BIOLOGY OF REPRODUCTION 68, 954–960 (2003) Published online before print 23 October 2002. DOI 10.1095/biolreprod.102.008938
Cytoplasmic Versus Nuclear Localization of Fos-Related Proteins in the Frog, Rana esculenta, Testis: In Vivo and Direct In Vitro Effect of a Gonadotropin-Releasing Hormone Agonist1 Gilda Cobellis, Rosaria Meccariello, Sergio Minucci, Carmela Palmiero, Riccardo Pierantoni,2 and Silvia Fasano Dipartimento di Medicina Sperimentale, sez. ‘‘F. Bottazzi,’’ 80138 Napoli, Italy ABSTRACT
have been found in testis, and their receptors (GnRH-R) have been localized in Leydig cells of several mammalian species [1, 2], excluding mouse [3]. Recently, gene expression of GnRH and GnRH-R has been evidenced in the gonads of human and several vertebrate species [4, 5]. As with the rodent model, amphibians have mainly been used to demonstrate direct gonadal action of GnRH molecular forms [1, 6]. In particular, in Rana esculenta testis, a GnRH agonist (GnRHa; buserelin) has been found to increase androgen production and spermatogonia (SPG) proliferation throughout specific receptors [7–11]. Antiproliferative activity of GnRH-like peptides has been demonstrated in tumor cells [12–16], but whether this effect is caused by receptor downregulation is still a matter of debate [17]. The possibility that GnRH interacts with the cell cycle has also been demonstrated [18]. Expression of the protooncogene c-fos has been associated with cell proliferation [19] and the physiology of mammalian and nonmammalian male germ cells [20–23]. Moreover, c-fos null mutation has been shown to have a fundamental role in mouse gametogenesis [24]. A relationship between GnRH and fos activity has been described in pituitary cells. In fact, in GT1-7 cells, GnRH induces a marked and transient increase of c-fos expression [25]. In addition, in T3-1 cells, GnRH stimulates extracellular regulated kinase (ERK) phosphorylation and c-fos expression [26]. From the above-quoted references on GnRH and fos activity, a relationship between the peptide and the protooncogene can be hypothesized for spermatogenesis. This would add a fundamental step in our understanding of the mechanism of GnRH action at the gonadal level. As far as we know, data related to possible direct functional interaction between gonadal GnRH-like substances and fos activity in the testis are nonexistent. Therefore, because of the presence of fos activity in primary (I) SPG of R. esculenta [23], the aim of the present study was to investigate the relationship between testicular GnRH-like substances and Fos. We have previously described a 68-kDa Fos protein as a nuclear phosphoprotein coming from a 52-kDa form stored in the cytoplasm of SPG [23]. In this respect, localization of cytoplasmic and nuclear Fos protein exclusively in SPG of frog testis during the annual sexual cycle [23, 27] strongly supports the use of this animal model. We asked whether nuclear localization of Fos-like material was related to GnRH-induced SPG proliferation. Localization of c-Fos in the cytoplasmic compartment, generally considered to be heretical, has also been confirmed in neurons of this species [28]. However, cytoplasmic localization of Jun and Fos (AP1 transcription factor) has also been found in
Evidence has been accumulated indicating that GnRH-like peptides are present in a variety of extrabrain areas of mammalian and nonmammalian vertebrates. A pioneer study carried out in the frog, Rana esculenta, demonstrated that testicular GnRH induced spermatogonial proliferation. Recently, we have shown that in proliferating spermatogonia (SPG) of frogs, a change of localization of the oncoprotein Fos, from the cytoplasm to the nucleus, occurs. This leads to the hypothesis that one or more testicular GnRH peptides may regulate SPG proliferation through Fos family proteins. Therefore, in vivo experiments in intact R. esculenta and in vitro incubations of testis fragments have been carried out using GnRH agonist (GnRHa; buserelin) and GnRH antagonist (D-pGlu1,D-Phe2,D-Trp3,6GnRH). Cytoplasmic and nuclear Fos-like protein localization has been found by Western blot analysis in testicular extracts. Immunocytochemistry confirmed that cytoplasmic immunostaining was restricted to SPG; change of localization into the nuclear compartment was observed after GnRHa treatment. Northern blot analysis showed that treatments of testis fragments with GnRHa did not modify testicular c-fos mRNA expression. On the contrary, a Fos-like protein of 52 kDa, while not affected in vivo, disappeared from testicular cytosolic extracts after in vitro treatment with GnRHa. Contemporaneously, a 55-kDa Fos-related signal appeared in nuclear extracts. The GnRH antagonist counteracted the effects of GnRHa. Furthermore, in vivo treatments showed that GnRHa acted negatively on a 43-kDa nuclear Fos-related signal and that gonadotropins caused the decrease of 52-kDa cytoplasmic signal. In conclusion, we show, to our knowledge for the first time, that Fos is regulated by GnRHa directly (not through the pituitary) at the testicular level. The main effect appears to be related to Fos translocation from cytoplasmic to nuclear compartments of SPG.
gonadotropin-releasing hormone, spermatogenesis, testis
INTRODUCTION
Gonadotropin-releasing hormone is a hypothalamic neurohormone best known for its action in triggering gonadotropin discharge from the pituitary in vertebrates [1]. The GnRH-like substances, probably coming from Sertoli cells, Supported by grants ex40% ‘‘GEREMIA,’’ MURST ex60%, CNR Target Project ‘‘Biotechnology,’’ and Regione Campania. Correspondence: Riccardo Pierantoni, Dipartimento Medicina Sperimentale, section ‘‘Filippo Bottazzi,’’ Via Costantinopoli 16, 80138 Napoli, Italy. FAX: 39 081 5667536; e-mail:
[email protected]
1
2
Received: 11 July 2002. First decision: 6 August 2002. Accepted: 27 September 2002. Q 2003 by the Society for the Study of Reproduction, Inc. ISSN: 0006-3363. http://www.biolreprod.org
954
GnRH EFFECT ON FOS TRANSLOCATION
other experimental models [22, 29], including tumor cells [30]. MATERIALS AND METHODS
Animals Intact male frogs (R. esculenta) were captured near Naples. The animals for in vivo and in vitro treatments were killed under anesthesia with MS222 (Sigma Chemical Co., St. Louis, MO), and testes were immediately prepared for histological observation or stored at 2808C until processed for protein preparation or RNA extraction. Brains were also stored until used as positive-control tissue. This research was approved by the Italian Ministry of University and Scientific and Technological Research.
In Vivo Experiments GnRHa (buserelin) treatments. Nine of 45 animals were immediately killed and considered as fresh controls to analyze putative effects of captivity that could arise during the experimental protocol [31]. The remaining frogs were divided into three experimental groups (n 5 12/group) and injected (for 2 wk on alternate days into the dorsal sac) with 100 ml of vehicle (Krebs Ringer bicarbonate buffer [KRB], pH 7.4) alone (captive control group), with 1025 M [D-ser(But)6]GnRH(1–9)nonapeptide-ethylamide (GnRHa; buserelin; a gift from Dr. J. Sandow, Hoechst, Frankfurt, Germany) in 100 ml of KRB (GnRHa group), or with 100 ml of KRB containing 1025 M GnRHa at 1 h after the injection of 100 ml of KRB containing 1023 M GnRH antagonist (D-pGlu1,D-Phe2,D-Trp3,6-GnRH; Peninsula Labs Ltd., St. Helens, U.K.; GnRHa 1 antagonist group). After 2.5 h from the last injection, testes were removed and used for Western blot analysis or immunocytochemistry. The doses were chosen on the basis of dose-response experiments carried out in previous studies [7, 8]. Hypophysis homogenate treatments. Animals (n 5 10) were injected with 100 ml of KRB (n 5 5; controls) or with one third of a hypophysis (n 5 5; treatment group) gently homogenized in 100 ml of KRB using a type B pestle. Injections were carried out for 2 wk on alternate days into the dorsal sac. After 2.5 h from the last injection, testes were removed and proteins extracted for Western blot analysis.
955
KCl, 12% glycerol, 0.1 mM EGTA, 0.5 mM dithiothreitol, and 0.5 mM spermidine) in the presence of protease inhibitors (4 mg/ml each of leupeptin, aprotinin, pepstatin A, chymostatin, PMSF, and 5 mg/ml of N-tosylL-phenylalanine chloromethyl ketone [TPCK]) and phosphatase inhibitors (0.3 nM okadaic acid and 1 mM sodium orthovanadate [Sigma]). After centrifugation at 800 3 g, the supernatant was removed and centrifuged on glycerol at 100 000 3 g for 1 h at 48C to obtain cytosolic extracts. The nuclear pellet, washed three times, was resuspended with 1.2 volumes (1.2 ml/mg pellet) of buffer (10 mM Hepes [pH 7.9], 1.5 mM MgCl2, 420 mM NaCl, 15% glycerol, 0.1 mM EGTA, 0.5 mM dithiothreitol, and 2 mM spermidine) in the presence of protease (4 mg/ml each of leupeptin, aprotinin, pepstatin A, chymostatin, PMSF, and 5 mg/ml of TPCK) and phosphatase (0.3 nM okadaic acid and 1 mM sodium orthovanadate) inhibitors and mixed at 48C for 30 min. Nuclei were pelleted by centrifugation at 800 3 g. The supernatants, containing nuclear proteins, were collected. Protein concentrations of cytoplasmic and nuclear extracts were determined by Lowry assay [36]. A similar protocol was used for the preparation of brain protein extracts.
Western Blot Analysis Twenty-five micrograms of proteins per lane were separated by SDSpolyacrylamide gel (8% or 10%) electrophoresis and immediately transferred to nitrocellulose (Amersham Pharmacia Biotech) at 280 mA for 2.5 h at 48C as already described [37]. The membranes, stained by Ponceau S (Sigma) as a control for protein transfer, were treated for 3 h to prevent nonspecific adsorption with blocking solution (5% [w/v] nonfat powdered milk, 0.25% Tween-20 in Tris-buffered saline [TBS; pH 7.6]) and incubated with the primary antibody (0.2 mg/ml, 1:1000; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) diluted in TBS-3% (w/v) nonfat powdered milk solution overnight at 48C on an orbital shaker. Filters, washed in TBS-0.1% Tween-20, were incubated with 1:1000 horseradish peroxidaseconjugated anti-goat immunoglobulin (Ig) G (DAKO, Glostrup, Denmark) in TBS-1% normal swine serum (NSS; DAKO) and then washed three times in TBS-0.25% Tween-20. The immune complexes were detected using the enhanced chemiluminescence-Western blot detection system (Amersham Pharmacia Biotech) following the manufacturer’s instructions. The membranes were stripped at 608C for 30 min in stripping buffer (100 mM 2-mercaptoethanol, 2% SDS, and 62.5 mM Tris-HCl [pH 7.6]) and reprobed to check specificity of immunoreaction as described below.
In Vitro Experiments
Immunocytochemistry
Testes were removed from 70 animals, washed in KRB (0 h; control group) and incubated for 1, 2, or 8 h in 10 ml of KRB with 1026 M GnRHa or with 1026 M GnRHa plus 1024 M GnRH antagonist at 248C. The agonist was added in the medium 30 min after the antagonist. Each experimental group consisted of 20 testes. The doses were chosen on the basis of dose-response experiments carried out in previous studies [7, 8]. Ten testes per time/treatment were processed for Northern blot analysis, and 10 testes per time/treatment were processed for nuclear and cytosolic protein extractions.
Frog testes, rapidly removed and fixed in Bouin fluid, were dehydrated in ethanol, cleared in xylene, and embedded in paraffin. Tissue sections (thickness, 5 mm) were processed by the peroxidase-antiperoxidase technique (PAP) as already described [37]. Four sections per animal per treatment were examined. The sections were treated for 20 min with H2O2 to block endogenous peroxidases. Incubations were performed at 48C in a moist chamber for 14 h with the primary antiserum. The antiserum was diluted 1:50 to 1:200 in 0.1 M PBS at pH 7.1 containing 1% NSS. After washing in PBS, the sections were incubated for 1 h at room temperature with a secondary antiserum (anti-Ig raised in sheep; DAKO) diluted 1:50 in PBS. After washing in PBS, the sections were incubated for 1 h with goat PAP complex (DAKO). The antigen was visualized using 3,39-diaminobenzidine tetrahydrochloride (Sigma) and 30% H2O2 in Tris-HCl (pH 7.6). Counting of immunopositive ISPG was expressed as number of cytoplasmic and nuclear immunopositive cells per total ISPG counted per section, multiplied by 100.
RNA Preparation and Northern Blot Analysis Total RNA was prepared as previously described [32] from frog testes using guanidium isothiocyanate [33]. A 1.1-kilobase (kb) v-fos probe was obtained by double digestion with PstI/SalI from pBR322 [34]. Northern blot analysis employed [a-32P]dCTP (3000 Ci/mmol) by random priming (Amersham Pharmacia Biotech, Buckinghamshire, U.K.) following the manufacturer’s instruction. Blots were prehybridized and probed in 50% (v/v) formamide, 53 SSPE (13 SSPE: 150 mM NaCl, 10 mM NaH2PO4, and 1 mM Na-EDTA), 23 Denhardt solution, 0.1% (w/v) SDS, 40 mg/ml of transfer RNA (Escherichia coli), and 100 mg/ml of denatured salmon sperm DNA overnight at 428C, and then blots were washed. Normalization employed a-32P-labeled 28S ribosomal DNA probe from Drosophila melanogaster. The relative amount of c-fos-like mRNA was determined by densitometric analysis of the autoradiographs of the Northern blots. Autoradiographs were quantified by scanning densitometric analysis (Ultroscan XL; LKB, Uppsala, Sweden).
Protein Extract Preparation Cytoplasmic and nuclear extracts were prepared using modifications of the method previously used by Dignam et al. [35]. Briefly, 100 mg of testicular tissue was gently homogenized using a type B pestle in seven volumes (w/v) of buffer (10 mM Hepes [pH 7.9], 1.5 mM MgCl2, 10 mM
Fos Antibody and Control of Immunoreaction Specificity The antibody used for Western blot analysis was polyclonal anti-c-Fos (sc-253-G; Santa Cruz Biotechnology) raised in goat against an acid sequence within a highly conserved domain in Fos family members (c-Fos, FosB, Fra-1, and Fra-2) of human origin identical to the corresponding mouse, rat, and chicken sequences. Therefore, we refer to Fos-like protein instead of c-Fos activity. Specificity of the signals has previously been assessed in frog testis extracts by immunoprecipitation [23]. Specificity has also been tested by extinguishing the reaction with an excess amount (1026 M) of the cognate peptide (sc-253P; Santa Cruz Biotechnology). Furthermore, brain extracts of R. esculenta have been used as positive control for Fos expression. A polyclonal sheep anti c-Fos (Serva, Heidelberg, Germany) was used, and aliquots of the antibody were preabsorbed with purified antigen to verify the specificity of immunocytochemical reactions. Pictures showing
956
COBELLIS ET AL.
FIG. 1. A) Western blot detection in the frog, R. esculenta, testis of Fos proteins in cytosolic (lanes 1–4) or nuclear (lanes 5– 8) extracts of fresh animals (lanes 1 and 5) and captive animals treated with KRB (lanes 2 and 6), GnRHa (lanes 3 and 7), and GnRHa 1 antagonist (Ant; lanes 4 and 8). Cytosolic and nuclear extracts of R. esculenta brain were used as positive control of Fos expression (lanes 9 and 10, respectively). B) Specificity of Fos signals was checked by Western blot analysis using Fos antibody preabsorbed with an excess amount (1026 M) of antigen. C) Graphed values of cytoplasmic (52 kDa) and nuclear (68 kDa) Fos signals. Values corresponding to the total amount of 521 68-kDa signals (inset) show no significant variations after GnRHa treatments. Data are expressed as the mean 6 SEM. Results are representative of three independent assays. a,bSame letter indicates no significant differences; different letters indicate P , 0.01.
sections of frog testes treated with preabsorbed antiserum are presented and have already been shown elsewhere [37].
Statistics Quantification of 52-, 55-, and 68-kDa signals detected by Western blot analysis of in vivo- and in vitro-incubated testis pieces was carried
out using a transmitting scanning densitometer (Ultroscan XL) and evaluating the integration of peak areas of each signal. The amounts of the signals of in vivo treatments are reported as cytoplasmic (52 kDa), nuclear (68 kDa), and total (52 1 68 kDa) values. The relative amounts of the signals of the in vitro experiments are expressed as fold-increases over the minimal value registered. The c-fos mRNA and 28S rRNA were quantified by densitometric analysis of Northern blot signals. Graphed values of relative amounts of c-fos mRNA were obtained by evaluating the integration of peak areas of c-fos mRNA/28S rRNA. Significance of differences was evaluated by ANOVA followed by the Duncan test for multigroup comparison.
RESULTS
In Vivo Treatments
FIG. 2. A) Western blot analysis of cytosolic protein extracts from testes of captive frogs, R. esculenta. Untreated (2) and treated (1) animals with hypophysis homogenate are shown. B) Graphed values of 52-kDa Fos signals show a marked decrease (P , 0.01) after hypophysis homogenate treatment. Data are expressed as the mean 6 SEM. Results are representative of three independent assays. a,bDifferent letters indicate P , 0.01.
The animals were treated in vivo to examine the effects of GnRHa on Fos expression in frog testis. Western blot analysis was used to study Fos-like protein pattern expression in both cytosolic and nuclear protein extracts. In the testis of fresh control animals, a band of 52 kDa was present in cytoplasmic extracts, whereas a band of 68 kDa was evidenced in nuclear extract (Fig. 1A, lanes 1 and 5). Cytoplasmic preparations of both captive control and captive treated frogs showed a single band of approximately 52 kDa (Fig. 1A, lanes 2–4). The signal was stronger in samples from captive frogs (P , 0.01) as compared to the fresh control group (Fig. 1, A [lane 1 vs. 2–4] and C). Furthermore, when captive frogs were treated with hypophysis homogenate, a marked decrease (P , 0.01) of the 52-kDa signal was observed (Fig. 2). In nuclear extracts, bands of 68 and 43 kDa were present in captive frogs (Fig. 1A, lanes 6 and 8). Densitometric analysis carried out on 52- 1 68-kDa signals demonstrated
GnRH EFFECT ON FOS TRANSLOCATION
957 FIG. 3. A) Localization of Fos-immunoreactive signal in the cytoplasm (arrow) and nuclei (arrowheads) of ISPG in the frog, R. esculenta. Magnification 3340. B) Negative control of immunoreaction was obtained by preabsorbing the antiserum with an excess amount (1026 M) of the cognate peptide. C) Number of Fos-immunopositive ISPG of animals (R. esculenta) treated with KRB, GnRHa, and GnRHa 1 antagonist (Ant; see Materials and Methods). Immunopositive ISPG were counted in four randomly chosen sections per testis per frog (n 5 12 frogs/experimental group). The number of cytoplasmic and nuclear immunopositive cells is expressed as the number of immunopositive cells per total ISPG counted per section, multiplied by 100. Values are expressed as the mean 6 SEM.
that GnRHa treatment did not affect Fos levels. In fact, the value corresponding to the total amount of the two signals (Fig. 1C, inset) did not show significant variations after GnRHa treatment. With respect to the 43-kDa signal, this was not present in the GnRHa-treated group (Fig. 1A, lane7). Interestingly, the GnRH antagonist counteracted the effect of GnRHa (Fig. 1A, lane 8). To support specificity of the antiserum in the Western blot analysis, the reaction was extinguished with an excess amount of the cognate peptide (Fig. 1B). Worthy of note is the detection of Fos signals in brain protein extracts (cytosolic and nuclear preparations) of R. esculenta comparable in size to those found in testis protein extracts (Fig. 1A, lanes 9 and 10), which was consistent with the results of an earlier study concerning cytosolic and nuclear c-Fos immunoreactivity in brains during the annual cycle [28]. Therefore, the detection in a different tissue of identical Fos signals, showing the same intracellular localization, may further support specificity. By immunocytochemistry, Fos immunostaining in the captive control group was observed in both cytoplasmic and nuclear compartments of ISPG (Fig. 3A). Specificity was suggested by extinguishing the reaction with an excess amount of the cognate peptide (Fig. 3B) [27, 37]. The number of ISPG with cytoplasmic staining was significantly higher compared to the number of ISPG showing nuclear staining in the control group (P , 0.01) (Fig. 3C). After GnRHa treatment, the number of ISPG showing nuclear immunopositivity increased and was higher compared to the number of ISPG with cytoplasmic staining (P , 0.05). The GnRH antagonist counteracted (P , 0.01) the effect of GnRHa (Fig. 3C).
Indeed, by Western blot analysis, a 52-kDa Fos-immunoreactive band was detected in cytoplasmic extract of the control group (Fig. 5A, lane 1). Eight hours after GnRHa stimulation, a gradual decrease (P , 0.01) of cytoplasmic Fos signal was observed (Fig. 5, A [lanes 3–5] and B). Concomitantly, the Fos signal increased (P , 0.01) in nuclear protein extracts (Fig. 5, A [lanes 2 and 6–8] and B). This effect was counteracted by GnRH antagonist (P , 0.05) (Fig. 5, A [lanes 9 and 10] and B). The size of nuclear Fos signal was higher (55 kDa) compared to the size of cytosolic Fos signal (52 kDa) (Fig. 5A), and this difference was not attributable to differences in
GnRHa Effect on Incubated Testis Pieces
Testis pieces, washed with KRB (0 h; control group), were incubated for 1, 2, or 8 h with GnRHa 1 antagonist. Northern blot analysis showed the detection of 1.9-kb c-fos mRNA (Fig. 4A), confirming previous results [32]. Treatment with GnRHa did not significantly modify c-fos expression (Fig. 4B). When protein extract was examined, testes treated with GnRHa showed the translocation of Foslike protein from cytoplasm into the nuclear compartment.
FIG. 4. A) Autoradiographs of Northern blots showing the effects of GnRHa and GnRHa 1 antagonist (Ant) treatments on c-fos mRNA concentration in the frog, R. esculenta, testis. B) The dynamics of steady-state c-fos-like mRNA levels were graphed after quantitative densitometric analysis of fos autoradiographic signals, corrected on the basis of 28S rRNA signals in the same blot. No significant variations were observed. Results are the mean 6 SEM. One out of three independent experiments is shown.
958
COBELLIS ET AL.
FIG. 5. A) Western blot analysis showing Fos signals in cytoplasmic (lanes 3–5) and nuclear (lanes 6–8) extracts from R. esculenta testes either untreated (control group) or treated with GnRHa 6 antagonist. B) Graphed values of relative amounts of 52to 55-kDa signals. Results are represented as the fold-increase over the minimal value found in the nuclear extract of control group. Data are expressed as the mean 6 SEM. Results are representative of three independent assays. Capital letters indicate significance of cytoplasmic Fos levels; small letters are related to nuclear Fos levels. Different letters indicate significant differences. A vs. B, C, and D, P , 0.01; c vs. a and b, P , 0.01; a vs. b, P , 0.05.
buffer composition. Indeed, in Figure 6, we show that the size difference available between cytoplasmic and nuclear Fos-like proteins was still present when buffer with the same saline concentration was used. DISCUSSION
Our results confirm previous data indicating the presence of Fos proteins in cytoplasmic and nuclear compartments of the frog, R. esculenta, testis with a molecular size of 43, 52, and 68 kDa [23]. Because of the exclusive localization of 52- and 68-kDa signals in ISPG [23], the present data concerning incubation of testis pieces show, to our knowledge for the first time in a vertebrate, that GnRHa induces the translocation of 52-kDa Fos protein into the nucleus of SPG acting directly (not through the pituitary) on the gonads. This suggests that GnRH-like substances, detected in gonads of several vertebrate species [3], may act in a paracrine/autocrine manner through Fos-mediated mechanisms. In the present study, an antiserum that recognizes all Fos family members [38] (c-Fos, Fos B, Fra1, and Fra2) was used. Therefore, it is not surprising to find 43-, 52-, and 68-kDa Fos signals. Specificity has previously been tested [23]. Moreover, frog brain, previously described by im-
FIG. 6. The 52- and 55-kDa signals in cytoplasmic (lane 1) and nuclear (lane 2) extracts from testes of the frog, R. esculenta, treated with GnRHa for 8 h. Both extracts have been adjusted at 80 mM NaCl. Western blot analysis shows that the nuclear Fos-immunoreactive protein is 55 kDa, whereas the cytoplasmic form is 52 kDa. One experiment representative of four SDS-PAGE is shown.
munocytochemistry (using a specific c-Fos antibody) to possess cytoplasmic and nuclear Fos immunoreactivity [28], was used as a positive control of Fos immunoreaction. Identical signals were observed in the brain, a positive control, and the testis. Similar data pertaining to the molecular weights of Fos family proteins have already been reported [39]. The 43kDa signal is well in the range of Fra proteins [40, 41], and unpublished data from our laboratory seems to confirm this hypothesis. Signals of 52 and 68 kDa correspond to the size range of c-Fos [42–45], described as a heterogeneous protein with an approximate molecular mass of 52 to 68 kDa or greater [42–46]. This heterogeneity is reported to be largely the result of phosphorylation [46, 47] within serine groups (Ser 362–364, 368–371, and 373–374), which regulates the transcriptional activity of the protein as AP1 complex [38]. In intact animals, 52-kDa Fos signal in testicular cytoplasmic extract is markedly strong in captive animals compared with fresh frogs. Because of the well-documented impairment of pituitary activity after capture [31], we tested whether hypophysis homogenate treatment decreased the Fos-immunoreactive band in captive animals. Injecting hypophysis homogenate, we showed that pituitary regulates Fos activity, decreasing the 52-kDa form signal in frog testis. It is worthy of note that in rat granulosa cells, LH and FSH [48–50] modulate c-fos expression. The 43-kDa signal, hypothesized to be a Fra protein ([23], unpublished results), appears in testicular nuclear preparations of captive frogs. Treatments with GnRHa, which triggers gonadotropin discharge [51], suppress the 43-kDa signal to levels that were similar to those of freshly caught animals. Treatment with GnRH antagonist reverses the effect, restoring the situation evidenced in captive control animals. Therefore, the appearance of 43-kDa protein can be related to signals triggered by captivity. In certain periods of the year (September–October), the 43-kDa protein has been detected in cycling animals showing a normal seminiferous epithelium [23]. In animals characterized by testes enriched with SPG only and devoid of other stages, the 43-kDa signal was not evidenced [23]. It was therefore concluded that only the 52- and 68-kDa signals were specific to SPG cells [23]. Detection of Fos by immunocytochemistry indicates that
GnRH EFFECT ON FOS TRANSLOCATION
after GnRHa in vivo, the protein is primarily localized in nuclei of ISPG. However, Western blot analysis indicates that GnRHa treatment does not interfere with the increase of nuclear Fos signal (68 kDa). This discrepancy may result from a low sensitivity of the method compared with the direct observation of nuclei carried out by immunocytochemistry. Change of immunolocalization (from cytoplasm to the nucleus) has previously been observed after estrogen treatments in the frog, R. esculenta [37]. This treatment has been postulated to induce SPG proliferation. Moreover, in a recent report [23], we demonstrated that increase in the nuclei of SPG of the 68-kDa form, coming from the 52kDa protein, induces SPG proliferation. Because GnRH, acting directly on the testis, also induces SPG mitotic activity in frogs [9], we used in vitro incubation of testis pieces to investigate effects of GnRHa on Fos proteins. Analysis of the in vitro treatment shows that GnRHa does not increase c-fos expression, but it is suggested that the peptide induces Fos protein translocation from the cytoplasm into the nuclear compartment. Indeed, after GnRHa stimulation, the 52-kDa signal progressively increases into the nuclear extracts, showing here a size of 55 kDa. Therefore, contrary to the in vivo treatment, in vitro stimulation permitted the detection of different forms, represented by the 52- to 55-kDa bands, which were also seen by Western blot analysis. Interference of gonadotropins and/or duration of the in vivo experiment may explain discrepancies with in vitro results. After 8 h of incubation, the bulk of the signal is nuclear, whereas in control and GnRHa 1 antagonist groups, scant nuclear staining is detectable. This suggests that Fos protein cannot be detected in the nucleus in the absence of GnRHa. Conversely, in the presence of active GnRHa, the lower-molecular-weight form (52 kDa) of Fos protein, detected in cytoplasmic extract, may be progressively converted into the higher-molecular-weight form (55 kDa) in the nucleus. It is important to remark that the increase of molecular weight in nuclear extracts is not an artifact caused by the different buffer compositions used for migration of cytosolic and nuclear proteins. Indeed, when we have used the same saline concentration, the difference in size is still available between cytosolic and nuclear Fos signals. This difference in size can be caused by the degree of phosphorylation, and Fos phosphorylation in frog SPG has already been shown to occur [23]. We know from previous experiments that the 68-kDa form is a nuclear phosphoprotein coming from a 52-kDa form stored in the cytoplasm [23]. We have never observed the 68-kDa protein in nuclear extracts after GnRHa in vitro treatments; thus, further research is required to study a putative shift from 55-kDa nuclear Fos protein into the 68-kDa form. In conclusion, we have demonstrated that GnRHa modulates Fos activity in both the pituitary and the testis. At the testicular level, GnRHa seems to induce Fos-related protein translocation from the cytoplasm into the nuclear compartment (e.g., in SPG). We cannot exclude that the effects observed in the present study are mediated by further unknown intratesticular factors (probably coming from Leydig cells) triggered by GnRHa. Finally, we suppose that testicular GnRH-like substances may exert such a role (mimicked in the present paper by GnRHa) in species, human included, possessing one or more gonadal GnRH peptides [1, 4]. The intriguing possibility that, in general, Fos translocation might serve as transduction signal in GnRHmodulated proliferation of normal and transformed cells can also be suggested.
959
REFERENCES 1. Chieffi G, Pierantoni R, Fasano S. Immunoreactive GnRH in hypothalamic and extrahypothalamic areas. Int Rev Cytol 1991; 127:1–55. 2. Sharpe RM, Fraser HM. HCG stimulation of testicular LHRH-like activity. Nature 1980; 287:642–643. 3. Gnessi L, Fabbri A, Spera G. Gonadal peptides as mediators of development and functional control of the testis: an integrated system with hormones and local environment. Endocr Rev 1997; 18:541–609. 4. Von Schalburg KR, Harrower WL, Sherwood NM. Regulation and expression of gonadotropin-releasing hormone in salmon embryo and gonad. Mol Cell Endocrinol 1999; 157:41–54. 5. White RB, Eisen JA, Kasten TL, Fernald RD. Second gene for gonadotropin-releasing hormone in humans. Proc Natl Acad Sci U S A 1998; 95:305–309. 6. Segal SJ, Adejuwon CA. Direct effect of LHRH on testicular steroidogenesis in Rana pipiens. Biol Bull 1979; 157:393–398. 7. Pierantoni R, Iela L, d’Istria M, Fasano S, Rastogi RK, Delrio G. Seasonal testosterone profile and testicular responsiveness to pituitary factors and gonadotropin releasing hormone during two different phases of sexual cycle of the frog (Rana esculenta). J Endocrinol 1984; 102:387–392. 8. Pierantoni R, Fasano S, Di Matteo L, Minucci S, Varriale B, Chieffi G. Stimulatory effect of a GnRH agonist (buserelin) in in vitro and in vivo testosterone production by the frog (Rana esculenta) testis. Mol Cell Endocrinol 1984; 38:215–219. 9. Minucci S, Di Matteo L, Pierantoni R, Varriale B, Rastogi RK, Chieffi G. In vivo and in vitro stimulatory effect of GnRH analog (HOE 766) on spermatogonial multiplication in the frog, Rana esculenta. Endocrinology 1986; 119:731–736. 10. Di Matteo L, Minucci S, Fasano S, Pierantoni R, Varriale B, Chieffi G. GnRH antagonist decreases androgen production and spermatogonial multiplication in frog (Rana esculenta): indirect evidence for the existence of GnRH or GnRH-like material receptors in the hypophysis and testis. Endocrinology 1988; 122:62–67. 11. Fasano S, de Leeuw R, Pierantoni R, Chieffi G, van Oordt PGWJ. Characterization of gonadotropin-releasing hormone (GnRH) binding sites in the pituitary and testis of the frog, Rana esculenta. Biochem Biophys Res Commun 1990; 168:923–932. 12. Qayum A, Gullick W, Clayton RC, Sikora K, Waxman J. The effects of gonadotropin-releasing hormone receptor analogues in prostate cancer cells are mediated through specific tumor receptor. Br J Cancer 1990; 62:96–99. 13. Harris NC, Dutlow C, Eiden KA, Dong KW, Roberts JL, Millar RP. Gonadotropin-releasing hormone gene expression in MDA-MB231and Zr-75-1 breast carcinoma cell line. Cancer Res 1991; 51: 2577–2581. 14. Emons G, Ortmann O, Becker M, Irmer G, Springer B, Laun R, Holzen F, Schultz KD, Schally AV. High affinity binding and direct antiproliferative effects of LHRH analogous in human ovarian cancer cell lines. Cancer Res 1993; 53:5439–5446. 15. Emons G, Schroder B, Ortmann O, Westphalen W, Schulz KD, Schally AV. High affinity binding and direct antiproliferative effects of luteinizing hormone-releasing hormone analogous in human endometrial cancer cell lines. J Clin Endocrinol Metab 1993; 77:1458–1464. 16. Yano T, Pinski J, Radulovic S, Schally AV. Inhibition of human epithelial ovarian cancer cell growth in vitro by agonistic and antagonistic analogues of luteinizing hormone-releasing hormone. Proc Natl Acad Sci U S A 1994; 91:1701–1705. 17. Kang SK, Choi KC, Cheng KW, Nathwani PS, Auersperg N, Leung PCK. Role of gonadotropin-releasing hormone as an autocrine growth factor in human ovarian surface epithelium. Endocrinology 2000; 141: 72–80. 18. Minucci S, Fasano S, Pierantoni R. Induction of S-phase entry by gonadotropin-releasing hormone agonist (buserelin) in the frog, Rana esculenta, primary spermatogonia. Comp Biochem Physiol C Pharmacol Toxicol Endocrinol 1996; 113:99–102. 19. Weisz A, Bresciani F. Estrogen induces expression of c-fos and c-myc proto-oncogenes in rat uterus. Mol Endocrinol 1988; 2:816–824. 20. Kierszenbaum AL. Mammalian spermatogenesis in vivo and in vitro: a partnership of spermatogenic and somatic cell lineages. Endocr Rev 1994; 15:116–134. 21. Fasano S, Minucci S, Chieffi P, Garnier DH, Cobellis G, Jegou B, Pierantoni R. Detection of proto-oncogene-like activity in the testis of Scyliorhinus canicula (Elasmobranchs). Neth J Zool 1995; 45:157– 159. 22. Chieffi P, Angelini F, Pierantoni R. Proto-oncogene activity in the
960
23.
24. 25. 26.
27.
28.
29.
30.
31.
32. 33. 34. 35.
COBELLIS ET AL.
testis of the lizard, Podarcis s.sicula, during the annual reproductive cycle. Gen Comp Endocrinol 1998; 108:173–181. Cobellis G, Meccariello R, Fienga G, Pierantoni R, Fasano S. Cytoplasmic and nuclear Fos protein forms regulate resumption of spermatogenesis in the frog, Rana esculenta. Endocrinology 2002; 143: 163–170. Johnson RS, Spiegelman BM, Papaioannou V. Pleiotropic effects of a null mutation in the c-fos proto-oncogene. Cell 1992; 71:577–586. Cesnjaj M, Krsmanovic LZ, Catt KJ, Stejilkovic SS. Autocrine induction of c-fos expression in GT1 neuronal cells by gonadotropinreleasing hormone. Endocrinology 1993; 133:3042–3045. Mulvaney JM, Zhang T, Fewtrell C, Roberson MS. Calcium influx trough L-type channels is required for selective activation of extracellular signal-regulated kinase by gonadotropin-releasing hormone. J Biol Chem 1999; 274:29796–29804. Chieffi P, Minucci S, Cobellis G, Fasano S, Pierantoni R. Changes in proto-oncogene activity in the testis of the frog, Rana esculenta. during the annual reproductive cycle. Gen Comp Endocrinol 1995; 99: 127–136. Cobellis G, Vallarino M, Meccariello R, Pierantoni R, Masini MA, Mathieu M, Pernas-Alonso R, Chieffi P, Fasano S. Fos localization in cytosolic and nuclear compartments in neurones of the frog, Rana esculenta, brain: an analysis carried out in parallel with GnRH molecular forms. J Neuroendocrinol 1999; 11:725–735. Xavier F, Lagarrigue S, Guillomot M, Gaillard-Sanchez I. Expression of c-fos and c-jun proto-oncogenes in ovine trophoblasts in relation to interferon-tau expression and early implantation process. Mol Reprod Dev 1997; 46:127–137. Neyns B, Katesuwanasing D, Vermeij J, Bourgain C, Vandamme B, Amfo K, Lissens W, De Sutter P, Hooge-Peters E, De Greve J. Expression of the jun family genes in human ovarian cancer and normal ovarian surface epithelium. Oncogene 1996; 12:1247–1257. Licht P, McCreery BR, Barnes R, Pang R. Seasonal and stress related changes in plasma gonadotropin, sex steroids and corticosterone in bull frog, Rana catesbeiana. Gen Comp Endocrinol 1983; 50:124– 145. Cobellis G, Pierantoni R, Fasano S. c-fos- and c-jun-like mRNA expression in frog (Rana esculenta) testis during the annual reproductive cycle. Gen Comp Endocrinol 1997; 106:23–29. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987; 162:156–159. Curran T, Peters G, van Beveren C, Teich NM, Verma IM. FBJ murine osteosarcoma virus: identification and molecular cloning of biologically active proviral DNA. J Virol 1982; 44:674–682. Dignam JD, Lebowitz RM, Roeder AG. Accurate transcription by
View publication stats
36. 37. 38. 39. 40.
41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51.
RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res 1983; 11:1475–1489. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 1955; 193:265–275. Cobellis G, Pierantoni R, Minucci S, Pernas-Alonso R, Meccariello R, Fasano S. c-fos activity in Rana esculenta testis: seasonal and estradiol induced changes. Endocrinology 1999; 140:3238–3244. Karin M, Liu Z, Zandi E. AP-1 functions and regulation. Curr Opin Cell Biol 1997; 9:240–246. Nestler EJ, Barrot M, Self DW. FosB: a sustained molecular switch for addiction. Proc Natl Acad Sci U S A 2001; 98:11042–11046. Cohen DR, Ferreira PCP, Gentz R, Franza BR Jr, Curran T. The product of a fos-related gene, fra-1, binds cooperatively to the AP-1 site with Jun: transcription factor AP-1 is comprised of multiple protein complexes. Genes Dev 1989; 3:173–184. Nishina H, Sato H, Suzuki T, Dato M, Iba H. Isolation and characterization of fra-2, an additional member of the fos gene family. Proc Natl Acad Sci U S A 1990; 87:3619–3623. Curran T, Miller AD, Zokas L, Verma IM. Viral and cellular fos proteins: a comparative analysis. Cell 1984; 36:259–268. Curran T, Teich NM. Candidate product of the FBJ murine osteosarcoma virus oncogene: characterization of a 55,000-dalton phosphoprotein. J Virol 1982; 42:114–122. Ofir R, Dwarki VJ, Rashid D, Verma IM. Phosphorylation of C terminus of Fos protein is required for transcriptional transrepression of c-fos promoter. Nature 1990; 348:80–82. Li SL, Cougnon N, Bresson-Be´poldin L, Zhao SJ, Schlegel WC. Fos mRNA and FOS protein expression is induced by Ca12 influx in GH3B6, pituitary cells. J Mol Endocrinol 1996; 16:229–238. Tratner I, Ofir R, Verma IM. Alteration of a cyclic AMP-dependent protein kinase phosphorylation site in the c-Fos protein augments its transforming potential. Mol Cell Biol 1992; 12:998–1006. Meek DW, Street AJ. Nuclear protein phosphorylation and growth control. Biochem J 1992; 287:1–15. Delidow BC, White BA, Peluso JJ. Gonadotropin induction of c-Fos and c-Myc expression and deoxyribonucleic acid synthesis in rat granulosa cells. Endocrinology 1990; 126:2302–2306. Pennybacker M, Herman B. Follicle-stimulating hormone increases cfos mRNA levels in rat granulosa cells via a protein kinase C-dependent mechanism. Mol Cell Endocrinol 1991; 80:11–20. Ness JM, Kasson BG. Gonadotropin regulation of c-fos and c-jun messenger ribonucleic acids in cultured rat granulosa cells. Mol Cell Endocrinol 1992; 90:17–25. McCreery BR, Licht P. Effects of gonadectomy and sex steroids on pituitary gonadotropin release and response to gonadotropin releasing hormone (GnRH) agonist in the bullfrog, Rana catesbeiana. Gen Comp Endocrinol 1984; 54:283–296.
