(Received for publication, December 3, 1996, and in revised form, April 15, 1997)
From the Department of Preventive Medicine,
§ the Second Department of Surgery, Kyoto Prefectural
University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto 602, Japan, ¶ Chugai Research Institute for Molecular Medicine, Nagai,
Niihari, Ibaraki 300-41, Japan, and
the Institute of Medical
Science, University of Tokyo, Minato-ku, Tokyo 108, Japan
Butyrate is a well known colonic luminal short
chain fatty acid, which arrests cell growth and induces differentiation
in various cell types. We examined the effect of butyrate on the expression of WAF1/Cip1, a potent inhibitor of
cyclin-dependent kinases, and its relation to growth arrest
in a p53-mutated human colon cancer cell line WiDr. Five millimolar
butyrate completely inhibited the growth of WiDr and caused
G1-phase arrest. WAF1/Cip1 mRNA was rapidly
induced within 3 h by treatment with 5.0 mM butyrate, and drastic WAF1/Cip1 protein induction was detected. Using several mutant WAF1/Cip1 promoter fragments, we found that the
butyrate-responsive elements are two Sp1 sites at 82 and
69
relative to the transcription start site. We also found that a TATA
element at
46 and two overlapping consensus Sp1 sites at
60 and
55 are essential for the basal promoter activity of
WAF1/Cip1. These findings suggest that butyrate arrests the
growth of WiDr by activating the WAF1/Cip1 promoter through
specific Sp1 sites in a p53-independent fashion.
Butyrate is one of the most abundant short chain fatty acids in the large intestine, generated by bacterial fermentation of dietary fibers (1). Butyrate shows potent effects on growth arrest and differentiation in vitro in various malignant tumor cell lines, such as breast cancer cells, hepatoma cells, and others (2-5). In colorectal cancer cells, butyrate inhibits cell growth and induces differentiation marker proteins such as alkaline phosphatase and carcinoembryonic antigen (6-9). Furthermore, butyrate arrests the cell cycle progression at the G1 phase (9) and decreases c-myc oncogene expression in human colon cancer cell lines (9, 10). However, the precise mechanism of growth suppression by butyrate in colon cancer cells has not been clarified.
WAF1/Cip1 protein potently inhibits the various G1 cyclin-dependent kinases activities (11-13) by suppressing the phosphorylation of retinoblastoma (RB) protein, thereby supposedly inhibiting the G1-S phase transition (11, 14). Besides its role as a kinase inhibitor, it has been reported recently that WAF1/Cip1 at low doses assembles kinase complexes and promotes a kinase activity (15). Furthermore, the transcription of the WAF1/Cip1 gene is directly activated by wild-type p53 protein (16). Thus, WAF1/Cip1 could play a key role as a downstream mediator of the p53-induced cell growth arrest.
Several studies have already shown the p53-independent induction of
WAF1/Cip1 by serum, transforming growth factor , and other differentiation-inducers (17-20). In addition, butyrate has been
reported to induce WAF1/Cip1 mRNA independently of p53
during differentiation of hematopoietic cells, hepatoma cells, and
colon cancer cells in vitro (18, 21). Butyrate can also
dephosphorylate the retinoblastoma protein in mouse fibroblasts (22).
To investigate the mechanism of butyrate-induced growth arrest, we used
a human colon cancer cell line WiDr harboring a point mutation in p53 at codon 273 (23) and examined the effect of butyrate on the expression
of the WAF1/Cip1 gene.
Our results demonstrate that WAF1/Cip1 mRNA is rapidly
induced upon butyrate treatment, although WiDr lacks the wild-type p53 gene. We then found that butyrate markedly induces
WAF1/Cip1 protein and causes G1-phase arrest. In addition,
we observed that butyrate can strongly activate the
WAF1/Cip1 promoter, and that the two p53-binding sites are
not required for the transcriptional activation by butyrate. Using a
series of mutant WAF1/Cip1 promoter constructs, we also
found in p53-negative cell lines WiDr and human osteosarcoma cell line
MG63 (24), that two Sp1 sites at 82 and
69 relative to the
transcription start site are involved in the activation of the
WAF1/Cip1 promoter by butyrate. Furthermore, the essential
elements for the WAF1/Cip1 promoter activity have been shown
to be two overlapping consensus Sp1 sites at
60 and
55 and TATA
sequence at
46.
Human colon adenocarcinoma cell line WiDr was a kind gift from Dr. R. Takahashi of Kyoto University and human osteosarcoma cell line MG63 was kindly provided by Dr. Y. Yanase of Wakayama Medical College. WiDr harbors a point mutation at codon 273 of p53 (23), and MG63 contains rearrangements in the p53 gene (24). Both cell lines were maintained in DMEM supplemented with 10% fetal calf serum and were incubated at 37 °C in a humidified atmosphere of 5% CO2 in air. For the cell growth study, WiDr cells were inoculated at a density of 5 × 104 cells in 35-mm-diameter dishes. A solution of 500 mM butyrate was prepared by adjusting the concentration of n-butyrate (Sigma Chemical Co.) in DMEM and was diluted to its final concentration in each culture dish. Two days after the inoculation, butyrate was added at various concentrations. From the second day to the sixth day from the inoculation, the number of viable cells was counted by a trypan blue dye exclusion test. This cell growth study was carried out in duplicate and repeated at least three times.
Plasmid PreparationThe human wild-type
WAF1/Cip1 promoter-luciferase fusion plasmid, WWP-Luc, was a
kind gift from Dr. B. Vogelstein (16). The 2.4-kilobase pair genomic
fragment containing the transcription start site was subcloned into the
HindIII site of the luciferase reporter vector, pGL3-Basic
(Promega), to generate pWWP. To generate pWPdel-p53, pWWP was treated
with an exonuclease III-based system, Kilo-Sequence deletion kit
(Takara), to remove about 1100 bases from the 5 end of the
WAF1/Cip1 promoter, and was religated. pWPdel-PstI was
generated by digesting pWWP with PstI and BglII and by religating (Fig. 6). pWPdel-SmaI was also generated by digesting
pWWP with SmaI and by religating (Fig. 6). To construct pWP101 containing four Sp1 sites termed Sp1-3, Sp1-4, and Sp1-5-6 (Fig.
8), complementary oligonucleotides corresponding to the sequence
between
101 and
61 of the WAF1/Cip1 promoter relative to
the transcription start site (16) were synthesized, annealed, and
cloned into the SacI and SmaI sites of
pWPdel-SmaI. Sequences of the oligonucleotides for pWP101 were
sense, 5
-CGGTGGGCCGAGCGCGGGTCCCGCCTCCTTGAGGCGGGCCC-3
, and
antisense, 5
-GGGCCCGCCTCAAGGAGGCGCGACCCGCGCTCGGCCCACCGAGCT-3
. Two
mutants of pWP101, pWP101-mtSp1-3 and pWP101-mtSp1-4 were also
constructed using synthesized oligonucleotides as follows (Fig. 8):
pWP101-mt Sp1-3: sense,
5
-CGGTGGGCCGAGCGCGGGTCAAAACTCCTTGAGGCGGGCCC-3
and
antisense,
5
-GGGCCCGCCTCAAGGAGTTTTGACCCGCGCTCGGCCCACCGAGCT-3
; pWP101-mt Sp1-4: sense,
5
-CGGTGGGCCGAGCGCGGGTCCCGCCTCCTTGAGTTTTGCCC-3
and
antisense,
5
-GGGCAAAACTCAAGGAGGCGGGACCCGCGCTCGGCCCACCGAGCT-3
. To
generate pWP101-mtSp1-5-6 and pWP101-mtTATA, oligonucleotides were
synthesized as follows and subcloned into the SmaI and
HindIII sites of pWP101 respectively (Fig. 8):
pWP101-mtSp1-5-6: sense,
-GGTTTTGGCGGTTGTATATCAGGGCCGCGCAGAGCTGCGCCAGCTGAGGTGTGAGCAGCTGCCGAAGTCAGA-3
, and antisense,
5
-AGCTTCTGACTTCGGCAGCTGCTCACACCTCAGCTGGCGCAGCTCTGCGCGGCCCTGATATACAACCGCCAAAACC-3
; and pWP101-mtTATA: sense,
5
-GGGCGGGGCGGTTGTGCGTCAGGGCCGCGCAGAGCTGCGCCAGCTGAGGTGTGAGCAGCTGCCGAAGTCAGA-3
and antisense,
5
-AGCTTCTGACTTCGGCAGCTGCTCACACCTCAGCTGGCGCAGCTCTGCGCGGCCCTGACGCACAACCGCCCCGCCC-3
. These sequences are identical to that of pWP101 except for the sequence
underlined (Fig. 8). To construct pWP124, complementary oligonucleotides corresponding to the sequence between
124 and
61 of the WAF1/Cip1 promoter, containing all six Sp1
consensus binding sites (16, 25, 26), were synthesized and cloned into
the KpnI and SmaI sites of pWPdel-SmaI
(Fig. 7). The extent of 5
deletions was determined by sequencing or
utilizing the restriction enzyme sites.
The luciferase-reporter plasmid, Sp1-luc, which contains the sequence
of
5-CGCGTGGGCGGAACTGGGCGGAGTTAGGGGCGGGA-3
,
consisting of three consensus Sp1 binding sites underlined from the
SV40 promoter, was a kind gift from Dr. Peggy J. Farnham (27). A vacant
vector pGL2-Basic was purchased from Promega and used for control
reporter plasmid.
Cells were removed from culture dishes by trypsinization and collected by centrifugation. After washing with Ca2+- and Mg2+-free PBS, cells were suspended in PBS containing 0.1% Triton X-100 to prepare nuclei. After the suspension was filtered through 50 µM nylon mesh, 0.1% RNase and 50 µg/ml propidium iodide were added. DNA contents in stained nuclei were analyzed with FACScan (Becton Dickinson). The suspension of cells was analyzed for each DNA histogram. The number of stained nuclei in each phase was measured according to the S-fit program in the FACScan.
RNA Isolation and Northern Blot AnalysisTotal RNA was isolated from cells grown in a 100-mm diameter dish to 80% confluency using an Isogen RNA isolation kit (Nippon gene), and 20 µg of total RNA per lane were examined by Northern blot analysis. The WAF1/Cip1 cDNA for a probe was obtained from a pCEP-WAF1 plasmid (a kind gift from Dr. B. Vogelstein) by digesting with NotI. Northern blot analysis was performed by following standard methods (28). The mRNA level was determined using a bio-imaging analyzer BAS 2000 (Fujix).
Protein Isolation and Western Blot AnalysisProtein extraction was performed from cells grown to 80-90% confluency. Cells were washed twice with cold PBS, and 300 µl of SDS sample buffer (50 mM Tris-HCl, pH 6.8, 10% glycerol, 1% SDS, 5% 2-mercaptoethanol, and 0.025% bromphenol blue) were directly added to the cells. The lysed cells were collected and kept for analysis. The protein extract was boiled for 3 min and loaded onto 12% (for WAF1/Cip1 detection) or 8% (for RB detection) polyacrylamide gel, electrophoresed, and transferred to nitrocellulose membrane. A monoclonal antibody to WAF1/Cip1 (PM-15091A from Pharmingen) or RB (PM-14001A from Pharmingen) was used as the primary antibody. The signal was then developed with the enhanced chemiluminescence system (ECL; Amersham Corp.).
Stable TransfectionThe human wild-type WAF1/Cip1 promoter-luciferase fusion plasmid, WWP-Luc, and a plasmid conferring neomycin resistance, pSV2neo, were cotransfected into WiDr cells by the Chen-Okayama method (29). Forty-eight h after transfection, Geneticin (G418 from Life Technologies, Inc.) was added to the medium (final concentration, 400 µg/ml), and the G418-resistant colonies were selected. The medium was changed every 2 days. About 3 weeks later, the G418-resistant cells were isolated as single colonies and separately expanded. For luciferase assay, cells were inoculated at a density of 5 × 105 cells in 100-mm-diameter dishes. The following day, butyrate was added, and cell lysates were collected at each indicated time.
Transient TransfectionWiDr cells were transfected by calcium phosphate coprecipitation technique. WiDr cells were inoculated at a density of 5 × 105 cells in 100-mm-diameter dishes. After 2 days, 8 µg of reporter plasmid DNA in calcium phosphate precipitates mixture were used for transfection for 8 h. Twenty-four h after the transfection, butyrate was added, and 48 h after the transfection, cell lysates were collected for a luciferase assay. MG63 cells were transfected by a lipofection technique. MG63 cells were seeded at a density of 1.5 × 105 cells in 35-mm-diameter dishes. The next day, cells were transfected with 2 µg/dish of reporter plasmid DNA in Lipofectamine (Life Technologies, Inc.) for 5 h. Twenty-four h after the transfection, the medium was changed, and 10.0 mM butyrate was added, and 48 h after the transfection, cell lysates were collected for the luciferase assay.
Luciferase AssayThe luciferase activities of the cell lysates were measured as described previously (30). Luciferase activities were normalized for the amount of the protein in cell lysates. All the luciferase assays were carried out at least in triplicate. Each experiment was repeated at least three times.
Electrophoretic Mobility Shift Assay (EMSA)1Nuclear extracts were prepared according to the method of Dignam et al. (31). In brief, MG63 cells were scraped and incubated in 10 mM Hepes-KOH buffer, pH 7.9, containing 10 mM KCl, 1.5 mM MgCl2, 0.5 mM dithiothreitol, 5 mM sodium fluoride, 5 mM sodium orthovanadate, and 0.5 mM phenylmethylsulfonyl fluoride on ice for 10 min. Cells were disrupted by Dounce homogenizer. After centrifugation, nuclei were resuspended in 20 mM Hepes-KOH buffer, pH 7.9, containing 400 mM NaCl, 1.5 mM MgCl2, 25% glycerol, 0.1 mM EDTA, 1 mM dithiothreitol, 5 mM sodium fluoride, 5 mM sodium orthovanadate, and 0.5 mM phenylmethylsulfonyl fluoride and incubated at 4 °C for 60 min. The mixture was centrifuged at 35,000 rpm for 30 min at 4 °C, and the supernatant was recovered as nuclear extracts. Nuclear extracts were dialyzed against 20 mM Hepes-KOH buffer, pH 7.9, containing 400 mM KCl, 20% glycerol, 0.1 mM EDTA, 1 mM dithiothreitol, 5 mM sodium fluoride, 5 mM sodium orthovanadate, and 0.5 mM phenylmethylsulfonyl fluoride.
Annealed oligonucleotides containing the sequence between 87 and
72
of the WAF1/Cip1 promoter (AGCTCGGGTCCCGCCTCCTT and TCGAAAGGAGGCGGGACCCG) were labeled with [
-33P]dCTP
using the Klenow fragment of Escherichia coli DNA polymerase I and were used as a probe, termed wt
87/
72. The reaction mixture for the EMSA contained 8 mM Tris-HCl, pH 7.9, 24 mM Hepes-KCl, pH 7.9, 120 mM KCl, 24% (v/v)
glycerol, 2 mM EDTA, 2 mM dithiothreitol, 1 mg
of poly(dI·dC) (Pharmacia Biotech, Inc.) and 8 mg of nuclear extract.
After preincubation for 5 min, 33P-labeled probe DNA was
added to the mixture, and the binding reaction was allowed to proceed
at room temperature for 20 min. The reaction mixture was further
incubated for 20 min in the presence or absence of anti-Sp1 or Sp3
antibody (Santa Cruz Biotechnology). The product was then resolved by
electrophoresis on a 6% polyacrylamide gel.
We first examined the effect of butyrate on the
proliferation of WiDr cells. Fig. 1 shows
the growth of WiDr cells in the presence of various concentrations of
butyrate. Butyrate showed little effect on the viability of the cells
as assessed by trypan blue dye exclusion test up to 5.0 mM.
However, 10.0 mM butyrate was slightly cytocidal. A
dose-dependent inhibition of the cell growth was observed
at concentrations of 0.625 mM or more. On day 6, the growth
of cells was inhibited to 71, 27, 16, 7, or 2% of the control level by
0.625, 1.25, 2.5, 5.0, or 10.0 mM butyrate, respectively
(Fig. 1).
Butyrate Increases WAF1/Cip1 mRNA and Protein Levels in WiDr Cells
To investigate whether WAF1/Cip1 is involved in
the butyrate-induced growth arrest in WiDr cells, butyrate-treated or
untreated WiDr cells were assayed for WAF1/Cip1 mRNA
expression by Northern blot analysis. In the untreated control cells,
WAF1/Cip1 expression was too weak to be detected, probably
due to the lack of wild-type p53 gene in the cells (Fig.
2A, lane 1). However, 24-h
exposure to butyrate caused distinct WAF1/Cip1 induction in
a dose-dependent fashion from 0.625 mM up to
5.0 mM (Fig. 2A, lanes 2-5). It is consistent
with the result that butyrate has inhibited the growth of WiDr
dose-dependently at 0.625 mM or more. As p53 is
mutated in WiDr cells (23), it is most likely that the induction of WAF1/Cip1 mRNA is mediated through a p53-independent
pathway. The time course study showed that WAF1/Cip1
mRNA was induced 3 h after the treatment with 5.0 mM butyrate and reached its peak at around 12 h after
the treatment, when mRNA levels were induced approximately 13-fold
compared with the control. The induction remained at least for 24 h (Fig. 2B). On the other hand, butyrate could not activate
other major cyclin-dependent kinase inhibitors such as
p27kip1 (32, 33) and p16INK4A (34) at a mRNA level
(data not shown). Thus, we conclude that butyrate specifically induces
WAF1/Cip1 at a mRNA level.
Next, we tried to elucidate whether the WAF1/Cip1 protein would also be
induced by treatment with butyrate in WiDr cells. Western blot analysis
showed that WAF1/Cip1 protein expression was hardly detected in the
untreated control cells (Fig. 3A,
lane 1), as expected from the result of Northern blot analysis. In contrast, treatment with 5.0 mM butyrate for 24 h
induced expression of WAF1/Cip1 protein (Fig. 3A, lane
2).
Taken together, these results indicate that butyrate specifically induces WAF1/Cip1 mRNA and consequently increases WAF1/Cip1 protein levels in WiDr cells through a p53-independent pathway. To confirm these results, another p53-negative osteosarcoma cell line MG63 (24) was used. Similarly in MG63 cells, butyrate effectively suppressed the cell growth (data not shown) and specifically induced WAF1/Cip1 mRNA in the absence or presence of cycloheximide (CHX) (Fig. 2C) but not that of p27kip1 and p16INK4A (data not shown). Furthermore, 24 h after the treatment with butyrate, the WAF1/Cip1 protein level was drastically increased (Fig. 3B, upper panel), and subsequently, the majority of the RB protein was converted into a hypophosphorylated form (Fig. 3B, lower panel).
Butyrate Arrests WiDr Cells at the G1 Phase in Cell Cycle ProgressionTo investigate the effect of butyrate on cell
cycle progression of WiDr cells, the DNA content of nuclei of WiDr
cells was measured by flow cytometric analysis. As shown in Fig.
4A, fluorescence analysis
revealed that a 24 h exposure to 2.5 or 5.0 mM
butyrate apparently decreased the population of S-phase cells in a
dose-dependent manner. S-fit analysis of the DNA histograms
also revealed that 5.0 mM butyrate caused the accumulation
of cells in the G1 phase from 42 to 71% (Fig.
4B). A time course study showed that cells started to
accumulate in G1 phase at least 16 h after the
addition of 5.0 mM butyrate, and the effect reached its
maximum at 24 h after the treatment (data not shown). Furthermore
in MG63 cells, 24 h treatment of butyrate also caused
G1 phase arrest with weak G2-M arrest (data not
shown).
Butyrate Stimulates WAF1/Cip1 Promoter Activity
Having
demonstrated that WAF1/Cip1 mRNA expression is
drastically induced by butyrate in WiDr cells lacking wild-type p53, we
subsequently investigated whether butyrate can stimulate activity of
the promoter of the WAF1/Cip1 gene. The effect of butyrate on the wild-type WAF1/Cip1 promoter-luciferase fusion
plasmid, WWP-Luc, was examined by transient transfection. Following a
24-h exposure to 5.0 mM butyrate, the luciferase activity
from the WWP-Luc plasmid was increased 55-fold compared with the
untreated control (Fig. 5A).
For further examination, we prepared a WiDr cell line stably
transfected with the WWP-Luc plasmid. As shown in Fig. 5B,
the luciferase activity was increased in a dose-dependent manner up to 8-fold by the treatment with 5.0 mM butyrate.
Time course study indicated that treatment with 5.0 mM
butyrate for 1 or 3 h caused a slight increase of the
WAF1/Cip1 promoter activity compared with the controls.
Six-h treatment with 5.0 mM butyrate significantly
increased the WAF1/Cip1 promoter activity, and 24-h treatment strongly stimulated the promoter activity up to 9.2-fold (Fig. 5C).
Analysis of the Butyrate-responsive Elements in the WAF1/Cip1 Promoter
Next, we tried to determine what regions of the
WAF1/Cip1 promoter are responsive to butyrate activation.
For this purpose, a series of 5 deletion constructs of the
WAF1/Cip1 promoter were generated (see "Materials and
Methods"). The resulting plasmids were transiently transfected into
WiDr cells, and luciferase activities following butyrate treatment were
measured relative to full-size WAF1/Cip1 promoter (pWWP).
The pWPdel-p53 plasmid, including a 1219-bp promoter fragment lacking
two p53 binding sites, was consistently fully activated by butyrate up
to 20-25-fold, a level comparable to that of the full-length
WAF1/Cip1 promoter (pWWP) in WiDr cells (data not shown),
suggesting that the two p53 binding sites are not required for the
transcriptional activation by butyrate. Furthermore, about 20-fold
activation by butyrate was still observed in the 210-bp promoter
fragment from pWPdel-PstI (data not shown; the location of the PstI
site is shown in Fig. 6). These results
suggest that butyrate-responsive elements exist within the 210-bp
region relative to the start site of transcription (Fig. 6).
This 210-bp region harbors four independent and two overlapping nearly
consensus binding sites for transcription factor Sp1 (16, 25, 26). We
termed them Sp1-1, Sp1-2, Sp1-3, Sp1-4, and Sp1-5-6 from the upstream
(Fig. 6). To determine whether these Sp1 binding sites are involved in
activation by butyrate, a series of 5 deletion plasmids, pWP124,
containing all of the six Sp1 binding sites, pWP101, lacking Sp1-1 and
Sp1-2, and pWPdel-SmaI, lacking Sp1-1 to Sp1-4 sites, were constructed
and assayed for luciferase activity in the absence or presence of
butyrate (Figs. 6 and 7). As shown in Fig.
7, luciferase activity of pWP101 as well
as pWP124 was activated about 26-31-fold by 5.0 mM
butyrate, a level similar to that of the activation by the full-size
promoter (pWWP). On the other hand, in pWPdel-SmaI the activation by
5.0 mM butyrate was only 3.9-fold. Furthermore, the basal
promoter activity of pWPdel-SmaI significantly decreased to 13.4% of
pWWP, whereas that of pWP124 or pWP101 did not significantly decrease. We then generated a series of mutants of pWP101 having mutations in the
various Sp1 sites or TATA element, and we termed them pWP101-mtSp1-3, pWP101-mtSp1-4, pWP101-mtSp1-5-6, and pWP101-mtTATA, respectively (Fig.
8). These constructs were transiently
transfected into WiDr cells, and their luciferase activities were
assayed in the absence or presence of 5.0 mM butyrate. As
shown in Fig. 8A, the basal activity of pWP101-mtSp1-3 was
reduced to 2.5% of that of pWP101, and the activation by butyrate in
pWP101-mtSp1-3 decreased to only 2.1-fold from 33.3-fold activation in
wild-type pWP101. Similarly, the basal activity of pWP101-mtSp1-4 was
reduced to 7.1% of pWP101, and the activation by butyrate decreased to
8.1-fold (Fig. 8A). On the other hand, the basal activity of
pWP101-mtSp1-5-6 or pWP101-mtTATA was reduced to background levels,
about 0.4% of pWP101, and the activation by butyrate was entirely
abolished in these constructs (Fig. 8A). As shown in Fig.
8B, in MG63, similar to the results in WiDr, the activation
by butyrate in pWP101-mtSp1-3 was 3.6-fold, and that in pWP101-mtSp1-4
was 15.1-fold, whereas 70.0-fold activation was detected in pWP101.
Taken together, we conclude that the Sp1-3 site located between
82
and
77 relative to the transcription start site is the main
butyrate-responsive element, and that the Sp1-4 site between
69 and
64 is also partially involved in the activation (Fig. 8). We also
show that the Sp1-5-6 site and TATA element are the most important core
promoter elements indispensable for the basal promoter activity of the
WAF1/Cip1 promoter (Fig. 8).
Using WAF1/Cip1 mutant constructs, we found that butyrate
activates the WAF1/Cip1 promoter through the effect at the
Sp1 sites. To confirm that Sp1 elements are indeed activated by
butyrate, the reporter plasmid Sp1-luc, containing SV40
promoter-derived three consensus Sp1 binding sites but no TATA box (see
"Materials and Methods"), was transfected into WiDr cells, and
activation of the promoter by butyrate was analyzed. As shown in Fig.
9, butyrate significantly activated the
Sp1-luc plasmid about 37-fold, whereas vacant vector pGL2-Basic,
lacking Sp1 sites, was not activated (Fig. 9).
Identification of Proteins Interacting with the Main Butyrate Responsive Element
To determine if Sp1 or other proteins can
interact with the main butyrate-responsive element, EMSAs were
performed using the oligonucleotide containing the wild-type Sp1-3
site, between 87 and
72 from the transcription start site. Nuclear
extracts were purified from either butyrate-treated or untreated MG63
cells. As shown in Fig. 10, two major
DNA-protein complexes were detected, which were competed away by an
excess of unlabeled homologous oligonucleotide (data not shown). To
elucidate whether the retarded bands represent the binding of Sp1 or
Sp3 (a member of Sp1 family), EMSA was performed with the nuclear
extracts preincubated with Sp1- or Sp3-antibody. In the presence of
Sp1-antibody, the upper complex was supershifted (lanes 2 and 5), and the lower complex was diminished in the presence
of Sp3-antibody (lanes 3 and 6). However, both
the mobility pattern and intensity were not changed by butyrate
treatment.
Mounting evidence indicates that mutations in p53 are among the most common genetic events in the development of human cancer (35, 36). On the other hand, WAF1/Cip1 is well known to be induced by wild-type p53 (16). Hence, it might be plausible that little or no expression of WAF1/Cip1 is also a common event in cancer cells. Therefore, it would be of great value to identify the p53-independent pathway of WAF1/Cip1 induction, which could lead to an alternative pathway to suppress the oncogenic progression.
In the present study, we have shown that treatment of either WiDr or
MG63 cells with butyrate specifically induces WAF1/Cip1 mRNA and protein, resulting in G1 arrest of the cell
cycle progression in a p53-independent manner. A series of mutation
analyses of the WAF1/Cip1 promoters have revealed that the
main butyrate-responsive element is the Sp1 site between 82 and
77
relative to the transcription start site (the Sp1-3 site in this
report), and the Sp1 site between
69 and
64 (Sp1-4 site) is also
partially involved in this activation. In addition, we found that
butyrate is capable of activating transcription from the luciferase
reporter plasmid including only three Sp1 sites. These results strongly
suggest that Sp1 is involved in the transcriptional activation of the
WAF1/Cip1 promoter in response to butyrate; in fact, EMSA
using MG63 cells revealed that Sp1 and Sp3 can specifically interact
with this main butyrate-responsive element, the Sp1-3 site. However,
the intensity and mobility pattern of the retarded bands were not
changed by butyrate, which means that activation of the
WAF1/Cip1 promoter by butyrate does not appear to be due to
increasing the binding of Sp1 or Sp3. Additionally, butyrate could not
affect the phosphorylation pattern of the Sp1 protein (data not shown),
and CHX did not block the WAF1/Cip1 mRNA induction by
butyrate in MG63 cells (Fig. 2C, lanes 5 and 6).
Hence, there is a possibility that Sp1-related or other unknown factors
pre-exist and will be subject to modulation, such as phosphorylation, and involved in the activation of WAF1/Cip1 promoter in
response to butyrate. On the other hand, very little is presently known of how the Sp1 modification affects transcription except for
phosphorylation or glycosylation (37). Thus, further studies will be
required to elucidate the mechanism of how butyrate modulates the
potent transcriptional function of Sp1.
Recently, several studies have reported the p53-independent induction
of WAF1/Cip1 (17-20). The promoter analysis of the
p53-independent pathways has also been reported (25, 26, 38, 39). Biggs et al. (26) have reported that the region between 122 and
61 from the transcription start site, including Sp1-1 to Sp1-4 sites in our paper, is required for both the basal activity and the full
activation of the WAF1/Cip1 promoter by phorbol esters and okadaic acid and suggested that Sp1 is involved in this activity by
using gel mobility shift assays (26). Furthermore, Datto et
al. (25) have identified the main transforming growth factor
-responsive element of the WAF1/Cip1 promoter, termed
T
RE, as an element including Sp1 site between
82 and
77 (Sp1-3
site), by using a series of deleted or mutated constructs. It is of
interest that T
RE in the WAF1/Cip1 promoter corresponds
to the main butyrate-responsive element including the Sp1 site between
82 and
77 (Sp1-3 site). Sp1 protein is a ubiquitously expressed
transcription factor that regulates a large number of constitutive and
induced mammalian genes by interacting with specific GC-rich elements
(GC boxes) (40, 41). It would thus be of great interest to clarify the mechanism by which butyrate and other WAF1/Cip1-inducing
factors such as TGF-
act on the Sp1 transcription factor.
In addition, we clearly showed that two overlapping Sp1 sites between
60 and
51 (Sp1-5-6 site) and TATA box are the most essential for
the WAF1/Cip1 promoter activity. This discrepancy with the
results of Biggs et al. (26) or Datto et al. (25) may be explained by the different cell lines or by small differences in
the sequences of generated plasmids (25, 26).
In summary, our results suggest that butyrate-induced growth arrest in WiDr cells is due to the p53-independent activation of WAF1/Cip1 promoter mediated through specific Sp1 sites in the promoter region. Recently, we proposed a novel approach for chemotherapy or chemoprevention against cancer, which we termed "gene-regulating chemotherapy or chemoprevention" (42). Our strategy is to activate the potent function of growth-inhibitory genes, which are activating targets of p53. The WAF1/Cip1 gene is one of the good candidates, because WAF1/Cip1 appears to be rarely mutated in human common tumors (43, 44), whereas the p53 gene is frequently mutated (35, 36). Therefore, in the future, clarification of the p53-independent activating pathway of the WAF1/Cip1 gene might contribute to the therapy or the prevention of cancer when p53 is mutated.
We thank Dr. K. Kawai for his continuous encouragement. We also thank Drs. R. Takahashi and E. Hara for useful advice on the Western blot technique.