(Received for publication, January 9, 1997, and in revised form, May 21, 1997)
From the Georgetown University Medical Center, Department of Pharmacology and the Lombardi Cancer Center, Washington, D.C. 20007
Transcription factor Sp1 is a phosphoprotein whose level and DNA binding activity are markedly increased in doxorubicin-resistant HL-60 (HL-60/AR) leukemia cells. The trans-activating and DNA binding properties of Sp1 in HL-60/AR cells are stimulated by cAMP-dependent protein kinase (PKA) and PKA agonists and inhibited by PKA antagonists as well as by the PKA regulatory subunit. Reporter gene activity under the control of the Sp1-dependent SV40 promoter is stimulated in insect cells transiently expressing Sp1 and PKA, and the DNA binding activity of recombinant Sp1 is activated by exogenous PKA in vitro. These results indicate that Sp1 is a cAMP-responsive transcription factor and that Sp1-dependent genes may be modulated through a cAMP-dependent signaling pathway.
Sp1 is a ubiquitous transcription factor that binds to the consensus sequence (G/A)(G/A)GGCG(G/T)(G/A)(G/A)(G/T) or GC box (1). It was initially identified as a HeLa cell-derived factor that activated six tandem Sp1 sites in the SV40 early promoter (2-4). Sp1 elements of varying affinity have been characterized in the HIV-11 (5), herpes simplex virus thymidine kinase (6), metallothionein IIA (7), and MDR1 (8) promoter regions, among others.
Sp1 is glycosylated (9) and is phosphorylated at its N terminus by
DNA-dependent protein kinase, a nuclear Ser/Thr kinase that
is stimulated by 3-termini in DNA (10). Nevertheless, phosphorylation
of Sp1 by DNA-dependent protein kinase does not affect
either its trans-activating or DNA binding activities (10). However, dephosphorylation of Sp1 has been suggested to enhance Sp1 DNA
binding activity (11-13), and thus, the roles of Sp1 phosphorylation and the target protein kinase involved in this process still remain unclear.
Type I cAMP-dependent protein kinase (PKA) is a tetrameric holoenzyme consisting of two regulatory cAMP-binding (R-I) subunits and two catalytic (C) subunits that dissociate upon binding cAMP (14). Dissociation of PKA results in its activation and the translocation of the catalytic subunit to the nucleus (15). Nuclear localization of PKA is essential to mediate induction of cAMP-regulated genes (16, 17) via phosphorylation of the cAMP response element (CRE)-binding protein, CREB (18, 19). Recently, another CRE, termed CRS, has been characterized in the promoter region of members of the CYP (20-22) and ferredoxin (23) genes. The CRS binds a transcription factor similar to Sp1 in size and sequence specificity, and the binding of this factor to the CRS is inhibited by an Sp1 consensus double-stranded oligonucleotide (24). The CRS confers high basal levels of transcription in adrenocortical tumor cells, and gene expression through the CRS is stimulated by forskolin, an activator of adenyl cyclase (24).
HL-60/AR leukemia cells (25) exhibit a multidrug-resistant (MDR) phenotype with constitutively high Sp1 and CREB DNA binding activities (11, 26). Reversion of drug resistance by the type I PKA antagonist 8-Cl-cAMP results in the down-regulation of CREB DNA binding activity but not in the levels of CREB and other cAMP-regulated transcription factors (26), suggesting that the presence of PKA-dependent transcription factors in these cells may be a prerequisite for maintenance of the MDR phenotype.
In the present study, we investigated whether the cAMP-dependent signaling pathway could modulate the DNA binding and trans-activating properties of Sp1. Our results indicate for the first time that Sp1 is activated by PKA and that a clearly defined cAMP signaling pathway may be responsible for up-regulating the DNA binding and trans-activating properties of this transcription factor.
HL-60 and HL-60/AR cells (25) were obtained from the American Type Culture Collection and Dr. James E. Gervasoni, Jr. (Columbia University), respectively. HL-60 cells were grown in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum (Biofluids, Inc.), 40 mM Hepes (pH 7.4), and 50 µg/ml gentamicin. HL-60/AR cells were maintained in the same medium with 1 µM doxorubicin but were diluted 10-fold with doxorubicin-free medium prior to use.
8-Cl-cAMP and doxorubicin were obtained from the Natural Products Branch, Developmental Therapeutics Program, National Cancer Institute. RpcAMP[S], SpcAMP[S], and Rp-8-Cl-cAMP[S] were obtained from Biolog Life Science (Bremen, Germany).
Mobility Shift AssayNuclear extracts were prepared from
5 × 106 cells as described (27). Nuclear extracts (7 µg of protein) were incubated for 30 min at room temperature with 2 fmol (680 Ci/mmol) of a double-stranded Sp1 consensus
oligodeoxynucleotide (Stratagene) that was end-labeled with
[-32P]ATP and T4 polynucleotide kinase. Mobility shift
assays were performed as described previously (11). Incubation was
carried out in 20 µl of binding buffer containing: 10 mM
Tris-HCl (pH 7.5), 100 mM KCl, 1 mM EDTA, 1 mM dithiothreitol, 12.5% glycerol, and 2 µg of
poly(dI-dC)·poly(dI-dC). Reaction mixtures with recombinant Sp1 also
contained 10 µM ZnSO4. After incubation, the
reaction mixture was loaded directly onto a 4% polyacrylamide gel and
separated by electrophoresis at 100 V for 4 h at 4 °C (28).
Autoradiography was performed by exposure of the dried gel to Fuji-RX
film.
When Sp1 DNA binding activity was measured in the presence of PKA, 0.5 ng of purified Sp1 (99% purity, Promega) was incubated with 40 units of the PKA catalytic subunit (Sigma) in the presence of 10 mM MgCl2, 20 mM Tris (pH 7.4), and 40 µM ATP.
ImmunoblottingNuclear extracts (50 µg of protein) were separated by SDS-polyacrylamide gel electrophoresis and blotted onto nitrocellulose as described previously (28). Sp1 was detected with a rabbit polyclonal antibody (generously provided by Dr. Robert Tjian, University of California at Berkeley) diluted 1:1,000 in 1% dry milk, 10 mM Tris-HCl (pH 8.0), 0.15 M NaCl, 0.05% Tween 20. Alkaline phosphatase-conjugated goat anti-rabbit IgG served as the secondary antibody, and CSPD® served as the chemiluminescent substrate (Tropix Inc.). Autoradiography was performed by exposure of the blot to Fuji RX film.
TransfectionHL-60 and HL-60/AR cells were suspended at a
concentration of 1.5 × 107 cells/0.5 ml RPMI 1640 containing 40 mM Hepes (pH 7.4). The cell suspension was
incubated for 15 min in a 0.4-cm electroporation cuvette with either 20 µg of pSV-CAT (Promega) or p(
71) CAT, 2 µg pRSV-
Gal (to
normalize for transfection efficiency), and 50 µg DEAE-Dextran
(Promega). Electroporation was carried out at room temperature using a
Gene Pulser (Bio-Rad) set at 300 V and 960 microfarad. Following
electroporation, cells were diluted with 12 ml of RPMI 1640 medium
containing 10% fetal bovine serum, and after 2 h, either 100 µM RpcAMP[S] or 100 µM SpcAMP[S] was added, and cells were incubated for 24 h. In some experiments, HL-60/AR cells were cotransfected in the presence of 2 µM
CdCl2 with control plasmid pOT1521 or pOT1521R-I
(generously provided by Dr. Yoon Sang Cho-Chung, National Cancer
Institute) containing the regulatory subunit of type I PKA under the
control of the mouse metallothionein I promoter (29).
Sf9 insect cells were transfected by lipofection (Lipofectin, Life
Technologies, Inc.). Cells were grown in serum-free SF900 medium (Life
Technologies, Inc.) under air at 27 °C and transfected with 10 µg
of pSV40-CAT, 2 µg of pPacSp1, and 10 µg of
pPacC. CAT activity was normalized for transfection
efficiency by cotransfecting cells with 2 µg of pADH-
Gal
containing the lacZ gene under the control of the
Drosophila melanogaster alcohol dehydrogenase promoter.
Cells were harvested by
centrifugation at 500 × g for 5 min, washed once with
cold phosphate-buffered saline, suspended in 90 µl of 250 mM Tris (pH 7.8) containing 0.5 mM
phenylmethylsulfonyl fluoride, and sonicated with ten 1-s bursts at
4 °C. Cell debris was removed by centrifugation at 10,000 × g for 10 min at 4 °C. 60 µl of extract were heated at
70 °C for 10 min and centrifuged at 10,000 × g for
10 min, and the supernatant was used for measuring CAT activity (30).
Cell lysate (50 µl) was incubated in a reaction mixture containing:
100 mM Tris (pH 7.8), 6 mM MgCl2,
75 mM KCl, 0.5 mM sodium acetate, 0.5 mM coenzyme A, 3.75 mM ATP, 50 µM
chloramphenicol, and 0.25 µCi of [14C] chloramphenicol
(40-60 mCi/mmol, New England Nuclear) for 4-20 h at 37 °C.
Acetylated chloramphenicol was recovered by three extractions with
xylene, and the amount of acetylated [14C]chloramphenicol
was determined by liquid scintillation spectrometry (31).
-Galactosidase activity was measured by incubating 30 µl of cell
lysate in 100 µl of reaction mixture containing: 2.5 mg/ml
o-nitrophenyl-
-D-galactopyranoside in 0.1 M sodium phosphate (pH 7.5), 1 mM
MgCl2, and 45 mM
-mercaptoethanol. The
reactions were carried out in 96-well plates for 20-45 min, and
absorbance was read at 410 nm in a microplate reader (Series 750, Cambridge Technology, Inc.).
The human Sp1 cDNA was cloned into the NotI and SmaI site in baculovirus expression vector pVL1392 under the control of the polyhedrin promoter (32). The recombinant virus was plaque purified and used to infect Sf9 cells as described previously (33). Sf9 cells (4 × 106) were infected with 5 plaque-forming units/cell of recombinant virus for 1 h, and cells were harvested 60 h after infection. Cells were washed once in ice-cold phosphate-buffered saline and suspended in 4 ml of buffer D containing 20 mM Hepes (pH 7.4), 0.5 mM EDTA, 0.5 mM EGTA, 10 mM 2-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, and 0.5 mg/ml aprotinin and homogenized three times with 20 strokes of a Dounce homogenizer at 10-min intervals. Cell homogenates were centrifuged at 100,000 × g for 1 h at 4 °C. The resulting supernatant was applied to a radial DEAE column (MemSep1010, Millipore) attached to a Pharmacia FPLC. Elution was carried out at 4 °C at a flow rate of 2 ml/min with a linear gradient of 0-1.0 M NaCl in buffer D, and 2-ml fractions were collected. Fractions containing Sp1 were determined by Western blotting and were pooled, diluted 3-fold with buffer D, and applied to a Mono Q HR 5/5 column (Pharmacia). Elution was carried out with a linear gradient of 0-0.4 M NaCl in buffer D at a flow rate of 1.0 ml/min. Sp1 was approximately 25% pure after Mono Q chromatography as determined by Rapid Coomassie Stain (Diversified Biotech) of 10% SDS-polyacrylamide gels.
For determining Sp1 DNA binding activity, the fractions obtained from the Mono Q chromatography step were incubated with PKA catalytic subunit for 10 min at 30 °C in a 50-µl reaction mixture containing 20 mM Tris-HCl (pH 7.4), 10 mM MgCl2, and 40 µM ATP (28), and gel shift assays were carried out as described above.
Phosphorylation of Sp1Cell extracts were prepared from
Sp1-expressing Sf9 cells as described above and incubated for 20 min at
30 °C with a buffer containing 20 mM Tris-HCl (pH 7.4),
10 mM MgCl2, and 4 µCi of [-32P]ATP (3000 Ci/mmol). In some instances, either 40 units of purified PKA catalytic subunit
(Sigma) or 0.5 µg of PKA
inhibitor, PKI (Sigma), was included in the assay. The reaction was
terminated with 5 × SDS-polyacrylamide gel electrophoresis sample
buffer, and samples were separated in a 10% polyacrylamide gel by
SDS-polyacrylamide gel electrophoresis. Autoradiography was performed
by exposing the dried gel to Fuji RX film. Phosphorylation of purified
Sp1 (99% pure, Promega) was carried out in a similar manner.
For gel shift assays, 30 ng (0.5 footprint unit) of purified Sp1
(Promega) were incubated for 20 min in the presence or the absence of
20 units of purified bovine heart PKA catalytic subunit (Sigma) and
40 µM ATP. For immunoprecipitation, purified Sp1 was
(Promega) was incubated as described for gel shift assays and incubated
overnight at 4 °C with 1 µg of Sp1 polyclonal antibody (Santa Cruz
Biotechnology) as described previously (34).
8-Cl-cAMP specifically targets the high affinity cAMP
binding sites in the regulatory R-I subunit of PKA (35), which results in the sustained activation and proteolysis of type I PKA and the
down-regulation of cAMP-dependent transcription factors
(26, 28). To determine if Sp1 DNA binding activity is regulated in a
similar fashion, HL-60/AR cells were treated for 48 h with 2.5 µM 8-Cl-cAMP, a 10% inhibitory concentration, and DNA
binding activity was measured by mobility shift analysis (Fig.
1A). Nuclear extracts from
HL-60/AR cells (Fig. 1A, lanes 3 and
4) exhibited increased basal Sp1 DNA binding activity
compared with wild type cells (Fig. 1A, lanes 1 and 2). Treatment of HL-60/AR cells with 8-Cl-cAMP markedly
reduced Sp1 DNA binding activity (Fig. 1A, lanes
5 and 6). Because the MDR phenotype in HL-60/AR cells
is maintained by exposure to doxorubicin, Sp1 DNA binding activity was
measured in the absence of selective pressure by maintenance of
HL-60/AR cells in drug-free medium (Fig. 1B). Growth of
cells in doxorubicin-free medium for 2 weeks resulted in the virtual disappearance of Sp1 activity (Fig. 1B, lane 1)
that was dramatically induced upon re-exposure to 1 µM
doxorubicin for either 24 or 48 h (Fig. 1B, lanes
2 and 3). In contrast, maintenance of HL-60/AR cells in
1 µM doxorubicin plus the addition of either the
nonhydrolyzable PKA antagonists RpcAMP[S] and Rp-8-Cl-cAMP[S] (36)
or 8-Cl-cAMP counteracted the up-regulation of Sp1 activity by
doxorubicin (Fig. 1B, lanes 4-6). The levels of
Sp1 as determined by Western blotting were not affected under these
conditions (results not shown).
HL-60/AR Cells Exhibit Increased SV40 Promoter Activity
To
determine whether the increased Sp1 DNA binding activity in HL-60/AR
cells correlated with Sp1-dependent transcriptional activation, HL-60/AR cells were transfected with a plasmid containing the CAT reporter gene under the control of the SV40 promoter, which
contains six tandem Sp1 response elements (2) (Fig.
2). CAT activity was increased 27-fold in
HL-60/AR cells compared with wild type cells, and transcriptional
activation correlated closely with DNA binding activity (Fig. 1). In
comparison, transcription from p(
71)-CAT, which contains a single
CRE (37), was enhanced 5-fold in resistant cells (Fig. 2).
We next determined if endogenous PKA is required for Sp1
trans-activation. In these experiments, HL-60/AR cells were
cotransfected with pSV40-CAT and pOT1521R-I containing the R-I subunit
of PKA (38) (Fig. 3). In the absence of
R-I overexpression, treatment of HL-60/AR cells with the
nonhydrolyzable and membrane permeable PKA agonist, SpcAMP[S] (36),
stimulated CAT activity approximately 2-fold, whereas RpcAMP[S]
attenuated CAT activity by 20%. Transfection of HL-60/AR cells with
increasing amounts of R-I produced a progressive inhibition of CAT
activity in the absence of drug treatment, and upon treatment with
RpcAMP[S], reporter gene activity was blocked completely. In
contrast, the inhibitory effect of R-I was reversed with SpcAMP[S].
Transfection with the empty vector, pOT1521, did not inhibit CAT
activity (results not shown).
To further assess the requirement for PKA in Sp1-dependent
trans-activation, Sf9 insect cells, which lack endogenous
Sp1 and have low PKA activity, were cotransfected with pSV40-CAT and
pPacSp1 in the presence or the absence of
pPacC, which contains the PKA catalytic subunit under
the control of the D. melanogaster actin 5C promoter (39)
(Fig. 4). No CAT activity was observed in
the absence of Sp1 transfection (results not shown), but a small degree of trans-activation was observed when Sp1 alone was
expressed. In contrast, CAT activity was stimulated 9-fold when cells
were transfected with pPacSp1 and pPacC
.
PKA Stimulates Sp1 DNA Binding Activity in Vitro
Because
these results strongly suggested that PKA was necessary for the
activation of Sp1 trans-activating and DNA binding activities, Sp1 was expressed in Sf9 insect cells using a recombinant baculovirus, and DNA binding activity was determined in the absence and
the presence of PKA (Fig. 5A).
Sp1 was partially purified by DEAE and Mono Q anion-exchange
chromatography that resulted in the elution of approximately equal Sp1
levels in fractions 17-20 as determined by immunoblotting (Fig.
5B). Fractions 17 and 18 were devoid of basal DNA binding
activity, but the addition of exogenous PKA resulted in a marked
stimulation of activity (Fig. 5A). On the other hand,
fractions 19 and 20 did exhibit DNA binding activity that was further
stimulated by PKA. Sp1 phosphorylation was also determined in
vitro with cell lysates from Sp1-expressing insect cells (Fig.
5C). Assays were conducted with uninfected Sf9 cell extract
in the presence of exogenous purified PKA (lane 1), and in
Sp1-containing extracts in the presence (lane 2) and the
absence (lane 3) of the PKA inhibitor, PKI, as well as in the presence of exogenous PKA (lane 4). These results
indicate that Sp1 is phosphorylated by both endogenous and exogenous
PKA, and that PKA can stimulate Sp1 DNA binding activity. These results also suggest that the presence of partially activated Sp1 in fractions 19 and 20 (Fig. 5A) is a result of its partial
phosphorylation by endogenous PKA.
The phosphorylation of Sp1 by PKA was also determined using purified
Sp1 (Fig. 6, A and
B). Phosphorylated Sp1 was readily detected before (Fig.
6A) and after (Fig. 6B) immunoprecipitation with
an Sp1 polyclonal antibody. It is interesting that PKA coprecipitated with Sp1, suggesting high affinity between these proteins. Sp1 incubated in the presence of PKA exhibited greater DNA binding activity
than in its absence (Fig. 6C), and incubation of PKA phosphorylated Sp1 with PP2A markedly reduced its DNA binding activity
(Fig. 6D).
The present study provides strong evidence that Sp1
trans-activating and DNA binding activities are modulated
through a cAMP/PKA signaling pathway. Momoi et al. (24) was
the first to show that a CRS with homology to the Sp1 response element
was involved in the cAMP-dependent transcriptional
activation of the bovine CYP 11A and human CYP
21B genes. Our data are consistent with these results and
demonstrate further that PKA can stimulate transcription from an
Sp1-dependent promoter in intact cells as well as activate the DNA binding activity of Sp1 in vitro. In previous
studies, Sp1 was shown to be phosphorylated at multiple sites in HeLa
cell nuclear extracts and to be a substrate in vitro for
DNA-dependent protein kinase; however,
DNA-dependent protein kinase did not affect the extent and
specificity of DNA binding or Sp1-dependent transcription
(40). In contrast, our data indicate that not only is Sp1
phosphorylated by PKA but that the DNA binding and trans-activating activities are stimulated as well. One
mechanism that may account for these results is that phosphorylation of the trans-activation domain of Sp1 results in increased Sp1
multimerization. Sp1 domain B confers high affinity DNA binding (41)
and contains a PKA consensus phosphorylation site at
Thr366. Although still speculative, PKA phosphorylation may
induce conformational changes similar to those that occur by
multimerization through this domain (42, 43). The tight association
between Sp1 and PKA is similar to the complex formed between PKA
catalytic subunit and NF-B (44) and gives further credence to this
mechanism of Sp1 regulation.
PKA activity remained unchanged in HL-60/AR cells regardless of whether cells were maintained in doxorubicin (26), suggesting that drug-mediated Sp1 activation may require an additional signaling mechanism. One possibility is that protein dephosphorylation is also regulated by PKA and that this process may be a rate-limiting factor in Sp1 deactivation. Maximal activation of Sp1 may require the inactivation of PP1 and PP2A by the PKA-mediated activation of a nuclear inhibitory protein such as NIPP-1 (45). This mechanism is analogous to the attenuation of CREB by PP1 and PP2A (46, 47). Phorbol esters also mediate inactivation of Sp1 DNA binding in HL-60/AR cells (11), and although PKC does not phosphorylate Sp1 in vitro (40), it does activate PP2A (48). These results are consistent with the inhibitory effect of phorbol esters on CRS-dependent transcription (49). Therefore, the mechanism of Sp1 inactivation may comprise multiple signaling pathways involving phosphorylation and dephosphorylation.
Sp1 activity was up-regulated by doxorubicin treatment, suggesting that MDR-associated drugs can directly influence trans-activation and the MDR phenotype. Because the promoter region in the Sp1 gene contains five Sp1 sites with three additional elements in the first intron (50), exposure of cells to doxorubicin may create a positive feedback loop that results in its autoregulation. This mechanism may also account for the increased expression of type I PKA, because the promoter of the R-I gene contains multiple Sp1 response elements (51). In addition, the MRP drug transporter gene, which is expressed in HL-60/AR cells, also contains multiple Sp1 elements (52), and therefore, there may be a close association between selective pressure and Sp1 activation as a mediator of resistance.
We are indebted to Dr. Robert Tjian,
University of California at Berkeley, for providing the Sp1 antibody
and plasmids pPacSp1 and pADH-Gal and to Dr. Yoon Sang
Cho-Chung, National Cancer Institute, for providing pOT1521R-I,
p
(
71)CAT, and the C
cDNA.