s t e r o i d s 7 1 ( 2 0 0 6 ) 256–265

avai lab le at www.sc iencedi rec t .com

journa l homepage: www.e lsev ier .com/ locate /s tero ids

Estrogen induces phospholipase A2 activation throughERK1/2 to mobilize intracellular calcium in MCF-7 cells

Warren Thomas ∗, Natasha Coen, Sheila Faherty, Cathal O Flatharta, Brian J. HarveyCharitable Infirmary Trust Molecular Medicine Laboratories, Royal College of Surgeons in Ireland,Beaumont Hospital, P.O. Box 9063, Dublin 9, Ireland

a r t i c l e i n f o

Article history:

Received 17 June 2005

Received in revised form 5 October

2005

a b s t r a c t

The principal secreted estrogen, 17�-estradiol rapidly activates signaling cascades that reg-

ulate important physiological processes including ion transport across membranes, cytoso-

lic pH and cell proliferation. These effects have been extensively studied in the MCF-7

estrogen-responsive human breast carcinoma cell line. Here, we demonstrate that a physi-

Accepted 31 October 2005

Available online 22 December 2005

Keywords:

Estrogen

Phospholipase A

Calcium

MAP kinase

ological concentration of 17�-estradiol caused a rapid, synchronous and transient increase

in intracellular calcium concentration in a confluent monolayer of MCF-7 cells 2–3 min after

treatment. This response was abolished when cells were pre-incubated with the phospho-

lipase A2 (PLA2) inhibitor quinacrine or with the cyclooxygenase inhibitor indomethacin.

The translocation of GFP-cPLA2� to perinuclear membranes occurred 1–2 min after 17�-

estradiol treatment; this translocation was concurrent with the transient phosphorylation

of cPLA2� at serine residue 505. The phosphorylation and translocation of cPLA2 were sen-

sitive to inhibition of the extracellular signal regulated kinase (ERK) signaling cascade and

occurred simultaneously with a transient activation of ERK. The phosphorylation of cPLA2

could be stimulated by membrane impermeable 17�-estradiol conjugated to bovine serum

albumen and was blocked by an antagonist of the classical estrogen receptor. Here we show,

for the first time, that PLA2 and the eicosanoid biosynthetic pathway are involved in the

17�-estradiol induced rapid calcium responses of breast cancer cells.

© 2005 Elsevier Inc. All rights reserved.

1. Introduction

The most active, secreted form of estrogen, 17�-estradiol(E2) modulates whole body physiology through endocrine,autocrine and paracrine mechanisms to elicit diverseresponses in different target tissues [1,2]. The physiologicalimportance of E2 is emphasized when normal circulatory E2

levels are perturbed, contributing to the development and pro-gression of many diseases. The role of E2 in normal bone recy-cling and the consequences of its withdrawal at menopauseleading to osteoporosis are established [3]. E2 also has pro-liferative effects on breast and primary reproductive tissues

∗ Corresponding author. Tel.: +353 1809 3825; fax: +353 1809 3778.E-mail address: wthomas@rcsi.ie (W. Thomas).

[4]. Recent epidemiological evidence links the reinstatementof E2 through hormone replacement therapy to an increasedrisk of breast cancer initiation and progression [5]. The con-tribution of E2 to breast cancer development has been thesubject of intense investigation and much information aboutthe E2 linked cell signaling responses have been gained bystudying its effects on model breast cancer cell lines. At thecellular level, E2 acts through two interlinked mechanisms:The genomic, transcriptional response mediated by classi-cal E2 receptors [6,7] and also diverse rapid, cell signalingprocesses whose receptor linkage is not yet fully elucidated[7–9].

0039-128X/$ – see front matter © 2005 Elsevier Inc. All rights reserved.doi:10.1016/j.steroids.2005.10.010

s t e r o i d s 7 1 ( 2 0 0 6 ) 256–265 257

The E2 stimulated changes in ion transport across cellmembranes that culminate in cytoplasmic alkalinization[10,11] and an increase in intracellular Ca2+ levels [12–14] makea major contribution to overall cellular physiology includ-ing proliferation. The proliferative effects of E2 are regulatedthrough multiple pathways; however, the sustained activationof the extracellular signal regulated kinase (ERK1/2) mitogenactivated protein kinase (MAPK) signaling pathway is the beststudied and has been implicated in the development of E2

deprivation resistant breast cancer [15]. At least in MCF-7 cellsthe initiation of ERK1/2 activation 3–5 min after E2 applica-tion appears to be calcium dependent [16–18]. The signalingpathways by which E2 raises intracellular calcium levels arenot fully elucidated; however, we have previously demon-strated a PKA and PKC� dependency in the rapid calciuminflux stimulated by E2 in the colon [19–21]. Significantly, PKC�

is involved in regulating the ERK1/2-dependent growth pro-motion of MCF-7 cells following E2 treatment [16]. E2 alsorapidly increases adenylate cyclase activity in breast can-cer cells, isolated uterine cells and also in the intact uteruswith a concurrent stimulation of CRE-dependent transcription[22].

The arachidonic acid (AA) based eicosanoid signaling path-way is also involved in the rapid E2 responses of the colon [23],embryonic membranes [24] and also in rapid effects of 1�,25-(OH)2D3 but not E2 on chondrocytes [25]. Both the inhibition ofphospholipase A2 (PLA2) with quinacrine and the inhibition oftb[psfPtrv

rpostacalsii[cpti[

[cw

2. Experimental

2.1. Cell culture

The MCF-7 human breast carcinoma cell line (ATTC, Ted-dington, United Kingdom) was maintained in DMEM:Ham’sF12 medium without phenol red (Sigma–Aldrich, Dublin, Ire-land) supplemented with non-essential amino acids, sodiumpyruvate, l-glutamine and 10% fetal bovine serum (InvitrogenLife Technologies, Paisely, United Kingdom) and propagated at37 ◦C in a humid atmosphere of 5% CO2. Prior to steroid treat-ment the cells were subjected to progressive serum depletionover 4 days (24 h at 5% serum, 24 h at 2% serum and 48 h serumfree) to achieve growth arrest.

2.2. Reagents and antibodies

E2 (Sigma–Aldrich) was initially dissolved in methanol at aconcentration of 10 mM then diluted to a final concentrationof 10 nM in serum free tissue culture medium. E2 conjugated tobovine serum albumen (E2-BSA) was obtained from Steraloids(Newport, RI) and diluted to a final concentration of 10 nMin serum free tissue culture medium immediately beforeeach experiment. The phospholipase A inhibitor quinacrinewas from Sigma–Aldrich as was the cyclooxygenase inhibitorindomethacin. The ER antagonist ICI 182,780 and the MAPK

he cyclooxygenase (COX) family enzymes with indomethacinlock the calcium response to E2 in the isolated colonic crypts10]. The different isoforms of PLA2 release AA from membranehospholipids while the COX enzymes catalyze the conver-ion of AA to prostaglandin H2 (PGH2) [26]. PGH2 is a substrateor prostaglandin E2 synthase (PGES) [27,28] that producesGE2 and other biologically active prostaglandins through fur-her metabolism. Prostaglandins regulate the inflammatoryesponse and other physiological processes through the acti-ation of G-protein coupled receptors [29].

PLA2 acts on the sn-2 bond of membrane phospholipids toelease AA and a lysophospholipid. Both of these products areotent secondary messengers [26,30]. The different isoformsf PLA2 are divided into three groups. The sPLA2 enzymes areecreted from the cell and require extracellular calcium forheir activity. The intracellular iPLA2 enzymes are membranessociated or cytosolic and their activation is independent ofalcium availability. The calcium dependent cPLA2 enzymesre cytosolic but translocate to intracellular membranes fol-owing activation. The cPLA2� isoform is ubiquitously and con-titutively expressed in most cells. Different stimuli that raisentracellular calcium levels promote the activation of cPLA2,ncluding ATP depletion [31] or treatment with noradrenalin32]. Calcium binding to the C2 domain promotes the translo-ation of cPLA2 and its association with vimentin, rather thanromoting activation [33]. In some experimental systems theranslocation of cPLA2 can also occur at resting [Ca2+]i, such asn response to okadaic acid or phorbolmyristylacetate (PMA)34,35].

Here, we investigated the regulation of the rapid rise inCa2+]i following E2 treatment in the MCF-7 cell line throughPLA2 activation and AA metabolism based on our previousork in the isolated colonic crypts.

kinase (MEK) inhibitor PD98059 were from Tocris (Avonmouth,United Kingdom). All other chemical reagents used in thisinvestigation were obtained from Sigma–Aldrich, unless oth-erwise specified.

The primary polyclonal antibodies used in thisinvestigation, anti-phosphoSer505 cPLA2, total cPLA2, anti-phosphoThr202/phosphoTyr204 ERK1/2 and total ERK1/2were from Cell Signalling Technology (Hitchin, United King-dom). The anti-rabbit IgG horseradish peroxidase conjugateused in Western detection was from Sigma–Aldrich. Theenhanced chemiluminescene (ECL) detection reagents werefrom Amersham (Little Chalfont, United Kingdom).

2.3. Analysis of cPLA2 phosphorylation

Subconfluent MCF-7 cells were propagated on 6 cm tissue cul-ture dishes and subjected to progressive serum depletion over96 h prior to treatment with 10 nM E2 or methanol vehiclecontrol over a period of 5 or 10 min. At the indicated timepoints, the plates were placed on ice and the cells washedwith ice-cold phosphate buffered saline (PBS). The cells werelysed in Laemmli sample buffer and the proteins from thelysates were separated on 8% SDS-PAGE gels then transferredonto PVDF membrane (Amersham). The membranes wereprobed with anti-phosphoSer505 cPLA2 or anti-cPLA2 poly-clonal rabbit antibody. Binding of the primary antibody wasdetected using an anti-rabbit horseradish peroxidase conju-gate and visualized using ECL. The increase in phosphory-lation was quantified by densitometry using Genesnap soft-ware (Synoptics Ltd., Cambridge, United Kingdom). The meandensitometry data from at least three independent experi-ments are presented as fold increases relative to the valuesof untreated controls. The analysis of variance (ANOVA) of

258 s t e r o i d s 7 1 ( 2 0 0 6 ) 256–265

densitometry values relative to untreated controls was per-formed using Dunnett’s modified t-test on one-way ANOVAdata, p < 0.05 was treated as significant (*p < 0.05, **p < 0.01,***p < 0.001).

2.4. Estrogen receptor involvement

The effect of inhibiting the classical ER was investigated byincubating the cells for 1 h with an ER antagonist, the anti-estrogen ICI 182,780 (Tocris) at 10 �M final concentration priorto the addition of E2. The effect of ICI 182,780 on the E2

induced phosphorylation of cPLA2 was determined as above.The role of membrane associated ER in initiating the path-way leading to cPLA2 activation was investigated by examiningthe effect of cell impermeable E2-BSA on cPLA2 phosphoryla-tion. E2-BSA does not cross the cell membrane and pathwaysstimulated by this agonist require a membrane receptor andtransmembrane signal transduction for initiation. E2-BSA fil-tered over a 5-kDa cut off Ultracel PL5 filter (Millipore, Wat-ford, United Kingdom) was also used to demonstrate thatcPLA2 activation was not attributable to E2 shed from the BSAcomplex.

2.5. ERK1/2 dependency of cPLA2 phosphorylation

The activation of ERK1/2 MAP kinase over the timescale ofthis experiment was investigated through Western blotting

2.7. Spectrofluorometric measurement of intracellularcalcium levels

The MCF-7 cells were grown on glass bottom dishes (Intra-cel, Royston, United Kingdom) for 24 h prior to spec-trofluorescence experiments and loaded with the calciumsensitive fluorescent dye FURA-2/AM (5 �M) for 30 min at22 ◦C in serum-free DMEM:F12. The cells were washedonce in Kreb’s solution (NaCl 140 mM, KCl 5 mM, MgCl21 mM, CaCl2 2 mM, HEPES 10 mM, glucose 10 mM, Tris–HCl10 mM, pH 7.4) and maintained in Kreb’s solution (200 �l)at room temperature (20–22 ◦C) for the duration of theexperiment. E2 and other agonists were diluted in Kreb’ssolution then added (800 �l) directly to cells by pipetteat the edge of the plate. The intracellular Fura-2 fluo-rescence was analyzed using a Diaphot 200 inverted epi-fluorescence microscope (Nikon, Welwyn Garden City, UnitedKingdom). The light from a Xenon lamp (Nikon) was fil-tered through alternating 340 and 380 nm interference fil-ters of 10 nm bandwidth (Nikon). The resulting fluorescencewas passed through a 400 nm dichroic mirror, filtered at510 nm and then collected using an intensified CCD cam-era system (Hammatsu, Sunayama, Japan). Images weredigitized and analyzed using Openlab2 software (Impro-vision Ltd., Coventry, United Kingdom). Data for 10 cellswithin the same field of view was simultaneously col-lected in each experiment. Measurements of [Ca2+]i are pre-

analysis of lysates using an activated ERK1/2-specific anti-body (anti-phosphoThr202/phosphoTyr204) followed by ECLdetection. The effect of blocking the Raf-MEK-ERK1/2 sig-nal transduction pathway on cPLA2 activation was inves-tigated by pre-incubating the cells with the specific MEKinhibitor PD98059 (20 �M) for 30 min prior to E2 treat-ment. The phosphorylation of cPLA2 and ERK1/2 follow-ing MEK inhibition were then investigated by Westernblotting.

2.6. cPLA2 transfection and confocal microscopy

The EGFP-cPLA2� expression vector used in this investiga-tion was as previously described [31]. Briefly, human cPLA2�

cDNA was ligated into the PstI site of pEGFP-C1 (Clontech,Mountain View, CA) and the cPLA2�-EGFP fusion protein wasexpressed under the control of the constitutive CMV IE pro-moter. Cells were seeded at 50% confluency in 8-well cham-ber slides and transfected in serum free medium with plas-mid DNA using Lipofectamine reagent (Invitrogen Life Tech-nologies) according to the manufacturer’s instructions. At12 h post transfection, when the cells were still subconflu-ent, they were treated with E2 or vehicle for 1.5 min. Othercells were also pretreated with PD98059 (20 �M) for 30 min orICI 182,780 (10 �M) for 1 h prior to E2 treatment. After treat-ment the cells were fixed for 30 min in 4% paraformaldehydein phosphate buffered saline then mounted in Vectashieldmountant (Vector Laboratories, Burlingame, CA). Images werecollected using a LSM510 confocal microscope (Zeiss, Wel-wyn Garden City, United Kingdom). Images from all treat-ment groups were captured at the same time using identi-cal image capture parameters at ×40 magnification under oilimmersion.

sented as mean values ± S.E.M. of at least five independentexperiments.

3. Results

3.1. Estrogen promotes the phosphorylation of cPLA2

We first determined whether E2 could stimulate the acti-vation of cPLA2 in the MCF-7 human breast cancer cellline on the time scale of the previously described rapidresponses of these cells. The regulation of cPLA2 activityis principally through phosphorylation dependent transloca-tion to substrate rich sites. The phosphorylation of cPLA2

augments its calcium dependent translocation and asso-ciation with the membranes of cellular organelles. Theenzymes responsible for AA metabolism are localized inthe perinuclear membranes and endoplasmic reticulum. Fol-lowing activation cPLA2 translocates to these intracellularmembranes where it releases AA for eicosanoid produc-tion and lysophospholipids. The phosphorylation of cPLA2

at Ser505 by ERK1/2 MAP kinase is pre eminent in theregulation of cPLA2 activity. Western blot analysis usinga phosphorylation site-specific polyclonal antibody demon-strated that basal level of cPLA2 phosphorylation at Ser505in the MCF-7 cells was extremely low (Fig. 1); however,E2 (10 nM) stimulated the rapid phosphorylation of cPLA2

at this site within 1 min of treatment. The phosphoryla-tion was transient and returned to basal levels within afurther 2 min. The transient phosphorylation of cPLA2 inresponse to E2 is the result of activation of at least oneprotein kinase-dependent signaling pathway, most probablyERK1/2.

s t e r o i d s 7 1 ( 2 0 0 6 ) 256–265 259

Fig. 1 – Phosphorylation of cPLA2 in response to E2. MCF-7cells were treated with 10 nM E2 or vehicle only.Phosphorylation of cPLA2 at Ser505 was examined over a10 min time course by Western blot using a site-specificantibody. Equivalent samples were probed with a totalcPLA2 antibody. The mean densitometry measurementsfrom three independent experiments ±S.E. mean (*p < 0.05,***p < 0.001) are presented as a bar chart with arepresentative Western blot from one of these experiments.

3.2. Estrogen promotes the translocation of cPLA2 toperinuclear membranes

Translocation of cPLA2 to perinuclear membranes is an indi-cator of its activation. We examined the effect of E2 on thesub-cellular distribution of enhanced green fluorescent pro-tein (EGFP) tagged cPLA2� by confocal microscopy. A plas-mid expressing EGFP tagged cPLA2� was transfected into sub-confluent MCF-7 cells with approximately 60% efficiency oftransfection. The fusion protein was diffusely and uniformlylocated throughout the cytoplasm of the transfected cells 12 hafter transfection. This appearance was also observed in cellstreated with vehicle for 1.5 min (Fig. 2A). After 1.5 min treat-ment with E2 the EGFP-cPLA2 relocalized to discrete perinu-clear regions in a majority of the cells (Fig. 2B). There wasa concurrent reduction in the uniformly distributed fluores-cence in the peripheral cytoplasm of the responding cells.The variability of the effect in the cellular population maybe attributed to the transfected cells being subconfluent atthe time of investigation and correlates with the non-uniformcalcium responses of MCF-7 cells that have not formed a con-fluent monolayer. The E2 dependent perinuclear translocationof EGFP-cPLA2� was blocked by inhibition of MEK (Fig. 2C),with PD98059 (20 �M) and partially blocked by inhibition of ER(Fig. 2D) with ICI 182,780 (10 �M). The translocation of EGFP-cPLA2� in response to E2 is thus dependent upon activation

Fig. 2 – Translocation of EGFP-cPLA2� in response to E2.MCF-7 cells expressing GFP-cPLA2� were treated withvehicle (A) or 10 nM E2 (B) and the sub-cellular localisationof the fusion protein was investigated by confocalmicroscopy 1.5 min after treatment. The translocation ofEGFP-cPLA2 in response to E2 was also investigatedfollowing pretreatment with the MEK inhibitor PD98059 (C)or the ER inhibitor ICI 182,780 (D).

of the Raf-MEK-ERK1/2 pathway through the stimulation of areceptor harboring the classical E2 binding site.

Cytoplasmic protrusion and the development of apop-totic bodies were observed in the nuclei of cells expressingthe cPLA2 fusion protein for periods greater than 24 h (datanot shown). This is in agreement with the observation thatincreased PLA2 activity promotes cellular apoptosis throughthe activation of the caspase family proteases [30].

3.3. Estrogen activates cPLA2 through a membraneassociated receptor

Classical estrogen receptors (ER) are cytosolic in distributionbut translocated to the nucleus upon activation. There is alsogrowing evidence for dimerization of a truncated, classical ERat the cytoplasmic membrane following ligand binding and itsparticipation in rapid signaling events [36]. The initiation ofcell signaling and transcription through ER is blocked by com-petitive inhibition with estrogen antagonists such as tamox-ifen and ICI 182,780. MCF-7 cells express ER� and ER� isoformsof the classical cytosolic receptor and the truncated form ofER�. In this study, preincubation of MCF-7 cells with the ERinhibitor ICI 182,780 (10 �M) blocked the transient phosphory-lation of cPLA2 in response to E2 demonstrating the involve-ment of an estrogen receptor in initiating this signaling event(Fig. 3A). It has been documented that E2 stimulates a rapidactivation of ERK1/2 MAP kinase in MCF-7 cells and that this

is inhibited through pre-treatment with ICI 182,780 [17].

260 s t e r o i d s 7 1 ( 2 0 0 6 ) 256–265

Fig. 3 – ER is involved in initiating the phosphorylation ofcPLA2. MCF-7 cells were pretreated with ICI 182,780 orvehicle control for 1 h prior to treatment with E2 (A). MCF-7cells were treated with E2 or cell impermeable E2-BSA (B).Phosphorylation of cPLA2 at Ser505 was examined over a5 min time course by Western blot. Equivalent sampleswere probed with a total cPLA2 antibody. The meandensitometry measurements from three independentexperiments ±S.E. mean (*p < 0.05, ***p < 0.001) arepresented as a bar chart with a representative Western blotfrom one of these experiments.

Steroids can freely diffuse across biological membranes toactivate intracellular receptors. Truncated forms of the estro-gen receptor are membrane associated and coupled throughG-proteins to intracellular signaling pathways. Activation ofsuch receptors has been demonstrated using impeded ligands,

such as E2 coupled to biotin or BSA. When comparing the meanvalues obtained from three independent experiments, E2-BSAstimulated the phosphorylation of cPLA2 at Ser505 over thesame time period as free E2 but with a reduced maximal activ-ity (Fig. 3B). E2-BSA washed over a 5-kDa cut off membranefilter also stimulated cPLA2 phosphorylation (data not shown).This observation confirms that cPLA2 is activated though asurface exposed receptors rather than the classical cytosolicreceptor that directly regulates ERE dependent transcription.

3.4. The activation of cPLA2 is ERK1/2 dependent

The phosphorylation of cPLA2 by a protein kinase is pre-requisite to its activation and translocation to intracellularmembranes. The phosphorylation of residue Ser505 is mostclosely linked to biological activity. Ser505 is flanked by theconsensus sequence for ERK1/2 directed phosphorylation andERK1/2 regulates cPLA2 activation in response to other ago-nists. Here, we determined that the phosphorylation of cPLA2

at Ser505 in response to E2 treatment could be abolished bypre-treatment with the MEK inhibitor PD98059 (20 �M) demon-strating that the activation of cPLA2 by E2 was dependenton the MEK/ERK1/2 signaling pathway (Fig. 4A). Preincuba-tion with PD98059 also inhibited the E2 induced activation ofERK1/2 at 30 s and 5 min after E2 treatment (Fig. 4B). The ini-tial phase of ERK1/2 activation around 30 s to 1 min after E2

treatment thus coincides with the observed transient peak in

cPLA2 phosphorylation. The second phase of ERK activationoccurs at least 2 min after the peak in cPLA2 phosphorylationand coincides the increase in [Ca2+]i.

3.5. Eicosanoid signaling regulates the estrogeninduced rapid calcium response

We confirmed that E2 induced a rise in [Ca2+]i and estab-lished that PLA2 activity played a role in this effect. Treatmentwith physiological concentrations of E2 (10 nM) caused a rapid,two- to three-fold increase in the cytosolic calcium levels ofall the cells in a confluent MCF-7 monolayer after 2–3 min(Fig. 5A). The cytosolic Ca2+ levels returned to basal levelsafter a further 3–4 min. Treatment with a vehicle control didnot stimulate such a calcium response. However, treatmentof these cells with the endoplasmic reticulum Ca2+-ATPaseinhibitor thapsigargin (1 �M) did elicit an immediate rise inintracellular calcium levels, demonstrating that the cells werecapable of calcium mobilization. A sub-confluent populationof MCF-7 cells showed considerable variation in their ability torespond to E2 treatment and those cells that did respond wereasynchronous in their calcium response (data not shown).A similar observation has been made by other workers [17].Pre-incubation of the cells with quinacrine an inhibitor of allPLA2 isoforms completely abolished the E2 induced calciumresponse in the MCF-7 cells, demonstrating that PLA2 activa-tion was essential for this effect (Fig. 5B). Indomethacin, aninhibitor of COX enzymes, also abolished the E2 induced cal-cium transient (Fig. 5B). Incubation of the cells with quinacrineor indomethacin alone did not change intracellular calciumlevels over the course of these experiments; however, the cellsdisplayed an attenuated response to thapsigargin by theseinhibitors. The eicosanoid signaling pathway thus mediates

s t e r o i d s 7 1 ( 2 0 0 6 ) 256–265 261

Fig. 4 – The Raf-MEK-ERK1/2 pathway stimulates cPLA2

activation in response to E2. MCF-7 cells were pretreatedwith the MEK inhibitor PD98059 before E2 addition or justtreated with E2. The effect of the inhibitor on cPLA2

phosphorylation was examined by Western blotting withanti-phosphoSer505 antibody (A). The efficacy of theinhibitor in blocking MEK phosphorylation of ERK1/2Western blotting using was demonstrated usinganti-phosphoThr202/phosphoTyr204 (B). Equivalentsamples were probed with a total cPLA2 or total ERK1/2antibody. The mean densitometry measurements fromthree independent experiments ±S.E. mean (*p < 0.05,***p < 0.001) are presented as a bar chart with arepresentative Western blot from one of these experiments.

Fig. 5 – Effect of E2 on [Ca2+]i. MCF-7 cells were treated with10 nM E2 (black line) and vehicle (red line) control (A). Theeffect of inhibition of the prostaglandin synthetic pathwayon the E2 induced increase in intracellular calciumconcentration [Ca2+]i. MCF-7 cells incubated with the PLA2

inhibitor quinacrine for 15 min (black line) and COXinhibitor indomethacin for 30 min (red line) beforetreatment with 10 nM E2 (B). Cells that did not respond to E2

in the presence of inhibitors were treated with 1 �Mthapsigargin to demonstrate their potential for Ca2+

mobilization. Measurement of [Ca2+]i was by Fura-2spectrafluorescence and was expressed as nM Ca2+. Thedata represent mean ± S.E. mean of at least fiveindependent experiments. The mean [Ca2+]i for 1 min (thepoint of E2 addition) 5 min and 10 min after commencementof the experiment for E2 treated, E2 and quinacrinepretreatment, E2 and indomethacin pretreatment andvehicle control ±S.E. mean (C). (For interpretation of thereferences to colour in this figure legend, the reader isreferred to the web version of the article.)

262 s t e r o i d s 7 1 ( 2 0 0 6 ) 256–265

the E2 induced transient increase in cytosolic Ca2+ levels inthe MCF-7 breast cancer cell line.

4. Discussion

The availability of free AA is dependent upon PLA2 activitywithin the cell. Calcium dependent and independent isoformsof this enzyme contribute to this activity; however, the releaseof AA in signaling processes is generally catalyzed by cPLA2

isoforms since iPLA2 isoforms principally play a homeostaticrole in membrane lipid remodeling [26]. Phosphorylation ofcPLA2 is essential for its activation and phosphorylation atSer505 appears to correlate most closely with increased bio-logical activity although phosphorylation at other sites hasalso been observed. MAP kinases (ERK1/2 and p38) phos-phorylate cPLA2 at residue Ser505 [37,38], MAPK interactingkinase I, a basotrophic kinase phosphorylates cPLA2 at Ser727and calmodulin dependent kinase II (CAMKII) phosphorylatescPLA2 at Ser515 [39]. We have demonstrated a role for ERK1/2in the E2 induced phosphorylation at Ser505 and subsequenttranslocation of cPLA2 to perinuclear membranes. However,the well-characterized activation of ERK1/2 by E2 after 5 mintreatment occurred subsequent to and was dependent uponthe rapid calcium response [17].

The data presented here show that transient ERK1/2 acti-vation resulting in cPLA2 phosphorylation is also detectable

from the ovarian follicle is also intracellular [47], while the cal-cium is taken up from the extracellular environment by cellsof the myometrium during estrogen induced uterine contrac-tion [48]. E2 induced calcium uptake by osteoblasts is bipha-sic resulting from the combined involvement of thapsigarginsensitive stores and influx through nifedipine and verapamilsensitive L-type membrane calcium channels [49,50]. Mem-brane impermeable E2 also stimulates the activation of L-typecalcium channels through membrane associated ER in proxi-mal renal tubules [51] and in prolactinoma pituitary cells by aprocess that is insensitive to internal store depletion [52].

The different sources of calcium reflect the different sig-naling pathways that regulate the increase in its cytosolicconcentration. We have shown that the E2 induced calciumresponse in isolated colonic crypts involves a PKC/PKA depen-dent activation of verapamil sensitive membrane ion channels[20]. A similar observation has been made for the pituitary cellcalcium influx [52]. The calcium influx in the crypts is sensi-tive to inhibition of eicosanoid signaling as we demonstratehere for the MCF-7 cell line. This would imply that the mobi-lization of intracellular calcium stores in the crypts leads tothe activation of store operated calcium channels in the cellmembrane, but which in the MCF-7 cells leads only to a briefcalcium transient. The role of AA metabolism was previouslyexamined in the rapid responses to E2 in the MCF-7 cell line[56]; here the activation of PKC� by E2 was mediated by a mem-brane ER that was insensitive to ICI 182,780 inhibition and did

before the calcium response, within 1 min of E2 treatment. Thesignaling pathways leading to ERK1/2 activation in responseto E2 treatment of MCF-7 cells have been extensively studied.PKC� regulates the delayed ERK1/2 activation that occurs 1 hafter E2 treatment [16]. The more rapid activation that occurs3–15 min after E2 treatment requires a rise in cytosolic calciumlevels and is sensitive to ER antagonists such as ICI 182,780[17,40]. This rapid activation is also dependent on membraneassociated ER that interacts with Shc [40,41]. The transientactivation of ERK1/2 we observed 30 s to 1 min after E2 treat-ment coincided with cPLA2 phosphorylation and mirrors thebiphasic ERK1/2-dependent activation of PKB observed in E2

treated HepG2 cells [42]. Here PKB activation peaked twice,once within 3 min and also again 10 min after hormone treat-ment. Furthermore, PKB activation was sensitive to inhibitionof MEK with PD98059. Inhibition of cPLA2 catalytic activity byquinacrine completely blocked the calcium response. Phos-phorylation of cPLA2 at other serine residues by other kinasesknown to be facets of the rapid E2 responses has also beenpostulated [33]. Phosphorylation of cPLA2 by PKC [43–45] andPKA [46] has also been reported; however, this has only beendemonstrated in vitro in the case of PKA. Our data demon-strates that cPLA2 is phosphorylated by ERK1/2 in response toE2 but does not preclude the involvement of other kinases inregulating its activity.

A transient rise in [Ca2+]i is a significant feature of therapid signaling responses stimulated by steroid hormones.The source of this calcium may be intracellular organellestores or influx from the extracellular environment throughmembrane calcium channels. In the case of the E2 inducedcalcium transient in MCF-7 cells, the source is the thapsigar-gin sensitive stores in the endoplasmic reticulum [17]. Thecalcium source for the transient in E2 treated granulosa cells

not involve AA release or conversion to other prostanoids. Thefemale gender specific increase in [Ca2+]i in E2 treated chon-drocytes is also the result of internal store release initiated byan ICI 182,780 insensitive ER and appears to be mediated byPLC [57]. Thus, the similar calcium responses may be initiatedthrough different signaling pathways in different cell types.Our observation that ERK1/2 regulates the calcium response inMCF-7 cells does not preclude the involvement of PLC and PKCactivation in synergistically regulating the response in this celltype since PLA2 is also a potential substrate for PKC-dependentphosphorylation. However, signaling processes linked to PLCrather than PLA are pre eminent in regulating the rapid E2 cal-cium response in chondrocytes.

The lysophospholipid products of PLA2 activity modu-late [Ca2+]i by regulating Ca2+ release/uptake by mitochon-dria [53] and the opening of P-type calcium channels inthe cell membrane [54]. AA can directly activate G-proteindependent signaling or be metabolized into other eicosanoidmolecules such as prostanoids, leukotrienes and thrombox-anes. Human amnion-like WISH cells rapidly release AA andPGE2 on E2 treatment through a signaling cascade linked tomembrane associated ER� [24]. Indomethacin inhibits the COXenzymes that catalyze the two major rate-limiting steps in themetabolism AA into other eicosanoid secondary messengerssuch as PGE2. Pre-incubation of MCF-7 cells for 15 min withindomethacin (1 �M) before E2 treatment, resulted in a par-tial block of the calcium response. A complete inhibition ofthe E2 response was achieved when the incubation time withindomethacin was extended to 30 min; however, the responseto thapsigargin was not blocked in the same way (Fig. 1B).This data supports a role for the conversion of AA to eitherprostaglandin E2 (PGE2) or thromboxane in the regulation ofthe rapid calcium response to E2 in the MCF-7 cells rather

s t e r o i d s 7 1 ( 2 0 0 6 ) 256–265 263

than a mechanism involving a lysophospholipid or AA itself.The action of PLA2 may serve as the alternative rate-limitingstep in the biosynthesis of lipids regulating this pathway sincethe response is too rapid to allow for changes in the expres-sion levels of the constitutively active COX enzymes. A changein the availability of the AA substrate thus couples the COXpathway to the calcium events. PGE2 is released following1�,25-(OH)2D3 treatment of chondrocytes and acts through theGq coupled EP1 receptor to activate PKC and PKA [25]. Bothof these kinase families are also implicated in the rapid cal-cium response of the colonic epithelium to E2 treatment andmay play a role in the calcium response to E2 in MCF-7 cells.Prostaglandin receptors are located in the cell membrane,however EP3� has also been isolated in an active conforma-tion from the endoplasmic reticulum [55]. The endoplasmicreticulum is also the major storage organelle for intracellularCa2+ from where it can be released into the cytosol. Endoplas-mic reticulum associated PGE2 receptors may thus provide amechanism for the rapid release of stored calcium in responseto cPLA2 activation.

The rapid signaling events stimulated by E2 result in theactivation of multiple signaling pathways (reviewed in [58]).These events are initiated by the binding of E2 to a mem-brane associated, G-protein coupled receptor. The identity ofthis receptor(s) has been the subject of speculation. However,there now appears to be three candidate receptors that eitherparticipate in different aspects of the rapid estrogen responseoemaEa4vb[ttdtcabecScrict

paaewtw

Fig. 6 – The role of cPLA2 in the E2 stimulated calciumresponse. The E2 induced rapid responses are mediated bymembrane associated, G-protein coupled receptors (R). Thebinding of E2 to these receptors stimulate cell signalingcascades that result in the activation of protein kinasessuch as PKA, PKC and ERK1/2 MAPK for which cPLA2 is aknown substrate. Phosphorylation of cPLA2 at Ser505 byERK1/2 results in its translocation to membrane boundorganelles where arachidonic acid (AA) andlysophospholipids are liberated from membranephospholipids. AA is converted to PGH2 by cyclooxygenase(COX) enzymes and PGH2 is converted to activeprostaglandins including PGE2 by membrane-associatedprostaglandin E2 synthase. The prostanoids produced bythis pathway rapidly stimulate the release of calcium frominternal stores and at least in MCF-7 cells this isprerequisite for the E2 dependent activation of ERK1/2MAPK.

be established; however, the rapid activation of these signal-ing pathways may contribute to E2 stimulated breast cancerprogression.

Acknowledgements

The plasmid expressing EGFP-cPLA2� was a gift from Dr. ASheridan (Massachusetts General Hospital and the Depart-ments of Medicine, Harvard Medical School, Charlestown,MA). This work was supported by the Wellcome Trust (ProgramGrant No. 040067/Z/93) and by the Higher Education Authority(HEA) of Ireland. The authors would like to acknowledge theassistance of Dr. Derek Carter in analysis of the intracellularcalcium data.

r synergize to activate particular cellular responses. Het-rologously expressed, full length ER� is associated with theembrane of transfected CHO cells and stimulated ERK1/2

ctivation in response to estrogen [59]. The dimerization ofR� is prerequisite to E2 induced EGFR transphosphorylationnd the subsequent activation of c-src and ERK1/2. A truncated6-kDa alternative splice variant of ER� has been described inascular endothelial cells that is recruited to the cell mem-rane in response to E2 and regulates the release of nitric oxide

36]. E2-BSA was more effective than free E2 in stimulatinghis response suggesting the surface exposure of this recep-or’s E2 binding site. A truncated form of ER� has also beenescribed in the MCF-7 cell lines and its abundance relativeo the full length 66-kDa receptor varies according to growthonditions [60]. Thirdly, a novel GPCR like protein GPR30, haslso been proposed as a non-classical E2 receptor with highinding affinities for E2 and ICI 182,780 [61]. GPR30 is widelyxpressed in E2-responsive tissues including breast cancerells, and appears to initiate G�s coupled signaling events.ince each of these receptors is expressed in breast cancerells and are sensitive to inhibition by anti-estrogens, theirole in initiating different aspects of the rapid E2 responses ismportant in elucidating the contribution of E2 to breast can-er progression and our studies do not distinguish betweenheir individual roles.

In summary, we have demonstrated that E2 activates therostanoid signaling pathway in MCF-7 breast cancer cellsnd this pathway regulates the rapid increase in [Ca2+]i. Thectivation of cPLA2 is mediated by a membrane associatedstrogen receptor that rapidly transduces a signal to ERK1/2,hich phosphorylates cPLA2 and promotes its translocation

o perinuclear membranes (Fig. 6). The precise mechanism byhich PGE2 modulates the cytosolic calcium flux remains to

264 s t e r o i d s 7 1 ( 2 0 0 6 ) 256–265

r e f e r e n c e s

[1] Carrer HF, Cambiasso MJ. Sexual differentiation of thebrain: genes, estrogen, and neurotrophic factors. Cell MolNeurobiol 2002;22(5–6):479–500.

[2] Britt KL, Findlay JK. Estrogen actions in the ovary revisited.J Endocrinol 2002;175(2):269–76.

[3] Kloosterboer HJ, Ederveen AG. Pros and cons of existingtreatment modalities in osteoporosis: a comparisonbetween tibolone, SERMs and estrogen (±progestogen)treatments. J Steroid Biochem Mol Biol 2002;83(1–5):157–65.

[4] Anderson E, Clarke RB, Howell A. Estrogen responsivenessand control of normal human breast proliferation. JMammary Gland Biol Neoplasia 1998;3(1):23–35.

[5] Santen RJ, Pinkerton J, McCartney C, Petroni GR. Risk ofbreast cancer with progestins in combination withestrogen as hormone replacement therapy. J ClinEndocrinol Metab 2001;86(1):16–23.

[6] Nilsson S, Makela S, Treuter E, Tujague M, Thomsen J,Andersson G, et al. Mechanisms of estrogen action.Physiol Rev 2001;81(4):1535–65.

[7] Losel R, Wehling M. Nongenomic actions of steroidhormones. Nat Rev Mol Cell Biol 2003;4(1):46–56.

[8] Segars JH, Driggers PH. Estrogen action and cytoplasmicsignaling cascades. Part I: membrane-associated signalingcomplexes. Trends Endocrinol Metab 2002;13(8):349–54.

[9] Driggers PH, Segars JH. Estrogen action and cytoplasmicsignaling pathways. Part II: the role of growth factors and

phosphorylation cascade. Proc Natl Acad Sci USA2002;99(23):14783–8.

[19] Doolan CM, Harvey BJ. Modulation of cytosolic proteinkinase C and calcium ion activity by steroid hormones inrat distal colon. J Biol Chem 1996;271(15):8763–7.

[20] Doolan CM, Condliffe SB, Harvey BJ. Rapid non-genomicactivation of cytosolic cyclic AMP-dependent proteinkinase activity and [Ca2+]i by 17beta-oestradiol in femalerat distal colon. Br J Pharmacol 2000;129(7):1375–86.

[21] Harvey BJ, Condliffe S, Doolan CM. Sex and salt hormones:rapid effects in epithelia. News Physiol Sci 2001;16:174–7.

[22] Aronica SM, Kraus WL, Katzenellenbogen BS. Estrogenaction via the cAMP signaling pathway: stimulation ofadenylate cyclase and cAMP-regulated gene transcription.Proc Natl Acad Sci USA 1994;91(18):8517–21.

[23] Harvey BJ, Alzamora R, Healy V, Renard C, Doolan CM.Rapid responses to steroid hormones: from frog skin tohuman colon. A homage to Hans Ussing. Biochim BiophysActa 2002;1566(1–2):116–28.

[24] Fiorini S, Ferretti ME, Biondi C, Pavan B, Lunghi L,Paganetto G, et al. 17Beta-Estradiol stimulatesarachidonate release from human amnion-like WISH cellsthrough a rapid mechanism involving a membranereceptor. Endocrinology 2003;144(8):3359–67.

[25] Sylvia VL, Del Toro Jr F, Hardin RR, Dean DD, Boyan BD,Schwartz Z. Characterization of PGE2 receptors (EP) andtheir role as mediators of 1alpha,25-(OH)2D3 effects ongrowth zone chondrocytes. J Steroid Biochem Mol Biol2001;78(3):261–74.

[26] Murakami M, Kudo I. Phospholipase A2. J Biochem (Tokyo)2002;131(3):285–92.

[27] Jakobsson PJ, Thoren S, Morgenstern R, Samuelsson B.Identification of human prostaglandin E synthase: amicrosomal, glutathione-dependent, inducible enzyme,constituting a potential novel drug target. Proc Natl AcadSci USA 1999;96(13):7220–5.

[28] Murakami M, Naraba H, Tanioka T, Semmyo N, NakataniY, Kojima F, et al. Regulation of prostaglandin E2biosynthesis by inducible membrane-associatedprostaglandin E2 synthase that acts in concert withcyclooxygenase-2. J Biol Chem 2000;275(42):32783–92.

[29] Breyer MD, Breyer RM. G protein-coupled prostanoidreceptors and the kidney. Annu Rev Physiol2001;63:579–605.

[30] Cummings BS, McHowat J, Schnellmann RG. PhospholipaseA2s in cell injury and death. J Pharmacol Exp Ther2000;294(3):793–9.

[31] Sheridan AM, Sapirstein A, Lemieux N, Martin BD, KimDK, Bonventre JV. Nuclear translocation of cytosolicphospholipase A2 is induced by ATP depletion. J BiolChem 2001;276(32):29899–905.

[32] Fatima S, Yaghini FA, Ahmed A, Khandekar Z, Malik KU.CaM kinase IIalpha mediates norepinephrine-inducedtranslocation of cytosolic phospholipase A2 to the nuclearenvelope. J Cell Sci 2003;116(Pt 2):353–65.

[33] Evans JH, Fergus DJ, Leslie CC. Regulation of cytosolicphospholipase A2 translocation. Adv Enzyme Regul2003;43:229–44.

[34] Gijon MA, Spencer DM, Siddiqi AR, Bonventre JV, Leslie CC.Cytosolic phospholipase A2 is required for macrophagearachidonic acid release by agonists that Do and Do notmobilize calcium. Novel role of mitogen-activated proteinkinase pathways in cytosolic phospholipase A2 regulation.J Biol Chem 2000;275(26):20146–56.

[35] Qiu ZH, Gijon MA, de Carvalho MS, Spencer DM, Leslie CC.The role of calcium and phosphorylation of cytosolicphospholipase A2 in regulating arachidonic acid release inmacrophages. J Biol Chem 1998;273(14):8203–11.

phosphorylation in estrogen signalling. Trends EndocrinolMetab 2002;13(10):422–7.

[10] Harvey BJ, Doolan CM, Condliffe SB, Renard C, Alzamora R,Urbach V. Non-genomic convergent and divergentsignalling of rapid responses to aldosterone and estradiolin mammalian colon. Steroids 2002;67(6):483–91.

[11] Winter DC, Taylor C, COS G, Harvey BJ. Mitogenic effectsof oestrogen mediated by a non-genomic receptor inhuman colon. Br J Surg 2000;87(12):1684–9.

[12] Doolan CM, Harvey BJ. A Galphas protein-coupledmembrane receptor, distinct from the classical oestrogenreceptor, transduces rapid effects of oestradiol on [Ca2+]i

in female rat distal colon. Mol Cell Endocrinol2003;199(1–2):87–103.

[13] Beyer C, Raab H. Nongenomic effects of oestrogen:embryonic mouse midbrain neurones respond with arapid release of calcium from intracellular stores. Eur JNeurosci 1998;10(1):255–62.

[14] Nakajima T, Iwasawa K, Oonuma H, Morita T, Goto A,Wang Y, et al. Antiarrhythmic effect and its underlyingionic mechanism of 17beta-estradiol in cardiac myocytes.Br J Pharmacol 1999;127(2):429–40.

[15] Osborne CK, Shou J, Massarweh S, Schiff R. Crosstalkbetween estrogen receptor and growth factor receptorpathways as a cause for endocrine therapy resistance inbreast cancer. Clin Cancer Res 2005;11(2 Pt 2):865s–70s.

[16] Keshamouni VG, Mattingly RR, Reddy KB. Mechanism of17-beta-estradiol-induced Erk1/2 activation in breastcancer cells. A role for HER2 AND PKC-delta. J Biol Chem2002;277(25):22558–65.

[17] Improta-Brears T, Whorton AR, Codazzi F, York JD, MeyerT, McDonnell DP. Estrogen-induced activation ofmitogen-activated protein kinase requires mobilization ofintracellular calcium. Proc Natl Acad Sci USA1999;96(8):4686–91.

[18] Wong CW, McNally C, Nickbarg E, Komm BS, Cheskis BJ.Estrogen receptor-interacting protein that modulates itsnongenomic activity-crosstalk with Src/Erk

s t e r o i d s 7 1 ( 2 0 0 6 ) 256–265 265

[36] Li L, Haynes MP, Bender JR. Plasma membrane localizationand function of the estrogen receptor alpha variant (ER46)in human endothelial cells. Proc Natl Acad Sci USA2003;100(8):4807–12.

[37] Kramer RM, Roberts EF, Um SL, Borsch-Haubold AG,Watson SP, Fisher MJ, et al. p38 mitogen-activated proteinkinase phosphorylates cytosolic phospholipase A2 (cPLA2)in thrombin-stimulated platelets. Evidence thatproline-directed phosphorylation is not required formobilization of arachidonic acid by cPLA2. J Biol Chem1996;271(44):27723–9.

[38] Lin LL, Wartmann M, Lin AY, Knopf JL, Seth A, Davis RJ.cPLA2 is phosphorylated and activated by MAP kinase. Cell1993;72(2):269–78.

[39] Muthalif MM, Benter IF, Uddin MR, Malik KU.Calcium/calmodulin-dependent protein kinase IIalphamediates activation of mitogen-activated protein kinaseand cytosolic phospholipase A2 in norepinephrine-inducedarachidonic acid release in rabbit aortic smooth musclecells. J Biol Chem 1996;271(47):30149–57.

[40] Song RX, McPherson RA, Adam L, Bao Y, Shupnik M,Kumar R, et al. Linkage of rapid estrogen action to MAPKactivation by ERalpha-Shc association and Shc pathwayactivation. Mol Endocrinol 2002;16(1):116–27.

[41] Zhang Z, Maier B, Santen RJ, Song RX. Membraneassociation of estrogen receptor alpha mediates estrogeneffect on MAPK activation. Biochem Biophys Res Commun2002;294(5):926–33.

[42] Marino M, Acconcia F, Trentalance A. Biphasicestradiol-induced AKT phosphorylation is modulated byPTEN via MAP kinase in HepG2 cells. Mol Biol Cell

of the lower urinary tract. Eur J Pharmacol1986;127(1–2):37–42.

[49] Lieberherr M, Grosse B, Kachkache M, Balsan S. Cellsignaling and estrogens in female rat osteoblasts: apossible involvement of unconventional nonnuclearreceptors. J Bone Miner Res 1993;8(11):1365–76.

[50] Le Mellay V, Grosse B, Lieberherr M. Phospholipase C betaand membrane action of calcitriol and estradiol. J BiolChem 1997;272(18):11902–7.

[51] Han HJ, Lee YH, Park SH. Estradiol-17beta-BSA stimulatesCa2+ uptake through nongenomic pathways in primaryrabbit kidney proximal tubule cells: involvement of cAMPand PKC. J Cell Physiol 2000;183(1):37–44.

[52] Bulayeva NN, Wozniak AL, Lash LL, Watson CS.Mechanisms of membrane estrogenreceptor-alpha-mediated rapid stimulation of Ca2+ levelsand prolactin release in a pituitary cell line. Am J PhysiolEndocrinol Metab 2005;288(2):E388–97.

[53] Dalton S, Hughes BP, Barritt GJ. Effects oflysophospholipids on Ca2+ transport in rat livermitochondria incubated at physiological Ca2+

concentrations in the presence of Mg2+, phosphate andATP at 37 ◦C. Biochem J 1984;224(2):423–30.

[54] Yang L, Andrews DA, Low PS. Lysophosphatidic acid opensa Ca++ channel in human erythrocytes. Blood2000;95(7):2420–5.

[55] Boer U, Neuschafer-Rube F, Moller U, Puschel GP.Requirement of N-glycosylation of the prostaglandin E2receptor EP3beta for correct sorting to the plasmamembrane but not for correct folding. Biochem J2000;350(Pt 3):839–47.

2003;14(6):2583–91.[43] Akiba S, Mizunaga S, Kume K, Hayama M, Sato T.

Involvement of group VI Ca2+-independent phospholipaseA2 in protein kinase C-dependent arachidonic acidliberation in zymosan-stimulated macrophage-like P388D1cells. J Biol Chem 1999;274(28):19906–12.

[44] Chen BC, Lin LL, Lin WW. Protein kinase Cepsilon-dependent pathway of extracellularsignal-regulated protein kinase activation by P2Y1 andP2Y2 purinoceptors that activate cytosolic phospholipaseA2 in endothelial cells. Eur J Pharmacol 1999;373(1):101–10.

[45] Wang HQ, Kim MP, Tiano HF, Langenbach R, Smart RC.Protein kinase C-alpha coordinately regulates cytosolicphospholipase A2 activity and the expression ofcyclooxygenase-2 through different mechanisms in mousekeratinocytes. Mol Pharmacol 2001;59(4):860–6.

[46] Leslie CC. Properties and regulation of cytosolicphospholipase A2. J Biol Chem 1997;272(27):16709–12.

[47] Morley P, Whitfield JF, Vanderhyden BC, Tsang BK,Schwartz JL. A new, nongenomic estrogen action: the rapidrelease of intracellular calcium. Endocrinology1992;131(3):1305–12.

[48] Batra S. Effect of estrogen and progesterone treatment oncalcium uptake by the myometrium and smooth muscle

[56] Boyan BD, Sylvia VL, Frambach T, Lohmann CH, Dietl J,Dean DD, et al. Estrogen-dependent rapid activation ofprotein kinase C in estrogen receptor-positive MCF-7breast cancer cells and estrogen receptor-negative HCC38cells is membrane-mediated and inhibited by tamoxifen.Endocrinology 2003;144(5):1812–24.

[57] Ekstein J, Nasatzky E, Boyan BD, Ornoy A, Schwartz Z.Growth-plate chondrocytes respond to 17beta-estradiolwith sex-specific increases in IP3 and intracellular calciumion signalling via a capacitative entry mechanism. Steroids2005.

[58] Levin ER. Cell localization, physiology, and nongenomicactions of estrogen receptors. J Appl Physiol2001;91(4):1860–7.

[59] Razandi M, Pedram A, Merchenthaler I, Greene GL, LevinER. Plasma membrane estrogen receptors exist andfunctions as dimmers. Mol Endocrinol 2004;18(12):2854–65.

[60] Flouriot G, Brand H, Denger S, Metivier R, Kos M, Reid G,et al. Identification of a new isoform of the humanestrogen receptor-alpha (hER-alpha) that is encoded bydistinct transcripts and that is able to repress hER-alphaactivation function 1. EMBO J 2000;19(17):4688–700.

[61] Thomas P, Pang Y, Filardo EJ, Dong J. Identity of anestrogen membrane receptor coupled to a G protein inhuman breast cancer cells. Endocrinology2005;146(2):624–32.