Regulation of pituitary inhibin/activin subunits and follistatin gene expression by

gonadotropin-releasing hormone in female rats

Petra Popovicsa,b

, Zoltan Rekasia, Alan J. Stewart

b, Magdolna Kovacs

a

aDepartment of Anatomy, Medical School, University of Pecs, Pecs,7624, Hungary.

bSchool of Medicine, University of St Andrews, St Andrews, KY16 9TF, UK.

Short title: Pituitary inhibin-activin-follistatin expression

Author correspondence to:

Magdolna Kovacs MD., PhD., DSc.

University of Pecs, Medical School, Department of Anatomy

12 Szigeti Str., Pecs, 7624-Hungary

e-mail: magdolna.kovacs@hotmail.com

phone: +36-72-536-001

Key Words: GnRH, LHR, FSHR, ovariectomy, estradiol

Page 1 of 35 Accepted Preprint first posted on 13 April 2011 as Manuscript JOE-10-0485

Copyright © 2011 by the Society for Endocrinology.

Abstract

Pituitary inhibin B, activin B, and follistatin are local regulators of FSH. Activin B is a

homodimeric molecule (βB-βB), while inhibin B contains an α and a βB subunit. The

regulation of gene expression of α, βB, and follistatin by local and endocrine hormones were

examined in pituitaries from female rats and in perifused pituitary cells by RT-PCR.

Ovariectomy (OVX) induced an elevation in the mRNA level of α and βB subunits and

follistatin. Short-term (4h) treatment of pituitary cells with GnRH decreased both the inhibin

α and the inhibin/activin βB subunit mRNA levels, while long-term treatment (20h) with

100nM GnRH stimulated the expression of both subunits. In contrast, the mRNA level of

follistatin was elevated after the short-term GnRH treatment. Long-term exposure of pituitary

cells to estradiol and inhibin B suppressed the mRNA expression of βB and had no effect on

the expression of α subunit and follistatin. Our results demonstrate that the increased

expressions of inhibin/activin subunits and follistatin in the post-ovariectomy period can be

induced by the lack of gonadal negative feedback and the consequently developing high

GnRH environment of the pituitary. This study reports first that GnRH administered in high

dose and for a long period stimulates the gene expression of inhibin/activin subunits and

thereby may contribute to the stimulatory effect of OVX on the expression of these genes.

Page 2 of 35

Introduction

Activin, inhibin and follistatin are key regulators of pituitary FSH production. These factors,

together with ovarian steroids and GnRH, define the diverse pattern of LH and FSH

secretion. Based on their ability to modulate FSH secretion, inhibin, activin, and follistatin

were first identified as FSH-regulating gonadal hormones (Carroll et al. 1989; Weiss et al.

1993). Later, it was discovered that these proteins are also produced in a wide range of tissues

including the pituitary gland (Meunier et al. 1988; Shimasaki et al. 1989).

The structurally related proteins inhibin and activin were originally isolated from ovary and

characterized as members of the transforming growth factor β (TGFβ) superfamily (Ling et

al. 1985; Miyamoto et al, 1985; Robertson et al. 1985; Vale et al. 1986). Inhibins are dimeric

proteins composed of a unique α subunit and one of the two β subunits (A or B), forming

inhibin A (αβA) or inhibin B (αβB) (Mason et al. 1985). Activins are homodimer or

heterodimer molecules comprised of two β subunits (βA and βB) to produce three isoforms of

the mature activin, such as activinA (βAβA), activin B (βBβB), and activin AB (βAβB) (Vale et

al. 1986). Activins stimulate FSH synthesis by increasing promoter activity of the FSHβ gene

(Huang et al. 2001) and by stabilizing the FSH mRNA (Attardi & Winters 1993). Inhibins

counteract the activity of activins in two different ways. First, there is a competition for the β

subunits during ligand assembly, as both inhibin and activin dimers assemble from a common

pool of β subunits (Chapman & Woodruff 2003). Second, inhibins can also bind to the type II

activin receptor, without stimulating any of the intracellular pathways but blocking the effect

of activin by preventing receptor dimerization (Lebrun & Vale 1997). Surprisingly, the

continuing search for the unique inhibin receptor was unsuccessful; only inhibin co-receptors,

Page 3 of 35

such as inhibin binding protein and betaglycan, could be identified. However, these receptors

lack the signaling motif in their intracellular domain (Bernard et al. 2002).

The monomeric glycoprotein follistatin was isolated in the late 1980s (Esch et al. 1987;

Robertson et al. 1987). It has two forms generated by alternative splicing (Shimasaki et al.

1988a, 1988b). These isoforms of follistatin derive from transcription of five or six exons

(Inouye et al. 1991). Later a third, intermediate form, which is generated by proteolytic

processing of the C-terminus, was also found (Sugino et al. 1993). The action of follistatin is

attributed to its ability to bind and biologically inactivate the activins and, at least to some

extent, the inhibins also, through the common β subunit (Nakamura et al. 1990; Shimonaka et

al. 1991). Follistatin has been detected in various tissues, and its expression generally

coincides with that of the inhibin/activin subunits (Meunier et al. 1988; Shimasaki et al.

1989). The expression of inhibins, activins, and follistatin in the pituitary gonadotroph and

folliculostellate cells indicates that these proteins may act as paracrine/autocrine modulators

of FSH production (Bilezikjian et al. 2004). Corrigan et al. demonstrated the

paracrine/autocrine effect of activin B by incubation of cultured rat anterior pituitary cells

with a monoclonal antibody specific for activin B. The antibody attenuated the basal

secretion of FSH in a concentration- and time-dependent manner, without influencing LH

secretion (Corrigan et al. 1991). Unlike activin B, activin A and inhibin A seem to have less

of a role in FSH secretion whilst both the mRNA and protein levels of the βA subunit are

undetectable in pituitary (Meunier et al. 1988; Bilezikjian et al. 1993).

Although several studies focused on this issue, our knowledge on the regulation of the

pituitary inhibin-activin-follistatin system is incomplete. It is well established that fast

frequency GnRH pulses stimulate the expression of pituitary follistatin, and that the increased

Page 4 of 35

level of follistatin downregulates FSH secretion (Kirk et al. 1994; Besecke et al 1996).

However, the regulation of the inhibin/activin subunit expression is not so clear (Besecke et

al 1996; Tebar et al. 2000; Bilezikjian et al. 1996). Bilezikjian et al. demonstrated that GnRH

inhibits the expression of βB subunit, while in the study by Tebar et al. an antagonist of

GnRH was shown to suppress the mRNA level of βB (Tebar et al. 2000; Bilezikjian et al.

1996). It is known from earlier studies, that a regulatory mechanism exists between pituitary

inhibin, activin, and follistatin. Activin is able to elevate βB subunit and follistatin mRNA

levels and this effect is inhibited by follistatin and inhibin (Bilezikjian et al. 1996).

To reveal the control mechanisms of the pituitary inhibin-activin-follistatin system, we

investigated the effects of endogenous GnRH/gonadotropin overproduction on the mRNA

expression of pituitary inhibin/activin subunits and follistatin in ovariectomized (OVX) rats.

The changes in serum and pituitary LH and FSH levels after ovariectomy were also detected.

In studies in vitro, we examined the specific effects of GnRH, recombinant human LH and

FSH, estradiol, and inhibin B on the gene expression of inhibin/activin subunits and

follistatin of perifused pituitary cells. The gene expression of LH and FSH receptors was also

investigated.

Materials and methods

Drugs and chemicals

Recombinant human (rh) LH and FSH were obtained from Serono (Geneva, Switzerland), rh

inhibin B from Thermo Scientific, Pierce Biotechnology (Rockford, IL), and 17-β-estradiol

was purchased from Calbiochem (Gibbstown, NJ). GnRH was a gift by János Seprıdi

Page 5 of 35

(Semmelweis University, Budapest). The drugs were diluted in Medium 199 (Sigma-Aldrich,

St. Louis, MO).

Animals and surgery

Adult female 2-3 month old Wistar-R-Amsterdam rats of 250-300g body weights were used

for all experiments. The rats were housed under controlled conditions (12h light/12h dark

schedule at 24oC) with food and water ad libitum. Two weeks before using them for

experiments, the animals were tested for estrous cycle by taking daily vaginal smears, and the

rats showing 2 consecutive 4-day cycles were used. Ovariectomy or sham operation was

performed through bilateral lumbar incision under isoflurane anesthesia. One group of OVX

rats were killed by decapitation on day 1 (acute OVX) and two groups on day 28 (chronic

OVX) after the surgery. To abolish cycle-dependent variations in the target gene expression,

the sham-operated control rats were sacrificed in estrous stage. Pituitaries were removed,

homogenized, and stored in Lysis buffer (NucleoSpin RNA II kit, Macherey-Nagel, Düren,

Germany) at -70oC until RNA was extracted and RT-PCR was performed. Pituitaries and

blood samples from the second group of chronic OVX and of sham-operated control rats

were used for LH and FSH determination by RIA. Blood samples (0.3 ml) were obtained

from the jugular vein of these rats under isoflurane anesthesia before decapitation. LH and

FSH were extracted from the pituitary homogenates by 0.1M HCL and were determined by

RIA. LH and FSH concentrations of the pituitaries were expressed as ug/pituitary. All groups

consisted of 5-6 animals. This study was approved by the local ethical committee for animal

experiments (No: BA02/2000-20/2006)

Experiments in vitro

Page 6 of 35

The superfused rat pituitary cell system was carried out as described by Csernus & Schally,

1991. Mixed populations of pituitary cells obtained from normal female 2-3 month old rats

showing regular ovarian cycle were used for these experiments. The pituitaries were digested

in medium containing 0.5% collagenase (Type 2, Worthington Biochemical Corporation,

Lakewood, NJ), dispersed into cell groups, mixed with Sephadex G-10 (Sigma-Aldrich,

Gillingham, UK), and put into three chambers of the superfusion system. Each chamber

contained pituitary cells from 3 rats, providing about 2-3x106 cells per channel. Each

experiment produced 3 data (2 treated and 1 control), and the experiments were repeated

twice. To obtain steady-state cells the treatments were started after perfusing the cells with

medium for 2h. The cells in chamber 1 and 2 were treated simultaneously with the same drug

dissolved in medium for various times (40 min, 4h, 6h, 8h or 20 h), and control cells in

chamber 3 were perfused with medium. GnRH was used in concentration of 10 nM and 100

nM, LH and FSH were applied at 1 IU/ml, inhibin B at 50 ng/ml, and 17β-estradiol at 100

nM. After stopping the perfusion, subcellular fractions were extracted from the cells by RA1

reagent (Sigma, Gillingham, UK) containing 1% β-mercapthoethanol. The cell extract was

separated from the Sephadex gel by filtering through NucleoSpin Filter units (Macherey-

Nagel Inc., Düren, Germany), and RNA was isolated as described below.

RNA extraction and isolation

For total RNA extraction we used NucleoSpin RNA II kit (Macherey-Nagel, Düren,

Germany). Each pituitary was homogenized in 350 µl lysis solution containing 1% β-

mercaptoethanol (Sigma-Aldrich, Gillingham, UK) followed by RNA isolation according to

the manufacturer’s protocol. The yield and quality of RNA samples were determined

Page 7 of 35

spectrophotometrically at 260 nm and 260/280 nm and 260/230 nm ratios. The isolated RNA

was stored at -70°C until RT-PCR was

performed.

Semi-quantitative RT-PCR

Activin/inhibin subunits, follistatin, and GnRH receptor (GnRHR) mRNA levels were

assessed using semiquantitative RT-PCR assays. Total RNA (1µg) was reverse transcribed to

cDNA using the Moloney murine leukemia virus reverse transcriptase enzyme (MMLV-RT,

Promega, Madison, WI). RT reaction was performed in a final volume of 25 µl containing 4

µM exo-resistant random primer, 0.5 mM dNTP mix (Fermentas, Vilnius, Lithuania), 1.6

units/µl RNasin Rnase inhibitor, 1 x MMLV RT buffer, 6.4 units/µl MMLV RT (Promega).

One µl of the RT product was used for each PCR amplification with 2 primer pairs that

would amplify: (i) cDNA of β-actin standard gene or (ii) cDNAs of inhibin α, inhibin/activin

βA, βB or follistatin (for primers sequences see Table 1). Each reaction contained 0.05

units/µl GoTaq Flexi DNA polymerase, 1x GoTaq enzyme buffer, 1.5 mM MgCl (Promega,

Madison, WI), 0.2 mM dNTP and 0.3 µM of each primer. The PCR amplification was

conducted with the following cycle profile: 95 oC for 15 s, 60

oC for 30 s, 72

oC for 45 s.

Cycle numbers for the different primer pair combinations were determined to terminate the

PCR reaction when the amplification of both the β-actin and the target are in the logarithmic

phase. The original copy number of the β-actin mRNA was higher than that of the

inhibin/activin subunits or follistatin, and therefore the addition of the standard gene primer

pair was always delayed (for cycle numbers see Table 1). The PCR products were separated

on Sybr Safe prestained 2% agarose gels for 13 minutes (iBase™ Power System, Invitrogen,

Carlsbad, CA). All PCR products were detected at the expected molecular weights (Fig 1).

Results were quantified using The Kodak Electrophoresis Documentation and Analysis

Page 8 of 35

System (Eastman Kodak Company, Rochester, NY). All data were normalized to the β-actin

PCR product and shown in percentage of sham-operated/vehicle-treated controls.

Detection of LHR and FSHR mRNA

To detect the expression of LH and FSH receptors (LHCGR and FSHR) in pituitary cells we

used specific primers (LHCGR: 5` ggc gcc cat ctc ttt ctt tgc cat c 3` forward and 5` ggc tta ctt

gct cct ggg aag cc 3` reverse, 273bp, FSHR: 5` gcc gct cat cac tgt gtc caa gg 3` forward and

5` gct ctt tcg ggc atg gaa gtt gtg g 3` reverse primer, 214bp) for the amplification of mRNA in

untreated perifused pituitary cells and in rat ovary. β-actin primers are shown in Table 1. A

negative control (NEC) sample in which the RT enzyme was absent during reverse

transcription was prepared from normal pituitary RNA. All PCR products were detected at

the expected molecular weights (Fig 1).

Radioimmunoassay (RIA)

LH and FSH concentrations of the sera and pituitary fractions were determined by RIA using

materials obtained from the National Hormone and Pituitary Program (Rockville, MD; Rat

LH-11 and FSH-11 antibody, LH-RP-3 and FSH-RP-3 reference preparation, LH-I-9 and

FSH-I-9 hormone for iodination). The determinations were performed in duplicates. The

sensitivity of the RIA was 0.03ng for LH and 0.12 ng for FSH. The inter-assay and intra-

assay variations were 10% or less. LH and FSH concentrations of the pituitaries were

expressed as ug/pituitary.

Statistical analysis

Page 9 of 35

Statistical analysis of data was performed by t-test or one-way ANOVA followed by Tukey-

test using the computer software Sigma Stat (Jandel, San Rafel, CA). Differences were

considered significant when P< 0.05.

Results

Effects of ovariectomy on the inhibin α, inhibin/activin βA, βB, and follistatin mRNA levels

The initial PCR standardizing experiments revealed that the pituitary inhibin/activin βA

mRNA level is much lower than the α or βB levels, as we had to employ a high number of

PCR cycles (36) to detect the PCR product (Table 1). Considering that we had to use 36

cycles to detect βA but only 26 cycles for βB, the cDNA for βA needs 10 cycles more than

the cDNA for βB to reach the detectable level. Because the copies of DNA increase in

exponential manner in the cycles of PCR, they increase to 210

(=1024) in 10 cycles. Thus, the

mRNA level of βA in the pituitary might be about 1000-fold lower than the level of βB.

Furthermore, we could not detect any changes in the βA mRNA level after OVX (data not

shown). As this subunit is less significant in local activin/inhibin assembly and is not

regulated by OVX-induced hormonal changes, the expression of βA subunit was not

examined in our further experiments. The levels of inhibin α, inhibin/activin, βB, and

follistatin mRNA expression were investigated on day 1 and day 28 after OVX by semi-

quantitative RT-PCR (Fig. 2a). On day 1, no significant changes were found in the

expression level of the 3 genes examined. On day 28 after OVX, the mRNA of α subunit was

increased by 27% (P<0.05), the level of βB by 51% (P<0.001), and the expression level of

follistatin by 57% (P<0.05), compared to sham-operated controls (Fig. 2a).

Page 10 of 35

Effects of ovariectomy on the serum and pituitary LH and FSH levels

Serum and pituitary hormone levels were measured by RIA 28 days after OVX (Fig. 2b). We

found a 46-fold increase in serum LH and a 17-fold elevation in serum FSH levels compared

to sham-operated animals (p<0.001). The concentration of pituitary LH was increased by 16-

fold and the concentration of FSH by 3.5 fold (p<0.001) (Fig. 2b).

The effect of GnRH on α, βB, follistatin, and GnRH receptor mRNA levels

Perifused pituitary cells were treated with 100nM GnRH continuously for 40 minutes or 4, 6,

8, and 20 hours, and semi-quantitative PCR was performed to detect changes in α, βB and

follistatin mRNA expressions (Fig. 3a). After the treatment of cells with GnRH for 40 min,

no changes could be detected in the mRNA expression of α and βB subunits. When the cells

were perfused with GnRH for 4 h, a small decrease was observed in both subunits mRNA

levels; a 34% decrease in α (P<0.05) and a 20% decrease in βB (P<0.01). Upon further

prolongation of the treatment to 6 h, the inhibitory effect of GnRH gradually attenuated and

then disappeared at 8 h of the perfusion. However, when the continuous administration of

GnRH was extended to 20 h, a significant elevation could be observed in both subunits

mRNA level. The level of α subunit increased by 100% (P<0.001) and the level of βB by

44% (P<0.001). In contrast, the mRNA value for follistatin was found to be increased by

34% (P<0.001) already after a short treatment of the cells with GnRH for 40 min and peaked

at 4 h (65%, P<0.05). Then the mRNA level decreased gradually and returned to the control

level at 20 h (Fig. 3a). The effect of 10nM GnRH was also investigated and was compared to

the effect of 100 nM GnRH (Fig. 3b). Continuous perfusion of the cells with 10nM GnRH for

20h caused a small increase (34%, P<0.001) in α subunit mRNA level. The mRNA

expression of βB and follistatin were not altered by this treatment (Fig. 3b). Furthermore,

Page 11 of 35

sustained treatment of pituitary cells with 100 nM GnRH for 20 h had no significant effect on

the mRNA expression of the GnRHR.

The effect of LH and FSH on inhibin/activin subunits and follistatin gene expression

To investigate whether the gonadotropins LH and FSH have a role in the regulation of the

pituitary inhibin, activin, and follistatin gene expression and thereby can contribute to the

changes in mRNA expression of these genes after OVX, pituitary cells were treated with

1IU/ml LH or FSH for 4 h and 20 h (Fig. 4a). The 1 IU/ml dose of the recombinant FSH is

equal to 73.3 ng/ml, which is consistent with the ovariectomy-induced serum FSH level. The

dose of LH (45.3 ng/ml) is somewhat lower than the serum LH level measured in OVX rats

but still 13-fold of the normal control level. Treatment of pituitary cells with LH for 20 h

caused a 33% decrease (P<0.01) in the mRNA expression of βB subunit and no significant

alterations in the gene expression of α subunit and follistatin (Fig. 4ab). The mRNA

expressions of both LHR and FSHR could be detected with PCR in control pituitary cells of

the perifusion experiments (Fig. 4c). The PCR products obtained from control rat pituitary

and rat ovary, which was used as positive control, are shown in Fig. 4c. The β-actin primers

were used in separate PCR reactions. Equal quantity of β-actin mRNA was found in the cells

from rat pituitary and rat ovary. A negative control sample (NEC), in which the RT enzyme

was absent during reverse transcription, was prepared from normal pituitary RNA (Fig. 4c).

The effect of estradiol and inhibin B on inhibin/activin subunits and follistatin gene

expression

Pituitary cells were treated either with 100 nM estradiol or 50 ng/ml inhibin B for 20 h, and

mRNA expressions for α, βB, and follistatin were measured by semi-quantitative RT-PCR

(Fig. 5a, b). Estradiol suppressed the mRNA level of βB subunit by 34% (P<0.05) but did not

alter the mRNA expression of inhibin-α and follistatin (Fig. 5a). Treatment of the cells with

Page 12 of 35

inhibin B for 20 h also reduced the βB mRNA level to 37% (p<0.05) (Fig. 5b). The mRNA

expressions of α subunit and follistatin were not modified either by inhibin B or estradiol

(Fig. 4a, b).

Discussion

The pituitary inhibin-activin-follistatin regulatory loop has a major role in the control of FSH

production and defines the diverse pattern of FSH and LH secretion (Bilezikjian et al 2004).

Activin and GnRH interplay to stimulate the synthesis and secretion of FSH (Weiss et al

1993), whereas follistatin and inhibin counteract the action of activin, by reducing its

biological activity and blocking its binding to the receptor (Nakamura et al. 1990; Inouye et

al. 1991; Shimonaka et al. 1991). Consequently, the hormones and local factors which

regulate the synthesis of pituitary inhibin, activin, and follistatin indirectly influence FSH

production.

The aim of this study was to reveal the hormones and local factors which play a role in the

regulation of pituitary inhibin/activin/follistatin gene expression and can be accounted for the

increased expression of these genes in long-term OVX rats. Although the impact of

ovariectomy on the mRNA expression of inhibin/activin subunits and follistatin has been

published before, the mechanisms mediating the effect of ovariectomy are not identified in all

details. In agreement with previous studies (Dalkin et al. 1998; Prendergast et al. 2004), we

found that inhibin α, inhibin/activin βB, and follistatin mRNA levels were increased after

OVX. However, the expression of βA subunit was hardly detectable in pituitary and was not

changed by OVX in our experiments. Moreover, others have been unable to detect the RNA

of the βA subunit with S1 nuclease analysis in rat pituitary (Meunier et al. 1988). The low

Page 13 of 35

level of gene expression and the fact that no change in βA expression level could be detected

after OVX suggests that the A forms of inhibin/activin have very little or no involvement in

the local FSH-regulatory loop.

According to the results of our previous study (Kovacs et al. 2001), we showed a substantial

elevation in LH and FSH secretion in long-term OVX rats. In this earlier study, we also

provided evidence for the pivotal role of GnRH in the stimulation of gonadotropin secretion

after OVX by demonstrating that GnRH receptor antagonist Cetrorelix entirely prevented the

stimulatory effect of OVX on LH secretion (Kovacs et al. 2001). The high GnRH

environment of pituitary was also shown in long-term OVX and adrenalectomized rats

(Sherwood & Fink, 1980). Despite the indirect inhibitory effect of adrenalectomy on GnRH

production through CRF, of which production increases after adrenalectomy, the GnRH

concentration in the pituitary portal blood was found to be substantially and constantly

elevated in these OVX rats (Li et al. 2010).

Considering that the level of various hormones, such as gonadal steroids and gonadotropins

in the circulating blood and GnRH in the pituitary portal blood, change after OVX, any of

these hormones could play a role in the regulation of pituitary inhibin/activin subunits and

follistatin gene expression. To reveal the regulatory agents which mediate the effect of

ovariectomy to these genes, we investigated the direct effect of the hormones and local

factors that change after OVX on the inhibin α, inhibin/activin βB, and follistatin mRNA

level of pituitary cells in vitro. Because our in vitro studies were motivated by the results of

previous studies in OVX rats (Dalkin et al. 1998; Prendergast et al. 2004), we performed

OVX experiments and determined serum gonadotropin levels of the OVX rats to demonstrate

the effects of OVX in our experimental conditions. To mimic the high GnRH environment of

the pituitary developing after OVX we used high doses of GnRH in vitro. As 1 nM is known

Page 14 of 35

to be a physiological dose of GnRH to release LH from rat gonadotrophs in vitro (Kovacs &

Schally 2001; Kovacs et al. 2001), we used 10-fold and 100-fold of the physiological dose.

To investigate the time-dependent effect of the high GnRH environment of pituitary on the

gene expression of inhibin/activin and follistatin, we applied high-dose GnRH perfusions of

the cells for various times from 20 min to 20 h. It is well known, that GnRH stimulates the

gene expression of follistatin in pituitary (Kirk et al. 1994). In contrast, the effect of GnRH

on the expression of inhibin and activin is not clear. Bilezikjian et al. found that a short-term

(2h) exposure of pituitary cells to GnRH downregulated the mRNA expression of βB subunit

(Bilezikjian et al. 1996). Similarly to this finding, we also showed an acute suppression of α

and βB subunit mRNA level by GnRH in our experiments. However, on the basis of these

results, it is hard to elucidate how the increased level of activin and inhibin develops in the

post-OVX period. We demonstrate in this study that although a short-term GnRH treatment

inhibits the gene expression of both α and βB subunits, the prolonged presence of GnRH has

a stimulatory effect on their production. These results are the first to show that GnRH has

diverse short- and long-term effects on the gene expression of inhibin/activin subunits and

thereby clarify the development of increased gene expression of activin and inhibin observed

in long-term OVX animals. A gradual upregulation of the GnRH receptor expression in

pituitary after OVX may enhance the stimulatory effect of GnRH on gonadotrophs (Kovacs

et al. 2001).

The gene expression of follistatin was stimulated by GnRH, except at 20 h, but it was not

altered by either estrogen or inhibin in our experiments. This observation supports the earlier

finding that the main regulator of follistatin is GnRH (Kirk 1994). The gradual attenuation of

follistatin gene expression during the long exposure of cells to GnRH can be explained by the

temporal changes in GnRH signaling. Earlier studies have shown that long term continuous

administration of GnRH was ineffective for stimulating gonadotropin β subunits whereas the

Page 15 of 35

α subunit level was elevated after a GnRH exposure of the pituitary cells for 24 hours

(Shupnik & Fallest 1994), indicating the existence of a constant and a pulse dependent

component in GnRH signaling (Vasilyev et al. 2002). Accordingly, the expression of

follistatin might rely on GnRH pulse frequency, whereas the expression of inhibin/activin

subunits can be stimulated by constant GnRH signaling. It has been reported that the

cytoplasmic carboxyl tail of the GnRH receptor, which is responsible for the agonist

dependent acute desensitization and internalization, is absent in the mammalian receptor

(Willars et al 1994). Moreover, we detected no change in the receptor mRNA expression after

a sustained treatment of pituitary cells with GnRH. However, despite the preserved

expression of the receptor, the receptor binding and signaling capacity can change during the

long-term exposure of the gonadotrophs to GnRH. Receptor uncoupling and temporal

impairment of the signaling pathways were reported to cause desensitization of the LH

releasing capacity of the gonadotrophs to GnRH stimulation (Chang et al 1988). In contrast,

the synthesis of gonadotropin α subunit increases after a sustained GnRH treatment (Shupnik

& Fallest 1994). This indicates that although the receptor-coupled intracellular pathways are

partly down-regulated by the constant activation, a component of the pathway remains

functional. Based on these findings, a temporal impairment of the signaling pathways can be

accounted for the time-dependent changes in mRNA expressions of follistatin in response to

sustained administration of GnRH for 10-20h in our experiments.

The secretion of gonadotropins is increased in long-term OVX animals (Ramirez & Sawyer

1974), but the local role of LH and FSH in the pituitary is not well established. It is still not

known if primary gonadotrophs contain functional LH or FSH receptors. Gonadotroph cell

line αT3 was shown to express LH receptors, and the activation of the receptors was able to

elevate gonadotropin α subunit mRNA level (Huang et al. 1995). Moreover, Lei et al. found

low mRNA expression of the LH receptor (Lei et al. 1993), and we also showed the presence

Page 16 of 35

of receptor transcript in pituitary. These findings together with the current observation that

long-term treatment of pituitary cells with LH suppressed the expression of βB suggest that

LH might be able to act in a paracrine/autocrine manner in pituitary. Similarly to pituitary,

Liu et al. demonstrated in human granulosa lutein cells that high levels of gonadotropins

inhibited the expression of the βB subunit (Liu et al. 2001). In contrast to LH, there is no

evidence for the presence of functional FSH receptors in the pituitary gonadotrophs.

Although we could detect the presence of the FSH transcript, the treatment of pituitary cells

with FSH had no effect on the expression of the 3 genes tested. The lack of FSH action to

influence the mRNA expression of inhibin/activin and/or follistatin in our experiments

indicates that only LH but not FSH may play a role in the autocrine/paracrine regulation of

these genes. It is possible that the lack of FSH action is due to the lack of functional receptors

for FSH on the gonadotrophs. Further experiments are needed to determine the protein

expression of FSH receptor and investigate possible responses of the cells to receptor

activation.

It was shown in an earlier study that activin is able to elevate βB mRNA expression, and

inhibin is a potent inhibitor of the auto-stimulatory effect of activin (Bilezikjian et al. 1996).

The existence of this auto-regulatory mechanism by activin/inhibin is supported by our

finding that inhibin B decreases the mRNA expression of βB. The treatment of pituitary cells

with estradiol also decreased the mRNA level of βB expression in our experiments. This

finding may help to explain why the replacement of estradiol in gonadectomized rats

prevented the increase in inhibin/activin βB expression (Dalkin et al. 1998). On the other

hand, because GnRH concentration is highly increased in the portal blood after OVX

(Sherwood & Fink 1980), and sustained GnRH administration at high dose stimulates the

expression of βB, the negative feedback effect of estradiol to hypothalamic GnRH might

Page 17 of 35

contribute to the direct negative effect of estradiol to prevent the increase of βB expression

after OVX. In physiological conditions, however, the concentration and time-dependent

reactivity of gonadotrophs is different from that found after OVX, and physiological

concentrations of estradiol can enhance the sensitivity of gonadotrophs to GnRH (Colin &

Jameson 1998). As a result, the stimulation of inhibin/activin expression might occur at lower

GnRH concentrations or physiological GnRH pulsation. Further investigation is required to

determine whether there is cooperation between GnRH and estradiol in the regulation of

inhibin/activin subunits expression in gonadotrophs in vivo. Our findings, in agreement with

earlier results, show that ovarian hormones, such as inhibin and estradiol, negatively regulate

the gene expression of pituitary inhibin and activin in the absence of GnRH (Bilezikjian et al.

1996; Dalkin et al. 1998). Based on these findings and the results from our in vitro

experiments, the stimulatory action of GnRH together with the lack of ovarian estradiol and

inhibin could cause the increase of βB gene expression in long-term OVX rats. As the

expression of α subunit was also stimulated by the long-term administration of GnRH but it

was not suppressed by estradiol or inhibin, the role of GnRH in the stimulation of the inhibin-

specific α subunit seems to be more significant than in the stimulation of βB.

In summary, our study demonstrates that the expression of pituitary activin/inhibin subunits

is regulated by local and peripheral hormones. Estradiol, inhibin B, LH, and short-term

GnRH treatment decrease the mRNA expression of βB subunit, while long-term GnRH

treatment stimulates βB expression. GnRH has similar effects on the inhibin α and

inhibin/activin βB subunits, whereas estradiol, inhibin B, and LH have no significant effect

on the gene expression of α subunit. Our results provide the first evidence that GnRH

administered in high dose and for a long period directly stimulates the gene expression of

pituitary inhibin/activin and thereby may participate in the stimulatory effect of ovariectomy

Page 18 of 35

on the expression of these genes.

Page 19 of 35

Declaration of interest

The authors declare no conflict of interest.

Funding

This work was supported by the Hungarian Research Grant of University of Pécs, Medical

School (AOK-KA-34039-8-2009) to M.K.

Acknowledgement

We are grateful to Izabella Orbán and Enikı Nagy for technical assistance.

Page 20 of 35

References

1. Attardi B & Winters SJ 1993 Decay of follicle-stimulating hormone-beta messenger

RNA in the presence of transcriptional inhibitors and/or inhibin, activin, or follistatin.

Molecular Endocrinology 7 668-680.

2. Bernard DJ, Chapman SC & Woodruff TK 2002 Minireview: Inhibin binding protein

(InhBP/p120), betaglycan, and the continuing search for the inhibin receptor.

Molecular Endocrinology 16 207-212.

3. Besecke LM, Guendner MJ, Schneyer AL, BauerDantoin AC, Jameson JL & Weiss J

1996 Gonadotropin-releasing hormone regulates follicle-stimulating hormone-beta

gene expression through an activin/follistatin autocrine or paracrine loop.

Endocrinology 137 3667-3673.

4. Bilezikjian LM, Blount AL, Leal AMO, Donaldson CJ, Fischer WH & Vale WW

2004 Autocrine/paracrine regulation of pituitary function by activin, inhibin and

follistatin. Molecular and Cellular Endocrinology 225 29-36.

5. Bilezikjian LM, Corrigan AZ, Blount AL & Vale WW 1996 Pituitary follistatin and

inhibin subunit messenger ribonucleic acid levels are differentially regulated by local

and hormonal factors. Endocrinology 137 4277-84.

6. Bilezikjian LM, Vaughan JM & Vale WW 1993 Characterization and the regulation

of inhibin/activin subunit proteins of cultured rat anterior pituitary cells.

Endocrinology 133 2545-2553.

7. Carroll RS, Corrigan AZ, Gharib SD, Vale W & Chin WW 1989 Inhibin, activin, and

follistatin: regulation of follicle-stimulating-hormone messenger ribonucleic acid

levels. Molecular Endocrinology 3 1969-1976.

Page 21 of 35

8. Chang JP, Graeter JS & Katt KJ 1988 Desensitization of pituitary gonadotropes by

mediators of LH release. Biochemical and Biophysical Research Communications 153

919-924.

9. Chapman SC & Woodruff TK 2003 Betaglycan localization in the female rat

pituitary: implications for the regulation of follicle-stimulating hormone by inhibin.

Endocrinology 144 5640-5649.

10. Colin IM & Jameson JL 1998 Estradiol sensitization of rat pituitary cells to

gonadotropin-releasing hormone: involvement of protein kinase C- and calcium

dependent signaling pathways. Endocrinology 139 3796-3802.

11. Corrigan AZ, Bilezikjian LM, Carroll RS, Bald LN, Schmelzer CH, Fendly BM,

Mason AJ, Chin WW, Schwall RH, Vale W 1991 Evidence for an autocrine role of

activin B within rat anterior pituitary cultures. Endocrinology 128 1682-1684.

12. Csernus V & Schally AV 1991 The dispersed cell superfusion system. In

Neuroendocrine Research Methods, pp 71-109. Ed B Greenstein. London: Harwood

Academic.

13. Dalkin AC, Haisenleder DJ, Gilrain JT, Aylor K, Yasin M & Marshall JC 1998

Regulation of pituitary follistatin and inhibin/activin subunit messenger ribonucleic

acids (mRNAs) in male and female rats: evidence for inhibin regulation of follistatin

mRNA in females. Endocrinology 139 2818-2823.

14. Esch FS, Shimasaki S, Mercado M, Cooksey K, Ling N, Ying S, Ueno N & Guillemin

R 1987 Structural characterization of follistatin: a novel follicle-stimulating hormone

release-inhibiting polypeptide from the gonad. Molecular Endocrinology 1 849-855.

15. Huang HJ, Sebastian J, Strahl BD, Wu JC & Miller WL 2001 Transcriptional

regulation of the ovine follicle-stimulating hormone-beta gene by activin and

Page 22 of 35

gonadotropin-releasing hormone (GnRH): involvement of two proximal activator

protein-1 sites for GnRH stimulation. Endocrinology 142 2267-2274.

16. Huang ZH, Lei ZM & Rao CV 1995 Immortalized anterior pituitary alpha t3

gonadotropes contain functional luteinizing hormone/human chorionic-gonadotropin

receptors. Molecular and Cellular Endocrinology 114 217-222.

17. Inouye S, Guo Y, DePaolo L, Shimonaka M, Ling N & Shimasaki S 1991

Recombinant expression of human follistatin with 315 and 288 amino acids: chemical

and biological comparison with native porcine follistatin. Endocrinology 129 815-22.

18. Kirk SE, Dalkin AC, Yasin M, Haisenleder DJ & Marshall JC 1994 Gonadotropin-

releasing hormone pulse frequency regulates expression of pituitary follistatin

messenger ribonucleic acid: a mechanism for differential gonadotrope function.

Endocrinology 135 876-880.

19. Kovacs M & Schally AV 2001 Comparison of mechanisms of action of luteinizing

hormone-releasing hormone (LHRH) antagonist cetrorelix and LHRH agonist

triptorelin on the gene expression of pituitary LHRH receptors in rats. Proceedings of

the National Academy of Sciences USA 98 12197-12202.

20. Kovacs M, Schally AV, Csernus B & Rekasi Z 2001 Luteinizing hormone-releasing

hormone (LH-RH) antagonist Cetrorelix down-regulates the mRNA expression of

pituitary receptors for LH-RH by counteracting the stimulatory effect of endogenous

LH-RH. Proceedings of the National Academy of Sciences USA 98 1829-1834.

21. Lebrun JJ & Vale WW 1997 Activin and inhibin have antagonistic effects on ligand-

dependent heteromerization of the type I and type II activin receptors and human

erythroid differentiation. Molecular and Cellular Biology 17 1682-1691.

Page 23 of 35

22. Lei ZM, Rao CV, Kornyei JL, Licht P & Hiatt ES 1993 Novel expression of human

chorionic gonadotropin/luteinizing hormone receptor gene in brain. Endocrinology

132 2262-70.

23. Li XF, Knox AM & O'Byrne KT 2010 Corticotrophin-releasing factor and stress-

induced inhibition of the gonadotrophin-releasing hormone pulse generator in the

female. Brain Research 1364 153-63.

24. Ling N, Ying SY, Ueno N, Esch F, Denoroy L & Guillemin R 1985 Isolation and

partial characterization of a Mr 32,000 protein with inhibin activity from porcine

follicular fluid. Proceedings of the National Academy of Sciences USA 82 7217-7221.

25. Liu J, Hyden-Granskog C & Voutilainen R 2001 Gonadotrophins inhibit and activin

induces expression of inhibin/activin betaB subunit mRNA in cultured human

granulosa-luteal cells. Molecular Human Reproduction 7 319-323.

26. Mason AJ, Hayflick JS, Ling N, Esch F, Ueno N, Ying SY, Guillemin R, Niall H &

Seeburg PH 1985 Complementary DNA sequences of ovarian follicular fluid inhibin

show precursor structure and homology with transforming growth factor-beta. Nature

318 659-663.

27. Meunier H, Rivier C, Evans RM & Vale W 1988 Gonadal and extragonadal

expression of inhibin alpha, beta A, and beta B subunits in various tissues predicts

diverse functions. Proceedings of the National Academy of Sciences USA 85 247-251.

28. Miyamoto K, Hasegawa Y, Fukuda M, Nomura M, Igarashi M, Kangawa K &

Matsuo H 1985 Isolation of porcine follicular fluid inhibin of 32k daltons.

Biochemical and Biophysical Research Communications 129 396-403.

29. Nakamura T, Takio K, Eto Y, Shibai H, Titani K & Sugino H 1990 Activin-binding

protein from rat ovary is follistatin. Science 247 836-838.

Page 24 of 35

30. Prendergast KA, Burger LL, Aylor KW, Haisenleder DJ, Dalkin AC & Marshall JC

2004 Pituitary follistatin gene expression in female rats: Evidence that inhibin

regulates transcription. Biology of Reproduction 70 364-370.

31. Ramirez VD & Sawyer CH 1974 Differential dynamic responses of plasma LH and

FSH to ovariectomy and to a single injection of estrogen in rat. Endocrinology 94

987-993.

32. Robertson DM, Klein R, Devos FL, Mclachlan RI, Wettenhall REH, Hearn MTW,

Burger HG & Dekretser DM 1987 The isolation of polypeptides with FSH

suppressing activity from bovine follicular fluid which are structurally different to

inhibin. Biochemical and Biophysical Research Communications 149 744-749.

33. Robertson DM, Foulds LM, Leversha L, Morgan FJ, Hearn MTW, Burger HG,

Wettenhall REH & Dekretser DM 1985 Isolation of inhibin from bovine follicular

fluid. Biochemical and Biophysical Research Communications 126 220-226.

34. Sherwood NM & Fink G 1980 Effect of ovariectomy and adrenalectomy on

luteinizing hormone-releasing hormone in pituitary stalk blood from female rats.

Endocrinology 106 363-367.

35. Shimasaki S, Koga M, Buscaglia ML, Simmons DM, Bicsak TA & Ling N 1989

Follistatin Gene expression in the ovary and extragonadal tissues. Molecular

Endocrinology 3 651-659.

36. Shimasaki S, Koga M, Esch F, Cooksey K, Mercado M, Koba A, Ueno N, Ying SY,

Ling N & Guillemin R 1988 Primary structure of the human follistatin precursor and

its genomic organization. Proceedings of the National Academy of Sciences USA 85

4218-4222.

37. Shimasaki S, Koga M, Esch F, Mercado M, Cooksey K, Koba A & Ling N 1988

Porcine follistatin gene structure supports two forms of mature follistatin produced by

Page 25 of 35

alternative splicing. Biochemical and Biophysical Research Communications 152

717-23.

38. Shimonaka M, Inouye S, Shimasaki S & Ling N 1991 Follistatin binds to both activin

and inhibin through the common beta-subunit. Endocrinology 128 3313-3315.

39. Shupnik MA & Fallest PC 1994 Pulsatile GnRH regulation of gonadotropin subunit

gene transcription. Neuroscience and Biobehavioral Reviews 18 597–599.

40. Sugino K, Kurosawa N, Nakamura T, Takio K, Shimasaki S, Ling N, Titani K &

Sugino H 1993 Molecular heterogeneity of follistatin, an activin-binding protein.

Higher affinity of the carboxyl-terminal truncated forms for heparan sulfate

proteoglycans on the ovarian granulosa cell. Journal of Biological Chemistry 268

15579-87.

41. Tebar M, de Jong FH & Sanchez-Criado JE 2000 Regulation of inhibin/activin

subunits and follistatin mRNA expression in the rat pituitary at early estrus. Life

Sciences 67 2549-2562.

42. Vale W, Rivier J, Vaughan J, Mcclintock R, Corrigan A, Woo W, Karr D & Spiess J

1986 Purification and characterization of an FSH releasing protein from porcine

ovarian follicular fluid. Nature 321 776-779.

43. Weiss J, Crowley WF, Halvorson LM & Jameson JL 1993 Perifusion of rat pituitary

cells with gonadotropin-releasing hormone, activin, and inhibin reveals distinct

effects on gonadotropin gene expression and secretion. Endocrinology 132 2307-

2311.

44. Vasilyev VV, Lawson MA, Dipaolo D, Webster NJG & Mellon PL 2002 Different

signaling pathways control acute induction versus long-term repression of LHß

transcription by GnRH. Endocrinology 143 3414-3426.

Page 26 of 35

45. Willars GB, Heding A, Vrecl M, Sellar R, Blomenrohr M, Nahorski SR & Eidne KA

1999 Lack of a C-terminal tail in the mammalian gonadotropin-releasing hormone

receptor confers resistance to agonist-dependent phosphorylation and rapid

desensitization. Journal of Biological Chemistry 274 30146-30153.

Page 27 of 35

Figure legends

Figure 1

Validation of the primers used in our experiments. A 100bp-DNA molecular weight marker is

shown ranged from 200bp to 700bp.The bands for inhibin α, inhibin/activin βB, follistatin,

GnRHR, β-actin, LHCGR, and FSHR were detected at the expected molecular weights.

Figure 2

a/ The effect of OVX on the mRNA levels of inhibin α, inhibin/activin βB, or follistatin on

day 1 and 28 after OVX. The horizontal line symbolizes the mRNA level in sham-operated

animals. b/ The effect of OVX on the serum and pituitary LH and FSH concentrations. Error

bars represent mean ± SEM of 5-6 animals in all groups. Asterisks indicate significant

difference between controls and operated animals. (***P<0.001; **P<0.01; *P<0.05).

Figure 3

a/ The effect of continuous treatment of pituitary cells with 100nM GnRH for various times

(40 min, 4 h, 6 h, 8 h, and 20 h) on the mRNA levels of inhibin α, inhibin/activin βB, and

follistatin. (n=4 for all time points) b/ mRNA levels of pituitary inhibin α, inhibin/activin βB,

and follistatin after a continuous perfusion of the cells with 10 nM (n=4) or 100 nM (n=4)

GnRH. c/ Representative gel pictures showing the expression of inhibin α, inhibin/activin βB

subunits, and follistatin relative to β-actin after a continuous perfusion of the cells with 100

nM GnRH for 20 h. Error bars represent mean ±SEM. The horizontal line symbolizes the

mRNA level in untreated perifused cells. Asterisks indicate significant difference between

vehicle- and hormone-treated cells. (***P <0.001; **P<0.01; *P<0.05).

Figure 4

a/ Effects of long-term treatment of pituitary cells with LH and FSH for 4h and 20h on the

mRNA level of pituitary inhibin α, inhibin/activin βB, and follistatin. Error bars represent

mean ±SEM. The horizontal line symbolizes the mRNA level in untreated perifused cells.

Page 28 of 35

Asterisks indicate significant difference between vehicle- and hormone-treated cells.

**P<0.01; n=4 for all data. b/ Representative gel picture of the mRNA expression of βB

subunit after a perfusion of the cells with 1IU LH for 20 h. c/ The PCR products obtained

from control rat pituitary and rat ovary (positive control) are shown. A negative control

sample (NEC), in which the RT enzyme was absent during reverse transcription was prepared

from normal pituitary RNA.

Figure 5

The effect of continuous treatment of pituitary cells with 100 nM estradiol (E2) for 20 h (a)

or 50 ng/ml inhibin for 20 h (b) on the mRNA level of inhibin α, inhibin/activin βB, and

follistatin, including representative gel pictures of the relative expression of inhibin βB

subunit after the treatments. Error bars represent mean ±SEM; n=4 for both estradiol and

inhibin. The horizontal line represents the mRNA level in untreated perifused cells. Asterisks

indicate significant difference between vehicle- and hormone-treated cells. (***P<0.001).

Page 29 of 35

Figure 1

Page 30 of 35

Figure 2

0

20

40

60

80

100

120

140

160

180

α βB FS

mR

NA

level

(% c

on

tro

l)

OVX 1 day OVX 28 days

***

*

*

a

0

20

40

60

80

100

120

140

160

180

200

220

LH FSH

seru

mco

ncen

trati

on

(ng

/ml)

control OVX 28 days

***

***

b

0

5

10

15

20

25

LH FSH

pit

uit

ary

co

ncen

trati

on

(µg

/pit

uit

ary

)

***

***

Page 31 of 35

Figure 3

50

75

100

125

150

175

200

225

0 5 10 15 20

mR

NA

level

(% c

on

tro

l)

Time (hour)

α βB follistatin

***

*

*

***

***

***

0

50

100

150

200

250

α βB follistatin

mR

NA

level

(% c

on

tro

l)

GnRH 1nM 20h

GnRH 100nM 20h

β-actin

β-actin

β-actinα

βB

follistatin

GnRH Control

Control

a b

c

100 nM GnRH

***

***

***

GnRH

GnRH

GnRH

GnRH ControlGnRH

GnRH ControlGnRH

β-actin

GnRHR

Page 32 of 35

Figure 4

0

20

40

60

80

100

120

140

160

α βB follistatin α βB follistatin

mR

NA

lev

el (%

co

ntr

ol)

1IU/ml 4h

1IU/ml 20h

LH FSH

β-actin

βB

LH Control

**

LH

LHCGR

FSHR

β-actin

pit. cells ovary NEC

a b

c

Page 33 of 35

Figure 5

0

20

40

60

80

100

120

α-subunit βB-subunit follistatinm

RN

A l

ev

el

(%c

on

tro

l)

0

20

40

60

80

100

120

α βB follistatin

mR

NA

lev

el

(% c

on

tro

l)

β-actin

βB -subunit

Inhib. ControlE2 Control

β-actin

βB -subunit

a b

***

E2 Inhib.

*

Page 34 of 35

Table 1 Primer sequences and PCR cycles used in semi-quantitative PCR

F: forward primer, R: reverse primer

mRNA Primers

PCR

product

(bp)

PCR

cycles

(target

gene)

PCR

cycles

(ref.

gene)

inhibin α

F 5´ TTC ATT TTC CAC TAC TGC CAT GGT

AGC 3`

R 5` GAT ACA AGC ACA GTG TTG TGT AAT

GA 3`

240 30 21

inhibin/

activin

β A

F 5’ AGA GGA CGA CAT TGG CAG GAG 3’

R 5’ AGA GGA CGA CAT TGG CAG GAG 3’

165 36 22

inhibin/

activin

β B

F 5’ AGG CAA CAG TTC TTC ATC GAC TTT

CGC 3’

R 5’AGC CAC ACT CCT CCA CAA TCA TGT T

3’

304 26 20

follistatin

F 5’ CCT ACT GTG TGA CCT GTA ATC 3’

R 5’ CTC CTC TTC CTC CGT TTC TTC 3’

422 30 22

β-actin

(ref. gene)

F 5` GTC ACC CAC ACT GTG CCC ATC T 3`

R 5´ ACA GAG TAC TTG CGC TCA GGA G 3´

542 - -

GnRHR

F 5` CCG CAA TGG TGG CAT GAA GCC TTC 3`

R 5` TAG AGT TCT CAG CCG TGC TCT TGG 3`

192 26 20

Page 35 of 35