227.full

Publish in

Documents

123 views

Please download to get full document.

View again

of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
227 Estrogen receptor (ER) expression and function in the pregnant human myometrium: estradiol via ERa activates ERK1/2 signaling in term myometrium Toni Welsh1,2, Matrika Johnson1, Lijuan Yi1, Huiqing Tan1, Roksana Rahman1, Amy Merlino1, Tamas Zakar2,3 and Sam Mesiano1 1 2 3 Department of Reproductive Biology, Case Western Reserve University, 11100 Euclid Avenue, Cleveland, Ohio 44106, USA Mothers and Babies Research Centre, University of Newcastle, Newcastle, New South Wales 2308, Australia
Tags
Transcript
  Estrogen receptor (ER) expression and function in the pregnant humanmyometrium: estradiol via ER a  activates ERK1/2 signaling in termmyometrium Toni Welsh 1,2 , Matrika Johnson 1 , Lijuan Yi 1 , Huiqing Tan 1 , Roksana Rahman 1 , Amy Merlino 1 ,Tamas Zakar 2,3 and  Sam Mesiano 1 1 Department of Reproductive Biology, Case Western Reserve University, 11100 Euclid Avenue, Cleveland, Ohio 44106, USA 2 Mothers and Babies Research Centre, University of Newcastle, Newcastle, New South Wales 2308, Australia 3 Department of Obstetrics and Gynaecology, John Hunter Hospital, Newcastle, New South Wales 2305, Australia(Correspondence should be addressed to S Mesiano; Email: sam.mesiano@case.edu) Abstract Estrogens are thought to promote labor by increasingthe expression of pro-contraction genes in myometrial cells.The specific estrogen receptors ((ERs: ER a  and ER b  (alsoknown as ESR1 and ESR2)) and G protein-coupled receptor 30 (GPR30; also known as G protein-coupled estrogenreceptor 1)) and signaling pathways that mediate these actionsare not clearly understood. In this study, we identified the ERsexpressedinthepregnanthumanmyometriumanddetermineda key extranuclear signaling pathway through which estradiol(E 2 ) modulates expression of the gene encoding the oxytocinreceptor (OXTR), a major pro-contraction protein. Usingquantitative RT-PCR, we found that  ER  a  and  GPR30  mRNAs were expressed in the human pregnant myometriumwhile  ER  b  mRNAwas virtually undetectable. While mRNAencoding ER a  was the predominant ER transcript in thepregnant myometrium, ER a  protein was largely undetectablein myometrial tissue by immunoblotting. Pharmacologicalinhibition of 26S proteasome activity increased ER a  proteinabundance to detectable levels in term myometrial explants,however, indicating rapid turnover of ER a  protein byproteasomal processing in the pregnant myometrium. E 2 stimulated rapid extranuclear signaling in myometrial explants,as evidenced by increased extracellularly regulated kinase(ERK1/2) phosphorylation within 10 min. This effect wasinhibited by pre-treatment with an ER antagonist, ICI182 780, indicating the involvement of ER a . Inhibition of ERK signaling abrogated the ability of E 2  to stimulate  OXTR  gene expression in myometrial explants. We conclude thatestrogenic actions in the human myometrium duringpregnancy, including the stimulation of contraction-associatedgene expression, can be mediated by extranuclear signalingthrough ER a  via activation of the ERK/mitogen-activatedprotein kinase pathway.  Journal of Endocrinology   (2012)  212,  227–238 Introduction Estrogens (mainlyestradiol (E 2 )) promote labor by stimulatingbiochemical and physical changes in myometrial cells thataugment uterine contractility and excitability (for review, seePepe & Albrecht (1995)). Studies on various species haveshown that E 2  increases the expression of genes in myometrialcells encoding gap junction proteins that form low-resistanceconnections between myometrial cells to promote synchro-nized contractions (Lye  et al  . 1993, Petrocelli & Lye 1993, Kilarski  et al  . 1996, 2000), receptors for uterotonic hormones such as oxytocin and prostaglandin F 2 a  (PGF 2 a ; Pinto  et al  .1966, 1967, Nissenson  et al  . 1978), and enzymes such asPG-endoperoxide synthase-2 (PTGS2), the main induciblerate-limiting enzyme for PG synthesis (Mesiano  et al  . 2002).Although it is generally accepted that these actions aremediated by estrogen receptors (ERs), the specific ERs andassociated signaling pathways are not well characterized in thepregnant human myometrium. Human ERs exist in threeprincipal forms: two classical nuclear ER subtypes, ER a  (alsocalled ESR1) and ER b (ESR2; Warner   et al  . 1999, Dechering et al  . 2000), and a seven-transmembrane G protein-coupledreceptor known as GPR30 (alternatively known as G protein-coupled estrogen receptor 1 (GPER); Revankar   et al  . 2005,Thomas  et al  . 2005). The purpose of this study was todetermine whether these ERs are expressed in the pregnanthuman myometrium and to investigate the signaling pathwaysthrough which they mediate estrogenic actions.In most species, the pro-labor influence of estrogens isachieved by a marked increase in circulating estrogenconcentrations prior to the onset of labor. In humanpregnancy, however, the myometrium is exposed to highlevels of estrogens in the form of E 2 , estrone, and estriol for most of pregnancy and systemic estrogen levels do not change 227  Journal of Endocrinology   (2012)  212,  227–238 DOI: 10.1530/JOE-11-03580022–0795/12/0212–227  q 2012 Society for Endocrinology  Printed in Great Britain   Online version via http://www.endocrinology-journals.org  significantly prior to labor onset (De Hertogh  et al  . 1975,Smith  et al  . 2009). It appears, therefore, that the humanmyometrium is refractory to the pro-contractile actions of estrogens for most of pregnancy and that the parturitionprocess involves increased myometrial responsiveness toestrogens. We have previously found that myometrial  ER  a mRNA levels rise at parturition and that the levels arestrongly correlated with the abundance of mRNAs encoding PTGS2  and the oxytocin receptor ( OXTR  ; Mesiano  et al  .2002). However, attempts to measure ER a  protein inpregnant human myometrium by immunohistochemistry(Geimonen  et al  . 1998, Winkler   et al  . 2002, Laudanski  et al  .2004) or immunoblotting (Geimonen  et al  . 1998) have beenlargely unsuccessful, and it is well documented that ER a expression is substantially reduced during pregnancycompared with the non-pregnant state (Geimonen  et al  .1998, Benassayag  et al  . 1999). Wu  et al  . (2000) reported thatER b but not ER a was detectable by immunoblotting in termhuman myometrium and suggested that ER b  is the principalmediator of estrogenic actions in the pregnant human uterus.However, we found that  ER  b  mRNA was barely detectableby the very sensitive real-time RT-PCR technique in termmyometrium (Mesiano  et al  . 2002). Thus, there is a lack of consensus regarding the expression of ER a  and ER b  in thepregnant human myometrium and their role in mediatingestrogen actions, which is further complicated by thepossibility that GPR30 is also involved in estrogen signalingin myometrial cells (Maiti  et al  . 2011).Estrogens typically affect target cell function via thegenomic mode of steroid hormone action, whereby ER a and ER b  function as ligand-activated transcription factorsto modulate the expression of specific genes. However, manypro-contraction genes that are regulated by E 2 , including OXTR   and connexin-43 ( GJA1 ), lack complete canonicalestrogen response elements (EREs) in their promoter regions(Geimonen  et al  . 1996, Gimpl & Fahrenholz 2001). This suggests that E 2  modulates the expression of some genes viaalternative mechanisms. One possibility is that E 2  regulatesthe expression of ERE-containing genes via the classicalgenomic mode of action and that the products of those genesthen regulate the expression of secondaryor tertiary gene sets.An alternative possibility is that estrogens activate extra-nuclear/non-genomic pathways that lead to downstreamchanges in gene expression (reviewed in Vasudevan & Pfaff (2008), Fox  et al  . (2009), and Prossnitz & Maggiolini (2009)). Unlike the genomic mode of action, the extranuclear modeaffects cell function directly by modulating cytoplasmicsignaling cascades that impact on the activity of multiplecellular processes and transcription factors. Extranuclear pathways activated by E 2  include the extracellularly regulatedkinase (ERK)/mitogen-activated protein kinase (MAPK) andphosphatidylinositol 3-kinase/protein kinase B (PKB/AKT)pathways, as well as Ca 2 C influx and G protein signaling(Vasudevan & Pfaff 2008, Fox  et al  . 2009, Prossnitz &Maggiolini 2009). ER a , ER b , and GPR30 have each beenfound to induce extranuclear signaling, although theinvolvement of the different receptors appears to be tissue-and cell-type specific (Vasudevan & Pfaff 2008, Fox  et al  .2009, Prossnitz & Maggiolini 2009). Rapid activation of the ERK/MAPK signaling pathway by E 2  has been extensivelystudied in breast cancer cells where this cascade is thought tomediate the proliferative response to estrogens (Migliaccio  et al  . 1996, Razandi  et al  . 2003, Song  et al  . 2002, 2004). Upon ligand binding, ER a  and ER b  can interact with numerousscaffolding proteins and signaling molecules in the cytoplasm,including SRC (Migliaccio  et al  . 1996, Barletta  et al  . 2004),MNAR (Barletta  et al  . 2004), SHC (Song  et al  . 2002, 2004), and G proteins (Kumar   et al  . 2007), and in concert with theseproteins stimulate downstream activation of ERK1/2 (alsoknown as p44/42 MAPK). GPR30 activates rapid non-genomic signaling via its coupled G proteins, which activateSRC and the downstream ERK pathway, often by way of extracellular release of heparin-bound epidermal growthfactor (HB-EGF) and transactivation of the EGF receptor (reviewed in Maggiolini & Picard (2010)). Madak-Erdogan et al  . (2008) have found that the extranuclear actions of E 2  mediated by ER a  controlled w 25% of all E 2 -responsivegenes in MCF7 breast cancer cells, highlighting theimportance of extranuclear ER a  signaling in mediatingligand-activated gene expression.Extranuclear signaling by estrogens is largely unexploredin pregnant human myometrium. In pregnant rats, theabundance of activated ERK in myometrial cells increaseswith gestation (Oldenhof   et al  . 2002, Serrano-Sanchez  et al  .2008) and labor onset (Li  et al  . 2003). In animals induced todeliver preterm (PT) by administration of the progesteronereceptor antagonist RU486, concurrent inhibition of ERKsignaling delayed preterm labor and decreased myometrialcontractility (Li  et al  . 2004). These data emphasize thepotential importance of this pathway in the regulation of myometrial smooth muscle contractility. In this study, wecharacterized the ERs expressed in the pregnant humanmyometrium and determined whether they activate thecytoplasmic ERK signaling cascade in response to E 2  inexplant cultures of term human myometrium. Materials and Methods Myometrial tissue collection Lower uterine segment myometrium was obtained followingcollection of written informed consent from women under-going cesarean section at MacDonald Women’s Hospital,University Hospitals Case Medical Center, Cleveland, OH(IRB approval #11-04-06), and MetroHealth MedicalCenter, Cleveland, OH (IRB approval #05-00287). A biopsyof myometrium ( w 0 . 5 cm 3 ) was excised from the upper margin of the lower segment incision after delivery of theplacenta. Tissue was collected from women delivering at PTand not in labor (NIL) (PT-NIL;  n Z 6; range 27w0d–34w2d), at PTand in labor (IL) (PT-IL;  n Z 6; range 24w1d–36 T WELSH  and others  .  Estrogen signaling in the pregnant uterus  228  Journal of Endocrinology   (2012)  212,  227–238 www.endocrinology-journals.org  w6d), at term and NIL (T-NIL;  n Z 19; range 37w3d–41w6d), and at term and IL (T-IL;  n Z 11; range 38w1d–40w4d). Cesarean sections were performed in non-laboringwomen for complications such as pre-eclampsia, breechpresentation, or previous cesarean section, while indicationsfor cesarean section during labor included fetal distress,maternal complications, or failure to progress. Samples frompregnancies with induced labor or clinical signs of chor-ioamnionitis were excluded. Tissues were rinsed in ice-coldPBS and portions of each myometrial specimen wereimmediately snap-frozen in liquid nitrogen or placed infixative (10% neutral-buffered formalin). Tissue samples usedfor explant studies were transported to the laboratory in PBSon ice for further processing. Non-pregnant myometriumwas collected from women undergoing surgery for benign gynecological conditions at MacDonald Women’sHospital (IRB approval #06-07-29) and snap-frozen inliquid nitrogen. Myometrial explant culture  Myometrial tissue samples collected from women at term andNIL were dissected into explants of 2–3 mm 3 and placedonto 0 . 2  m m polycarbonate filters (GE Water & ProcessTechnologies, Trevose, PA, USA) floating on 2 ml of phenolred-free, serum-free DMEM (Sigma) containing 2 mmGlutaMAX (Invitrogen), 100 U/ml penicillin, and100  m g/ml streptomycin in 6-well plates and cultured over-night (16–18 h) at 37  8 C in 5% CO 2 , 95% air (six to sevenexplants per well). Media were refreshed and the explantswere incubated with media alone or U0126 (10  m M) (CellSignaling,Beverly,MA,USA),ICI182 780(1  m M;Sigma),or vehicle for 30 min. E 2 , 1 or 10 nm (Sigma), phorbol myristateacetate (PMA; 100 nm), or vehicle was then added for varioustimes, and the explants were snap-frozen in liquid nitrogen.Theviabilityof myometrial explants in this culturesystem wasconfirmed by histological analysis showing few necrotic cellsand maintenance of smooth muscle-specific marker expression and RNA integrity (data not shown). For 26Sproteasome inhibition experiments, explants were placed intoculture and stimulated without delay with vehicle or MG132(10  m M; Sigma) for 24 h. Nuclear and cytoplasmic proteinextracts were then immediately prepared using the Nuclear Extract Kit (Active Motif, Carlsbad, CA, USA) according tothe manufacturer’s instructions. Cell culture and transfection The hTERT-HM cell line, a telomerase-immortalized myo-metrial smooth muscle cell line derived from the fundus of anon-pregnant, pre-menopausal woman (Condon  et al  . 2002),was provided by Prof. William Rainey, Medical College of Georgia. Cells were cultured in complete media: phenol red-free DMEM containing 10% charcoal-stripped FCS, 2 mmGlutaMAX, 100 U/ml penicillin,and 100  m g/ml streptomycinat 37  8 C in5% CO 2 , 95% air. Cells at 90–95% confluency werecollected by trypsinization, centrifuged, and resuspended inAmaxa Nucleofector Mammalian Smooth Muscle CellSolution (Lonza, Walkersville, MD, USA) at a concentrationof 2 ! 10 6 cells/100 m l solution. siRNA targeted against exon 8of   ER  a  (Silencer Select Pre-Designed siRNA ID s4824;Ambion) or Silencer Select Negative Control #1 siRNA(Ambion) was added to the transfection solution at 375 or 750 nmeach.Thetransfectionsolution(100  m l)was transferredto an electroporation cuvette and nucleofection was achievedusing the Amaxa Nucleofector Device with program A33. Thetransfection solution was then diluted in 8 ml complete mediaand plated at 0 . 5 ! 10 6 cells/35 mm dish. A subset of cells wereco-transfected with an expression plasmid encoding full-lengthER a (1 m g plasmidDNA/0 . 5 ! 10 6 cells) provided by DrPeter Kushner, University of California at San Francisco. The cellswere cultured for 48 h before lysis at 4  8 C using CelLytic MTLysis Buffer (Sigma) supplemented with protease and phospha-tase inhibitors (Sigma). T47D and MCF7 human breastcarcinomacellswereobtainedfromtheAmericanTypeCultureCollection (Manassas, VA, USA) and were used as positivecontrols for ER expression. Immunoblotting  Whole tissue protein extracts were prepared from myo-metrium by homogenization in CelLytic MT Lysis Buffer supplemented with protease and phosphatase inhibitors onice. Tissue homogenates were centrifuged at 10 000  g  at 4  8 Cfor 10 min and the supernatants werecollected and assayed for total protein content using the Quick Start Bradford ProteinAssay (Bio-Rad). Cell lysates were incubated with reducingsample buffer and subjected to SDS–PAGE (20–200  m gprotein/lane) using Pierce 10% Precise Protein Gels (ThermoFisher Scientific, Rockford, IL, USA). Proteins weretransferred to polyvinyl difluoride (PVDF) membranes andnon-specific binding was blocked using either 5% BSA (for phospho-antibodies) or 5% skimmed milk (for non-phospho-antibodies) in 20 mm Tris, 150 mm NaCl, pH 7 . 5,containing 0 . 1% Tween-20 (TTBS) for 1 h at roomtemperature. Membranes were incubated with primaryantibodies (Table 1) diluted in blocking buffer overnight at4  8 C, then washed, and incubated with secondary antibody(anti-mouse or anti-rabbit IgG–HRP; 1:3000; Cell Signaling)at room temperature for 1 h. The membranes were washed,incubated with Amersham ECL chemiluminescent reagent(GE Healthcare, Piscataway, NJ, USA), and exposed toautoradiography film (Denville Scientific, South Plainfield,NJ, USA). The resultant bands were quantified using digitaldensitometry (Fujifilm MultiGauge v 3.0 Software; FujiPhoto Film Co., Tokyo, Japan). Where required, the PVDFmembranes were dried, re-wet in methanol, and re-probedwith antibodies directed against loading control proteins or nuclear markers. The effective stripping of each primaryantibody between blots was confirmed by a lack of signalupon incubation with secondary antibody alone withchemiluminescent detection. Estrogen signaling in the pregnant uterus   .  T WELSH  and others 229 www.endocrinology-journals.org  Journal of Endocrinology   (2012)  212,  227–238  RNA extraction and quantitative RT-PCR  RNA was extracted from myometrial tissues and T47D cellsusing Trizol reagent (Invitrogen) according to the manufac-turer’s instructions. Total RNA extracts were treated withDNase (Ambion) to remove residual DNA and quantified byu.v. spectrometry. RNA integrity was determined by agarosegel electrophoresis. Total RNA (1  m g) was reverse transcribedto cDNA using the Superscript II or III First-Strand SynthesisSystems for RT-PCR (Invitrogen) with random hexamer primers. Real-time PCR was performed using an ABIPRISM 7500 Sequence Detector with SYBR Greenfluorescence detection (Applied Biosystems, Carlsbad, CA,USA). PCR primers (Table 2) were designed using Primer Express Software (Applied Biosystems). The abundance of specific mRNAs in myometrial tissues and explants weredetermined relative to the abundance of h-caldesmon ( hCaD  )mRNA using the  D C  t  calculation (Livak & Schmittgen2001). We had determined in previous experiments that thelevels of   hCaD   mRNA (which is expressed specifically insmooth muscle) do not change significantly in the myome-trium with advancing gestational age (GA) or labor onset(linear regression performed on logarithmically transformeddata showed no relationship between  hCaD   mRNA and GA( r  2 Z 0 . 0286,  P  Z 0 . 464;  n Z 21) or labor status ( r  2 Z 0 . 0009, P  Z 0 . 897;  n Z 21)). PCR efficiency for each amplicon wascalculated using LinRegPCR version 11.1 Software (Ruijter  et al  . 2009). T47D cell line cDNA served as a positive controlfor   ER  a , ER  b , and  GPR30   expression. Data analysis The distribution of data was assessed using the Skewness andKurtosis test for normality. Comparisons between two groupsof normally distributed data were performed by Student’s  t  -test (paired or unpaired as appropriate). Phospho-ERKexpression in myometrial explants was assessed by repeatedmeasures ANOVA and differences between treatments wereanalyzed by paired  t  -test with Bonferroni’s correction for multiple comparisons. Linear regression was performed toexamine the relationships between a number of variables( ER  a  mRNA,  ER  b  mRNA (logarithmically transformeddata),  GPR30   mRNA (logarithmically transformed data),GPR30 protein, pERK1 expression (square-root transformeddata), pERK2 or total ERK expression, or the ratio of pERK1 and pERK2 to total ERK (square-root transformeddata)) and GA or labor. Spearman’s rank correlation was usedto test the relationship between  GPR30   mRNA and proteindata. Explant  OXTR   mRNA data were examined usingrepeated measures ANOVA with Dunnett’s  post-hoc   test. Thedata were analyzed using GraphPad Prism version 5.03(GraphPad Software, Inc., San Diego, CA, USA) or Stataversion 11 (StataCorp LP, College Station, TX, USA) and P  ! 0 . 05 was considered significant for all analyses. Results ER expression in the pregnant human myometrium ER  a  and  GPR30   mRNAs were readily detectable inpregnant human myometrium (Fig. 1). However,  ER  b mRNA was detected at extremely low levels by quantitativeRT-PCR (qRT-PCR). T47D cell cDNA served as a positivecontrol for   ER  b  expression and confirmed the functionalityof our PCR primers. Therewas no relationship between  ER  b mRNA abundance and GA or labor status (linear regressionperformed on log-transformed data;  ER  b  vs GA,  r  2 Z 0 . 018, Table 1  Primary antibodies used for western blotting Antibody target Clone/designation Species Supplier Dilution ER a  F10 Mouse Santa Cruz (Santa Cruz, CA, USA) 1:1000ER a  HC20 Rabbit Santa Cruz 1:1000ER a  C-542 Mouse Calbiochem (Rockland, MA, USA) 1:1000GPR30 LS-A4272 Rabbit LifespanBiosciences(Seattle,WA,USA) 1:500GAPDH 6C5 Mouse Santa Cruz 1:50000LSD1 C69G12 Rabbit Cell Signaling 1:1000Phospho-ERK1/2 9101 Rabbit Cell Signaling 1:1000Total ERK1/2 9102 Rabbit Cell Signaling 1:2000 Table 2  Real-time PCR primer sequences TargetmRNAAmpliconsize  (bp) GenBankaccession no.Forwardprimer  (5 0 –3 0 ) Reverseprimer  (5 0 –3 0 ) ER  a  66 NM_000125.3 TGAAAGGTGGGATACGAAAAGAC CATCTCTCTGGCGCTTGTGTT ER  b  101 NM_001437.2 CACGTCAGGCATGCGAGTAA GCATTCAGCATCTCCAGCAG GPR30   78 NM_001505.2 TGCACCTTCATGTCGCTCTT GCGGTCGAAGCTCATCCA OXTR   73 NM_000916.3 CTGGACGCCTTTCTTCTTCGT GAAGGCCGAGGCTTCCTT hCaD   79 NM_033138.2 CAGAGAGGGCAAGGTTGGAA GCTCGTTCATCTGCTATCTTTTTG T WELSH  and others  .  Estrogen signaling in the pregnant uterus  230  Journal of Endocrinology   (2012)  212,  227–238 www.endocrinology-journals.org
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks