Novel Transcriptional Regulation of the Human CYP3A7
Gene by Sp1 and Sp3 through Nuclear Factor 
B-like Element*
Tetsuya
Saito,
Yoshiki
Takahashi,
Hisashi
Hashimoto, and
Tetsuya
Kamataki
From the Laboratory of Drug Metabolism, Graduate School of
Pharmaceutical Sciences, Hokkaido University, N12W6, Kita-ku, Sapporo,
Hokkaido 060-0812, Japan
Received for publication, July 2, 2001
 |
ABSTRACT |
Human CYP3A7 and CYP3A4 are expressed in fetal
and adult livers, respectively, although the 5'-flanking regions of the
two genes show 90% homology. The purpose of this study was to clarify the mechanism(s) responsible for the transcriptional regulation of the
CYP3A7 gene in human hepatoma HepG2 cells that showed fetal phenotypes. Transfection studies using a series of the
CYP3A7 or CYP3A4 promoter-luciferase chimeric
genes identified a nuclear factor 
B (NF-B)-like element between
nucleotides 
2326 and 2297 that conferred the transcriptional
activation of the CYP3A7 gene. A 1-base pair
mismatch within the corresponding region of the CYP3A4 gene
was sufficient for a differential enhancer activity. A gel shift assay
using nuclear extracts from HepG2 cells showed that Sp1 and Sp3 bound
to the NF-B-like element of the CYP3A7 but not
CYP3A4 gene. Specific activation of the CYP3A7
promoter by Sp1 and Sp3 was confirmed by a co-transfection of the
p3A7NF-B or p3A4NF-B reporter gene with Sp1 or Sp3 expression
plasmid into Drosophila cells, which lacked endogenous Sp
family. Additionally, introduction of mutations into binding sites for
hepatocyte nuclear factor 3
, upstream stimulatory factor 1, and a basic transcription element in the proximal promoter attenuated
luciferase activity to 20% of the level seen with the intact
CYP3A7 promoter. Thus, we conclude that the expression of
the CYP3A7 gene in HepG2 cells is cooperatively regulated
by Sp1, Sp3, hepatocyte nuclear factor 3, and upstream stimulatory
factor 1.
| |
INTRODUCTION |
Cytochrome P-450 (CYP)1
is responsible for the metabolism of a wide variety of xenobiotics and
endogenous substrates (1). Multiple forms of CYP have been shown to be
present in the liver (2). Among the forms of CYPs, CYP3A is the most
abundant form in human livers (3, 4). The human CYP3A subfamily
consists of four members: CYP3A4 (5-7), CYP3A5 (8, 9), CYP3A7 (10), and CYP3A43 (11).
Human CYP3A7 catalyzes the 16
-hydroxylation of
dehydroepiandrosterone 3-sulfate and the 6-hydroxylation of
testosterone (12, 13). This enzyme was first purified and cloned by us from human fetal livers (10, 14). Our studies using Chinese hamster
lung fibroblast cells stably transfected with a CYP3A7 expression
plasmid and using transgenic mice that carried a CYP3A7 cDNA
indicated that the enzyme was also responsible for the bioactivation of
some carcinogens such as aflatoxin B1 and sterigmatocystin (15-17).
The nucleotide sequence of the CYP3A7 (10) cDNA shares 94.2, 88.5, and 82.2% identities with those of CYP3A4 (5-7), CYP3A5 (8, 9), and
CYP3A43 (11) cDNAs, respectively. Northern blot analysis using
CYP3A7 or CYP3A4 cDNA as a probe and total RNAs prepared from human
fetal and adult livers demonstrated that CYP3A7 and CYP3A4 mRNAs
were expressed specifically in human fetal and adult livers,
respectively (18). It has also been reported that a trace amount of
CYP3A7 mRNA could be detected in some adult livers (19). CYP3A7
mRNA was detectable at 50-60 days of gestation (20). This
expression reached a maximal level within a week after birth and then
declined (21). Reciprocally, the expression of CYP3A4 mRNA was not
seen in human fetal livers and began to increase after birth (21). The
molecular mechanism(s) underlying the age-related expression of the
CYP3A7 and CYP3A4 genes, however, has not yet
been elucidated.
As reported previously, we isolated genomic clones containing the
5'-flanking region and the 13 exons spanning ~30 kb of the CYP3A7 or CYP3A4 gene (22, 23). The structures of
exon-intron junctions of the CYP3A7 and CYP3A4
genes were highly conserved between the two genes. In addition, a
5'-flanking region from 1 kb to the transcriptional start site of the
CYP3A7 gene was 91% identical to that of the
CYP3A4 gene. In the same series of studies, we found
characteristic sequences termed "HFLaSE" (P-450 HFLa-specific
element) and "NFSE" (P-450NF-specific element) in the
5'-flanking region of the CYP3A7 and CYP3A4
genes, respectively. However, we could not detect any age-related
factors in gel shift assays using these elements as probes and nuclear
extracts prepared from human fetal and adult livers. Thus, it appeared
that a possible cis-acting element(s) responsible for the
developmental transcription of the CYP3A7 or
CYP3A4 gene was located in regions with minor differences or
far upstream regions of the two genes.
The expression of the CYP3A genes is inducible by a wide
variety of clinically important drugs, including rifampicin,
dexamethasone, and clotrimazole (24). Recently, a human orphan nuclear
receptor termed PXR, steroid and xenobiotic receptor, or
pregnane-activated receptor that mediates the induction of CYP3As has
been found (25-27). The PXR/retinoid X receptor heterodimer has been
reported to interact with a PXR-responsive element that contains the
everted repeat of an imperfect AGGTCA motif separated by 6 nucleotides (called as ER6). This element was identified in the proximal promoters of the CYP3A4 and CYP3A7 genes (25-28). Because
the loss of the expression of PXR protein does not alter the
constitutive expression of murine CYP3A proteins (29), a transcription
factor(s) other than PXR is assumed to contribute to the constitutive
expression of the CYP3A genes in mammals.
In an attempt to analyze the function of a promoter expressed in human
fetal livers, human hepatoblastoma HepG2 cells have been used, because
the cells show fetal hepatic phenotypes such as the increased synthesis
of -fetoprotein, fetal aldolase, and pyruvate kinase and the reduced
synthesis of albumin (30). Successful results obtained so far suggest
that HepG2 cells possess transcription factors necessary for the gene
expression seen in the fetus. Thus, in the present study, HepG2 cells
were employed and transiently transfected with a reporter plasmid in
which the 5'-flanking region of the CYP3A7 or the
CYP3A4 gene was fused to the luciferase gene to identify a
possible cis-acting element(s) specifically responsible for
the expression of the CYP3A7 gene.
In this paper, we provide lines of evidence showing that the expression
of CYP3A7 in HepG2 cells is coordinately regulated by HNF-3, USF1,
and the Sp (specificity protein) family
members, which bind to the NF-B-like element of the
CYP3A7 gene.
| |
EXPERIMENTAL PROCEDURES |
Cell Culture--
Human hepatoma cells, HepG2 and HuH-7, were
purchased from Riken (Tsukuba, Japan). Drosophila SL2 cells
were purchased from Invitrogen (Groningen, The Netherlands). HepG2 and
HuH-7 cells were maintained in Dulbecco's modified Eagle's medium
(Nissui Pharmacy, Tokyo, Japan) supplemented with 10% fetal bovine
serum (BioWhittaker, Walkersville, MD), nonessential amino acids (ICN, Aurora, OH), and 1 mM sodium pyruvate (Life Technologies,
Inc.) at 37 °C in 5% CO2. Drosophila
SL2 cells were maintained in Schneider's medium (Life Technologies,
Inc.) supplemented with 10% fetal bovine serum at 26 °C in 5%
CO2.
RNA Preparation and RT-PCR--
Total RNAs were prepared from
human hepatoma cells according to an acid guanidinium
thiocyanate/phenol/chloroform method (31). To examine the expression
levels of CYP3A7 and CYP3A4 mRNAs in HepG2 and HuH-7 cells, RT-PCR
was performed as described previously (32). Briefly, total RNA (1 µg)
and oligodeoxythymidylic acid primer (0.5 µg) were mixed. The
RNA-primer mixture was incubated at 70 °C for 10 min and then cooled
on ice. Subsequently, Moloney murine leukemia virus RT (20 units)
(Toyobo, Tokyo, Japan), RNase inhibitor (20 units) (Takara, Tokyo,
Japan), and 0.5 mM each of four deoxynucleoside
triphosphates were added to the RNA-primer mixture and then incubated
at 42 °C for 50 min. PCR was performed in a solution containing
cDNA synthesized in a solution containing the above RT reaction
mixture (1 µl), 1.5 mM MgCl2, 0.2 mM each of four deoxynucleoside triphosphates, each primer
(50 pmol), AmpliTaq Gold polymerase (2.5 units) (PerkinElmer Life
Sciences), and 10× AmpliTaq reaction buffer (5 µl) (PerkinElmer Life
Sciences). The reaction was performed for 30 cycles at 94 °C for 1 min, at 55 °C for 1 min, and at 72 °C for 2 min. The PCR products
were subjected to a 2% agarose gel and then visualized by ethidium bromide staining. The sequences of primers used for RT-PCR are as
follows: CYP3A7, 5'-CTATGATACTGTGCTACAGT-3' and
5'-TCAGGCTCCACTTACGGTCT-3'; CYP3A4, 5'-CCAAGCTATGCTCTTCACCG-3'
and 5'-TCAGGCTCCACTTACGGT GC-3'; and glyceraldehyde-3-phosphate
dehydrogenase, 5'-ACCACAGTCCATGCCATCAC-3' and 5'-TCCACCAC
CCTGTTGCTGTA-3'.
Construction of Plasmids--
The oligonucleotide primers used
for the synthesis of DNA fragments or for a site-directed mutagenesis
are as follows: 3A7-S, 5'-TTCACTTGGCCACTGGAAGT-3';
3A7-HindIII, 5'-TGAGAAGCTTCACTACTTTCCTTCCT-3'; 3A4-S,
5'-TGGACAGCCTGTCCACTGAT-3'; 3A4-HindIII,
5'-TGAGAAGCTTCACTACTTTCCTTACT-3'; 3A7/4-5.0 XhoI,
5'-CCTCGAGGT CTATAAAGTAT-3'; 3A7/4-4.5 XhoI,
5'-CCTCGAGGCTCTGGCCACTA-3'; 3A7/4-4.0 XhoI,
5'-CCTCGAGGGAGCTGTTGGTC-3'; 3A7/4-3.5, 5'-CCTCGAGGTGCCC TGAACAC-3';
3A7/4-3.0, 5'-CCTCGAGGTAGATACCACGT-3'; 3A7/4-AS,
5'-ACCTCTCTTCTGGGAAGCTT-3'; 3A7-2.5 XhoI,
5'-CCTCGAGGCCATGTGCTTAGGGT ACAAA-3'; 3A7-2.47 XhoI, 5'-CCTCGAGGAGACCAAGAATAATGTCTGGGAG CACAATA-3'; 3A7-2.41
XhoI, 5'-CCTCGAGGTAATACAGGAAATGAG-3'; 3A7-2.35
XhoI, 5'-CCTCGAGGATCTTCCTTGACACAG-3', 3A7-2.30
XhoI, 5'-CCTCGAGGC ATGTTAGAAGATGTTAC-3'; 3A7-2.0
XhoI, 5'-CCTCGAGGAAAGGCTCTTTGTTTAGGTG-3'; 3A7-1.5
XhoI, 5'-CCTCGAGGATGTACCAGAATTCCCTGGA-3'; 3A7-1.0
XhoI, 5'-CCTCGAGGCATGCAGTATTTCCAGAGAG-3'; 3A7-140
XhoI, 5'-CCTCGAGGTGTGTGATTATTTGCCAACT-3'; 3A7-90
XhoI, 5'-CCTCGAGG CTGCAGGCAGAGCACGGGGGCCCTGCTAC-3'; 3A7-35
XhoI, 5'-CCTCGAG GCTCCAGCATATA-3'; 3A4-2.5 XhoI,
5'-CCTCGAGGCTATGTGCTTAGGGTACAAA-3'; 3A4-2.0 XhoI,
5'-CCTCGAGGAAAGGCTCTCGGTTTAGGTG-3'; 3A4-1.5 XhoI, 5'-CCTCGAGGATGTATCAGAATTCCCTGGA-3'; 3A4-1.0 XhoI,
5'-CCTCGAGGCATGCAGTATTTCCAGAGAG-3'; 3A4-140 XhoI,
5'-CCTCGAGGTG TGTGATTCTTTGCCAACT-3'; 3A4-90 XhoI, 5'-CCTCGAGGCTGCAGGCAGA GCACAGGTGGCCCTGCTAC-3'; 3A7/4-2.25
XhoI, 5'-CCTCGAGGTCTGCCA CTTAATTCCAC-3'; 3A7NF-B-S,
5'-CAGCTCTCAGTAGGCAAGTCCCTACA TGTT-3'; 3A7NF-B-AS,
5'-AACATGTAGGGACTTGCCTACTGAGAGCTG-3'; 3A4NF-B-S, 5'-CAGCTCTCAGTAGTCAAGTCCCTACAT- GTT3'; 3A4NF-B-AS,
5'-AACATGTAGGGACTTGACTACTGAGAGCTG-3'; 3A4-140/120-S, 5'-TGTGTG
ATTCTTTGCCAACT-3'; 3A4-140/120-AS, 5'-AGTTGGCAAAGAATCACACA-3';
3A7NF1mut.-S, 5'-TGTGTCATTATTTGCGAACT-3'; 3A7NF1mut.-AS, 5'-AGTTCGCA
AATAATGACACA-3'; 3A4E-box-S, 5'-CTGCAGGCAGAGCACAGGTGGCCCT GCTAC-3';
3A4E-box-AS, 5'-GTAGCAGGGCCACCTGTGCTCTGCCTGCAG-3'; 3A7USFmut.-S, 5'-CTGCAGGCAGAGCATAAGGGCCCTGCTAC-3'; 3A7USFmut.-AS,
5'-GTAGCAGGGCCCTTATGCTCTGCCTGCAG-3'; 3A7/4BTEmut.-S, 5'-CTCCAGCG
AGGCCTCCTTCT-3'; and 3A7/4BTEmut.-AS, 5'-AGAAGGAGGCCTCGC- TGGAG-3'.
Reporter plasmid p3A7/5.5, which contained the 5'-flanking region
from nucleotides 5564 to +105 (abbreviated to 5564/+105) relative
to the transcriptional start site (22) of the CYP3A7 gene,
was constructed by ligation of the following three DNA fragments: (i)
an XhoI/HindIII fragment (5564/2800) from the
HFLa11-44 (22); (ii) a HindIII fragment (2800/+105)
obtained by means of PCR using oligonucleotide primers, 3A7-S and
3A7-HindIII, and the HFLa11-44 (22) as a template; and
(iii) an XhoI/HindIII fragment from a control
luciferase reporter plasmid, Basic Vector 2 (Toyoinki, Tokyo, Japan).
Reporter plasmid p3A4/5.5, which contained the 5'-flanking region
(-5551/+105) of the CYP3A4 gene (23), was constructed by
ligation of the following three DNA fragments: (i) an
XhoI/EcoRV fragment (5551/2700) from the
NF-32 (23); (ii) a EcoRV/HindIII fragment
(2700/+103) obtained by PCR using oligonucleotide primers, 3A4-S and
3A4-HindIII, and the NF-32 (23) as a template; and (iii)
an XhoI/HindIII fragment from Basic Vector 2. To
construct reporter plasmids p3A7/5.0, p3A7/4.5, p3A7/4.0,
p3A7/3.5, and p3A7/3.0, PCR was performed using 3A7/4-5.0
XhoI, 3A7/4-4.5 XhoI, 3A7/4-4.0
XhoI, 3A7/4-3.5 XhoI, or 3A7/4-3.0
XhoI as a sense primer and 3A7/4-AS as an antisense primer.
Synthesized fragments were digested with restriction enzymes XhoI and HindIII. Resultant fragments were
inserted into the XhoI/HindIII site of
p3A7/5.5. To generate reporter plasmids p3A4/5.0, p3A4/4.5, p3A4/4.0, p3A4/3.5, and p3A4/3.0, PCR was carried out using 3A7/4-5.0 XhoI, 3A7/4-4.5 XhoI, 3A7/4-4.0
XhoI, 3A7/4-3.5 XhoI, or 3A7/4-3.0
XhoI as a sense primer and 3A7/4-AS as an antisense primer.
Synthesized fragments were cleaved with XhoI and
HindIII. Resultant fragments were inserted into the
XhoI/HindIII site of p3A4/5.5. To generate
reporter plasmids p3A7/2.5, p3A7/2.47, p3A7/2.41, p3A7/2.35,
p3A7/2.30, p3A7/2.0, p3A7/1.5, p3A7/1.0, p3A7/140, p3A7/90,
p3A7/35, p3A4/2.5, p3A4/2.0, p3A4/1.5, p3A4/1.0, p3A4/140,
and p3A4/90, PCR was performed using 3A7-2.5 XhoI,
3A7-2.47 XhoI, 3A7-2.41 XhoI, 3A7-2.35
XhoI, 3A7-2.30 XhoI, 3A7-2.0 XhoI,
3A7-1.5 XhoI, 3A7-1.0 XhoI, 3A7-140
XhoI, 3A7-90 XhoI, 3A7-35 XhoI,
3A4-2.5 XhoI, 3A4-2.0 XhoI, 3A4-1.5
XhoI, 3A4-1.0 XhoI, 3A4-140 XhoI, or
3A4-90 XhoI as a sense primer and 3A7-HindIII or
3A4-HindIII as an antisense primer. Synthesized fragments
were digested with XhoI and HindIII. Resultant
fragments were inserted into the XhoI/HindIII
site of Basic Vector 2. Reporter plasmids p3A7/580, p3A4/580,
p3A7/360, p3A4/360, p3A7/62, and p3A4/62 were generated by the
cleavage of p3A7/1.0 or p3A4/1.0 with restriction enzymes
SacI, BglII, and PstI, respectively.
To construct reporter plasmids p3A7Ewild and p3A4Ewild, a region from
2.50 to 2.25 kb was amplified by PCR using 3A7-2.5 XhoI
or 3A4-2.5 XhoI as a sense primer and 3A7/4-2.25
XhoI as an antisense primer. Synthesized fragments were
digested with XhoI and then inserted into the
XhoI site of p3A7/62. The direction of the
inserts was confirmed by a sequence analysis (ABI PRISMTM
377; PerkinElmer Life Sciences). Reporter plasmids p3A7Emut. and
p3A4Emut. were produced by site-directed mutagenesis using 3A4NF-B-S, 3A4NF-B-AS, 3A7NF-B-S, and 3A7NF-B-AS as mutated primers. To construct p3A7NF-B or p3A4NF-B, double-stranded 3A7NF-B or 3A4NF-B was introduced into the SmaI site
of p3A7/62. The copy number of the introduced oligonucleotides was
confirmed by a sequence analysis. Reporter plasmids
p3A7/140/HNF-3mut., p3A7/2.5/HNF-3mut., p3A7/140/NF1mut.,
p3A7/2.5/NF1mut., p3A7/140/USFmut., p3A7/2.5/USFmut.,
p3A7/140/BTEmut., and p3A7/2.5/BTEmut. were produced by
site-directed mutagenesis using 3A4-140/120-S, 3A4-140/120-AS, 3A7NF1mut.-S, 3A7NF1mut.-AS, 3A7USFmut.-S, 3A7USFmut.-AS,
3A7/4BTEmut.-S, or 3A7/4BTEmut.-AS as mutated primers. To produce
reporter plasmids p3A7E/2.0, p3A7E/360, p3A7E/62, and p3A7E/35,
PCR was performed using 3A7-2.5 XhoI as a sense primer and
3A7/4-2.25 XhoI as an antisense primer. Synthesized
fragments were digested with XhoI and then inserted into the
XhoI site of p3A7/2.0, p3A7/360, p3A7/62, and
p3A7/35.
Expression plasmids pPacSp1 (33) and pPacUSp3 (34), were generous gifts
from Dr. Robert Tjian (University of California at Berkeley, Berkeley,
CA) and Dr. Guntram Susuke (Institute fur Molekülarbiologie und
Tumorforschung, Marburg, Germany), respectively. To construct an
expression plasmid for HNF-3 (named pAcHNF-3), HNF-3 cDNA
was isolated from CMV-HNF-3 (35), a generous gift from Dr. Robert H
Costa (University of Illinois, Chicago, IL) by
SpeI/EcoRV digestion and was inserted into the
EcoRV site of pAc5.1/V5-His (Invitrogen, Groningen, The
Netherlands). To construct an expression plasmid for USF1 (named
pAcUSF1), USF1 cDNA was isolated from USF-SR (36) by
XhoI/XbaI digestion and was inserted into the
XhoI/XbaI site of pAc5.1/V5-His. To construct the
HNF-3 expression plasmid (named pCI-HNF-3) for in
vitro transcription and translation, HNF-3 cDNA was
isolated from CMV-HNF-3 by SpeI/EcoRV digestion and was inserted into the SmaI site of pCI-Neo
carrying T7 promoter (Promega, Madison, WI).
Transient Transfection and Dual-Luciferase Assay--
HepG2
cells (2 × 106 cells) were plated onto a 60-mm dish
and then transfected with a reporter plasmid (2-8 µg) and pRL-SV40 (0.1 µg) (Promega) as an internal control by using the methods of
calcium phosphate co-precipitation (37). Four hours after the DNA
transfection, the cells were treated with 20% glycerol for 1.5 min.
Transfection of DNA into SL2 cells was performed as described
previously by Kudo (38). After 36 h, the cells were washed with
phosphate-buffered saline, followed by a dual-luciferase assay
according to the manufacturer's instructions (Promega).
In Vitro Transcription and Translation--
In vitro
transcription and translation assays were carried out using a rabbit
reticulocyte lysate system (Toyoinki, Tokyo, Japan). Briefly, the
pCI-HNF-3 expression plasmid under control of T7 promoter (1 µg)
was added to a reaction mixture (50 µl) containing 50% of a rabbit
reticulocyte lysate, 20 µM complete amino acid mixture, a
ribonuclease inhibitor (40 units), and T7 RNA polymerase (20 units).
The reaction mixture was incubated at 30 °C for 90 min. The reaction
mixture (4 µl) was subjected to a gel shift assay.
Gel Shift Assay--
Nuclear extracts were prepared
from HepG2 cells according to the method of Dignam et al.
(39). The gel shift assay was performed with double-stranded synthetic
oligonucleotides labeled with [
-32P]ATP (Amersham
Pharmacia Biotech) and T4 polynucleotide kinase (Takara). The binding
reaction was carried out with a reaction mixture (10 µl) containing
25 mM Hepes (pH 7.9), 4% Ficoll, 40 mM KCl,
0.5 mM dithiothreitol, 0.1 mM EGTA, 1 mM MgCl2, 5% glycerol, poly(dI-dC) (0.5 µg),
nuclear extracts (10 µg), and a 32P-labeled probe DNA (5 fmol). The mixture was incubated at 24 °C for 30 min. The samples
were resolved on a 4% nondenaturing polyacrylamide gel in 0.5×
Tris-boric acid-EDTA disodium salt at 100 V at room temperature and
visualized by BAS-2500 Imaging Analyzer (Fuji Film, Tokyo, Japan).
Antibodies to Sp1, Sp3, NF-B p50, and USF1 were purchased from Santa
Cruz Biotechnology (Santa Cruz, CA). Antibodies to HNF-3 (35) and NF1
were kindly provided by Dr. Robert H. Costa (University of Illinois,
Chicago, IL) and Dr. Naoko Tanese (New York University, New York, NY),
respectively. The supershift assay was performed using these antibodies
as follows. After incubation of probe DNAs with nuclear extracts as
described above, the antibodies were added to the reaction mixture and
incubated at 24 °C for 1 h. The products were then analyzed by
a gel shift assay. The sequence of oligonucleotides used as probes is
as follows: 3A7NF-B, 5'-CAGCTCTCAGTAGGCAAGTCCCTACATGTT-3';
3A4NF-B, 5'-CAGCTCTCAGTAGTCAAGTCCCTACATGTT-3'; 3A7NF-Bmut.,
5'-CAGCTCTCAGTAGGAAAGTCCCTACATGTT-3'; Ig,
5'-CAACAGAGGGGACTTTCCGAGGCCATCTG-3'; interleukin 2, 5'-CTAACAAAGAGGGATTTCACCTACAT-3'; H2k,
5'-CAGGGCTGGGGATTCCCCATCTCCACAGG-3'; SV40 core C,
5'-GTTAGGGTGTGGAAAGTCCCCAGGCTC-3'; acute phase response element,
5'-CCACAGTTGGGATTTCCCAACCTGACCAGA-3'; Sp1, 5'-ATTCGAT CGGGGCGGGGCGAGC-3'; C/EBP, 5'-TGCAGATTGCGCAATCTGCA-3'; 3A7-136/99, 5'-GTGTGT- GATTATTTGCCAACTGCCAACTGCCGAGGTGGAGAAGCCTC-3';
3A4-139/102, 5'-GTGTGTGATTCTTTGCCAACTTCCAAGGTGGAGAAGCCTC-3';
3A7-136/117, 5'-TGTGTGATTATTTGCCAACT-3'; 3A4-139/120,
5'-TGTGTGATTCTTTGCCAACT-3'; 3A7-119/100, 5'-ACTGCCGAGGTGGAGAAGCC-3';
3A4-122/103, 5'-ACTTCCAAGGTGGAGAAGCC-3'; 3A7-103/80,
5'-AGCCTCTTCCGACTGCAGGCAGAG-3'; 3A4-106/83,
5'-AGCCTCTTCCAACTGCAGGCAGAG-3'; 3A7-90/63,
5'-CTGCAGGCAGAGCACGGGGGCCCTGCTAC-3'; 3A4-93/65,
5'-CTGCAGGCAGAGCACAGGTGGCCCTGCTAC-3'; HNF-3,
5'-TCTGATTATTGACTTAGTCAAG-3'; NF1, 5'-TATTTTGGATTGAAGCCAATATGATA-3'; 3A7NF1mut., 5'-TGTGTCATTATTTGCGAACT-3'; 3A4NF1mut.,
5'-TGTGTCATTCTTTGCGAACT-3'; 3A7USFmut.,
5'-CTGCAGGCAGAGCATAAGGGCCCTGCTAC-3'; adenovirus major late promoter,
5'-GATCCGTAGGCCACGTGACCCGGG-3'; 3A7/4BTE, 5'-CTCCAGCCCTGCCTCCTT CT-3';
and 3A7/4BTEmut., 5'-CTCC- AGCGAGGCCTCCTTCT-3'.
| |
RESULTS |
Transcriptional Activities of the CYP3A7 and CYP3A4 Genes in Human
Hepatoma Cells--
The expression levels of CYP3A7 and CYP3A4
mRNAs in HepG2 or HuH-7 cells were determined by RT-PCR. CYP3A7 but
not CYP3A4 mRNA was detectable in both hepatoma cells (Fig.
1). Because the specific expression of
CYP3A7 mRNA in HepG2 or HuH-7 cells was in accordance with that
seen in human fetal livers (18), these cells seem to be a useful
in vitro model to analyze the function of the promoter of
the CYP3A7 gene expressed in human fetal livers.
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 1.
RT-PCR for analyzing the expression of CYP3A7
and CYP3A4 mRNAs in human hepatoma cells. PCR was performed
for 30 cycles with total RNAs (1 µg) prepared from HepG2 and HuH-7
cells. The expression of glyceraldehyde-3-phosphate dehydrogenase was
also determined as an internal control. The positions of the expected
sizes of PCR products are indicated by arrows.
|
|
To identify a possible cis-acting element(s) specifically
responsible for the transcriptional regulation of the CYP3A7
gene, HepG2 cells were transiently transfected with a series of
5'-truncated promoter-luciferase chimeric genes as shown in Fig.
2. A reporter plasmid p3A7/5.5
contained a region from nucleotides 5564 to +105 of the
CYP3A7 gene, whereas a reporter plasmid p3A4/5.5 had a
region from nucleotides 5551 to +103 of the CYP3A4 gene. Luciferase activity seen with p3A7/5.5 was 28-fold higher than that
seen with p3A4/5.5 (Fig. 2), although 5'-flanking regions up to 5.5
kb were 90% identical between the CYP3A7 and
CYP3A4 genes (data not shown). This result indicates that a
possible enhancer(s) is present specifically in the region from
5.5 kb to +105 of the CYP3A7 gene. To further clarify
the region involved in the transcription of the gene, the 5'-flanking
sequence of the CYP3A7 or CYP3A4 gene was
successively deleted. The deletion constructs were then transfected
into HepG2 cells. Deletion to 2.5 kb of the CYP3A7 gene
increased the luciferase activity by 1.8-fold as compared with the
level seen with p3A7/5.5. Thus, a negative regulatory element(s) was
assumed to be present in the region from 3.0 to 2.5 kb. Deletion to
2.0 kb decreased the luciferase activity to 18% of the level seen
with p3A7/2.5. Although the luciferase activity with p3A7/1.0 was
one-fifth of that with p3A7/2.5, the luciferase activity was still
10-fold higher than that with p3A4/1.0 or Basic Vector 2. These data indicate that positive regulatory elements locate in the regions from
2.5 to 2.0 kb and from 1.0 kb to +105 of the CYP3A7
gene.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 2.
Transcriptional activity of the 5'-flanking
region of the CYP3A7 and CYP3A4 genes
in HepG2 cells. The construction of deletion mutants is described
under "Experimental Procedures." HepG2 cells (2 × 106 cells) were transiently transfected by the calcium
phosphate co-precipitation method (37) with a reporter plasmid (8 µg). The cells were harvested 36 h after DNA transfection, and
luciferase activity was assayed. The numbers given to the deletion
mutants indicate the 5'-end of the 5'-flanking sequence of the
CYP3A7 or CYP3A4 gene counted negatively from the
transcriptional start site (22, 23). The luciferase activity represents
the average ± S.D. from at least three independent experiments.
The mean value obtained with p3A7/2.5 was defined as 100%.
|
|
Identification of the Upstream Enhancer of the CYP3A7 Gene--
A
second set of successive deletion mutants was transfected into HepG2
cells as shown in Fig. 3A.
Although deletion of the region from 2.50 to 2.35 kb of the
CYP3A7 gene decreased the luciferase activity to 80% of the
level seen with p3A7/2.5 containing the region up to 2.50 kb of the
CYP3A7 gene, further deletion to 2.30 kb lowered the
luciferase activity to 18% of the level seen with p3A7/2.5. The
result indicates that the region from 2.35 to 2.30 kb of the
CYP3A7 gene functions as an enhancer. The sequence of the
5'-flanking region from 2.35 to 2.30 kb of the CYP3A7
gene was found to contain a 1-base pair mismatch as compared with that
of the corresponding region of the CYP3A4 gene, namely
guanine base for the CYP3A7 gene and thymine base for the
CYP3A4 gene (Fig. 3B). To examine whether or not
the 1-base pair mismatch between the CYP3A7 and
CYP3A4 genes causes difference in transcriptional activity,
the nucleotide sequence was changed from a guanine base at position
2313 of the CYP3A7 gene to a thymine base and conversely
from thymine base at position 2318 of the CYP3A4 gene to
guanine base by site-directed mutagenesis (Fig. 3C). The
intact or the mutant 5'-flanking region from 2.50 to 2.25 kb of the
CYP3A7 or CYP3A4 gene was then linked to the CYP3A7 minimal promoter (62/+105) possessing BTE and TATA
box. Reporter plasmids p3A7Ewild and p3A4Ewild possessed the intact CYP3A genes, whereas reporter plasmids p3A7Emut. and
p3A4Emut. had the mutated CYP3A genes. As shown in Fig.
3C, the luciferase activity with p3A7Ewild was ~5-fold
higher than that with p3A4Ewild. Luciferase activity with p3A7Emut. was
one-fifth the level seen with p3A7Ewild. On the other hand, the
luciferase activity with p3A4Emut. was 6-fold higher than that with
p3A4Ewild. These results indicate that the 1-base pair mismatch leads
to the difference in enhancer activity between the CYP3A7
and CYP3A4 genes. Comparing the nucleotide sequence from
2.35 to 2.30 kb of the CYP3A7 gene with the elements
reported so far, we found that the sequence of the region containing
the 1-base pair mismatch partially overlapped with those of
NF-B-binding sites (Fig. 3D). The sequence of the NF-B-like elements of the CYP3A7 and CYP3A4
genes had 1- and 2-base pair changes, respectively, as compared with
the most typical NF-B-binding sequence found in the Ig gene (40)
and SV40 core C (41). To compare the enhancer activity of the
NF-B-like elements between the CYP3A7 and
CYP3A4 genes, two copies of the NF-B-like element of the
CYP3A7 or CYP3A4 gene were linked to the
CYP3A7 minimal promoter containing a region from nucleotides
62 to +105 (Fig. 3E). Luciferase activity with p3A7NF-B
was 5-fold higher than that with p3A4NF-B, although p3A4NF-B
showed some enhancer activity as compared with p3A7/62 carrying the
minimal promoter. This result was consistent with the data shown in
Fig. 3C, indicating that the NF-B-like element of the
CYP3A7 gene was required for the transcriptional activation
of the CYP3A7 gene.
View larger version (45K):
[in this window]
[in a new window]
|
Fig. 3.
Identification and characterization of an
NF-B-like enhancer in the CYP3A7
gene. A, transcriptional activity of the region
from 2.5 to 2.0 kb of the CYP3A7 gene in HepG2 cells.
The activity represents the average ± S.D. from at least three
independent experiments. The mean value obtained with p3A7/2.5 was
defined as 100%. B, nucleotide sequences from 2.5 to
2.3 kb of the CYP3A7 and CYP3A4 genes. The
arrows indicate the 5'-end of the 5'-flanking sequence of
the CYP3A7 gene in deletion constructs, p3A7/2.5,
p3A7/2.47, p3A7/2.41, p3A7/2.35, and p3A7/2.30. The 1-base pair
mismatch (G/T) between the two genes in the CYP3A7 enhancer
region is indicated by a box. C, effects of the 1-base pair
substitution between the CYP3A7 and CYP3A4 genes
on an enhancer activity in HepG2 cells. The construction of reporter
plasmids is described under "Experimental Procedures." HepG2 cells
(2 × 106 cells) were transiently transfected with a
reporter plasmid (2 µg), p3A7/62, p3A7Ewild, p3A7Emut., p3A4Ewild,
and p3A4Emut. The mutated sequence is underlined. The data
are presented as a fold induction relative to the activity with
p3A7/62. The activities represent the averages ± S.D. from at
least three independent experiments. D, alignment of the
NF-B-binding sites conserved in the CYP3A7 and other
genes. The sequences are aligned with decameric nucleotides, GGGRNNYCC,
known to be a consensus binding motif for NF-B (89, 90). The
right column indicates the constitutive factors bound to
each sequence. The mutated sequence is underlined. An
italic letter indicates a different nucleotide
sequence between the CYP3A7 and CYP3A4 genes.
E, enhancer activity of the NF-B-like element in the
CYP3A7 or CYP3A4 gene in HepG2 cells. The
construction of reporter plasmids is described under "Experimental
Procedures." HepG2 cells (2 × 106 cells) were
transiently transfected with a reporter plasmid (2 µg), p3A7/62,
p3A7NF-B, and p3A4NF-B. The results are presented as fold
induction relative to the activity with p3A7/62, and the values
represent the averages ± S.D. from at least three independent
experiments.
|
|
Sp1 and Sp3 as Possible Factors Binding to the NF-B-like Element
of the CYP3A7 Gene--
Transcription factors capable of interacting
with the NF-B-like element of the CYP3A7 or
CYP3A4 gene were investigated by a gel shift assay (Fig.
4A). Using nuclear extracts
prepared from HepG2 cells, three shifted bands (complexes A, B, and C)
appeared with the NF-B-like element of the CYP3A7 gene
(defined this probe DNA as CYP3A7 NF-B-like element),
whereas one shifted band (complex D) appeared with the NF-B-like
element of the CYP3A4 gene (defined this probe DNA as
CYP3A4 NF-B-like element). The three shifted bands
(complexes A, B, and C) were not eliminated by the addition of a
100-fold molar excess of the CYP3A4 NF-B-like element,
although the intensity of the shifted band A was weakened. On the other hand, the shifted band D was abolished by the addition of the CYP3A7 or CYP3A4 NF-B-like element. Thus, the
shifted band A, which appeared with the CYP3A7 NF-B-like
element, is likely to be partially overlapped with the shifted band D. These results indicate that the three complexes A, B, and C, but not D,
are complexes that bind specifically to the CYP3A7
NF-B-like element. To identify a constitutive factor(s) bound to the
CYP3A7 NF-B-like element, we carried out a gel shift
assay using competitors having NF-B sequences found in the Ig
(40), interleukin 2 (42), major histocompatibility complex H2K (43),
angiotensinogen (44) genes, and SV40 core C (41) (Fig. 4B).
We also employed competitors having Sp1 and C/EBP consensus sequences,
because Sp1 and C/EBP were reported to bind to a subset of
NF-B-binding sites (44, 45). The formation of the three complexes,
A, B, and C, was partially inhibited by the NF-B sequences of the
Ig, interleukin 2, H2k, angiotensinogen genes, and SV40 core C. The
addition of Sp1 consensus sequence strongly inhibited the formation of
all of the three complexes, A, B, and C. A faint band, which was not competed out by the Sp1 consensus sequence (Fig. 4B), is
likely to be complex D. To examine the binding specificity of the Sp family, a gel shift assay was performed using an Sp1 consensus site as
a probe and the NF-B-like element of the CYP3A7 or
CYP3A4 gene as a competitor (Fig. 4C). The
binding of the Sp family to the most typical Sp1 consensus sequence was
inhibited by increasing the amounts of the CYP3A7
NF-B-like element added, whereas the addition of the
CYP3A4 NF-B-like element was ineffective. Thus, it
appeared that constitutive factors specifically bound to the CYP3A7 NF-B-like element were Sp family members. To
further confirm the binding of the Sp family to the CYP3A7
NF-B-like element, a supershift assay using antibodies to Sp1 and
Sp3, which are abundant forms of the Sp family in various types of
cells (46), was performed. Antibodies to the p50 subunit of NF-B
complex were also used because the p50 homodimer but not p50/p65
heterodimer is reported to be retained in the nuclei without stimuli
(47). As shown in Fig. 4D, antibodies to Sp1 and Sp3
supershifted complex A and complexes B and C, respectively. Complex D
was not affected by antibodies to Sp1, Sp3, or the p50 subunit of
NF-B complex. These results indicate that both Sp1 and Sp3 bind to
the CYP3A7 NF-B-like element but not to the
CYP3A4 NF-B-like element.
View larger version (56K):
[in this window]
[in a new window]
|
Fig. 4.
Interaction of Sp1 and Sp3 with the
CYP3A7 NF-B-like
element. A, gel shift assay with the NF-B-like
elements in the CYP3A7 and CYP3A4 genes. Nuclear
extracts (10 µg) from HepG2 cells were incubated with the
32P-labeled NF-B-like elements of the CYP3A7
and CYP3A4 genes. Competition analysis was carried out by
the addition of a 100-fold molar excess of the unlabeled
CYP3A7 or CYP3A4 NF-B-like element.
DNA-binding complex was resolved by a 4% polyacrylamide gel. The
arrows indicate the specific complexes A, B, C, and D. N.S., nonspecific band; N.E., nuclear extracts.
B, effects of various competitors on the binding of nuclear
factors to the CYP3A7 NF-B-like element. Nuclear extracts
(10 µg) from HepG2 cells were incubated with the
32P-labeled CYP3A7 NF-B-like element in the
presence of a 100-fold molar excess of various competitors.
C, effects of the CYP3A7 and CYP3A4
NF-B-like elements on the binding of nuclear factors to an Sp1
consensus sequence. Nuclear extracts (10 µg) from HepG2 cells were
incubated with the 32P-labeled Sp1 consensus sequence in
the presence of 100-2000-fold molar excess of the unlabeled Sp1
consensus sequence, the NF-B-like element of the CYP3A7
or CYP3A4 gene. The region of DNA-binding complexes is
indicated by bracket. D, supershift assay using antibodies
to Sp1, Sp3, or NF-B p50. The 32P-labeled NF-B-like
element of the CYP3A7 or CYP3A4 gene was
incubated with nuclear extracts (10 µg) prepared from HepG2 cells in
the presence of antibodies (1 µl) to Sp1, Sp3, or NF-B p50. The
supershifted bands are indicated by the bracket.
|
|
Transactivation of the Promoter of the CYP3A7 Gene by Sp1 and Sp3
through the NF-B-like Element--
To determine the relative
contribution of each Sp isoform to the CYP3A7 enhancer
activity, Drosophila SL2 cells that lacked an endogenous Sp
family (33) were transiently transfected with a reporter plasmid,
p3A7/62, p3A7NF-B, or p3A4NF-B, together with an Sp1 or Sp3
expression plasmid (Fig. 5). The levels
of basal transcriptional activities of p3A7/62, p3A7NF-B, and
p3A4NF-B in Drosophila SL2 cells were nearly the same
(data not shown). The transcriptional activities of the p3A7/62 and
p3A4NF-B were slightly increased by Sp1 or Sp3, whereas the
transcriptional activity of p3A7NF-B was increased ~20-fold when
an Sp expression plasmid, pPacSp1 (0.1 µg) or pPacUSp3 (0.1 µg),
was co-transfected into SL2 cells in combination with p3A7NF-B.
These results indicate that both Sp1 and Sp3 act as the activators of
the CYP3A7 gene through the NF-B-like element identified
in the present study.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 5.
Activation by Sp1 and Sp3 of the
CYP3A7 gene in Drosophila SL2
cells. A reporter plasmid (2 µg), p3A7/62, p3A7NF-B, or
p3A4NF-B was co-transfected into SL2 cells (1 × 106 cells) with increasing amounts of expression plasmids
for Sp1 (closed bars) or Sp3 (open bars). The
cells were harvested 36 h after transfection, and the luciferase
activity was assayed. The results are expressed as fold induction
relative to the activity obtained by co-transfection with the
expression vector pPac as a control. The activity represents the
average ± S.D. from at least three independent experiments.
|
|
Identification of the Proximal Promoter of the CYP3A7 Gene--
A
proximal promoter region from 1.0 kb to +105 of the CYP3A7
gene also showed higher luciferase activity than did that of the
CYP3A4 gene (Fig. 2). To identify a possible regulatory
element(s) in the proximal promoter of the CYP3A7 gene, the
sequences of the CYP3A7 and CYP3A4 genes were
successively deleted as shown in Fig.
6A. The maximal luciferase
activity was seen with the p3A7/140 reporter plasmid. The luciferase
activity with the p3A7/140 was 5-fold higher than that with the
p3A4/140 harboring the corresponding region of the CYP3A4
gene. Deletion to 90 decreased the luciferase activity to 50% of the
level seen with the p3A7/140, although the luciferase activity was
still higher than that with the p3A4/90. Further deletion to 62 of
the CYP3A7 gene attenuated the luciferase activity to 10%
of the level seen with the p3A7/140. These results indicate that
the region from nucleotides 140 to 62 is also a necessary region
for the transcriptional activation of the CYP3A7 gene.
View larger version (50K):
[in this window]
[in a new window]
|
Fig. 6.
Characterization of the proximal promoter of
the CYP3A7 gene in HepG2 cells. A,
deletion analysis of a region from 1.0 kb to +105 of the
CYP3A7 gene in HepG2 cells. The construction of deletion
mutants is described under "Experimental Procedures." The
numbers given to deletion mutants indicate the 5'-end of the
5'-flanking sequence of the CYP3A7 or CYP3A4 gene
counted negatively from the transcriptional start site. The activity
represents the average ± S.D. from at least three independent
experiments. The mean value obtained with p3A7/140 was defined as
100%. B, nucleotide sequences between nucleotides 136 and
+3 of the CYP3A7 and CYP3A4 genes. The
transcription start sites of the CYP3A7 and
CYP3A4 genes (22, 23) are assigned as +1 shown by
arrows. BTE and TATA box are boxed. Putative
cis-acting elements are indicated by brackets.
Fragments used for gel shift assay, e.g. 136/99 for the
CYP3A7 gene and (139/102) for the CYP3A4
gene, are underlined. C, gel shift assay with the 136/99
and 139/102 fragments of the respective CYP3A7 and
CYP3A4 genes. Nuclear extracts (10 µg) from HepG2 cells
were incubated with the 32P-labeled 136/99 or
139/102 fragment in the presence of a 100-fold molar excess of a
competitor. DNA-binding complexes are shown by arrows.
N.S., nonspecific band; N.E., nuclear extracts.
D, gel shift assay with the 90/63 and 93/65
fragments of the respective CYP3A7 and CYP3A4
genes. Nuclear extracts (10 µg) from HepG2 cells were incubated with
the 32P-labeled 90/63 or 93/65 fragment in the
presence of a 100-fold molar excess of a competitor. DNA-binding
complexes (G, H, I, and L) are shown by arrows.
N.S., nonspecific band; N.E., nuclear
extracts.
|
|
To compare the binding pattern of transcription factors, the proximal
promoter region of the CYP3A7 gene was cut to yield five DNA
fragments: 136/99, 136/117, 119/100, 103/80, and
90/63 (Fig. 6B). The proximal promoter region of the
CYP3A4 gene was also divided into the corresponding DNA
fragments: 139/102, 139/120, 122/103, 106/83, and
93/65 (Fig. 6B). These fragments were then used as
probes or competitors for a gel shift assay. As can be seen in Fig.
6C, different DNA-binding complexes (complexes E, F, and G
for CYP3A7 and complex I for CYP3A4) appeared
with the 136/99 fragment of the CYP3A7 and the
139/102 fragment of the CYP3A4, whereas a broad and
faint complex H was detected with the fragments of both genes.
Complexes E, F, G, and H were decreased by the addition of the
136/117 fragment of the CYP3A7 gene, whereas complex I
was diminished by the 122/103 fragment of the CYP3A4
gene. As shown in Fig. 6D, a complex J appeared with both
the 90/63 fragment of the CYP3A7 and the 93/65
fragment of the CYP3A4 genes. No DNA-binding complexes
appeared with the 103/80 and 106/83 fragments (data no shown).
Searching for a possible binding site(s) of reported transcription
factor, we found that the putative binding sites for liver-enriched
transcription factors (HNF-3, C/EBP, and NF1), transcription repressor
YY1, and the E-box-like elements were located within the 136/117, 119/100, and 90/63 fragments, respectively (Fig.
6B).
HNF-3, NF1, USF1, and Sp1/Sp3 as Proteins Binding to the
Proximal Promoter Region of the CYP3A7 Gene--
To identify the
factors detected with the 136/99 fragment of the
CYP3A7 gene, a gel shift assay using the 136/117
fragment of the CYP3A7 gene and the 139/120 fragment of
the CYP3A4 gene as a probe and nuclear extracts prepared
from HepG2 cells was performed (Fig.
7A). In accordance with the
data shown in Fig. 6C, DNA-binding complexes E, F, and G
appeared with the 136/117 fragment of the CYP3A7 gene as
a probe. Complex H was detected with the same migration with the
136/117 and 139/120 fragments of both the CYP3A7 and
CYP3A4 genes. To further characterize a nuclear factor(s)
binding to the CYP3A7 proximal promoter, a gel shift assay
was performed using mutated probe, NF1mut., and various possible
competitors, HNF-3, NF1, and C/EBP (Fig. 7B). As shown in
Fig. 7C, NF1mut. was designed to have two mutations in the NF1-binding site of the CYP3A7 and CYP3A4 genes.
Three shifted bands (E, F, and G) present in the 136/117 of the
CYP3A7 gene disappeared in the presence of HNF-3. Supporting
this result, the binding of nuclear factors to HNF-3 was also
diminished by the addition of the 136/117 fragment of the
CYP3A7 gene. The result indicates that HNF-3 or an
HNF-3-related factor(s) interacts with the 136/117 fragment of the
CYP3A7 gene. Complex H seen with the 136/117 and
139/120 fragments of the CYP3A7 and CYP3A4 genes disappeared in the presence of NF1. The addition of NF1mut., HNF-3, or C/EBP was ineffective. This result was in agreement with the
fact that the putative NF1-binding sequence was conserved in the
proximal promoter region of the CYP3A7 and CYP3A4
genes (Fig. 7C). To confirm the binding of HNF-3 and NF1 to
the 136/117 fragment, a gel shift assay was performed using
antibodies to HNF-3, NF1 or C/EBP (Fig. 7D). The formation
of complex G was diminished by the addition of antibodies to HNF-3,
although complexes E and F were unaffected by the addition of
antibodies to either HNF-3 or HNF-3. To distinguish HNF-3 from
NF1, antibodies to NF1 were used for a supershift assay. The formation
of complex H decreased when antibodies to a N-terminal peptide
conserved among NF1 isoforms were added to nuclear extracts prepared
from HepG2 cells (Fig. 7D). Thus, complex H was confirmed to
be derived from NF1. Because HNF-3- and NF1-binding sites in the
promoter region of the CYP3A7 gene perfectly overlapped, it
seemed likely that HNF-3 and the members of the NF1 family bound to
the HNF-3/NF1-binding site in a mutually exclusive manner. Detailed
analysis of complex E and F is currently under investigation. To
further confirm the specific binding of HNF-3 to the
CYP3A7 gene, the 136/117 and 139/120 fragments of
the CYP3A7 and CYP3A4 genes were incubated with
in vitro translated HNF-3 protein. As shown in Fig.
7E, HNF-3 bound to the 136/117 fragment of the
CYP3A7 gene to generate a clear band. The binding of
HNF-3 was competed out by an HNF-3 binding site and was inhibited by
antibodies to HNF-3. These results indicate that HNF-3 is a
component of complex G.
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 7.
The binding of nuclear factors to
the 136/117 fragment of the CYP3A7 gene in HepG2
cells. A, gel shift assay with the 136/117 and
139/120 fragments of the respective CYP3A7 and
CYP3A4 genes. Nuclear extracts (10 µg) from HepG2 cells
were incubated with the 32P-labeled 136/117 or
139/120 fragment. Competition analysis was carried out by the
addition of a 100-fold molar excess of unlabeled 136/117 and
139/120 fragments of the CYP3A7 and CYP3A4
genes, respectively. The arrows indicate the DNA-binding
complexes. N.S., nonspecific band; N.E., nuclear
extracts. B, effects of various competitors on the binding
of nuclear factors to the 136/117 or 139/120 fragment. Nuclear
extracts (10 µg) from HepG2 cells were incubated with the
32P-labeled 136/117, 139/120, or HNF-3 in the
presence of a 100-fold molar excess of a competitor. DNA-binding
complexes are indicated by arrows. C, oligonucleotides used
for gel shift assays. An italic letter indicates
a nucleotide change between the 136/117 of the CYP3A7
gene and the 139/120 fragment of the CYP3A4 gene.
Changed nucleotides in mutant probes are underlined.
Consensus binding sequences for HNF-3, NF1 or C/EBP (from TFSEARCH,
pdap1.trc.rwcp.or.jp/research/db/TFSEARCH.html) are aligned with the
136/117 and 139/120 fragments of the respective
CYP3A7 and CYP3A4 genes. D, supershift
assay with antibodies to HNF-3, HNF-3, NF1, or C/EBP.
32P-Labeled 136/117 or 139/120 fragment was
incubated with nuclear extracts (10 µg) from HepG2 cells in the
presence or absence of antibodies (1 µl) to HNF-3, HNF-3, NF1,
or C/EBP. Supershifted bands are shown by arrows. E, gel
shift assay with in vitro synthesized HNF-3 proteins.
32P-Labeled 136/117 and 139/120 fragments of the
respective CYP3A7 and CYP3A4 genes were incubated
with HNF-3 proteins produced by in vitro transcription
and translation in the presence or absence of a competitor HNF-3 or
antibodies to HNF-3 (anti-HNF-3).
|
|
We also confirmed that complex I seen with the 122/103 fragment of
the CYP3A4 gene was identical to YY1 (data not shown). The
functional role of YY1 in the transcriptional regulation of the
CYP3A4 gene will be described elsewhere.
E-box is reportedly a target element for several transcription factors
belonging to a basic helix-loop-helix family (48). Among them, USF
appears to be the most predominant basic helix-loop-helix factor in the
liver (49). The E-box-like elements of the CYP3A7 and
CYP3A4 genes resembled the USF-binding elements of the rat -fibrinogen gene (50) and the human -globin locus control region
(51), respectively (Fig. 8A).
To examine whether the E-box-like element of the CYP3A7 gene
is indeed necessary for the formation of complex J, mutations were
introduced into the E-box-like element of the 90/63 fragment of the
CYP3A7 gene as shown in Fig. 8A. The formation of
complex J with the 90/63 fragment of the CYP3A7 gene was
not completely abolished by the addition of the mutated E-box,
mCYP3A7 (Fig. 8B). Additionally, the formation of
complex J was inhibited by the addition of a fragment having the
USF-binding sequence of the adenovirus major late promoter (52) (Fig.
8B). As expected, complex J was supershifted by antibodies
against USF1 when the E-boxes of the CYP3A7 and CYP3A4 genes were used as probes (Fig. 8B). These
results indicate that the E-boxes of the CYP3A7 and
CYP3A4 genes are recognized by USF1 in HepG2 cells.
View larger version (45K):
[in this window]
[in a new window]
|
Fig. 8.
Interaction of USF1 with the E-box of the
CYP3A7 and CYP3A4 genes in HepG2
cells. A, alignment of the binding sites for USF1. The
sequences are aligned with the 6-base pair domain CANNTG, which is a
consensus binding motif for USF1 (48). Changed nucleotides in mutant
oligonucleotide are underlined. B, gel shift assay with the
90/63 and 93/65 fragments of the CYP3A7 and
CYP3A4 genes. 32P-Labeled 90/63 or 93/65
fragment was incubated with nuclear extracts (10 µg) from HepG2 cells
in the presence or absence of a competitor or antibodies (1 µl) to
USF1. The DNA-binding complexes are shown by arrow
J. N.S., nonspecific band; N.E.,
nuclear extracts.
|
|
We also performed a gel shift assay using BTE as a probe, because the
sequence of this element is highly conserved among the P-450 genes and
is believed to contribute to the basal promoter activity of the P-450
genes (53, 54). The sequence of the BTE region of the CYP3A7
and CYP3A4 genes perfectly fitted to degenerate Sp1
consensus sequence (Fig. 9A).
As shown in Fig. 9B, three shifted bands (K, L, and M)
appeared with the BTE region of the CYP3A7 and
CYP3A4 genes. All three complexes disappeared after the
addition of Sp1 consensus sequence. A mutated BTE, which had mutations
within the core of an Sp1-binding sequence, did not affect the
formation of the complexes. The addition of antibodies to Sp1 and Sp3
supershifted complex K and complexes L and M, respectively (Fig.
9B). These results indicate that the Sp family members
interact with the BTE of the CYP3A7 and CYP3A4
genes in HepG2 cells.
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 9.
Nuclear factors binding to the BTE of the
CYP3A7 and CYP3A4 genes.
A, alignment of the BTE of the CYP3A7 or
CYP3A4 gene. The sequences of BTEs in the human
CYP3A and the rat CYP1A1 genes are aligned with a
10-base pair domain GRGGCRGGGW, which is a consensus binding motif for
Sp1 (from TFSEARCH). Changed nucleotides in the mutant BTE are
underlined. Italic letters represent nucleotide
changes in BTEs of the human CYP3A and the rat
CYP1A1 genes. B, identification of a factor(s)
binding to BTE. 32P-Labeled BTE was incubated with nuclear
extracts (10 µg) from HepG2 cells in the presence or absence of a
competitor or antibodies (1 µl) to NF-B p50, Sp1, or Sp3. The
DNA-protein complexes are shown by arrows. A bracket
indicates the supershifted band. N.S., nonspecific band;
N.E., nuclear extracts.
|
|
Functional cis-Acting Elements for the CYP3A7 Proximal
Promoter--
To examine the roles of the HNF-3-binding site,
NF1-binding site, USF1-binding site, and BTE on the promoter activity
of the CYP3A7 gene, mutations were introduced into these
sites of a p3A7/140 reporter construct by site-directed mutagenesis
(Fig. 10A). The introduction
of the mutation in the HNF-3-binding site, the USF1-binding site, or
BTE reduced the promoter activity by 40, 40, or 50%, respectively,
relative to the level seen with the p3A7/140, whereas the mutation
in the NF1-binding site increased the promoter activity. These results
indicate that the HNF-3- and USF1-binding sites and BTE are required
for the full activity of the proximal promoter. To further determine
the relative contribution of each transcription factor in the
transcriptional activation of the CYP3A7 gene,
Drosophila SL2 cells were co-transfected with expression
plasmids, pPacSp1, pPacUSp3, pAcHNF-3, or pAcUSF1 in combination
with a reporter plasmid p3A7/140 (Fig. 10B). Transfection
with Sp1, Sp3, or USF1 activated the proximal promoter of the
CYP3A7 gene by ~10-fold. Although HNF-3 alone
stimulated the promoter activity, the activity was enhanced by
combination of HNF-3 with USF1, Sp1/Sp3, or both USF1 and Sp1/Sp3.
Co-transfection with HNF-3, Sp1, and USF1 enhanced the promoter
activity by 43.6-fold. Thus, we conclude that transcription factors
Sp1/Sp3, HNF-3, and USF1 were critical factors for the maximal
activity of the CYP3A7 proximal promoter.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 10.
Functional roles of binding sites for
HNF-3, NF1, and USF1 and of BTE in the
CYP3A7 promoter activity. A, effects
of mutations in binding sites for HNF-3, NF1, and USF1 and in BTE on
the activity of the CYP3A7 proximal promoter in HepG2 cells.
The construction of mutant plasmids is described under "Experimental
Procedures." HepG2 cells were transiently transfected with a reporter
plasmid (8 µg). The values represent the averages ± S.D. from
at least three independent experiments. The mean value obtained with
p3A7/140 is defined as 100%. B, expression plasmids,
pPacSp1 (1 µg), pPacUSp3 (1 µg), pAcHNF-3 (1 µg), and/or
pAcUSF1 (1 µg) were co-transfected to SL2 cells in combination with
p3A7/140 (4 µg). The results are expressed as fold induction
relative to the activity obtained by co-transfection with the
expression vector pPac as a control. The values represent the
averages ± S.D. from at least three independent
experiments.
|
|
Synergism of Enhancer and Promoter--
It was of interest to
clarify a relationship between the NF-B-like enhancer and the
proximal promoter of the CYP3A7 gene. Thus, we fused the
distal enhancer region (2500/2250) encompassing the NF-B-like
element to promoters successively deleted as shown in Fig.
11A. Deletion from
nucleotides 2000 to 360 increased the luciferase activity by
1.5-fold. Deletion of the HNF-3-binding site, the USF1-binding site,
and BTE in the proximal promoter region did not appreciably affect the
enhancer-dependent activation of the CYP3A7
gene, when the NF-B-like enhancer was located close to TATA box
(Fig. 11A). Thus, the minimal requirement for the
NF-
TOP
ABSTRACT
INTRODUCTION
