| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
in the Human Myometrium during Pregnancy and Labor Involves Transcriptional Regulation by Cyclic Adenosine 3',5'-Monophosphate and Binding of Phosphorylated Nuclear Proteins to Multiple GC Boxes within the Promoter
Department of Obstetrics and Gynecology, University of Newcastle upon Tyne, Royal Victoria Infirmary, Newcastle upon Tyne NE1 4LP, United Kingdom
Address all correspondence and requests for reprints to: Dr. Robert J. Phillips, Department of Obstetrics and Gynecology, University of Newcastle upon Tyne, 4th Floor, Leazes Wing, Royal Victoria Infirmary, Newcastle upon Tyne NE1 4LP, United Kingdom. E-mail: g.n.europe-finner{at}ncl.ac.uk or r.j.phillips{at}ncl.ac.uk.
Abstract
Gs
is the G protein subunit that stimulates adenylyl cyclase activity in the myometrium during pregnancy, raising intracellular levels of the smooth muscle relaxant cAMP. The promoter region of the gene encoding Gs is GC rich and contains multiple putative binding sites for the specificity protein (Sp) transcription factor family. In electrophoretic mobility shift assays, four of these Sp sites were bound by recombinant Sp1 protein. Binding was dependent on phosphorylation of Sp1 by protein kinase A. Phosphorylated Sp1–4 proteins were observed in extracts of cultured human myometrial cells, but in electrophoretic mobility shift assays Gs promoter sequence binding by Sp1 was not apparent. Instead, these assays showed phosphorylation-dependent Gs promoter binding by lower molecular weight myometrial proteins that could not be supershifted by antibodies specific to Sp1–4 proteins. To investigate the regulation of Gs expression, the GC-rich promoter region was used to direct transcription of a firefly luciferase reporter gene in transient transfection assays of primary human myometrial cell cultures, COS-7 and HEK 293 cells. Reporter gene expression was found to follow a biphasic response to forskolin and 8-bromo-cAMP, with an initial, concentration-dependent increase in luciferase activity, followed by a prolonged decrease. In myometrial cells, this pattern was also seen in response to treatment with human chorionic gonadotropin.
PRETERM LABOR IS now the primary cause of neonatal death, as a consequence of failure to reduce either its rate of incidence or the level of neonatal mortality resulting from prematurity (1). The lack of progress in preventing preterm labor correlates with the lack of understanding of the biochemical mechanisms regulating uterine contractile activity. One of the most significant physiological adaptations of the uterus to pregnancy is the development of a relative state of myometrial smooth muscle inactivity known as quiescence. During labor the quiescent state ends, and a series of powerful uterine contractions expels the neonate. It is becoming increasingly apparent that these processes involve the differential expression of specific genes that are responsible for controlling the activity of the uterus during pregnancy and parturition.
In previous studies defining the expression of GTP-binding protein (G protein) species in the human myometrium during fetal maturation we have shown that up-regulation of the adenylyl cyclase (AC) stimulatory G protein (Gs) occurs in the human myometrium during pregnancy (2) and that this is associated with a concomitant increase in Gs-coupled AC activity (3). During term or preterm labor, levels of Gs protein and Gs-coupled AC activity in the myometrium are substantially reduced (3, 4). These observations suggest that the myometrial Gs-AC pathway is involved in the maintenance of uterine quiescence throughout gestation, but at present the mechanisms responsible for increased expression of Gs in the myometrium over this period remain undefined.
The promoter region of the human Gs gene has an extremely high GC nucleotide content and contains four canonical 5'-CCGCCC-3' and two noncanonical 5'-CCCTCCCC-3' specificity protein (Sp) transcription factor-binding sites as well as a 5'-CGTCA-3' hemipalindromic cAMP response element (5). The first Sp protein to be identified, Sp1, is a ubiquitously expressed, 105-kDa transcription-activating, DNA-binding protein with three zinc fingers. Sp1 is glycosylated and multiply phosphorylated, apparently by more than one protein kinase. Although phosphorylation of Sp1 by DNA-dependent protein kinase does not appear to affect the extent or specificity of DNA binding (6), the cAMP-dependent protein kinase A (PKA) has been shown to stimulate the binding activity of Sp1 in vitro (7, 8, 9). Of the other proteins in the human Sp family, Sp3 and Sp4 have extensive homology with Sp1, whereas Sp2 is less similar.
There is now a large number of reports of transcriptional regulation involving Sp proteins, in particular Sp1 and Sp3. In several of these cases, the cAMP signaling pathway has also been implicated. cAMP-dependent expression involving Sp1-, Sp3-, and Sp1-like proteins has been shown for the human genes for surfactant protein A2 (10), progesterone receptor (11), apolipoprotein AI (12), biglycan (13), ferredoxin (14), cytochrome P450c17 (15), and MDR1 (16). For the progesterone receptor (11), apolipoprotein AI (12), biglycan (13), ferredoxin (14), and MDR1 (16) genes, cAMP responsiveness was shown to be dependent on PKA, and it has been reported that phosphorylation of Sp1 by PKA increases the trans-activating and DNA-binding properties of this transcription factor (8). The presence of multiple putative Sp transcription factor-binding sites in the promoter of the human Gs gene suggests a mechanism for the regulation of Gs expression by Sp protein activity, which may allow control to be exerted by PKA. However, at present there is little evidence to indicate that this occurs in vivo in the myometrium and other tissue types.
In this present investigation we have demonstrated that phosphorylated forms of each of the Sp1 through Sp4 proteins are expressed in cultured human myometrial cells. We have shown that recombinant Sp1 protein, when phosphorylated by PKA, bound specifically to four of the six potential Sp-binding sites from the Gs promoter in vitro. Using myometrial cell nuclear extracts, no Sp DNA-binding activity was seen, but binding of phosphorylated, low molecular weight proteins to four of the six potential Gs Sp-binding sites was observed. In addition, in reporter gene assays transcription directed by the Gs promoter was found to respond similarly to treatment of cells with forskolin, 8-bromo-cAMP, and human chorionic gonadotropin (hCG). The response was biphasic, with an initial rise in transcription, followed by a prolonged decrease. The effect was dose dependent, with higher concentrations of treatments reducing the levels of transcription. These data indicate that transcriptional control of myometrial Gs may be exerted by a cAMP-dependent mechanism.
Materials and Methods
Cell culture
Primary adherent cell cultures of human myometrial smooth muscle cells, used up to passage 3, were established from myometrial tissue samples taken from the lower uterine segment during elective cesarean sections, as described previously (17). Written consent was obtained from all women, and ethical approval was granted by the Newcastle and North Tyneside Health Authority ethics committee. Adherent cell cultures of the African green monkey kidney cell line COS-7 and the human embryonic kidney cell line HEK 293 were also established using routine cell culture techniques. All cultures were incubated in DMEM (Life Technologies, Inc., Gaithersburg, MD) supplemented with 10% (vol/vol) fetal bovine serum, 1 U/ml penicillin, and 1 ng/ml streptomycin, at 37 C in a humidified atmosphere of 95% air/5% CO2.
Immunoprecipitation
Myometrial cells grown in 75-cm2 culture flasks were incubated for 1 h in phosphate-free DMEM (Life Technologies, Inc.) at 37 C. [32P]Orthophosphate (833 µCi; 10 mCi/ml; Amersham Pharmacia Biotech, Piscataway, NJ) was added to each T75 flask, and cultures were incubated for 24 or 72 h at 37 C. To harvest the cells, they were washed twice with PBS and incubated for 5 min with trypsin/EDTA at 37 C. Cells were collected by centrifugation, washed with PBS, and resuspended in lysis buffer [50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% sodium deoxycholate, 1% Triton X-100, 0.5 mM dithiothreitol, 0.2 mM EDTA, 100 nM okadaic acid, and 5 nM calyculin A, supplemented with protease inhibitor cocktail]. After 15 min on ice, the lysate was cleared by centrifugation at 12,300 x g for 10 min at 4 C. For each immunoprecipitation reaction, cleared lysate containing 100 µg protein was incubated with 50 µl Pansorbin cells (Calbiochem, La Jolla, CA) on ice for 30 min, then centrifuged at 7,500 x g for 3 min. The supernatant was transferred to a fresh tube, and 4 µg anti-Sp1 (PEP 2), -Sp2 (K-20), -Sp3 (D-20), or -Sp4 (V-20) polyclonal antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) were added. After 90 min on ice, 50 µl Pansorbin cells were added, and the mixture was incubated for 16 h at 4 C on a gently rocking platform. The immunoprecipitated complexes were collected by centrifugation at 7,500 x g for 3 min and washed three times by resuspension and centrifugation in 600 µl Pansorbin wash buffer [20 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.5% Nonidet P-40, and 5 mM EDTA]. The washed, pelleted complexes were resuspended in SDS-PAGE loading buffer, and the proteins were separated by SDS-PAGE in 10% gels. Gels were dried onto 3MM paper (Whatman, Clifton, NJ) and exposed to X-OMAT LS film (Eastman Kodak Co., Rochester, NY) for visualization of radiolabeled proteins by autoradiography.
Preparation of nuclear extracts
Cultured myometrial cells were harvested by trypsin digestion, washed with ice-cold PBS, and resuspended by gentle pipetting in 4 packed cell vol of ice-cold hypotonic lysis buffer [10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonylfluoride, 100 nM okadaic acid, and 5 nM calyculin A]. After 15-min incubation on ice, 0.25 original vol of 10% Nonidet P-40 were added, and cells were vortexed vigorously for 10 sec and centrifuged at 13,000 x g for 1 min. The supernatant, containing cytoplasmic protein, was removed, and the pellet, containing nuclei and cell debris, was resuspended in 0.5 original vol of ice-cold, high salt extraction buffer [20 mM HEPES (pH 7.9), 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonylfluoride, 100 nM okadaic acid, and 5 nM calyculin A]. The suspension was incubated at 4 C with constant agitation for 15 min, then centrifuged at 12,300 x g for 5 min at 4 C. The supernatant, containing nuclear extract, was split into aliquots and stored at -75 C. Fresh myometrial tissue nuclear extract was also prepared in a similar manner after homogenization.
Electrophoretic mobility shift assay (EMSA)
All oligonucleotides used and the Gs promoter sequence from which they were derived are detailed in Fig. 1
. Oligonucleotides (20 pmol each; Life Technologies, Inc.) were 5'-end labeled using T4 PNK (Promega Corp., Madison, WI) and 50 µCi [
-32P]ATP (Amersham Pharmacia Biotech), annealed to form double-stranded pairs, and purified using ProbeQuant G50 microcolumns (Amersham Pharmacia Biotech). Unlabeled competitor oligonucleotide pairs were prepared by annealing 2 nmol of each oligo to give a final concentration of 10 µM.
|
Plasmid construction
A 461-bp fragment of the Gs gene promoter, from -789 to -329 in relation to the translation initiation site (5) (see Fig. 1A
), was amplified by PCR from human genomic DNA using pfu proofreading thermostable DNA polymerase (Promega Corp.). Note that position -329 is the furthest downstream transcriptional start site according to Kozasa et al. (5), whereas the region from -329 to 0 is the 5'-untranslated region of exon 1 of Gs. This fragment was cloned directly into the pTOPO II topoisomerase-linked cloning vector (Invitrogen, San Diego, CA) and sequenced to confirm fidelity, then subcloned by XhoI/KpnI digestion and ligation into the pGL3-E vector (Promega Corp.). The resulting plasmid, pGs2GL3, thus carries the Gs promoter fragment directing transcription of the firefly luciferase gene with upstream enhancer elements to increase overall transcriptional levels.
Transient DNA transfections
Cells were grown in 35-mm culture dishes to 70% confluence. For each dish, 3 µl lipofectin transfection reagent (Life Technologies, Inc.) were mixed with 100 µl OptiMEM reduced serum medium (Life Technologies, Inc.; not supplemented with antibiotics), and the mixture was incubated at room temperature for 30 min. Sterile, highly purified, plasmid DNA was diluted to 10 µg/ml in OptiMEM medium. For each dish, 1 µg diluted DNA was mixed with the lipofectin/OptiMEM mixture and incubated for an additional 15 min before the addition of 0.8 ml supplemented DMEM. Culture medium was removed from the cells, and they were then washed with 1 ml untreated OptiMEM medium, which was replaced with the DNA/lipofectin/medium mixture. Cultures were incubated at 37 C for 5 h to allow transfection to occur, the transfection mix was replaced with 3 ml supplemented DMEM, and cells were incubated for an additional 16 h before treatment. Note that in each set of experiments a sufficient number of cells were transfected with the Gs promoter luciferase reporter gene in one batch before being split into individual culture vessels and treated. Transfection efficiencies using the ß-galactosidase-encoding plasmid pcDNA3.1/His/lacZ (Invitrogen) were between 20–25% for each cell type.
Luciferase assay
For single point experiments, cultures were chemically treated and incubated as indicated. Culture medium was then aspirated, and cells were washed with PBS and lysed by addition of 500 µl passive lysis buffer (Promega Corp.) directly to the culture dish, followed by incubation for 15 min at room temperature on a rocking platform. Lysate (20 µl) was transferred to a 1.5-ml microcentrifuge tube and mixed with 100 µl luciferase assay reagent (Promega Corp.), then firefly luciferase activity was measured for 10 sec in a Turner Designs TD 20-20 luminometer (Palo Alto, CA).
For multiple time point experiments, culture medium was aspirated and replaced with 2 ml fresh medium containing 10 µg beetle luciferin (Promega Corp.). After 5 min, luciferase activity was measured by placing the 35-mm culture dish directly in the luminometer and reading with an integration period of 30 sec. The culture medium was replaced with fresh medium, with or without added chemical treatments, and returned to 37 C incubation. At the next time point, beetle luciferin was added to the culture medium at a final concentration of 5 µg/ml. Luciferase activity was again measured, the cultures were treated as before, and the cycle was repeated for the appropriate number of time point measurements.
Immunoblotting
For Western blotting, treated myometrial cells were lysed directly by addition of SDS-PAGE loading buffer [62.5 mM Tris-HCl (pH 6.8), 2% sodium dodecyl sulfate, 25% glycerol, and 0.01% bromophenol blue]. Proteins were denatured at 100 C for 5 min, separated by SDS-PAGE in 10% gels, and transferred to polyvinylidene difluoride membranes, which were blocked by overnight incubation at 4 C in 4% fat-free dried milk in PBS. Blocked membranes were washed in PBS and incubated for 2 h at room temperature with the primary, Gs-specific RM/1 antibody, (NEN Life Science Products, Boston, MA) at a 1:1000 dilution in 4% fat-free dried milk in PBS. Membranes were then extensively washed with PBS and incubated for 1 h at room temperature with the secondary, horseradish peroxidase-conjugated goat antirabbit IgG antibody (DAKO Corp., Carpenteria, CA) at a 1:3000 dilution in 4% fat-free dried milk in PBS. Membranes were washed in PBS once more, then treated with enhanced chemiluminescence reagent (Amersham Pharmacia Biotech) and exposed to flash-sensitized Hyperfilm ECL (Amersham Pharmacia Biotech). Immunodetected bands were scanned using a UMAX PS 2400 scanner coupled to the Intelligent Quantifier software package from BioImage (Ann Arbor, MI). Blots were then reprobed using the SW/1 primary antibody, recognizing an epitope common to the five different Gß-subunits, and Gß protein levels were measured to allow identification of differences in protein loading between gel lanes.
Results
Phosphorylated Sp are present in cultured human myometrial cells
Proteins from myometrial cells in culture were radiolabeled by incubation of cell cultures with [32P]orthophosphate to allow identification of phosphorylated Sp protein species by immunoprecipitation, as shown in Fig. 2. Using specific antibodies, phosphorylated forms of each of the Sp1–4 were precipitated. Sp1 antibody precipitated four phosphorylated proteins of approximately 110, 83, 70, and 43 kDa. Sp2 antibody precipitated a faint phosphorylated protein at 83 kDa and a highly phosphorylated protein at 50 kDa. Sp3 antibody precipitated phosphorylated proteins of approximately 175 and 110 kDa, whereas Sp4 antibody precipitated phosphorylated proteins of 110, 100, and 34 kDa. In the absence of Sp antibodies, the protein A-containing precipitating agent Pansorbin did not precipitate any phosphorylated protein species, indicating the specificity of the immunoprecipitation reaction.
|
Recombinant Sp1 was assayed for binding to the Sp1 consensus sequence by mobility shift assay. This binding, which was manifested by the formation of a single, low mobility, radiolabeled complex, was not observed with the recombinant protein unless PKA was included in the reaction mixture (Fig. 3A). The stimulatory effect of PKA could not be replaced by substitution with PKC. In the absence of Sp1, PKA showed no DNA binding activity of its own.
|
promoter
Oligonucleotide pairs containing potential Sp-binding sites from the Gs promoter were tested in the mobility shift assay with PKA-treated recombinant Sp1. Figure 3B shows Sp1 binding to the consensus sequence, to a known Sp1-binding sequence from the connexin 43 (Cx43) promoter, and to the -396 to -375 region of the Gs promoter (G2). In each case, an Sp1-specific supershift was demonstrated. The sequence specificity of the binding was shown by competition with excess unlabeled oligonucleotide pairs. Binding to each labeled probe was effectively eliminated by inclusion of the same sequence as competitor (Fig. 3B, lanes 2, 7, and 12), but binding was unaffected by inclusion of the consensus sequence for the unrelated transcription factor, Oct-1 (Fig. 3B, lanes 4, 9, and 14). When another Gs promoter sequence, G1, was used as the competitor (Fig. 3B, lanes 3, 8, and 13), only Sp1 binding to the G2 probe was affected, and this was only reduced and not eliminated (lane 8). This implies that Sp1 had some binding affinity toward the G1 competitor sequence, but that this was lower than the affinity toward the probe sequences and was sufficiently low that a 100-fold excess of G1 competitor did not noticeably impair the mobility shift of the Sp1 consensus and Cx43 sequences. In support of this, a low level of sequence-specific binding of the recombinant Sp1 protein to the radiolabeled G1 probe sequence was observed (data not shown).
Recombinant Sp1 protein was assayed for binding to a number of other sequences from the Gs promoter. Figure 3C shows that of four further potential Sp-binding sites assayed, only two, G3 and G6, were bound by Sp1. No binding was seen using a probe containing a hemipalindromic cAMP response element (CRE) sequence from the Gs promoter.
Phosphorylated, low molecular weight proteins from myometrial cells bind Gs promoter sequences
Nuclear extracts were prepared from cultured myometrial cells and tested in EMSAs. Mobility shift at a position equivalent to that with recombinant Sp1 was not seen using any of the protein extracts from several different cell cultures. An alternative pattern of bandshifting was observed with several oligonucleotide probes, with the positions of the two protein-DNA complexes indicating binding by much smaller proteins than the recombinant Sp1 (Fig. 4A). Sequence-specific binding, which was disrupted by competition with the same sequence, but not with the unrelated Oct-1 sequence, was seen with the G2, G3, G5, G6, and Cx probes, but not with the mutated G2mut sequence. No binding was observed with the Sp consensus, G1, G4, and CRE probes (data not shown). The same two protein-DNA complexes (I and II), which were closely related in size, were seen with each of the G2, G3, G5, G6, and Cx DNA probes, but their relative levels varied, so that for G2 the smaller complex I was dominant, whereas for G3 the more slowly migrating complex II prevailed. The G3 probe also gave rise to a higher molecular weight complex III, which was seen at a position intermediate between those of the recombinant Sp1 and complex II. None of the protein-DNA complexes seen using myometrial protein extracts was supershifted by any of the Sp1–4 antibodies (data not shown).
|
GC sequences was further examined by attempting to disrupt the complex formed with the G2 probe by competition, as shown in Fig. 4B. Although complex formation was slightly diminished by the inclusion of the G2mut competitor, the only sequences to completely prevent the mobility shift were G2 itself and Cx. This experiment also demonstrated that the protein involved in this binding reaction is phosphorylated, as incubation of the myometrial cell nuclear extract with potato acid phosphatase before the mobility shift assay resulted in a loss of complex formation. The same results were seen when nuclear extracts from cultured myometrial cells were replaced with similarly prepared extracts from homogenized myometrial tissue samples obtained from pregnant nonlaboring women during elective cesarean section (see Fig. 4C). Similar binding was observed with nuclear extracts prepared from term laboring myometrium (data not shown).
The Gs promoter directs transcription that is regulated by cAMP
A region of the Gs promoter containing the six GC boxes and the hemipalindromic CRE was used to direct transcription of the firefly luciferase gene in transient transfection assays of human myometrial cells and COS-7 and HEK 293 cells. In these assays the effects of treating the cultured cells with differing concentrations of forskolin, 8-bromo-cAMP, and hCG on the transcriptional activity of the pGs2GL3 reporter construct were monitored by measuring levels of cellular luciferase activity. In control transfection reactions using the promoterless pGL3-E parent vector, there was no detectable luciferase activity, indicating that the activity seen in pGs2GL3 transfections was due to the promoter function of the inserted Gs sequence.
Multiple time point analyses of the effects of increasing concentrations of forskolin on the pGs2GL3 reporter in COS-7 cells are shown in Fig. 5A. These data indicate that a biphasic response occurred with the addition of 50 µM forskolin, such that there was an initial rise in transcriptional activity up to 3.5 h, followed by a decrease in activity to below control levels. Increasing the forskolin concentration to 100 µM resulted in a smaller increase in activity, which decreased after approximately 1.5 h to lower levels of activity than those observed for 50 µM forskolin. In contrast, the addition of 200 µM forskolin resulted in an immediate decrease in transcriptional activity of the reporter gene compared with the control, which was further decreased after 1 h to much lower levels than those found for the lower concentrations of forskolin. A similar pattern of response was seen on treatment of the COS-7 cells with the cell-permeable cAMP analog 8-bromo-cAMP, as shown in Fig. 5B. A biphasic response was seen with 100 µM 8-bromo-cAMP, with an initial increase followed by a decrease in luciferase activity compared with the levels in untreated control cells. With 200 µM and 1 mM 8-bromo-cAMP treatment, a marginal initial rise was then followed by a progressive fall in luciferase activity.
|
shows the results of assays of COS-7 and HEK 293 cells transiently transfected with pGs2GL3 and incubated for 6 h with forskolin or 8-bromo-cAMP. These results indicate that similar dose effects of 8-bromo-cAMP and forskolin occurred in both cell types. Addition of either 10 µM 8-bromo-cAMP or 10 µM forskolin caused an increase in promoter activity, but treatment with higher concentrations (0.1 and 1 mM 8-bromo-cAMP and 50 µM forskolin) resulted in decreased activity relative to that in untreated control cells.
|
. Addition of 50 µM forskolin resulted in an initial rise in transcriptional activity that occurred within 1 h, after which there was a prolonged fall to below the levels in control cells, continuing for at least 10 h. Increasing the forskolin concentration to 200 µM caused an immediate decrease in transcriptional activity, which remained below the levels in both control and 50 µM forskolin-treated cells. Incubation with 10 nM hCG, which acts via receptor-coupled Gs stimulation of ACs, resulted in an increase in transcriptional activity that was maintained for at least 2 h and then progressively decreased to control levels within 10 h. When the concentration of hCG was increased to 100 nM, luciferase activity was much lower than the control level within 1 h, with no initial increase in transcription seen.
|
expression was also observed at the protein level in vivo in primary cultures of human myometrial cells incubated with hCG, as shown in Fig. 8. Incubation with 10 nM hCG for 5 h resulted in increased Gs protein expression, as measured by Western blotting with the specific RM/1 antibody, whereas incubation with 100 nM hCG resulted in no apparent increase in protein expression (Fig. 8A), which correlates with the in vitro Gs reporter assay. Moreover, the effect of hCG was negated in the presence of the highly specific PKA inhibitors KT5720 and HA1077 (Fig. 8B), indicating the involvement of PKA in this process.
|
Immunoprecipitation experiments indicated that phosphorylated forms of each of the Sp1–4 proteins are expressed in human myometrial cells, with apparent molecular masses ranging between 43–175 kDa. Not all of the phosphorylated species observed corresponded to Sp proteins detailed by the antibody manufacturer, but several alternative forms of Sp proteins have also been described. The Sp1 antibody used here has been seen to detect a range of proteins between 60–120 kDa in human leiomyosarcoma cells (13). Alternative splicing of murine Sp1 mRNA has been shown to produce 48- and 80-kDa forms of the protein (18), and in human HL60 cells, the action of the serine protease myeloblastin was found to convert all full-length Sp1 protein to a truncated 30-kDa fragment capable of DNA binding at GC elements (19). In addition to the 115-, 80-, and 78-kDa forms of Sp3 known to be produced by the use of alternative translational start sites (20), a 28-kDa form of the protein has been seen in decidualized human endometrial stromal cells (21). If these proteins do not represent cross-reactive non-Sp1–4 species they may therefore represent previously uncharacterized alternative Sp isoforms. The use of radiolabeling by phosphorylation in vivo may act to enhance the detection of protein moieties not normally seen in Western blotting. For instance, Sp2 and Sp4 were considerably under represented compared with Sp1 and Sp3 in Western blots of unlabeled myometrial cell proteins (data not shown), but their phosphorylated forms were equivalently represented. It is possible that the total levels of Sp1 and Sp3 isoforms in these cells are higher than those of Sp2 and Sp4, but that there are similar levels of phosphorylated moieties of each of the proteins.
Mobility shift assays using recombinant Sp1 showed that PKA-activated Sp1 specifically bound multiple GC boxes within the promoter region of the Gs gene, resulting in a single high molecular weight protein-DNA complex. The human Cx43 promoter contains several CCTCC motifs that have been shown to bind Sp1 and are required for maximal promoter activity in human myometrial cells (22). Paired oligonucleotides corresponding to two CCTCC sequences in tandem from the Cx43 promoter region were therefore used as a positive control in EMSAs alongside the Gs GC sequences, resulting in an identical bandshift. This implies that despite the tandem binding sequences, only a single Sp1 complexes with each Cx43 probe. Although there may only be one functional Sp1-binding site on the probe, there may also be a physical limitation to multiple protein binding due to the size of the oligonucleotide pair. The requirement for the specific phosphorylation of Sp1 by PKA for binding to Gs and Cx43 promoter sequences supports previous results (7, 8) showing that Sp1 binding activity is greatly increased by PKA.
When recombinant Sp1 protein was replaced with nuclear protein extracts from myometrial cells in culture in EMSAs, doublets of much lower molecular weight protein-DNA complexes were observed with four of the Gs GC boxes, as well as with the Cx43 probe. Binding to the G3 Gs sequence also resulted in a further bandshift complex. The Gs and the Cx43 myometrial protein-DNA complexes were not supershifted by incubations with Sp1–4 antibodies, indicating that the nuclear proteins from myometrial cells in culture that bound these GC boxes were not immunologically related to the Sp1–4 proteins. Additionally, no bandshifting was seen using the Sp1 consensus sequence as the probe, and an excess of unlabeled Sp1 consensus sequence was unable to compete with protein-DNA complex formation using the G2 probe. This implies that the proteins possessed different DNA sequence recognition properties from Sp1, which would not be the case if they were simply isoforms of Sp1.
Several examples of Sp-like proteins interacting with Sp-binding sites have previously been described. In extracts from myometrially derived leiomyosarcoma cells, an Sp1-like binding activity was seen using a DNA probe containing a CCCTCCCC sequence responsible for biglycan gene cAMP responsiveness (13). Although the complex formed with this probe was supershifted with an Sp1 antibody, recombinant Sp1 protein did not bind to the sequence. The human Cx43 gene promoter contains several potential Sp-binding sites, and proteins from human myometrial cells have been seen to form three bandshift complexes in EMSAs using a 50-bp Cx43 sequence incorporating three CCTCC motifs (22). Two of these complexes were supershifted by Sp1 antibody, but the third was not. A CRE in the surfactant protein A2 gene containing a GGGGTGGGG sequence was found to bind both Sp1 and a 55-kDa protein that was not recognized by an Sp1 antibody and was not competed away by an excess of the consensus Sp1 GC box sequence (10). Another GT element in the cystathionine ß-synthase gene bound both Sp1 and Sp3 in addition to a smaller protein that was not recognized by antibodies to either protein, but was competed by an excess of Sp1 consensus sequence (23). The rat LH receptor gene promoter contains two adjacent sequence elements, a GT box that was found to bind Sp1 and a canonical GC box that bound several proteins from CHO cells that were not recognized by Sp1 antibody, including a high mobility protein doublet (24). The identities of these Sp-like proteins are not known, although other members of the Krüppel-like zinc finger DNA-binding protein family, such as LKLF and BTEB1, have been suggested (10, 24).
Despite the observation of Sp in cultured myometrial cell extracts and the binding of recombinant Sp1 to Gs and Cx43 promoter sequences, there was no apparent cultured myometrial cell Sp binding to these sequences in EMSAs. It is unlikely that this was due to higher affinity of Sp-like proteins for the probe DNA preventing binding of Sp, as even in the absence of Sp-like protein binding, such as with the Sp1 consensus sequence, Sp bandshifts were not observed. Alternatively, it is possible that the Sp identified in myometrial cells in culture had a reduced ability to bind to their DNA target sites. This could be due to interaction with inhibitory proteins, such as the 20- and 74-kDa proteins that have been found to inhibit DNA binding and trans-activation by Sp1 (25, 26), or could be due to physical modulation of the Sp themselves, such as by alteration of levels of phosphorylation or glycosylation. Although phosphorylated Sp moieties were identified, the protein kinase(s) responsible for the phosphorylation was not determined. Whereas phosphorylation of Sp1 by PKA has been shown to stimulate its DNA-binding activity in vitro (7, 8, 9), phosphorylation by DNA-dependent protein kinase does not appear to affect the extent of such binding (6), and phosphorylation by casein kinase II has been shown to decrease the binding activity of Sp1 (27). The cultured myometrial cell nuclear proteins identified in these EMSAs were also apparently phosphorylated, and treating the extracts with a phosphatase greatly reduced the observed DNA binding. Again, the phosphorylating kinase is not known, but in light of the cAMP dependence of the Gs transcriptional regulation described below, PKA is a candidate.
The DNA-binding properties of recombinant Sp1 and the proteins from myometrial cells in culture were different, but had some overlap. Only the G2, G3, and G6 Gs probes and the Cx43 probe formed DNA-protein complexes with both protein types. These probes had in common the sequence CCTCC, as did all of the probes recognized by the proteins from myometrial cells in culture, but this element was also present in the G4 probe, which was not bound in any of the EMSAs. Of the sequences recognized by recombinant Sp1, all except the Cx probe contained a GGGCGG motif, but this motif was also present in the G5 probe that was not bound. This suggests that DNA-protein binding may involve the sequences surrounding these motifs or something other than the primary sequences of the probes. These factors may also influence the relative affinities of the protein moieties in the myometrial nuclear extract for the various probes, as demonstrated by the differences in the relative intensities of the observed complexes I and II seen in Fig. 4A. Likewise, the competition experiment shown in Fig. 4B demonstrated that the affinities of the proteins from myometrial cells in culture for the G2 probe were only matched by those for the Cx sequence, as none of the other competitors could affect G2 binding, although the slight interference seen with the altered G2 sequence G2m as competitor suggested that the alteration had not completely disrupted the relevant binding motif.
Transfection of COS-7, HEK 293, and human myometrial cells with the Gs promoter reporter construct incorporating the six GC boxes studied in the EMSAs detailed above indicated that transcriptional activity of the Gs promoter might be modulated by cAMP. In these experiments addition to the culture medium of agents that promoted an increase in cellular cAMP levels was found to have a concentration-dependent effect on transcriptional activity of the reporter construct. In all cases incubations with forskolin or 8-bromo-cAMP at low concentrations resulted in a biphasic effect, with initially increased transcriptional activity, followed by a reduction to levels less than those in untreated control cells. At higher concentrations of these reagents, the initial rise in transcription was reduced or abolished, resulting in an earlier fall below control levels. A similar effect was observed on treatment of myometrial cells with hCG, where 10 nM resulted in an initial increase in promoter activity, followed by a prolonged decline to control levels, whereas 100 nM eliminated the initial increase in activity, with luciferase activity dropping below control levels within 1 h. This effect is presumably due to stimulation of intracellular levels of cAMP promoted by hCG receptor-coupled Gs activation.
The concentration-dependent response to hCG of Gs transcription in transfected myometrial cells was also duplicated in in vivo measurements of Gs protein levels by Western blotting. Incubation of primary cultures of myometrial cells with 10 nM hCG for 5 h resulted in increased cellular levels of Gs protein relative to those in untreated control cells, but incubation with 100 nM hCG resulted in Gs levels not significantly different from the control levels. The stimulatory effect of low concentrations of hCG on Gs levels was apparently mediated through the action of PKA, as the addition of specific PKA inhibitors blocked the effect, maintaining Gs levels at or below the control levels.
Sp proteins and the cAMP signaling pathway have previously been associated with differential regulation of genes in response to a range of stimulus concentrations. In the retinoic acid-induced differentiation of mouse F9 teratocarcinoma cells, Gs expression is stimulated, but although low concentrations of cAMP analogs and forskolin had little effect on this process, high concentrations caused inhibition of Gs expression (28). The human endothelial nitric oxide synthase gene was induced by Sp1, but coexpression of an amino-terminal deletion variant of Sp3 corresponding to the 78- and 80-kDa Sp3 isoforms modulated this effect in a biphasic pattern. Low concentrations of the Sp3 variant decreased the Sp1-stimulated expression of endothelial nitric oxide synthase, but high concentrations resulted in a synergistic stimulation to approximately 5 times the level of promoter activity seen with Sp1 alone (29). In rat ovarian granulosa cells, responses to the induction of cAMP by FSH treatment have been shown to vary with time, in a biphasic manner. Phosphorylation of the CRE-binding protein showed an early increase less than 2 h after treatment, followed by a decrease in phosphorylation after 6 h and then a secondary increase to maximal levels at 48 h (30). An almost identical pattern of response was seen in these cells with the expression of the serum/glucocorticoid inducible-protein kinase (sgk) gene (31). The biphasic induction of sgk was found to depend on the presence of Sp1 and Sp3, which bound to the sgk promoter (31) and also on the action of PKA (32).
The biphasic response of Gs transcriptional activity to cAMP was seen here in three cell types and may be indicative of a common regulatory mechanism. In human myometrial cells this would appear to involve binding of nuclear factors to several sequence elements in the Gs promoter, possibly at CCTCC motifs. This binding requires protein phosphorylation, which may be mediated by the cAMP-dependent PKA. The concentration-dependent pattern of response of Gs expression may also suggest the presence of a possible regulatory feedback loop whereby agonist receptor-stimulated cAMP production initially results in increased Gs transcription to accommodate the stimulus with further AC activity. The resultant increase in intracellular cAMP levels then produces downstream effects on genetic regulation, mediated by PKA, until a threshold cAMP level is reached, at which point Gs transcription is down-regulated, thus reducing AC activity and ensuring cellular cAMP homeostasis.
In conclusion, we provide evidence that expression of Gs in human myometrial cells in culture is regulated by the intracellular cAMP concentration and involves binding of phosphorylated factors that are immunologically unrelated to the Sp1–4 protein family to specific motifs within the promoter region of the gene. To achieve a better understanding of the mechanism by which Gs expression is regulated in vivo in the human myometrium during pregnancy will necessitate the further identification of these proteins.
Acknowledgments
Footnotes
This work was supported by Grant 058020 from the Wellcome Trust.
Abbreviations: AC, Adenylyl cyclase; CRE, cAMP response element; Cx, connexin; EMSA, electrophoretic mobility shift assay; Gs, stimulatory G protein; hCG, human chorionic gonadotropin; PKA, protein kinase A; PKC, protein kinase C; Sp, specificity protein.
Received May 1, 2002.
Accepted August 15, 2002.
References
s in pregnancy. Endocrinology 132:2484–2490s in human myometrium in term and preterm labor: a mechanism for parturition. J Clin Endocrinol Metab 79:1835–1839[Abstract]
gene. Proc Natl Acad Sci USA 85:2081–2085-hydroxylase/17,20 lyase) in human adrenal NCI-H295A cells. Mol Endocrinol 15:1277–1293 expression by retinoic acid and parathyroid hormone-related protein in F9 teratocarcinoma cells. J Biol Chem 265:20081–20084This article has been cited by other articles:
 |
|
 |
|
  N. R. Chapman, I. Smyrnias, D. O. C. Anumba, G. N. Europe-Finner, and S. C. Robson Expression of the GTP-Binding Protein (G{alpha}s) Is Repressed by the Nuclear Factor {kappa}B RelA Subunit in Human Myometrium Endocrinology, November 1, 2005; 146(11): 4994 - 5002. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M S Fernandes, V Pierron, D Michalovich, S Astle, S Thornton, H Peltoketo, E W-F Lam, B Gellersen, I Huhtaniemi, J Allen, et al. Regulated expression of putative membrane progestin receptor homologues in human endometrium and gestational tissues J. Endocrinol., October 1, 2005; 187(1): 89 - 101. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Soloff, Y.-J. Jeng, M. Ilies, S. L. Soloff, M. G. Izban, T. G. Wood, N. I. Panova, G. V.N. Velagaleti, and G. D. Anderson Immortalization and characterization of human myometrial cells from term-pregnant patients using a telomerase expression vector Mol. Hum. Reprod., September 1, 2004; 10(9): 685 - 695. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-H. Cheng, B. D. Richardson, M. A. Hubert, and S. Handwerger Isolation and Characterization of the Human Syncytin Gene Promoter Biol Reprod, March 1, 2004; 70(3): 694 - 701. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
, Untreated control; 
, 50 µM forskolin; 
, 100 µM forskolin; •, 200 µM forskolin. B, Transfected COS-7 cells were treated with 8-bromo-cAMP, and luciferase activity was measured in intact cells at intervals after treatment by addition of luciferin to the culture medium. 
