1759 (2006) 340–347www.elsevier.com/locate/bbaexp

Biochimica et Biophysica Acta

Upregulation of human angiotensinogen (AGT) gene transcription byinterferon–gamma: Involvement of the STAT1-binding motif

in the AGT promoter

Sudhir Jain a, Mehul Shah b, Yanna Li a, Govindaiah Vinukonda a,Pravin B. Sehgal b, Ashok Kumar a,⁎

a Department of Pathology, New York Medical College, Rm 455, Basic Science Building, Valhalla, NY 10595, USAb Department of Cell Biology and Anatomy, New York Medical College, Valhalla, NY 10595, USA

Received 1 February 2006; received in revised form 19 July 2006; accepted 24 July 2006Available online 28 July 2006

Abstract

Mechanisms to maintain blood pressure in the face of infection are critical to survival. The angiotensinogen (AGT) gene locus is an importantcomponent of this response. Thus the AGT gene, expressed predominantly by liver cells, is known to be a positive acute phase reactant. We havepreviously demonstrated activation of the AGT promoter in hepatocytes through the IL6/STAT3 signaling mechanism. We have now investigatedwhether IFN-gamma, a cytokine also induced in response to diverse infections, can regulate AGT gene expression, and have elucidated the molecularmechanism involved. IFN gamma treatment up-regulated AGT mRNA level and promoter activity in Hep3B hepatocytes. Sequential deletion of thepromoter from the 5′ side suggested themajor IFN gamma responsiveDNA element to be between−303 and−103. This region contained a candidateSTAT1-binding site between −271 and −279. EMSA and chromatin immuno-precipitation (ChIP) assays confirmed that IFN-gamma treatmentinduced the binding of STAT1 to this element. Reporter constructs containing this AGT promoter derived element in a multimerized context but not amutant version were responsive to IFN gamma. Moreover mutating this STAT1 element in the context of the wild-type AGT holo promoter reducedresponsiveness to IFN gamma. In contrast to the clear synergism between dexamethasone and IL 6 in the upregulation of the AGT promoter (throughinteraction betweenGR and STAT3), the combination of IFN gammawith IL 6 or with dexamethasone did not further increase AGT promoter activitysuggesting that the IFN gamma/STAT1 pathway represents a separate signaling mechanism. These data highlight the redundancy in cytokine-mediated host response pathways aimed at the maintenance of blood pressure during infection.© 2006 Elsevier B.V. All rights reserved.

Keywords: Cytokine; Transcription factor; Gene regulation; Inflammation

1. Introduction

Mechanisms to maintain blood pressure in the face ofinfection are critical to survival. The angiotensinogen (AGT)gene locus is an important component of this response and isassociated with human essential hypertension [1,2]. Singlenucleotide polymorphisms in the promoter of the hAGT geneaffect the transcription of this gene and increased transcription ofthe hAGT gene may lead to hypertension. It is therefore impor-tant to understand factors involved in transcriptional regulationof the hAGT gene. AGT is an acute phase protein and its ex-

⁎ Corresponding author. Tel.: +1 914 594 4398; fax: +1 914 594 4163.E-mail address: ashok_kumar@nymc.edu (A. Kumar).

0167-4781/$ - see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.bbaexp.2006.07.003

pression is positively regulated by cytokines during the hostresponse to inflammation [3]. AGT is primarily synthesized inthe liver and adipose tissue, and to a lesser extent in the kidney,brain, heart, adrenal, and vascular walls [4,5]. AGT is first con-verted by renin to produce a decapeptide, angiotensin-I, which isthen converted to angiotensin-II by the removal of a C- terminaldipeptide by angiotensin-converting enzyme (ACE) [3]. Angio-tensin II (Ang II) is one of the most potent vasoactive hormonesand regulates a variety of physiological responses, includingfluid homeostasis, aldosterone production, renal function andcontraction of vascular smooth muscle (VSM) [6,7].

The syndrome of “septic shock”, by its very definition, ischaracterized by a reduction in blood pressure (hypotension).The host response prior to reaching the hypotensive state of

341S. Jain et al. / Biochimica et Biophysica Acta 1759 (2006) 340–347

“shock” is to orchestrate mechanisms to preserve blood pressure.Consistent with the now established theme of multiple, over-lapping and redundant cytokine-mediated pathways underlyingthe host response, mechanisms to maintain blood pressureinclude upregulation of AGT gene expression through the TNFand IL-1/NF-κB and the IL-6/STAT3 signaling pathways [8,9].Moreover, glucocorticoids, which are known to be beneficial inseptic shock [10,11], have a synergistic effect with IL-6/STAT3in the transcriptional upregulation of AGT transcription throughthe glucocorticoid receptor (GR)-binding site in the AGTpromoter [12]. In the present study, we asked whether IFN-γ, acytokine also induced in response to diverse infections, con-tributed towards maintenance of blood pressure by an effect onAGT gene expression. We report here that IFN-γ treatmentincreases transcription of the hAGT gene to an extent compa-rable to the effect of IL-6, and provide functional evidence thatthe signaling pathway involves the STAT1-binding site locatedbetween −271 and −279 in the AGT promoter.

2. Materials and methods

2.1. Cell culture

Human hepatoma cells (Hep3B) were grown as monolayers in Dulbecco'smodified Eagle's medium (DMEM) supplemented with 10% fetal calf serum,100 U/ml penicillin, and 100 μg/ml streptomycin in an atmosphere of 5% CO2.Hep3B cells were grown to confluence in 100 mm Petri dishes for RNApreparation and to 80% confluence in six well plates for transient transfections at37 °C. For the treatments, cultures were washed twice with phosphate-bufferedsaline (PBS), replenished with serum-free medium and then exposed to IFN-γ(10 ng/ml), IL-6 (10 ng/ml), and Dex (100 nM) alone or in combination fordifferent periods of time as indicated in respective figure legends.

2.2. Plasmid construction

The reporter construct pHAGT1300luc, and its deletion constructspHAGT303luc, pHAGT217luc and pHAGT103luc were constructed by PCRamplification of human AGT gene 5′ flanking region using TATGCTAGC-GAGGAGTCCCTATCTATAGGAACA, TATGCTAGCACACACCTAGGGA-GATGCTCCCGTTTCTGG, TATGCTAGCGCTCACTCTGTTCAGCAGT-GAAACTC and TATGCTAGCCAAGTGATGTAACCCTCCTCTCCAG asthe respective forward primers and CCGGCTCGAGATACCCTTCTGCTGT-AGTAC as the reverse primer. The amplified fragments contained the nucleotidesequence −1206 to +40, −303 to +40, −217 to +40 and −103 to +40respectively and subsequently were subcloned in the pGL3 basic vector that lackseukaryotic promoter and enhancer sequences (Promega, Madison, WI). Theplasmid 1.3lucIFNmut was obtained by site specific mutagenesis using theStratagene site-directed mutagenesis kit (Stratagene, TX, USA). The oligonu-cleotide sequence used for mutation was CGTCGCTGGGTGCCT (mutatednucleotides are italicized). Reporter constructs (950M4)4-luc and (950M1)4-lucwere constructed by tetramerization of oligonucleotide CGTTTCTGGGAACCTand CGTCGCTGGGTGCCT respectively and blunt ended ligation of thesemultimers in the SmaI site of pGL3 promoter vector.

2.3. Transient transfections

Transfections in Hep3B cell cultures in 6-well plates were carried out usingthe LipofectAMINE reagent (Qiagen, Velencia, CA) and the manufacturer'sprotocol. Briefly, 250 ng of reporter constructs and 50 ng of RSV β-gal was usedin each experiment. Four hours later the medium was changed. After 24 h oftransfection, cells were treated for an additional 6 h with IFN-γ (10 ng/ml), IL-6(10 ng/ml), and Dex (100 nM) alone or in combination as indicated in respectivefigure legends. Whole cell extracts were prepared by extraction with 200 μl oflysis buffer (Promega, Madison, USA). An aliquot of the cell extract was used to

measure luciferase activity using a luciferase assay system (Promega) andTurners Design Luminometer TD 20/20 as described by the manufacturer.Luciferase activity was normalized with the β-gal activity. β-gal activity wasdetermined as described previously [13,14].

2.4. Gel mobility shift assay

The probes for electrophoretic mobility shift assay (EMSA) were chemicallysynthesized, annealed and radiolabeled at the 5′-ends by polynucleotide kinaseusing [γ-32P] ATP. DNA fragments (20,000–50,000 cpm), 1–2 μg of poly(dI–dC), and 5–10 μg of the nuclear extracts were incubated in a solutioncontaining 10 mM HEPES (pH 7.5)–50 mM KCl–5 mM MgCl2–0.5 mMEDTA–1 mM DTT–12.5% glycerol on ice for 30 min and separated byelectrophoresis through a 5% polyacrylamide gel in the cold room. After 2–3 h,the gel was dried under vacuum and protein–nucleic acid complexes wereidentified by autoradiography. For supershift assay, 1 μl of antibodywas added tothe reaction mixture that was incubated for 30 min and analyzed by EMSA.Radioactive oligonucleotides were purified by Chroma spin columns (BDBiosciences, CA). Nuclear extracts for gel mobility shift assays were prepared bymodification of a previously described method [15]. Antibodies against STAT1were purchased from Santa Cruz Biotechnologies (Santa Cruz, CA).

2.5. Oligonucleotides

Double stranded oligonucleotides containingwild type (950M4) andmutated(950M1) IFN-γ responsive sequences were obtained by annealingCGTTTCTGGGAACCT and CGTCGCTGGGTGCCT with their respectivecomplementary oligonucleotides. An oligonucleotide containing consensusSTAT1 binding site (SIE oligo GTGCATTTCCCGTAAATCTTGTCTACA)was purchased from Santa Cruz Biotechnology, CA, USA. Two non specificoligonucleotides (NS1 and NS2) were obtained by annealing oligonucleotidesCTAGTATTATTGACTTAGGATC and ATTCGATCGGGGCGGGGCGACCwith their respective complimentary oligonucleotides.

2.6. Real-time-PCR (TaqMan PCR)

RNA used for TaqMan PCR was prepared from cells using the RNA easyminiprep kit from Qiagen (Qiagen Sciences, MD), followed by PCR with theTaqMan Universal PCR Master mix. The primers used were 5′CTTCACTGA-GAGCGCCTGC3′ as forward and 5′GAGACCCTCCACCTTGTCCA3′ asreverse primer for human AGT. Human glyceraldehyde-3-phosphate dehydro-genase (GAPDH) was chosen as the endogenous control gene [16]. We used theTaqMan Universal Master Mix and the TaqMan reagents for GAPDH. The twostep PCR conditions for both hAGTand GAPDHwere 2 min at 50 °C, 10 min at95 °C, 40 cycles with 15 s at 95 °C and 1 min at 60 °C. Threshold cycle numbers(Ct) were determined with sequence detector software (version 1.6; AppliedBiosystems) and transformed using theΔCt orΔΔCt method as described by themanufacturer. Results were expressed as fold induction by normalizing the datato the control conditions as previously described [16].

2.7. ChIP assays

The chromatin immunoprecipitation (ChIP) assay was performed using theChIP assay kit from Upstate Biotechnology (Lake Placid, N.Y.). Cells (3–4 million) were plated in 100 mm plates and after 24 h were treated with IFN-γfor 15–20 min in serum free medium. They were then fixed with 1%formaldehyde for 30 min, washed with chilled PBS, scraped and collected in1.5 ml Eppendorf tube followed by their lysis. The DNA was fragmented bysonication and 10 μl of the chromatin solution was saved as input. A 1 μg amountof anti-STAT1 antibody or rabbit immunoglobulin G was added to the tubescontaining 900 μl of chromatin solution and incubated overnight at 4 °C. Theantibody complexeswere capturedwith proteinA-agarose beads and subjected toserial washes (as described in manufacturer's protocol). The chromatinfraction was fur- ther reverse crosslinked at 65 °C for 4–6 h. The DNAwas thenpurified using Qiagen miniprep columns. The immunoprecipitated DNA(5 μl) and the input DNA (5 μl) were subjected to 35 cycles of PCR using−314AGT For (5′-CTCAGGCTGTCACACACCTA-3′) as a forward and

342 S. Jain et al. / Biochimica et Biophysica Acta 1759 (2006) 340–347

−82AGT Rev (5′-GAGAGGAGGGTTACATCA-3′) as a reverse primer. Thisamplification resulted in 234 bp region spanning the STAT binding site (−314 to−82) of the proximal promoter of human AGT gene. A control PCR (negativecontrol) was also performed from a non specific region, away from the STATsiteusing forward primer 5′ TATGCTAGCGAGGAGTCCCTATCTATAGGAACA3′ (AGTlucFor) and reverse primer 5′ GGGCATGACAGAGACCTTGG 3′(793AGT Rev) to show the specificity of STAT site in the CHIP assay. PCRproducts were then analyzed on an agarose gel.

2.8. Statistical analysis

The paired t-test was used to compare relative luciferase activities of reporterconstructs in transient transfection experiments. All transient transfectionsexperiments were performed in triplicate in three independent experiments(n=9). Additionally we used 1-Way ANOVA for multiple group comparisons.Statistically significant results are marked by an asterisk (P<0.05).

3. Results

3.1. Interferon-γ treatment increases the AGT mRNA level inhuman liver cells

In order to examine the effect of IFN-γ on AGT mRNA levelin human liver cells, we treated Hep3B cells in the absence andpresence of IFN-γ for 4, 12 and 24 h. The AGT mRNA levelswere quantitated by real time RT-PCR using the Taqmanprocedure. Results of this experiment (Fig. 1) showed that 12 hof IFN-γ treatment increased the AGT mRNA level by about2.3 fold and 24 h treatment increased the mRNA level by about2.45 fold to the basal level (mean of three experiments).

3.2. Interferon-γ treatment increases the promoter activity of areporter construct containing 1.2 kb of the 5′-flanking region ofthe hAGT gene

We next examined the effect of IFN-γ treatment on thetranscriptional activity of reporter constructs containing hAGTgene promoter. For this purpose, we performed transient trans-

Fig. 1. IFN-γ increases AGT mRNA levels in human liver cells. Hep3B cellswere treated in the absence or in presence of IFN-γ for 4, 12 and 24 h. The AGTmRNA levels were quantitated by real time RT-PCR using Taqman procedure.The bars shows the AGT mRNA levels on IFN-γ treatment at different timeintervals (mean±S.E. of three experiments). Asterisks denote P<0.05.

fection of reporter construct pHAG1.3-luc (containing 1223 bpof the 5′-flanking region and 23 bp of the first exon of hAGTgene attached to the luciferase gene) in Hep3B cells. In order todetermine the optimum dose of IFN-γ in transfection experi-ments, we used 5, 10, and 20 ng of IFN-γ/ml and analyzed thepromoter activity after 24 h of treatment. Results of theseexperiments (Fig. 2A) showed that 10 ng/ml of IFN-γ treatmentincreased the promoter activity of phAGT1.3-luc by about four-fold and therefore this dose was used in future transfectionexperiments.

3.3. Nucleotide sequence located between −103 and −303 ofthe hAGT gene promoter contains IFN-γ responsive element(IFNRE)

In order to identify the hAGT gene promoter sequence that isresponsible for IFN-γ induced promoter activity, we synthe-sized deletion constructs containing different regions of thehAGT gene promoter and used them in transient transfections.Results of this experiment (Fig. 2B) showed that IFN-γtreatment induced the promoter activity of the deletion constructcontaining 303 bp of the hAGT gene promoter by about 4.5-foldbut did not increase the promoter activity of a reporter constructcontaining 103 bp of the hAGT gene promoter. There is anotherputative STAT1 binding site (TTCCTGGAA) located between−163 and −173 that may be contributing to 3 fold induction ofpromoter activity of another deletion construct containing 217bp of the hAGT gene after IFN treatment. However thissequence also contains a consensus Ets binding site and alonewas not responsive to the effect of IFN-γ (data not shown).Taken together, results of these experiments suggested that IFN-γ-responsive element (IFNRE) is located between −103 and−303 bp of the 5′-flanking region of the hAGT gene.

3.4. Nucleotide sequence located between −271 and −279 ofthe hAGT gene contains the IFNRE

Nucleotide sequence of the hAGT gene promoter locatedbetween −271 and −279 has homology with STAT1 binding site(Fig. 3A) and since IFN-γ is known to activate gene tran-scription through STAT1, we performed gel shift assays usingradiolabeled oligonucleotides M4 (containing STAT1 bindingsite) and M1 (containing mutated STAT1 binding site). The gelshift assay was performed using untreated and IFN-γ treatedHep3B nuclear extract. Results of this experiment are shown inFig. 3B. The radiolabeled oligonucleotide M4 produced a newcomplex (marked with an arrow) in the presence of IFN-γtreated nuclear extract as compared to the untreated nuclearextract (compare lanes 5 and 3). This complex was not formedwhen mutated radiolabeled oligonucleotides M1 was used(compare lanes 12 and 4). The complex obtained in the presenceof IFN-γ treated extract (lane 5) was competed out in thepresence of a 10-fold excess of cold oligonucleotide M4 (lane 6)but not in the presence of mutated oligonucleotideM1 (lane 7) orother non specific oligonuleotides NS1 and NS2 (lanes 8 and 9).The intensity of this complex was also reduced in the presence ofSTAT1 antibody (lane 10) but not in the presence of pre-immune

Fig. 2. IFN-γ increases the promoter activity of a reporter construct containing1.2 Kb hAGT promoter or its deletion constructs. (A) Reporter constructpHAG1.3-luc (containing 1223 bp of the 5′-flanking region and 23 bp of the firstexon of hAGT gene attached to the luciferase gene) was transfected in Hep3Bcells. Luciferase activity wasmeasured after treating the cells with different dosesof IFNγ. Graph shows the promoter activity of phAGT1.3-luc. (B) Deletionconstructs containing different regions of the hAGT gene promoter weresynthesized and used in transient transfections. The graph shows that IFNγtreatment induced the promoter activity of the deletion construct containing up to−217 bp but did not increase the promoter activity of a reporter constructcontaining 103 bp of the hAGT gene promoter. Luciferase activity (mean±S.E.;n=9; normalized with respect to β-galactosidase as a transfection control) isexpressed in terms of that observed in untreated cultures. Asterisks denoteP<0.05.

343S. Jain et al. / Biochimica et Biophysica Acta 1759 (2006) 340–347

sera (PIS) (lane 11). We also performed a control gel shift assayusing radiolabeled SIE oligonucleotide (that contains thepreviously characterized STAT-1 binding site) and Hep3Bnuclear extract. The SIE oligonucleotide formed a major com-plex with the same mobility as M4 (lane 15) in the presence ofIFN-γ treated extract. This complex was not formed whenuntreated Hep3B nuclear extract was used (lane 14). The com-plex obtained with the SIE oligonucleotide and IFN-γ treatedextract (lane 15) was competed out with a 10 fold excess of coldSIE oligonucleotide (lane 16); 10 and 20 fold excess of cold M4oligonucleotide (lanes 17 and 18); but not with a 20 fold excessof mutated oligonucleotide M1 (lane 19). The intensity of thecomplex formed with SIE oligonucleotide was also reduced inthe presence of STAT1 antibody (lane 20) but not in the presenceof PIS (lane 21). Results of this experiment suggested that theIFNRE located between −271 and −279 of the hAGT gene

promoter binds to STAT1 which is activated by IFN-γ treatmentof Hep3B cells.

3.5. Chromatin immunoprecipitation assay shows that IFN-γtreatment increases the binding of STAT1 to IFNRE of thehAGT gene in Hep3B cells

In order to confirm that IFN-γ treatment increases theexpression of the hAGT gene in Hep3B cells through increasedbinding of STAT1 with the IFNRE between −271 and −279, weperformed chromatin immunoprecipitation (ChIP) assay. Thisassay investigates the interaction of transcription factors withtarget gene promoters in native chromosomal DNA. The protein:DNA complex derived from either control Hep3B cells or IFN-γ- treated Hep3B cells was immunoprecipitated by anti-STAT1or anti-STAT3 antibodies. The protein:DNA complex was thenreverse cross-linked and amplified using hAGT promoterspecific primers. The PCR-amplified product was then analyzedby gel electrophoresis. Results of this experiment (Fig. 4)showed that STAT1 bound 3 fold more strongly to the hAGTgene promoter in IFN-γ-treated Hep3B cells (Panel I, lane 2) ascompared with untreated cells (Panel I, lane 1). On the otherhand no significant difference was observed when anti-STAT3antibody was used in immunoprecipitation of the DNA:proteincomplex of IFN-γ-treated or untreated Hep3B cells (Panel II,lanes 1 and 2). Panel III shows a control experiment demon-strating that equal amount of input DNA was used for ampli-fication from control and IFN-γ-treated cells. Panel IV showsanother control verifying that almost negligible amplificationwas detected in the absence of anti-STAT1 antibody. Panel Vshows a negative control (no amplification) from an upstreamregion away from the STAT binding site. All the Chip resultswere normalized to the corresponding input values.

3.6. IFN-γ increases the expression of a reporter constructcontaining four copies of hAGT IFNRE as compared to thesame reporter construct containing mutated IFNRE

In order to confirm that the nucleotide sequence locatedbetween −271 and −279 of the hAGT gene promoter is indeedan IFNRE, we synthesized reporter constructs where four copiesof either oligonucleotide M4 or M1 were attached upstream ofthe luciferase gene in pGL3 promoter vector. These reporterconstructs were then transiently transfected in Hep3B cells andthe cultures treated with IFN-γ. Results of this experiment (Fig.5A) showed that whereas the promoter activity of M4-luc wasincreased by about 48-fold on IFN-γ treatment, the promoteractivity of M1-luc did not increase as a result of this treatment.

3.7. Mutation of IFNRE in pHAGT1.3-luc reduces IFN-γinduced promoter activity on transient transfection inHep3B cells

In order to show that IFNRE in the context of complete 1.3 kbpromoter of the hAGT gene is responsible for IFN-γ-inducedpromoter activity in human liver cells, we mutated IFNRE in thereporter construct pHAGT1.3-luc by site-specific mutagenesis

Fig. 3. Nucleotide sequence located between −271 and −279 of the hAGT gene contains the IFNRE. (A) Homology between −271 and −279 bp of the hAGT genepromoter with STAT-1 binding site. (B) Gel shift assay using radio-labeled oligonucleotides M4 (containing STAT-1 binding site) and M1 (containing mutated STAT-1binding site). The gel shift assay was performed in the presence of Hep3B nuclear extract (with and without IFN-γ treatment). Lanes 1 and 2 show the free radiolabeledM1 andM4 probes in the absence of nuclear extract. Lanes 3 and 4 show the complexes obtained with oligonucleotides M1 andM4 in the presence of untreated Hep3Bcell extract. Lane 5 shows the complex obtained with oligonucleotide M4 in the presence of IFN-γ treated Hep3B cell extract (the major complex is marked with anarrow), lanes 6 to 9 show competition of this complex with M4, M1 and non specific oligonucleotides NS1 and NS2 respectively; lanes 10 and 11 show the gel shiftassay in the presence of STAT-1 antibody and PIS respectively. Lane 12 shows the binding of M1 oligonucleotide using nuclear extract from IFN-γ treated Hep3Bcells. Lane 13 shows radiolabeled SIE probe in the absence of nuclear extract; lane 14 shows the complex formed with SIE oligonucleotide in the presence of untreatedHep3B cell extract; lane 15 shows the complex obtained with SIE oligonucleotide in the presence of IFN-γ treated cell extract, lanes 16 to 19 show competition of thiscomplex using cold SIE, M4 and M1 oligonucleotides; lanes 20 and 21 show the gel shift assay in the presence of STAT-1 antibody and PIS respectively.

344 S. Jain et al. / Biochimica et Biophysica Acta 1759 (2006) 340–347

as described in Materials and methods. The resulting mutatedconstruct mutpHAGT1.3-luc along with the wild-type reporterconstruct was then transiently transfected in Hep3B cells, treatedin the absence or presence of IFN-γ and promoter activity wasanalyzed. Results of this experiment (Fig. 5B) showed thatwhereas the promoter activity of pHAGT1.3-luc was increased

by 5-fold on IFN-γ treatment, the promoter activity ofmutpHAGT1.3-luc was increased only about 1.5-fold. Takentogether results of this experiment confirmed that nucleotidesequence located between − 271 and − 279 is mainly responsiblefor IFN-γ-induced promoter activity of the hAGT gene inhuman liver cells.

Fig. 4. ChIP assay showing an increase in the binding of STAT1 to IFNRE of thehAGT gene after IFN-γ treatment of Hep3B cells. The figure shows the bindingof STAT1 to the hAGT gene promoter in IFN-γ treated Hep3B cells (Panel I, lane2) as comparedwith untreated cells (Panel I, lane 1). Panel II, lanes 1 and 2 showsbinding of STAT3 antibody with the DNA:protein complex of IFN-γ treated oruntreated Hep3B cells. Panel III shows a control experiment demonstrating thatequal amount of input DNAwas used for amplification from control and IFN-γtreated cells. Panel IV shows another control that almost negligible amplificationoccurred in the absence of anti-STAT1 antibody. Panel V shows a negativecontrol (no amplification) from an upstream region from the STAT binding site.

Fig. 5. Mutations in IFNRE downregulate the IFN-γ induced expression ofreporter constructs containing the wild type hAGT promoter or a multimerizedIFNRE. (A) Reporter constructs with four copies of either oligonucleotide M4 orM1 were attached upstream of the luciferase gene in pGL3 promoter vector andthen transiently transfected into Hep3B cells in the absence or presence ofIFN-γ. Bars in panel A show the promoter activity of M4-luc and M1-luc incomparison to their basal luciferase activity. (B) IFNRE was mutated in thereporter construct pHAGT1.3-luc by site-specific mutagenesis as described inMaterials and methods and then transiently transfected into Hep3B cells, in theabsence or presence of IFN-γ and promoter activity was analyzed. The stripedbars in panel B show the promoter activity of pHAGT1.3-luc and mutpHAGT1.3-luc after IFN-γ treatment in comparison to their basal activity.Luciferase activity (mean±S.E.; n=9; normalized with respect to β-galactosi-dase as a transfection control) is expressed in terms of that observed in untreatedcultures. Asterisks denote P<0.05.

345S. Jain et al. / Biochimica et Biophysica Acta 1759 (2006) 340–347

3.8. IFN-γ and IL-6 independently increase the expression ofthe reporter constructs containing IFNRE

In order to investigate the interaction of IFN-γ- and IL-6-induced pathways we treated Hep3B cells with IFN-γ and IL-6alone and in combination after transfecting them with the luci-ferase reporter constructs containing IFNRE. Fig. 6A and Brespectively shows the IFN-γ and IL-6 induction of holo-promoter with IFNRE and an IFNRE tetramer. With both re-porter constructs, induction with either IFN-γ and IL-6 alone orin combination provided near-maximal activation.

3.9. Dex does not further increase IFN-γ induced upregulationof the reporter constructs containing IFNRE

Previous studies have already shown an increase in IL-6-induced activation of the AGT promoter (and constructs con-taining the STAT3 binding site between −278 and −269) withDex treatment [12]. We carried out similar studies with IFN-γ-induced regulation of reporter constructs containing the IFNRE.Hep3B cells were transfected with respective reporter constructscontaining the IFNRE and after 24 h the cultures were treatedwith IFN-γ and/or IL-6 in the presence or absence of Dex. Fig.7A and B shows the results from these experiments for the 1.3 kbAGT promoter containing IFNRE and a tetramer of IFNRErespectively. The data show that while Dex further upregulatesIL-6 induced activity (confirming our previous studies) [12], thisis not observed with the combination of IFN-γ and Dex.

4. Discussion

The maintenance of blood pressure is a critical component ofthe host response to infection. A plethora of cytokines is released

into the circulation following acute infection and injury. Anemergent general principle is that these cytokines have over-lapping and redundant effects as part of the host's attempt todiversify protective mechanisms. Among these, IL-6 has alreadybeen shown to be an important upregulator of the AGT gene; thisupregulation takes place via STAT3 signaling [9]. TNF-α andIL-1 also upregulate rat AGT gene expression; this signalingtakes place via NF-κB [17]. In the present study we askedwhether IFN-γ could also contribute to themaintenance of bloodpressure by upregulating AGT gene expression.

We observed that IFN-γ increased the expression of AGTmRNA. In investigating the underlying mechanism we dis-covered a STAT1-binding site in the AGT promoter locatedbetween nucleotide −271 and −279. Reporter constructs con-taining this site could be upregulated by IFN-γ. In contrast,

346 S. Jain et al. / Biochimica et Biophysica Acta 1759 (2006) 340–347

mutations placed within this site, either in isolated site constructsor within the context of the holo-promoter, blocked the ability ofIFN-γ to enhance AGT promoter activity. Using ChIP assays weobserved the increased association of STAT1 in intact nucleiprepared from IFN-γ-treated Hep3B cells with the IFNREelement identified by us in the present study.

Glucocorticoids are well known for their therapeutic use inchronic inflammatory diseases. In acute inflammation, particu-larly sepsis, low dose glucocorticoid regimens have shown pro-mising protective results [10,18–22]. We have already shownthe role of glucocorticoids in synergism with IL-6 in theupregulation of the hAGT promoter [12]. In the present work weinvestigated the contribution of glucocorticoids to IFN-γinduced activation of the hAGT promoter. Our results show(a) the induction of the AGT promoter activity by IL-6 or IFN-γalone but no further upregulation by their combination, and (b)while Dex further upregulated IL-6 induced activation, thisphenotype was not observed in combination with IFN-γ. The

Fig. 6. IFN-γ and IL-6 equivalently increase the expression of theAGTpromoter.Hep3B cells were treated with IFN-γ and IL-6 alone and in combination aftertransfecting them with AGT promoter/luciferase constructs as indicated. PanelsA and B respectively show the IFN-γ and IL-6 induction of an IFNRE tetramerand the 1.2 kb holo AGT promoter. Luciferase activity (mean±S.E.; n=9;normalized with respect to β-galactosidase as a transfection control) is expressedin terms of that observed in untreated cultures. Asterisks denote P<0.05.

Fig. 7. Dex does not further upregulate IFN-γ induced activation of the AGTpromoter. Hep3B cells were transfected with the indicated AGT reporter con-structs and after 24 h treated with IFN-γ and IL-6 in presence or absence of Dex.Panels A and B show respectively the results from these experiments for atetramer of the IFNRE and for the 1.2 kb holo-AGT promoter. Luciferase activity(mean±S.E.; n=9; normalized with respect to β-galactosidase as a transfectioncontrol) is expressed in terms of that observed in untreated cultures. Asterisksdenote P<0.05.

data suggest that the IL-6/STAT3/GR and IFN-γ/STAT1 aredistinct alternate pathways to activate AGT gene expression.

The present data identify a new component of the hostresponse aimed at maintaining blood pressure in the face ofinfection and septic shock. Clearly, the latter condition ensueswhen the host's defence mechanisms to maintain blood pressureare overwhelmed by vasodilatory events. The discovery of theIFN-γ/STAT1 pathway leading to upregulation of AGT geneexpression, together with the previously characterized TNF/IL-1/NF-kappaB and IL-6/STAT3 pathways, highlights theoverlap and redundancy of cytokine pathways used by the hostto maintain blood pressure under such circumstances.

Acknowledgments

This work was supported by research grants HL49884 and59547 from NHLBI, and from Phillip-Morris Incorporated (toAK) and research grant HL 73301 (to PBS).

347S. Jain et al. / Biochimica et Biophysica Acta 1759 (2006) 340–347

References

[1] W.G. Walker, P.K. Whelton, H. Saito, R.P. Russell, J. Hermann, Relationbetween blood pressure and renin, renin substrate, angiotensin II,aldosterone and urinary sodium and potassium in 574 ambulatory subjects,Hypertension 1 (1979) 287–291.

[2] X. Jeunemaitre, F. Soubrier, Y.V. Kotelevtsev, R.P. Lifton, C.S. Williams,A. Charru, S.C. Hunt, P.N. Hopkins, R.R. Williams, J.M. Lalouel, et al.,Molecular basis of human hypertension: role of angiotensinogen, Cell 71(1992) 169–180.

[3] P. Corvol, X. Jeunemaitre, Molecular genetics of human hypertension: roleof angiotensinogen, Endocr. Rev. 18 (1997) 662–677.

[4] D.J. Campbell, J.F. Habener, Angiotensinogen gene is expressed anddifferentially regulated in multiple tissuesof the rat, J. Clin. Invest. 78(1986) 31–39.

[5] D.J. Campbell, J.F. Habener, Cellular localization of angiotensinogengene expression in brown adipose tissue and mesentery: quantificationof messenger ribonucleic acid abundance using hybridization in situ,Endocrinology 121 (1987) 1616–1626.

[6] K.K. Griendling, M. Ushio-Fukai, B. Lassegue, R.W. Alexander, Angio-tensin II signaling in vascular smooth muscle. New concepts, Hypertension29 (1997) 366–373.

[7] M.J. Peach, D.E. Dostal, The angiotensin II receptor and the actions ofangiotensin II, J. Cardiovasc. Pharmacol. 16 (Suppl. 4) (1990) S25–S30.

[8] M. Jamaluddin, T. Meng, J. Sun, I. Boldogh, Y. Han, A.R. Brasier,Angiotensin II induces nuclear factor (NF)-kappaB1 isoforms to bindthe angiotensinogen gene acute-phase response element: a stimulus-specific pathway for NF-kappaB activation, Mol. Endocrinol. 14 (2000)99–113.

[9] C.T. Sherman, A.R. Brasier, Role of signal transducers and activators oftranscription 1 and -3 in inducible regulation of the human angiotensino-gen gene by Interleukin-6, Mol. Endocrinol. 15 (2001) 441–457.

[10] R.A. Balk, Steroids for septic shock: back from the dead? (Pro), Chest 123(2003) 490S–499S.

[11] D. Keh, T. Boehnke, S. Weber-Cartens, C. Schulz, O. Ahlers, S. Bercker,H.D. Volk, W.D. Doecke, K.J. Falke, H. Gerlach, Immunologic andhemodynamic effects of “low-dose” hydrocortisone in septic shock: adouble-blind, randomized, placebo-controlled, crossover study, Am. J.Respir. Crit Care Med. 167 (2003) 512–520.

[12] S. Jain, Y. Li, S. Patil, A. Kumar, A single-nucleotide polymorphism inhuman angiotensinogen gene is associated with essential hypertension andaffects glucocorticoid induced promoter activity, J. Mol. Med. 83 (2005)121–131.

[13] C.S. Narayanan, Y. Cui, A. Kumar, DBP binds to the proximal promoterand regulates liver-specific expression of the human angiotensinogen gene,Biochem. Biophys. Res. Commun. 251 (1998) 388–393.

[14] S. Jain, X. Tang, C.S. Narayanan, Y. Agarwal, S.M. Peterson, C.D. Brown,J. Ott, A. Kumar, Angiotensinogen gene polymorphism at −217 affectsbasal promoter activity and is associated with hypertension in African-Americans, J. Biol. Chem. 277 (2002) 36889–36896.

[15] J.D. Dignam, R.M. Lebovitz, R.G. Roeder, Accurate transcriptioninitiation by RNA polymerase II in a soluble extract from isolatedmammalian nuclei, Nucleic Acids Res. 11 (1983) 1475–1489.

[16] K. Gorzelniak, S. Engeli, J. Janke, F.C. Luft, A.M. Sharma, Hormonalregulation of the human adipose-tissue renin–angiotensin system:relationship to obesity and hypertension, J. Hypertens. 20 (2002) 965–973.

[17] A.R. Brasier, J. Li, Mechanisms for inducible control of angiotensinogengene transcription, Hypertension 27 (1996) 465–475.

[18] A. Eklund, A. Leppaniemi, E. Kemppainen, V. Pettila, Vasodilatory shockin severe acute pancreatitis without sepsis: is there any place forhydrocortisone treatment? Acta Anaesthesiol. Scand. 49 (2005) 379–384.

[19] D. Annane, V. Sebille, C. Charpentier, P.E. Bollaert, B. Francois, J.M.Korach, G. Capellier, Y. Cohen, E. Azoulay, G. Troche, P. Chaumet-Riffaut, E. Bellissant, Effect of treatment with low doses of hydrocortisoneand fludrocortisone on mortality in patients with septic shock, JAMA 288(2002) 862–871.

[20] J. Briegel, H. Forst, M. Haller, G. Schelling, E. Kilger, G. Kuprat, B.Hemmer, T. Hummel, A. Lenhart, M. Heyduck, C. Stoll, K. Peter, Stressdoses of hydrocortisone reverse hyperdynamic septic shock: a prospective,randomized, double-blind, single-center study, Crit. Care Med. 27 (1999)723–732.

[21] P.E. Bollaert, C. Charpentier, B. Levy, M. Debouverie, G. Audibert, A.Larcan, Reversal of late septic shock with supraphysiologic doses ofhydrocortisone, Crit. Care Med. 26 (1998) 645–650.

[22] C.L. Sprung, P.V. Caralis, E.H. Marcial, M. Pierce, M.A. Gelbard, W.M.Long, R.C. Duncan, M.D. Tendler, M. Karpf, The effects of high-dosecorticosteroids in patients with septic shock. A prospective, controlledstudy, N. Engl. J. Med. 311 (1984) 1137–1143.