|Home | About | Journals | Submit | Contact Us | Français|
Mitogen-activated protein kinase (MAPK) cascades regulate a wide variety of cellular processes that ultimately depend on changes in gene expression. We have found a novel mechanism whereby one of the key MAP3 kinases, Mekk1, regulates transcriptional activity through an interaction with p53. The tumor suppressor protein p53 down-regulates a number of genes, including the gene most frequently mutated in autosomal dominant polycystic kidney disease (PKD1). We have discovered that Mekk1 translocates to the nucleus and acts as a co-repressor with p53 to down-regulate PKD1 transcriptional activity. This repression does not require Mekk1 kinase activity, excluding the need for an Mekk1 phosphorylation cascade. However, this PKD1 repression can also be induced by the stress-pathway stimuli, including TNFα, suggesting that Mekk1 activation induces both JNK-dependent and JNK-independent pathways that target the PKD1 gene. An Mekk1-p53 interaction at the PKD1 promoter suggests a new mechanism by which abnormally elevated stress-pathway stimuli might directly down-regulate the PKD1 gene, possibly causing haploinsufficiency and cyst formation.
Autosomal dominant polycystic kidney disease (ADPKD)3 is an inherited disorder that affects ~1 in 500–1,000 individuals, and accounts for ~1 in 10 cases of end stage kidney failure (1,–7). ADPKD patients develop numerous large fluid-filled cysts from kidney tubules and collecting ducts, and frequently exhibit a variety of non-renal manifestations such as liver or pancreatic cysts, cerebral aneurysms, cardiac developmental abnormalities, or hypertension. Approximately 85% of ADPKD cases are caused by mutations in the PKD1 gene, with the remainder being caused by mutations in the PKD2 gene. PKD1 encodes polycystin-1, a large membrane protein that regulates a number of signaling pathways involved in cell cycle control, cell differentiation, and apoptotic cell death.
PKD1 loss-of-function and/or decreased expression leading to haploinsufficiency have been shown to cause renal cyst formation in ADPKD and Pkd1 mutant mice (8,–13). Because cyst formation may be caused by decreased expression of the PKD1 gene, it is important to identify transcriptional mechanisms that up-regulate gene expression, as they may ameliorate the disease; and that down-regulate gene expression as they may initiate or exacerbate the disease.
We have demonstrated that the human PKD1 gene can be up-regulated by β-catenin via a TCF/LEF consensus motif (14), and by retinoic acid via Sp1-binding GC-box motifs in the PKD1 proximal promoter (15). Conversely, we and others have shown that the PKD1 promoter is negatively regulated by Ets-1/Fli-1 (16) and by the tumor suppressor protein p53 (17). The p53 protein is widely known for its role in cell cycle control, DNA repair, and programmed cell death; and its inactivation by genetic mutation is the most frequent alteration in human cancers (18,–20). p53 regulates transcription of target genes by directly binding DNA elements and functioning as an activator or repressor depending on the target gene (21). The mechanisms of p53 transcriptional repression are not well understood (e.g. Ref. 22).
The mammalian mitogen-activated protein kinases (MAPKs) include at least three subgroups: ERKs (extracellular signal-regulated kinases), p38 MAPKs, and JNKs (the c-Jun N-terminal kinases, also known as stress-activated protein kinases or SAPKs) (23,–25). Each is activated through a phosphorylation cascade initiated by activation of a MAP kinase kinase kinase (MAPKKK or MAP3K), which phosphorylates a MAP kinase kinase (MAPKK), which in turn phosphorylates a MAPK. Following activation, the MAPKs translocate to the nucleus to regulate the activity of transcription factors controlling a wide range of genes. One of the key MAP3K components of the stress-activated JNK pathway is Mekk1. Upon stimulation, Mekk1 phosphorylates Mkk4/Mkk7 (MAPKKs), which then phosphorylate and activate JNK (26,–31). Signaling initiated with the typically membrane-associated Mekk1 ends with activation of the transcription factor AP-1 (activator protein-1), which is a homo- or heterodimer of c-Jun with c-Fos or ATF2; or with other Mekk1-JNK responsive transcription factors including p53. The small GTPases, Ras, Rac, and cdc42, and their downstream effectors JNK and AP-1 have been linked to the PKD renal cystic phenotype and polycystin-1 function (32, 33) suggesting that the Mekk1 pathway may be involved in the regulation of PKD1 expression.
In the present study, we discovered that Mekk1 directly represses transcription of the PKD1 gene. However, Mekk1 kinase activity is not required for this repression, suggesting a novel mechanism. Multiple lines of evidence revealed that the Mekk1 effect is mediated in the nucleus through interaction with the tumor suppressor protein p53 via an atypical p53 DNA binding motif in the PKD1 proximal promoter. A physical association between Mekk1 and p53 has not been reported previously. Mekk1 was found to reduce endogenous mRNA levels for the PKD1, bradykinin receptor B2, and IEX-1 genes, suggesting a general mechanism. As such, these results have identified a new transcriptional mechanism involving the nuclear localization of Mekk1 and its kinase-independent regulation of p53 target gene transcription.
A 3.3-kb human PKD1 promoter fragment (−3,346 to +33) in the promoterless luciferase reporter vector pGL3-Basic (Promega) was originally generated in our lab (14). Promoter deletion constructs (1.3 kb, 2.0 kb, 1.7 kb, 1.0 kb, and 200 bp) cloned in pGL2-Basic and/or pGL3-Basic were previously described (14,–16). The mammalian expression construct for the activated CA-Mekk1 (pFC-MEKK) was from Stratagene. This construct represents the C-terminal end (293 amino acids) of full-length (1,493 amino acids) Mekk1 (23). p53 and dominant-negative p53 mutant were generous gifts of Dr. Samir El-Dahr (Tulane University Health Sciences Center). HCT116 p53−/− cells were a gift from the Vogelstein lab (The Johns Hopkins University).
The Ets and Sp1 binding site mutations in the 200-bp PKD1 proximal promoter region were described earlier (15, 16). Mutations in the p53 binding site were introduced using the QuikChange site-directed mutagenesis kit (Stratagene) and the primer pairs: 5′-GAA GGG GGC GGA taC caa AaC CGC CCC GCC C-3′ and 5′-GGG CGG GGC GGt Ttt gGt aTC CGC CCC CTT C-3′. The lowercase nucleotides represent the mutations. Following verification by sequencing, the 200-bp mutant fragments were subcloned into pGL3.
HEK293T and COS-1 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) with 4.5 g/liter of glucose containing 10% (v/v) heat-inactivated fetal bovine serum (FBS) and antibiotic (100 IU/ml of penicillin, 100 μg/ml of streptomycin). HCT116 p53−/− cells were maintained in McCoy's medium containing 10% FBS and antibiotic, and M-1 cells were grown in DMEM/F-12 supplemented with 5% FBS, antibiotic, 1.5 mm Hepes, and 24 mm NaHCO3. All cells were grown at 37 °C supplied with 5% CO2. Following overnight culturing in six-well plates, the cells were transfected using either the calcium phosphate method (HEK293T and HCT116 cells) (34) or with Lipofectamine (Invitrogen) (COS-1 and M-1 cells) according to the manufacturer's instructions. A β-galactosidase expression plasmid (50 ng) or Renilla luciferase expression vector (10 ng) was included to monitor transfection efficiency and for normalization. After 6 h, the DNA-containing medium was removed, and the cells were incubated with DMEM containing 2% serum and were harvested at 40 h as described earlier (15). In some experiments, transfected cells were incubated with DMSO or inhibitors dissolved in DMSO before harvesting. Luciferase assays were carried out in 1× reporter lysis buffer (Promega) using Luminol (Promega) as a substrate, and β-galactosidase assays were performed using o-nitrophenyl-β-d-galactopyranoside (Sigma) as a substrate. Protein concentrations were determined with the BCA protein assay kit (Pierce). The measured luciferase activity in each sample was normalized to β-galactosidase activity, Renilla luciferase, or total protein. These normalized luciferase activities (RLU) were plotted using Microsoft Excel as average ± S.D. of triplicate samples from typical experiments. Experiments were repeated at least three times. Statistical significance was determined by pairwise comparisons using Student's t test.
Transfection with siRNA was carried out using Lipofectamine 2000 (Invitrogen) following the manufacturer's instructions. Briefly, 200 bp of PKD1 promoter DNA (0.5 μg) with the vector or CA-Mekk1 expression plasmids (0.2 μg of DNA) with 80 pmol of scrambled siRNA, human p53 siRNA, or Mekk1 siRNA duplex in FBS and antibiotic-free medium was incubated, added to cells, and incubated overnight at 37 °C. The medium was then replaced with fresh DMEM containing 2% serum and antibiotics, and the cells were harvested at 40 h. M-1 cells grown in DMEM/F-12 supplemented with 5% FBS, antibiotic, 1.5 mm Hepes, and 24 mm NaHCO3 at 37 °C were transfected with pEGFPN3 vector or kinase-dead CA-Mekk1 (DC) in the same vector using Lipofectamine and selected for neomycin (G418) resistance at a concentration of 400 μg/ml as before (35).
Mekk1−/− and Mekk1+/+ mouse embryonic fibroblast cells were cultured in DMEM with 4.5 g/liter of glucose containing 10% (v/v) heat-inactivated fetal bovine serum (FBS) and antibiotic (100 IU/ml of penicillin, 100 μg/ml of streptomycin). Following overnight plating, cells in growth medium were treated with PMA (0, 0.5, and 1.0 μm), H2O2 (0, 0.1, and 0.2 mm), and TNFα (20 ng/ml) for 6 or 24 h, and harvested for RNA isolation or ChIP analysis as described below.
HEK293T cells were transfected with the control vector, Mekk1, or CA-Mekk1 for 40 h. Total RNA was extracted with TRIzol (Invitrogen) according to the manufacturer's protocols, and the samples were treated with DNase I (Ambion). RT-PCR was carried out with 1 μg of RNA in a total volume of 25 μl using random primers and SuperScript reverse transcriptase (Invitrogen) as described previously (36). Primers specific for the ribosomal protein L7 were: forward, 5′-GCT TCG AAA GGC AAG GAG GAA GC-3′ and reverse, 5′-TCC TCC ATG CAG ATG ATG C-3′ giving a 440-bp product. Primers specific for PKD1 were: forward, 5′-CGC CGC TTC ACT AGC TTC GAC-3′ and reverse, 5′-ACG CTC CAG AGG GAG TCC AC-3′, giving a 260-bp product; primers specific for α-fetoprotein were: forward, 5′-GCT GCA AAC TGA CCA CGC TG-3′ and reverse, 5′-CCA ATA ACT CCT GGT ATC C-3′, giving a 239-bp product; primers specific for bradykinin receptor B2 were: forward, 5′-GCA GCA GAC CTG ATC CTG G-3′ and reverse, 5′-GAT CAC CAA GCT GTA GAG-3′, giving a 243-bp product; and primers specific for IEX-1 were: forward, 5′-GGT GAG TAT CGC CGA AGT G-3′ and reverse, 5′-CTG AGG TCC AGA GCG TAG TC-3′, giving a 339-bp product. PCR of PKD1 and IEX-1 was performed for 30 cycles of 94 °C for 30 s, 62 °C for 45 s, and 72 °C for 45 s, with a final extension of 7 min at 72 °C. α-Fetoprotein and bradykinin receptor were done for 30 cycles of 94 °C for 30 s, 56 °C for 45 s, and 72 °C for 45 s, with a final extension of 7 min at 72 °C.
Amplified PCR fragments were electrophoresed on 2% agarose gels containing ethidium bromide, and the bands were analyzed by NIH Image software. Quantified band intensity was normalized to values for ribosomal protein L7 or L15 mRNA and plotted as relative units. The data were plotted with Microsoft Excel as average ± S.D. of three independent experiments.
Expression of various constructs in cells was confirmed by 10% SDS-PAGE, followed by immunoblotting using the primary antibodies: rabbit anti-p53 (Santa Cruz, sc-6243, 1:1,500), rabbit anti-Mekk1 (Santa Cruz, sc-252, 1:1,500), mouse anti-β-tubulin (Santa Cruz, sc-5286, 1:1,500), rabbit anti-poly(ADP-ribose) polymerase-1 (Santa Cruz, sc-74469, 1:1,500), rabbit anti-β-actin (Sigma, A2066, 1:2,500), and rabbit anti-JNK (Santa Cruz, sc-474, 1:1,500). Alkaline phosphate-conjugated secondary anti-rabbit (Sigma, A8025) or anti-mouse (Sigma, A7434) antibodies were used at 1:10,000 dilution. Following equilibration in chemiluminescence buffer (0.1 m diethanolamine, 1 mm MgCl2, pH 9.5) for 5 min, the membrane was incubated with the substrate CDP-Star (Amersham Biosciences) for 5 min and exposed to film (RPI). For re-probing, the blots were stripped of primary and secondary antibodies using 20 mm Tris-HCl, pH 6.8, containing 1% SDS and 100 mm 2-mercaptoethanol at 55 °C for 60 min with shaking at 125–135 rpm, followed by extensive washing and blocking in 5% nonfat dry milk in TBST for at least 2 h.
Nuclear extracts were prepared from HEK293T or COS-1 cells using NE-PER Nuclear and Cytoplasmic Extraction Reagent Kit (Pierce) following the manufacturer's instructions. 30–60 μg of cytosolic extract (CE) or nuclear extract (NE) was mixed in 400 μl of 20 mm Tris-HCl, pH 8.0, 150 mm NaCl, 1 mm EDTA, 1% Triton X-100 and protease inhibitor mixture and incubated with 5 μl of anti-p53 (Santa Cruz, sc-6243) or anti-Mekk1 (Santa Cruz, sc-252) antibodies on a rotating wheel at 6 °C for 4 h. The antibodies were captured with Protein A/G PLUS-agarose (Santa Cruz, sc-2003), washed, and analyzed by immunoblotting using either anti-Mekk1 or anti-p53 antibodies. AP-conjugated Clean-Blot IP Detection Reagent (Thermo Scientific, 21233) was used as the secondary antibody.
Approximately 1 × 105 cells/chamber were transfected for 40 h with CA-Mekk1, which was tagged with GFP at the C-terminal end. The cells were fixed with 4% formaldehyde (30 min at 4 °C), permeabilized with 0.1% Triton X-100 in PBS, and incubated with anti-p53 (Santa Cruz, sc-6243, 1:300) or anti-Mekk1 (Santa Cruz, sc-252, 1:300) overnight. Following washing three times with PBS, the cells were incubated with FITC-conjugated goat anti-rabbit IgG (Sigma, 1:200) for 1 h. The coverslipped cells were examined under a fluorescence microscope.
Biotin labeling by A-tailing, binding reaction, gel separation, and detection were carried out as follows. 250 pmol of blunt-ended duplex probe in 20 μl of 1× Taq polymerase buffer containing 0.1 mm Biotin-14-dATP and 2.5 units of Taq polymerase was incubated for 30 min at 70 °C. 1 μl of 0.5 m EDTA was added to inactivate polymerase, and the biotinylated probe was ethanol precipitated. Binding reactions were carried out by incubating 25 pmol of biotinylated probe in a total volume of 15 μl containing 1× binding buffer (20 mm Hepes, pH 7.4, 25 mm KCl, 2.5 mm MgCl2, 0.2 mm EDTA, 1 mm DTT, 0.5 mm PMSF, 12.5% glycerol), 1 μg of poly(dI-dC), and 10 μg of nuclear extract, with or without unlabeled annealed primers (25 pmol) at room temperature for 30 min.
Following binding of the biotinylated probes with nuclear extract prepared from HEK293T cells transfected with CA-Mekk1 for 30 min, the reaction mixture was diluted with 400 μl of 1× PBS containing 0.5% Nonidet P-40, 0.5% Triton X-100, 0.1% SDS, 1 mm EDTA and protease inhibitor mixture. Streptavidin-agarose (30 μl) was added followed by rotation on a wheel for 4 h at 6 °C. Agarose-bound streptavidin was collected by centrifugation for 1 min at 16,300 × g. The pulldown product was washed 3 times with 10 mm Tris-HCl, pH 8.0, containing 150 mm NaCl, 0.5% Nonidet P-40, 0.5% Triton X-100, 0.1% SDS, and 1 mm EDTA and boiled for 5 min in SDS sample buffer prior to SDS-PAGE and Western blotting. The membrane was probed simultaneously with anti-p53 (Santa Cruz, sc-6243, 1:1,500) and anti-Mekk1 (Santa Cruz, sc-252, 1:1,500) antibodies. The membrane was stripped and re-probed with anti-c-Jun antibody (Santa Cruz, sc-1694). The bipartite p53 probe (see Fig. 3) is a duplex of primer 5′-GGT CGC GCT GTG GCG AAG GGG GCG GAG CCT GCA C-3′ and its complementary sequence. The nonspecific probe is a duplex of primer 5′-CGG CCG ACT CTG CAG TGC GAC G-3′ and its complementary sequence, which is located 60 bp upstream of the 200-bp PKD1 proximal promoter.
Untransfected or transfected HEK293T cells or HCT116 p53−/− cells in a T-25 flask or p60 plates were cross-linked for 10 min in 1% formaldehyde in PBS at room temperature. Cross-linking was quenched with 0.125 m glycine for 10 min at room temperature. Cells were washed twice with ice-cold PBS, harvested by centrifugation for 2 min at 5,000 × g, and resuspended in 0.5 ml of ChIP lysis buffer (10 mm Tris-HCl, pH 8.0, containing 10 mm NaCl, 0.5% Nonidet P-40, 1 mm EDTA, and protease inhibitors mixture), and incubated on ice for 15 min. Crude nuclear extract collected by centrifugation at 2,500 × g for 5 min was washed once with 0.2 ml of PBS, resuspended in 0.5 ml of ChIP lysis buffer without Nonidet P-40, and sonicated on ice 3 times for 15 s with 30 s cooling (60% efficiency). The sonicated samples were added with 1% Triton X-100, incubated on ice for 5 min, and centrifuged at maximum speed for 5 min. The supernatants were recovered and pre-cleared with PBS-washed Protein A/G for 1 h. The pre-cleared supernatants were divided and immunoprecipitated with 5 μl of rabbit preimmune antisera (Sigma), rabbit anti-RXR antibody (Santa Cruz, sc-553), rabbit anti-p53 (Santa Cruz, sc-6243), and rabbit anti-Mekk1 (Santa Cruz, sc-252) with overnight rotation at 6 °C. Following 3 washing with wash buffer (50 mm Tris-HCl, pH 8.0, containing 0.15 m NaCl, 0.5% Triton X-100, 0.5% Nonidet P-40, 0.1% SDS, 1 mm EDTA), the cross-links were reversed in elution buffer (1% SDS, 0.1 m NaHCO3, 0.4 m NaCl) for 3 h at 67 °C. The DNA from the supernatants was purified using the MinElute Reaction Cleanup Kit (Qiagen). PCR of the input and specific and nonspecific immunoprecipitated DNA was performed for 35 cycles of 94 °C for 45 s, 62 °C for 45 s, 72 °C for 45 s, with a final extension of 7 min at 72 °C using forward PKD1, the forward primer, 5′-CTG CTG CCG ACC CTG TGG AG-3′ and reverse primer, 5′-GCG GCG CGG GGC GGA CGG-3′ (135 bp product); and forward human IEX-1, the forward primer, 5′-GAG GCA GGA AAA CGC TTG-3′ and reverse primer, 5′-GAC TTG ACA TGC ACA ATC CTA G-3′. The conditions used to amplify the 200-bp wild-type and p53-mutant PKD1 proximal promoters in pGL3 were 30 cycles of 94 °C for 45 s, 60 °C for 45 s, 72 °C for 45 s, using the forward primer GL3F, 5′-CAT GCA AAA TAG GCT GCT CC-3′ and reverse primer GL3R, 5′-GGC TTT ACC AAC AGT ACC G-3′.
Total RNA was prepared as described above and reverse transcription was carried out with 1 mg of total RNA using random hexamers at 42 °C for 1 h using a Promega kit. Quantitative PCR were run with 5 ng of the original input RNA (in 5 ml), 10 pmol of each primer for either PKD1 or L7 and GoTaq qPCR master mixture (Promega) in a total reaction volume of 25 ml, and the products were detected by SYBR Green in a Bio-Rad CFX96 cycler. The products were confirmed by melting temperatures. The qPCR conditions for PKD1 and L7 amplification were: 95 °C for 3 min and then 40 cycles of three steps: 30 s at 95 °C, 62 s at 62 °C, and 45 s at 72 °C. The “cycle threshold” Ct values were selected from the linear part of the fluorescence signal between 300 and 400 RFU. ΔCt was determined as: ΔCt = average PKD1 Ct − average L7 Ct, and gene-fold expression was calculated according to the Livak method, 2−ΔΔCt. Quantification of ChIP DNA was done in a similar way, except that the reaction was carried out under the same PCR conditions without final extension, using the same primer set with 1:100 dilution of the DNA.
We and others have shown that the product of the PKD1 gene, polycystin-1, activates JNK and AP-1 (32, 33). To determine whether the PKD1 gene is feedback-regulated by AP-1, we searched the PKD1 promoter region for the presence of canonical AP-1 sites using TESS (Transcription Element Search System) and identified a number of sites within a region 3.3-kb upstream of the PKD1 transcription start site. One element at −2,934 bp (TGAGTCA) showed high homology with other well characterized AP-1 sites. To examine whether any of these AP-1 sites were functional, a 3.3-kb PKD1 promoter-luciferase reporter (Fig. 1A) was transiently transfected into HEK293T cells together with a constitutively active CA-Mekk1 C-terminal construct. As a positive control for the CA-Mekk1 response, we used a promoter construct containing seven tandem AP-1 (7× AP-1) response elements upstream of a luciferase reporter. As shown in Fig. 1B, the expected increase in the AP-1 response was observed in CA-Mekk1 expressing cells. By contrast, the 3.3-kb PKD1 promoter showed markedly reduced activity. This observation suggested either that AP-1 was acting as a negative regulator of the PKD1 promoter, or that the CA-Mekk1 construct was working by another mechanism.
To localize the region in the promoter responsible for this repression, we made use of a number of deletion constructs of the PKD1 promoter (Fig. 1A). As shown in Fig. 1B, there was a similar CA-Mekk1 inhibitory response with all of the deletion constructs, suggesting that the CA-Mekk1 effect was mediated by the most proximal 200-bp fragment. Previously we showed that this DNA region has high basal promoter activity, and contains a number of functionally important transcription elements (14,–16), but does not contain a canonical AP-1 site. The 200-bp PKD1 promoter fragment responded to CA-Mekk1 in a dose-dependent fashion resulting in almost complete repression of basal activity (Fig. 1C).
CA-Mekk1 is comprised only of the C-terminal-most 293 amino acids of Mekk1. To examine whether the full-length protein elicited similar repression, we carried out co-transfections with a full-length Mekk1 construct in human HEK293T cells, monkey kidney cells (COS-1), and mouse cortical collecting duct cells (M-1). As shown in Fig. 1D, Mekk1 had a significant inhibitory effect on the 200-bp proximal promoter in all three cell types, similar to that seen with CA-Mekk1. The expression of the truncated and full-length Mekk1 proteins is shown by Western blot (Fig. 1D, inset). It had been demonstrated previously that the full-length 195–196-kDa Mekk1 generates an 80–90-kDa C-terminal product by caspase-3 cleavage (37, 38). The anti-Mekk1 antibody specific to the C-terminal 22 amino acids of Mekk1 is able to recognize the C-terminal products of full-length Mekk1, as well as the CA-Mekk1 protein. Thus, it is quite possible that the C-terminal proteolytic fragment of full-length Mekk1 causes the observed Mekk1 inhibition of the 200-bp PKD1 promoter.
It is well established that a sequence of activation events through the Mekk1-Mkk4/7-JNK phosphorylation cascade leads to transcriptional regulation of a number of target genes (25, 28, 29). One of the primary targets of this pathway is AP-1, which consists of homodimers of c-Jun or heterodimers of c-Jun with c-Fos or ATF2. c-Jun and ATF2 are direct substrates of c-Jun N-terminal kinase (JNK). Active JNK, by targeting different transcription factors, can activate or repress transcription. Mekk1 can also signal to MEK-ERK and p38. To determine whether the Mekk1-Mkk4/7-JNK signaling cascade or another MAPK pathway was involved in Mekk1-mediated repression, HEK293T cells were transfected with the 200-bp PKD1 promoter and CA-Mekk1 and were incubated with inhibitors of MEK-ERK (PD98059), p38 (SB202190), and JNK (SP600125 and HIV-TAT-PP-JBDIP) (39). The 7× AP-1 luciferase construct was used as a positive control for JNK inhibition. As expected, both of the JNK inhibitors effectively abrogated CA-Mekk1-mediated induction of AP-1 (Fig. 2, A and B), whereas neither the MEK inhibitor nor the p38 inhibitor had an effect (Fig. 2A). Surprisingly, none of the inhibitors was able to reverse the CA-Mekk1 inhibitory effect on the 200-bp PKD1 promoter suggesting that the JNK pathway is not involved in Mekk1-mediated repression of the PKD1 promoter, nor is any other known Mekk1-associated MAPK pathway.
Although it appeared that the known MAPKs were not involved, we considered the possibility that some other unknown downstream target could be involved. As such, we tested whether Mekk1 kinase activity was required for its inhibition of the PKD1 promoter. To address this, we exploited kinase-dead CA-Mekk1 (DC) and Mekk1 (DM) constructs, both carrying a single mutation at the same locus, replacing the kinase active-site lysine with methionine (K1253M) (40). As shown in Fig. 2C, the kinase-inactive isoforms were unable to induce AP-1 promoter activity, but were still able to repress PKD1 promoter activity, indicating that Mekk1 kinase activity, necessary for activation of downstream targets by phosphorylation, was not required for this transcriptional repression. These kinase-dead variants were expressed at levels equal to their active isoforms in transfected cells (Fig. 2D). As such, these results confirmed the observations obtained with the inhibitors and support the idea that the MAPK cascades are not involved in Mekk1-mediated repression of the PKD1 promoter, nor is any alternative pathway requiring Mekk1 kinase phosphorylation, thereby pointing to a novel mechanism of transcriptional repression by Mekk1.
Previously, we had shown that the 200-bp PKD1 proximal promoter region contains functional Sp1 and Ets motifs (Fig. 3A) (15, 16). Sp1 (115 kDa) is a ubiquitous transcription factor known to activate target gene expression. A smaller (80 kDa) Sp1 variant produced by internal transcription initiation can act as a potent inhibitor of Sp1/Sp3-mediated transcription (41). Our previous observation that Sp1 alone (or together with RXR) activates the 200-bp promoter via one or more Sp1 sites suggested that the Sp1 sites are not involved in mediating the Mekk1-inhibitory effect. Such was found to be the case when we introduced single and combination mutations into these sites and tested their activity by transient transfection in HEK293T cells (data not shown). Ets family transcription factors have been shown to be activators and repressors (42). To determine whether the Mekk1-inhibitory effect was mediated by any of the Ets motifs, we again introduced single and combination mutations into these sites and investigated their activity. None of these mutations was able to suppress the CA-Mekk1-mediated repression effect (data not shown).
Van Bodegom et al. (17) identified an atypical p53 site in the PKD1 promoter extending from −106 to −73, which mediates p53-dependent transcriptional repression (italics, Fig. 3A). To determine whether this p53 site could be involved, multiple mutations were introduced into the p53 site (p53 Mut), replacing consensus base pairs (RRRC(A/T)(T/A)GYYY) in the 3′ half-site. Two of these nucleotides play a key role in regulating p53 DNA-binding activity (43). As shown in Fig. 3B, these mutations in the PKD1 promoter substantially reduced inhibition by CA-Mekk1, as well as by p53, suggesting that the p53 site may be involved in the Mekk1-mediated repression. The somewhat lower basal activity in the mutant compared with the wild-type promoter could be attributed to the multiple mutations affecting nearby or sandwiched Sp1 sites, as a number of studies (44,–46) have identified the importance of 3′-flanking sequences immediately adjacent to an Sp1 motif, in determining functionality.
Previous studies showed that Mekk1 can increase p53 protein levels through JNK-dependent (47, 48) and JNK-independent (38) mechanisms. The data above provide evidence that a Mekk1-mediated, JNK-independent mechanism may be involved in the repression of PKD1 promoter activity. To determine whether p53 is involved in this mechanism, a dominant-negative (E258K) p53 mutant lacking the DNA-binding capability was tested. As shown in Fig. 4A, the mutant p53 prevented CA-Mekk1 repression of PKD1 promoter activity in a dose-dependent fashion, suggesting that mutant p53 is blocking the effect of endogenous p53 at the promoter. As shown in Fig. 4B, the levels of p53 were somewhat higher in the transfected cells.
To further test whether p53 is required for Mekk1-mediated repression, cells were transfected with a p53-specific siRNA. In the absence of transfected CA-Mekk1, p53 siRNA increased transcriptional activity of the promoter, consistent with knockdown of endogenous p53 (Fig. 4C, left, p53 knockdown is shown in D). Mekk1-specific siRNA also increased promoter activity (Fig. 4C, left, Mekk1 knockdown is shown in E). The repression of PKD1 promoter activity seen with transfected CA-Mekk1 was completely abolished by p53 siRNA (Fig. 4C, right). Furthermore, the repression seen with transfected p53 was abolished by Mekk1 siRNA. Together, these results suggest that there is a mutual interdependence between Mekk1 and p53 to effect repression of the PKD1 promoter.
To further determine whether these two factors are mutually dependent at the PKD1 promoter, p53-null HCT116 cells were transfected with the Mekk1 constructs. As shown in Fig. 4F, CA-Mekk1, Mekk1, and kinase-dead Mekk1 failed to cause repression in p53-null cells. If anything, the activity of the PKD1 promoter was slightly induced in both CA-Mekk1 and Mekk1 expressing cells. Together, these and previous results support the idea that the Mekk1-repressive effect is mediated by p53 via the p53 binding site in the 200-bp PKD1 promoter region.
If Mekk1 were to partner with p53 to repress the PKD1 promoter, it could do so by direct interaction in the nucleus. As such, we have used cell fractionation to test whether Mekk1 is able to translocate to the nuclear compartment. To establish the fidelity of the CE and NE we used the markers α-tubulin and poly(ADP-ribose) polymerase-1 (Fig. 5A). Transfected Mekk1 (CA-Mekk1, full-length Mekk1, and kinase-dead full-length Mekk1) could be detected in the nuclear extracts of HEK293T cells (Fig. 5A) as well as in the cytosolic extracts. Mekk1 nuclear transport was also observed in transfected COS-1 cells (Fig. 5A) and in p53-null HCT116 cells (data not shown). Importantly, a 195-kDa band representing full-length Mekk1 in vector-transfected (V) cells was observed in nuclear extracts, showing that endogenous, full-length Mekk1 is present in nuclei as well as in the cytoplasm.
Nuclear localization of Mekk1 was also observed by fluorescence microscopy in HEK293T cells transfected with GFP-tagged CA-Mekk1 (Fig. 5B, a and b). As there are indications that GFP may target fusion proteins to the nucleus, we also demonstrated Mekk1 in transfected HEK293T cells using immunofluorescence (Fig. 5B, c-f). In both experiments, a significant fraction of the transfected cells exhibited nuclear localization of Mekk1.
To evaluate whether Mekk1 associates with p53, we performed co-immunoprecipitation with anti-Mekk1 and anti-p53 antibodies. The data in Fig. 5C demonstrate that Mekk1 and p53 indeed interact with each other in both the cytosolic fraction and the nuclear fraction, and that this interaction involves the C-terminal region of Mekk1.
To determine whether Mekk1 binds the endogenous PKD1 promoter, we carried out chromatin immunoprecipitation (ChIP)-PCR assays using an anti-Mekk1 antibody. The locations of the forward and reverse PCR primers are shown in Fig. 3A. In this analysis (Fig. 6) the PCR products were sequentially run on the gel, first loading the products from control cells to look at endogenous proteins, followed 15 min later by the products from CA-Mekk1-transfected cells. As shown in Fig. 6A (lane 5 versus 2, lower bands), a PCR product similar in size to that obtained with the input DNA (0.14 kb) was seen with anti-Mekk1 antibodies, indicating that endogenous Mekk1 is bound to a region of the promoter of that size, upstream of the transcription start site. As expected, p53 was also localized to this region (lane 4, lower band). The increased band intensity observed in cells transfected with CA-Mekk1 (lane 5, upper band) is most likely due to increased Mekk1 binding to the promoter. CA-Mekk1 expression also increased p53 binding (lane 4, upper band), supporting the idea that the two proteins interact at the promoter and mutually stabilize their cooperative binding.
Because both Mekk1 and p53 appear to bind the PKD1 promoter to elicit similar repression effects, we asked whether p53 is necessary for endogenous Mekk1 to bind the promoter. To do this, p53-null HCT116 cells were subjected to ChIP analysis using anti-p53 and anti-Mekk1 antibodies. The presence of RXR at the promoter was used as a positive control (15). In Fig. 6B (upper panel) PCR products similar in size to those obtained with the input DNA (0.14 kb) were seen following ChIP with anti-RXR in both the vector-transfected (lane 4, lower band) and p53-transfected cells (lane 4, upper panel), indicating that RXR binds independently of p53. By contrast, amplified products from anti-Mekk1-immunoprecipitated DNA were only detected in the p53-transfected cells (lane 6), indicating that Mekk1 does not bind the PKD1 proximal promoter in the absence of p53. Similar results were observed for the p53 site of the IEX-1 promoter (Fig. 6B, lower panel) (49), where the presence of p53 and Mekk1 at the IEX-1 promoter was only seen in cells expressing p53 (lanes 5 and 6) supporting the idea that p53 is required for Mekk1 to bind p53 promoter sites in both genes.
To determine whether mutations in the p53 binding site can prevent Mekk1 binding, ChIP analysis using vector-specific primers flanking the 200-bp promoter was carried out on transfected constructs containing wild-type p53 and mutated p53 PKD1 promoter. As shown in Fig. 6C, PCR products of the same size obtained with the input DNA (0.27 kb) were seen in anti-p53 and anti-Mekk1 immunoprecipitated samples from cells transfected with the wild-type (WT) but not the mutated (Mut) promoter, further confirming the importance of p53 binding for Mekk1 binding.
Further confirmation that Mekk1 binds to this promoter fragment together with p53 was obtained with DNA pulldown assays (supplemental Fig. S1). These assays were carried out with nuclear extracts and a biotinylated 200-bp promoter fragment or a probe containing just the bipartite p53 site. Both the 200-bp fragment and the bipartite p53 probe were able to bring down CA-Mekk1, as well as p53, from nuclear extracts of HEK293T cells expressing CA-Mekk1. Neither probe pulled down c-Jun. Taken together, these results demonstrate that Mekk1 can bind to the 200-bp PKD1 proximal promoter via the p53 site.
To test whether expression of the endogenous PKD1 gene also responds to Mekk1, total RNA was isolated from HEK293T cells transfected with CA-Mekk1 or full-length Mekk1, and was analyzed by RT-PCR. As shown in Fig. 7, A–D, PKD1 mRNA levels underwent a decrease of up to ~70% in CA-Mekk1 or Mekk1 expressing cells in comparison to the control ribosomal protein L7 mRNA. These results show that the endogenous PKD1 gene is down-regulated by Mekk1 and thus may be a physiological target of the Mekk1 protein.
We established mouse M-1 cell lines stably expressing either kinase-dead CA-Mekk1-GFP (DC) or the control vector (V). Immunoblot analysis (Fig. 7E) showed increased levels of p53 in the kinase-dead CA-Mekk1-expressing cells. RT-PCR analysis on total RNA from the stably transfected cells, using primers specific for mouse Pkd1, revealed very significantly reduced Pkd1 mRNA levels in the kinase-dead CA-Mekk1 expressing cells (Fig. 7, F and G). Increased p53 levels were seen not only in the kinase-dead CA-Mekk1 expressing cells (Fig. 7E) but also in cells transiently transfected with kinase-active or kinase-dead full-length or truncated Mekk1 (supplemental Fig. S2).
To address whether Mekk1 could also be associated with p53-mediated repression of other genes, we carried out RT-PCR analysis on total RNA isolated from HEK293T cells transfected with CA-Mekk1, using primers specific for α-fetoprotein, bradykinin receptor B2 (BDRB2), and IEX-1. These genes were previously reported to be repressed by p53 (22, 49). As shown in Fig. 7H, mRNA levels for BDRB2 and IEX-1, but not α-fetoprotein, were significantly reduced in CA-Mekk1-transfected cells, suggesting that a Mekk1-p53 interaction may be involved in transcriptional repression of other genes in addition to PKD1, in a gene-specific manner.
To test whether extracellular stimuli can evoke p53-dependent repression of endogenous PKD1 gene expression in untransfected cells, we used a well known stress pathway inducer H2O2, and two activators of Mekk1, the phorbol ester PMA and TNFα. HEK293T (p53+/+) and HCT116 (p53−/−) cells were treated with these compounds and PKD1 mRNA was quantified using RT-qPCR. As shown in Fig. 8A, H2O2 mildly but significantly repressed PKD1 expression, and both PMA and TNFα strongly and significantly repressed PKD1 expression. The effect with TNFα was not seen at 24 h suggesting that it is transient. Repression was not observed in HCT116 (p53−/−) cells demonstrating that it is dependent on p53 (Fig. 8B). Instead of repression there was significant induction of the PKD1 promoter in p53-null cells. In parallel, ChIP analysis demonstrated increased chromatin binding for both p53 and Mekk1 at the PKD1 promoter following treatment of HEK293T cells with H2O2, PMA, and TNFα (Fig. 8C), indicating that endogenous Mekk1 is recruited to the PKD1 promoter along with p53. The involvement of endogenous Mekk1 in PKD1 repression was further tested using Mekk1-null mouse embryonic fibroblast cells. As shown in Fig. 8, D and E, there were higher levels of PKD1 mRNA (Fig. 8D) or PKD1 promoter-reporter activity (Fig. 8E) in Mekk1−/− compared with Mekk1+/+ cells. Endogenous PKD1 expression could also be induced by Mekk1 siRNA knockdown (supplemental Fig. S3). All these results taken together strongly suggest that the endogenous PKD1 gene is a target of Mekk1-p53 repression, and that PKD1 gene expression can be down-regulated through this mechanism by stimulation of the Mekk1-p53 pathway.
In this study, we report that the PKD1 gene is a target of Mekk1, thus establishing a formal link between a MAPK pathway component and PKD1 gene regulation. This connection was not entirely unexpected as MAPK pathways are involved in many cellular processes including cell proliferation, differentiation, migration, and apoptosis, all considered functions of PKD1. Mekk1 is at the top of a phosphorylation cascade that leads to activation of JNK, and Mekk1 can activate the ERK (50) and p38 pathways and NFκB in response to pro-inflammatory, growth stimulatory, or stress-response signals (51,–53). What was unexpected was the discovery that Mekk1 repression of the PKD1 promoter is mediated by a kinase-independent binding interaction with p53 at the PKD1 promoter.
Evidence that the PKD1 gene is a target of Mekk1 regulation came from the following observations: 1) PKD1 promoter-reporter assays demonstrated that transfected Mekk1 constructs repressed transcriptional activity of the 200-bp PKD1 proximal promoter in three different cell lines: human HEK293T, monkey COS-1, and mouse M-1 cells. 2) Ectopic Mekk1 expression markedly reduced endogenous PKD1 gene expression in human and mouse cells. 3) ChIP-PCR assays showed the presence of Mekk1 at the endogenous PKD1 promoter within a 135-bp fragment located within 169 bp of the start site of transcription.
It is widely known that signals from Mekk1 are propagated by sequential activation of Mek4/7 and JNK, which then activates or inhibits a number of transcription factors. However, we found that pharmacological inhibitors of the three major MAPK pathways: SP600125 and JBDIP (JNK pathway), PD98059 (MEK-ERK pathway), and SB202190 (p38 pathway), failed to reverse Mekk1 repression of the PKD1 promoter. We also showed that Mekk1 kinase activity, which is necessary for phosphorylation of its downstream targets, the key mechanism by which Mekk1 exerts its effects, was dispensable. This was shown by using kinase-dead mutants of Mekk1, which were found to be just as inhibitory as their kinase-active counterparts. These results suggested that classical MAPK signaling and substrate phosphorylation are not involved in Mekk1 repression of PKD1, thus pointing to a novel mechanism of transcriptional regulation. Although it has been shown that Mekk1 can function as an E3 ligase to regulate ERK (54) and c-Jun (55) protein levels through ubiquitation and proteasomal degradation, we have no evidence supporting such a mechanism in the regulation of PKD1 promoter activity.
Mutation of the p53 binding site within the 200-bp PKD1 proximal promoter prevented Mekk1 repression of the promoter. Further support was obtained by DNA pulldown assays, whereby Mekk1 was brought down by a DNA fragment harboring the p53 binding site from the PKD1 proximal promoter. These findings demonstrated that the p53 site in the 200-bp PKD1 promoter is involved in Mekk1-mediated transcriptional repression.
As shown earlier (17) and more recently by Van Bodegom et al. (56), the PKD1 gene is a physiological target of p53, whereby p53 directly binds the proximal-most p53 element in the PKD1 promoter to repress transcription. Thus, one possibility was that Mekk1 and p53 might function together to repress the promoter. This idea was supported by multiple lines of evidence that made use of a dominant-negative p53, siRNA knockdown of p53, and p53-null cells. Similar increases in basal activity of the 200-bp promoter observed with either p53 siRNA or Mekk1 siRNA suggest an interdependent mechanism for the two.
Our data suggest that Mekk1 binds the PKD1 promoter through p53 by acting as a co-repressor. Although p53 is a known DNA-binding protein, Mekk1 is not. Co-immunoprecipitation of Mekk1 and p53 from nuclear lysates suggest their close, physical interaction and thus the possibility that Mekk1 binds promoter-bound p53. This was supported by ChIP analysis, which showed that Mekk1 cannot bind the endogenous PKD1 promoter in the absence of p53 binding. It also appears that p53 does not function on the PKD1 promoter without Mekk1, because p53 was unable to repress promoter activity following Mekk1 siRNA knockdown. The IEX-1 (49) and BDRB2 (22) genes are also repressed by p53. Here we showed a marked reduction in the mRNA levels for these two genes following Mekk1 expression. Thus, a similar Mekk1-p53 repression mechanism may be operational for other genes as well.
Our studies place Mekk1 in a prominent position as a major upstream regulator of p53 function (Fig. 9). Others have shown Mekk1-mediated, JNK-dependent p53 protein stabilization, and JNK-dependent p53 phosphorylation and transcriptional activation of p53 target genes (Fig. 9, left) (47, 48). We now show that Mekk1 can directly regulate p53 target genes by translocating to the nucleus to act as an essential p53 co-repressor in close, physical association with DNA-bound p53 (Fig. 9, right). The kinase-dead mutants of Mekk1 were seen to increase p53 protein levels (Fig. 7E and supplemental Fig. S2). But because these increases in p53 were somewhat greater with kinase-active Mekk1 (supplemental Fig. S2), it is possible that both JNK-dependent and JNK-independent pathways regulate p53 protein levels (Fig. 9, left and right). In p53-positive HCT116 cells, γ-irradiation was found to cause significant reductions in PKD1 mRNA, as compared with irradiated isogenic p53-null cells (17). γ-Irradiation, which in many cells is a poor activator of JNK, is thought to increase p53 through a JNK-independent pathway (48, 57). As shown in Fig. 7E, an increase in p53 was seen in cells stably expressing kinase-dead CA-Mekk1, and this was associated with a significant decrease in PKD1 mRNA. Thus, the Mekk1 repressive effect may involve a combination of increased p53 protein and Mekk1-p53 transcriptional repression at the promoter. Importantly, extracellular stimuli thought to target the stress-response pathway (H2O2, PMA, and TNFα) were shown to recruit Mekk1 to the PKD1 promoter and to down-regulate PKD1 mRNA levels in a p53-dependent fashion.
The p53 protein can function as a transcriptional activator or repressor depending on the target gene. In fact, PKD1 and BDRB2, which have multiple p53 sites, can be activated or repressed by p53 (22, 56). The mechanisms of p53-mediated repression are not well understood and vary widely. They include: complex formation with the co-repressor mSin3a and histone deacetylases (for the PKD1, Map4, and stathmin genes (56, 58)); indirectly through intermediate factors such as p21/CDKN1A (for the survivin gene and 11 other p53-repressed genes (59), and p21 and E2F (for the hTERT telomerase gene (60)); interaction with Sp1 (for the cyclin B1 gene (61) and the telomerase gene (62)); and active repression and displacement of the overlapping transactivator HNF-3 (for the α-fetoprotein gene (63)). We now add another mechanism for p53-mediated repression in which Mekk1 acts as a nuclear p53 co-repressor (for the PKD1, IEX-1, and BDRB2 genes).
Recent evidence has shown that expression of the PKD1 gene can be modulated by both positive and negative p53 regulation, possibly to prevent under- or overexpression of the gene (17, 56). Our data now suggest that signals acting on Mekk1, such as oxidative stress or TNFα, could tip the balance at the PKD1 gene, resulting in sufficient down-regulation of gene expression to cause PKD1 haploinsufficiency and cyst formation (Fig. 9). Stress-pathway signals have been implicated in cyst formation in ADPKD (1,–7, 64), including TNFα, which has been found in human cyst fluid (64). Notably, TNFα has been shown to cause cyst formation in embryonic kidney organ cultures and in vivo through a Pkd2-dependent mechanism, and has been considered a potential target for PKD therapy (64).
Whether Mekk1 acts widely on most p53-targeted promoters, and whether Mekk1-p53 can activate as well as repress gene expression, remain to be elucidated. Nevertheless, it is intriguing to think that a Mekk1-p53 functional interaction might explain how Mekk1 variants could have a role in human cancer susceptibility (65) as well as in PKD.
We thank Dr. S. El-Dahr (Departments of Pediatrics and Physiology, Tulane University Health Science Center) for the p53 constructs, and Dr. B. Vogelstein (Johns Hopkins University Medical Center) for HCT116 p53-null cells.
3The abbreviations used are: