Cell Tissue Res. 202, 251-261 (1979)
Cell and Tissue Research 9 by Springer-Verlag 1979
Developmental Correlation Between Hypothalamic Somatostatin and Hypophysial Growth Hormone* Douglas S. Gross and Joan D. Longer Temple University School of Medicine, Department of Anatomy, Philadelphia, and Swarthmore College, Department of Biology, Swarthmore, USA
Summary. The objective of the present study was to determine, by means of immunocytochemistry, the age in fetal development at which G H is first detectable in the pituitary gland and somatostatin in the median eminence, and to correlate temporally the development of these two hormones throughout the remainder of pregnancy. Mice were studied at 15-19 days of gestation with the peroxidase-antiperoxidase (PAP) technique of Sternberger. Somatotropes in the pars distalis were initially detected at 16 days of gestation and by 17 days they were a prominent component of the parenchymal cell population of the hypophysis. These cells were ovoid and distributed uniformly throughout the pars distalis; many were located adjacent to sinusoidal capillaries. Their number and staining intensity increased by 19 days. Somatostatin was not consistently observed in the median eminence until 19 days of gestation. Reaction product indicative of the presence o f somatostatin in presumptive nerve endings was located on the ventral surface of the median eminence and in the external lamina of the infundibulum in proximity to the superficial portal capillaries. Results of the present investigation support the concept that the potential for neuroendocrine control of G H secretion exists in the mouse by the end of fetal development. Several hypotheses concerning the temporal relationship between the appearance of somatostatin in the hypothalamus and of G H in the anterior pituitary gland are discussed. Key words: Somatostatin - Growth hormone - Mouse Fetus - Immunocytochemistry. Although the presence of growth hormone-containing cells (somatotropes, GHcells) in the fetal pituitary has been reported by several investigators (Sano and Dr. Douglas S. Gross, Department of Human Anatomy, University of California School of Medicine, Davis, CA 95616 * Supported by a BiomedicalResearch Support Grant (NIH RR 5417).Appreciation is extended to the National Pituitary Agency, NIAMDD for the followingradioiodination-gradehormones: hGH, rPRL, rTSH, rFSH and hCG Send offprint requests to:
0302-766X/79/0202/0251/$02.20
252
D.S. Gross and J.D. Longer
Sasaki, 1969; Baker and Jaffe, 1975; Dearden and Holmes, 1976; Sdtfil6 and Nakane, 1976), evidence is scarce concerning the ability of the fetal brain to regulate the production of G H . Morphological studies of the hypothalamus and hypophysis led to the proposal that in the fetal rat (Glydon, 1957) and rabbit (Campbell, 1966), neurovascular control of the pituitary by the brain is unlikely, largely because capillary loops of the primary portal vascular plexus do not penetrate the median eminence prenatally. In the adult, a primary capillary bed in the median eminence is connected by portal vessels to a secondary capillary bed in the anterior pituitary. Hypothalamic neurosecretory cells terminate on capillaries of the primary plexus and the neurohormones secreted reach the pituitary gland by way of this portal vascular system (Harris, 1948). Late in gestation, capillaries on the surface of the fetal median eminence become connected with sinusoidal capillaries of the developing pituitary gland by portal vessels (Enemar, 1961; Halfisz et al., 1972; Dearden and Holmes, 1976; Gross and Baker, 1977). Because neurosecretory neurons terminate on the capillaries of this fetal primary plexus (Eurenius and Jarskar, 1971; Beauvillain, 1973), it was proposed that a functional neurovascular link between the fetal hypothalamus and anterior pituitary exists late in gestation. Additional evidence has been presented which suggests that the potential for neuroendocrine control of the pituitary-adrenal (Jost et al., 1970), pituitary-thyroid (Conklin et al., 1973), and pituitary-gonadal (Gross and Baker, 1977, 1978) systems exists late in fetal life. Since growth hormone plays a crucial role in growth and metabolism in the neonate and adult, increased understanding of the regulation of G H prior to birth is desirable. Secretion of G H by the adult pituitary gland is believed to be regulated by both a releasing hormone (growth hormone-releasing factor, G H R F ) and a releaseinhibiting hormone, (somatotropin-release-inhibiting hormone, SRIH) commonly known as somatostatin. In order to suggest that hypothalamic control of G H secretion exists in the fetus it is necessary to demonstrate (a) somatotropes in the pituitary gland and (b) a growth hormone-regulating hormone (or hormones) in the median eminence. Although dual control for G H secretion probably exists, only somatostatin has been isolated and characterized and can be reliably identified in the hypothalamus. Therefore, the purpose of the present investigation was to determine the age in fetal development at which GH-containing cells are first detectable in the pituitary gland and somatostatin in the median eminence with the use of the highly sensitive technique of immunocytochemistry, and to correlate temporally these two hormones throughout the remainder of gestation in the mouse.
Materials and Methods
Swiss-Webster mice, maintained in a breeding colony in air-conditioned quarters with a controlled lighting schedule of 14h light/10 h darkness, were used in this study. Femalesweremated overnight and checked for the presence of vaginal plugs the next morning. If a plug was present, the female was removed to a separate cage and considered to be in day one of pregnancy. The averagegestation period was 19 days. The numbers of fetal mice studied during gestation were9 at 15 days, 10 at 16 days, 10 at 17 days, 5 at 18 days, and 10 at 19 days. The fetal micewere removedfrom the uterus and decapitated. The cranium containing the intact brain and pituitary was immersedin Bouin's fluid for 48 to 72h, and the
Somatostatin and Growth Hormone During Development
253
tissue was then embedded in Paraplast and sectioned serially at 4 ~tm. To aid in the identification of brain regions suitable for immunocytochemical study every twentieth tissue section was stained with cresyl violet. The postinfundibular region of the median eminence was most often examined for the presence of somatostatin because in the adult mouse (Gross, unpublished) and rat (Brownstein et al., 1975; Dub6 et al., 1975! H6kfelt et al., 1975; King et al., 1975; Pelletier et al., 1975), this is the area of its highest concentration. Other regions of the median eminence were also examined, and midcoronal sections through the pituitary were selected to demonstrate the presence of somatotropes.
Immunoeytochemistry Sections were labeled for somatostatin and GH by the unlabeled antibody, peroxidase-antiperoxidase (PAP) technique of Sternberger et al. (1970). Briefly, this involves the sequential application of the following substances to the tissue sections: (1) rabbit antiserum to the hormone in question, in this case either synthetic somatostatin or highly purified human GH; (2) goat antiserum to rabbit IgG; (3) rabbit PAP complex (supplied by L.A. Sternberger); (4) a solution of 3,Y-diaminobenzidine and hydrogen peroxide. All antisera were applied for 30 rain at room temperature. In addition, sections were treated with normal goat serum before the application of anti-somatostatin or anti-GH, anti-IgG, and PAP, as described by Petrali et al. (1974) to reduce nonspecific staining. The somatostatin antiserum (anti-SRIH B 173) was generated against the cyclic form of the hormone and used at a dilution of 1/50. The growth hormone antiserum (anti-hGH B85) was generated against highly purified human GH and used at a dilution of 1/200. Both antisera were provided by B.L. Baker of the University of Michigan. The dilutions employed were those that kept background staining at a minimum while still producing maximal specific staining intensity. The preparation and specificity of these antisera for the demonstration of somatostatin and GH in adult animals has been described previously (Baker and Yu, 1976; Baker and Gross, 1978, respectively). In order to confirm the specificity of the irmnunocytochemical technique for localization of GH and somatostatin in fetal tissues several control procedures were performed. Immunostaining was eliminated by preabsorption of anti-hGH B85 with 5ng/~tl hGH, while preabsorption of anti-hGH B85 with 500ng/lal of radioiodination grade rPRL, rTSH, rFSH, hGG or fll-Z4ACTH had no effect on immunocytochemical staining intensity. Similarly, preabsorption of anti-SRIH B173 with 5ng/lal synthetic somatostatin prevented all staining, while preabsorption of anti-SRIH B173 with up to 500ngAd synthetic arginine-vasopressin, oxytocin, thyrotropin-releasing hormone (TRH) or gonadotropin-releasing hormone (GnRH) was without effect. Also, when normal rabbit serum was applied in place of either anti-hGH B85 or anti-SRIH B173, no immunostaining was observed. In order to prevent false negative observations due to failure of the immunocytochemical technique, an adult pituitary or hypothalamic section known to contain either GH or somatostatin was included as a control each time a set of fetal tissue sections was stained. The presence of GH and somatostatin was consistently demonstrated in the adult control tissues (Figs. 1a, 6b).
Results Growth Hormone I n the a d u l t p i t u i t a r y g l a n d s o m a t o t r o p e s are d i s t r i b u t e d u n i f o r m l y t h r o u g h o u t the p a r s distalis. T h e y a r e o v o i d w i t h d e n s e l y - s t a i n e d c y t o p l a s m a n d lightly, o r u n s t a i n e d n u c l e i (Fig. 1 a). T h e specificity o f G H l o c a l i z a t i o n in t h e s e cells w a s d e m o n s t r a t e d by t h e l a c k o f i m m u n o s t a i n i n g in a d j a c e n t s e c t i o n s t r e a t e d w i t h a n t i h G H p r e a b s o r b e d w i t h h G H (Fig. I b). N o s o m a t o t r o p e s w e r e o b s e r v e d at 15 d a y s o f g e s t a t i o n (Fig. 2). N o n - s p e c i f i c s t a i n i n g was s e e n in b l o o d cells in s i n u s o i d a l c a p i l l a r i e s o f t h e p a r s distalis at this stage. T h e s e cells w e r e also l a b e l e d in t h e c o n t r o l sections. It s h o u l d be n o t e d t h a t t h e s e c a p i l l a r i e s w e r e s m a l l a n d f e w in n u m b e r at 15 days.
Figs. 1-5. are of sections labeled immunocytochemically with anti-hGH B 85, Figs. 6-10. of sections labeled with anti-somatostatin B 173. Abbreviations." HG hypophysial cleft; M E median eminence; PC portal capillary; P1 pars intermedia; V third ventricle Fig. 1. a Region of pars distalis of adult mouse. Note ovoid somatotropes, with densely stained cytoplasm and unstained nuclei. Several large basophils unstained (arrows). • 600. b Control section adjacent to that shown in Fig. I a, labeled with anti-hGH preabsorbed with 5 ng/~tl purified hGH. Note absence of immunostained somatotropes. Cells indicated in Fig. 2a are seen (arrows). x 600 Fig. 2. Pars distalis, 15 days of gestration. Note nonspecific staining of blood cells in sinusoidal capillary
(arrow), and absence of reaction product in parenchymal cells, x 600 Fig. 3. Pars distalis, 16 days of gestation. Note several cells (arrows) with slightly more reaction product than in general background, demonstrating initial appearance of GH. x 600
Fig. 4. Pars distalis, 17 days of gestation. Note several somatotropes abutting on sinusoidal capillaries, others deeply within cell clusters (arrow). • 600 Fig. 5. Pars distalis, 19 days of gestation. Note increase in number of somatotropes. Several large densely-stained cells present (arrows). x 600 Fig. 6. a Adult median eminence. Somatostatin in external lamina, in proximity to hypophysial portal capillaries located between ventral surface o f median eminence and pars tuberalis. • 140. b Control section adjacent to that shown in Fig. 6a, labeled with anti-SRIH preabsorbed with 5 ng/~tl synthetic somatostatin. Immunostaining completely prevented. • 140 Fig. 7. Postinfundibular region of median eminence, 16 days of gestation. No somatostatin detectable. • 240
Fig. 8. a Postinfundibular median eminence, 17 days of gestation. Note small amount of somatostatin near ventral surface (arrow). x240. b Control section adjacent to that shown in Fig. 8a, labeled with anti-SRIH preabsorbed with somatostatin, Note absence of reaction deposit in median eminence (arrow). x 240 Fig. 9. Caudal region of median eminence, 19 days of gestation. Note accumulation of somatostatin in external lamina of infundibulum and at ventral surface of postinfundibular median eminence (arrows). Distribution of somatostatin similar to that of adult (see Fig. 6a). x 240 Fig. 10. Ventral surface of postinfundibular median eminence, 19 days of gestation. Somatostatin detected (arrow) near ventral surface of median eminence, in close proximity to superficial primary plexus of hypothalamo-hypophysial portal capillaries. • 600
Somatostatin and Growth Hormone During Development
257
At 16 days a small amount of growth hormone was identified in a few scattered cells of the pars distalis of all fetuses, as indicated by an increase in the staining intensity of these cells (Fig. 3). Such a staining pattern was absent in sections labeled with antiserum preabsorbed with hGH. Although the difference between 15- and 16-day fetuses was not dramatic, a slight increase in staining was judged present in every 16 day fetus studied. By the 17th day of gestation somatotropes were readily identified (Fig. 4). Numerous well-stained cells were observed throughout the pars distalis. They were distributed widely although in several specimens fewer such cells were present in the anteromedian wedge. The staining intensity of the cells in the lateral lobes of the pars distalis often appeared greater than that of the more medially located cells. As in the adult, somatotropes are small ovoid cells possessing a rim of darkly stained hormone-containing cytoplasm around a light nucleus (Fig. 4). Approximately one-half of the GH-cells could be seen abutting on sinusoidal capillaries. The pituitary gland of the 17 day-old fetus contained more prominent sinusoidal capillaries than at 15 days. The distribution of somatotropes at 18 days was similar to that observed at 17 days. There was a small increase in the number and staining intensity of labeled cells. By 19 days of gestation the pituitary gland had become flatter compared to the previous day and it exhibited an increased number of somatotropes (Fig. 5). Several of these cells stain more darkly and possess a greater amount of hormonecontaining cytoplasm than those seen at 17 days (Fig. 5). In all of the fetuses studied there was no apparent difference in number, size, or distribution of GH-cells between the sexes.
Somatostatin
In the adult hypothalamus somatostatin was found in axons and nerve endings in the external lamina of the median eminence, in proximity to the hypophysial portal capillaries (Fig. 6a). The absence of immunostaining in adjacent sections treated with anti-SRIH preabsorbed with synthetic somatostatin indicated that the immunocytochemical localization of this hormone was specific (Fig. 6b). Somatostatin was not detected in the median eminence in any of the fetuses studied at either 15 or 16 days of gestation (Fig. 7). At 17 days somatostatin was observed in only 2 of the 10 animals studied. The amount detected was extremely small and the foci containing somatostatin were located bilaterally on the ventral surface of the postinfundibular median eminence (Fig. 8 a). Characteristically, the reaction deposit indicative ofsomatostatin in presumptive axons and nerve endings was more diffuse than that seen when G n R H was immunostained in the fetal median eminence (Gross and Baker, 1977), Because the amount of somatostatin detected in fetal tissues was so small, control procedures were especially important to demonstrate specificity. Sections labeled with antisera preabsorbed with somatostatin always exhibited a lack of staining (Fig. 8b). S omatostatin was detected in 2 of the 5 fetuses studied at 18 days of gestation. It was present dorsal to the tuberoinfundibular sulci and along the ventral surface of the median eminence, similar to its localization at 17 days.
258
D.S. Gross and J.D. Longer
At 19 days of gestation somatostatin was observed in 8 of the 10 animals studied. It was located over the tuberoinfundibutar sulci, at the ventral surface of the postinfundibular median eminence and in a diffuse band in the external lamina of the infundibulum (Fig. 9). There is considerable variation in the quantity of somatostatin detected in the 19-day fetuses, but the amount is consistently greater than that seen at the earlier ages studied. Immunoreactive deposits were often identified at the ventral surface of the median eminence, adjacent to the superficial, or supratuberal, plexus of capillaries forming the primary plexus of the fetal hypothalamo-hypophysial portal system (Fig. 10). Somatostatin was not detected in neuronal cell bodies of any fetus examined.
Discussion
In this study, GH was first detected in the hypophysial pars distalis at 16 and was prominent by 17 days of gestation in the mouse. This is in general agreement with the results of other investigators. Blazquez et al. (1974) and Rieutort (1974), using radioimmunoassays, detected G H in the pituitary and plasma of rat fetuses at 19 days of pregnancy. Using immunocytochemistry, S6t/tl6 and Nakane (1976) observed somatotropes at 18 days in the fetal rat. In the mouse, somatotropes have been identified on the basis of their fine structural characteristics at 19 days of gestation (Sano and Sasaki, 1969; Dearden and Holmes, 1976). The present study is the first report in which somatotropes are identified in the fetal mouse pituitary gland by means of immunocytochemistry. The PAP technique is one of the most sensitive methods of detecting hormones in tissue (Petrali et al., 1974), and this sensitivity may explain the earlier detection of GH in our study. The apparent concentration of GH in the fetal pars distalis increases from 16 to 19 days, but there is still considerably less than in the adult. Another difference between the fetal and adult pituitary is that the adult shows sex-related differences in the somatotropes, with the male pituitary containing more and larger cells than the female (Baker and Gross, 1978). This distinction does not exist in the fetal tissues, perhaps because gonadal steroids of fetal origin are not yet circulating in sufficient quantity to affect the pituitary, or because steroid receptors on somatotropes are not yet functional. It is also possible that a high concentration of maternal estrogen masks the effects of small quantities of sex hormones produced by the fetal gonads. This investigation is the first to study the appearance of somatostatin in the fetus of a laboratory animal. Our results indicate that somatostatin is present in the median eminence of some mice as early as 17 days of gestation, but is not consistently detected until 19 days. In the adult rat, somatostatin appears to be constained in intra-axonal granules (Pelletier et al., 1977). The appearance of somatostatin in the median eminence of the mouse fetus can be correlated with the development of nerve endings containing small dense-core granules. These nerve endings first come in contact with the perivascular space surrounding capillaries of the supratuberal plexus at 16 (Eurenius and Jarskar, 1971) or 18 (Beauvillain, 1973) days of gestation. The proximity of somatostatin-containing elements in the ventral median eminence to the developing portal vascular system suggests that this hormone may gain access to the developing pituitary gland during gestation. In its
Somatostatinand Growth HormoneDuring Development
259
ventral position, somatostatin is in close contact with the superficial capillaries of the developing supratuberal plexus which is present in the mouse as early as 16 days of gestation (Dearden and Holmes, 1976). This plexus is connected to sinusoidal capillaries in the pars distalis by portal veins at either 16 (Enemar, 1961) or 17 (Dearden and Holmes, 1976) days of gestation. This morphological evidence suggests that somatostatin may affect pituitary production and/or secretion of GH at this stage in development. Many fetal GH-cells were observed abutting on sinusoidal capillaries where they may be affected by somatostatin in portal blood. The appearence of somatostatin in the median eminence coincides with the appearance of monoamines in axons in this brain region at 17 days of gestation in the mouse (Bjrrklund et al., 1968). It has been reported that such biogenic amines exert control over GH secretion in both the fetal and adult rat, but whether such amines act directly on the pituitary or exert effects via the hypothalamus has not been determined (Stuart et al., 1976). Certain hypotheses can be offered concerning the relationship between somatostatin, which is not detectable consistently in the median eminence until 19 days of gestation, and GH, which appears in the anterior pituitary gland as early as 16 days of gestation. The possibility exists that a releasing hormone, present in the median eminence prior to the 19th day of gestation, could have access to the developing pituitary gland by way of an intact hypothalamo-hypophysial portal system or by diffusion. This putative growth hormone-releasing factor (GHRF) could affect GH production and perhaps differentiation of somatotropes before somatostatin exerts its effect. A similar hypothesis has been suggested by Kaplan et al. (1976) concerning the ontogenesis of somatostatin in the human fetus. They postulated that early in gestation GH production is controlled by a releasing hormone because GH concentration in the pituitary and plasma of anencephalic fetuses is low. Rieutort and Jost (1976) reported similar findings in anencephalic rat fetuses. In normal human fetuses Kaplan et al. (1976) found that serum GH levels decrease as gestation progresses and they suggested that this could be due to increased somatostatin secretion, decreased releasing hormone secretion, or both. Confirmation of this hypothesis awaits identification of the putative GHRF, and development of techniques for its measurement. Another hypothesis which fits our results is that somatotropes may develop in the fetal pituitary before the influences of regulating hormones from the hypothalamus are exerted. This could explain the appearance of somatostatin in the median eminence later than the appearance of GH in the pituitary. Experimental evidence exists concerning the ability of the pituitary gland to undergo partial differentiation independent of hypothalamic influences (Daikoku et al., 1973; Nemesk~ry et al., 1976; Srt~16 and Nakane, 1976). It must be noted that, although the immunocytochemical technique is able to indicate the presence of hormones in tissue, it can give only limited information about their release. Nonetheless, the presence of somatostatin in the median eminence, of GH in the anterior pituitary, and of an intact portal vascular system is consistent with the assumption that these hormones are being secreted and that this neuroendocrine axis is functional during fetal development. Although it has not yet been tested on fetal rat or mouse pituitaries, exogenous somatostatin has been found to decrease the secretion of GH by human fetal pituitaries in vitro (Goodyer
260
D.S. Gross and J.D. Longer
et al., 1977). Gross and Baker (1979) have found that another neuroendocrine system is also intact at this time. They determined that gonadotropin-releasing hormone and luteinizing hormone both appear in the mouse at 17 days of gestation, suggesting the potential for hypothalamo-pituitary-gonadal function in the fetus. However, for the somatostatin-GH axis to be functional, there must be evidence for effects of GH on fetal growth and/or metabolism. The literature, however, contains conflicting reports on the role of growth hormone in fetal growth, due, in part, to species variation 0ost, 1954, 1962; Eguchi, 1961; Heggestad and Wells, 1965; Liggins and Kennedy, 1968). However, several investigators agree that GH affects many metabolic processes in prenatal animals (Jost, 1966; Blazquez, 1974; Honnebier and Swaab, 1974). Thus, although the function of GH in fetal growth is a subject of controversy, GH does appear to play a role in fetal development. This fact, along with the data obtained from the present study, suggests that a functional hypothalamohypophysial system for GI-I secretion exists during fetal development in the mouse. References Baker, B.L., Gross, D.S.: Cytology and distribution of secretory cell types in the murine hypophysis as demonstrated by immunocytochemistry. Am. J. Anat. 153, 193-216 (1978) Baker, B.L., Jaffe, R.B.: The genesis of cell types in the adenohypophysis of the human fetus as observed with immunocytochemistry. Am. J. Anat. 143, 137-162 (1975) Baker, B.L., Yu, Y.Y.: The influence of hypophysectomy of the stores of somatostatin in the hypothalamus and pituitary stem. Proc. Soc. Exp. Biol. Med. 151, 599-602 (1976) Beauvillain, J.C.: Structure fine de l'bminence m~diane de souris au cours de son ontog~n6se. Z. Zellforsch. 139, 201 215 (1973) Bj6rklund, A., Enemar, A., Falck, B.: Monoamines in the hypothalamo-hypophysial system of the mouse with special reference to the ontogenetic aspects. Z. Zellforsch. 89, 590-607 (1968) Bl~izquez, E., Simon, F.A., Blb,zquez, M., Fo/~, P.P.: Changes in serum growth hormone levels from fetal to adult age in the rat. Proc. Soc. Exp. Biol. Med. 147, 780-783 (1974) Brownstein, M., Arimura, A., Sato, H., Schally, A.V., Kizer, J.S.: The regional distribution of somatostatin in the rat brain. Endocrinology 96, 1456-1461 (1975) Campbell, H.J.: The development of the primary portal plexus in the median eminence of the rabbit. J. Anat. 100, 381-387 (1966) Conklin, P.M., Schindler, W.J., HoU, S.F.: Hypothalamic thyrotropin releasing factor and pituitary responsiveness during development in the rat. Neuroendocrinology 2, 304-314 (1973) Daikoku, S., Kinutani, M., Watanabe, Y.G.: Role of the hypothalamus on development of the hypophysis: an EM study. Neuroendocrinology 11, 284-305 (1973) Dearden, N.M., Holmes, R.L.: Cyto-differentiation and portal vascular development in the mouse adenohypophysis. J. Anat. 121, 551-569 (1976) Dub6, D., Leclerc, R., Pelletier, G., Arimura, A., Schally, A.V.: Immunohistochemical detection of growth hormone-release inhibiting hormone (somatostatin) in the guinea pig brain. Cell Tissue Res. 161, 385-390 (1975) Eguchi, Y.: Atrophy of the fetal mouse adrenal following decapitation in utero. Endocrinology 68, 716719 (1961) Enemar, A.: The structure and development of the hypophysial portal system in the laboratory mouse, with particular regard to the primary plexus. Ark. Zool., II. Set. 13, 203-252 (1961) Eurenius, L., Jarskar, R.: Electron microscopic studies on the development of the external zone of the mouse median eminence. Z. Zellforsch. 122, 488-502 (1971) Glydon, R.J.: The development of the blood supply of the pituitary in the albino rat, with special reference to the portal vessels. J. Anat. 91, 237-244 (1957) Goodyer, C.G., St. George Hall, C., Gudya, H., Robert, F., Giroud, C.J.P.: Human fetal pituitary in culture: hormone secretion and response to somatostatin, LHRH, TRH, and dbcAMP. J. Clin. Endocrinol. Metab. 45, 73-84 (1977)
Somatostatin and Growth Hormone During Development
261
Gross, D.S., Baker, B.L.: Immunohistochemical localization of gonadotropin-releasing hormone (GnRH) in the fetal and early postnatal mouse brain. Am. J. Anat. 148, 195-215 (1977) Gross, D.S., Baker, B.L. : Developmental correlation between hypothalamic gonadotropin-releasing hormone and hypophysial luteinizing hormone. Am. J. Anat. 154, 1-10 (1979) Hal~isz, B., Kosaras, B., Lengvari, I.: Ontogenesis of the neurovascular link between the hypothalamus and the anterior pituitary in the rat. In: Median Eminence: Structure and Function. (K.M. Knigge, D.E. Scott, A. Weindl, eds.). Basel: S. Karger (1972) Harris, G.W.: Neural control of the pituitary gland. Physiol. Rev. 28, 139-179 (1948) Heggestad, C.B., Wells, L.J.: Experiments on the contribution of somatotrophin to prenatal growth in the rat. Acta Anat. 60, 348-361 (1965) Hrkfelt, T.M., Efendic, S., Hellerstrom, C., Johansson, O., Luft, R., Arimura, A.; Cellular localization of somatostatin in endocrine-like cells and neurons of the rat. Acta Endocrinol. 80, Supp 100 (1975) Honnebier, WJ., Swaab, D.F.: Influence of ~-melanocyte-stimulating hormone (c~-MSH), growth hormone (GH), and fetal extracts in intrauterine growth of fetus and placenta in the rat. J. Obstet. Gynecol. Brit. Comm. 81, 439-447 (1974) Jost, A.: Hormones in the development of the foetus. Cold Spring Harbor Symp. Quant. Biol. 19, 167181 (1954) Jost, A.: Endocrine factors of foetal development. Triange 5, 189-193 (1962) Jost, A.: Anterior pituitary function in fetal life. In: The Pituitary Gland. Vol. II. (G.W. Harrison, and B.T. Donovan,eds.) Berkeley: University of California Press 1966 Jost, A., Dupouy, J.P., Geloso-Meyer, A.: Hypothalamo-hypohysial relationships in the fetus. In: The Hypothalamus. (L. Martini, M. Motta, and F. Fraschini, eds.) New York: Academic Press 1970 Kaplan, S.L., Grumbach, M.M., Aubert, M.L.: The ontogenesis of pituitary hormones and hypothalamic factors in the human fetus; maturation of central nervous system regulation of anterior pituitary function. Recent Prog. Horm. Res. 32, 161-234 (1976) King, J.C., Arimura, A., Gerall, A.A., Fishback, J.B., Elkind, K.E.: Growth-hormone release inhibiting hormone (GHIH) pathway of the rat hypothalamus revealed by the unlabeled antibody peroxidaseantiperoxidase method. Cell Tissue Res. 160, 423-427 (1975) Liggins, G.C., Kennedy, P.G.: Effects of electrocoagulation of the fetal lamb hypophysis in growth and development. J. Endocrinol. 40, 371-381 (1968) Nemeskrry, A., Nrmeth, A., Srt~il6, G., Vigh, S., Halfisz, B.: Cell differentiation in the fetal rat anterior pituitary in vitro. Cell Tissue Res. 170, 263-273 (1976) Pelletier, G., Leclerc, R., Dubr, D., Labrie, F., Puviani, R., Arimura, A., Schally, A.V.: Localization of growth hormone-release-inhibiting hormone (somatostatin) in rat brain. Am. J. Anat. 142, 397-400 (1975) Pelletier, G., Dubr, D., Puviani, R.: Somatostatin: electron microscope immunohistochemical localization in secretory neurons of rat hypothalamus. Science 196, 1469-1470 (1977) Petrali, J.P., Hinton, D.M., Moriarty, G.C., Sternberger, L.A.: The unlabeled antibody enzyme method of immunocytochemistry. Quantitative comparison of sensitivities with and without peroxidaseantiperoxidase complex. J. Histochem. Cytochem. 22, 782-801 (1974) Rieutort, M. : Pituitary content and plasma levels of GH in fetal and weanling rats. J. Endocrinol. 60, 261-268 (1974) Rieutort, M., Jost, A.: Growth hormone in encephalectomized rat fetuses with comments on the effects of anesthetics. Endocrinology 98, 1123-1129 (1976) Sano, M., Sasaki, F.: Embryonic development of the mouse anterior pituitary studied by electron microscopy. Z. Anat. Entwickl. 129, 195-222 (1969) Srtfil6, B., Nakane, P.K.: Functional differentiation of the fetal anterior pituitary cells in the rat. Endocrinol. Exp. 10, 155-166 (1976) Sternberger, L.A., Hardy, Jr., P.H., Cuculis, J.J., Meyer, H.B.: The unlabeled antibody enzyme method of immunohistochemistry. Preparation and properties of soluble antigen-antibody complex (horseradish peroxidase-antihorseradish peroxidase) and its use in identification of spirochetes. J. Histochem. Cytochem. 18, 315-333 (1970) Stuart, M., Lazarus, L., Smythe, GA., Moore, S., Sara, V.: Biogenic amine control of growth hormone secretion in the fetal and neonatal rat. Neuroendocrinology 22, 337-342 (1976)
Accepted May 29, 1979
Cell and Tissue Research 9 by Springer-Verlag 1979
Developmental Correlation Between Hypothalamic Somatostatin and Hypophysial Growth Hormone* Douglas S. Gross and Joan D. Longer Temple University School of Medicine, Department of Anatomy, Philadelphia, and Swarthmore College, Department of Biology, Swarthmore, USA
Summary. The objective of the present study was to determine, by means of immunocytochemistry, the age in fetal development at which G H is first detectable in the pituitary gland and somatostatin in the median eminence, and to correlate temporally the development of these two hormones throughout the remainder of pregnancy. Mice were studied at 15-19 days of gestation with the peroxidase-antiperoxidase (PAP) technique of Sternberger. Somatotropes in the pars distalis were initially detected at 16 days of gestation and by 17 days they were a prominent component of the parenchymal cell population of the hypophysis. These cells were ovoid and distributed uniformly throughout the pars distalis; many were located adjacent to sinusoidal capillaries. Their number and staining intensity increased by 19 days. Somatostatin was not consistently observed in the median eminence until 19 days of gestation. Reaction product indicative of the presence o f somatostatin in presumptive nerve endings was located on the ventral surface of the median eminence and in the external lamina of the infundibulum in proximity to the superficial portal capillaries. Results of the present investigation support the concept that the potential for neuroendocrine control of G H secretion exists in the mouse by the end of fetal development. Several hypotheses concerning the temporal relationship between the appearance of somatostatin in the hypothalamus and of G H in the anterior pituitary gland are discussed. Key words: Somatostatin - Growth hormone - Mouse Fetus - Immunocytochemistry. Although the presence of growth hormone-containing cells (somatotropes, GHcells) in the fetal pituitary has been reported by several investigators (Sano and Dr. Douglas S. Gross, Department of Human Anatomy, University of California School of Medicine, Davis, CA 95616 * Supported by a BiomedicalResearch Support Grant (NIH RR 5417).Appreciation is extended to the National Pituitary Agency, NIAMDD for the followingradioiodination-gradehormones: hGH, rPRL, rTSH, rFSH and hCG Send offprint requests to:
0302-766X/79/0202/0251/$02.20
252
D.S. Gross and J.D. Longer
Sasaki, 1969; Baker and Jaffe, 1975; Dearden and Holmes, 1976; Sdtfil6 and Nakane, 1976), evidence is scarce concerning the ability of the fetal brain to regulate the production of G H . Morphological studies of the hypothalamus and hypophysis led to the proposal that in the fetal rat (Glydon, 1957) and rabbit (Campbell, 1966), neurovascular control of the pituitary by the brain is unlikely, largely because capillary loops of the primary portal vascular plexus do not penetrate the median eminence prenatally. In the adult, a primary capillary bed in the median eminence is connected by portal vessels to a secondary capillary bed in the anterior pituitary. Hypothalamic neurosecretory cells terminate on capillaries of the primary plexus and the neurohormones secreted reach the pituitary gland by way of this portal vascular system (Harris, 1948). Late in gestation, capillaries on the surface of the fetal median eminence become connected with sinusoidal capillaries of the developing pituitary gland by portal vessels (Enemar, 1961; Halfisz et al., 1972; Dearden and Holmes, 1976; Gross and Baker, 1977). Because neurosecretory neurons terminate on the capillaries of this fetal primary plexus (Eurenius and Jarskar, 1971; Beauvillain, 1973), it was proposed that a functional neurovascular link between the fetal hypothalamus and anterior pituitary exists late in gestation. Additional evidence has been presented which suggests that the potential for neuroendocrine control of the pituitary-adrenal (Jost et al., 1970), pituitary-thyroid (Conklin et al., 1973), and pituitary-gonadal (Gross and Baker, 1977, 1978) systems exists late in fetal life. Since growth hormone plays a crucial role in growth and metabolism in the neonate and adult, increased understanding of the regulation of G H prior to birth is desirable. Secretion of G H by the adult pituitary gland is believed to be regulated by both a releasing hormone (growth hormone-releasing factor, G H R F ) and a releaseinhibiting hormone, (somatotropin-release-inhibiting hormone, SRIH) commonly known as somatostatin. In order to suggest that hypothalamic control of G H secretion exists in the fetus it is necessary to demonstrate (a) somatotropes in the pituitary gland and (b) a growth hormone-regulating hormone (or hormones) in the median eminence. Although dual control for G H secretion probably exists, only somatostatin has been isolated and characterized and can be reliably identified in the hypothalamus. Therefore, the purpose of the present investigation was to determine the age in fetal development at which GH-containing cells are first detectable in the pituitary gland and somatostatin in the median eminence with the use of the highly sensitive technique of immunocytochemistry, and to correlate temporally these two hormones throughout the remainder of gestation in the mouse.
Materials and Methods
Swiss-Webster mice, maintained in a breeding colony in air-conditioned quarters with a controlled lighting schedule of 14h light/10 h darkness, were used in this study. Femalesweremated overnight and checked for the presence of vaginal plugs the next morning. If a plug was present, the female was removed to a separate cage and considered to be in day one of pregnancy. The averagegestation period was 19 days. The numbers of fetal mice studied during gestation were9 at 15 days, 10 at 16 days, 10 at 17 days, 5 at 18 days, and 10 at 19 days. The fetal micewere removedfrom the uterus and decapitated. The cranium containing the intact brain and pituitary was immersedin Bouin's fluid for 48 to 72h, and the
Somatostatin and Growth Hormone During Development
253
tissue was then embedded in Paraplast and sectioned serially at 4 ~tm. To aid in the identification of brain regions suitable for immunocytochemical study every twentieth tissue section was stained with cresyl violet. The postinfundibular region of the median eminence was most often examined for the presence of somatostatin because in the adult mouse (Gross, unpublished) and rat (Brownstein et al., 1975; Dub6 et al., 1975! H6kfelt et al., 1975; King et al., 1975; Pelletier et al., 1975), this is the area of its highest concentration. Other regions of the median eminence were also examined, and midcoronal sections through the pituitary were selected to demonstrate the presence of somatotropes.
Immunoeytochemistry Sections were labeled for somatostatin and GH by the unlabeled antibody, peroxidase-antiperoxidase (PAP) technique of Sternberger et al. (1970). Briefly, this involves the sequential application of the following substances to the tissue sections: (1) rabbit antiserum to the hormone in question, in this case either synthetic somatostatin or highly purified human GH; (2) goat antiserum to rabbit IgG; (3) rabbit PAP complex (supplied by L.A. Sternberger); (4) a solution of 3,Y-diaminobenzidine and hydrogen peroxide. All antisera were applied for 30 rain at room temperature. In addition, sections were treated with normal goat serum before the application of anti-somatostatin or anti-GH, anti-IgG, and PAP, as described by Petrali et al. (1974) to reduce nonspecific staining. The somatostatin antiserum (anti-SRIH B 173) was generated against the cyclic form of the hormone and used at a dilution of 1/50. The growth hormone antiserum (anti-hGH B85) was generated against highly purified human GH and used at a dilution of 1/200. Both antisera were provided by B.L. Baker of the University of Michigan. The dilutions employed were those that kept background staining at a minimum while still producing maximal specific staining intensity. The preparation and specificity of these antisera for the demonstration of somatostatin and GH in adult animals has been described previously (Baker and Yu, 1976; Baker and Gross, 1978, respectively). In order to confirm the specificity of the irmnunocytochemical technique for localization of GH and somatostatin in fetal tissues several control procedures were performed. Immunostaining was eliminated by preabsorption of anti-hGH B85 with 5ng/~tl hGH, while preabsorption of anti-hGH B85 with 500ng/lal of radioiodination grade rPRL, rTSH, rFSH, hGG or fll-Z4ACTH had no effect on immunocytochemical staining intensity. Similarly, preabsorption of anti-SRIH B173 with 5ng/lal synthetic somatostatin prevented all staining, while preabsorption of anti-SRIH B173 with up to 500ngAd synthetic arginine-vasopressin, oxytocin, thyrotropin-releasing hormone (TRH) or gonadotropin-releasing hormone (GnRH) was without effect. Also, when normal rabbit serum was applied in place of either anti-hGH B85 or anti-SRIH B173, no immunostaining was observed. In order to prevent false negative observations due to failure of the immunocytochemical technique, an adult pituitary or hypothalamic section known to contain either GH or somatostatin was included as a control each time a set of fetal tissue sections was stained. The presence of GH and somatostatin was consistently demonstrated in the adult control tissues (Figs. 1a, 6b).
Results Growth Hormone I n the a d u l t p i t u i t a r y g l a n d s o m a t o t r o p e s are d i s t r i b u t e d u n i f o r m l y t h r o u g h o u t the p a r s distalis. T h e y a r e o v o i d w i t h d e n s e l y - s t a i n e d c y t o p l a s m a n d lightly, o r u n s t a i n e d n u c l e i (Fig. 1 a). T h e specificity o f G H l o c a l i z a t i o n in t h e s e cells w a s d e m o n s t r a t e d by t h e l a c k o f i m m u n o s t a i n i n g in a d j a c e n t s e c t i o n s t r e a t e d w i t h a n t i h G H p r e a b s o r b e d w i t h h G H (Fig. I b). N o s o m a t o t r o p e s w e r e o b s e r v e d at 15 d a y s o f g e s t a t i o n (Fig. 2). N o n - s p e c i f i c s t a i n i n g was s e e n in b l o o d cells in s i n u s o i d a l c a p i l l a r i e s o f t h e p a r s distalis at this stage. T h e s e cells w e r e also l a b e l e d in t h e c o n t r o l sections. It s h o u l d be n o t e d t h a t t h e s e c a p i l l a r i e s w e r e s m a l l a n d f e w in n u m b e r at 15 days.
Figs. 1-5. are of sections labeled immunocytochemically with anti-hGH B 85, Figs. 6-10. of sections labeled with anti-somatostatin B 173. Abbreviations." HG hypophysial cleft; M E median eminence; PC portal capillary; P1 pars intermedia; V third ventricle Fig. 1. a Region of pars distalis of adult mouse. Note ovoid somatotropes, with densely stained cytoplasm and unstained nuclei. Several large basophils unstained (arrows). • 600. b Control section adjacent to that shown in Fig. I a, labeled with anti-hGH preabsorbed with 5 ng/~tl purified hGH. Note absence of immunostained somatotropes. Cells indicated in Fig. 2a are seen (arrows). x 600 Fig. 2. Pars distalis, 15 days of gestration. Note nonspecific staining of blood cells in sinusoidal capillary
(arrow), and absence of reaction product in parenchymal cells, x 600 Fig. 3. Pars distalis, 16 days of gestation. Note several cells (arrows) with slightly more reaction product than in general background, demonstrating initial appearance of GH. x 600
Fig. 4. Pars distalis, 17 days of gestation. Note several somatotropes abutting on sinusoidal capillaries, others deeply within cell clusters (arrow). • 600 Fig. 5. Pars distalis, 19 days of gestation. Note increase in number of somatotropes. Several large densely-stained cells present (arrows). x 600 Fig. 6. a Adult median eminence. Somatostatin in external lamina, in proximity to hypophysial portal capillaries located between ventral surface o f median eminence and pars tuberalis. • 140. b Control section adjacent to that shown in Fig. 6a, labeled with anti-SRIH preabsorbed with 5 ng/~tl synthetic somatostatin. Immunostaining completely prevented. • 140 Fig. 7. Postinfundibular region of median eminence, 16 days of gestation. No somatostatin detectable. • 240
Fig. 8. a Postinfundibular median eminence, 17 days of gestation. Note small amount of somatostatin near ventral surface (arrow). x240. b Control section adjacent to that shown in Fig. 8a, labeled with anti-SRIH preabsorbed with somatostatin, Note absence of reaction deposit in median eminence (arrow). x 240 Fig. 9. Caudal region of median eminence, 19 days of gestation. Note accumulation of somatostatin in external lamina of infundibulum and at ventral surface of postinfundibular median eminence (arrows). Distribution of somatostatin similar to that of adult (see Fig. 6a). x 240 Fig. 10. Ventral surface of postinfundibular median eminence, 19 days of gestation. Somatostatin detected (arrow) near ventral surface of median eminence, in close proximity to superficial primary plexus of hypothalamo-hypophysial portal capillaries. • 600
Somatostatin and Growth Hormone During Development
257
At 16 days a small amount of growth hormone was identified in a few scattered cells of the pars distalis of all fetuses, as indicated by an increase in the staining intensity of these cells (Fig. 3). Such a staining pattern was absent in sections labeled with antiserum preabsorbed with hGH. Although the difference between 15- and 16-day fetuses was not dramatic, a slight increase in staining was judged present in every 16 day fetus studied. By the 17th day of gestation somatotropes were readily identified (Fig. 4). Numerous well-stained cells were observed throughout the pars distalis. They were distributed widely although in several specimens fewer such cells were present in the anteromedian wedge. The staining intensity of the cells in the lateral lobes of the pars distalis often appeared greater than that of the more medially located cells. As in the adult, somatotropes are small ovoid cells possessing a rim of darkly stained hormone-containing cytoplasm around a light nucleus (Fig. 4). Approximately one-half of the GH-cells could be seen abutting on sinusoidal capillaries. The pituitary gland of the 17 day-old fetus contained more prominent sinusoidal capillaries than at 15 days. The distribution of somatotropes at 18 days was similar to that observed at 17 days. There was a small increase in the number and staining intensity of labeled cells. By 19 days of gestation the pituitary gland had become flatter compared to the previous day and it exhibited an increased number of somatotropes (Fig. 5). Several of these cells stain more darkly and possess a greater amount of hormonecontaining cytoplasm than those seen at 17 days (Fig. 5). In all of the fetuses studied there was no apparent difference in number, size, or distribution of GH-cells between the sexes.
Somatostatin
In the adult hypothalamus somatostatin was found in axons and nerve endings in the external lamina of the median eminence, in proximity to the hypophysial portal capillaries (Fig. 6a). The absence of immunostaining in adjacent sections treated with anti-SRIH preabsorbed with synthetic somatostatin indicated that the immunocytochemical localization of this hormone was specific (Fig. 6b). Somatostatin was not detected in the median eminence in any of the fetuses studied at either 15 or 16 days of gestation (Fig. 7). At 17 days somatostatin was observed in only 2 of the 10 animals studied. The amount detected was extremely small and the foci containing somatostatin were located bilaterally on the ventral surface of the postinfundibular median eminence (Fig. 8 a). Characteristically, the reaction deposit indicative ofsomatostatin in presumptive axons and nerve endings was more diffuse than that seen when G n R H was immunostained in the fetal median eminence (Gross and Baker, 1977), Because the amount of somatostatin detected in fetal tissues was so small, control procedures were especially important to demonstrate specificity. Sections labeled with antisera preabsorbed with somatostatin always exhibited a lack of staining (Fig. 8b). S omatostatin was detected in 2 of the 5 fetuses studied at 18 days of gestation. It was present dorsal to the tuberoinfundibular sulci and along the ventral surface of the median eminence, similar to its localization at 17 days.
258
D.S. Gross and J.D. Longer
At 19 days of gestation somatostatin was observed in 8 of the 10 animals studied. It was located over the tuberoinfundibutar sulci, at the ventral surface of the postinfundibular median eminence and in a diffuse band in the external lamina of the infundibulum (Fig. 9). There is considerable variation in the quantity of somatostatin detected in the 19-day fetuses, but the amount is consistently greater than that seen at the earlier ages studied. Immunoreactive deposits were often identified at the ventral surface of the median eminence, adjacent to the superficial, or supratuberal, plexus of capillaries forming the primary plexus of the fetal hypothalamo-hypophysial portal system (Fig. 10). Somatostatin was not detected in neuronal cell bodies of any fetus examined.
Discussion
In this study, GH was first detected in the hypophysial pars distalis at 16 and was prominent by 17 days of gestation in the mouse. This is in general agreement with the results of other investigators. Blazquez et al. (1974) and Rieutort (1974), using radioimmunoassays, detected G H in the pituitary and plasma of rat fetuses at 19 days of pregnancy. Using immunocytochemistry, S6t/tl6 and Nakane (1976) observed somatotropes at 18 days in the fetal rat. In the mouse, somatotropes have been identified on the basis of their fine structural characteristics at 19 days of gestation (Sano and Sasaki, 1969; Dearden and Holmes, 1976). The present study is the first report in which somatotropes are identified in the fetal mouse pituitary gland by means of immunocytochemistry. The PAP technique is one of the most sensitive methods of detecting hormones in tissue (Petrali et al., 1974), and this sensitivity may explain the earlier detection of GH in our study. The apparent concentration of GH in the fetal pars distalis increases from 16 to 19 days, but there is still considerably less than in the adult. Another difference between the fetal and adult pituitary is that the adult shows sex-related differences in the somatotropes, with the male pituitary containing more and larger cells than the female (Baker and Gross, 1978). This distinction does not exist in the fetal tissues, perhaps because gonadal steroids of fetal origin are not yet circulating in sufficient quantity to affect the pituitary, or because steroid receptors on somatotropes are not yet functional. It is also possible that a high concentration of maternal estrogen masks the effects of small quantities of sex hormones produced by the fetal gonads. This investigation is the first to study the appearance of somatostatin in the fetus of a laboratory animal. Our results indicate that somatostatin is present in the median eminence of some mice as early as 17 days of gestation, but is not consistently detected until 19 days. In the adult rat, somatostatin appears to be constained in intra-axonal granules (Pelletier et al., 1977). The appearance of somatostatin in the median eminence of the mouse fetus can be correlated with the development of nerve endings containing small dense-core granules. These nerve endings first come in contact with the perivascular space surrounding capillaries of the supratuberal plexus at 16 (Eurenius and Jarskar, 1971) or 18 (Beauvillain, 1973) days of gestation. The proximity of somatostatin-containing elements in the ventral median eminence to the developing portal vascular system suggests that this hormone may gain access to the developing pituitary gland during gestation. In its
Somatostatinand Growth HormoneDuring Development
259
ventral position, somatostatin is in close contact with the superficial capillaries of the developing supratuberal plexus which is present in the mouse as early as 16 days of gestation (Dearden and Holmes, 1976). This plexus is connected to sinusoidal capillaries in the pars distalis by portal veins at either 16 (Enemar, 1961) or 17 (Dearden and Holmes, 1976) days of gestation. This morphological evidence suggests that somatostatin may affect pituitary production and/or secretion of GH at this stage in development. Many fetal GH-cells were observed abutting on sinusoidal capillaries where they may be affected by somatostatin in portal blood. The appearence of somatostatin in the median eminence coincides with the appearance of monoamines in axons in this brain region at 17 days of gestation in the mouse (Bjrrklund et al., 1968). It has been reported that such biogenic amines exert control over GH secretion in both the fetal and adult rat, but whether such amines act directly on the pituitary or exert effects via the hypothalamus has not been determined (Stuart et al., 1976). Certain hypotheses can be offered concerning the relationship between somatostatin, which is not detectable consistently in the median eminence until 19 days of gestation, and GH, which appears in the anterior pituitary gland as early as 16 days of gestation. The possibility exists that a releasing hormone, present in the median eminence prior to the 19th day of gestation, could have access to the developing pituitary gland by way of an intact hypothalamo-hypophysial portal system or by diffusion. This putative growth hormone-releasing factor (GHRF) could affect GH production and perhaps differentiation of somatotropes before somatostatin exerts its effect. A similar hypothesis has been suggested by Kaplan et al. (1976) concerning the ontogenesis of somatostatin in the human fetus. They postulated that early in gestation GH production is controlled by a releasing hormone because GH concentration in the pituitary and plasma of anencephalic fetuses is low. Rieutort and Jost (1976) reported similar findings in anencephalic rat fetuses. In normal human fetuses Kaplan et al. (1976) found that serum GH levels decrease as gestation progresses and they suggested that this could be due to increased somatostatin secretion, decreased releasing hormone secretion, or both. Confirmation of this hypothesis awaits identification of the putative GHRF, and development of techniques for its measurement. Another hypothesis which fits our results is that somatotropes may develop in the fetal pituitary before the influences of regulating hormones from the hypothalamus are exerted. This could explain the appearance of somatostatin in the median eminence later than the appearance of GH in the pituitary. Experimental evidence exists concerning the ability of the pituitary gland to undergo partial differentiation independent of hypothalamic influences (Daikoku et al., 1973; Nemesk~ry et al., 1976; Srt~16 and Nakane, 1976). It must be noted that, although the immunocytochemical technique is able to indicate the presence of hormones in tissue, it can give only limited information about their release. Nonetheless, the presence of somatostatin in the median eminence, of GH in the anterior pituitary, and of an intact portal vascular system is consistent with the assumption that these hormones are being secreted and that this neuroendocrine axis is functional during fetal development. Although it has not yet been tested on fetal rat or mouse pituitaries, exogenous somatostatin has been found to decrease the secretion of GH by human fetal pituitaries in vitro (Goodyer
260
D.S. Gross and J.D. Longer
et al., 1977). Gross and Baker (1979) have found that another neuroendocrine system is also intact at this time. They determined that gonadotropin-releasing hormone and luteinizing hormone both appear in the mouse at 17 days of gestation, suggesting the potential for hypothalamo-pituitary-gonadal function in the fetus. However, for the somatostatin-GH axis to be functional, there must be evidence for effects of GH on fetal growth and/or metabolism. The literature, however, contains conflicting reports on the role of growth hormone in fetal growth, due, in part, to species variation 0ost, 1954, 1962; Eguchi, 1961; Heggestad and Wells, 1965; Liggins and Kennedy, 1968). However, several investigators agree that GH affects many metabolic processes in prenatal animals (Jost, 1966; Blazquez, 1974; Honnebier and Swaab, 1974). Thus, although the function of GH in fetal growth is a subject of controversy, GH does appear to play a role in fetal development. This fact, along with the data obtained from the present study, suggests that a functional hypothalamohypophysial system for GI-I secretion exists during fetal development in the mouse. References Baker, B.L., Gross, D.S.: Cytology and distribution of secretory cell types in the murine hypophysis as demonstrated by immunocytochemistry. Am. J. Anat. 153, 193-216 (1978) Baker, B.L., Jaffe, R.B.: The genesis of cell types in the adenohypophysis of the human fetus as observed with immunocytochemistry. Am. J. Anat. 143, 137-162 (1975) Baker, B.L., Yu, Y.Y.: The influence of hypophysectomy of the stores of somatostatin in the hypothalamus and pituitary stem. Proc. Soc. Exp. Biol. Med. 151, 599-602 (1976) Beauvillain, J.C.: Structure fine de l'bminence m~diane de souris au cours de son ontog~n6se. Z. Zellforsch. 139, 201 215 (1973) Bj6rklund, A., Enemar, A., Falck, B.: Monoamines in the hypothalamo-hypophysial system of the mouse with special reference to the ontogenetic aspects. Z. Zellforsch. 89, 590-607 (1968) Bl~izquez, E., Simon, F.A., Blb,zquez, M., Fo/~, P.P.: Changes in serum growth hormone levels from fetal to adult age in the rat. Proc. Soc. Exp. Biol. Med. 147, 780-783 (1974) Brownstein, M., Arimura, A., Sato, H., Schally, A.V., Kizer, J.S.: The regional distribution of somatostatin in the rat brain. Endocrinology 96, 1456-1461 (1975) Campbell, H.J.: The development of the primary portal plexus in the median eminence of the rabbit. J. Anat. 100, 381-387 (1966) Conklin, P.M., Schindler, W.J., HoU, S.F.: Hypothalamic thyrotropin releasing factor and pituitary responsiveness during development in the rat. Neuroendocrinology 2, 304-314 (1973) Daikoku, S., Kinutani, M., Watanabe, Y.G.: Role of the hypothalamus on development of the hypophysis: an EM study. Neuroendocrinology 11, 284-305 (1973) Dearden, N.M., Holmes, R.L.: Cyto-differentiation and portal vascular development in the mouse adenohypophysis. J. Anat. 121, 551-569 (1976) Dub6, D., Leclerc, R., Pelletier, G., Arimura, A., Schally, A.V.: Immunohistochemical detection of growth hormone-release inhibiting hormone (somatostatin) in the guinea pig brain. Cell Tissue Res. 161, 385-390 (1975) Eguchi, Y.: Atrophy of the fetal mouse adrenal following decapitation in utero. Endocrinology 68, 716719 (1961) Enemar, A.: The structure and development of the hypophysial portal system in the laboratory mouse, with particular regard to the primary plexus. Ark. Zool., II. Set. 13, 203-252 (1961) Eurenius, L., Jarskar, R.: Electron microscopic studies on the development of the external zone of the mouse median eminence. Z. Zellforsch. 122, 488-502 (1971) Glydon, R.J.: The development of the blood supply of the pituitary in the albino rat, with special reference to the portal vessels. J. Anat. 91, 237-244 (1957) Goodyer, C.G., St. George Hall, C., Gudya, H., Robert, F., Giroud, C.J.P.: Human fetal pituitary in culture: hormone secretion and response to somatostatin, LHRH, TRH, and dbcAMP. J. Clin. Endocrinol. Metab. 45, 73-84 (1977)
Somatostatin and Growth Hormone During Development
261
Gross, D.S., Baker, B.L.: Immunohistochemical localization of gonadotropin-releasing hormone (GnRH) in the fetal and early postnatal mouse brain. Am. J. Anat. 148, 195-215 (1977) Gross, D.S., Baker, B.L. : Developmental correlation between hypothalamic gonadotropin-releasing hormone and hypophysial luteinizing hormone. Am. J. Anat. 154, 1-10 (1979) Hal~isz, B., Kosaras, B., Lengvari, I.: Ontogenesis of the neurovascular link between the hypothalamus and the anterior pituitary in the rat. In: Median Eminence: Structure and Function. (K.M. Knigge, D.E. Scott, A. Weindl, eds.). Basel: S. Karger (1972) Harris, G.W.: Neural control of the pituitary gland. Physiol. Rev. 28, 139-179 (1948) Heggestad, C.B., Wells, L.J.: Experiments on the contribution of somatotrophin to prenatal growth in the rat. Acta Anat. 60, 348-361 (1965) Hrkfelt, T.M., Efendic, S., Hellerstrom, C., Johansson, O., Luft, R., Arimura, A.; Cellular localization of somatostatin in endocrine-like cells and neurons of the rat. Acta Endocrinol. 80, Supp 100 (1975) Honnebier, WJ., Swaab, D.F.: Influence of ~-melanocyte-stimulating hormone (c~-MSH), growth hormone (GH), and fetal extracts in intrauterine growth of fetus and placenta in the rat. J. Obstet. Gynecol. Brit. Comm. 81, 439-447 (1974) Jost, A.: Hormones in the development of the foetus. Cold Spring Harbor Symp. Quant. Biol. 19, 167181 (1954) Jost, A.: Endocrine factors of foetal development. Triange 5, 189-193 (1962) Jost, A.: Anterior pituitary function in fetal life. In: The Pituitary Gland. Vol. II. (G.W. Harrison, and B.T. Donovan,eds.) Berkeley: University of California Press 1966 Jost, A., Dupouy, J.P., Geloso-Meyer, A.: Hypothalamo-hypohysial relationships in the fetus. In: The Hypothalamus. (L. Martini, M. Motta, and F. Fraschini, eds.) New York: Academic Press 1970 Kaplan, S.L., Grumbach, M.M., Aubert, M.L.: The ontogenesis of pituitary hormones and hypothalamic factors in the human fetus; maturation of central nervous system regulation of anterior pituitary function. Recent Prog. Horm. Res. 32, 161-234 (1976) King, J.C., Arimura, A., Gerall, A.A., Fishback, J.B., Elkind, K.E.: Growth-hormone release inhibiting hormone (GHIH) pathway of the rat hypothalamus revealed by the unlabeled antibody peroxidaseantiperoxidase method. Cell Tissue Res. 160, 423-427 (1975) Liggins, G.C., Kennedy, P.G.: Effects of electrocoagulation of the fetal lamb hypophysis in growth and development. J. Endocrinol. 40, 371-381 (1968) Nemeskrry, A., Nrmeth, A., Srt~il6, G., Vigh, S., Halfisz, B.: Cell differentiation in the fetal rat anterior pituitary in vitro. Cell Tissue Res. 170, 263-273 (1976) Pelletier, G., Leclerc, R., Dubr, D., Labrie, F., Puviani, R., Arimura, A., Schally, A.V.: Localization of growth hormone-release-inhibiting hormone (somatostatin) in rat brain. Am. J. Anat. 142, 397-400 (1975) Pelletier, G., Dubr, D., Puviani, R.: Somatostatin: electron microscope immunohistochemical localization in secretory neurons of rat hypothalamus. Science 196, 1469-1470 (1977) Petrali, J.P., Hinton, D.M., Moriarty, G.C., Sternberger, L.A.: The unlabeled antibody enzyme method of immunocytochemistry. Quantitative comparison of sensitivities with and without peroxidaseantiperoxidase complex. J. Histochem. Cytochem. 22, 782-801 (1974) Rieutort, M. : Pituitary content and plasma levels of GH in fetal and weanling rats. J. Endocrinol. 60, 261-268 (1974) Rieutort, M., Jost, A.: Growth hormone in encephalectomized rat fetuses with comments on the effects of anesthetics. Endocrinology 98, 1123-1129 (1976) Sano, M., Sasaki, F.: Embryonic development of the mouse anterior pituitary studied by electron microscopy. Z. Anat. Entwickl. 129, 195-222 (1969) Srtfil6, B., Nakane, P.K.: Functional differentiation of the fetal anterior pituitary cells in the rat. Endocrinol. Exp. 10, 155-166 (1976) Sternberger, L.A., Hardy, Jr., P.H., Cuculis, J.J., Meyer, H.B.: The unlabeled antibody enzyme method of immunohistochemistry. Preparation and properties of soluble antigen-antibody complex (horseradish peroxidase-antihorseradish peroxidase) and its use in identification of spirochetes. J. Histochem. Cytochem. 18, 315-333 (1970) Stuart, M., Lazarus, L., Smythe, GA., Moore, S., Sara, V.: Biogenic amine control of growth hormone secretion in the fetal and neonatal rat. Neuroendocrinology 22, 337-342 (1976)
Accepted May 29, 1979
