The effects of activin A and follistatin on the release of follicle-stimulating hormone (FSH), luteinizing hormone (LH), growth hormone (GH) and prolactin (PRL) from dispersed pituitary cells of the bullfrogs Rana catesbeiana were studied. Activin A stimulated the release of FSH, GH, and PRL dosedependently, but not that of LH. Follistatin suppressed the activin-induced FSH, GH, and PRL release, but did not affect the basal secretion of those hormones. From the results obtained in this experiment, together with the previously obtained findings that activin B enhanced the release of FSH, LH, GH, and PRL, we conclude that activin A, in addition to activin B, influences the function of multiple types of pituitary cells in the bullfrog.
Evidence has accumulated that in mammals, different types of anterior pituitary cells regulate the function of the neighboring cells and/or themselves by secreting paracrine and/or autocrine factors. (Denef et al., 1989; Denef and Van Bael, 1998; Schwarts and Cherney, 1992). In amphibians, we have also demonstrated that pituitary hormones or substances present in the pituitary cells act in a paracrine and/or autocrine fashion to regulate the release of pituitary hormones. In the bullfrog (Rana catesbeiana), the α-subunit of glycoprotein hormones stimulates prolactin (PRL) release (Oguchi et al., 1996), PRL enhances the release of luteinizing hormone (LH) (Oguchi et al., 1997), and endothelin-3 existing in gonadotrophs (Suzuki et al., 1997a) stimulates the release of growth hormone (GH) and PRL (Suzuki et al., 1997b). Proopiomelanocortin (POMC)-derived peptides, namely, the N-terminal peptide of POMC (NPP), and adrenocorticotropic hormone (ACTH) enhance the release of LH (Aida et al., 1997), GH and PRL (Aida et al., 1999).
Recently, the presence of immunoreactive activin/inhibin βB in bullfrog gonadotrophs and thyrotrophs has been dem onstrated. Investigation of the effects of activin B and inhibin B on the release of gonadotropins from dispersed anterior pituitary cells revealed that activin B enhanced the release of both follicle-stimulating hormone (FSH) and LH, and that inhibin B blocked the activin B-induced enhancement of FSH and LH release without suppressing their basal secretion (Uchiyama et al., 2000), suggesting the possibility that activin B is one of the paracrine and/or autocrine factors in the bullfrog pituitary.
Activin and inhibin were first purified from mammalian follicular fluid as factors that regulate the secretion of FSH from pituitary cells in culture (Ling et al., 1985; 1986a,b; Mason et al., 1985; Rivier et al., 1985; Vale et al., 1986). Inhibin, a suppressor of FSH secretion, is composed of an a chain covalently linked to a β chain; whereas activin, a stimulator of FSH secretion, is composed of a dimer of the β chain. Since there are two structurally related β chains, activin has three isoforms (βAβA as activin A, βAβB as activin AB, and βBβB as activin B).
In the present experiment, we studied the effect of activin A on the release of FSH, LH, GH and PRL from dispersed pituitary cells of the bullfrog. The effect of follistatin, a potent inhibitor of activin, on the release of the four pituitary hormones in the presence or absence of activin A was also investigated.
MATERIALS AND METHODS
Adult male bullfrogs, weighing approximately 600g were supplied from Oouchi (Misato, Japan) during summer and were kept in a laboratory condition for 10–15 days.
Human recombinant activin A and follistatin were provided by Ajinomoto Corporation (Kawasaki, Japan). Mammalian gonadotropin-releasing hormone (mGnRH) and thyrotropin-releasing hormone (TRH) were purchased from Peptide Institute (Osaka, Japan) and Sigma-Aldrich Corporation. (Tokyo, Japan), respectively.
Cell incubation and hormone assay
Preparation of anterior pituitary cells of the bullfrog was performed according to the procedures described elsewhere (Oguchi et al., 1996). The enzymatically dispersed anterior pituitary cells were suspended in 70% Medium 199 (M199; Nissui Pharmaceutical, Tokyo, Japan) containing 0.1% BSA (Fraction V; Sigma, St. Louis, MO). Two hundred microliters containing 60,000 pituitary cells were plated in each well of a 96-multiwell plate (Corning Inc., NY) and incubated at 25°C in a humidified atomosphere of 95% air-5% CO2. After preincubation for 24 hr, the medium was replaced with 70% M199 containing a test substance i.e. activin A or follistatin either singly or in combination. When necessary, hGnRH or TRH was also used in order to secure the responsiveness of pituitary cells. Incubation was continued for 24 hr, since dispersed pituitary cells of the bullfrog are known to release various pituitary hormones continuously during this period (Oguchi et al., 1997; Aida et al., 1997). After incubation, the medium was collected from each well and centrifuged, and the supernatant was subjected to radioimmunoassay (RIA) for bullfrog FSH (Polzonetti-Magni et al., 1998), LH (Oguchi et al., 1997), GH (Kobayashi et al., 1991) and PRL (Yamamoto and Kikuyama, 1982).
The values, given as the mean±SEM, were expressed as ng/104cells and analyzed by using analysis of variance and Duncan's multiple range test. Differences at P<0.05 were considered significant.
Effect of activin A on the release of FSH and LH.
Activin A stimulated the release of FSH from the dispersed pituitary cells dose-dependently in the range of 0.01–10 ng/ml (Fig. 1A). Activin A at the concentration of 10 ng/ml and 10−7M mGnRH, which was employed as a ref rence secreta gogue, were equipotent in terms of stimulating the release of FSH during the 24-hr incubation. On the other hand, activin A in the same dose range did not exert any influence upon the release of LH (Fig. 1B). As shown in Fig. 2, follistatin alone (10ng/ml) did not affect the basal release of either FSH or LH. However, it completely suppressed the activin (1 ng/ml)-induced FSH release at the concentration of 1–10 ng/ml. (Fig. 2A); whereas the combination of activin A and follistatin did not affect the release of LH (Fig. 2B).
Effect of activin A on the release of GH and PRL.
TRH used as a reference secretagogue significantly stimulated the release of both GH and PRL during the 24-hr incubation period. Activin A stimulated the release of GH markedly (Fig. 3A); and PRL, rather moderately (Fig. 3B). In either case, the response to activin A was dose-dependent in the range of 0.01–10 ng/ml. The effect of follistatin on the activin-induced GH and PRL release was then studied. As shown in Fig. 4, follistatin completely suppressed the activin (1 ng/ml)-induced release of both GH and PRL at the dose of 10 ng/ml. When the follistatin concentration was reduced to 1 ng/ml, the inhibitor blocked the release of GH partially but the release of PRL completely. Follistatin did not affect the basal secretion of either GH or PRL at the concentration of 10 ng/ml.
The present experiment indicated that activin A has a stimulatory effect on the release of hormones from multiple types of pituitary cells. We focused on the release of LH, FSH, GH, and PRL because RIA systems for these four hormones of the bullfrog are available in our laboratory. Activin A stimulated the release of FSH but not that of LH. This is in accord with the results obtained in mammals. Activin selectively stimulates FSH secretion but has little or only a slight effect on LH secretion in cultured and perifused rat pituitary cells (Ling et al., 1986b; Vale et al., 1986; Ying et al., 1988; Carroll et al., 1991), although there are reports that activin enhances GnRH-induced LH as well as FSH release from rat (Weiss et al., 1993) and macaque (Zhengwei et al., 1998) pituitaries. In the bullfrog, we observed that activin B stimulates both FSH and LH from dispersed pituitary cells (Uchiyama et al., in press). In the goldfish, Ge et al. (1992) reported that activin A promotes GTH-II secretion. More recently, they found that recombinant goldfish activin B stimulated the expression of GTH-Iβ subunit mRNA but suppressed that of GTH-IIβ mRNA (Yan et al., 1999).
In this experiment, we found that activin A stimulated not only FSH but also GH and PRL release, GH cells being the most sensitive to activin A among LH, GH, and PRL cells. In mammals, there are some reports on the effects of activin on the release of GH and PRL. Incubation of rat anterior pituitary cells with activin A results in a reduction in basal GH secretion (Billestrup et al., 1990). Also, activin A suppresses GH-releasing hormone-induced GH release (Billestrup et al., 1990; Kitaoka et al., 1988). On the other hand, activin A has no effect on basal PRL secretion but restrains TRH-induced PRL release (Kitaoka et al., 1988). In lower vertebrates, only one report on the effect of activin on GH and PRL cell function has been published. In goldfish, activin A stimulated GH release from perifused pituitary fragments (Ge and Peter, 1994). We observed that activin B enhanced the release of GH and PRL from the bullfrog pituitary in vitro (unpublished data). Thus, we can conclude that activins exert diverse effects on the pituitary function, according to the cell types and species.
In lower vertebrates, information about the localization of activins in the pituitary is rather scanty, although genes for the activin/inhibin βA and βB chains have been cloned in Xenopus (Thomsen et al., 1990; Dohrmann et al., 1993) and the goldfish (Ge et al., 1993). According to Ge and Peter (1994), activin/inhibin subunits (βA and βB) do not occur in gonadotrophs but are confined to somatotrophs in the goldfish pituitary. Since the inhibin α-subunit is not expressed in somatotrophs but in other cells, they concluded that somatotrophs may contain activin(s). More recently, we demonstrated that immunoreactive βB exists in thyrotrophs, somatotrophs and gonadotrophs of the Xenopus pituitary (Uchiyama et al., 1996) and in thyrotrophs and gonadotrophs of the bullfrog pituitary (Uchiyama et al., in press). However, further studies are needed to determine which isoforms of activins and inhibins are actually contained in these immunoreactive cells.
We thank Dr. Y. Eto of Ajinomoto Corporation for the gifts of human recombinant activin A and follistatin and Prof. K. Wakabayashi of Gunma University for supplying the goat anti-rabbit IgG serum (HAC-RBA2-05GTP91). Thanks are also extended to Drs. A. Oguchi and T. Aida for their help during the course of the experiments. This work was supported by INSERM-JSPS exchange program to H.V. and S.R. and research grants from Waseda University, Asahi Glass Research Foundation, Uehara Memorial Foundation, and the Ministry of Education, Science, and Culture of Japan to S.K.