BIOLOGY OF REPRODUCTION 75, 697–704 (2006)Published online before print 19 July 2006.DOI 10.1095/biolreprod.106.053439

The Interleukin 1beta-Induced Expression of Human Prostaglandin F2alpha

ReceptorMessenger RNA in Human Myometrial-Derived ULTR Cells Requires the TranscriptionFactor, NFkappaB1

Dean B. Zaragoza,2,3 Robyn R. Wilson,3 Bryan F. Mitchell,3 and David M. Olson3,4,5

Departments of Obstetrics & Gynecology,3 Pediatrics,4 and Physiology,5 Perinatal Research Centre, University ofAlberta, Edmonton, Alberta, Canada T6G 2S2

ABSTRACT

The molecular mechanisms that regulate the expression ofgenes involved in parturition are poorly understood. The mRNAexpression of the prostaglandin F2alpha receptor (PTGFR), auterine activating gene, is increased at labor and is required foruterine contractile activity in numerous animal models, althoughthe signaling pathways responsible for this increased expressionhave not been identified. Proinflammatory cytokines have beenproposed to regulate the expression of the uterine activatinggenes via activation of the nuclear transcription factor,NFkappaB, and initiate labor. However, it is uncertain whetheruterine PTGFR is regulated this way. In this report, wedemonstrate for the first time that treatment of immortalizedhuman myometrial-derived ULTR cells with the proinflamma-tory cytokine IL1beta causes an increase in PTGFR mRNA levels.Furthermore, IL1beta treatment increased the nuclear levels ofthe RELA subunit of NFkappaB and increased binding of RELA tothe NFkappaB DNA-binding site. Inhibition of NFkappaBactivation with either the proteasome inhibitor MG132 orphenethyl caffeiate reduced PTGFR mRNA levels, whichindicates that this transcription factor is important for basaltranscription. Furthermore, this inhibition prevented IL1betainduction of PTGFR mRNA, which confirms that NFkappaB isrequired for the IL1beta-induced increase in PTGFR. Theseresults are consistent with the proposal that proinflammatorycytokines directly regulate uterine activation genes and that thetranscription factor NFkappaB is involved in both basal andIL1beta-stimulated transcription of the PTGFR gene.

cytokines, gene regulation, parturition, signal transduction, uterus

INTRODUCTION

Prostaglandin F2a (PGF

2a) is involved in several reproduc-tive processes, including a number of physiological eventsassociated with parturition. Noort et al. [1] have reported anassociation between elevated human PGF

2a plasma levels andplacental separation. In cows with retained fetal membranesdue to the non-separation of the fetal and maternal cotyledons,placental tissue concentrations of PGF

2a were found to besignificantly lower than in normally separated cotyledons at 6 hafter parturition [2]. In the human deciduas, PGF

2a stimulatesthe production of MMP2 and MMP9 and decreases the

production of the progesterone receptor isoforms A, B, and C[3, 4], and these actions indirectly lead to the termination ofpregnancy. Uterine involution is very closely associated withplasma levels of PGF

2a in cows and sheep [5, 6], and theadministration of PGF

2a promotes uterine involution in cows[7]. However, its best known function is to stimulate thecontractions of labor, since the myometrium contracts inresponse to exogenous PGF2a both in vivo and in vitro [8–10].

PGF2a action is mediated by its receptor, PTGFR, which is a

seven transmembrane, Gaq protein-coupled receptor. PTGFR isone of several uterine-activating genes involved in preparingthe uterus for the contractions of labor and other labor-associated functions. It is found in all the intrauterine tissues[11], which suggests different biological roles in each tissue,and in line with that notion, recent studies have revealedconsiderable regulation of the levels of PTGFR in the non-pregnant, pregnant, and parturient uterus. In human, rat, andmouse studies, myometrial PTGFR mRNA was found to beelevated at term and/or preterm birth, and specific antagonismof its action in sheep and mice delayed preterm birth andprolonged gestation [12–16]. Interestingly, studies have shownthat PTGFR mRNA expression is decreased in the myometri-um of nonlaboring pregnant women compared to nonpregnantwomen. This result is consistent with both the presence ofrepressor and enhancer regions in the promoter [17], andsuggests that PTGFR plays a role in pregnancy maintenanceand termination. In the mouse, elevated levels of PTGFRmRNA without a concomitant increase in PGF

2a are associatedwith preterm birth [14, 18, 19]. However, these changes are notexclusive to the myometrium; recently, we observed anincrease in PTGFR mRNA and protein expression in humandecidua at term delivery [20]. Therefore, the uterine tissuelevels of PTGFR mRNA are related to pregnancy maintenanceand termination. The question remains as to how PTGFRexpression is regulated.

The finding that the levels of proinflammatory cytokinesincrease in the amniotic fluid and maternal serum at the onsetof term and preterm labor [21–26] has led to studies showingtheir involvement in the regulation of uterine gene expressionat the time of labor [27, 28]. Only one paper has suggested alink between a cytokine and PTGFR by showing thatinterleukin (IL) 1B stimulates PTGFR mRNA expression after24 h of exposure to cultured human granulosa-luteal cells [29].

Considerable evidence accumulated since 1999 has shownthat the nuclear transcription factor nuclear factor kappa B(NFKB) is involved in many aspects of PG synthesis and actionin intrauterine tissues. The NFKB family of transcription factorsis associated with inflammation and can be activated by the pro-inflammatory cytokines associated with labor to affect PGsynthesis in uterine-derived tissues and cells. This process maybe mediated by tumor necrosis factor [30, 31], IL1B [27], andlipopolysaccharide [32], although this latter effect is not always

1Supported by the Canadian Institutes for Health Research (The Institutefor Human Development, Child and Youth Health) and the CIHR Groupfor Perinatal Health and Disease.2Correspondence: FAX: 780 492 1308; e-mail: dzaragoz@ualberta.ca

Received: 25 April 2006.First decision: 19 May 2006.Accepted: 10 July 2006.� 2006 by the Society for the Study of Reproduction, Inc.ISSN: 0006-3363. http://www.biolreprod.org

697

consistent [33]. It appears that NFKB mediates IL1B action atseveral levels of the PG synthesis-receptor cascade, includingsecretory type II phospholipase A

2[33] and PG endoperoxide H

synthase (PTGS, cyclooxygenase)-2 [27, 31, 34].The possibility that a proinflammatory cytokine can regulate

PTGFR transcriptional regulation via NFKB is high, since wehave identified NFKB-binding sites at four separate positions inthe human PTGFR promoter: at �3835 and �3017 (unpub-lished) and within intron 1 at positions 744 and 777 [17]. Thepurpose of this present study was to test the possibility that theproinflammatory cytokine IL1B regulates PTGFR mRNAexpression through NFKB activation in the immortalizedhuman myometrial-derived ULTR cell line, which has beenpreviously utilized for studying IL1B signaling [27, 30].

MATERIALS AND METHODS

Cell Culture

ULTR cells were provided by Dr. J.K. McDougall’s laboratory (FredHutchison Cancer Research Centre, Seattle, WA) [35]. Cells were cultured inDulbecco Modified Eagle Medium (DMEM) that contained 10% fetal bovineserum (Invitrogen, Burlington, ON, Canada) and 13 antimycotic (100 U/mlpenicillin G sodium, 100 lg/ml streptomycin sulfate, 0.25 lg/ml amphotericinB) at 378C in 5% CO

2. Cells were grown in either T75 flasks for RNA and

nuclear extractions or in 8-chamber slides for immunofluorescence. All the cellculture experiments utilized cells from passage 17 to 25. Following growth to;80% confluence, the growth medium was aspirated and the cells were washedthree times with phosphate-buffered saline (PBS). For treatments, the cells werecultured in serum-free medium (DMEM with 13 antimycotic) for an additional24 h at 378C in 5% CO

2. Following starvation, the cells were cultured in serum-

free medium that contained either 0.6 nM (10 ng/ml) IL1B or vehicle (H2O).

The inhibition experiments involved pretreating cells with or without 17.6 lM(5 lg/ml) phenethyl caffeiate (CAPE) or 5 and 10 lM MG132 (EMDBiosciences, Calbiochem, San Diego, CA) 1 h prior to treatment with IL1B(0.585 nM) or vehicle control (H

2O). The duration of IL1B or vehicle treatment

is indicated (1, 3, 6, 12 or 24 h) in the figure legends.

RNA Extraction

Total RNA was extracted with Trizol reagent (Invitrogen, Burlington, ON,Canada) as per the manufacturer’s instructions. Cells were lysed with Trizol(3.5 ml/flask) and incubated at room temperature for 5 min. To assess variationintroduced by the RNA extraction procedure, each flask was divided into threeseparate RNA isolations by transferring 1 ml of the lysed cells into threeseparate microfuge tubes. Following the addition of 200 ll chloroform, thesamples were shaken vigorously by hand for 30 sec, incubated at roomtemperature for 2 min, and centrifuged at 12 000 3 g for 15 min at 48C. Theaqueous layer was transferred to new tubes and RNA was precipitated with anequal volume of isopropanol. RNA was precipitated by incubation for 1 h at�208C, followed by centrifugation at 12 000 3 g for 10 min at 48C. The RNApellets were washed with 75% ethanol (1 ml) and resuspended in 1 mM sodiumcitrate solution (pH 6.4) (Ambion, Austin, TX). The RNA was treated withDNase using the Turbo DNAfree Kit (Ambion), to remove contaminatinggenomic DNA. RNA purity and concentration were determined by A

260and

A280

measurements, and sample integrity was verified by denaturing agaroseelectrophoresis.

Nuclear Protein Extractions

Following treatment, nuclear proteins were extracted as described bySchreiber et al [36]. Cells were washed three times with PBS and scraped into1.5-ml microfuge tubes. Cells were pelleted by centrifugation at 1500 3 g for 3min at 48C. The supernatant was removed and the pellet was gentlyresuspended in 400 ll of ice-cold buffer A [10 mM Hepes (pH 7.9), 10 mMKCl, 0.1 mM EDTA, 0.1 mM EGTA, 1mM DTT, 0.5 mM PMSF, 10 llprotease inhibitor cocktail (Sigma-Aldrich Canada Ltd., Oakville, ON, Canada)per 107 cells]. Resuspended cells were incubated on ice for 15 min, 10% NP-40was added, and the sample was vortexed for 10 sec. Samples were centrifugedfor 30 sec at 14 000 3 g. The pellet was resuspended in 50 ll of ice-cold bufferC [20 mM Hepes (pH 7.9), 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mMDTT, 1 mM PMSF, and 25 ll/ml protease inhibitor cocktail]. Samples wereplaced on an orbital shaker for 20 min at 48C. Following centrifugation at 12000 3 g for 10 min, the supernatants were removed and assayed for protein

concentration by the Bradford Assay (Bio-Rad Laboratories, Mississauga, ON,Canada). Nuclear extracts were stored at �808C.

Immunofluorescence

Following treatment, cells were washed three times with PBS and fixed with500 ll/well fixation and permeabilization solution (3.7% formaldehyde, 0.5%Triton X-100) for 15 min at room temperature. Each chamber was washed twicewith 500 ll PBS and blocked in 1% BSA in 13 PBS for 30 min. The cells werethen incubated with the anti-RELA polyclonal antibody (Santa CruzBiotechnology, Santa Cruz, CA) at a dilution of 1:100 in blocking buffer for2 h at room temperature. Following two washes with PBS, the cells wereincubated with a 1:200 dilution of anti-rabbit IgG (rhodamine-conjugated) inblocking buffer for 30 min at room temperature in the dark. After PBS washingthree times in the dark, the gasket was removed from the slide and 50 ll of a 1:3dilution of Vectashield mounting solution with DAPI (Vector Laboratories,Burlingame, CA) was applied to the top edge of the slide. A coverslip uniformlydistributed the solution across the slide. Images were acquired with an OlympusBX40 fluorescent microscope equipped with a SPOT2 color digital camera.

Western Blotting

Nuclear extracts (5 lg) and a prestained protein ladder (MBI Fermentas,Burlington, ON, Canada) were electrophoresed on a 10% SDS-polyacrylamidegel and transferred to a nitrocellulose membrane by the standard methods. Themembrane was incubated overnight with blocking buffer (7% dried milk in0.1% TBS-Tween) at 48C on a platform rocker. The anti-NFKB polyclonalantibody (Santa Cruz Biotechnology) was diluted 1:1000 in blocking buffer(5% dried milk in 0.1% TBS-Tween) and incubated with the membrane for 2 hat room temperature with gentle rocking. The membrane was washed threetimes (10 min each) with 0.1% TBS-Tween and incubated with horseradishperoxidase-conjugated goat anti-rabbit IgG at a dilution of 1:4000 in blockingbuffer for 1.5 h at room temperature. The membrane was washed four times (15min each) with 0.1% TBS-Tween and proteins were detected with the ECLdetection reagents (Amersham Biosciences, Baie d’Urfe, QC, Canada) andexposure to Fuji RX x-ray film.

Electrophoretic Mobility Shift Assay (EMSA)

Binding of nuclear protein to the RELA DNA-binding site was analyzedusing the Nushift Kit (ActiveMotif, Carlsbad, CA). Each labeling reactioncontained 50 ng (2 ll) of wild-type oligonucleotide (50-AGCTTGGGG-TATTTCCAGCCG-30; the NFKB site is underlined), 2 ll of 103 kinasebuffer, 4 ll (40 lCi) of c[-32P]ATP (3000 Ci/mM), 1 ll (10 U) of T4polynucleotide kinase, and 11 ll of water. The reaction was incubated for 30min at 378C and stopped with the addition of 5 ll of 1% SDS/100 mM EDTA.The labeled probe was gel-purified on a 10% polyacrylamide gel and specificactivity was determined by scintillation counting. For each binding reaction, anextract/antibody premix was prepared that contained 11 lg nuclear extracts, 4ll Binding Buffer B (43), 2 ll Stabilizing Solution D, with or without 4 ll anti-RELA antibody, with or without 2 ll NFKB peptide, and water to a totalvolume of 16 ll. The mix was incubated for 30 min on ice. During theincubation, a probe premix was prepared for each reaction that contained 2 llBinding Buffer C2 (43), 1 ll Stabilizing Solution D, 100 000 cpm of labeledprobe, with or without 2 ll of cold wild-type oligonucleotide, with or without 2ll of mutant oligonucleotide (50-AGCTTGGCATAGGTCCAGCCG-30 (themutated NFKB site is underlined), and water to 8 ll. After 30 min of extract/antibody premix incubation, 8 ll of probe premix was added and the entirereaction was incubated for 30 min at 48C. The binding reaction was then loadedonto a 5% native polyacrylamide gel and electrophoresed in 13 TGE (25 mMTris base per 190 mM glycine per 1mM EDTA; pH 8.3) at 200 V for 2 h at 48C.Following electrophoresis, the gel was dried and analyzed by autoradiography.

Real-Time RT-PCR

Total RNA (1000 ng) was reverse-transcribed using Supercript II reversetranscriptase (Invitrogen) in a volume of 20 ll, using the protocol supplied bythe manufacturer. The cDNA obtained was used in subsequent PCR reactions.To exclude contamination of cDNA samples with genomic DNA, correspond-ing reactions, in which Superscript II was replaced with water (the No RTcontrols), were performed with randomly chosen samples. The followinghuman PTGFR-specific primers were designed from the sequence withaccession no. L24470: forward primer, 5 0-(1009)TCCTGTATTTGTTG-GAGCCCATTTCTGGTTAC(1040)-3 0; and reverse primer, 5 0-(1123)TCCATGTTGCCATTCGGAGAGCAAAAAG(1096)-30. The HPLC-purifiedhGAPDH primers were the same as those found in the Applied Biosystems

698 ZARAGOZA ET AL.

(Foster City, CA) GAPDH TaqMan Gold RT-PCR kit (P/N N808-0233;forward primer: 50-(1457)GAAGGTGAAGGTCGGAGTC(1476)-30; reverseprimer: 50-(3412)GAAGATGGTGATGGGATTTC(3392)-30.

Each 50-ll reaction contained 2 ll of cDNA, 25 ll of 23 SYBR greenMaster Mix (Applied Biosystems), 1 ll of forward primer and 1 ll of reverseprimer, 1 ll AmpErase (to control for PCR product contamination), and waterto a total reaction volume of 50 ll. Real time RT-PCR (iCycler; Bio-RadLaboratories) was performed under the following conditions: 2 min at 508C, 10min at 958C, followed by 40 cycles of 20 sec at 958C and 1 min at 608C. Tocontrol for the amplification of non-specific products, melt curve analysis wasperformed following amplification by measuring fluorescence while increasingthe temperature every 12 sec by 0.58C, from 558C to 958C. No amplification ofnon-specific products was observed with any set of primers. The PCR productsof each sample were checked by gel electrophoresis, and sequenced to verifythat the correct product was amplified.

Human PTGFR mRNA measurements were normalized to GAPDH mRNA.Standard curves for both PTGFR and GAPDH were generated by serial dilutionsof pooled cDNA samples. The amplification efficiency for each primer set wasdetermined by converting the slope of the standard curve using the algorithm E¼10–1/slope. For each gene, the mean threshold cycle (from triplicate reactions) wascorrected for the efficiency of the reaction and expressed relative to a controlsample for each experiment [37]. The human PTGFR levels were then expressedrelative to the GAPDH levels using the following formula:

Ratio ¼ ED CtðControl�SampleÞ

FP

ED CtðControl�SampleÞ

GAPDH

The mean of each treatment group was determined. The number of samples (N)for each treatment group is stated in the figure legends. The results for all themRNA measurements were analyzed by either a two-way analysis of variance(post-hoc test using the Holm-Sidak method) or the Student t-test when only twogroups were compared. Significance was achieved at P , 0.05.

RESULTS

Treatment of ULTR Cells with IL1B IncreasesPTGFR mRNA Levels

To examine whether IL1B regulates PTGFR mRNA levels,ULTR cells were first treated with IL1B (0.06 nM or 0.6 nM) for24 h and analyzed for PTGFR mRNA abundance by real-timeRT-PCR. Both doses of IL1B resulted in an increase in PTGFRexpression, but the greatest effect was seen with the 0.6 nM dose(data not shown). For this reason, the remaining experimentsused this dose. The levels of human PTGFR mRNA over 24 h inULTR cells treated with 0.6 nM IL1B are shown in Figure 1.Two-way ANOVA indicated significant effects of both time (P, 0.001) and treatment (P¼ 0.002) on PTGFR mRNA levels.Pairwise multiple comparisons procedures (Holm-Sidak meth-od) indicated that the 12 h controls were significantly differentfrom the 0 h controls (P , 0.05) but not the 6 h controls. The 24 hcontrols were significantly different from both the 0 h and 6 hcontrols (both P , 0.05). Furthermore, the 12 h and 24 htreatments were significantly different from both the 0 h controls(P , 0.001) and 6 h treatment (P , 0.001). Finally, the PTGFRmRNA levels were significantly higher in the treated cells at both12 h and 24 h than in the time-matched controls. Thus, IL1Bpositively regulates human PTGFR mRNA levels.

The RELA Subunit of NFKB Translocates to the Nucleusin ULTR Cells Treated with IL1B

We examined potential mediators of the IL1B effect onPTGFR expression. Previous studies have demonstrated thatNFKB is a primary mediator of the effects of this cytokine. Inaddition, we have previously identified potential NFKB siteswithin the PTGFR promoter, which suggests that this family oftranscription factors plays a role in modulating PTGFRtranscription in myometrial cells. Therefore, we examined byimmunofluorescence the localization of the RELA subunit ofNFKB in ULTR cells in response to IL1B treatment. Treatment

of ULTR cells with IL1B resulted in an increase in nuclearRELA at both 12 h and 24 h compared to control cells. Figure2a illustrates that the RELA fluorescent signal (red) matchedthe DAPI nuclear fluorescence (blue) (Fig. 2a, panels 5 and 6).In contrast, control cells had very little nuclear fluorescence buthad a rather strong cytoplasmic signal (Fig. 2a, panel 1). Thisresult was verified by analyzing the RELA protein levels byWestern blotting of nuclear extracts derived from control andIL1B-treated ULTR cells (Fig. 2b). The nuclear RELA levelswere higher in IL1B-treated ULTR cells (0.6 nM) than incontrol cells. Taken together, these results demonstrate thattreatment of ULTR cells with IL1B results in nucleartranslocation of the RELA subunit of NFKB.

Binding of RELA to the NFKB Response Element is Higherin ULTR Cells Treated with IL1B

We used the electrophoretic mobility shift assays (EMSA) todetermine whether IL1B treatment of ULTR cells also resultedin increased binding of RELA to the NFKB response element.Validation of the assay is shown in Figure 3a. Labeled unboundprobe that corresponded to the wild-type NFKB site is shown atthe bottom of each gel in all the lanes. Nuclear extracts fromboth IL1B-treated ULTR cells and TPA-treated Jurkat cells(control provided with the kit) exhibited binding activity to thelabeled NFKB probe, as shown by decreased mobility (shiftedband, Fig. 3a, lanes 1 and 6). In addition, while this shifted bandwas eliminated by including unlabeled wild-type probe in thebinding reaction (Fig. 3a, lane 2), it was not eliminated whenunlabeled mutant NFKB-binding site was included (Fig. 3a,lane 3). These results indicate that the binding activity isspecific for the wild-type NFKB site. To identify RELA as partof the binding complex, a polyclonal antibody to the RELAsubunit of NFKB was included in the binding reaction. Thisresulted in a further decrease in mobility (supershifted band,Fig. 3a, lanes 4 and 7). Finally, a peptide that corresponds to theeptitope of the RELA antibody was included in the bindingreaction. This peptide competed with the shifted complex forthe anti-RELA antibody and eliminated the supershifted band(Fig. 3a, lane 5). Therefore, the binding activity of the ULTRnuclear extracts consists of a complex that contains RELA.

FIG. 1. Human PTGFR mRNA levels in ULTR cells treated with vehicleor 0.6 nM (10 ng/ml) IL1B for 0, 1, 3, 6, 12, and 24 h (N ¼ 3 for eachgroup, mean 6 SEM). *, P , 0.001 compared with time-matched controls;**, P , 0.05 compared with 0 h controls; ***, P , 0.05 compared to 6 hwithin the same treatment group.

IL1beta REGULATES PTGFR mRNA THROUGH NFkappaB 699

Nuclear extracts from control and IL1B-treated (0.6 nM for24 h) ULTR cells were compared for binding to the wild-typeNFKB probe (Fig. 3b). Although there were no apparentdifferences in the intensities of the shifted bands between thetwo groups, there was a large increase in the anti-RELAantibody-shifted complex in the nuclear extracts derived fromtreated cells (supershifted band is undetectable in Fig. 3b, lanes2 and 4 compared with Fig. 3b, lanes 6, 8, and 10), whichindicates increased binding of RELA to the NFKB sitefollowing IL1B treatment. The fact that inclusion of the anti-RELA antibody did not completely eliminate the shiftedcomplex may indicate some nonspecific binding activity in theextracts, an inadequate amount of anti-RELA antibody in thebinding reaction, or the presence of other non-RELA-containing NFKB complexes. Nevertheless, these resultsdemonstrate that RELA binding to the NFKB response elementis elevated in ULTR cells treated with IL1B.

The Proteasome Inhibitor MG132 Prevents Increased RELANuclear Abundance with IL1B Treatment, and Reducesthe PTGFR mRNA levels in Both Control and IL1B-TreatedULTR Cells

To test whether the activation of the NFKB pathway byIL1B is responsible for the increase in PTGFR mRNA levels,we used the proteasome inhibitor MG132 to prevent nucleartranslocation of RELA. Inhibition of the 26S proteasome

prevents the degradation of NFKBI, thereby keeping thenuclear localization signal of the NFKB subunits masked [38].ULTR cells were pretreated with either MG132 (5 or 10 lM) orvehicle control for 1 h prior to treatment with either 0.6 nMIL1B or control for 24 h. Nuclear extracts were prepared andanalyzed for RELA abundance (Fig. 4a). While the IL1B-induced increase in RELA was observed without MG132treatment (Fig. 4a, lanes 1, 2, 5, and 6), preincubation of ULTRcells with 5 lM MG132 partially prevented this increase, andpreincubation with 10 lM MG132 completely prevented thisincrease (Fig. 4a, lanes 3, 4, 7, and 8). These results suggestthat MG132 prevents IL1B-induced nuclear translocation ofthe RELA subunit of NFKB.

We next examined whether inhibiting the IL1B-inducedincrease in nuclear RELA levels with MG132 also preventedthe IL1B-induced increase in PTGFR mRNA levels. ULTRcells were pretreated with either 10 lM MG132 or vehiclecontrol for 1 h prior to treatment with either 0.6 nM IL1B orcontrol for 12 or 24 h. RNA was extracted and the PTGFRmRNA levels were measured by real-time RT-PCR (Fig. 4b).Once again in the absence of MG132, IL1B treatmentincreased the PTGFR mRNA levels ;2-fold at 12 and 24 h,whereas pretreatment with the proteasome inhibitor decreasedthe PTGFR mRNA levels in both the control and IL1B-treatedcells. Furthermore, there was no significant difference betweenthe control and IL1B-treated cells when the cells werepretreated with MG132. This suggests that inhibition of the

FIG. 2. Immunofluorescence of ULTR cells. A) Fluorescent signals for RELA (red) and nucleus (blue) in ULTR cells treated with vehicle (panels 1–4) or 0.6nM (panels 5–8) IL1B for 24 h. Panels 1, 2, 5, and 6 represent immunofluorescence in the presence of the anti-RELA antibody, while panels 3,4, 7 and 8represent negative controls, in which the anti-RELA antibody was omitted. B) Western blot of RELA protein levels in nuclear extracts of ULTR cells treatedwith vehicle or 0.6 nM (10 ng/ml) IL1B for 24 h (N¼ 3 for each group).

700 ZARAGOZA ET AL.

NFKB pathway decreases PTGFR mRNA levels in controlcells and prevents IL1B induction of PTGFR mRNA levels.The reduction in PTGFR mRNA levels observed with MG132was not due to cell death, as trypan blue staining confirmed thattreatment of ULTR cells with either 5 or 10 lM MG132 did notsignificantly alter cell viability (data not shown).

Phenethyl Caffeiate (CAPE) Inhibits RELA Bindingto the NFKB Response Element in ULTR Cells Treatedwith IL1B and Reduces PTGFR mRNA Levels in BothControl and IL1B-Treated Cells

In order to address specifically the role of NFKB in theIL1B-induced upregulation of PTGFR, ULTR cells werepretreated with 17.6 lM (5 lg/ml) CAPE prior to treatmentwith IL1B, and assayed for RELA nuclear abundance, RELAbinding activity, and PTGFR mRNA levels. Figure 4cillustrates the nuclear RELA protein levels in these cells. Onceagain, the nuclear RELA levels increased with IL1B treatmentat both 12 h (Fig. 4c, lanes 1 and 2) and 24 h (Fig. 4c, lanes 5and 6). Pretreatment of ULTR cells with CAPE had no effecton these increases at either 12 h (Fig. 4c, lanes 3 and 4) or 24 h(Fig. 4c, lanes 7 and 8).

CAPE inhibited RELA binding to the NFKB responseelement at 12 h of IL1B treatment (Fig. 4d). Increases in theamount of the anti-RELA antibody supershifted complex weredetected with IL1B treatment at both 12 h (Fig. 4d, lanes 1 and2) and 24 h (Fig. 4d, lanes 5 and 6). However, pretreatmentwith CAPE prevented the IL1B-induced increase in thesupershifted complex at 12 h (Fig. 4d, lanes 3 and 4) but not24 h (Fig. 4d, lanes 7 and 8).

We next examined whether pretreatment with CAPE alsoprevented the IL1B-induced increase in PTGFR mRNA levels,as was the case with MG132. ULTR cells were pretreated witheither 17.6 lM (5 lg/ml) CAPE or vehicle control for 1 h priorto treatment with either 0.6 nM IL1B or control for 12 or 24 h.Again, in the absence of inhibitor, IL1B treatment increased thePTGFR mRNA levels ;2-fold at 12 and 24 h, whereaspretreatment with the binding inhibitor decreased the PTGFRmRNA levels in both the control and IL1B-treated cells (Fig.4b). In addition, there was no significant difference betweenthe control and IL1B-treated cells at either 12 h or 24 h whenthe cells were pretreated with CAPE. This suggests thatinhibition of NFKB binding decreases PTGFR mRNA levels incontrol cells and prevents IL1B induction of PTGFR mRNAlevels. Furthermore, the reduction in PTGFR mRNA levelsobserved with CAPE was not due to cell death, as trypan bluestaining confirmed that treatment of ULTR cells with 17.6 lMCAPE did not significantly alter cell viability (data not shown).

DISCUSSION

Several immortalized cell lines have been developed to aidin the study of uterine gene expression and function. TheULTR cell line maintains smooth muscle morphology,expresses smooth muscle a-actin [35], and retains many ofthe features of primary human myometrial cells. For example,these cells possess Ca2þ-activated K

þchannels, and similar to

primary cultured human myometrial cells and intact tissue,increase [Ca2þ]

iand inositol phosphate formation in response

to the uterotonin oxytocin [39–42]. These cells are well suitedto the study of cytokine signaling, as they have similarresponse as primary myometrial cells. For instance, IL1B andTNF stimulate the production of both PGE

2and 6-keto-PGF

1ain ULTR cells and in primary myometrial cells [43–45]. Theuse of this cell line has also uncovered important details of how

IL1B and interferon-c (IFNG) regulate PG synthesis throughNFKB [27, 30]. Based on these observations, as well as ourown observations that ULTR cells express many factorsthought to be important for regulating myometrial contractility(estrogen and progesterone receptors, oxytocin receptors,connexin-43), it is evident that ULTR cells provide areasonable first model for investigating the role of NFKB inthe regulation of PTGFR expression by IL1B in uterine tissues.

In amnion cells, the NFKB sites on the PTGS2 promoter arecritical for IL1B-stimulated PTGS2 expression [46], and in themyometrium, NFKB is necessary for the IL1B-inducedincrease in PTGS2 expression [27, 47]. Furthermore, laborhas been shown to be associated with increased NFKB activityin the amnion, where it is proposed to act as an antagonist ofthe progesterone receptor [34]. An inhibitor of NFKB, SN-50,

FIG. 3. A) Verification of EMSA of nuclear extracts from ULTR cellstreated with 0.6 nM IL1B for 24 h (lanes 1–5) and a control extract isolatedfrom Jurkat cells treated with TPA (lanes 6 and 7). B) DNA-binding activityof RELA in ULTR cells treated with vehicle (lanes 1–4) or 0.6 nM IL1B for24 h (lanes 5–10). Binding reactions were performed with or without theanti-RELA antibody for each nuclear extract.

IL1beta REGULATES PTGFR mRNA THROUGH NFkappaB 701

was able to delay preterm birth when administered into theamniotic fluid of mice [48], and infusion of sulfasalazine, ananti-inflammatory agent and NFKB inhibitor, decreased uterineelectromyographic activity in pregnant ewes induced to enterpreterm labor with RU486, the progesterone receptor blocker(I.R. Young, personal communication). These studies demon-strate the participation of NFKB in preterm and term labor.

Although NFKB is composed of dimeric complexes formedfrom the REL family of proteins, it is classically detected as theNFKB1/RELA heterodimer [49]. In unstimulated cells, NFKBis sequestered in the cytoplasm through binding to NFKBI(previously known as IjB), which masks the nuclearlocalization signal of the particular NFKB subunit [49].Treatment with cytokines leads to phosphorylation of NFKBI,which causes its ubiquitination and degradation by the 26Sproteasome, thereby unmasking the nuclear translocation signalthat allows nuclear translocation of the NFKB subunit, where itmodulates gene expression through binding to various NFKBresponse elements [49]. Activation of the NFKB pathway iscommonly blocked through the use of either the 26Sproteasome inhibitor MG132, which prevents the degradationof NFKBI, or phenethyl caffeiate (CAPE), which preventsbinding of NFKB subunits to their DNA response elements[38, 50].

PTGFR is one of several genes involved in activating theuterus for the contractions of labor. This has been demonstratedin mouse, rat, and human studies [12–14, 19]. In sheep,although PTGFR expression does not change with the onset oflabor in every study, antagonism of PTGFR expressionprevents RU486-induced preterm labor [16, 51, 52]. Takentogether, these studies suggest that PTGFR plays a prominentrole in the process of labor, and that understanding its

regulation is an important step toward understanding theinitiation of the uterine contractions of labor.

Our results show that IL1B increases the expression of theuterine activating gene PTGFR in an immortalized humanmyometrial cell line, and are consistent with studies in humangranulosa-luteal cells, in which IL1B induces both PTGS2 andPTGFR mRNAs in time- and concentration-dependent man-ners [29]. PTGS2, which is another uterine activation gene, isalso upregulated in human myometrial cells by IL1B, whichsuggests that this cytokine coregulates both prostaglandinsynthesis and contractile prostaglandin responsiveness inuterine myocytes [27, 47]. These findings are consistent withthe notion that proinflammatory cytokines stimulate laborthrough increased expression of uterine activation genes.Although, the increase in PTGFR mRNA reported in thepresent study is modest compared to PTGS2 induction, a ;2-fold increase may still have significant effects in themyometrium. In fact, mouse studies illustrate that a ;2-foldincrease in PTGFR expression in the absence of an increase inPGF2a level is sufficient for labor to occur [14, 19].

Although the reason for the time-dependent increase inPTGFR mRNA levels is unclear, we do know that it is not dueto growth arrest brought on by either serum withdrawal orconfluence. Time course experiments in which control cellswere not serum starved also show an increase in PTGFRmRNA levels at 12 h and 24 h (data not shown). Furthermore,differing cell densities (40–100%) at the start of a 24-htreatment had no effect on either the PTGFR mRNA levels inthe controls or the magnitude of the IL1B-induced increase inPTGFR mRNA levels (data not shown).

We have shown that treatment of ULTR cells with IL1Bresults in nuclear translocation of the RELA subunit of NFKB,as well as increased binding activity of RELA to a consensus

FIG. 4. The effect of the MG132 and CAPEon NFKB activation and PTGFR mRNAlevels in ULTR cells treated with IL1B. A)Western blot of RELA protein levels innuclear extracts of ULTR cells pretreatedwith or without 5 lM or 10 lM MG132 1 hprior to treatment with or without 0.6 nMIL1B for 24 h. B) PTGFR mRNA levels (mean6 SEM) in nuclear extracts of ULTR cellspretreated with or without 0.6 nM (10 lM)MG132 or 17.6 lM (5 lg/ml) CAPE 1 h priorto treatment with or without 0.6 nM (10 ng/ml) IL1B for 12 and 24 h. The results shownare a compilation of all the experiments,where N ¼ 11 for the control and IL1Bgroups control at 12 h, N ¼ 7 for the controland IL1B groups control at 24 h; and N ¼ 3for all the MG132- and CAPE-pretreatedgroups. *, P , 0.05 compared to time-matched controls. C) Western blot of RELAprotein levels in nuclear extracts of ULTRcells pretreated for 1h with or without 17.6lM CAPE and then treated with or without0.6 nM IL1B for 12 h or 24 h. D) EMSA forRELA binding in nuclear extracts of ULTRcells pretreated for 1 h with or without 17.6lM CAPE and then with or without 0.6 nMIL1B for 12 or 24 h.

702 ZARAGOZA ET AL.

NFKB-binding site. This is consistent with previous studies ofimmortalized human myometrial cells, which show completedegradation of NFKBI within 30 min of IL1B treatment,thereby allowing nuclear translocation of the NFKB subunit[27, 47]. The NFKBI levels returned to baseline levels by 60min, which suggests that IL1B-induced NFKB activation isshort-lived. We also observed activation of NFKB after just 30min of treatment with IL1B (data not shown). However, thisactivation lasted for up to 24 h of IL1B treatment, as thenuclear abundance and DNA-binding activity of RELA werestill higher than those of the controls at 24 h of IL1B treatment(Figs. 2 and 3). Inhibition of the NFKB pathway by eitherpreventing RELA nuclear translocation (with MG132) orinhibiting RELA DNA-binding activity (with CAPE) reducedthe PTGFR mRNA levels to well below the control levels. Thissuggests that NFKB is either a part of the basal transcriptionmachinery for PTGFR in ULTR cells or is required for theexpression of basal transcription factors. In addition to theeffect on basal PTGFR mRNA levels, both MG132 and CAPEprevented IL1B induction of PTGFR mRNA. This is consistentwith the hypothesis that the increase in PTGFR mRNA levelsinduced by IL1B requires NFKB; however, it is possible thatthe lack of induction is secondary to the effect on basaltranscription, i.e., the basal machinery might have beeninhibited to an extent that no stimulation could occur.

MG132 affects several pathways in its role as a proteasomeinhibitor. Therefore, it is possible that the effects on IL1Binduction of PTGFR may not be specific to NFKB activation.However, the CAPE experiments strengthen our claim that theIL1B induction of PTGFR requires NFKB. In contrast toMG132, CAPE is a specific inhibitor of NFKB activation anddoes not appear to block nuclear translocation, sincepretreatment of U937 cells with CAPE did not inhibit TNF-dependent phosphorylation and degradation of NFKBIA(previously known as IjBa) [50]. Our experiments indicatethat although CAPE does not block RELA nuclear transloca-tion in myometrial cells, it does inhibit RELA DNA-binding at12 hours of IL1B treatment. Furthermore, the inhibition ofNFKB binding is specific, as the binding of other transcriptionfactors, including activator protein 1, POU2F1 (previouslyknown as OCT-1), and TBP (TFIID), to their respectivebinding sites is not affected by CAPE treatment [50]. OurEMSA experiments also show that after 24 h treatment withIL1B, CAPE no longer had an inhibitory affect on RELADNA-binding, yet the PTGFR mRNA levels were still reducedcompared to untreated control ULTR cells (Fig. 4b and d). Thereason for this result is unclear and it may simply reflect short-lived CAPE activity in culture. Alternatively, inhibition ofNFKB may have affected the expression of certain genes thatin turn regulate PTGFR expression. Despite the recovery ofNFKB binding activity at 24 h of IL1B treatment, the return tocontrol levels of PTGFR mRNA may require more time asother NFKB-regulated genes recover.

Since NFKB is known to influence the transcription of alarge number of genes, the use of inhibitors makes it difficult toassess whether the NFKB-dependant induction of PTGFRmRNA with IL1B is direct or indirect [53]. Our results suggestthat NFKB is required for IL1B induction of PTGFR, but asdiscussed above, it is possible that this induction occursthrough other NFKB-regulated genes rather than through thedirect binding of NFKB to the PTGFR promoter. However, thepresence of NFKB-binding sites in the PTGFR promoter atpositions �3835 and �3017 (Yamamoto et al. unpublishedresults), as well as within intron 1 at positions 744 and 777suggests a direct effect [17].

The choice of NFKB subunit to examine in this study wasan important consideration because it is the specific combina-tion of subunits that determines the specificity of transcrip-tional activation [49]. Chapman et al have identified changes inNFKB subunit composition associated with pregnancy andlabor, with NFKB1 homodimers predominating in non-pregnant myometrium samples and RELA:NFKB1 heterodi-mers predominating in pregnant and laboring samples [54].They also found that all of the subunits were significantlyreduced in laboring myometrial samples, while the spontaneouslaboring levels of RELA, NFKB1, and REL were significantlyhigher in the upper uterine segment compared to the loweruterine segment, which suggests a possible role for thesesubunits in the contraction of the upper uterine segment duringlabor [54]. Furthermore, RELA was the only subunit that wasnot reduced in pregnant myometrium when compared to non-pregnant controls [54]. For these reasons, we believe we arejustified in our choice of analyzing RELA in our experiments.Nevertheless, an examination of other NFKB subunits in IL1B-treated ULTR cells may be informative.

While the PTGFR regulatory mechanism identified here is asignificant result, the in vivo situation is obviously morecomplicated. The EMSAs demonstrate only the in vitro proteinDNA interaction and any selectivity provided when the site iswithin its native promoter is lost [55]. Therefore, anyregulatory effects of chromatin structure are not accountedfor in this study. Indeed, modification of chromatin structurehas been found to play a prominent role in IL1-inducedstimulation of PTGS2 in human myometrial cells [47]. Ourstudy is only a first step towards understanding PTGFRregulation in human uterine cells, and future efforts should bedirected towards specific uterine cells and a consideration ofchromatin structure as well as other agonists.

ACKNOWLEDGMENTS

The authors thank Dr. J.K. McDougall (Fred Hutchison CancerResearch Centre, Seattle, WA) for the kind gift of ULTR cells, and Dr.Xin Fang for expert technical assistance.

REFERENCES

1. Noort WA, van Bulck B, Vereecken A, de Zwart FA, Keirse MJ. Changesin plasma levels of PGF2 alpha and PGI2 metabolites at and after deliveryat term. Prostaglandins 1989; 37:3–12.

2. Takagi M, Fujimoto S, Ohtani M, Miyamoto A, Wijagunawardane MP,Acosta TJ, Miyazawa K, Sato K. Bovine retained placenta: hormonalconcentrations in fetal and maternal placenta. Placenta 2002; 23:429–437.

3. Ulug U, Goldman S, Ben-Shlomo I, Shalev E. Matrix metalloproteinase(MMP)-2 and MMP-9 and their inhibitor, TIMP-1, in human term deciduaand fetal membranes: the effect of prostaglandin F(2alpha) andindomethacin. Mol Hum Reprod 2001; 7:1187–1193.

4. Goldman S, Weiss A, Almalah I, Shalev E. Progesterone receptorexpression in human decidua and fetal membranes before and aftercontractions: possible mechanism for functional progesterone withdrawal.Mol Hum Reprod 2005; 11:269–277.

5. Lindell JO, Kindahl H, Jansson L, Edqvist LE. Postpartum release ofprostaglandin F2alpha and uterine involution in the cow. Theriogenology1982; 17:237–245.

6. Thompson FN, Page RD, Cook CB, Caudle AB. Prostaglandin F2 alphametabolite levels in normal and uterine-infected postpartum cows. Vet ResCommun 1987; 11:503–507.

7. Lindell JO, Kindahl H. Exogenous prostaglandin F2 alpha promotesuterine involution in the cow. Acta Vet Scand 1983; 24:269–274.

8. Senior J, Marshall K, Sangha R, Baxter GS, Clayton JK. In vitrocharacterization of prostanoid EP-receptors in the non-pregnant humanmyometrium. Br J Pharmacol 1991; 102:747–753.

9. Senior J, Sangha R, Baxter GS, Marshall K, Clayton JK. In vitrocharacterization of prostanoid FP-, DP-, IP- and TP-receptors on the non-pregnant human myometrium. Br J Pharmacol 1992; 107:215–221.

10. Senior J, Marshall K, Sangha R, Clayton JK. In vitro characterization of

IL1beta REGULATES PTGFR mRNA THROUGH NFkappaB 703

prostanoid receptors on human myometrium at term pregnancy. Br JPharmacol 1993; 108:501–506.

11. Alfaidy N, Xiong ZG, Myatt L, Lye SJ, MacDonald JF, Challis JR.Prostaglandin F2alpha potentiates cortisol production by stimulating11beta-hydroxysteroid dehydrogenase 1: a novel feedback loop that maycontribute to human labor. J Clin Endocrinol Metab 2001; 86:5585–5592.

12. Brodt-Eppley J, Myatt L. Changes in expression of contractile FP andrelaxatory EP2 receptors in pregnant rat myometrium during late gestation,at labor, and postpartum. Biol Reprod 1998; 59:878–883.

13. Brodt-Eppley J, Myatt L. Prostaglandin receptors in lower segmentmyometrium during gestation and labor. Obstet Gynecol 1999; 93:89–93.

14. Cook JL, Shallow MC, Zaragoza DB, Anderson KI, Olson DM. Mouseplacental prostaglandins are associated with uterine activation and thetiming of birth. Biol Reprod 2003; 68:579–587.

15. Peri KG, Quiniou C, Hou X, Abran D, Varma DR, Lubell WD, ChemtobS. THG113: a novel selective FP antagonist that delays preterm labor.Semin Perinatol 2002; 26:389–397.

16. Hirst JJ, Parkington HC, Young IR, Palliser HK, Peri KG, Olson DM.Delay of preterm birth in sheep by THG113.31, a prostaglandin F2alphareceptor antagonist. Am J Obstet Gynecol 2005; 193:256–266.

17. Zaragoza DB, Wilson R, Eyster K, Olson DM. Cloning and character-ization of the promoter region of the human prostaglandin F2alphareceptor gene. Biochim Biophys Acta 2004; 1676:193–202.

18. Matsumoto T, Sagawa N, Yoshida M, Mori T, Tanaka I, Mukoyama M,Kotani M, Nakao K. The prostaglandin E2 and F2 alpha receptor genes areexpressed in human myometrium and are down-regulated duringpregnancy. Biochem Biophys Res Commun 1997; 238:838–841.

19. Cook JL, Zaragoza DB, Sung DH, Olson DM. Expression of myometrialactivation and stimulation genes in a mouse model of preterm labor:myometrial activation, stimulation, and preterm labor. Endocrinology2000; 141:1718–1728.

20. Makino S, Zaragoza DB, Mitchell BF, Yonemoto H, Olson DM. DecidualActivation: Abundance and localization of prostaglandin F2alpha (FP)mRNA and protein and uterine activation proteins in human decidua atpreterm and term birth. Placenta 2006; (in press).

21. Romero R, Yoon BH, Mazor M, Gomez R, Diamond MP, Kenney JS,Ramirez M, Fidel PL, Sorokin Y, Cotton D, Sehgal P. The diagnostic andprognostic value of amniotic fluid white blood cell count, glucose,interleukin-6, and gram stain in patients with preterm labor and intactmembranes. Am J Obstet Gynecol 1993; 169:805–816.

22. Dudley DJ, Hunter C, Mitchell MD, Varner MW. Clinical value ofamniotic fluid interleukin-6 determinations in the management of pretermlabour. Br J Obstet Gynaecol 1994; 101:592–597.

23. Romero R, Mazor M, Sepulveda W, Avila C, Copeland D, Williams J.Tumor necrosis factor in preterm and term labor. Am J Obstet Gynecol1992; 166:1576–1587.

24. Romero R, Brody DT, Oyarzun E, Mazor M, Wu YK, Hobbins JC, DurumSK. Infection and labor. III. Interleukin-1: a signal for the onset ofparturition. Am J Obstet Gynecol 1989; 160:1117–1123.

25. Romero R, Manogue KR, Mitchell MD, Wu YK, Oyarzun E, Hobbins JC,Cerami A. Infection and labor. IV. Cachectin-tumor necrosis factor in theamniotic fluid of women with intraamniotic infection and preterm labor.Am J Obstet Gynecol 1989; 161:336–341.

26. Opsjln SL, Wathen NC, Tingulstad S, Wiedswang G, Sundan A, WaageA, Austgulen R. Tumor necrosis factor, interleukin-1, and interleukin-6 innormal human pregnancy. Am J Obstet Gynecol 1993; 169:397–404.

27. Belt AR, Baldassare JJ, Molnar M, Romero R, Hertelendy F. The nucleartranscription factor NF-kappaB mediates interleukin-1beta-induced ex-pression of cyclooxygenase-2 in human myometrial cells. Am J ObstetGynecol 1999; 181:359–366.

28. Sooranna SR, Engineer N, Loudon JA, Terzidou V, Bennett PR, JohnsonMR. The mitogen-activated protein kinase dependent expression ofprostaglandin H synthase-2 and interleukin-8 messenger ribonucleic acidby myometrial cells: the differential effect of stretch and interleukin-1beta.J Clin Endocrinol Metab 2005; 90:3517–3527.

29. Narko K, Ritvos O, Ristimaki A. Induction of cyclooxygenase-2 andprostaglandin F2alpha receptor expression by interleukin-1beta in culturedhuman granulosa-luteal cells. Endocrinology 1997; 138:3638–3644.

30. Hertelendy F, Molnar M, Romero R. Interferon gamma antagonizesinterleukin-1beta-induced cyclooxygenase-2 expression and prostaglandinE(2) production in human myometrial cells. J Soc Gynecol Investig 2002;9:215–219.

31. Kniss DA, Rovin B, Fertel RH, Zimmerman PD. Blockade NF-kappaBactivation prohibits TNF-alpha-induced cyclooxygenase-2 gene expres-sion in ED27 trophoblast-like cells. Placenta 2001; 22:80–89.

32. Lappas M, Yee K, Permezel M, Rice GE. Lipopolysaccharide and TNF-

alpha activate the nuclear factor kappa B pathway in the human placentalJEG-3 cells. Placenta 2006; 27:568–575.

33. Lappas M, Permezel M, Georgiou HM, Rice GE. Regulation ofphospholipase isozymes by nuclear factor-kappaB in human gestationaltissues in vitro. J Clin Endocrinol Metab 2004; 89:2365–2372.

34. Allport VC, Pieber D, Slater DM, Newton R, White JO, Bennett PR.Human labour is associated with nuclear factor-kappaB activity whichmediates cyclo-oxygenase-2 expression and is involved with the‘functional progesterone withdrawal’. Mol Hum Reprod 2001; 7:581–586.

35. Perez-Reyes N, Halbert CL, Smith PP, Benditt EP, McDougall JK.Immortalization of primary human smooth muscle cells. Proc Natl AcadSci U S A 1992; 89:1224–1228.

36. Schreiber E, Matthias P, Muller MM, Schaffner W. Rapid detection ofoctamer binding proteins with ‘mini-extracts’, prepared from a smallnumber of cells. Nucleic Acids Res 1989; 17:6419.

37. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001; 29:e45.

38. Andela VB, Rosier RN. The proteosome inhibitor MG132 attenuatesretinoic acid receptor trans-activation and enhances trans-repression ofnuclear factor kappaB. Potential relevance to chemo-preventive interven-tions with retinoids. Mol Cancer 2004; 3:8.

39. Anwer K, Oberti C, Perez GJ, Perez-Reyes N, McDougall JK, Monga M,Sanborn BM, Stefani E, Toro L. Calcium-activated Kþ channels asmodulators of human myometrial contractile activity. Am J Physiol 1993;265:C976–985.

40. Maher E, Bardequez A, Gardner JP, Goldsmith L, Weiss G, Mascarina M,Aviv A. Endothelin- and oxytocin-induced calcium signaling in culturedhuman myometrial cells. J Clin Invest 1991; 87:1251–1258.

41. Schrey MP, Read AM, Steer PJ. Oxytocin and vasopressin stimulateinositol phosphate production in human gestational myometrium anddecidua cells. Biosci Rep 1986; 6:613–619.

42. Tasaka K, Masumoto N, Miyake A, Tanizawa O. Direct measurement ofintracellular free calcium in cultured human puerperal myometrial cellsstimulated by oxytocin: effects of extracellular calcium and calciumchannel blockers. Obstet Gynecol 1991; 77:101–106.

43. Sato TA, Keelan JA, Mitchell MD. Regulation of prostaglandinproduction in an immortalized human myometrial cell line by cytokinesand non-steroidal anti-inflammatory drugs. Eur J Obstet Gynecol ReprodBiol 2002; 100:158–162.

44. Grammatopoulos DK, Hillhouse EW. Basal and interleukin-1beta-stimulated prostaglandin production from cultured human myometrialcells: differential regulation by corticotropin-releasing hormone. J ClinEndocrinol Metab 1999; 84:2204–2211.

45. Pollard JK, Mitchell MD. Intrauterine infection and the effects ofinflammatory mediators on prostaglandin production by myometrial cellsfrom pregnant women. Am J Obstet Gynecol 1996; 174:682–686.

46. Allport VC, Slater DM, Newton R, Bennett PR. NF-kappaB and AP-1 arerequired for cyclo-oxygenase 2 gene expression in amnion epithelial cellline (WISH). Mol Hum Reprod 2000; 6:561–565.

47. Soloff MS, Cook DL Jr., Jeng YJ, Anderson GD. In situ analysis ofinterleukin-1-induced transcription of cox-2 and il-8 in cultured humanmyometrial cells. Endocrinology 2004; 145:1248–1254.

48. Condon JC, Jeyasuria P, Faust JM, Mendelson CR. Surfactant proteinsecreted by the maturing mouse fetal lung acts as a hormone that signals theinitiation of parturition. Proc Natl Acad Sci U S A 2004; 101:4978–4983.

49. Perkins ND. The Rel/NF-kappa B family: friend and foe. Trends BiochemSci 2000; 25:434–440.

50. Natarajan K, Singh S, Burke TR Jr., Grunberger D, Aggarwal BB. Caffeicacid phenethyl ester is a potent and specific inhibitor of activation ofnuclear transcription factor NF-kappa B. Proc Natl Acad Sci U S A 1996;93:9090–9095.

51. Gyomorey S, Lye SJ, Gibb W, Challis JR. Fetal-to-maternal progressionof prostaglandin H(2) synthase-2 expression in ovine intrauterine tissuesduring the course of labor. Biol Reprod 2000; 62:797–805.

52. Palliser HK, Hirst JJ, Ooi GT, Rice GE, Dellios NL, Escalona RM,Parkington HC, Young IR. Prostaglandin e and f receptor expression andmyometrial sensitivity at labor onset in the sheep. Biol Reprod 2005; 72:937–943.

53. Verma IM, Stevenson JK, Schwarz EM, Van Antwerp D, Miyamoto S.Rel/NF-kappa B/I kappa B family: intimate tales of association anddissociation. Genes Dev 1995; 9:2723–2735.

54. Chapman NR, Europe-Finner GN, Robson SC. Expression and deoxy-ribonucleic acid-binding activity of the nuclear factor kappaB family in thehuman myometrium during pregnancy and labor. J Clin Endocrinol Metab2004; 89:5683–5693.

55. Saccani S, Pantano S, Natoli G. Modulation of NF-kappaB activity byexchange of dimers. Mol Cell 2003; 11:1563–1574.

704 ZARAGOZA ET AL.