| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 281, Issue 25, 16951-16961, June 23, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1
2
From the
Department of Medicine and Cancer Center, University of California at San Diego, La Jolla, California 92093 and the Section of Gastroenterology, Boston University School of Medicine, Boston, Massachusetts 02118
Received for publication, March 6, 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
and 
, differ in their N-terminal 100 amino acids and are splice variants of the two most 5' exons of the PKG I gene, which appear to be flanked by separate GC-rich promoters (6). PKG I is expressed highly in vascular smooth muscle cells (VSMCs), platelets, and cerebellum, whereas PKG I is prevalent in other parts of the central nervous system, the adrenal gland, and uterus (1, 6).
Within a given cell type or tissue, PKG I expression varies greatly, depending on growth conditions. A transient decrease in PKG I mRNA occurs in VSMCs exposed to mitogens (7), and some but not all investigators have observed lower PKG I expression in actively proliferating, subconfluent VSMC cultures compared with postconfluent cultures (8, 9). In early passage VSMCs and cardiomyocytes and in intact blood vessels, NO or cGMP or cAMP analogs decrease PKG I mRNA and protein expression by decreasing transcription; inflammatory cytokines down-regulate PKG I in VSMCs by inducing NO synthase and increasing cGMP production (10–13). After vascular injury in vivo, PKG I expression is transiently reduced in proliferating neointimal VSMCs compared with the normal vessel wall (14–16). During myoblast differentiation, PKG I expression is up-regulated by the transcription factor FoxO1a to orchestrate myoblast fusion (17).
Rho family proteins, including RhoA and Rac1, cycle between an active, GTP-bound state and an inactive, GDP-bound state; their activities are regulated downstream of cell-cell adhesion receptors and various mitogen receptors via guanine nucleotide exchange factors and GTPase-activating proteins (18–21). Rho proteins regulate actin organization, muscle contractility, cell motility, cell cycle progression, and gene expression (19). Active RhoA induces actin polymerization and stress fiber formation; it increases the expression of smooth muscle-specific genes through actin-regulated cooperation between serum response factor (SRF) and transcription factors of the myocardin family (22, 23). RhoA can also repress gene expression (e.g. RhoA down-regulates transcription of the cyclin-dependent kinase inhibitor p21Waf1/Cip1 and inhibits cytokine-induced transcription of the inducible NO synthase gene), but mechanism(s) of transcriptional inhibition by RhoA remain largely unknown (24–27).
PKG I plays a major role in vascular homeostasis and determines vascular responses to NO. Activation of PKG I mediates NO-induced smooth muscle relaxation and inhibition of VSMC proliferation and migration (1). However, little is known about the mechanisms regulating PKG I expression in VSMCs. We found that PKG I expression is controlled by RhoA and Rac1 activity and that RhoA regulation of the PKG I promoter is mediated, at least in part, through binding of the Krüppel-like transcription factor KLF4 to Sp1 consensus sites in the proximal promoter.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
promoter (positions –430/+35 relative to the transcription start site (6)) was amplified by PCR using genomic DNA from the leukemic cell line HL-60 as a template. This promoter construct does not include the recently described binding sites for FoxO1a or upstream regulatory factors 1/2 (17, 28). The PCR product was cloned into the pSVOA luciferase vector (29); the pSVOA parent vector was chosen because of low basal luciferase expression. Truncated promoter constructs (–80/+35, –430/+22, and –430/+5) were generated by PCR using appropriate primers. Point mutations in putative Sp1 binding sites were introduced into the full-length promoter using the QuikChange site-directed mutagenesis kit (Stratagene) according to the manufacturer's instructions, with the following oligodeoxynucleotides (oligodNTs; sense strand, underlined letters indicate mutated bases): 5'-CCGCCGCCGCCGCCAAACGAGAAAAAGTTTC-3'(mutated A-site located at +1), 5'-GAAAAAGTTTCGCGGAAAGGCTCAGTGAAAAA-3' (mutated B-site at +22), 5'-GAGGGGGACGAGGGAAAGGGTCTCAGGGGAG-3' (mutated C-site at –294), and 5'-TCAGGGGAGGAAGGAAAGCTCTAATTGGTT-3' (mutated D-site at –272). All PCR products were sequenced.
The bacterial expression vector pGEX-mPAK3 (amino acids 65–137) was from R. Cerione (30), and the vector for KLF5 (IKLF) was from J. B Lingrel (31). Vectors for human KLF4 (GKLF), Rho-related proteins, Rho effectors, and the control plasmid pTK-Gal were described previously (32, 33).
The anti-C-terminal PKG I antibody was from StressGen; an antibody specific for PKG I, antibodies for the Rho isoforms A–C, and antibodies for Sp1 and Sp3 were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Antibodies specific for Rac1 and glucose-6-phosphate dehydrogenase were from Upstate%20Biotechnology">Upstate Biotechnology, Inc., and Sigma, respectively; anti-HA epitope and anti-KLF5 antibodies were from Covance and Orbigen, respectively. Antibodies recognizing human and rat KLF4 and anti-EE epitope antibodies were described previously (34, 35). Calpeptin, jasplakinolide, and latrunculin B were from Calbiochem, and 8-para-chlorophenylthio-cGMP (8-CPT-cGMP) was from Biolog.
Cell Culture, DNA Transfection, and Reporter Gene Assays—CS54 rat pulmonary artery smooth muscle cells were from A. Rothman (36), primary bovine aortic smooth muscle cells were from Clonetics (used at passages 3–5), and REF52 rat embryonal fibroblasts were from J. Feramisco. All cells, including HEK 293 cells, were routinely cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (33). In DNA transfection experiments, cells were grown to 
70% confluence and transfected using LipofectamineTM Plus (Invitrogen); cells were harvested after 24 h in serum-containing medium, and firefly luciferase and -galactosidase activities were measured as described previously (33).
siRNA Transfection—An siRNA for green fluorescent protein (GFP) was from Dharmacon. For RhoA siRNA, we used the oligoribonucleotides 5'-GAAGUCAAGCAUUUCUGUCTT-3' and 5'-GACAGAAAUGCUUGACUUCTT-3', which were produced by Qiagen. The sequence for KLF4 siRNA has been described previously (37). Each pair of oligoribonucleotides was annealed at a concentration of 20 µM and introduced into cells by transfection with either OligofectamineTM or LipofectamineTM 2000 (Invitrogen) according to the manufacturer's protocol.
Western Blots and PKG Activity Assay—SDS-PAGE and Western blotting were performed as described (33). Western blots were developed using horseradish peroxidase-coupled secondary antibodies and enhanced chemiluminescence. PKG activity was measured in cell extracts as the difference between Leu-Arg-Arg-Ala-Ser-Leu-Gly (Kemptide) phosphorylation in the presence and absence of 3 µM 8-bromo-cGMP; assays were performed in the presence of 100 µM [
-32PO4]ATP, 300 µM Kemptide, and a 0.3 µM concentration of the specific protein kinase A inhibitor PKI (33).
Quantitative Reverse Transcription-PCR (RT-PCR)—Total cytoplasmic RNA was extracted and subjected to reverse transcription with oligo(dT) primers (35). For the semiquantitative RT-PCR specific for PKG I shown in Fig. 1B, the first strand cDNA was amplified with intron-spanning primers generating a 486-bp product as described (35). Primers specific for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) served as a control for equal mRNA loading. Quantitative RT-PCR was performed using an Mx3000 real time PCR detection system (Stratagene) using IQTM SYBR Green Supermix (Bio-Rad) according to the manufacturer's protocol with a 0.2 µM concentration of the following primers: 5'-GCGTTCCGGAAGTTCACTAA-3' (PKG I, sense) and 5'-TTGATGATGCAGCTCTCCTTC-3' (PKG I, anti-sense); 5'-ACTAACCGTTGGCGAGAGGAAC-3' (KLF4, sense) and 5'-TGGGATAGCGAGTTGGAAAGG-3' (KLF4, antisense); 5'-CGTGGTTCACACCCATCACAAAC-3' (GAPDH, sense) and 5'-GCAAGTTCAACGGCACAGTCAAG-3' (GAPDH, antisense). DNA was denaturated at 95 °C for 30 s, with annealing and extension occurring at 60 °C for 1 min; each primer pair generated a single product as determined by analyzing melting curves after a 40-cycle control reaction. Standard curves were generated by plotting Ct values versus the amount of input RNA for each primer pair and demonstrated similar amplification efficiencies. Relative changes in mRNA expression were analyzed using the 2–
Ct method (38), with GAPDH serving as an internal reference to correct for differences in RNA extraction or reverse transcription efficiencies.
Measurement of Rho and Rac Activity—The activation state of Rho was measured using the Rho binding domain of rhotekin (rhotekin-RBD) to isolate Rho·GTP; the amount of RhoA·GTP or RhoB·GTP bound to rhotekin-RBD-coated beads was assessed by Western blotting with RhoA- or RhoB-specific antibodies, respectively (39). The activation state of Rac was measured using the Rac/CDC42-binding domain of murine p21-activated kinase-1 (PAK-RBD) and quantitating the isolated Rac·GTP by Western blotting with a Rac1-specific antibody (30, 40). Cells were extracted in ice-cold lysis buffer containing 2% Nonidet P-40, 10 mM MgCl2, and protease and phosphatase inhibitors (40, 41). Lysates were cleared by centrifugation at 10,000 x g for 2 min and were diluted to 1 mg/ml protein concentration. To 0.8 mg of cell lysate protein was added either 20 µg of rhotekin-RBD or 10 µg of PAK-RBD bound to glutathione-agarose (via glutathione S-transferase tag), and the mixture was incubated with gentle rocking at 4 °C for 60 min. Beads were washed with lysis buffer four times and eluted in SDS-sample buffer for Western blot analysis (41).
EMSAs—Nuclear extracts were incubated with 5'-end-labeled double-stranded oligodNT probes and analyzed by nondenaturing PAGE/autoradiography; for supershift experiments, 10 µg of nuclear extract protein were preincubated for 30 min with 1 µg of the indicated antibody (35). OligodNTs corresponding to –3 to +29 relative to the transcription start site of PKG I (5'-CCGCCGCCCGAGAAAAAGTTTCGCGGAGGGGC-3' for upper strand, with putative Sp1 sites A and B in boldface type), were synthesized, annealed, gel-purified, and used as the probe designated Sp1(AB). The Sp1 consensus site oligodNT (5'-ATTCGATCGGGGCGGGGCGAGC-3') was from Promega, and the oligodNT corresponding to the human V promoter initiator (Inr) sequence was described previously (35).
Statistical Analysis—The data in bar graphs represent the mean ± S.D. of at least three independent experiments performed in duplicate. Blots, autoradiographs, or photographs represent typical experiments reproduced at least three times with similar results. Statistical analyses were performed using a Student's t test, with a two-tailed value of p < 0.05 considered significant.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
; using an isotype-specific antibody, we were able to detect low levels of PKG I in CS54 cells, with minimal differences between subconfluent and postconfluent cells (data not shown). Similar to the increased PKG I protein, we found an increase in PKG activity from 413 ± 39 pmol/min/mg in subconfluent cells to 816 ± 104 pmol/min/mg protein in postconfluent CS54 cells (n = 3). Cornwell et al. (8) reported comparable PKG activities and differences between subconfluent and postconfluent cultures in primary rat aortic VSMCs.
Semiquantitative RT-PCR with PKG I-specific primers showed about 3-fold higher PKG I mRNA levels in postconfluent CS54 cells compared with subconfluent cells, whereas GAPDH mRNA levels were similar under both conditions (Fig. 1B). To quantitate PKG I mRNA expression more accurately, we used fluorescence-based, real time RT-PCR and measured PKG I relative to GAPDH expression. In CS54 cells and BASMCs, relative PKG I mRNA levels were 2–3-fold higher in postconfluent cells compared with subconfluent cells; in REF52 cells, PKG I mRNA levels increased 10-fold when cells went form a subconfluent to a postconfluent state (Fig. 1C).
To determine whether the observed differences in PKG I expression affected cGMP signaling in VSMCs, we examined phosphorylation of vasodilator-stimulated phosphoprotein (VASP), a physiologically important substrate of PKG I in VSMCs (42, 43). When subconfluent or postconfluent CS54 cells were treated for 10 min with increasing concentrations of a membrane-permeable cGMP analog, half-maximal VASP phosphorylation on Ser239, the preferred PKG I phosphorylation site, occurred at about 3 µM 8-CPT-cGMP in postconfluent cells, whereas in subconfluent cells, it required >10 µM (Fig. 1D). The left upper panel shows a Western blot developed with a VASP phospho-Ser239-specific antibody; cGMP concentrations of >10 µM are not shown. Note that additional phosphorylation of VASP on Ser157 induces a gel shift with the appearance of a lower mobility species (42). Equal loading was demonstrated by blotting with an anti-tubulin antibody (Fig. 1D, left lower panel). When CS54 cells were incubated with 10 µM 8-CPT-cGMP, postconfluent cells demonstrated faster kinetics of VASP Ser239 phosphorylation compared with subconfluent cells (Fig. 1D, right upper panel). Thus, increased PKG I expression in postconfluent VSMCs led to faster VASP phosphorylation at lower cGMP concentrations, indicating enhanced cGMP signal transduction efficiency.
|
In contrast to RhoA and B, Rac1·GTP was higher in postconfluent cells compared with subconfluent cells (Fig. 1F, top), with similar amounts of total Rac1 found at both densities (bottom). Thus, cell density-dependent changes in Rac1 activation correlated with changes in PKG I expression, whereas RhoA and -B activation showed an inverse correlation.
Regulation of the PKG I Promoter by RhoA and Rac1—Since Rho-related proteins regulate gene transcription, we examined the effect of RhoA and Rac1 on PKG I promoter activity. Sequences from –430 to +35 relative to the transcription start site of PKG I were cloned upstream of a firefly luciferase reporter gene to generate the construct pSVOA-PKG(–430/+35). When transfected into subconfluent CS54 cells, the construct produced about 500-fold higher luciferase activity than the promoterless parent construct pSVOA (Fig. 2A). Co-transfection of an expression vector encoding constitutively active RhoA (RhoAV14) inhibited transcription from the PKG promoter by 44 ± 15%, whereas co-transfection of a dominant negative RhoA (RhoAN19) increased transcription by 51 ± 10% (Fig. 2A; p < 0.05 for the comparison between luciferase activity in the presence of RhoA vectors versus empty vector). In contrast, co-transfection of a constitutively active Rac (Rac1V12) increased promoter activity by 2.6-fold, whereas a dominant negative Rac (Rac1N17) inhibited promoter activity by 62 ± 11% (p < 0.05; Fig. 2A). The RhoA and Rac1 constructs had no significant effect on the promoterless luciferase parent vector (data not shown). The effect of constitutively active RhoBV14 on the PKG I promoter was comparable with the effect of RhoAV14. Similar results were obtained in REF52 cells (not shown).
Expression of the RhoA and Rac1 constructs was analyzed by Western blotting (Fig. 2B, RhoA (top) and Rac1 (bottom); the epitope-tagged proteins migrate with a higher apparent molecular weight compared with the endogenous proteins) (41). Taking into consideration a transfection efficiency of 25%, epitope-tagged Rac1 was expressed at levels similar to the endogenous protein, whereas epitope-tagged RhoA was expressed at higher levels. The effect of Rac1 on the PKG I promoter may be more potent than the effect of RhoA, because, in addition to its own effect, Rac1 can modulate RhoA; active Rac1 down-regulates RhoA activity, whereas down-regulation of Rac1 can increase RhoA activity (44, 45). To exclude the possibility that the epitope tag might influence the results, we tested both EE and HA epitope-tagged Rac1V12; both constructs expressed similar levels of Rac1V12 protein and activated the PKG I promoter to the same degree (not shown).
Regulation of the PKG I Promoter by RhoA Effectors—Since transcriptional effects of RhoA have been studied in much greater detail than those of Rac1 and since Rac1 activation can change RhoA activity (44–46), we decided to focus our subsequent work on analyzing the effects of RhoA on the PKG I promoter.
Transcriptional effects of RhoA appear to be mediated by multiple effector pathways, including Rho-kinase (ROK), and the protein kinase C-related kinases PKN and PKR2, but the contribution of each pathway may vary depending on the cell type (33, 47). We previously showed that all three RhoA effectors can trans-activate an SRF-dependent promoter, with PKN and PRK2 being most active in VSMCs but ROK having little effect (33). In contrast, we found that luciferase expression from the PKG I promoter was inhibited by constitutively active ROK and PKN in CS54 cells; the degree of inhibition was similar to the inhibition seen with RhoAV14 (Fig. 2C). A constitutively active PRK2 had no effect on PKG I promoter activity (Fig. 2C). Fig. 2D shows that all three constructs were expressed in CS54 cells.
Effect of Calpeptin-induced RhoA Activation on PKG I Expression—In REF52 cells, inhibition of the protein tyrosine phosphatase Shp-2 by calpeptin selectively activates RhoA without associated changes in Rac or CDC42 activity (48). We used this system to determine whether activation of endogenous RhoA could suppress PKG I expression. When REF52 cells were serum-starved for 48 h and treated with calpeptin for 2 h, RhoA·GTP levels increased about 3-fold (Fig. 3A, top panel; the second panel shows that calpeptin had no effect on total RhoA levels). Phalloidin staining of F-actin showed increased stress fibers consistent with RhoA activation in calpeptin-treated cells (supplemental Fig. 1A), as found by Schoenwaelder et al. (48). In the calpeptin-treated cells, PKG I protein decreased by >50%, whereas G6PDH expression was not significantly altered (Fig. 3A, bottom two panels). Calpeptin inhibited luciferase expression from the PKG I promoter but had no significant effect on the promoterless pSVOA parent vector or on a co-transfected -galactosidase construct (Fig. 3B). Calpeptin had no effect on cell viability in REF52 cells, but it was toxic to CS54 cells; therefore, its effect on PKG expression could not be studied in CS54 cells. In conclusion, RhoA activation by calpeptin suppressed PKG I expression.
Effect of siRNA-mediated Down-regulation of RhoA on PKG I Expression—To determine whether reducing RhoA levels in CS54 cells could modulate PKG I expression, we used an siRNA approach. CS54 cells were plated at low or high densities and transfected with siRNA oligoribonucleotides specific for RhoA or GFP, the latter serving as a control. After 48 h, cells were harvested and examined by Western blotting for expression of PKG I and RhoA (Fig. 3C). The RhoA-specific siRNA decreased total RhoA protein by >95% in subconfluent cells and by >50% in postconfluent cells (Fig. 3C, middle). The silencing of RhoA expression in subconfluent cultures increased PKG I expression to a level similar to that found in postconfluent cells (Fig. 3C, top). In post-confluent cells, PKG I expression could not be further increased by the RhoA-specific siRNA, probably because siRNA delivery was not as efficient as in subconfluent cells (Fig. 3C, middle) and because RhoA activity was already very low in these cells (Fig. 1E). Therefore, the change in RhoA·GTP levels induced in postconfluent cells was probably not sufficient to cause a detectable increase in the already high PKG I expression. Equal protein loading was assessed by Western blotting for G6PDH, which also demonstrated that G6PDH expression was not influenced by the RhoA-specific siRNA (Fig. 3C, bottom). Thus, siRNA-mediated down-regulation of RhoA increased PKG I expression in subconfluent cells with high RhoA activity.
|
promoter (49). To determine whether these sites are required for RhoA-mediated inhibition of the PKG I promoter, we generated a series of truncations and site-directed mutations of the (–430/+35) promoter (Fig. 4, A and B, show schematic representations of the constructs). CS54 cells transfected with a truncated (–430/+22) promoter construct produced about 30% less luciferase activity compared with the full-length promoter, but reporter gene activity was inhibited by active RhoAV14 to the same degree (Fig. 4A, compare filled and striped bars, representing promoter activity in the presence and absence of RhoAV14, respectively). This truncation removed a putative AP1 consensus sequence and the Sp1-B site, suggesting that they are not required for the RhoA effect. Further truncation of the promoter to +5 removed both Sp1-A and -B sites and reduced promoter activity by 75% compared with the (–430/+35) construct; this construct was no longer inhibited by RhoA. These results suggest that the Sp1-A site is important for RhoA-mediated transcriptional suppression but do not rule out that other sequences could contribute to the RhoA effect.
|
promoter abolished transcriptional inhibition by RhoAV14 (Fig. 4B; compare –430/+35 Sp1(Amut), Sp1(Bmut), and Sp1(ABmut) with the parental –430/+35 construct). Mutation of site A reduced basal promoter activity by about 40%, whereas mutation of site B had no effect on basal promoter activity. These results suggest that both Sp1-A and -B sites cooperate in mediating the effect of RhoA; mutation of the B site appears to interfere with the function of the A site, whereas deletion of the B site did not have this effect (Fig. 4, compare A and B). Site-directed mutation of two additional putative Sp1 binding sites at –294 and –272 (Sp1-C and -D sites) reduced basal activity of the PKG I promoter by about 40% and diminished, but did not abolish, the effect of RhoAV14, suggesting that these two sites may contribute to the negative transcriptional effects of RhoA.
Cell Density-regulated Binding of Nuclear Proteins to the PKG I Promoter Sp1(AB) Site—Sp1- and Krüppel-like factors (KLFs) comprise a family of highly related zinc finger proteins that regulate a large number of promoters through interaction with GC-rich elements (50). We performed electrophoretic mobility shift assays (EMSAs) to determine whether cell density-induced changes in PKG I expression correlated with changes in specific trans-activating factors binding to the PKG I promoter. Using an oligodNT probe corresponding to nucleotides –3 to +29 of the PKG I promoter, which contains the Sp1-A and -B sites important for RhoA regulation of the promoter (Fig. 4), we observed one major protein-DNA complex (C1) with nuclear extracts from CS54 cells. Significantly more probe was shifted by extracts from postconfluent cells compared with subconfluent cells (Fig. 5A, compare lanes 3 and 4, probe incubated with extracts from subconfluent versus postconfluent cells). The complex was competed by a 50-fold excess of unlabeled probe (lanes 2 and 5); it was also competed by a 50-fold excess of unlabeled Sp1 consensus sequence oligodNT (lane 6). Protein binding was not competed by the same amount of unrelated oligodNTs encoding NF-
B or AP1 consensus sequences (data not shown). Similar results were obtained with a probe encoding nucleotides –3 to +21 of the PKG I promoter, corresponding to the isolated Sp1-A site. Equal loading and quality of nuclear extracts was demonstrated with a probe encoding the Inr sequence of the human V promoter, which produces multiple protein-DNA complexes unrelated to Sp1 and Krüppel-like factors (35) (Fig. 4A, lanes 9 and 10, with competition shown in lanes 8 and 11).
To identify proteins bound to the Sp1(AB) probe, we preincubated nuclear extracts of postconfluent CS54 cells with antibodies specific for various members of the Sp1/KLF transcription factor family prior to performing EMSAs (Fig. 4B, lanes 1–7). EMSAs with an Sp1 consensus oligodNT probe served as controls (lanes 8–14). With the Sp1 consensus oligodNT probe, four specific protein-DNA complexes were formed, which were all competed by a 50-fold excess of unlabeled probe (lane 10) or the PKG promoter Sp1(AB) sequence (lane 11). The Sp1(AB) sequence competed most efficiently with the fastest migrating complex (arrow), which co-migrated with the major complex C1 formed by the Sp1(AB) probe. Preincubation of nuclear extracts with antibodies specific for Sp1 or Sp3 did not detectably alter protein binding to the Sp1(AB) probe (lanes 5 and 6) but inhibited formation of specific protein-DNA complexes in EMSAs with the Sp1 consensus oligodNT (lanes 12 and 13), suggesting that the latter contained Sp1 and Sp3 protein, respectively. Control IgG was without effect (lanes 7 and 14). Thus, Sp1 and Sp3 did not appear to contribute to the main protein-DNA complex C1 formed by the PKG promoter Sp1(AB) probe; however, competition experiments suggested that C1 contained protein(s) related to the Sp1/KLF family.
|
|
|
When we examined KLF4 protein levels in subconfluent and postconfluent CS54 and REF52 cells, we found similar amounts of KLF4 protein under both conditions (Fig. 6C, lanes 1 and 2 show results for subconfluent and postconfluent CS54 cells; lane 3 shows KLF4-transfected cells for comparison). Nuclear extracts from post- and subconfluent CS54 cells contained the same amount of KLF4 protein, but the former produced significantly more KLF4-containing protein-DNA complexes with the PKG I promoter probe compared with the latter (Fig. 5A). These results suggest that KLF4 DNA binding activity is regulated by cell density and are consistent with the finding that active RhoA reduces KLF4 DNA binding (Fig. 6B).
RhoA regulates the trans-activation potential of SRF through changes in actin dynamics; activation of SRF requires RhoA-induced actin polymerization, which leads to the recruitment of monomeric (G)-actin-binding SRF cofactors (56). To determine whether RhoA regulation of KLF4 may also involve changes in actin polymerization, we used the actin-binding drugs jasplakinolide and latrunculin B. Jasplakinolide stabilizes filamentous (F)-actin (57) and induces prominent actin stress fibers in CS54 cells (supplemental Fig. 1B). In postconfluent CS54 cells, jasplakinolide treatment decreased binding of nuclear extracts to the PKG I promoter Sp1(AB) site (Fig. 6D, top, compare lanes 1 and 2). This is consistent with the negative effect of RhoAV14 on KLF4 DNA binding (Fig. 6B). In contrast, latrunculin B sequesters G-actin monomers (57), thereby preventing actin polymerization and reducing stress fibers in CS54 cells (supplemental Fig. 1B). When subconfluent CS54 cells were treated with latrunculin B, binding of nuclear proteins to the PKG promoter Sp1(AB) site was increased (Fig. 6D, top, compare lanes 3 and 4; the bottom shows complexes formed with an Inr probe and demonstrates equal loading). These experiments suggest that RhoA may regulate KLF4 DNA binding through regulation of actin dynamics. Increased actin polymerization induced by jasplakinolide or RhoAV14 suppressed KLF4 DNA binding, whereas decreased actin polymerization in latrunculin B-treated cells had the opposite effect.
KLF4 Trans-activation of the PKG I Promoter—Next, we examined the effect of KLF4 overexpression on PKG I promoter activity. Transfection of a KLF4 expression vector into CS54 cells increased luciferase expression from the promoter by 2.3 ± 0.5-fold, but mutations in either the Sp1-A or Sp1-A/B sites rendered the promoter unresponsive to KLF4, and a mutation in the Sp1-B site greatly reduced responsiveness (Fig. 7A, p < 0.05 for the effect of KLF4 on the wild type promoter). The modest effect of KLF4 on the wild type PKG I promoter was not increased by transfecting higher amounts of KLF4 vector DNA, possibly because of nearly saturating amounts of endogenous KLF4. This led us to examine the effect of KLF4 in HEK 293 cells, which express very low levels of endogenous KLF4, possibly allowing for larger effects of the transfected protein (54). In 293 cells, KLF4 was able to transactivate the PKG I promoter 6-fold (Fig. 7B). Co-transfection of active RhoAV14 decreased KLF4-stimulated PKG I promoter activity by about 60% (Fig. 7B; RhoAV14 had no effect on KLF4 expression levels). In contrast to KLF4, transfection of KLF5 had no significant effect on PKG I promoter activity (Fig. 7C, top; KLF5 expression is shown at the bottom). These results indicate that KLF4 binds to the proximal PKG I promoter and trans-activates the promoter through the Sp1(AB) site and that KLF4 DNA binding and trans-activation potential are regulated by RhoA.
To determine whether KLF4 levels affect expression of endogenous PKG I mRNA in VSMCs, we performed siRNA knockdown experiments. CS54 cells were transfected twice with either a KLF4-specific or a control (GFP-specific) siRNA and were harvested 48 h after reaching confluence. Quantitative RT-PCR showed that KLF4 mRNA levels were reduced by 68 ± 3% in cells treated with KLF4 siRNA; this was associated with a significant decrease in nuclear extract binding to the PKG I promoter Sp1(AB) site (Fig. 7D; the upper panel shows KLF4 mRNA expression relative to GAPDH mRNA expression; the lower panels show EMSAs performed with either a Sp1(AB) probe (complex C1) or an Inr probe). We found that PKG I mRNA levels were reduced by 46 ± 10% (p < 0.05) in cells treated with KLF4 siRNA (Fig. 7D, right columns). Similar but less dramatic effects were observed with a second KLF4-specific siRNA (not shown). Thus, reduction of KLF4 expression in VSMCs reduced PKG I expression.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
promoter. We observed lower RhoA·GTP and higher Rac1·GTP levels in postconfluent cultures of VSMCs compared with subconfluent cultures; similar differences in RhoA and Rac1 activation have been observed in postconfluent versus subconfluent epithelial and endothelial cells (58–61). In epithelial cells, cadherin engagement (by cell-cell contact or by cadherin antibody cross-linking) decreases RhoA activity through activation of p190 Rho-GAP, and interaction of cadherin with p120 catenin leads to Rac1 activation through increased Rac guanine nucleotide exchange activity (59, 61–63). Our results suggest that the activation states of RhoA and Rac1 can explain cell density-dependent changes in PKG I expression and thus identify a novel mechanism for cell density-dependent regulation of gene expression. Moreover, we showed that the difference in PKG I expression between subconfluent and postconfluent cells was sufficient to affect cGMP signaling efficiency.
Experiments with calpeptin and RhoA siRNA demonstrated that endogenous RhoA can suppress PKG I expression. Among several RhoA effectors tested, constitutively active ROK and PKN were able to mimic the suppressive effect of constitutively active RhoAV14 on the PKG I promoter, but PRK2 was not. Preliminary results suggest that pharmacological inhibition of ROK is not sufficient to increase PKG I expression in CS54 cells.4 RhoA-mediated activation of SRF-dependent transcription has been studied in considerable detail over the past decade (22, 56, 57, 64), but much less is known about RhoA-mediated suppression of gene transcription. The present work adds PKG I to a relatively short list of genes negatively regulated by RhoA, which includes the cyclin-dependent kinase inhibitor p21Waf1/Cip1, inducible NO synthase, the scavenger receptor CD36, and RhoB (24–27, 65, 66).
Among the promoters negatively regulated by RhoA, only the p21Cip1/Waf1 promoter, which contains six Sp1 binding sites targeted by different stimuli, has been studied in detail (25). The cyclin-dependent kinase inhibitor p21Cip1/Waf1 is a growth arrest gene, and similar to PKG I, p21Cip1/Waf1 expression increases when cells are contact-inhibited by high density (67). RhoA suppression of p21Cip1/Waf1 requires several Sp1 consensus sequences, which bind Sp1, Sp3, and several KLF family members; however, the mechanism of RhoA regulation of Sp1/KLF transcription factors is not known (25, 54, 68). Constitutively active RhoA suppresses the p21Cip1/Waf1 promoter to a similar degree as the PKG I promoter (25). Whereas the effect of RhoA on the two promoters may appear modest, the RhoA-induced change in p21Cip1/Waf1 protein expression affects cell growth (24), and the RhoA/Rac effect on PKG I expression is sufficient to affect cGMP signaling. The human PKG I promoter contains four putative Sp1 binding sites, and we found the Sp1(A) site at the transcription start to be most important, with the other sites contributing to maximal transcriptional effects of RhoA. The sequence of the Sp1(A) site resembles the major Rho-regulated Sp1 site of the p21Cip1/Waf1 promoter, whereas the other putative Sp1 sites in the PKG I promoter have slightly different sequences (6, 68).
Using a probe corresponding to the combined Sp1(AB) site in the PKG I promoter, we observed a single major protein-DNA complex that contained KLF4; abundance of the complex was increased by overexpression of KLF4 and reduced by siRNA-mediated knockdown of the transcription factor. The complex did not contain detectable amounts of Sp1 or Sp3, although on long autoradiograph exposures, a separate, low abundance complex was detectable that co-migrated with the Sp1-containing complex formed by the Sp1 consensus oligodNT probe (data not shown). Binding of nuclear extract proteins to the Sp1(AB) site was cell density-dependent, but we found no significant difference in KLF4 protein expression between subconfluent and postconfluent CS54 cells, suggesting that KLF4 DNA binding activity is regulated by cell density. Correspondingly, active RhoAV14 suppressed KLF4 DNA binding activity at the PKG I promoter without altering KLF4 protein levels. KLF4 DNA binding may be regulated by post-translational modification(s) or by cooperation with other proteins.
KLF4 can serve as an activator or inhibitor of transcription, depending on the promoter context and/or cooperation with other transcription factors (50, 69). For example, KLF4 transactivates the iNOS promoter in cooperation with p65 (RelA) and the p21Cip1/Waf1 promoter in cooperation with p53 (54, 70), but it directly suppresses the p53 promoter and inhibits ornithine decarboxylase promoter activity by competing with Sp1 (32, 69). We found that KLF4 activation (as well as RhoA suppression) of the PKG I promoter required the Sp1(AB) site and that DNA binding and trans-activation potential of KLF4 was inhibited by RhoAV14. Separate reports describe that p21Cip1/Waf1 and iNOS promoter activities are activated by KLF4 (54, 70) and suppressed by RhoA (25, 26), but to our knowledge, this work is the first to demonstrate RhoA regulation of KLF4. Experiments with jaspakinolide and latrunculin B indicate that KLF4 DNA binding activity is regulated by changes in actin dynamics and suggest that RhoA may suppress KLF4 DNA binding by inducing actin polymerization. Determination of whether KLF4 is directly affected by changes in the F-actin/G-actin ratio or whether actin monomers regulate additional protein(s) that cooperate with KLF4 requires further study. In the case of RhoA regulation of SRF, G-actin binds to and controls the activity of the SRF coactivator MAL (56). Analogously, G-actin may regulate the activity of a KLF4 coactivator or corepressor.
KLF4 trans-activation of the PKG I promoter was similar in magnitude to its effect on the p21Cip1/Waf1 promoter (54). In postconfluent CS54 cells, siRNA-mediated KLF4 knockdown significantly decreased PKG I mRNA expression. KLF4 knock-out mice die shortly after birth due to severe defects in skin barrier function, making analysis of cGMP/PKG I signaling in knock-out VSMCs difficult; however, it would be interesting to compare PKG I expression levels between wild type and KLF4–/– tissues (71). Based on our results, we conclude that KLF4 is a major trans-acting factor binding to the PKG I promoter Sp1(AB) site, although we cannot rule out cooperation of KLF4 with other transacting factors.
RhoA signaling pathways in VSMCs are abnormally activated in various forms of hypertension, and RhoA activation is implicated in increased vascular resistance (72–75). It is tempting to speculate that increased RhoA activation may explain the decreased PKG I expression found in VSMCs of spontaneously hypertensive rats and that low PKG levels may contribute to impaired NO/cGMP-dependent VSMC relaxation in these animals (18, 72, 73). It has been suggested that RhoA is activated by peptide hormones, such as angiotensin II and thrombin, in proliferating, dedifferentiated VSMCs found in the neointima of injured vessels (72, 75), and several investigators found decreased PKG I expression in neointimal cells compared with normal vessel wall (14–16). Thus, RhoA regulation of PKG I expression may be of major physiological importance and explain PKG changes in VSMCs in hypertension and vascular injury.
|
FOOTNOTES |
|---|
The on-line version of this article (available at http://www.jbc.org) contains supplemental Fig. 1.
1 Supported by the Deutsche Forschungsgemeinschaft.
2 To whom correspondence should be addressed: Dept. of Medicine, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA 92093-0652. Tel.: 858-534-8805; Fax: 848-534-1421; E-mail: rpilz{at}ucsd.edu.
3 The abbreviations used are: NO, nitric oxide; BASMCs, bovine aortic smooth muscle cells; 8-CPT-cGMP, 8-para-chlorophenylthio-cGMP; EMSA, electrophoretic mobility shift assay; GFP, green fluorescent protein; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; G6PDH, glucose-6-phosphate dehydrogenase; KLF, Krüppel-like factor; oligodNT, oligodeoxyribonucleotide; PKG, cGMP-dependent protein kinase; PKN, protein kinase N; PRK, protein kinase C-related kinase; RBD, Rho/Rac-binding domain; ROK, Rho-kinase; RT, reverse transcription; SRF, serum response factor; VASP, vasodilator-stimulated phosphoprotein; VSMCs, vascular smooth muscle cells; siRNA, small interfering RNA; Inr, initiator.
4 Y. Zeng and R. B. Pilz, unpublished observation.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
