Inhibition of Mineralocorticoid-Mediated Transcription by NF-κB

tt

s

Archives of Biochemistry and BiophysicsVol. 383, No. 1, November 1, pp. 38–45, 2000doi:10.1006/abbi.2000.2045, available online at http://www.idealibrary.com on

Inhibition of Mineralocorticoid-Mediated Transcriptionby NF-kB1

Venkatadri Kolla2 and Gerald Litwack3

Department of Biochemistry and Molecular Pharmacology, Jefferson Medical College,Thomas Jefferson University, Philadelphia, Pennsylvania 19107

Received March 28, 2000, and in revised form July 31, 2000

c

Nuclear factor-kB (NF-kB) is a ubiquitous transcrip-tion factor that regulates the expression of multipleinflammatory and immune response genes and plays acritical role in host defense and in chronic inflamma-tory diseases. The mineralocorticoid receptor (MR) be-longs to the steroid/thyroid hormone receptor super-family of ligand-induced transcription factors. Wedemonstrate a dose-dependent, mutual transcrip-tional antagonism between NF-kB and MR in transientransfection experiments. We also show that the an-agonism is limited to the p65 subunit of NF-kB het-

erodimer but not p50. Transient cotransfection exper-iments with MR deletion constructs reveal the neces-sity of various N-terminal MR domains for thisphenomenon. Inhibition of NF-kB by IkB relieves therepression of NF-kB function by MR. These data sug-gest that the p65 subunit of NF-kB interacts with MRindirectly and transrepresses MR activation. © 2000

Academic Press

Key Words: NF-kB; MR; GR; aldosterone; TA; tran-cription; transfection.

NF-kB4 is a ubiquitous transcription factor that reg-ulates the expression of multiple inflammatory andimmune response genes (1–3). It is widely expressedand plays a key role in host defense and in chronicinflammatory diseases, such as asthma, inflammatorybowel disease, and rheumatoid arthritis. NF-kB wasoriginally identified as a factor that regulates the ex-

1 Supported by Research Grant AM44441 from the NIH.2 Supported by Training Grant T-32 DK07705 from the NIH.3 To whom correspondence and reprint requests should be ad-

dressed. Fax: (215) 503-5393. E-mail: [email protected] Abbreviations used: NF-kB, nuclear factor kB; MR, mineralocor-

ticoid receptor; Aldo, aldosterone; GRE, glucocorticoid response ele-ment; SEM, standard error mean.

38

pression of k-light chains in murine B-lymphocytes (1,4, 5). NF-kB is a heterodimeric transcription factorcomprised by p50 and p65 subunits, which belong tothe Rel family of transcriptional activators (6, 7). How-ever, a total of five NF-kB subunits that form dimershave been identified in mammalian cells; RelA or p65,RelB, c-Rel, p50, and p52. The most common and best-haracterized form of NF-kB is the RelA-p50 het-

erodimer. Each dimer combination exhibits differencesin DNA binding affinity and transactivation potential(8). These have a conserved 300-amino acid Rel homol-ogy domain or region (RHD/RHR) that is responsiblefor dimerization, nuclear localization, and DNA bind-ing (3, 9). NF-kB is normally present in the cytoplasmin an inactive form bound to and tightly regulated by asecond structurally related set of an inhibitory phos-phoproteins called IkB (10). Multiple forms of IkB havebeen identified such as IkBa, IkBb, IkB-e, IkBg, IkBd,Bcl-3, and the Drosophila protein cactus (11, 12). Thesemodulate the function of NF-kB heterodimer (9, 13–15). All IkB proteins have common 5–7 repeats ofhighly conserved 33 amino acid sequences calledankryn repeats that are required for interactions withthe RHD/RHR and NF-kB proteins. IkB proteins bindto NF-kB and mask its nuclear localization signal thussequestering the dimers in the cytoplasm of unstimu-lated cells (9, 13, 16). Phosphorylation of complexedIkB results in the dissociation of IkB from NF-kB fol-lowed by the translocation of NF-kB to the nucleus,where it binds to cis-acting Rel binding sites with theconsensus sequence 59 GGGPuNNPyPyCC 39 (9, 17). Inmost cells, NF-kB dimers remain sequestered in thecytoplasm until activated by environmental stimuli,including protein kinase C activators (phorbol esters,platelet-activating factor), proinflammatory cytokines(interleukin-1b, tumor necrosis factor-a, IL-11, IL-17),lipopolysaccharides, oxidants, ultraviolet, and ionizing
radiations (18). Viruses such as adenovirus, rhinovi-
0003-9861/00 $35.00Copyright © 2000 by Academic Press

All rights of reproduction in any form reserved.

fUmwJ

at

39TRANSCRIPTIONAL ANTAGONISM OF NF-kB WITH MINERALOCORTICOID RECEPTOR

rus, HIV, and hepatitis B could also activate NF-kB toregulate their own expression (3, 11).

The biologically active human mineralocorticoid hor-mone, aldosterone, plays a major role in regulatingsodium and potassium homeostasis in the kidney, sal-ivary glands, sweat glands, and gut (19). It participatesin the control of blood pressure and is implicated insome pathological disorders. Aldosterone exerts its ef-fects by acting through a ligand-activated transcriptionfactor, the mineralocorticoid receptor (20, 21). The min-eralocorticoid receptor (MR/NR3C2) (22) is a memberof the steroid/thyroid hormone receptor superfamily ofligand inducible transcription factors (23–25). Othernuclear hormone receptors in this family include glu-cocorticoid (GR/NR3C1) (22), thyroid (TR/NR1A1,NR1A2) (22), retinoic acid (RARs/NR1B1,2,3), and vi-tamin D receptors (NR1I1) (22) as well as numerousorphan receptors for which no ligands are known (25).These receptors display a modular structure consistingof five to six regions (A–F), with N-terminal A/B regionharboring an autonomous activation function, whereasthe C and E regions correspond to the DNA bindingdomain (DBD) and the ligand binding domain (LBD)(20). MR and GR exhibit a high degree of homology inthe DNA binding domain (DBD) (94%) and ligand bind-ing domain (LBD) but not in their N-terminal regions(15%). Both MR and GR bind two classes of adrenalsteroid hormones and differ greatly in their physiologiceffects. The latent MR and GR reside in the cytoplasmin a complex comprised by heat shock proteins (hsp90)and other stress family proteins (26). The binding ofhormone induces critical conformational changes insteroid receptors (SRs) that cause them to dissociatefrom an inhibitory hormone receptor complex resultingin activation of the receptor, which is thought to be amultistep process (27). The activated MR and GRtranslocate to the nucleus where the receptors, as ho-modimers, bind to a specific palindromic DNA se-quence known as a glucocorticoid response element(GRE) associated with target genes, and modulatetheir hormone-mediated transcription through specificinteraction with the transcription machinery (27, 28).

Modulation of genetic and cellular responses at thelevel of transcription by physiological or environmentalstimuli may involve synergistic or antagonistic inter-actions of transcription factors such as nuclear factor I(NF-I) and Sp1 with target gene promoter sequences(29, 30). A functional synergy between estrogen recep-tor (ER) and Sp1 has been shown where a protein–protein interaction was observed in an estrogen-in-duced transactivation pathway (31). Recently, we havedemonstrated the upregulation of MR and GR expres-sion by transcription factor Sp1 and the nuclear factorI (NF-I) (32). However, the interaction between MRand NF-kB has not been established. In an effort to
understand the molecular mechanism involved and the
role played by the MR in transcriptional antagonismwith NF-kB, we examined the interactions of MR andits deletion mutants, with two subunits, p50 and p65 ofNF-kB heterodimer, and identified regions responsiblefor the repression of MR transactivation.

MATERIALS AND METHODS

Reagents and plasmids. All hormones were purchased fromSigma Chemical Co. RU38486 was a gift from Roussel Uclaf (Ro-mainville, France). The human MR (NR3C2) expression plasmidsand pTK-CAT were obtained from Dr. Ronald M. Evans (Salk Insti-tute, La Jolla, CA) and the late Dr. Violet Daniel (Weizmann Insti-tute of Science, Rehovot, Israel), respectively. Expression plasmids ofp50 and p65 subunits and the inhibitory IkBa were a generous giftrom Dr. Sankar Ghosh (Howard Hughes Medical Institute, Yaleniversity School of Medicine, New Haven, CT). Gal4 reporter plas-id was generously provided by Raymond Penn and pGBT9 vectoras obtained from Dr. Alnemri (Kimmel Cancer Institute, Thomasefferson University, Philadelphia, PA).Statistical analyses. To determine the statistical significance of

ll the reported data we performed student t test analysis for threeo five sets of luciferase values normalized with b-gal. In all of our

experiments we found a P value ,0.05, suggesting that the reportedresults are statistically significant.

Oligonucleotides used as primers and clones. The oligonucleo-tides used in this study were commercially synthesized by Gibco-BRL Life Technologies and were used in cloning into pGL3-tkvector as well as for screening the clones by PCR. GRE31: 59TGTACAGGATGTTCTCTAGCGACTAGCTAGTTGTACAGGATGTTCT39; N-terminal HindIII primer: 59 CCCAAGCTTATGGAGACCAAAGGC-TAC 39; C-terminal SalI primer: 59 GGGGTCGACTCACTTCCGGTG-GAAGTA 39; C-terminal XhoI primer: 59 GGGCTCGAGTCACTTC-CGGTGGAAGTA 39; Bgl600R-Reverse primer: 59 CCCAGATCT-CAGGTGTTGGAAAGA 39; ScaIF- Forward primer: 59CCCAGATCTGTGGGATCTATCTGTAGC 39; ScaIR- Reverse primer: 59CCCAGATCTGTTGTTTGGAATAGCACC 39; MscIF- Forward primer: 59CCCAGATCTGGCCAGCTTAATATTGTC 39; MscIR- Reverse primer: 59CCCAGATCTGGCCATTTCCTTTGTAAAATTTC 39; B-402F- 59 GCTG-GATCCAGTACATTGCGG 39; T7- Primer 59 TAATACGACTCACTAT-AGGG 39; SP6 Primer- 59 ATTTAGGTGACACTATAG 39; Gal-DBD 59 KpnI Primer- 59 CCCGGTACCATGAAGCTACTGTCTTCTATC 39; Gal-DBD39 EcoRI I Primer- 59 CCCGAATTCTACAGTCAACTGTCTTTGACC 39;MR 600 KpnI Reverse Primer- 59 CCCGGTACCCAGGTGTTGGAAA-GATTGG 39; MR 670 EcoRI I Forward Primer- 59 CCCGAATTCTGTG-GCAGCTGCAAAG 39; Gal4-DBD was amplified by PCR from pGBT9vector (Clontech) with the above custom-made primers.

Cell culture and transfections. Monkey kidney cell line, COS-1,and MCF-7 were cultured in Dulbecco’s modified Eagle’s medium(DMEM) supplemented with charcoal-treated fetal bovine serum(10%). T-84, a colon-derived cell line was cultured in DMEM/F-12medium supplemented with stripped 10% fetal bovine serum. Tran-sient transfections of COS-1, MCF-7, and T-84 cells with variousreporter genes were performed as described in standard protocols(33) and by SuperFect (Qiagen). Cells (2 3 105) were electroporatedwith 5 mg of luciferase reporter construct (GRE31-Luc), 2 mg of MRexpression vector (RshMR), and 0.5 mg of b-galactosidase plasmid.Transfection with deletion constructs were carried out keepingequimolar amounts of expression plasmids created in KVhMR. Thesewere used to generate the data for Fig. 6B. For luciferase assayswhole cell extracts were prepared according to the manufacturer’sinstructions (Promega).

Cloning. pGL3-TK was constructed by excising the TK promoter(168 bp) from the pTK-CAT vector and cloning into the pGL3-basic
vector. GRE31 was cloned at the BglII site of pGL3-TK (21). MRwild-type cDNA fragment was amplified by PCR with designed cus-

vsm

mg

S1SBpAbtmIm

40 KOLLA AND LITWACK

tom made primers (with overhanging restriction enzyme sites ofHindIII and XhoI) and ultimately cloned in a eukaryotic expressionvector pcDNA3a (KVhMR). The deletion constructs of MR were madewithin KVhMR either by single/double digestions or by PCR ampli-fications followed by ligations. PCR amplifications were done usingcustom primers designed at required positions without changing thereading frames in the final construct after ligation. The hMR 59-328was created after digesting the cDNA with ScaI–ScaI. Subsequentreligation generated an uninterrupted 715-amino acid sequence inopen reading frame (34). Similarly, digestion of hMR cDNA withMscI (amino acids 149–391) followed by religation generated a 742-amino acid hMR in an open reading frame. The wild-type and dele-tion mutant receptor cDNAs were cloned at HindIII–XhoI sites ofpDNA3a. Recombinant MR with Gal4-DBD was constructed by PCRamplifying Gal4-DBD, MR except DNA-BD followed by ligation inproper orientation and in-frame. All other deletion constructs used inthis study were made as described earlier (34). Both restrictionenzyme analyses and DNA sequencing confirmed all the deletion andwild-type constructs.

Whole cell extract preparation. COS-1 and T-84 whole cell ex-tracts, expressing MR (35) and either p65 or 50 were prepared asdescribed by standard procedures. In brief, cells were harvested andhomogenized by 20 strokes using a dounce homogenizer in ice-coldbuffer containing 20 mM Tris Cl (pH 7.4), 600 mM KCl, 20% glycerol,2 mM dithiothreitol (DTT), 5 mM phenylmethylsulphonylfluoride(PMSF), 5 mg/ml leupeptin, and 5 mg/ml antipain. The homogenatewas centrifuged at 100,000g for 30 min at 4°C and protein concen-tration was estimated by Coomassie Plus protein assay reagent kit(PIERCE).

b-galactosidase assays and protein estimations. b-galactosidaseassays were performed to normalize for the variations in transfectionefficiencies (33). Briefly, equal amounts of protein from all sampleswere incubated at 37°C with 0.1 M sodium phosphate, 4 mg/mlONPG (o-nitrophenyl-b-galactopyranoside), and 1003 Mg buffer in aolume of 300 ml (33). The reactions were stopped with either 1 Modium carbonate or 1 M Tris solution and measured spectrophoto-etrically at 420 nm. b-Galactosidase values were used as an inter-

nal reference in transfection experiments and normalized againstthe luciferase numbers. Protein estimations were done by CoomassiePlus protein assay reagent kit (Pierce), a modified Bradford colori-metric method. Readings were recorded at 595 nm using BSA (bovineserum albumin) as standard.

Antibodies. Polyclonal MR antibody (aMR- raised against N-ter-inus and the DNA binding domain of the human MR antibody) was

enerated by immunization of rabbits with 50 mg of conjugatedprotein each for four times. Ab titers were assessed by enzyme-linkedimmunosorbent assay (ELISA) and aliquots of serum were stored at280°C (35). MR-specific Ab was also obtained from Affinity BioRe-agent Inc. and Santa Cruz Biotechnology. Polyclonal p65 and p50antibodies were obtained from Santa Cruz Biotechnology Inc.

Western blot analyses. The expressions of p65, p50, and MR wereexamined by Western blot analyses as described (35, 36). Totalprotein (70–100 mg) from transfected COS-1 cells was resolved on 8%

DS–PAGE and transferred electrophoretically (100 V constant forh) to Hybond ECL nitrocellulose filter paper (Amersham Life

cience). The filter paper was blocked overnight in 10% nonfat milk.lots were incubated in 5% nonfat milk with either polyclonal p65 or50 or MR primary antibody (1:1000) for 1 h at room temperature.fter stringent washing, they were incubated with secondary anti-ody (1:3000, Anti-rabbit Ig horseradish peroxidase conjugated an-ibody, Amersham Life Science) for 1 h and developed by chemilu-iniscent ECL as described (Amersham Inc., Arlington Heights, IL).

mmunoprecipitations were carried out as per the protocols recom-ended by Santa Cruz Biotechnology, Inc.

RESULTS AND DISCUSSION

Mutual Transcriptional Inhibition between MR andNF-kB

Earlier studies have demonstrated that NF-kB andGR mutually transrepress transactivation functions(37, 38). However, any interaction between NF-kBand the human mineralocorticoid receptor (MR) isunknown. To characterize the mutual interaction be-tween MR and NF-kB we transiently cotransfectedCOS-1 cells with a reporter construct GRE31-Luc,MR, pSVb-gal, and p65 subunit of NF-kB since thissubunit has been shown to be transcriptionally ac-tive in many cell types and inducible. GRE31-Lucreporter was constructed by inserting two copies ofGREs (Glucocorticoid Response Elements) upstreamof the thymidine kinase promoter in a basic lucif-erase vector pGL-3. Transfection experiments re-vealed the transcriptional repression of ligand (aldo-sterone)-activated MR by p65 (Fig. 1A). We also per-formed experiments in the absence of MR and withbasic minimal thymidine kinase promoter where weobserved no activity (data not presented). In this setof experiments we kept a fixed amount of MR expres-sion plasmid and varied amounts of p65 expressionplasmid as depicted in Fig. 1A. Under normal condi-tions MR exhibited over 12-fold induction with thereporter construct, whereas increasing amounts ofp65 subunit significantly reduced the activity of MR.To rule out the possibility that the observed transre-pression of MR by p65 might be nonspecific, we per-formed similar experiments with other reporter con-structs (GRE-CAT and MMTVCAT) and obtained ex-actly the same dose dependent effect of MRtransactivation by p65 regardless of the reporterconstruct used (data not shown). Western blot anal-yses demonstrated that the effect is not due to analteration of the expressed MR amount (Fig. 1B).Western blots were also performed to demonstratethat the increase in p65 plasmid (1–3 mg) did indeedresult in a progressive increase in p65 protein(Fig. 1C). These results suggest that the p65 sub-unit of NF-kB represses the transcriptional activityof MR.

To test whether the above phenomenon is mutualwe next examined the dose-dependent effect of MRwith GRE31-Luc reporter construct by keeping fixedamount of p65 expression plasmid and varyingamounts of MR expression plasmid. Transfection ex-periments indicate that increasing amounts of MRshowed proportionately better transcriptional induc-tion in the presence of aldosterone, whereas no effectwas observed in the absence of ligand (Fig. 2A). Wealso performed Western blot analysis to demonstrate
that the p65 subunit is expressed constitutively at

euiw

cjm(

e

tt

e

W


one level and not due to the variation (Fig. 2B).These results suggest that the transcriptional re-pression of MR by p65 subunit of NF-kB is mutual.These results also correlate with the results reportedfor other steroid receptors including glucocorticoid

FIG. 1. (A) NF-kB represses the MR transactivation of the GRE-Luc reporter. COS-1 cells were transiently transfected with areporter construct GRE-Luc (5 mg), b-gal (0.5 mg), MR (RshMR)xpression plasmid (2 mg), and various amounts of p65 expression

plasmid as indicated in the figure by Superfect. Total amount ofDNA in each transfection was adjusted to 12 mg with pGL3 basicDNA. After 24 h cells were treated with or without aldosterone(100 nM) and RU38486 (1 mM) followed by incubation at 37°C.Whole-cell extracts were prepared and used in luciferase assays asdescribed under Materials and Methods. The results are the meanof five independent experiments each run in triplicate and all thevalues were normalized to b-galactosidase activity. To determinehe statistical significance of these data, we performed student test analysis where the P value was ,0.05. The error bars indicate

6SEM. (B) Western blot analysis of MR. To determine proteinlevels of the MR corresponding to 0 to 3 mg of the p65 subunitxpression plasmid in the above experiment, 100 mg of total cel-

lular extract from aldosterone (100 nM)-treated cells was sub-jected to SDS–PAGE followed by Western blot as described underMaterials and Methods. The membrane was probed with a pri-mary polyclonal MR antibody (Santa Cruz Biotechnology) and asecondary HRP conjugated antibody and developed by ECL kit asdescribed under Materials and Methods. (C) Western blot analysisof p65. To determine the progressive increase in p65 plasmid (1–3mg) did indeed result in a progressive increase in p65 protein

estern blot analysis was performed using 100 mg of total cellularextract from aldosterone (100 nM)-treated cells as describedabove. The membrane was probed with a primary polyclonal p65antibody (Santa Cruz Biotechnology) and a secondary HRP con-jugated antibody and developed by ECL kit as described underMaterials and Methods.

receptor (37–39).bM

The p50 Subunit of NF-kB Is Not Required forRepression of MR Transactivation But p65 IsEssential

We next examined the role of p50 subunit of NF-kBin the repression of MR by performing transfectionexperiments with GRE-Luc reporter, p50 expressionplasmid, and MR expression plasmid. Cotransfectionexperiments revealed that there was no inhibition ofligand activated MR transactivation where we usedp50 subunit. Moreover, in the absence of MR (data notshown) and/or aldosterone, no transcriptional activa-tion was observed (Fig. 3A). As shown in Fig. 3A thep50 subunit of NF-kB heterodimer did not inhibit thetransrepression of ligand-activated transcription re-gardless of the amount of p50 expression plasmidtransfected. We also carried out Western blot analysisto assess the level of MR expression in the above trans-fection experiments performed for luciferase assays.

FIG. 2. (A) Mutual transcriptional antagonism between MR andthe p65 subunit of NF-kB. COS-1 cells were transiently transfectedby Superfect with a reporter construct GRE-Luc (5 mg), b-gal (0.5mg), various amounts of MR (RshMR) expression plasmid and p65expression plasmid (2 mg) as indicated in the figure. Total amount ofDNA in each transfection was adjusted to 12 mg with pGL3 BasicDNA. After 24 h cells were treated with or without aldosterone (100nM) and RU38486 (1 mM) followed by incubation at 37°C. Whole Cellxtracts were prepared and used in luciferase assays as describednder Materials and Methods. The results are the mean of five

ndependent experiments each run in triplicate and all the valuesere normalized to b-galactosidase activity. To determine the sta-

tistical significance of these data, we performed student t test anal-ysis where the P value was ,0.05. The error bars indicate 6SEM. (B)Western blot analysis of p65. To determine protein levels of p65corresponding to 0 to 5 mg of MR expression plasmid 100 mg of totalellular extract from aldosterone (100 nM)-treated cells was sub-ected to Western blot as described in the legend of Fig. 1B. The

embrane was probed with a primary polyclonal p65 antibodySanta Cruz Biotechnology) and a secondary HRP conjugated anti-
ody and developed by ECL kit as described under Materials andethods.

ea

t

(m(bM


Our results demonstrate that the expression of MR isnot altered irrespective of the p50 expression plasmidused (Fig. 3B). Further, there was no significant differ-ence in MR activation in the presence or absence of p50subunit expression plasmid suggesting that the p50subunit is not required in the mutual transcriptionalrepression.

We then performed similar experiments usingequimolar amounts of p65 and p50 or p65 subunitexpression plasmids in cotransfection experiments todetermine whether the formation of heterodimer hasany role in this mutual transcriptional antagonism.Results presented in Fig. 4 are exactly the same as weobserved in our previous experiments and not differentwhere we used only the p65 subunit. Moreover, addi-tion of equimolar amounts of p50 did not alter theactivation pattern of ligand activated MR. This sug-gests that the mutual transcriptional antagonism ofMR has occurred only with the p65 subunit of NF-kBbut not with the p50 subunit and does not require

FIG. 3. (A) Effect of the p50 subunit on mutual transcriptionalantagonism. COS-1 cells were transiently transfected with a re-porter construct GRE-Luc (5 mg), b-gal (0.5 mg), MR (RshMR) expres-sion plasmid (2 mg), and various amounts of the p50 expressionplasmid as indicated in the figure by Superfect. Cells treatments andpreparation of whole cell extracts were exactly the same as in pre-vious figures. The results are the mean of five independent experi-ments each run in triplicate and all the values were normalized tob-galactosidase activity. To determine the statistical significance ofthese data, we performed student t test analysis where the P valuewas ,0.05. The error bars indicate 6SEM. (B) Western blot analysisof p50. To determine that despite an increase in p50 protein expres-sion, there was no repression of MR Western blot analysis wascarried out using 100 mg of total cellular extract from aldosterone100 nM)-treated cells as described in the legend of Fig. 1C. Theembrane was probed with a primary polyclonal p50 antibody

Santa Cruz Biotechnology) and a secondary HRP-conjugated anti-ody and developed by ECL kit as described under Materials andethods.

subunit heterodimerization.d,

Mutual Transcriptional Repression between MR andp65 Is Relieved by I-kBa

To further confirm this mutual transcriptional an-tagonism between MR and the specificity of the p65subunit of NF-kB, we performed similar cotransfectionexperiments in the presence of the inhibitory rel sub-unit I-kBa. Increasing expression of this inhibitoryI-kBa should cause an increased binding to NF-kBespecially to the p65 subunit thereby making limitedamounts available to transrepress MR. It has also beenshown that glucocorticoids modulate NF-kB functiondirectly by increasing expression of I-kBa (40, 41). Inorder to investigate this we next determined whetheroverexpression of I-kBa relieves p65 mediated repres-sion of the ligand activated MR. Our results presentedin Fig. 5A (COS-1 cells) and Fig. 5B (T-84 cells) indi-cate that p65 mediated repression of MR transactiva-tion of GRE-Luc reporter is completely blocked by co-transfection of I-kBa. Similar results were obtainedwhen we used MCF-7 cells, indicating that the effect isindependent of cell type (data not presented). We alsoobserved no effect of I-kBa on GRE-Luc. Earlier reportsindicate that I-kBa relieves p65 mediated repressionfor GR, ER, and PR completely (37, 38), while for ARonly half of the transactivation was recovered (37, 38).Hence, I-kBa blocked the p65 subunit and therebyrelieved the repression of ligand activated MR. This

FIG. 4. The p50 subunit of NF-kB is not required for mutualtranscriptional antagonism. MR transactivation of the GRE-Luc wasdetermined in the presence of increasing amounts of NF-kB het-rodimer. COS-1 cells were transiently transfected by Superfect withreporter construct GRE-Luc (5 mg), b-gal (0.5 mg), MR (RshMR)

expression plasmid (2 mg) and various amounts of the p50 and p65subunit expression plasmids in equimolar amounts and p65 alone asindicated in the figure. Cell treatments and preparation of whole cellextracts for luciferase assays were exactly the same as in previousfigures. The results are the mean of four independent experimentseach run in triplicate and all the values were normalized to b-galac-osidase activity. To determine the statistical significance of these
ata, we performed student t test analysis where the P value was0.05. The error bars indicate 6SEM.

d

D

e

si tio

ude


confirms the role and specificity of the p65 subunit inthe mutual transcriptional antagonism with MR.

Multiple N-Terminal MR Domains Are Required forMutual Antagonism

In order to localize the regions of the MR that areinvolved in transcriptional repression by the p65 sub-unit of NF-kB, a number of MR deletion mutants wereconstructed and used in the transrepression study. Thedeletion mutants of MR were created either by restric-tion enzyme digestion or by PCR followed by the liga-tion of the fragments in an expression vector pcDNA3in-frame. Subsequently, these deletion mutants wereused in transfection experiments along with a reporterconstruct GRE-Luc and the p65 subunit of NF-kB, todetermine precise regions of MR involved in the trans-repression phenomenon (Fig. 6A). Our results demon-strate that various N-terminal domains are involved inthe mutual transcriptional antagonism by NF-kB (Fig.6B). As an experimental control tk-promoter linked toa luciferase gene was cotransfected with wild-type MRexpression plasmid where we observed no activity inthe presence or absence of p65. We also observed noactivity in the absence of ligand aldosterone (data notpresented). All other deletion constructs were trans-fected with a reporter construct GRE-Luc. As shown in

FIG. 5. I-kBa blocks the p65-mediated mutual transcriptional represcribed in the legend of Fig. 1 with a reporter construct GRE-Lu

expression plasmid, and I-kBa as indicated in the figure. Total amouNA. After 24 h cells were treated with or without aldosterone (100

extracts were prepared and used in luciferase assays as described unxperiments each run in triplicate and are presented as percent

b-galactosidase activity. To determine the statistical significance of t,0.05. (B) Role of I-kBa in p65-mediated mutual transcriptional reprame as in A, except the use of colon-derived cells instead of COS-1n triplicate and are presented as percentage of control transactiva

determine the statistical significance of these data, we performed st

Fig. 6B DNA-binding domain (DBD) and the ligand-

binding domain (LBD) of MR are essential for tran-scriptional repression with NF-kB. Previous reportsalso indicate that the DNA binding domain of GR,especially the two zinc finger domains are essential fortranscriptional repression, removal of which com-pletely releases the inhibition of NF-kB (37, 38). How-ever, in the case of MR the role of zinc finger domainswas not investigated since the mutations we created inthis region fall out of the reading frame. We believe thezinc finger domains in the human MR might haveplayed role similar to the case of GR with NF-kB sinceboth MR and GR exhibit high sequence homology.

Previously, it has been shown that the N-terminaldomain of hMR is involved in the transcriptional acti-vation without affecting aldosterone binding (34).However, it is not clear whether this region also isinvolved in mutual transcriptional antagonism withNF-kB. Our results demonstrate that N-terminal dele-tion constructs 59–328, 59–392, 150–328, and 1–402are essential for transactivation and involved in p65mediated transcriptional repression (Fig. 6B). How-ever, we detected moderate induction with deletionconstructs 59–328 and 150–328. These results confirmthe involvement of N-terminal regions in the mutualtranscriptional antagonism. We also demonstratedthat homologous down regulation of the hMR is not

on. (A) COS-1 cells were transiently transfected exactly the same as5 mg), b-gal (0.5 mg), MR (RshMR) expression plasmid (2 mg), p65of DNA in each transfection was adjusted to 12 mg with pGL3 basic

) and RU38486 (1 mM) followed by incubation at 37°C. Whole-cellr Materials and Methods. The results are the mean 6 SEM of three

of control transactivation after the values were normalized toe data, we performed student t test analysis where the P value wasion in T-84 cells. Transfections and cell treatments were exactly thels. The results are the mean 6 SEM of three experiments each runn after the values were normalized to b-galactosidase activity. Tont t test analysis where the P value was ,0.05.

essic (ntnM

deagehesesscel

inhibited by the p65 subunit like GR (data not shown)

t

or asep e er


(37, 38) and supports the idea of either direct or indi-rect physical interaction between MR and NF-kB.However, other transcription factors required for thisphenomenon are yet to be investigated.

In conclusion, the p65 subunit of NF-kB represses

FIG. 6. (A) Schematic representation of various deletion mutants u(B) Various N-terminal domains are essential for mutual transcriptransiently transfected with a reporter construct GRE-Luc (5 mg), b-gTotal amount of DNA in each transfection was adjusted to 12 mg withf extracts for luciferase assays were exactly the same as in previous fiun in triplicate and all the values were normalized to b-galactosiderformed student t test analysis where the P value was ,0.05. Th

he transcriptional activity of MR and this phenome-

non is mutual. Interestingly, I-kBa blocked the p65subunit and relieved the repression of ligand activatedMR that confirms the role and specificity of the p65subunit in the mutual transcriptional antagonism.Further, we observed no significant difference in MR

in this study. The numbers represent the exact deletion positions.nal antagonism with the p65 subunit of NF-kB. COS-1 cells were(0.5 mg), MR deletion constructs (2 mg), and p65 expression plasmid.L3 Basic DNA. Treatment of cells with aldosterone and preparation

res. The results are the mean of three independent experiments eachactivity. To determine the statistical significance of these data, weror bars indicate 6SEM.

sedtioalpGgu

activation in the presence or absence of the p50 sub-

1

11


unit, suggesting that the p50 subunit is not involved inthe mutual transcriptional repression. Mutationalanalyses revealed the involvement of various N-termi-nal regions in this mutual transcriptional antagonism;however, a precise nucleotide/amino acid sequencecould not be specified. We also demonstrated that ho-mologous down-regulation of the hMR is not inhibitedby the p65 subunit as in the case with GR (data notshown) (37, 38). These findings correlate with the re-ports of other steroid receptors, including glucocorti-coid receptor (37, 38, 42), even though there are somevariations. We plan to identify various transcriptionalfactors that might be involved in the interactions be-tween NF-kB and the human mineralocorticoid recep-tor which may clarify various stress related issues andthe mechanisms of mineralocorticoid actions.

ACKNOWLEDGMENT

Authors are thankful to Dr. Srinivasula, S.M., for providing vari-ous reagents and useful discussion in performing Immunoprecipita-tion experiments.

REFERENCES

1. Baeuerle, P. A., and Baltimore, D. (1996) Cell 87(1), 13–20.2. Barnes, P. J., and Adcock, I. M. (1997) Trends Pharmacol. Sci.

18(2), 46–50.3. Barnes, P. J. (1997) Int. J. Biochem. Cell Biol. 29(6), 867–870.4. Sen, R., and Baltimore, D. (1986) Cell 46(5), 705–716.5. Siebenlist, U., Franzoso, G., and Brown, K. (1994) Annu. Rev.

Cell Biol. 10, 405–455.6. Fujita, T., Nolan, G. P., Ghosh, S., and Baltimore, D. (1992)

Genes Dev. 6(5), 775–787.7. Liou, H. C., Nolan, G. P., Ghosh, S., Fujita, T., and Baltimore, D.

(1992) Embo J. 11(8), 3003–3009.8. Delfino, F., and Walker, W. H. (1999) Mol. Cell. Endocrinol.

157(1–2), 1–9.9. Miyamoto, S., and Verma, I. M. (1995) Adv. Cancer Res. 66,

255–292.10. Verma, I. M., Stevenson, J. K., Schwarz, E. M., Van Antwerp, D.,

and Miyamoto, S. (1995) Genes Dev. 9(22), 2723–2735.11. Ghosh, S., May, M. J., and Kopp, E. B. (1998) Annu. Rev. Immu-

nol. 16, 225–260.2. Whiteside, S. T., and Israel, A. (1997) Semin. Cancer Biol. 8(2),

75–82.3. Baldwin, A. S., Jr. (1996) Annu. Rev. Immunol. 14, 649–683.4. Rothwarf, D. M., Zandi, E., Natoli, G., and Karin, M. (1998)

Nature 395(6699), 297–300.15. Karin, M. (1999) J. Biol. Chem. 274(39), 27339–27342.
16. Nolan, G. P., and Baltimore, D. (1992) Curr. Opin. Genet. Dev.
2(2), 211–220.

17. Thanos, D., and Maniatis, T. (1995) Cell 80(4), 529–532.18. Schreck, R., Rieber, P., and Baeuerle, P. A. (1991) Embo J. 10(8),

2247–2258.19. Horisberger, J. D., Lemas, V., Kraehenbuhl, J. P., and Rossier,

B. C. (1991) Annu. Rev. Physiol. 53, 565–584.20. Fagart, J., Wurtz, J. M., Souque, A., Hellal-Levy, C., Moras, D.,

and Rafestin-Oblin, M. E. (1998) Embo J. 17(12), 3317–3325.21. Kolla, V., and Litwack, G. (2000) Mol. Cell. Biochem. 204, 35–40.22. Committee, N. R. N. (1999) Cell 97, 161–163.23. Arriza, J. L., Weinberger, C., Cerelli, G., Glaser, T. M., Handelin,

B. L., Housman, D. E., and Evans, R. M. (1987) Science237(4812), 268–275.

24. Pearce, D., and Yamamoto, K. R. (1993) Science 259(5098),1161–1165.

25. Tsai, M. J., and O’Malley, B. W. (1994) Annu. Rev. Biochem. 63,451–486.

26. Hong, H., Kohli, K., Garabedian, M. J., and Stallcup, M. R.(1997) Mol. Cell. Biol. 17(5), 2735–2744.

27. Trapp, T., and Holsboer, F. (1996) Trends Pharmacol. Sci. 17(4),145–149.

28. Beato, M., Herrlich, P., and Schutz, G. (1995) Cell 83(6), 851–857.

29. Truss, M., and Beato, M. (1993) Endocrinol. Rev. 14(4), 459–479.

30. Liu, M., and Freedman, L. P. (1994) Mol. Endocrinol. 8(12),1593–1604.

31. Porter, W., Saville, B., Hoivik, D., and Safe, S. (1997) Mol.Endocrinol. 11(11), 1569–1580.

32. Kolla, V., and Litwack, G. (1999) Mol. Cell Biol. Res. Commun.1(1), 44–47.

33. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) MolecularCloning: A Laboratory Manual, 2nd ed., Cold Spring HarborLaboratory Press, Cold Spring Harbor, New York.

34. Govindan, M. V., and Warriar, N. (1998) J. Biol. Chem. 273(38),24439–24447.

35. Kolla, V., Robertson, N. M., and Litwack, G. (1999) Biochem.Biophys. Res. Commun. 266(1), 5–14.

36. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seid-man, J. G., Smith, J. A., and Struhl, K. (Eds) (1989) CurrentProtocols in Molecular Biology, Wiley, New York.

37. McKay, L. I., and Cidlowski, J. A. (1998) Mol. Endocrinol. 12(1),45–56.

38. McKay, L. I., and Cidlowski, J. A. (1999) Endocrinol. Rev. 20(4),435–459.

39. Nolan, G. P., Ghosh, S., Liou, H. C., Tempst, P., and Baltimore,D. (1991) Cell 64(5), 961–969.

40. Kalkhoven, E., Wissink, S., van der Saag, P. T., and van derBurg, B. (1996) J. Biol. Chem. 271(11), 6217–6224.

41. Kalkhoven, E., Valentine, J. E., Heery, D. M., and Parker, M. G.(1998) Embo J. 17(1), 232–243.

42. Scheinman, R. I., Gualberto, A., Jewell, C. M., Cidlowski, J. A.,and Baldwin, A. S., Jr. (1995) Mol. Cell. Biol. 15(2), 943–953.

Inhibition of Mineralocorticoid-Mediated Transcription by NF-κB

Documents

Transcript of Inhibition of Mineralocorticoid-Mediated Transcription by NF-κB