMG-CoA synthase, whereas SREBP-2 is primarily responsible for the sterol-mediated inducible transcription of these genes in whole animal.
We have also observed differences between the two SREBPs in their binding to the identified transcription sequences in the gel-shift mobility assays (Fig. 6). Both the LDLR-SRE-1 and the HSS-SRE-1 sequences are shifted by the two SREBPs. However, while SREBP-2 shifts the oligonucleotide containing the SRE-3 sequence, the binding to SREBP-1 is very weak (for FPP-SRE-3) or completely absent (for Inv-SRE-3). In fact, there are distinct differences between the strength of the signal for FPP-SRE-3 and Inv-SRE-3. Clearly, the FPP-SRE-3 gene sequence generates a stronger signal in the gel mobility shift assays as compared with the HSS gene sequence. However, there is an indication that the study of binding of specific sequences to SREBPs by gel shift mobility assays may have limited significance in the understanding of the activation mechanism of the HSS promoter since the SRE-1(8/10) sequence showed differential mobility shift effects for the two SREBPs. It strongly binds to SREBP-2 and shows a very weak signal with SREBP-1. However, this 8-bp sequence by itself was reported to be non-functional for the LDLR (9) and its replacement mutation in the HSS promoter (pHSSm7-Luc) retained promoter activity and its response to sterols (Fig. 3).
The presence of multiple sterol regulatory elements in the HSS promoter may explain the high sterol-mediated transcriptional regulation of this gene that we have previously reported (5). The detailed interaction between these regulatory elements and the corresponding transcription factors is currently under investigation.