Immunohistochemical distribution of regulatory peptides in the human fetal adenohypophysis

Abstract

We have studied here the cellular distribution of several regulatory peptides in hormone-producing cells of the human pituitary during the fetal period. Immunohistochemistry was used to show the expression of several regulatory peptides, namely Angiotensin-II, Neurotensin and Galanin, at successive gestational stages and their co-localization with hormones in the human fetal adenohypophysis. Somatotrophs, gonadotrophs and thyrotrophs were differentiated earliest. At gestational week 9, Angiotensin-II immunoreactivity was co-localized only with growth hormone immunoreactivity in somatotrophs, one of the first hormone-producing cells to differentiate. This co-localization remained until week 37. Neurotensin immunoreactivity was present in gonadotrophs and thyrotrophs in week 23, after FSH and TSH hormone differentiation. Galanin immunoreactivity was present in all hormone-producing cell types except corticotrophs. The different pro-opiomelanocortin-derived peptides were detected at different stages of gestation and adrenocorticotrophic hormone immunoreaction was the last to be detected. Our results show an interesting relationship between regulatory peptides and hormones during human fetal development, which could imply that these peptides play a regulatory role in the development of pituitary function.

Introduction

The pituitary is the main neuroendocrine gland in vertebrates. Pituitary cell differentiation, growth and secretory function are dependent not only on hypothalamic factors but on locally produced and released paracrine and autocrine factors. These comprise a wide range of substances, some of which are called regulatory peptides. The mechanism by which peptides act when co-localized with hormone-producing cells is not well known. In mammals, the presence of several peptides has been shown in adenohypophysial cells, particularly in rats (Bello et al. 1992, 1999, 2004; Cimini, 2000). Other immunohistochemical studies determined the presence of peptides in the human pituitary (Roth & Krause, 1990; Gehlert et al. 1994; Takahashi et al. 1997; Grouzmann et al. 1998). Co-localization studies have shown coexpression of galanin (Gal) in adrenocorticotrophic cells (Vrontakis et al. 1990; Hsu et al. 1991; Leung et al. 2002). Reyes et al. (2007) showed that neurotensin (NT) is present in the corticotrophs of a pituitary secreting microadenoma. NT-immunoreactivity (NT-ir) was also present in gonadotrophs (FSH-ir) and thyrotrophs (TSH-ir) in healthy tissue. Other developmental studies of pituitary peptide expression, Gal, vasoactive intestinal peptide (VIP) or neuropeptide Y (NPY) (Reyes, 2002), Gal (Cimini, 2000, 2003) and NT (Bello et al. 2004) contributed new data concerning their relationship with different pituitary hormones in mammals.

Immunohistochemical studies on the cytogenesis, development and differentiation of the pituitary gland have also been carried out in several vertebrate groups (Chatelain et al. 1979; Batista et al. 1989; Japon et al. 1994; Allaerts et al. 1999), although in humans these studies are scarce (Baker & Jaffe, 1975; Asa et al. 1986; Tachibana et al. 1994), with apparently no studies of peptides in the developmental period.

Due to the known importance of peptides in the regulation of several adult pituitary cell functions, and their apparently different roles during development following research results in other mammals, our aim was to study the development of NT, Gal and Angiotensin-II (Ang-II) in human pituitary cells and correlation with the development of hormone-producing cells.

Materials and methods

Pituitary samples

Twenty-four human fetuses were obtained from spontaneous abortions, legal therapeutic terminations of pregnancy and intrauterine death. Gestational age was determined from the clinical histories. Different age groups were used: 9, 19, 23, 30, 36 and 37 weeks of gestation. Any fetus with serious abnormalities revealed by the autopsy report was excluded. The use of this material and this study itself were approved by the Ethics Committee of the University Hospital, Tenerife, Canary Islands, Spain.

The pituitaries were fixed in phosphate buffer (pH 7.4, 0.1 m) containing 4% paraformaldehyde and 0.2% picric acid and then immersed overnight in sodium veronal buffer (pH 7.4, 0.1 m) with 20% sucrose. The samples were embedded in Tissue-Tek and frozen in isopentane cooled with liquid nitrogen. Horizontal sections (8–10 µm) were obtained with a cryostat, collected on gelatine-coated slides, allowed to dry and rehydrated in sodium veronal buffer.

Immunohistochemistry

An indirect immunoenzymatic technique was used. After rehydration, blocking of non-specific binding with casein and of endogenous peroxidase with hydrogen peroxide was carried out. The sections were then incubated overnight at room temperature with the following specific antisera: anti-ACTH (1–24) and anti-hβFSH, both at dilution 1 : 1000, anti-hβTSH, anti-hGH, anti-αMSH, anti-βMSH, anti-βEnd and anti-rPRL, all at dilution 1 : 800, anti-Ang-II, anti-Gal and anti-NT at dilution 1 : 200 and, after several washes in sodium veronal buffer, sections were incubated with peroxidase conjugated to goat anti-rabbit IgG (Jackson Immunochemical, Baltimore, MD, USA). Peroxidase activity was revealed in Tris-HCl buffer (pH 7.6; 0.005 m) containing 0.04% 4-chloro-1-naphthol (Sigma) and 0.01% hydrogen peroxide.

To demonstrate coexistence, an elution-restaining procedure was used (Tramu et al. 1978), allowing a second immunocytochemical sequence to be carried out using the following antibodies: anti-hβFSH (1 : 1000), anti-hβTSH (1 : 800), anti-hGH (1 : 800) or anti-rPRL (1 : 800). Peroxidase activity was revealed using 0.005% 3,3′-diaminobenzidine tetrahydochloride (Sigma) and 0.004% hydrogen peroxide in Tris-HCl buffer.

Antisera

Antisera used in this study, provided by Dr G Tramu from Laboratoire de Neurocytochimie Fonctionnelle, Talence, France, were developed in rabbits against the adenohypophysis hormones and regulatory peptides: ACTH (1–24), αMSH, βMSH, βEnd, hβTSH, hGH, hβFSH, rPRL, NT, Gal and Ang-II. Their immunological properties have been described in previous research (Dubois MP, 1972; Dubois P, 1972; Tramu & Dubois, 1977; Hemming et al. 1986; Bello et al. 1991, 1992; Anglade et al. 1994; Jamali & Tramu, 1999).

The specificity of the immunostaining was assessed by replacing the specific antiserum by normal serum or following preabsorption of the antisera with the corresponding antigens.

The specificity of the elution procedure was assessed by replacing the specific antiserum of the second immunoreaction by buffer or normal serum (1 : 50, 60 min).

Quantitative analysis

Anterior pituitary sections were examined by light microscopy (Leitz Laborlux S, Germany). The immunoreactive cells per unit area (1000 µm²) were counted using Leica Q-Win software (Analysis Image System Leica Q-Win, Barcelona, Spain). Three to five fields from the same section were measured in each of five sections per case, and the mean was calculated. Both the immunostained cytoplasm and a non-stained nucleus were present in the cells considered in the count; immunoreactive broken cells or subcellular artefacts were excluded by the program itself on tuning it to a particular size of object. Only cell profiles with both immunostained cytoplasm and a non-stained nucleus were counted. Other stained fragments and artefacts were excluded by the program itself, which is set to a particular minimum size of object. The criterion used to consider a cell to be immunoreactive was the intensity of immunostaining, which was evaluated in arbitrary units of grey levels ranging from 1 (black) to 256 (white). Cells with values higher than those of the background of their control sections were considered to be reactive. This same analysis determined the number of cells displaying immunoreactivity for each hormone or peptide at different embryonic stages.

Results

In each period we studied the sequence of hormones and peptides present.

Gestation week 9

GH, TSH and FSH immunoreactivity were detected in pituitary cells (Fig. 1a–c), indicating that somatotrophs, thyrotrophs and gonadotrophs are the first to be differentiated in human pituitary development. At the same time, Ang-II-ir cells (Fig. 1d) were observed that were also GH-ir (Fig. 1e₁,e₂).

Fig. 1.

Section of human pituitary embryo at 9 weeks of gestation (E.9) showing GH-ir (a), TSH-ir (b), FSH-ir (c) and Ang-II-ir (d) cells. Section of human pituitary gland at E.9 showing the GH-ir cells (e₁) that also show Ang-II-ir (e₂). Scale bars: a,c,d = 20 µm; b = 30 µm; e₁,e₂ = 10 µm.

Gestation weeks 9–23

The number of GH-ir (Fig. 2a), TSH-ir and FSH-ir cells increased (Fig. 7a). At week 23, PRL-ir and βEnd-ir were located in a few pituitary cells at the anterior lobe periphery (Fig. 2c,d). Ang-II-ir (Fig. 2b) cell number increased (Fig. 7b), and a few NT-ir (Fig. 2e) and Gal-ir cells (Fig. 2f) were present for the first time.

Fig. 2.

Section of human pituitary at 23 weeks of gestation (E.23) showing GH-ir (a), Ang-II-ir (b), PRL-ir (c), βEnd-ir (d), NT-ir (e) and Gal-ir (f) cells. Scale bars: a,b,e,f = 25 µm; c,d = 10 µm.

Fig. 7.

Number of cells displaying immunoreactivity for adenohypophysial hormones (a) and regulatory peptides (b) in human pituitary at different embryonic stages. All values are the mean of immunostained cells number per unit area ± SEM in each adenohypophysis region. Measurements were made on five sections per sample (n = 5).

Gestation week 30

The PRL-ir cell number increased (Fig. 7a). GH-ir and Ang-II-ir decreased (Fig. 3a,b) and αMSH (Fig. 3c), βMSH (not shown) and ACTH-ir (Fig. 3d) cells were seen for the first time.

Fig. 3.

Section of human pituitary at 30 weeks of gestation (E.30) showing GH-ir (a), Ang-II-ir (b), αMSH-ir (c) and ACTH-ir (d) cells. Scale bars: a,b,d = 20 µm; c = 30 µm.

Gestation weeks 36 and 37

At the age usually corresponding to birth, GH-ir cells increased (Figs 4a and 7a); as did PRL-ir (Figs 4d and 7a) and all cells immunoreactive for Pro-opiomelanocortin (POMC)-derived peptides, βEnd (Figs 4e, 7a), αMSH, βMSH and ACTH (Fig. 7a), in the anterior and intermediate lobes. Ang-II-ir (Fig. 4b) decreased (Fig. 7b); this peptide was present not only in GH-ir cells but also in other cell types not yet determined. There was also a notable increase in NT-ir (Fig. 4f) and Gal-ir (Fig. 4g) cells (Fig. 7b); NT-ir was present in thyrotrophs (Fig. 5a₁,a₂) and gonadotrophs (Fig. 5b₁,b₂), although there were many NT-ir cells that did not correspond to these two cell types, while Gal-ir was present in gonadotrophs (Fig. 6a₁,a₂), thyrotrophs (Fig. 6b₁,b₂), somatotrophs (Fig. 6c₁,c₂) and lactotrophs (Fig. 6d₁,d₂).

Fig. 4.

Section of human pituitary at 36 weeks of gestation (E.36) showing GH-ir (a), Ang-II-ir (b), FSH-ir (c), PRL-ir (d), βEnd-ir (e), NT-ir (f) and Gal-ir (g) cells. Scale bars: a–g = 10 µm.

Fig. 5.

Section of human pituitary gland at E.36 showing the NT-ir cells (a₁) that also show TSH-ir (a₂) and NT-ir cells (b₁) that also show FSH-ir (b₂). Scale bar = 7 µm.

Fig. 6.

Section of human pituitary gland at E.36 showing the Gal-ir cells (a₁) that also show FSH-ir (a₂), Gal-ir cells (b₁) that also show TSH-ir (b₂), Gal-ir cells (c₁) that also show GH-ir (c₂) and Gal-ir cells (d₁) that also show PRL-ir (d₂). Scale bar = 5 µm.

Discussion

This study of human pituitary cells during development focuses on the relationship between the hormones and the regulatory peptides during differentiation. To our knowledge, the presence of regulatory peptides in hormone-producing cells during the developmental period in human pituitary cells has not previously been reported. Our results show that in the human pituitary, GH-ir cells were the earliest differentiated, co-localization with Ang-II occurring at the same time. These cells were observed at week 9; for Asa et al. (1986) these cells were also the first detected, at week 8, but at the same time as adrenocorticotrophin and β endorphin. Other authors demonstrated an early presence of GH cells in an electron microscopy study (Tachibana et al. 1994) and reported that one ultrastructural type of GH cells was present in the first half of gestation and a second type in the second half. GH is one of the regulatory factors of fetal growth (Waters & Kaye, 2002), its receptors being widespread (Hill et al. 1992; Werther et al. 1993) and expressed at week 8 (Simard et al. 1996). By contrast, human pituitary produces GH from the first trimester of gestation onwards.

The presence of Ang-II-ir in only the GH-ir cells until the last week of gestation strongly suggests a role in GH function, above all in the first half of the developmental period; the stimulatory effect of Ang-II on GH secretion in humans was demonstrated previously (Coiro et al. 1998).

In most species studied (Gasc & Sar, 1981; Thommes et al. 1987), excluding some birds, corticotrophs (ACTH-ir) are the first cells to be differentiated (Batista et al. 1989; Japon et al. 1994; Reyes, 2002). In humans, fewer data (Baker & Jaffe, 1975; Asa et al. 1986) are available; note, however, that there were major differences in sensitivity of technique and antibodies at the time these previous studies were made. The POMC-derived peptide antibodies in our study were used previously in other mammals and we demonstrate that the results were different in humans (Reyes, 2002). In addition, with the same antibody in human pituitary ACTH-secreting microadenoma, we showed the presence of ACTH-ir and, as expected, no ACTH-ir was present in healthy tissue while the other POMC-derived peptides were present (Reyes et al. 2007). Our results show that the first POMC-derived peptide detected in these cells was β endorphin-ir, at week 23 of embryonic development, while the other POMC-derived peptides, ACTH, αMSH and βMSH, were not observed until week 30 in the anterior and intermediate lobes. In contrast to studies in other vertebrate groups (Batista et al. 1989; Japon et al. 1994; Reyes, 2002), where corticotrophs became differentiated at different times in the anterior and intermediate lobes, in humans we found simultaneous corticotroph differentiation in both lobes.

Gal-ir coexists with ACTH-ir in a small group of rat pituitary cells (Cimini, 1996) and in adult human corticotrophs (Vrontakis et al. 1990; Hsu et al. 1991). Cimini (1996) suggests that Gal can modulate corticotrophin release in vitro; however it has been localized mainly in lactotrophs (Ren et al. 1999) and appears to be regulated by oestrogen (Shen et al. 1995) in female rats. By contrast, Hyde et al. (1998) observed Gal-ir in somatotrophs and thyrotrophs of male rats; in addition, a role in the proliferation and regulation of prolactin secretion has been suggested (Wynick et al. 1998) for this peptide. However, in the fetal rat, coexistence of Gal with most of the pituitary hormones has been observed (Cimini, 2000, 2003). Our results in human fetal pituitary are similar; Gal-ir was present in gonadotrophs, thyrotrophs, somatotrophs and lactotrophs, which suggests a similar role of galanin in rat and human fetal development. In adult humans, however, it is confined to corticotrophs (Vrontakis et al. 1990; Hsu et al. 1991).

In the present study, GH-ir, TSH-ir and FSH-ir cells were detected in a 9-week fetus; Asa et al. (1986) reported the presence of the β subunit of thyroid-stimulating hormone, follicle-stimulating hormone and luteinizing hormone in a 12-week-old fetus; Baker & Jaffe (1975) reported the presence of gonadotrophs and somatotrophs at week 10, although they did not detect TSH-ir until week 13.

Although NT-ir was present in FSH-ir and THS-ir cells, peptide and hormone immunoreactivity was not detected at the same time. This location was the same as that observed in adult human pituitary (Reyes et al. 2007). NT-ir was present in postnatal and adult rats in these same cell types (Bello et al. 1992, 1999, 2004); in this species NT was not present in embryos and sexual dimorphism in the distribution of NT-ir in pituitary cells (Bello et al. 1992) was established at week 5 of postnatal life (Bello et al. 2004), suggesting a common function for NT before sexual maturity. A relationship with the reproductive axis is also seen in rats (Bello et al. 1992, 1999). In human fetuses, we detected NT-ir in FSH- and TSH-ir cells of both genders, as observed in adults (Reyes et al. 2007). However, in human fetuses NT-ir was detected in cells other than TSH-ir or FSH-ir cells, as yet undefined. These results suggest a role for NT in the development of pituitary function not present in rats that may continue in the adult gland.

In conclusion, immunoreactivities for NT, Gal and Ang-II peptides were co-localized with pituitary hormones in different cells in human pituitary during the developmental period. This distribution pattern appears to be different from that observed in the adult pituitary for Gal and Ang-II, which suggests a regulatory role in hormone-producing cells of the anterior pituitary during development. Clarification of this regulatory role would greatly assist in understanding the functions of peptides involved in pituitary activity.

The scarce research in this area in humans inevitably shows discrepancies due to the greater technical sensitivity and specificity of antibodies now available. Further work is required to clarify these differences in results.