Biochimica et Biophysica Acta 1833 (2013) 511–519

Contents lists available at SciVerse ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r .com/ locate /bbamcr

Cardiac glycoside ouabain induces activation of ATF-1 and StAR expression byinteracting with the α4 isoform of the sodium pump in Sertoli cells

Raimund Dietze a, Lutz Konrad b, Mazen Shihan a, Ulrike Kirch a, Georgios Scheiner-Bobis a,⁎a Institut für Veterinär-Physiologie und -Biochemie, Fachbereich Veterinärmedizin, Justus-Liebig-Universität Giessen, Germanyb Zentrum f. Frauenheilkunde und Geburtshilfe, Fachbereich Medizin, Justus-Liebig-Universität Giessen, Germany

⁎ Corresponding author at: Institut für Veterinär-PhysioloVeterinärmedizin, Justus-Liebig-Universität Giessen, FrankfGermany. Tel.: +49 641 9938172; fax: +49 641 9938179.

E-mail address: Georgios.Scheiner-Bobis@vetmed.un(G. Scheiner-Bobis).

0167-4889/$ – see front matter © 2012 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.bbamcr.2012.11.022

a b s t r a c t

a r t i c l e i n f o

Article history:Received 27 June 2012Received in revised form 20 November 2012Accepted 26 November 2012Available online 6 December 2012

Keywords:OuabainSertoli cellATF-1StARAlpha4 isoformSodium pump

Sertoli cells express α1 and α4 isoforms of the catalytic subunit of Na+,K+-ATPase (sodium pump). Our recentfindings demonstrated that interactions of the α4 isoform with cardiotonic steroids (CTS) like ouabain inducesignaling cascades that resemble the so-called non-classical testosterone pathway characterized by activationof the c-Src/c-Raf/Erk1/2/CREB signaling cascade. Here we investigate a possible physiological significance ofthe activated cascade. The results obtained in the current investigation show that the ouabain-induced signalingcascade also leads to the activation of the CREB-related activating transcription factor 1 (ATF-1) in the Sertoli cellline 93RS2 in a concentration- and time-dependent manner, as demonstrated by detection of ATF-1 phosphory-lated on Ser63 in western blots. The ouabain-activated ATF-1 protein was found to localize to the cell nuclei. Thesodium pumpα4 isoformmediates this activation, as it is ablated when cells are incubatedwith siRNA to theα4isoform. Ouabain also leads to increased expression of steroidogenic acute regulator (StAR) protein, which hasbeen shown to be a downstream consequence of CREB/ATF-1 activation. Taking into consideration that CTSare most likely produced endogenously, the demonstrated induction of StAR expression by ouabain establishesa link between CTS, the α4 isoform of the sodium pump, and steroidogenesis crucial for male fertility andreproduction.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

The Na+,K+-ATPase (sodium pump) is a plasma membrane-boundenzyme that catalyzes the transport of 3 Na+ ions out of the cell and2K+ into the cell in a reaction that utilizes ATP. The enzyme consistsof α and β subunits. These two subunits, which are absolutely requiredfor ion transport, exist in various isoforms: a total of four α subunitisoforms (α1–α4) and three β subunit isoforms (β1–β3) have beenidentified thus far. In some cells an additional γ subunit from the familyof the FXYD proteins is involved in sodium pump regulation by inter-acting with α and β subunits.

While the α1 isoform of the sodium pump is ubiquitously ex-pressed in all types of animal cells, some of the other α isoforms arefound together with the α1 isoform only in certain cell types. In par-ticular, the α4 isoform seems to have the most restricted pattern ofexpression. This isoform has been exclusively detected in testes andwas demonstrated specifically in spermatocytes and mature sperm[1,2]. Its function has been associated with sperm motility, fertility,and capacitation [3–6].

gie und -Biochemie, Fachbereichurter Str. 100, D-35392 Giessen,

i-giessen.de

rights reserved.

These data notwithstanding, spermatocytes do not seem to be theexclusive location for α4 expression. We recently identified the α4isoform also in Sertoli cells, which are neithermotile nor involved in ca-pacitation [7]. In these cells the interaction of the cardiac glycoside(CTS) ouabain with this isoform triggers the non-classical testosteronesignaling pathway characterized by the activation of the c-Src/c-Raf/Erk1/2/CREB signaling cascade [8,9].

Activation of various signaling cascades by CTS/sodium pump inter-actions has been frequently demonstrated and is meanwhile widely ac-cepted [10,11]. There is very little known, however, about whetheractivation of signaling cascades by CTS acting on the α4 isoform resultsin significant physiological events. The investigation presented here ad-dresses this possibility. By using the Sertoli cell line 93RS2 we investi-gate whether the ouabain-triggered activation of the non-classicaltestosterone pathway has any physiological consequences for the cells.

The results presented here show for the first time a CTS-induced ac-tivation of the activating transcriptional factor 1 (ATF-1) within the nu-cleus, which is mediated through the α4 isoform of the sodium pump.The CTS-induced signaling cascade stimulates the expression of thesteroidogenic acute regulator (StAR) protein, a crucial regulator of ste-roidogenesis [12]. Thus, taking into consideration that CTS are mostlikely produced endogenously [13–15], their interactions with the α4isoform might be of physiological significance for steroidogenesis,male fertility, and reproduction. This possibility could be addressed infuture investigations.

512 R. Dietze et al. / Biochimica et Biophysica Acta 1833 (2013) 511–519

2. Materials and methods

2.1. Cell culture

The Sertoli cell line 93RS2 [16] was cultured in a humidified incuba-tor at 37 °C in 5% CO2. ThemediumusedwasDMEM/Ham's F-12 (1/1 byvol.) with L-glutamine supplemented with 10% (by vol.) fetal bovineserum (FBS) and 1% penicillin/streptomycin. It was replaced every twodays. To harvest cells, the medium was removed by aspiration. Cellswere then washed with Dulbecco's phosphate-buffered saline (PBS)without Ca2+ and Mg2+ and subsequently incubated with accutase(0.25%) for 4 min at 37 °C. Cell culture reagents were from PAA Labora-tories (Pasching, Austria).

2.2. Preparation of cell lysates

Cell lysateswere prepared as described [7]. Briefly, 1×105 cells of theSertoli cell line 93RS2 were grown in 10-cm culture dishes as describedabove. Cells were then serum-starved (i.e. 0.5% serum) for 24 h prior tothe addition of ouabain. Incubationwith the CTSwas allowed to proceedfor various times. The medium was then aspirated, and the cells werewashed with 5 ml PBS and lysed in 600 μl of cell lysis buffer (CellSignaling Technology, Frankfurt, Germany) containing 1 μM PMSF, 1×protease inhibitor cocktail (Roche, Mannheim, Germany), and 2 μg/mlpepstatin. All lysis steps were carried out on ice. After 5 min cells weredetached from the dish with a cell scraper. The mixture was then trans-ferred into reaction vials and sonicated 20 times for 1 s with intervals of1 s. Lysates were centrifuged at 4 °C and 13,000 ×g for 20 min. Proteinin the supernatant was determined as described [7]. Aliquots of the su-pernatant were stored at −20 °C.

2.3. SDS-PAGE and immunodetection of isolated proteins

A total of 10 μg protein from cell lysates was separated by SDS-PAGE on gels containing 10% or 12% acrylamide and 0.3% N,N′-methylene-bis-acrylamide. Biotinylated proteins (Cell Signaling Tech-nology, Frankfurt, Germany) served as molecular weight markers.Proteins were then electro-blotted onto nitrocellulose membranes(Schleicher & Shull, Dassel, Germany) at 500 mA for 30 min. Detectionof various proteins was carried out by following the recommendationsof the providers of the primary antibodies (Cell Signaling Technology)and the enhanced chemiluminescence kit (ECL; GE HealthCare,Munich,Germany). After exposure to a film, protein bands were quantified by adigital documentation system (Biostep, Jahnsdorf, Germany) and thePhoretix TotalLab gel image analysis software (Biostep).

In order to re-probe the western blot with a different antibody, ni-trocellulose sheets were incubated in 50 ml 100 mM glycine, pH 2.1 3times for 15 min and then washed twice in 50 ml TBS-T for 10 mineach. Incubation with the new primary and secondary antibodies wascarried out as described above.

2.4. Use of siRNA to silence either α1 or α4 mRNA expression

Silencing expression of either α1 or α4 mRNA was performedby using commercially available siRNA (Stealth™ RNAi; Invitrogen,Karlsruhe, Germany) as described previously [7] and by following thetransfection protocol of the provider. The oligonucleotides used were5′-UGU UUC UUG GCU UAU GGC AUC CGA A-3′ and 5′-UUC GGA UGCCAU AAG CCA AGA AAC A-3′ against the α1 isoform mRNA, and5′-GCC UCA UCA GGA GUC UCU UCC CAU A-3′ together with 5′-UAUGGG AAG AGA CUC CUG AUG AGG C-3′ against the α4 isoform mRNA.

Control cells were either treated with Lipofectamine 2000 andStealth™ RNAi Negative Control LO GC (negative control for either α1or α4 Na+,K+-ATPase Stealth™ RNAi). Transfection efficiency wasestimated by the Block-iT™ Transfection Kit (Invitrogen, Karlsruhe,Germany) according the protocol of the provider. After 72 h of

incubation cells were treated with 100 nM ouabain for 30 min; lysateswere isolated and the proteins of the cell lysates were analyzed bywestern blotting as described above.

2.5. Detection of active ATF-1 within the nucleus

A total of 1×104 cells was seeded in each well of a 12-well plateand cultured to 60–70% confluency as described above. Each of thefollowing media changes was followed by washing the cells twicewith 500 μl PBS. Unless otherwise specified the following procedureswere carried out at room temperature. All incubation steps withUV-sensitive reagents were performed in the dark.

Ouabain (10 nM)was added to each well and incubation proceededfor various periods of time. After aspiration of the ouabain-containingmedium, cells were fixed in 3.7% formaldehyde for 15 min. The formal-dehyde solutionwas then removed and thefixed andwashed cellswereincubated in blocking solution (3% BSA/0.6% Triton X-100 in TBS) for1 h. Thereafter, the incubationmediumwas replaced by TBS containing1% BSA, 0.6% Triton X-100, and the primary antibody against phosphor-ylated ATF-1 (Epitomics/Biomol, Hamburg, Germany) at a dilution of1:150 by vol. Incubation continued for 48 h at 4 °C.

For fluorescence staining, the previous incubation medium wasaspirated and replaced with TBS containing 1% BSA, 0.6% TritonX-100, and the secondary antibody (goat anti-rabbit IgG) labeledwith AlexaFluor 488 at a dilution of 1:250 by vol. Incubation contin-ued for 1 h at room temperature. The solution was then removedand the cells washed as stated above. To stain the nuclei, cells wereincubated for 10 min in methanol containing 0.5 μg/ml DAPI. For con-trols, the first antibody was omitted. The samples were overlayedwith 500 μl PBS and stored at 4 °C for further investigation. Imageswere obtained by an inverse Olympus IX81 microscope equipped withthe correspondingfluorescence system (Olympus, Hamburg, Germany).Fluorescence of single nuclei was measured by using the softwareprogram ImageJ (freely available at http://rsbweb.nih.gov/ij/). All cellsin the optical field were considered. Data points were transferred toand analyzed by the software program GraphPad Prism4 (GraphPadSoftware, Inc., La Jolla, CA, USA).

2.6. Detection of phospho-ATF-1 after silencing α1 or α4 isoforms

Cells were treated with siRNA to silence either the α1 or the α4subunit isoform as described above. Control cells were treated withLipofectamine 2000 and Stealth™ RNAi Negative Control LO GC.After 3 days of incubation, 10 nM ouabain was added to each sampleand incubation was continued for 30 min. Controls were incubatedfor the same time without ouabain. Cells were then fixed and usedto detect phospho-ATF-1 as described in the previous paragraph.

2.7. RT-PCR for the detection of StAR-specific mRNA/cDNA

93RS2 cells were grown as described above. Ouabain (10 nM) wasthen added and incubation proceeded for 24 h. Total mRNA was iso-lated by following the protocol of the commercially available RNeasyMini kit (Qiagen, Hilden, Germany). Reverse transcription and PCRamplification were carried out by the OneStep RT-PCR kit (Qiagen).A total of 60 ng of mRNA, 1 μl (20 pmol/ml) of each primer, 10 μl offive-fold concentrated buffer (20 mmol/l Tris HCl, 100 mmol/l KCl,12.5 mmol/l MgCl2, 10 mmol/l dNTPs), 2 μl of a mixure of Ominiscriptand Sensiscripts reverse transcriptases and HotStar Taq DNA polymer-ase in a final volume of 25 μl were incubated in aMasterCycler Gradient(Eppendorf, Hamburg, Germany) at 50 °C for 30 min to accomplishreverse transcription. For PCR amplification the mixture was heated at95 °C for 15 min, followed by 40 cycles of denaturation at 94 °C for1 min, annealing at 58 °C for 1 min, and cDNA extension at 72 °C for1 min. After amplification, a final extension at 72 °C was performedfor 10 min. For specific amplification of StAR forward and reverse

513R. Dietze et al. / Biochimica et Biophysica Acta 1833 (2013) 511–519

primers were 5′TTCAAGCTGTGTGCTGGGAGCTCC3′ and 5′GTGGCCATGCCTGCCAGCAC3′. They amplify a 593-bp fragment of the StAR cDNA.

2.8. Statistical analysis

Optical density of protein bands of the various Western blots wascorrected for any gel loading differences by taking into considerationthe optical density of actin that was detected in Western blots run inparallel.

Data were analyzed by Prism4 and by applying one-way ANOVAwith repeated measures and Dunnett's comparison of all data to thecontrol. Significance was accepted at pb0.05.

3. Results

3.1. Ouabain-induced activation of ATF-1

Testosterone triggers a reaction cascade involving c-Src/Ras/Raf/Erk1/2/CREB activation, which is referred to as the non-classical pathway oftestosterone action [9]. Earlier investigations demonstrated that low con-centrations of ouabain induce activation of the same cascade in Sertoli

ouabain (nM)

ouabain (nM)

0 1 10 100

MW

40 kDa

30 kDa p-ATF-1

actin 40 kDa

50 kDa

0 1 10 1000

1

2

AT

F-1

act

ivat

ion

(%)

**30 min

*

A)

B)

C)

D)

0 1 10 1000

1

2

AT

F-1

act

ivat

ion

(%

) ** 15 min * *

Fig. 1. ATF-1 activation by ouabain. 93RS2 cells were treated for either 15 or 30 minwith the indicated concentrations of ouabain. Cell lysates were then prepared and10 μg of protein was separated on SDS polyacrylamide gels and subsequently probed ina western blot using a monoclonal antibody against either actin, as a loading control (A),or phosphorylated (activated) ATF-1 (p-ATF-1) (B). Thewestern blots show typical resultsobtained after 30 min of incubationwith ouabain. Activation of ATF-1was significant after15 (C) or 30 min (D) at all ouabain concentrations tested (n=8; mean±SEM; *p≤0.05;**p≤0.01).

cells [7]. Since CREB often forms heterodimers with other transcriptionfactors like CREM or ATF-1, we investigated whether low concentrationsof ouabain would also lead to the stimulation of the transcription factorATF-1.

At low nanomolar concentrations ouabain stimulated the activation(phosphorylation at Ser63) of ATF-1 (Fig. 1). Incubation of the cells foreither 15 or 30 min with as little as 1 nM ouabain led to a prominentsignal in the western blot that was significantly different from the con-trol obtained in the absence of the CTS. Activation of ATF-1 showed anouabain concentration-dependent effect. Ouabain concentrationshigher than 100 nMwere not tested in order to prevent possible inter-actions of the steroid with the α1 isoform of the sodium pump, whichdisplays a lower affinity for the steroid (Ki=1.5±1.3×10−4 M) thanthe α4 isoform (Ki=1.6±1×10−9 M) [17]. For the same reason, allother measurements were carried out at either 10 or 100 nM ouabain.

The stimulation of ATF-1 phosphorylation by 10 nM ouabain wastime dependent (Fig. 2). Maximum phosphorylation of ATF-1 was ob-served at 10 min of incubation and declined thereafter, although itstill remained significantly higher than the basal phosphorylation ofthe control (at 0 min).

3.2. Ouabain-induced activation of ATF1 is mediated through the α4subunit isoform of the sodium pump

The low concentrations of ouabain needed to induce activation ofATF-1 are consistent with the involvement of the α4 isoform of the so-diumpump in thepropagation of the signaling events, since this isoformin rodents binds ouabain with a much higher affinity than the α1 iso-form [17]. Thus, we designed experiments to test which of these twoisoforms is specifically involved in ATF-1 activation. Fig. 3 shows thatthe transfection of 93RS2 Sertoli cells with the control siRNA did notaffect the ability of the cells to respond to ouabain. Incubation of thecells with 100 nMouabain for 10 min resulted in a significant activationof ATF-1 over the untreated control. The same observation was madewith cells that had been treated with siRNA against the α1 prior totheir incubation with ouabain, which indicates that the α1 isoform is

MW

0 3 10 30min

p-ATF-1

40 kDa

30 kDa

0 3 10 300.0

0.5

1.0

1.5

AT

F-1

act

ivat

ion

(%)

min

* *

**

40 kDa

50 kDa actin

A)

B)

C)

Fig. 2. Time-dependent activation of ATF-1 by ouabain. 93RS2 cells were incubated forthe indicated periods with 10 nM ouabain. Proteins in cell lysates (10 μg) were sepa-rated by SDS gel electrophoresis and blotted onto nitrocellulose, which was subse-quently probed in a western blot using a monoclonal antibody against either actin(A) or p-ATF-1 (B). The western blots show typical results obtained after incubationwith 10 nM ouabain. C) Activation of ATF-1 was significant at all times tested, with amaximum obtained after 10 min of ouabain treatment (n=12; mean±SEM; *p≤0.05;**p≤0.01).

0

1

2

3

AT

F-1

act

ivat

ion

(%

)

30 kDa

40 kDa

p-ATF-1

40 kDa

50 kDa 60 kDa

actin

ouabain 100 nM + + +

siRNA

ouabain 100 nM

siRNA

c α1 α4

α1 α4

α1 α4

* *

A)

B)

C)

MW

MW

ouabain 100 nM + + +

siRNA c

+ + +

c

Fig. 3. Ouabain effect on ATF-1 activation after selectively silencing α1 orα4 mRNA ex-pression. 93RS2 Sertoli cells were treated with siRNA directed against the α1 or α4isoforms of the sodium pump as described in “Materials and methods”. Control cellswere treated with negative control siRNA. After 72 h of incubation cells were treatedwith 100 nM ouabain for 30 min and were used to isolate lysates as described above.Cell lysate proteins (10 μg) were then separated by SDS-PAGE and probed in westernblots with antibodies recognizing either actin (A) or p-ATF-1 (B). Treatment of cellswith either control siRNA (c) or siRNA against the α1 isoform (α1) did not lead to aloss of ouabain-induced ATF-1 activation. When cells were treated with siRNA silenc-ing the α4 isoform expression (α4), ouabain did not activate ATF-1. C) Statistical anal-ysis of results (n=3; mean±SEM; *p≤0.05).

514 R. Dietze et al. / Biochimica et Biophysica Acta 1833 (2013) 511–519

not required for the ouabain-induced signaling pathway. When cellswere treated with siRNA to silence the α4 mRNA expression, however,they did not respond to 100 nM ouabain and displayed a phospho-ATF-1 level similar to that of the untreated control. Thus, it is apparentlythe α4 isoform that mediates the ouabain-induced signaling in Sertolicells.

3.3. Detection of activated ATF-1 within the nucleus

ATF-1 is a member of the ATF/CREB family, which belongs to theleucine zipper superfamily of proteins. Phosphorylation of these pro-teins is important for the binding of transcriptional co-activators andfor the binding to their DNA promoter regions to drive transcription[18]. In order to verify whether the ouabain-induced ATF-1 activationseen in the various western blots occurs in the nucleus, cells werestimulated for various times with 10 nM ouabain and then labeledwith a primary rabbit antibody against phospho-ATF-1 and a second-ary goat anti-rabbit IgG labeled with AlexaFluor 488. Nuclei werestained with DAPI. Fig. 4 shows that phospho-ATF-1-specific fluores-cence in the nuclei rapidly increased after the addition of 10 nMouabain. After 10 or 30 min of incubation, all nuclei were more orless fluorescent in comparison with only 27.6±5% fluorescent nucleiunder control conditions (without ouabain).

Ouabain-induced phospho-ATF-1 formation does not occur aftercells are treated with siRNA to prevent α4 isoform expression. Therewas no specific phospho-ATF-1 fluorescence detectable in the nucleiunder these conditions (compare Fig. 5E and F). Treatment of the cells

with siRNA against the α1 isoform did not hinder the ouabain-induced activation of ATF-1 (Fig. 5C and D), which was demonstratedby the fact that the AlexaFluor 488 fluorescence was present withinthe nuclei to the same extent as in the nuclei of cells treated with con-trol siRNA (Fig. 5A and B). These results are in good agreement withthe results shown in Fig. 4 and further confirm the α4 isoform as themediator of the ouabain-induced activation of ATF-1.

3.4. Ouabain stimulates expression of StAR-specific mRNA and StARprotein

In steroidogenic cells StAR expression is controlled by CREB/ATF-1transcription factors [19–21]. The activation of these transcriptionfactors, however, does not occur exclusively through cAMP-mediatedsignaling, as onemight expect by the participation of CREB, but throughvarious signaling cascades, including cAMP-independent signalinglike that propagated through the activation of Erk1/2 [22–24]. SinceStAR expression has been detected in Sertoli cells in several studies[19,25,26], we investigated in initial experiments whether the observedouabain activation of ATF-1 would result in increased expression ofStAR-specific mRNA in the Sertoli cell line 93RS2. The RT-PCR resultsconfirmed the existence of a StAR-specific, rather faint amplificate inthe absence of ouabain (Fig. 6A) which clearly becomes stronger after24 h of incubation with 10 nm ouabain.

The increased biosynthesis of StAR-specificmRNA is associatedwithincreased biosynthesis of StAR protein expression. Fig. 6C and D showsthe effect of 10 nMouabain on StAR protein expression. After 24 h of in-cubation with the cardiac glycoside, StAR expression was stimulated4-fold over the expression seen in the control, whichwas run in parallelin the absence of ouabain (Fig. 6C, D). Significant stimulation of StARprotein expressionwas seen after 5 h of incubationwith 10 nMouabain(Fig. 6D). Fig. 6B is the gel loading control: it shows thewestern blot de-tection of actin in the same nitrocellulose membrane shown in Fig. 6Cafter the stripping of the antibodies used for the detection of the StARprotein.

3.5. Stimulation of CREB and ATF-1 by ouabain

CREB and ATF-1 belong to the same family of transcription factorsand can act as homo- or heterodimers in order to influence gene expres-sion [18]. In a recent publicationwe demonstrated activation of CREB inSertoli cells by ouabain [7]. Since Erk1/2 and CREB/ATF-1 activationhave been shown to promote the interaction of p-CREB/p-ATF-1 withthe StAR gene promoter and the transcription of the StAR gene [22,23],we were interested to investigate whether both transcription factorswould still be active 24 h after the addition of ouabain. This possibilitywas examined by western blotting of lysates from 93RS2 cells treatedfor 24 h with 10 nM ouabain. Ouabain was omitted in control cells.The western blots were probed with an antibody that recognizes bothphospho-CREB and phospho-ATF-1. As shown in Fig. 7, 24 h after theaddition of ouabain phosphorylation of both transcription factors is sig-nificantly higher than the basal phosphorylation of the control. Theseresults are consistent with the results showing stimulation of StAR ex-pression by ouabain (Fig. 6) and support the notion that StAR expres-sion is driven by CREB/ATF-1 heterodimers [23].

4. Discussion

Expression of the α4 subunit isoform of the sodium pump wasoriginally detected in testis [2,3,27] and in particular in the middlepiece of the flagellum [5] or the head of spermatozoa [28]. Its functionwas thought to be associated with motility of sperm cells [3,5] or withtheir capacitation [28]. Our recent finding that the α4 subunit isoformis also present in rat and human Sertoli cells and our demonstrationthat ouabain via the α4 mediates the signaling cascade Src/c-Raf/Erk1/2/CREB [7], which is equivalent to the non-classical testosterone

A) phospho ATF-1 B) DAPI C) Merge

control control

10 min 10 min

control

10 min40 µm

control 3 min 10 min 30 min0

10

20

Ave

rag

e F

ITC

flu

ore

scen

ce(a

rbit

ary

un

its)

*

**

**

D)

Fig. 4. Time-dependent activation of ATF-1 within the nucleus of 93RS2 Sertoli cells. A) Activated ATF-1 was identified at different times after treatment with 10 nM ouabain byusing the primary antibody against p-ATF-1 and a secondary antibody labeled with AlexaFluor 488. B) The number and position of nuclei were determined by simultaneousDAPI staining. C) Merged A and B images. D) The relative fluorescence of every single nucleus from various photomicrographs was measured by using the software program ImageJ(http://rsbweb.nih.gov/ij/). At least 62 nuclei were considered for the analysis of data (n=62–91; mean±SEM; *p≤0.05; **p≤0.01). The values for 3 and 30 min were taken fromphotomicrographs not shown in the figure.

515R. Dietze et al. / Biochimica et Biophysica Acta 1833 (2013) 511–519

signaling cascade [9], indicate a possible involvement of the α4 sub-unit isoform in additional physiological processes beyond sperm mo-tility or capacitation.

Ouabain and other CTS-induced signaling cascades have been de-scribed in a great number of publications in the past decade [29]. Nev-ertheless, aside from the induction of apoptosis in tumor cell lines andpossibly also in tumors [30–33], there is not much known about addi-tional physiologically significant effects of CTS-induced signaling. Itwas therefore the aim of the study presented here to investigatewheth-er the ouabain-triggered non-classical testosterone signaling pathwaymight have any physiological consequences.

Transcription factors that bind to cAMP-responsive element (CRE)promoters of various genes induce the transcription of a great varietyof genes. In Sertoli cells CREB/CRE-inducible transcription is essentialfor the survival of spermatocytes and the production ofmature sperma-tozoa [34]. Activation of CREB, however, does not necessarily require anincreased concentration of cAMP and subsequent stimulation of proteinkinase A; CREB activation can also occur via different signaling cascades[18,22,24]. In Sertoli cells CREB activation is triggered by testosteronevia the activation of the c-Src/c-Raf/Erk1/2 signaling cascade, referredto as the non-classical testosterone signaling pathway [35–37]. Theprocess of spermatogenesis and the maturation of spermatogonia to

spermatozoa depend on the activation of Erk1/2 and other mitogen-activated protein kinases (MAPK) [38,39]. In addition, Erk1/2 activationis an absolute requirement for the production of haploid spermatozoa[40,41].

Erk1/2 activation and activation of CREB by phosphorylation atSer133 is mediated in Sertoli cells also by the interaction of the CTSouabain with the α4 isoform of the sodium pump [7]. In the currentinvestigation we demonstrate that ouabain additionally stimulatesthe CREB-related transcriptional factor ATF-1 in a concentration- andtime-dependent manner (Figs. 1 and 2). This finding is in good agree-ment with the results of others showing that ATF-1 activation isdownstream of Erk1/2 activation [42–45]. The activation of ATF-1(phosphorylation at Ser63) takes place within the nucleus (Fig. 4)and is mediated through the α4 subunit isoform (Fig. 3) at concentra-tions of ouabain within the range of the binding affinity of this isoformfor ouabain [17].

Both CREB and ATF-1 are members of the bZIP superfamily of tran-scription factors and can stimulate transcription by binding as eitherhomo- or heterodimers to the CRE region of various promoters [46].Phosphorylation of either CREB or ATF-1 transcription factors iscrucial for the activation of transcription [46], although possibly fordifferent reasons: phosphorylation of CREB at Ser133 by various

control siRNA; no ouabain control siRNA; + 10 nM ouabain

A) B)

D)

F)E)

C)

40 µm

α1 siRNA; no ouabain α1 siRNA; +10 nM ouabain

α4 siRNA; no ouabain α4 siRNA; +10 nM ouabain

Fig. 5. Ouabain effect on ATF-1 activation within the nucleus after selectively silencing α1 or α4 mRNA expression. 93RS2 Sertoli cells were treated with either control siRNA orsiRNA directed against the α1 or α4 isoforms of the sodium pump as described in “Materials and methods”. Cells were then incubated for 30 min in the absence or presence of10 nM ouabain and were fixed and probed for phospho-ATF-1 as described in “Materials and methods”. In the absence of ouabain cells treated with control siRNA (A), siRNA againstthe α1 isoform (C) or siRNA against the α4 isoform (E) shows a modest, basal level of activation of ATF-1, indicated by a few of the nuclei containing a green fluorescent signal. Inthe presence of 10 nM ouabain cells treated with control siRNA (B) or siRNA against the α1 isoform (D) respond with a clear activation (phosphorylation) of ATF-1 within theirnuclei. This activation does not occur when cells have been treated with siRNA against the α4 isoform of the sodium pump (F).

516 R. Dietze et al. / Biochimica et Biophysica Acta 1833 (2013) 511–519

kinases is not required for the binding of the factor to CRE but ratherfor the interaction of CREB with the transcriptional co-activator CBP(CREB binding protein) [47,48], a process that leads to transcriptionalactivation [49]; phosphorylation of ATF-1 at Ser63 is important forthe binding of that factor to the CRE of DNA promoters and forthe stimulation of transcription [50–52]. Phosphorylation of ATF-1at Ser63, however, is protein kinase A independent [53], supportingthe notion that the observed ATF-1 activation is the product of thestimulation of the non-classical, cAMP-independent testosterone sig-naling pathway c-Src/c-Raf/Erk1/2/CREB [7–9]; this signaling path-way can now be revised to include the activation of ATF-1.

Based on the literature cited above, the ouabain- andα4-mediatedphosphorylation of CREB at Ser133 [7] and of ATF-1 at Ser63, whichoccur within the nucleus, most likely result in the formation oftranscription-stimulating heterodimers. Our results shown in Fig. 7support this notion. The question we asked was, does this activationof the two transcription factors translate into the expression of phys-iologically relevant proteins in the Sertoli cell line?

Erk1/2 and CREB/ATF-1 activation have been shown to promotethe interaction of p-CREB/p-ATF-1 with the StAR gene promoter andthe transcription of the StAR gene [22,23]. Consistent with thesefindings, we demonstrate here that ouabain-induced stimulation of

control 3 h 5 h 15 h 24 h0

1

2

3

4

time of incubation

StA

R e

xpre

ssio

n (r

elat

ive)

*

**

**

ouabain (10 nM) incubation (h) 5 5 15 15 3 5 15

MW

30 kDa

20 kDaStAR

C)

B)

ouabain (10 nM) incubation(h) 5 15 15 3 15

MW

actin50 kDa

40 kDa

D)

600 593 bp800

bp

incubation (h)

A)

— — — — + + +

— — — — + + +

5 5

240

Fig. 6. Ouabain-induced expression of StAR-specific mRNA and protein. A) Cells wereincubated with 10 nM ouabain for the indicated periods of time. mRNA was then iso-lated and a StAR-specific fragment of 593 bp was amplified by RT-PCR. Although thefragment is rather faint, it is clearly present to a greater extent in cells treated withouabain for 24 h than in cells treated for 0 or 5 h. B) The blot shown in (C) wasstripped and then probed with an antibody against actin in order to obtain a gel load-ing control. C) Proteins in cell lysates (18 μg) were separated by SDS gel electrophore-sis, blotted onto nitrocellulose, and probed with an antibody against StAR. Due to thenature of the experiment it was not possible to include all samples taken at the varioustimes in a single western blot. The one shown is therefore just a representative westernblot showing the expression of StAR protein in some control and some ouabain-treatedcells. It is apparent that after 5 and 15 h the expression of StAR is higher than in thecontrols (some in duplicate). D) Similar western blots that could not be included inthe figure for reasons of space were used for the statistical analysis shown (n=3–4;mean±SEM; *p≤0.05; **p≤0.01).

40 kDa

30 kDa

50 kDa

MW

p-ATF-1

p-CREB

A)

40 kDa

30 kDa

50 kDa

actin

10 nM ouabain

MWB)

0

100

200

300C

RE

B/A

TF

-1ac

tivat

ion

(% c

ontr

ol)

10 nM ouabain

CREB ATF-1

**

C)

10 nM ouabain

- + - +

- +

- +

Fig. 7. Detection of phospho-CREB and phospho-ATF-1 after 24 h of incubation withouabain. Cell lysates were isolated after 24 h of treatment with 10 nM ouabain. Ouabainwas omitted in the controls. Proteins in cell lysates (10 μg) were separated by SDS gelelectrophoresis, blotted onto nitrocellulose, and probed with antibodies. A) By using anantibody that recognizes both phospho-CREB and phospho-ATF, it can be seen that phos-phorylation levels of both transcription factors are high after 24 h of incubation with oua-bain (2–2.5-fold higher than the phosphorylation levels of the 0 h-control). B) The blotshown in (A) was stripped and then probed with an antibody against actin in order toobtain a gel loading control. It is apparent that changes in p-ATF-1 and p-CREB levelsseen in (A) are not due to changes in protein expression. C) Statistical analysis of results(n=3; mean±SEM; *p≤0.05).

517R. Dietze et al. / Biochimica et Biophysica Acta 1833 (2013) 511–519

CREB and ATF-1 results in the expression of significant quantities ofStAR-specific mRNA and StAR protein (Fig. 6).

StAR is a key component of a protein complex that regulates ste-roidogenesis [12]. It stimulates themetabolism of cholesterol to steroidhormones by enhancing its transfer from the outermitochondrialmem-brane to cytochrome P45011A1 (P450scc) in the inner membrane [12].StAR is expressed in cells of the gonads and adrenals and its biosynthe-sis has been shown to depend on GnRH [54–56] and on CREB activation[19,22,23]. There is an on-going debate about whether StAR is ex-pressed in Sertoli cells and, if so, whether its role is associated with

the production of pregnenolone. Thus, immunohistochemical investiga-tions reveal that the steroidogenic protein P450ssc (cytochome P450side chain cleavage; CYP11A1) is lacking in Sertoli cells of goettingenminiature pigs [57], suggesting the inability of these cells to convertcholesterol into pregnenolone. Other publications, however, contradictthis assumption. Thus, immunohistochemical investigations of humantestis identify the StAR protein not only in Leydig but also in Sertolicells [58]. FSH treatment of rat Sertoli cells stimulates the expressionof StAR, identified by immunofluorescence [25]. Similarly, the stimula-tion of immature rat Sertoli cells in culture by dbcAMP leads to expres-sion of StAR as well P450scc (CYP11A1) that can easily be detected inWestern blots [19]. The stimulation of the expression of these two pro-teins is associated with increased pregnenolone production [19]. In avery recent paper, StAR and CYP11A1 expression on the mRNA levelwas identified in rat primary cultures of Sertoli and germ cells aftertheir exposure to atrazine [26].

518 R. Dietze et al. / Biochimica et Biophysica Acta 1833 (2013) 511–519

The Sertoli cell line 93RS2 used here might not be absolutely iden-tical to regular Sertoli cells; however, they do retain their ability toexpress StAR under certain conditions. The fact that the expressionof this protein is triggered by ouabain might reveal a new facet inthe process of steroidogenesis.

Ouabain and other CTS are viewed by many investigators as beingendogenously produced hormones [10,29]. The highest immunoreac-tivity for ouabain has been identified in adrenals, hypophysis, and hy-pothalamus [59–61]. Since adrenalectomy leads to the reduction ofcirculating ouabain levels, the adrenals seem to be themajor productionsite for CTS [60,62,63]. Taking into consideration these findings and ourrecent results, we suggest that the interaction of endogenously pro-duced ouabain with the α4 isoform of Sertoli cells induces signalingevents that result in StAR expression and could possibly influence thesteroidogenic activity of these cells as well the development of sper-matogenic cells. Thus, the focus of our forthcoming studies will centeron testing this hypothesis by investigating the effects of ouabain onuptake and processing of cholesterol in mitochondria of Sertoli cellsand the influence of the CTS on the induction of meiotic markers inco-cultures of Sertoli and spermatogenic cells. The fact that digoxin-like immunoreactivity has been detected in human seminal fluid [64]supports the concept of CTS acting as hormones on cells of the repro-ductive system and encourages further investigations in this direction.

Acknowledgements

The work was supported by the DFG, grant number Sche 307/6-1.

References

[1] O.I. Shamraj, J.B. Lingrel, A putative fourth Na+, K+-ATPase α-subunit is expressedin testis, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) 12952–12956.

[2] G. Blanco, G. Sanchez, R.J. Melton, W.G. Tourtellotte, R.W. Mercer, The alpha4 iso-form of the Na, K-ATPase is expressed in the germ cells of the testes, J. Histochem.Cytochem. 48 (2000) 1023–1032.

[3] A.L. Woo, P.F. James, J.B. Lingrel, Sperm motility is dependent on a unique isoformof the Na, K-ATPase, J. Biol. Chem. 275 (2000) 20693–20699.

[4] A.L. Woo, P.F. James, J.B. Lingrel, Roles of the Na, K-ATPase alpha4 isoform and theNa+/H+ exchanger in sperm motility, Mol. Reprod. Dev. 62 (2002) 348–356.

[5] G. Sanchez, A.N. Nguyen, B. Timmerberg, J.S. Tash, G. Blanco, The Na, K-ATPasealpha4 isoform from humans has distinct enzymatic properties and is importantfor sperm motility, Mol. Hum. Reprod. 12 (2006) 565–576.

[6] T. Jimenez, J.P. McDermott, G. Sanchez, G. Blanco, Na, K-ATPase alpha4 isoform isessential for sperm fertility, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 644–649.

[7] L. Konrad, R. Dietze, U. Kirch, H. Kirch, A. Eva, G. Scheiner-Bobis, Cardiotonic steroidstrigger non-classical testosterone signaling in Sertoli cells via the alpha4 isoform ofthe sodium pump, Biochim. Biophys. Acta 1813 (2011) 2118–2124.

[8] C. Fix, C. Jordan, P. Cano,W.H.Walker, Testosterone activatesmitogen-activated pro-tein kinase and the cAMP response element binding protein transcription factor inSertoli cells, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 10919–10924.

[9] W.H. Walker, Molecular mechanisms of testosterone action in spermatogenesis,Steroids 74 (2009) 602–607.

[10] W. Schoner, G. Scheiner-Bobis, Endogenous cardiac glycosides: hormones usingthe sodium pump as signal transducer, Semin. Nephrol. 25 (2005) 343–351.

[11] Z. Xie,Membrane transporters and signal transduction, CellMol. Biol. (Noisy-le-grand)52 (2006) 1–2.

[12] D.M. Stocco, The role of the StAR protein in steroidogenesis: challenges for thefuture, J. Endocrinol. 164 (2000) 247–253.

[13] J.M. Hamlyn, Discovery of endogenous ouabain— a newmammalian hormone, in:E. Bamberg, W. Schoner (Eds.), The Sodium Pump, Springer, New York, 1994,pp. 722–731.

[14] R. Schneider, R. Antolovic, H. Kost, B. Sich, U. Kirch, M. Tepel, W. Zidek, W. Schoner,Proscillaridin A immunoreactivity: its purification, transport in blood by a specificbinding protein and its correlation with blood pressure, Clin. Exp. Hypertens. 20(1998) 593–599.

[15] A.Y. Bagrov, E.G. Lakatta, O.V. Fedorova, Marinobufagenin, an endogenousbufadienolide sodiumpump inhibitor, in: K. Taniguchui, S. Kaya (Eds.), Na/K-ATPaseand Related ATPAses, Esevier Science, Amsterdam, 2000, pp. 647–654.

[16] C. Jiang, S.J. Hall, K. Boekelheide, Development and characterization of a prepu-bertal rat Sertoli cell line, 93RS2, J. Androl. 18 (1997) 393–399.

[17] K. Wagoner, G. Sanchez, A.N. Nguyen, G.C. Enders, G. Blanco, Different expressionand activity of the alpha1 and alpha4 isoforms of the Na, K-ATPase during ratmale germ cell ontogeny, Reproduction 130 (2005) 627–641.

[18] G. Servillo, M.A. Della Fazia, P. Sassone-Corsi, Coupling cAMP signaling to tran-scription in the liver: pivotal role of CREB and CREM, Exp. Cell Res. 275 (2002)143–154.

[19] S.L. Ford, A.J. Reinhart, Y. Lukyanenko, J.C. Hutson, D.M. Stocco, Pregnenolone syn-thesis in immature rat Sertoli cells, Mol. Cell. Endocrinol. 157 (1999) 87–94.

[20] P.R. Manna, M.T. Dyson, D.M. Stocco, Role of basic leucine zipper proteins in tran-scriptional regulation of the steroidogenic acute regulatory protein gene, Mol.Cell. Endocrinol. 302 (2009) 1–11.

[21] N. Yivgi-Ohana, N. Sher, N. Melamed-Book, S. Eimerl, M. Koler, P.R. Manna, D.M.Stocco, J. Orly, Transcription of steroidogenic acute regulatory protein in the rodentovary and placenta: alternativemodes of cyclic adenosine 3′, 5′-monophosphate de-pendent and independent regulation, Endocrinology 150 (2009) 977–989.

[22] S.L. Gyles, C.J. Burns, B.J. Whitehouse, D. Sugden, P.J. Marsh, S.J. Persaud, P.M. Jones,ERKs regulate cyclic AMP-induced steroid synthesis through transcription of the ste-roidogenic acute regulatory (StAR) gene, J. Biol. Chem. 276 (2001) 34888–34895.

[23] B.F. Clem, E.A. Hudson, B.J. Clark, Cyclic adenosine 3′,5′-monophosphate (cAMP)enhances cAMP-responsive element binding (CREB) protein phosphorylationand phospho-CREB interaction with the mouse steroidogenic acute regulatoryprotein gene promoter, Endocrinology 146 (2005) 1348–1356.

[24] D.M. Stocco, X. Wang, Y. Jo, P.R. Manna, Multiple signaling pathways regulatingsteroidogenesis and steroidogenic acute regulatory protein expression: morecomplicated than we thought, Mol. Endocrinol. 19 (2005) 2647–2659.

[25] C.W. Gregory, R.M. DePhilip, Detection of steroidogenic acute regulatory protein (stAR)in mitochondria of cultured rat Sertoli cells incubated with follicle-stimulating hor-mone, Biol. Reprod. 58 (1998) 470–474.

[26] S.O. Abarikwu, A.B. Pant, E.O. Farombi, The protective effects of quercetin on the cyto-toxicity of atrazine on rat Sertoli-germ cell co-culture, Int. J. Androl. 35 (2012) 590–600.

[27] A.L. Woo, P.F. James, J.B. Lingrel, Characterization of the fourth alpha isoform ofthe Na, K-ATPase, J. Membr. Biol. 169 (1999) 39–44.

[28] L.D. Newton, S. Krishnakumar, A.G. Menon, J.P. Kastelic, F.A. van der Hoorn, J.C.Thundathil, Na+/K+ATPase regulates sperm capacitation through a mechanisminvolving kinases and redistribution of its testis-specific isoform, Mol. Reprod.Dev. 77 (2010) 136–148.

[29] W. Schoner, G. Scheiner-Bobis, Endogenous and exogenous cardiac glycosides:their roles in hypertension, salt metabolism, and cell growth, Am. J. Physiol. CellPhysiol. 293 (2007) C509–C536.

[30] D.J. McConkey, Y. Lin, L.K. Nutt, H.Z. Ozel, R.A. Newman, Cardiac glycosides stim-ulate Ca2+ increases and apoptosis in androgen independent, metastatic humanprostate adenocarcinoma cells, Cancer Res. 60 (2000) 3807–3812.

[31] P. Kometiani, L. Liu, A. Askari, Digitalis-induced signaling by Na+/K+-ATPase inhuman breast cancer cells, Mol. Pharmacol. 67 (2005) 929–936.

[32] A. Kulikov, A. Eva, U. Kirch, A. Boldyrev, G. Scheiner-Bobis, Ouabain activates sig-naling pathways associated with cell death in human neuroblastoma, Biochim.Biophys. Acta 1768 (2007) 1691–1702.

[33] T. Mijatovic, F. Dufrasne, R. Kiss, Cardiotonic steroids-mediated targeting of theNa(+)/K(+)-ATPase to combat chemoresistant cancers, Curr. Med. Chem. 19(2012) 627–646.

[34] M. Scobey, S. Bertera, J. Somers, S. Watkins, A. Zeleznik, W. Walker, Delivery of acyclic adenosine 3′,5′-monophosphate response element-binding protein (creb)mutant to seminiferous tubules results in impaired spermatogenesis, Endocrinology142 (2001) 948–954.

[35] F. Rahman, H.C. Christian, Non-classical actions of testosterone: an update, TrendsEndocrinol. Metab. 18 (2007) 371–378.

[36] W.H. Walker, Non-classical actions of testosterone and spermatogenesis, Philos.Trans. R. Soc. Lond. B Biol. Sci. 365 (2010) 1557–1569.

[37] W.H. Walker, Testosterone signaling and the regulation of spermatogenesis,Spermatogenesis 1 (2011) 116–120.

[38] T. Almog, Z. Naor, Mitogen activated protein kinases (MAPKs) as regulators of sper-matogenesis and spermatozoa functions, Mol. Cell. Endocrinol. 282 (2008) 39–44.

[39] M.W. Li, D.D. Mruk, C.Y. Cheng, Mitogen-activated protein kinases in male repro-ductive function, Trends Mol. Med. 15 (2009) 159–168.

[40] C. Sette, M. Barchi, A. Bianchini, M. Conti, P. Rossi, R. Geremia, Activation of themitogen-activated protein kinase ERK1 during meiotic progression of mousepachytene spermatocytes, J. Biol. Chem. 274 (1999) 33571–33579.

[41] S. Di Agostino, F. Botti, A. Di Carlo, C. Sette, R. Geremia,Meiotic progression of isolatedmouse spermatocytes under simulated microgravity, Reproduction 128 (2004)25–32.

[42] B. Liu, J. Yu, L. Taylor, X. Zhou, P. Polgar, Microarray and phosphokinase screeningsleading to studies on ERK and JNK regulation of connective tissue growth factorexpression by angiotensin II 1a and bradykinin B2 receptors in Rat1 fibroblasts,J. Cell. Biochem. 97 (2006) 1104–1120.

[43] E. Laag, M. Majidi, M. Cekanova, T. Masi, T. Takahashi, H.M. Schuller, NNK activatesERK1/2 and CREB/ATF-1 via beta-1-AR and EGFR signaling in human lung adenocar-cinoma and small airway epithelial cells, Int. J. Cancer 119 (2006) 1547–1552.

[44] Z. He, J. Jiang, M. Kokkinaki, N. Golestaneh, M.C. Hofmann, M. Dym, Gdnfupregulates c-Fos transcription via the Ras/Erk1/2 pathway to promote mousespermatogonial stem cell proliferation, Stem Cells 26 (2008) 266–278.

[45] M. Delidaki, M. Gu, A. Hein, M. Vatish, D.K. Grammatopoulos, Interplay of cAMP andMAPK pathways in hCG secretion and fusogenic gene expression in a trophoblastcell line, Mol. Cell. Endocrinol. 332 (2011) 213–220.

[46] D. Rosenberg, L. Groussin, E. Jullian, K. Perlemoine, X. Bertagna, J. Bertherat, Role ofthe PKA-regulated transcription factor CREB in development and tumorigenesis ofendocrine tissues, Ann. N. Y. Acad. Sci. 968 (2002) 65–74.

[47] J.C. Chrivia, R.P. Kwok, N. Lamb, M. Hagiwara, M.R. Montminy, R.H. Goodman,Phosphorylated CREB binds specifically to the nuclear protein CBP, Nature 365(1993) 855–859.

[48] R.P. Kwok, J.R. Lundblad, J.C. Chrivia, J.P. Richards, H.P. Bachinger, R.G. Brennan,S.G. Roberts, M.R. Green, R.H. Goodman, Nuclear protein CBP is a coactivator forthe transcription factor CREB, Nature 370 (1994) 223–226.

519R. Dietze et al. / Biochimica et Biophysica Acta 1833 (2013) 511–519

[49] B.L. Kee, J. Arias, M.R. Montminy, Adaptor-mediated recruitment of RNA polymeraseII to a signal-dependent activator, J. Biol. Chem. 271 (1996) 2373–2375.

[50] R.P. Rehfuss, K.M. Walton, M.M. Loriaux, R.H. Goodman, The cAMP-regulatedenhancer-binding protein ATF-1 activates transcription in response to cAMP-dependent protein kinase A, J. Biol. Chem. 266 (1991) 18431–18434.

[51] M. Kobayashi, A. Shimomura, M. Hagiwara, K. Kawakami, Phosphorylation ofATF-1 enhances its DNA binding and transcription of the Na, K-ATPase alpha 1subunit gene promoter, Nucleic Acids Res. 25 (1997) 877–882.

[52] P. Gupta, R. Prywes, ATF1 phosphorylation by the ERK MAPK pathway is requiredfor epidermal growth factor-induced c-jun expression, J. Biol. Chem. 277 (2002)50550–50556.

[53] A. Shimomura, Y. Ogawa, T. Kitani, H. Fujisawa,M.Hagiwara, Calmodulin-dependentprotein kinase II potentiates transcriptional activation through activating transcrip-tion factor 1 but not cAMP response element-binding protein, J. Biol. Chem. 271(1996) 17957–17960.

[54] Y.M. Lin, M.Y. Liu, S.L. Poon, S.F. Leu, B.M. Huang, Gonadotrophin-releasinghormone-I and -II stimulate steroidogenesis in prepubertal murine Leydig cellsin vitro, Asian J. Androl. 10 (2008) 929–936.

[55] P. Singh, A. Krishna, Effects of GnRH agonist treatment on steroidogenesis andfolliculogenesis in the ovary of cyclic mice, J. Ovarian Res. 3 (2010) 26.

[56] F. Rosati, N. Sturli, M.C. Cungi, M. Morello, F. Villanelli, G. Bartolucci, C. Finocchi, A.Peri, M. Serio, G. Danza, Gonadotropin-releasing hormone modulates cholesterolsynthesis and steroidogenesis in SH-SY5Y cells, J. Steroid Biochem. Mol. Biol. 124(2011) 77–83.

[57] Q.Weng,M.S.Medan, G.Watanabe, T. Tsubota, Y. Tanioka, K. Taya, Immunolocalizationof steroidogenic enzymes P450scc, 3betaHSD, P450c17, and P450arom in Gottingenminiature pig testes, J. Reprod. Dev. 51 (2005) 299–304.

[58] S.E. Pollack, E.E. Furth, C.B. Kallen, F. Arakane, M. Kiriakidou, K.F. Kozarsky, J.F.Strauss III, Localization of the steroidogenic acute regulatory protein in humantissues, J. Clin. Endocrinol. Metab. 82 (1997) 4243–4251.

[59] J.M. Hamlyn, M.P. Blaustein, S. Bova, D.W. DuCharme, D.W. Harris, F. Mandel, W.R.Mathews, J.H. Ludens, Identification and characterization of a ouabain-like com-pound from human plasma, Proc. Natl. Acad. Sci. U. S. A. 88 (1991) 6259–6263.

[60] J.M. Hamlyn, Z.R. Lu, P. Manunta, J.H. Ludens, K. Kimura, J.R. Shah, J. Laredo, J.P.Hamilton, M.J. Hamilton, B.P. Hamilton, Observations on the nature, biosynthesis,secretion and significance of endogenous ouabain, Clin. Exp. Hypertens. 20 (1998)523–533.

[61] S. Li, C. Eim, U. Kirch, R.E. Lang, W. Schoner, Bovine adrenals and hypothalamusare a major source of proscillaridin A- and ouabain-immunoreactivities, Life Sci.62 (1998) 1023–1033.

[62] B.R. Boulanger,M.P. Lilly, J.M. Hamlyn, J. Laredo, D. Shurtleff, D.S. Gann, Ouabain is se-cretedby the adrenal gland of the awakedogs, Am. J. Physiol. 264 (1993) E413–E419.

[63] F. Masugi, T. Ogihara, T. Hasegawa, K. Sagakuchi, Y. Kumahara, Normalization ofhigh plasma level of ouabain-like immunoreactivity in primary aldosteronismafter removal of adenoma, J. Hum. Hypertens. 2 (1988) 17–20.

[64] A. Vadazs, P. Jakobi, J. Stoler, A.Makler, N. Krivoy, Endogenous digoxin-like immuno-reactivity measured in seminal fluid from a normal male population, Gynecol.Obstet. Invest. 33 (1992) 236–238.