Estradiol-17 inhibits GnRH-induced Ca2+ in gonadotropes to regulate negative feedback on LH

release

Javed Iqbal, Olivier Latchoumanin, Ika P Sari, Richard J Lang,

Harold A Coleman, Helena C Parkington and Iain J Clarke1

Dept. of Physiology, Monash University, Clayton, Vic 3800, Australia

Running title: Estradiol -17 inhibits Ca2+ in gonadotropes

Keywords (not in the title): Pituitary, second messengers, mitogen-activated protein kinase,

nifedipine, thapsigargin

Abbreviations: Akt, Serine/Threonine kinase; CREB, cAMP responsive element binding protein; E2,

Estradiol-17 ; ER, estrogen receptor; GPR30, G protein-coupled membrane protein receptor 30;[Ca2+]i, intracellular free calcium, IP3, inositol 1, 4, 5- triphosphate 3; MAPK, mitogen-activated protein kinase; OVX, ovariectomized; pERK-1/2, phosphorylated ERK-1/2; PI3K, phosphatidylinositol 3-Kinase; PSS, physiological saline solution, TTX, tetrotodoxin

1Corresponding author: Prof. Iain Clarke, Dept. of Physiology, P.O. Box 13F, Monash University,

Clayton, Victoria 3800, Australia. Telephone: +61-3-9905-2554; Fax: +61-3-9905-2566.

Email: iain.clarke@med.monash.edu.au

DISCLOSURE STATEMENT The authors have nothing to disclose.

Endocrinology. First published ahead of print May 28, 2009 as doi:10.1210/en.2009-0092

Copyright (C) 2009 by The Endocrine Society

2

ABSTRACT 28

In pituitary gonadotropes, estrogens have biphasic actions to cause an initial negative feedback 29

followed by a positive feedback on LH secretion, but the mechanisms involved are not clearly 30

understood. To investigate the feedback effects of estrogen, we used mixed ovine pituitary cell 31

cultures (48-72h) which were treated with 10-9M estradiol-17β (E2) or vehicle followed by a pulse of 32

10-9M GnRH. Medium was collected for LH assay and cells extracted to determine activation of 33

mitogen-activated protein kinase (MAPK) (phosphorylated ERK-1/2). E2 treatment for 5 min reduced 34

GnRH-induced LH release and caused phosphorylation of ERK-1/2. E2 alone also caused 35

phosphorylation of ERK-1/2, similar to the response evoked by GnRH alone. GnRH increased 36

cytoplasmic [Ca2+]i and this was abolished by 2 min pretreatment with E2 or E-bovine serum albumen 37

conjugate. Blockade of Ca2+ channels with nifedipine had no effect on the initial peak of GnRH-38

induced increase in [Ca2+]i, but reduced its duration by 27 ± 6%. Depletion of intracellular Ca2+ stores 39

with thapsigargin prevented GnRH-induced increase in [Ca2+]i. Thapsigargin (10-7M) or nifedipine 40

(10-5M) pretreatment (15 min) of cells lowered GnRH-induced LH secretion by 30 ± 6% and 50% ± 41

4%, respectively. We conclude that inhibition of the GnRH-induced increase in [Ca2+]i in 42

gonadotropes by E2 is a likely mechanism for the negative feedback effect of E2 on LH secretion 43

involving a rapid non-genomic effect of E2. Activation of the MAPK pathway by E2 may be the 44

mechanism for the time-delayed positive feedback effect on LH secretion at the level of the 45

gonadotrope. 46

Words:24947

3

INTRODUCTION 48

In addition to the classical genomic actions mediated through nuclear estrogen receptors (ER) (1), 49

estrogens also cause rapid non-genomic intracellular signalling in a variety of cells (2). These rapid 50

actions may be initiated at the plasma membrane (3) through G-protein coupled receptor mechanisms 51

(4), but there is also evidence of rapid effects involving estrogen interaction with cytoplasmic ER (3). 52

The rapid effects of estrogens involve a variety of intracellular signalling pathways, altering ion 53

channel function and activating second messengers, including cyclic adenosine 54

monophosphate/protein kinase A (cAMP/PKA), mitogen activated protein kinase 55

(MAPK)/extracellular signal regulated kinase (ERK-1/2), and phosphatidylinositol 3-kinase (PI3-K) 56

/Akt pathways in neurons (2, 4). Estrogens also cause rapid phosphorylation of the transcription 57

factors cAMP responsive element binding protein (CREB) through activation of phospholipase C 58

(PLC)/protein kinase C (PKC), MAPK/ Ca2+-activated calmodulin-kinase or Ras / ERK pathways (2). 59

Pituitary luteinizing hormone (LH) synthesis and secretion is regulated by pulsatile release 60

of gonadotropin releasing hormone (GnRH) from the hypothalamus (5), which binds to its G-protein 61

coupled receptor (GnRH-R) on pituitary gonadotropes (6). GnRH-R activation leads to activation of 62

PLC and the generation of inositol-1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). IP3 acts on its 63

receptor on the endoplasmic reticulum which results in a rapid release of stored Ca2+ into the 64

cytoplasm (7-9). Free intracellular [Ca2+]i levels are also elevated by Ca2+ influx through voltage 65

sensitive Ca2+ channels (10-12). A rise in [Ca2+]i is essential for exocytosis of LH vesicles (14). In 66

addition to this, DAG activates PKC, which increases the Ca2+ sensitivity of the secretory process in 67

the gonadotropes of rats (13) and regulates other processes including GnRH self-priming, cytoskeletal 68

rearrangement, and activation of Rab-GTP binding proteins to promote exocytosis (6). Thus, 69

activation of PLC, production of IP3 and elevation of [Ca2+]i are key events in gonadotropes in 70

response to GnRH.Estradiol-17ß (E2) has biphasic effects on LH secretion at the level of the pituitary 71

gonadotrope, with a rapid initial suppression of LH release (negative feedback) followed by a 72

‘positive feedback’ event many hours later which is the cause of the preovulatory LH surge (14, 15). 73

Using an in vivo ovine model, we have shown that neither LHβ mRNA expression nor GnRH-R are 74

reduced during the negative feedback phase of E2 action (16, 17). Studies in vitro (18) and in vivo (19) 75

using E-conjugates indicate that the acute suppression of LH secretion is a non-genomic mechanism, 76

4

but post-receptor mechanisms involved in the acute suppression of LH by E2 have not been detailed. 77

Our in vivo studies demonstrate phosphorylation of ERK-1/2, CREB and Akt within 15 min, 30 min 78

and 90 min, respectively, following i.v. injection of E2 in ovariectomized, hypothalamo-pituitary 79

disconnected ewes in which pulsatile LH secretion was driven by constant pulsatile administration of 80

GnRH (20). Information regarding the effects of E2 on second messenger systems activated by GnRH 81

are limited (14, 21, 22) and it is not clear how such signalling may relate to rapid E2 negative feedback 82

on LH secretion. 83

In the present study we show that both GnRH and E2 cause phosphorylation of similar second 84

messengers, so it seems unlikely that these cellular responses are involved in the negative feedback 85

response. Thus, it was hypothesized that E2 may reduce the GnRH-induced increase in [Ca2+]i and in 86

this way explain E2-elicited short-term negative feedback effect. This is the first report of the rapid 87

effect of E2 on [Ca2+]i in gonadotropes in any species. 88

89

MATERIALS AND METHODS 90

All procedures were conducted with prior approval of the institutional Animal Ethics 91

Committee under the requirements of the Australian Code of Practice for the Care and Use of 92

Animals for Scientific Purposes. 93

Preparation of ovine pituitary cultures 94

Adult Corriedale ewes were ovariectomized (OVX) at least one month before the experiment 95

to remove endogenous steroid hormones. The animals were euthanized by overdose (0.5 ml/kg body 96

weight, i.v.) of sodium pentobarbital (Lethobarb,; May and Baker Pty. Ltd., Melbourne, Australia) and 97

pituitaries were collected and placed in sterile Dulbecco’s modified Eagle medium (DMEM; Sigma-98

Aldrich Inc, MO, USA) containing 0.5% bovine serum albumin (BSA). Primary cultures were 99

established as described previously (23-25) in 24 well plates or on 9 mm round glass cover slips pre-100

coated with 0.01% poly-L-lysine (Sigma-Aldrich). The cells were cultured at a density of 0.5 x 101

106/well in 1 ml DMEM containing bovine 10% fetal calf serum (FCS) and antibiotic, and incubated 102

in a humidified atmosphere (37°C, 5 % CO2). Culture medium was replenished after 24 h and cells 103

were studied after 48 and/or 72 h in culture. 104

5

Experiment 1: E2 effects on LH secretion and activation of MAPK (ERK- 1/2) pathways in 105

gonadotropes106

On the day of experimentation, the medium was removed and the cells were incubated in 107

serum free DMEM containing 0.1% BSA for 1-2 h. Cells from the same animal were treated with 108

either vehicle or E2 (10−9 M) for 0, 5, 15, 30, and 60 min (duplicate wells) followed by stimulation of 109

LH release by a 15 min GnRH (10−9 M) pulse and medium was then collected for assay. LH levels 110

were determined by an established radioimmunoassay (26). 111

The time course of rapid activation of MAPK (phosphorylation of ERK-1/2) by E2 without or 112

with GnRH was determined. Cells (duplicate wells) were treated with vehicle, GnRH (10−9 M), E2113

(10−9 M) or E2 plus GnRH for 0, 5, 15, 30 and 60 min. The medium was removed and cells were lysed 114

with lysis buffer (1 ml per well). Well contents were immediately removed and centrifuged at 13,000 115

RPM for 20 min at 4ºC and the supernatant was collected and protein concentration determined by 116

Bradford assay (20). Phosphorylation of ERK-1/2 (pERK-1/2) was quantified by Western blot 117

analysis (see below). 118

Immunohistochemical detection of pERK-1/2 119

To determine whether E2-induced pERK-1/2 was limited to gonadotropes only or pERK-1/2 was 120

localized to other cell types as well, cells plated on glass cover slips were treated with either E2 (10−9121

M) or vehicle for 5 min, washed with 0.1 M PBS and fixed with 4% paraformaldehyde solution in 0.1 122

M phosphate-buffered saline (PBS) for 10 min and processed for double labelling 123

immunofluorescence histochemistry to measure the induction of pERK-like immunoreactivity..These 124

experiments were replicated in cultures from 3 different animals with all treatments in duplicate. . 125

Immunohistochemical detection of pERK-1/2 in gonadotropes and somatotropes was carried 126

out as described previously (20). To identify gonadotropes we used a mouse monoclonal anti-LH 127

(1:5000) antibody which recognises LHβ and does not cross react with other pituitary hormones (JF 128

Roser, University of California, Davis, USA). To identify somatotropes we used a mouse monoclonal 129

anti-growth hormone (1:10, kindly donated by Dr. M. Brandon, University of Melbourne, Australia). 130

6

Labeling of pERK-1/2 in both cell types was achieved with a rabbit monoclonal antibody (1:200) 131

(Cell Signalling Technology, MA, USA). Visualisation was achieved with goat anti-mouse Alexa 546 132

(1:500, red) for LH and GH and goat anti-rabbit Alexa 488 (1:500, green) for pERK-1/2 (Molecular133

Probes, OR, USA). Cell nuclei were counter-stained with DAPI and the immmunostaining was 134

viewed with a confocal microscope. Images were captured using a 40X and 60X oil immersion 135

objectives. These antibodies have been extensively characterised in our laboratory (20, 27, 28) and 136

negative controls included omission of the primary and/or secondary antibody.137

Western blot measurement of pERK- 1/2 138

Western blot analysis for the determination of pERK1/2 was performed as described 139

previously (20). Detection of pERK-1/2 and total ERK-1/2 was achieved with a rabbit polyclonal 140

anti-pERK1/2 antibody (1:2000) and a rabbit polyclonal anti ERK-1/2 (1:1000) (Cell Signalling) with 141

horseradish peroxidase conjugated anti-rabbit IgG second antibody (Antibodies Australia, Monash 142

University, Australia). Detection was accomplished using enhanced chemiluminescence reagent (GE 143

Healthcare, Chalfont, Buckingham, UK). Full range rainbow markers (GE Healthcare) were used to 144

determine the molecular weights of proteins, and β-actin was used as the control protein. The films 145

were scanned using a Canon scanner (Akimachi, Kunisaki-shi, Japan) and multi-gauge software 146

(Fujifilm Corp. Minatoku, Tokyo, Japan) was used to determine the band densities. 147

Experiment 2. E2 effects on GnRH-induced rise in [Ca2+]i in gonadotropes 148

The short-term effects of E2 (cells from 7 sheep) and E-BSA (cells from 5 sheep) on the 149

GnRH-stimulated increase in [Ca2+]i were determined and the source of this increase in [Ca2+]i was 150

probed using a pharmacological approach. Cells were prepared on cover slips and the experiments 151

were performed in a dark room. The cells were washed for 5 min with physiological saline solution 152

(PSS, containing 117 mM NaCl, 5 mM KCl, 2 mM MgCl2, 0.5 mM KH2PO4, 5 mM NaHCO3, 1.8 153

mM CaCl2, 10 mM HEPES, 10 mM glucose, 0.1% BSA, pH 7.35, 33°C) and were then loaded with 154

Fluo-4-AM (5×10−6 M) in PSS containing 0.01% pluronic for 30 min at room temperature (23°C). 155

The cells on cover slips were transferred to a Warner tissue bath mounted on an Olympus IX71 156

inverted confocal microscope and continuously superfused with PSS at 33°C for 20 min to allow 157

cleavage of Fluo-4-AM by cytoplasmic esterases. The cells were illuminated by a Kr/Ar laser at 488 158

7

nm, and the light passed through a Yokogawa CSU22 Nipkow spinning disc to a high sensitivity 159

electron-multiplying iXon CCD camera (Andor Inc. Belfast, Ireland). Full frames were collected at 2 160

s intervals. Control recordings were made over 5 min and the cells were then treated with 2×10-9 M 161

GnRH for 30 s (designated 0 min, exposure 1). E2 or E-BSA conjugate were added to the superfusate 162

at 9 min and GnRH was again applied for 30 s at 11 min (exposure 2). This was followed at 14 min 163

by a 30 s application of PSS containing 40 mM KCl (isosmotic substitution of NaCl) to confirm 164

integrity of the cells. For the time-matched controls, the ethanol (10% in water) vehicle used for E2165

was added to the superfusing PSS (0.1 μl/ml) at 9 min. To determine the source of GnRH-induced 166

increase in [Ca2+]i and the mechanisms by which E2 inhibited the increase in [Ca2+]i , blockers 167

including nifedipine (10-5M, to block L-type Ca2+ channels), ryanodine (10−5 M, to block Ca2+-168

induced Ca2+ release from intracellular stores), thapsigargin (10−7 M) or cylopiazionic acid (10−5 M to 169

deplete intracellular Ca2+ stores) were included in the PSS from 2 min onwards until the end of 170

experimentation. Cells from 5 sheep were analysed and all drugs were purchased from Sigma-171

Aldrich. 172

173

Experiment 3: Effects of nifedipine and thapsigargin on GnRH-stimulated LH secretion from 174

gonadotropes 175

Cells were cultured as described for experiment 1 and incubated in serum free medium for 176

experimentation. Either vehicle or thapsigargin (10-7 M) or nifedipine (10−5 M) was added (2 177

wells/treatment) followed 15 min later by a GnRH (10-9 M) pulse. After 2h exposure to GnRH, the 178

medium was harvested and stored at -20C until assayed for LH. 179

180

Experiment 4: Test of specificity of E2 rapid effects in gonadotropes181

To determine the specificity of the rapid effects of E2 on GnRH-induced LH secretion, 182

activation of MAPK (ERK-1/2), and GnRH-induced [Ca2+]i in gonadotropes, we examined the effects 183

of cortisol. Pituitary cell cultures in serum-free medium (3 replicates/animal; n=5 sheep) were pre-184

treated with cortisol (500 ng/ml) for 0, 0.5, 1, 2, 4, and 24h followed by 2h GnRH treatment. Medium 185

was collected for LH assay and cells were extracted to determine the effects of cortisol on 186

phosphorylation of ERK-1/2. Cell extracts were analysed by Western blot (as above). Cortisol effects 187

8

on GnRH-induced [Ca2+]i was also determined as described in Experiment 2.The effects of cortisol on 188

GnRH-induced [Ca2+]i and phosphorylation of ERK-1/2 are presented as supplementary data (figures 189

1-4). 190

Statistical analysis 191

The data were analysed using analysis of variance (ANOVA), with the least significant 192

differences method as a post hoc test. Data are presented as means (±SEM) and n indicates the 193

number of sheep studied. 194

195

RESULTS 196

Experiment 1: E2 effects on LH secretion and phosphorylation of ERK-1/2 197

Application of GnRH (10−9 M) for 15 min significantly (p<0.01) stimulated LH release, but 198

treatment with E2 (10−9 M) alone (not shown) or vehicle had no effect (Fig.1). Pre-treatment with E2199

(10−9 M) for 5 min blocked (p< 0.001) GnRH-induced LH release from the pituitary gonadotropes 200

(Fig. 1). Cortisol treatment of cultured cells did not reduce LH secretion within 0.5h but did so after 201

1h of treatment (Suppl. Fig. 1). These data show that there is no rapid (within 5mins) effect of 202

cortisol on the response to GnRH in gonadotropes, unlike the rapid effect observed with E2.203

204

Western blot analysis showed that treatment with either E2 (10−9 M) or GnRH (10−9 M) alone 205

induced pERK-1/2 within 5 min, and combined treatment of cells with GnRH plus E2 caused a similar 206

elevation of pERK-1/2 (Fig. 2). The level of pERK-1/2 returned to basal levels 30 min following 207

GnRH treatment, but was reduced after 15 min in cells treated with E2 alone or with combined E2 plus 208

GnRH treatment. The pERK-1/2 levels remained slightly higher in vehicle-treated cells as compared 209

with application of E2 or GnRH alone, or in combination (Fig. 2). Cortisol did not affect pERK-1/2 210

levels inmixed cell cultures, unlike the effect observed with E2 (Suppl. Fig. 2) 211

212

Double labelling immunofluorescence histochemistry revealed that 5 min exposure to E2213

(10−9 M) rapidly induced pERK–1/2–like immunoreactivity (-ir) in gonadotropes, visualized as 214

significantly brighter pERK–1/2–like–ir (Fig. 3A, panel B) compared with vehicle treatment (Fig. 3A, 215

9

panel F). All cells which were positive for pERK–1/2–like–ir also labelled with anti-LH antibody 216

(Fig. 3A, panels A, E). We did not observe any non-gonadotrope cells labelled for pERK–1/2–like–ir. 217

In addition, no pERK–1/2–like–ir was seen in GH–like–ir positive (somatotropes) cells (Fig. 3B). 218

219

Experiment 2: E2 effects on GnRH induced rise in [Ca2+]i220

GnRH (2×10−9 M) application for 30 s caused a rapid and transient rise in [Ca2+]i (Fig. 4A) 221

which was blocked in a dose-dependent manner by 2 min pre-treatment with E2 (10−9 M) (Fig. 4B & 222

C, n=7). Pre-treatment with E-BSA conjugate also reduced GnRH-stimulated rise in [Ca2+]i in a dose-223

dependent manner (Fig. 4C, n=5).E2 was, however, 10 fold more potent than E-BSA (pD2 [-log EC50] 224

8.72 ± 0.08 for E2 n= 7 sheep versus 7.67±0.15, n= 5 sheep for E-BSA, p<0.002) (Suppl. Fig. 3). 225

226

The source of the GnRH-induced rapid rise in [Ca2+]i was probed. Nifedipine (10−5 M) had no 227

effect on the peak amplitude of the [Ca2+]i transient evoked by GnRH, but it reduced the “tail” of the 228

response, reducing the integrated area under the curve of the [Ca2+]i response by 27 ± 6 % (6 plates 229

from n = 3 sheep, p<0.001) (Fig. 5A). Ryanodine (10−5 M) or caffeine (10−2 M) treatment was used 230

to release the ryanodine-sensitive stored Ca2+ and these agents were without effect on the GnRH-231

induced increase in [Ca2+]i (not shown). Thapsigargin (10−7 M) or cyclopiozenic acid (10−5 M), which 232

deplete all intracellular Ca2+ stores, completely blocked the GnRH-induced rise in [Ca2+]i (Fig 5B). 233

Cortisol (500ng/ ml) was without effect on the increase in [Ca2+]i evoked by GnRH (Suppl. Figs 4 & 234

5). 235

236

Experiment 3. Effects of nifedipine and thapsigargin on GnRH-induced LH secretion 237

GnRH (10−9 M) treatment increased (p<0.001) LH release and this response was reduced (p<0.05)by 238

thapsigargin (10−7 M) or nifedipine (10−5 M) pretreatment (30 ± 6% and 50 ± 4% respectively) (Fig. 239

6). 240

10

DISCUSSION 241

In this study we demonstrate that the suppression of GnRH-induced LH secretion by E2 is rapid and 242

involves inhibition of the rise in [Ca2+]i evoked by GnRH in ovine gonadotropes. This mechanism is 243

most likely associated with the rapid onset, negative feedback effect of E2 on LH secretion. We used 244

sheep pituitary cells in which ERα is substantially confined to the gonadotropes (27). There is also 245

minimal expression of ERβ in the ovine pituitary (29) and no other pituitary cell type expresses ERα246

or ERβ in this species (30). Therefore, the rapid effects of E2 on intracellular second messengers 247

signalling molecules is localised to this cell type alone in mixed pituitary cell cultures and the effects 248

are mediated by ERα, as has been reported in rats (31). This is further substantiated by the specific 249

immunohistochemical localisation of pERK-1/2-like-ir to gonadotropes following E2 treatment in the 250

mixed cell preparations. LH secretion is influenced by the other steroid hormones such as 251

progesterone (32, 33) and cortisol (34), but these hormones do not have rapid or biphasic effects on 252

LH secretion as does estradiol (14) and testosterone has minimal direct pituitary effect on the LH 253

secretion in the sheep (35), suggesting that the effects of E2 we observed in gonadotropes in the 254

present study are specific. 255

256

Rapid activation of second messengers, including MAPK/ERK-1/2 cascades, by E2 has been 257

shown in ovine gonadotropes (20). Consistent with this, the immunohistochemical and western blot 258

data of the present study showed up-regulation of pERK-1/2 within 5 min following E2 treatment, 259

which was short-lived (< 30 min). GnRH also increased pERK-1/2 in cultured cells and this was not 260

suppressed by pre-treatment with E2. The amplitude of the pERK-1/2 response to E2 alone was the 261

same as that to GnRH alone, as has been shown in other studies (36), suggesting that phosphorylation 262

of this second messenger is not involved in the rapid negative feedback of GnRH by E2. However, 263

treatment with E2 alone or with GnRH plus E2 resulted in the pERK-1/2 rise being more transient. 264

Thus, the transience of the ERK-1/2 phosphorylation by E2 may be important in relaying some 265

information which is essential for the time-delayed positive feedback event, which is a delayed event. 266

Further in vivo studies are in progress to test this hypothesis. This interpretation must be tempered by 267

the knowledge that, in normal animals, gonadal steroids do not act on gonadotropes in the absence of 268

background GnRH. Accordingly, the longer response phase seen when GnRH alone is applied to cells 269

11

may represent an artificial circumstance that never occurs in the animals experiencing normal ovarian 270

cycles. 271

272

Debate on the site and mechanisms of initiation of the rapid effects of E2 continues, with one 273

proposal being that liganded ERα acts on the internal face of the plasma membrane (3) to invoke a 274

mechanism via Ca2+ channels and cAMP-activated second messengers (4). It has also been proposed 275

that GPR30, as a membrane receptor, mediates the rapid effects of E2 (37), but this is not consistent 276

with the model proposed by Levin and associates (3). Our earlier study (20) and the present results 277

show that E2 has a rapid effect in ovine gonadotropes, reducing [Ca2+]i, increasing pERK-1/2 and 278

reducing LH secretion within 5 min. This prompts the question as to which of the above models is 279

operative in these cells. Our previous studies in ovariectomized sheep in vivo showed that E2280

suppresses LH secretion within 20 min (negative feed back). The present studies show that E-BSA 281

conjugate blocks the GnRH-induced elevation of [Ca2+]i within 2 min, providing support for the 282

existence of membrane ER on gonadotropes that can bind the conjugate and elicit a response without 283

entry into the cell. 284

285

GnRH-induced LH release is a Ca2+ dependent process, effected through activation of PLC, 286

leading to the generation of IP3 and DAG. IP3 leads to rise in [Ca2+]i as consequence of release from 287

internal stores (6, 38-40). GnRH also induces pERK-1/2 in gonadotropes (36). DAG activates PKC 288

which results in Ca2+ influx via voltage-gated Ca2+ channels (6, 38, 39, 41). Thus, the relative extent 289

to which the elevation of [Ca2+]i depends upon release from intracellular stores compared with influx 290

of Ca2+ into the cell is important. Previous studies in cells from rodents have indicated that the rapid 291

increase in [Ca2+]i in response to GnRH is due to release of Ca2+ from intracellular stores (10-12, 42, 292

43). Previous reports (24) and the present data indicate that this mechanism also occurs in ovine 293

gonadotropes. The rapid elevation in [Ca2+]i is required for the release of LH in response to GnRH 294

(10-12, 38, 43), but there is also a sustained elevation of [Ca2+ ]i (11), which is most likely responsible 295

for the activation of second messengers. In support of this notion, sustained release of LH induced by 296

continuous infusion of GnRH is prevented by blockade of the ERK-1/2 pathway (36). The present 297

data show that the rapid elevation of [Ca2+]i is due solely to mobilization of intracellular stores, but 298

12

the ‘tail’ of the response is reduced by nifedipine, demonstrating that a component is dependent on 299

Ca2+ influx through L-type Ca2+ channels. We hypothesized that the mechanism of the early negative300

feedback by E2 is inhibition of the GnRH-induced rise in [Ca2+]i and the present data support this. The 301

effect appears to be due to blockade of the GnRH-induced release of Ca2+ from intracellular stores, 302

since this is the predominant source of [Ca2+]i following GnRH stimulation. 303

304

In gonadotropes, we found that the GnRH-induced rapid rise in [Ca2+]i was unaffected by 305

nifedipine or ryanodine receptor blockade, but was prevented when intracellular stores were depleted 306

by thapsigargin or cylopiazionic acid.. Together, these data indicate that IP3 is responsible for the 307

elevation in [Ca2+ ]i, as has been shown in previous studies (10-12, 43). Interestingly, nifedipine, 308

which caused a 50% reduction in GnRH-stimulated LH release, also reduced the GnRH-induced rise 309

in the integrated [Ca2+]i signal by 27%. This was achieved by curtailing the duration, rather than the 310

peak of the response (see Fig. 6). These data suggest that the sustained phase involves Ca2+ influx via 311

L-type voltage-activated Ca2+ channels, as reported previously (42-44). 312

In summary, using a number of in vitro approaches, we have examined the intracellular 313

mechanism of rapid E2 effects in ovine pituitary gonadotropes. E2 alone or together with GnRH 314

rapidly activates MAPK/ERK-1/2 pathways indicated by the induction of pERK-1/2, and E2 inhibited 315

the GnRH-induced rise in [Ca2+]i. These data provide evidence for a mechanism to explain the rapid316

negative feedback effects of E2 on LH secretion. Elucidation of the means by which E2 inhibits the 317

rise in [Ca2+]i requires further studies. Activation of MAPK/ERK-1/2 pathway may be relevant to the 318

time-delayed positive feedback mechanism of E2 on LH secretion, since the activation of this 319

messenger is caused by both GnRH and E2.320

321

ACKNOWLEDGEMENTS 322

The authors thank Bruce Doughton, Jessica Thomas and Lynda Morrish for technical assistance. Dr. 323

A. Parlow of the National Hormone and Peptide Program provided LH assay reagents, for analysing 324

the plasma samples for LH. This work was supported by the Australian Research Council. We also 325

thank Antibodies Australia, Monash University, Australia for supplying the goat anti rabbit –HRP 326

secondary antibody. 327

13

REFERENCES 328329

1. Nilsson S, Makela S, Treuter E, Tujague M, Thomsen J, Andersson G, Enmark E, 330Pettersson K, Warner M, Gustafsson JA 2001 Mechanisms of estrogen action. 331Physiological reviews 81:1535-1565 332

2. Vasudevan N, Pfaff DW 2007 Membrane-initiated actions of estrogens in 333neuroendocrinology: emerging principles. Endocr Rev 28:1-19 334

3. Pedram A, Razandi M, Levin ER 2006 Nature of functional estrogen receptors at the 335plasma membrane. Mol Endocrinol 20:1996-2009 336

4. Kelly MJ, Qiu J, Ronnekleiv OK 2003 Estrogen modulation of G-protein-coupled receptor 337activation of potassium channels in the central nervous system. Ann N Y Acad Sci 1007:6-16 338

5. Clarke IJ 2002 Two decades of measuring GnRH secretion. Reproduction (Cambridge, 339England) Supplement 59:1-13 340

6. Anderson L 1996 Intracellular mechanisms triggering gonadotrophin secretion. Reviews of 341reproduction 1:193-202 342

7. Broad LM, Braun FJ, Lievremont JP, Bird GS, Kurosaki T, Putney JW, Jr. 2001 Role 343of the phospholipase C-inositol 1,4,5-trisphosphate pathway in calcium release-activated 344calcium current and capacitative calcium entry. J Biol Chem 276:15945-15952 345

8. Ma HT, Venkatachalam K, Li HS, Montell C, Kurosaki T, Patterson RL, Gill DL 2001 346Assessment of the role of the inositol 1,4,5-trisphosphate receptor in the activation of 347transient receptor potential channels and store-operated Ca2+ entry channels. J Biol Chem 348276:18888-18896 349

9. Patterson RL, Boehning D, Snyder SH 2004 Inositol 1,4,5-trisphosphate receptors as signal 350integrators. Annual review of biochemistry 73:437-465 351

10. Stojilkovic SS, Catt KJ 1992 Calcium oscillations in anterior pituitary cells. Endocr Rev 35213:256-280 353

11. Stojilkovic SS, Kukuljan M, Iida T, Rojas E, Catt KJ 1992 Integration of cytoplasmic 354calcium and membrane potential oscillations maintains calcium signaling in pituitary 355gonadotrophs. Proceedings of the National Academy of Sciences of the United States of 356America 89:4081-4085 357

12. Stojilkovic SS, Torsello A, Iida T, Rojas E, Catt KJ 1992 Calcium signaling and secretory 358responses in agonist-stimulated pituitary gonadotrophs. J Steroid Biochem Mol Biol 41:453-359467360

13. Zhu H, Hille B, Xu T 2002 Sensitization of regulated exocytosis by protein kinase C. 361Proceedings of the National Academy of Sciences of the United States of America 99:17055-36217059363

14. Clarke IJ 2002 Multifarious effects of estrogen on the pituitary gonadotrope with special 364emphasis on studies in the ovine species. Archives of physiology and biochemistry 110:62-73 365

15. Clarke IJ, Cummins JT, de Kretser DM 1983 Pituitary gland function after disconnection 366from direct hypothalamic influences in the sheep. Neuroendocrinology 36:376-384 367

16. Mercer JE, Clements JA, Funder JW, Clarke IJ 1988 Luteinizing hormone-beta mRNA 368levels are regulated primarily by gonadotropin-releasing hormone and not by negative 369estrogen feedback on the pituitary. Neuroendocrinology 47:563-566 370

17. Clarke IJ, Cummins JT, Crowder ME, Nett TM 1988 Pituitary receptors for 371gonadotropin-releasing hormone in relation to changes in pituitary and plasma gonadotropins 372in ovariectomized hypothalamo/pituitary-disconnected ewes. II. A marked rise in receptor 373number during the acute feedback effects of estradiol. Biol Reprod 39:349-354 374

18. Arreguin-Arevalo JA, Nett TM 2005 A nongenomic action of 17beta-estradiol as the 375mechanism underlying the acute suppression of secretion of luteinizing hormone. Biol 376Reprod 73:115-122 377

19. Arreguin-Arevalo JA, Nett TM 2006 A nongenomic action of estradiol as the mechanism 378underlying the acute suppression of secretion of luteinizing hormone in ovariectomized ewes. 379Biol Reprod 74:202-208 380

20. Iqbal J, Latchoumanin O, Clarke IJ 2007 Rapid in vivo effects of estradiol-17beta in ovine 381pituitary gonadotropes are displayed by phosphorylation of extracellularly regulated kinase, 382serine/threonine kinase, and 3',5'-cyclic adenosine 5'-monophosphate-responsive element-383binding protein. Endocrinology 148:5794-5802 384

14

21. McArdle CA, Poch A 1992 Dependence of gonadotropin-releasing hormone-stimulated 385luteinizing hormone release upon intracellular Ca2+ pools is revealed by desensitization and 386thapsigargin blockade. Endocrinology 130:3567-3574 387

22. McArdle CA, Schomerus E, Groner I, Poch A 1992 Estradiol regulates gonadotropin-388releasing hormone receptor number, growth and inositol phosphate production in alpha T3-1 389cells. Mol Cell Endocrinol 87:95-103 390

23. Chen C, Heyward P, Zhang J, Wu D, Clarke IJ 1994 Voltage-dependent potassium 391currents in ovine somatotrophs and their function in growth hormone secretion. 392Neuroendocrinology 59:1-9 393

24. Heyward PM, Clarke IJ 1995 A transient effect of estrogen on calcium currents and 394electrophysiological responses to gonadotropin-releasing hormone in ovine gonadotropes. 395Neuroendocrinology 62:543-552 396

25. Xu R, Zhao Y, Chen C 2002 Growth hormone-releasing peptide-2 reduces inward rectifying 397K+ currents via a PKA-cAMP-mediated signalling pathway in ovine somatotropes. The 398Journal of physiology 545:421-433 399

26. Lee VW, Cumming IA, de Kretser DM, Findlay JK, Hudson B, Keogh EJ 1976 400Regulation of gonadotrophin secretion in rams from birth to sexual maturity. I. Plasma LH, 401FSH and testosterone levels. Journal of reproduction and fertility 46:1-6 402

27. Tobin VA, Pompolo S, Clarke IJ 2001 The percentage of pituitary gonadotropes with 403immunoreactive oestradiol receptors increases in the follicular phase of the ovine oestrous 404cycle. Journal of neuroendocrinology 13:846-854 405

28. Iqbal J, Pompolo S, Considine RV, Clarke IJ 2000 Localization of leptin receptor-like 406immunoreactivity in the corticotropes, somatotropes, and gonadotropes in the ovine anterior 407pituitary. Endocrinology 141:1515-1520 408

29. Scott CJ, Rawson JA, Pereira AM, Clarke IJ 1998 The distribution of estrogen receptors 409in the brainstem of female sheep. Neuroscience letters 241:29-32 410

30. Scanlan N, Skinner DC 2002 Estradiol modulation of growth hormone secretion in the ewe: 411no growth hormone-releasing hormone neurons and few somatotropes express estradiol 412receptor alpha. Biol Reprod 66:1267-1273 413

31. Glidewell-Kenney C, Hurley LA, Pfaff L, Weiss J, Levine JE, Jameson JL 2007 414Nonclassical estrogen receptor alpha signaling mediates negative feedback in the female 415mouse reproductive axis. Proceedings of the National Academy of Sciences of the United 416States of America 104:8173-8177 417

32. Clarke IJ, Cummins JT 1984 Direct pituitary effects of estrogen and progesterone on 418gonadotropin secretion in the ovariectomized ewe. Neuroendocrinology 39:267-274 419

33. Clarke IJ, Cummins JT, Crowder ME, Nett TM 1989 Long-term negative feedback 420effects of oestrogen and progesterone on the pituitary gland of the long-term ovariectomized 421ewe. The Journal of endocrinology 120:207-214 422

34. Breen KM, Davis TL, Doro LC, Nett TM, Oakley AE, Padmanabhan V, Rispoli LA, 423Wagenmaker ER, Karsch FJ 2008 Insight into the neuroendocrine site and cellular 424mechanism by which cortisol suppresses pituitary responsiveness to gonadotropin-releasing 425hormone. Endocrinology 149:767-773 426

35. Tilbrook AJ, de Kretser DM, Cummins JT, Clarke IJ 1991 The negative feedback effects 427of testicular steroids are predominantly at the hypothalamus in the ram. Endocrinology 428129:3080-3092 429

36. Yang D, Caraty A, Dupont J 2005 Molecular mechanisms involved in LH release by the 430ovine pituitary cells. Domestic animal endocrinology 29:488-507 431

37. Filardo EJ, Thomas P 2005 GPR30: a seven-transmembrane-spanning estrogen receptor 432that triggers EGF release. Trends Endocrinol Metab 16:362-367 433

38. Weck J, Fallest PC, Pitt LK, Shupnik MA 1998 Differential gonadotropin-releasing 434hormone stimulation of rat luteinizing hormone subunit gene transcription by calcium influx 435and mitogen-activated protein kinase-signaling pathways. Mol Endocrinol 12:451-457 436

39. Kraus S, Naor Z, Seger R 2001 Intracellular signaling pathways mediated by the 437gonadotropin-releasing hormone (GnRH) receptor. Archives of medical research 32:499-509 438

40. Naor Z, Benard O, Seger R 2000 Activation of MAPK cascades by G-protein-coupled 439receptors: the case of gonadotropin-releasing hormone receptor. Trends Endocrinol Metab 44011:91-99 441

15

41. Conn PM, Rogers DC, Seay SG 1983 Structure-function relationship of calcium ion channel 442antagonists at the pituitary gonadotrope. Endocrinology 113:1592-1595 443

42. Smith CE, Wakefield I, King JA, Naor Z, Millar RP, Davidson JS 1987 The initial phase 444of GnRH-stimulated LH release from pituitary cells is independent of calcium entry through 445voltage-gated channels. FEBS letters 225:247-250 446

43. Tse A, Hille B 1993 Role of voltage-gated Na+ and Ca2+ channels in gonadotropin-releasing 447hormone-induced membrane potential changes in identified rat gonadotropes. Endocrinology 448132:1475-1481 449

44. Kile JP, Amoss MS, Jr. 1988 Role of Ca2+ and Na+ on luteinizing hormone release from the 450calf pituitary. The American journal of physiology 255:E469-474 451

452453

16

FIGURE LEGENDS 454455

Figure 1. Effects of vehicle, GnRH and E2 on LH secretion from ovine pituitary gonadotropes in 456

vitro. Cells were treated with E2 or vehicle and followed by a 15 min GnRH (10−9 M) pulse. GnRH 457

significantly (p<0.001) stimulated the LH secretion whereas pretreatment of cells with E2 (10−9 M) 458

significantly (p<0.001) blocked GnRH stimulated LH secretion from the pituitary gonadotropes in 459

vitro. The data are present as mean (±SEM) (n=5 sheep). 460

461

Figure 2. Effects of E2 on the activation (phosphorylation) of ERK-1/2 in ovine pituitary 462

gonadotropes (n= 5 sheep). Mixed pituitary cells were cultured for 72 h and then treated with E2 (10−9463

M) or vehicle for various time points followed by stimulation with a 15 min pulse of GnRH (10−9 M). 464

Cell protein extracts were analysed by Western blot analysis. Panel A presents data for one of the 465

representative replicates of one sheep. E2 and GnRH treatments significantly (p<0.01) increased 466

pERK-1/2 protein levels within 5 min after the treatment, which returned to basal levels within 30 467

min. The E2 application significantly suppressed the pERK-1/2 after 60 min as compared to the 468

GnRH alone or in combination with E2.469

470

Figure 3. E2 rapidly induced pERK-1/2 -like immunoreactivity (ir) was specific to pituitary 471

gonadotropes (Part A). Pituitary mixed cell cultures were treated with E2 or vehicle (10% ethanol in 472

water) for 5 min, washed with 0.1 M ice cold phosphate buffer and fixed with 4% paraformaldehyde 473

in 0.1 M phosphate buffer for 10 min and processed for double labelling immunfluorescence 474

histochemistry. Panel A-D shows E2 induced pERK-1/2 -ir in LH-ir positive gonadotrope (A), pERK-475

ir in gonadotropes (B) and DAPI (C) and D is a merged image of A, B and C. Panel E-H shows no 476

effects of vehicle on basal pERK-ir in LH-ir positive gonadotrope (E), pERK-ir in gonadotrope (F) 477

and DAPI (G) and H is a merged image of E, F and G. The data are presented for one of the 478

replicates from one sheep. E2 did induced pERK- 1/2 in other pituitary cell types. Part B of the figure 479

shows example of a somatotrope immunostained with mouse monoclonal anti-growth hormone 480

antibody (GH-ir) (A, red) which did not immunostained for pERK- 1/2-ir (B, green). Panel C is a 481

DAPI stained cell nuclei (blue) and panel D is a merged image of A, B and C. Asterisks indicate the 482

same place in all panels and white arrow in panel C indicates DAPI staining of a cell did not 483

17

immunostain either with anti-pERK-1/2 or anti-GH antibodies. These experiments were repeated in 5 484

sheep with 3 replicates per sheep with similar results. Scale bar in Part A= 10μm and in Part B = 485

20μm. 486

487

Figure 4. Rapid effects of E2 on intracellular free calcium [Ca2+]i. Panel A (control) shows an 488

example of measurement of [Ca2+]i in a pituitary gonadotrope in response to the first GnRH pulse 489

(2x10−9 M) (Response #1) and following a second GnRH pulse 11 min later (Response #2). Panel B 490

provides an example of GnRH induced [Ca2+]i (Response #1, control) followed by 2 min E2 treatment 491

9 min later and the second pulse of GnRH in the presence of E2 (Response #2). E2 treatment rapidly 492

blocked the GnRH induced rise in [Ca2+]i in ovine pituitary gonadotropes. Panel C shows E2 (data 493

from 7 sheep) and E-BSA conjugate (data from 5 sheep) blocked of GnRH (10−9 M) induced increase 494

in [Ca2+]i in pituitary gonadotropes in a dose-dependent manner suggesting that rapid inhibition of the 495

GnRH response is initiated from outside of the cell. 496

497

Figure 5. Nifedipine (10−5 M) had no effect on the peak amplitude of the increase in [Ca2+]i evoked in 498

gonadotropes by GnRH, while it reduced the duration and the integrated area under the curve (A). 499

Thapsigargin (10−7 M) treatment of the cells (which deplete all intracellular Ca2+ stores) completely 500

blocked the GnRH induced rise in [Ca+2]i (B). 501

502

Figure 6. Effects of vehicle, GnRH without and with thapsigargin or nifedipine on LH secretion from 503

ovine pituitary gonadotropes in vitro. Cells were treated with Vehicle, nifedipine or (10−5 M), 504

thapsigargin (10−7 M) for 15 min and stimulated with GnRH (10−9 M) for 2 h. Medium was collected 505

and assayed for LH secretion by RIA. Thapsigargin and nifedipine and treatment reduced GnRH-506

stimulated LH release suggesting a role of Ca2+ influx through voltage-gated Ca2+ channels during the 507

sustained phase of LH release. The data are present as mean (±SEM) (n=5 sheep). 508

509

Time (min)

LH

(n

g/m

l)

0

10

20

30

0 5 15 30 60

b

d d d

a

d

c a a ac

c cc

a

Vehicle

GnRH

E2+GnRH

a, b, c NSa vs. d, p≤ 0.001d vs. c, p ≤ 0.001

Figure 1

Figure 2

A

B

Time (min)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

0 5 15 30 60

Arb

itra

ry U

nit

s n

orm

aliz

ed

to 1

00%

of

con

tro

l

** ****

**

VehicleGnRH

E2

- - - - + - - - + - - - + - - - + - - -- - - - + + - + + + - + + + - + + + - +

- - - - + - + + + - + + + - + + + - + +

pERK-1/2

Total ERKβ-Actin

0min 5min 15min 30min 60min

VehicleGnRH

E2

E2+GnRH

**p<0.01

Figure 3

E2

V

LH-like-ir pERK-1/2-like-ir DAPI MergedA

A B C D

E F G H

pERK-1/2-like-ir Merged

** *

*

B

GH-like-ir

**

A C DDAPI

B

20μm

10μm

Gonadotropes

Somatotropes

A

B

Figure 4

GnRH 10-9M

30 s

9000

7000

5000

Flu

ore

scen

ce in

ten

sity

GnRH 10-9M

9000

7000

5000

Flu

ore

scen

ce in

ten

sity

Flu

ore

scen

ce:

% r

esp

on

se #

1

0

20

40

Vehicle E2

60

E-BSA

0 10-9 10-8 10-7 10-9 10-8 10-7

Figure 4

*

C

**

*

*

Thapsigargin 10-7M

Control

30 s

GnRH 10-9M

9000

7000

5000

Flu

ore

scen

ce in

ten

sity

Control

GnRH 10-9M

9000

7000

5000

30 sF

luo

resc

ence

in

ten

sity

Nifedipine 10-5M

Figure 5

A

B

0

50

100

150

200

250

V GnRH Thapsigargin+

GnRH

Nifedipine+

GnRH

LH

(n

g/m

l)

p<0.001

NS

p<0.05

Figure 6