Cell Tissue Res (1992) 267:251-260
Cell&Tissue Research 9 Springer-Verlag 1992
Immunohistochemical localization of keratin in bull, goat, and sheep anterior pituitary glands Takaki Shimada Department of Anatomy,Jikei UniversitySchool of Medicine, 3-25-8 Nishi-Shinbashi, Minato-ku,Tokyo, 105 Japan Received January 31, 1991 / Accepted September 10, 1991
Summary. Immunohistochemical localization of keratin, an intermediate filament protein, was studied in bull, goat, and sheep anterior pituitary glands, i.e., in animals of the order Artiodactyla. Strong immunoreactivity was detected in the cells of the marginal layer of bull and goat, as well as in cysts or large follicles in the anterior lobe of all 3 species. In addition, a number of stellateshape cells were immunoreactive for keratin and were distributed throughout the anterior lobe. The localization of keratin-positive cells in light-microscopic preparations correlated precisely with the localization of folliculo-stellate cells in adjacent ultrathin sections. In ultrastructural studies, many slender and elliptical membranous components which were different from smooth endoplasmic reticulum were observed in the cytoplasm of the some keratin-positive cells. Some of the folliculostellate cells in the 3 species were also immunoreactive for the c~ subunit of S-100 protein, which exists in some epithelial cells. On the other hand, immunolocalization of glial fibrillary acidic protein, a glial cell marker, could not be demonstrated in the anterior pituitary glands of the 3 species studied. These results suggest that keratinpositive folliculo-stellate cells express epithelial-like characteristics. Key words: Pituitary gland, pars anterior (distalis)- Folliculo-stellate cells - Keratin S 100 protein - Cow, goat, sheep (Artiodactyla)
Folliculo-stellate (FS) cells of the vertebrate pituitary gland have been studied both by ultrastructural and immunohistochemical methods. Ultrastructurally, for the most part, FS cells lack secretory granules and tend to be clustered. They form follicles which bear microvilli and/or cilia on their apical cell surfaces. Different hypotheses have been proposed regarding the functional roles of FS cells, such as ACTH secretory
cells (Farquhar 1957), supportive cells (Salazar 1968), dedifferentiated glandular cells (Horvath et al. 1974), phagocytic cells (Dingemans and Feltkamp 1972; Fukuda 1973), and stem cells (Yoshimura et al. 1977). However, their origin and functions have not been fully clarified. By immunohistochemistry, FS cells exhibit immunoreactivity for several different proteins. In particular, the S 100 protein has been a useful marker in the adult rat (Cocchia and Miani 1980; Nakajima et al. 1980; Shirasawa et al. 1983) and in the human (Nakazato et al. 1982; Lauriola et al. 1984; Takahashi et al. 1984; Girod et al. 1985). S 100 protein, however, is not a universal marker for FS cells in all species of animals (Shirasawa et al. 1984, 1988). In the goat, pituitary growth hormone (GH) cells were intensely immunostained with S 100 protein antiserum (Shirasawa et al. 1984); however, FS cells were immunoreactive for the S 100e component (Shirasawa 1986), which is the subunit of S 100 protein (Isobe et al. 1983). Furthermore, FS cells in the human pituitary were found to be immunoreactive for both S 100 protein and glial fibrillary acidic protein (GFAP) (Velasco et al. 1982; Morris and Hitchcock 1985). GFAP, which is contained in the human FS cells, is one of the intermediate filament proteins. Intermediate filaments are well-known as 10-rim filaments and as proteins forming a part of the cytoskeleton. Biochemical and immunological analyses have enabled intermediate filaments to be classified into 5 main groups (Lazarides 1980). Keratin is another type of intermediate filament protein. Keratin proteins have been identified in many diverse types of epithelia (Schlegel et al. 1980). In the study reported here, immunoreactive keratin was detected in bull, goat and sheep pituitary glands by immunohistochemistry. Pituitary keratin-containing cells were compared with pituitary hormone-containing cells, and with cells containing S 100~ protein and GFAP, the markers of the FS cells. Furthermore, the ultrastructure of the keratin-positive cells was observed using the superimposition technique.
252
Materials and methods
Tissues and superimposition technique Adult male bull (n=5), goat (n=3), and sheep (n=3) pituitary glands were fixed in sublimate-formalin (saturated sublimate solution:formalin, 7:3) for 16 h at 4~ C. After dehydration in an ethanol series, they were embedded in Paraplast, and sectioned at a thickness of 3-4 pm. In addition, frozen sections of bull pituitary gland were prepared as follows. They were dipped into OCT compound (Tissue Tek, Miles Scientific, Naperville, USA), and frozen in acetone cooled with dry ice. After sectioning using a cryostat, (5-6 gm in thickness), they were fixed for 10 min in methanol at - 2 0 ~ and then were fixed additionally in acetone at - 2 0 ~ for 30 sec. Both the Paraplast and cryostat sections were subsequently used for immunohistochemical studies. For electron-microscopic studies, small pieces of the pituitary gland from each animal were fixed in 2% glutaraldehyde, 0.1 M phosphate buffer, pH 7.4, at 4~ C for 3 h, and then postfixed with 1% osmic acid in Millonig solution (Millonig 1962), pH 7.4, at 4~ C for 2 h. After dehydration in an ethanol series, the pituitary glands were embedded in Epon-Araldite. Ultrathin sections were prepared for electron-microscopic observations, with adjacent thick sections cut for light-microscopic immunostaining, according to the superimposition technique described by Yoshimura and Nogami (1981). This procedure is a slight modification of the method of Beauvillain et al. (1975). Epon-Araldite thick sections were treated with a 14% methanol solution of sodium methoxide, and then were washed with methanol. Subsequently, the sections were immersed in a 1% aqueous solution of hydrogen peroxide for bleaching. The ultrathin sections were stained with uranyl acetate and lead citrate, and observed with a JEOL 100B electron microscope.
Immunohistochemical studies Immunohistochemical staining was carried out by the avidin-biotin-peroxidase complex method (Hsu et al. 1981). The primary antibodies used were anti-human skin keratin (Bio-Science Products, AG, Switzerland), a monoclonal antibody against human GFAP (Amersham International plc, England), and a polyclonal antibody against cow GFAP (Dakopatts, Denmark). The specificity of the anti-human skin keratin antibody was confirmed by the absorption test described below. In addition, other primary antibodies used were: (1) polyclonal antibodies against ovine GH, ovine prolactin, porcine ACTH 1-39, rat LHfl, and rat TSHfl antisera; and (2) a monoclonal antibody against the ~ subunit of S-100 protein (Japan Immunoresearch Lab. Co., Japan). Their specificities have been described previously (Yoshimura et al. 1981 ; Shirasawa and Yoshimura 1982; Shirasawa 1986).
Absorption tests Absorption tests were carried out on serial Paraplast sections of normal pituitary glands from the 3 species to check the specificity
Fig. l a, b. Adjacent Paraplast sections of bull anterior pituitary gland immunostained using anti-keratin antibody (a), and preabsorbed antibody (b). Immunoreactivity was lost by preabsorption. C Cyst. x 200
of the human skin keratin antibody. The antibody (1:200) was incubated with 10 ~tg of purified human skin keratin protein (Sigma, St. Louis, Mo., USA) at 4~ C for 3 days. As shown in Fig. 1, sections stained with the preabsorbed antibody did not demonstrate immunoreactivity. The antibody to keratin preincubated with 10 gg of bovine S 100 protein (Wako Chemical Industries Ltd., Osaka, Japan) and the antibodies to GFAP preincubated with 10 gg of purified human skin keratin protein did not lose their immunoreactivity.
Results
Ligh t-microscopic findings A n u m b e r o f cells reacted with the a n t i b o d y to keratin in pituitary glands o f bull and goat (Figs. 2, 3, respectively). These keratin-positive cells were stellate in shape with elongated cytoplasmic processes, and had a tendency to f o r m clusters. T h e y were distributed evenly t h r o u g h o u t the anterior lobe. I m m u n o r e a c t i v i t y was noted only in the c y t o p l a s m and the staining intensity was variable. In the anterior lobe o f sheep, similar stellate-shaped cells i m m u n o r e a c t i v e for keratin were also observed (Fig. 4). However, these cells were distributed singly t h r o u g h o u t the gland and there was little tendency to f o r m clusters. In pituitary glands o f bull and goat, the marginal layer o f the pars anterior was also i m m u n o reactive for keratin (Fig. 5) and consisted o f a layer o f cells which varied in thickness. Some ovine marginal cells possessed i m m u n o r e a c t i v i t y for keratin as well (Fig. 6). I m m u n o s t a i n i n g o f these marginal layer cells was stronger t h a n that observed in the stellate cells o f the bull and g o a t anterior lobe. Cysts or large follicles, which often contained colloid, were frequently observed in the anterior lobe in all o f the animals studied. Cells lining the cysts were strongly immunoreactive for keratin (Figs. 5, 7). The i m m u n o r e activity was observed in the cytoplasm only and the staining intensity was similar to that f o u n d in the marginal cells. The distribution patterns, cell size, and shapes o f keratin-positive cells were c o m p a r e d by i m m u n o h i s t o c h e m istry with that o f the pituitary h o r m o n e - p r o d u c i n g cells, G H , P R L , L H , TSH, and A C T H on adjacent sections (Figs. 8-12). G H , L H , and T S H cells were oval in shape, whereas keratin-positive cells were stellate-shaped (Figs. 8, 10, 11). P R L and A C T H cells were p o l y g o n a l ; however, their patterns o f distribution were different f r o m that o f keratin-positive cells (Figs. 9, 12). Cells i m m u n o r e a c t i v e with anti-S 100c~ a n t i b o d y were scattered t h r o u g h o u t the anterior pituitary glands o f the 3 animals studied. In the bull anterior pituitary gland, stellate-shaped cells and endothelial cells o f b l o o d vessels showed immunoreactivity, while in the g o a t and sheep anterior pituitary glands, immunopositive cells were stellate-shaped with elongated cytoplasmic processes (Figs. 13 b, 14b). I m m u n o r e a c t i o n p r o d u c t s were f o u n d in their c y t o p l a s m and nucleus. Their distribution was not u n i f o r m and they had a tendency to accumulate in the peripheral area o f the sheep anterior lobe, and in the area beneath the marginal layer o f the bull anteri-
253
Figs. 2--7. Paraplast sections immunostained for keratin. A number of stellate-shaped cells (Figs. 2, 3, 4), marginal cells (Figs. 5, 6) and cysts (Figs. 5, 7) exhibit immunoreactivity. ML Marginal layer', C cyst; I intermediate lobe; A anterior lobe. Figs. 2, 5 bull; Fig. 3 goat; Figs. 4, 6, 7 sheep. Figs. 2, 3, 4 x 320; Fig. 5 x 160; Figs. 6, 7 (counter-stained with hematoxylin) • 320
or lobe. There were few S 100~-positive stellate-shaped cells in the other regions of the gland. Some, but not all, of the marginal cells also contained S 100~ protein in anterior pituitary glands of bull and goat. Adjacent sections immunostained with anti-S 100~ and anti-keratin antibodies revealed that some ceils contained both S 100e and keratin proteins (Figs. 13, 14), while others were immunostained for only one of the proteins. There were no immunoreactive G F A P cells detected in anterior pituitary glands of the 3 species by the use of monoclonal or polyclonal anti-GFAP antibodies (Fig. 15a). However, in intermediate lobes of bull and goat, some stellate-shaped cells exhibited immunostainability for G F A P (Fig. 15b). The staining intensity was the same with either the monoclonal or polyclonal antibodies.
Electron-microscopicfindings of keratin-positive cells Ultrastructural characteristics of cells immunoreactive for keratin were observed using the superimposition technique (Figs. 16, 17). The ultrastructural counterparts
of the keratin-positive cells were agranular and irregular in shape, with elongated cytoplasmic processes and were identified as folliculo-stellate cells. The cytoplasmic processes contacted the parenchymal basement membrane that separated the glandular epithelium from the connective tissue. Keratin-positive cells were often clustered and formed follicles. In the bull anterior lobe, the follicular lumen was filled with colloid (Fig. 16) - an observation which was less frequently made in the anterior lobe of the goat. These cells had many microvilli which were closely interdigitated with each other. Intercellular junctions were well-developed between neighboring agranular cells (Figs. 18, 19, 20 a), and they were also observed between keratin-positive cells and granular cells (Fig. 19). Filamentous components were observed in the cytoplasm of the keratin-positive cells (Fig. 18). At the apical side of the FS cells, densely stained membranes were present in the adjacent FS cells. They were interdigitated with each other, and often presented round profiles with included cytoplasmic components (Figs. 18, 19). Some cells contained numerous slender and elliptical membranous components (Figs. 19, 20b). They were
254
Figs. 8--12. Adjacent Paraplast sections of bull anterior pituitary gland immunostained for keratin
(Figs. 8-12a), GH (Fig. 8b), prolactin (Fig. 9b), LH (Fig. 10b), TSH (Fig. l l b ) , and ACTH (Fig. 12b). Immunoreactive cells for keratin are different from hormone-producing cells (counter-stained with hematoxylin), x 500
255
Fig. 13a, b. Adjacent Paraplast sections of goat pituitary immunostained for keratin (a), and for S 100e (b). Some cells contain both proteins (arrow). A cyst (C) contains keratin, but not S 100~. x 500 Fig. 14a, b. Adjacent Paraplast sections on sheep pituitary gland immunostained for keratin (a) and S 100c~(b). Some cells are immunoreactive for both proteins (arrow). x 500 Fig. 15a, b. Frozen section of bull pituitary gland immunostained with monoclonal antibody to GFAP. Immunoreactivity detected in intermediate lobe (a, /), but not in anterior lobe (a, A). Stellate-shaped immunoreactive cells are seen in intermediate lobe (b). a x100;b x200
surrounded by unit membranes and contained electron dense material. The membranes of these components resembled the densely stained membranes observed at the apical side of the FS cell, but their ultrastructural characteristics were different from smooth endoplasmic reticulure (SER). Some FS cells either contained a few of these membranous components, or lacked them entirely. However, immunostainability to keratin had no relation to the existence of the above-mentioned special membranous components. The marginal layer cells of the bull pituitary gland, which demonstrated strong immunoreactivity for keratin, contained bundles of filaments (Fig. 21).
Discussion
The review of Moll et al. (1982) indicates that 19 different cytokeratin polypeptides with molecular weights ranging from 40 000-70 000 daltons have been identified in humans. Keratin, which is one of the intermediate filament proteins, has been detected immunohistochemically, especially in the human pituitary gland. Asa and
her colleagues showed that keratin was identified in the cells lining cystic structures in the pars intermedia (Asa et al. 1981), and in squamous cell nests in the pars tuberalis (Asa et al. 1983). H6fler et al. (1984a) reported that G H cells and A C T H cells were strongly positive for keratin, while P R L cells were weakly immunoreactive, and FS cells were unreactive. Ironside et al. (1987), using a polyclonal antibody to keratin, failed to detect keratin immunoreactivity in hormone-producing cells, but found immunostainability only in the follicular structures of the pars intermedia. In the present studies using anterior pituitary glands of bull, goat, and sheep, immunoreactive keratin was detected in the marginal layer cells and cystic structures. The immunoreactivity of cystic structures is in accordance with the findings in the human pituitary gland of Asa et al. (1981 and 1983) and of Ironside et al. (1987). Interestingly, those investigators used different monoclonal and polyclonal antibodies to keratin, which caused some disharmonious results of the immunohistochemical studies. Because keratin consists of several different types of polypeptide, immunoreactivities are different when using various antibodies (Moll et al. 1982). The antibody used in this study
256
Fig. 16. Cluster of bull FS cells in which one cell corresponds to an immunostained cell (arrow). Inset shows adjacent section immunostained with antibody to keratin. CF Colloidal follicle. x4050. Inset: x660 Fig. 17. Ultrastructure of goat FS ceils. Two immunoreactive cells correspond to cells present on electron micrograph demarcated by arrow and arrowhead. FS cells are clustered and display elongated cytoplasmic processes. Inset shows adjacent section immunostained with antibodies to keratin. • Inset: •
exhibited immunoreactivities in various epithelia such as stratified squamous epithelia, sweat glands, the urinary tract, small intestine, and bile duct making this antibody useful for detecting keratin in various epithelial cells and tissues by immunohistochemistry (not shown). G F A P , which is another type o f intermediate filament protein, also has been detected in the h u m a n pituitary
gland (Velasco et al. 1982; H6fler et al. 1984b; Morris and Hitchcock 1985). H6fler et al. (1984b) reported that G F A P and S 100 protein were detected in FS cells of the h u m a n anterior pituitary gland, and that 50% of the S 100-reactive FS cells reacted with the antibody to G F A P . In the present study, there were no immunoreacrive G F A P cells in the anterior pituitary glands of all
257
Fig. 18. Fine structure of bull FS cells illustrated with filaments (arrow) in cytoplasm, x 30000 Fig. 19. Fine structure of bull FS cells illustrating intercellular junctions (arrow) between neighboring FS (FS) and glandular (G) cells. Whorls of dense membrane with included cytoplasmic material appear at apical side of FS cells (arrowhead). Slender and elliptical membranous components containing electron-dense material are seen in FS cell cytoplasm (inset). x 15000. Inset: • 99000 3 species. The present result is in accordance with that of Stoeckel et al. (1981) who observed a strong reaction of pars intermedia stellate cells in the rat with antiserum to G F A P , while t h e reaction was negative in the FS cells of the remainder of the adenohypophysis. S 100 protein is one of the useful markers of FS cells in rat pituitary glands. On the contrary in the goat, FS
cells did not react with the antibody to S 100 protein, while G H cells were intensely and T S H and P R L cells faintly immunostained (Shirasawa et al. 1984). G H cells of sheep also contained S 100 protein (Shirasawa 1986). However, FS cells of goat and sheep pituitary glands reacted with an antibody to S 100c~ protein (Shirasawa 1986). S 100 protein is composed of c~ and ~ subunits
258
Fig. 20a, b. Electron-micrographs of sheep FS cells (FS) in which intercellular junctions are welldeveloped (a, arrow). Special membranous components containing electron-dense material which differ from the smooth endoplasmic reticulum (SER) are visible (b, arrowhead), a x 7500; b x 40500
(Isobe et al. 1983). In the present observations, the distribution of S 100~-reactive cells was not uniform in goat and sheep anterior pituitary glands. Some FS cells reacted with both antibodies to S 100c~ and keratin; however, keratin-positive cells did not always exhibit S 100c~ immunoreactivity. In goat and sheep pituitary glands,
S 100~ identified a smaller population of FS cells than keratin. There are several hypotheses about the origin of FS cells in the anterior pituitary gland. S 100 protein was first thought to be unique to the nervous tissues. It has been suggested that FS cells are of neuroectodermal ori-
259
Fig. 21. Bundles of filaments (arrowhead) observed in cytoplasm of bull marginal cells. intercellular junctions (arrow) are well-developed between neighboring cells. • 17730
gin and glial-like, because of their immunoreactivities for S 100 protein (Cocchia and Miani 1980; N a k a j i m a et al. 1980) and for G F A P (Velasco et al. 1982; Morris and Hitchcock 1985). In the fetal rat pituitary gland, follicles originate f r o m the forming epithelial cords (Yoshida 1966; Svalander 1974). Recently, it has been demonstrated that S 100 protein also exists in non-nervous tissues. According to H a i m o t o et al. (1987), S 100~ protein is contained in various epithelial cells of nonnervous tissues such as the sweat gland, m a m m a r y gland ducts, gall bladder, small intestine, and urinary bladder. Keratin also exists in the above-mentioned epithelial cells (Moll et al. 1982) that contain the S 100c~ protein. In the present observations, immunohistochemical characteristics of FS cells are in h a r m o n y with those of epithelial cells. The anterior pituitary FS cells of animals in the order Artiodactyla exhibit immunoreactive keratin and S 100c~, while there is no immunostainability to G F A P . It is suggested from this work that keratin-positive marginal cells, cystic structures, and FS cells express characteristics of epithelial cells more intensely than those of glial cells. The ultrastructural characteristics of the slender and elliptical m e m b r a n o u s components resemble those of the densely-stained m e m b r a n e which appears at the apical side of the FS cells. Thus, the elliptical profiles appear to be a part of the apical membrane. Forte et al. (1977) observed the dense m e m b r a n e in oxyntic cells, stimulated with histamine, at the early phase of the return
to resting state. It is possible that the dense m e m b r a n e and the elliptical m e m b r a n o u s profiles observed in this study are related to the m e m b r a n e uptake observed in oxyntic cells by Forte et al. (1977).
Acknowledgements. The author thanks Professor H. Ishikawa, Dr. H. Nogami (The Jikei University School of Medicine, Department of Anatomy), Dr. F. Nakamura (Hokkaido University), and Dr. N. Shirasawa (Shinsyu University), for their guidance, and Dr. D.C. Herbert (University of Texas Health Science Center at San Antonio, Department Cellular and Structural Biology), for his careful reading of the manuscript.
References Asa SL, Kovacs K, Bilbao JM, Penz G (1981) Immunohistochemical localization of keratin in craniopharyngiomas and squamous cell nests of the human pituitary. Acta Neuropathol (Berl) 54:257 260 Asa SL, Kovacs K, Bilbao JM (1983) The pars tuberalis of the human pituitary: a histologic, immunohistochemical, ultrastructural and immunoelectron microscopic analysis. Virchows Arch [A] 399: 49-59 Beauvillain JC, Tramu G, Dubois MP (1975) Characterization by different techniques of adrenocorticotropin and gonadotropin producing cells in lerot pituitary (Eliomys quercinus): a superimposition technique and an immunocytochemical technique. Cell Tissue Res 158 : 301-317 Cocchia D, Miani N (1980) Immunocytochemical localization of the brain-specific S-100 protein in the pituitary gland of adult rat. J Neurocytol 9:771 782
260 Dingemans KP, Feltkamp CA (1972) Nongranulated cells in the mouse adenohypophysis. Z Zellforsch 124:387 405 Farquhar MG (1957) "Corticotrophs" of the rat adenohypophysis as revealed by electron microscopy. Anat Rec 127:291 Forte TM, Machen TE, Forte GJ (1977) Ultrastructural changes in oxyntic cells associated with secretory function: a membranerecycling hypothesis. Gastroenterology 73 : 941-955 Fukuda T (1973) Agranular stellate cells (so-called follicular cells) in human fetal and adult adenohypophysis and in pituitary adenomas. Virchows Arch [A] 359:19-30 Girod C, Trouillas J, Dubois MP (1985) Immunocytochemical localization of S-100 protein in stellate cells (folliculostellate cells) of the anterior lobe of the normal human pituitary. Cell Tissue Res 241 : 505-511 Haimoto H, Hosoda S, Kato K (1987) Differential distribution of immunoreactive S100-c~and Sl00-fl proteins in normal nonnervous human tissues. Lab Invest 57:489-498 H6fler H, Denk H, Walter GF (1984a) Immunohistochemicaldemonstration of cytokeratins in endocrine cells of the human pituitary gland and in pituitary adenomas. Virchows Arch [A] 404:359 368 H6fler H, Walter GF, Denk H (1984b) Immunohistochemistry of folliculo-stellate cells in normal human adenohypophyses and in pituitary adenomas. Acta Neuropathol (Berl) 65:35-40 Horvath E, Kovacs K, Penz G, Ezrin C (1974) Origin, possible function and fate of "follicular cells" in the anterior lobe of the human pituitary: An electron microscopic study. Am J Pathol 77:199 212 Hsu SM, Rain L, Fanger H (1981) Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase technique : A comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29 : 577 580 Ironside JW, Royds JA, Jefferson AA, Timperley WR (1987) Immunolocalisation of cytokeratins in the normal and neoplastic human pituitary gland. J Neurol Neurosurg Psychiatry 50: 5765 Isobe T, Ishioka N, Masuda T, Takahashi Y, Ganno S, Okuyama T (1983) A rapid separation of S100 subunits by high performance liquid chromatography: the subunit compositions of S100 proteins. Bioehem Int 6 :419-426 Lauriola L, Cocchia D, Sentinelli S, Maggiano N, Maira G, Michetti F (1984) Immunohistochemiealdetection of folliculo-stellate cells in human pituitary adenomas. Virchows Arch [B] 47:189-197 Lazarides E (1980) Intermediate filaments as mechanical integrators of cellular space. Nature 283:249-256 Millonig G (1962) Further observation on a phosphate buffer for osmium solutions in fixation, in: Electron Microscopy, Fifth International Congress for Electron Microscopy, vol 2. Academic Press, New York, p 8 Moll R, Franke WW, Schiller DL, Greiger B, Krepler R (1982) The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell 31:11-24 Morris C, Hitchcock E (1985) Immunocytochemistry of folliculostellate cells of normal and neoplastic human pituitary gland. J Clin Pathol 38:481 488 Nakajima T, Yamaguchi H, Takahashi K (1980) S 100 protein
in folliculostellate cells of the rat pituitary anterior lobe. Brain Res 191:523-531 Nakazato Y, Yamaguchi H, Takahashi K, Mori T (1982) Immunohistochemical localization of nervous tissue proteins S-100, GFAP and astroprotein in extraneural tissue, In: Peters H (ed) Protides of biological fluids. Pergamon Press, Oxford, pp 47 50 Salazar H (1968) Ultrastructural evidence for the existence of a non-secretory, sustentacular cell in the human adenohypophysis. Anat Rec 160:419-420 Schlegel R, Banks-Schlegel S, Pinkus GS (1980) Immunohistochemical localization of keratin in normal human tissues. Lab Invest 42:91-96 Shirasawa N (1986) Differentiation of pituitary folliculo-stellate cells. In: Yoshimura F, Gorbman A (eds) Pars distalis of the pituitary gland: structure, function and regulation. Elsevier, Amsterdam, pp 175 181 Shirasawa N, Yoshimura F (1982) Immunohistochemical and electron microscopical studies of mitotic adenohypophysial ceils in different ages of rats. Anat Embryol 165:51-61 Shirasawa N, Kihara H, Yamaguchi S, Yoshimura F (1983) Pituitary folliculo-stellate cells immunostained with S-100 protein antiserum in postnatal, castrated and thyroidectomized rat. Cell Tissue Res 231:235-249 Shirasawa N, Yamaguchi S, Yoshimura F (1984) Granulated folliculo-stellate cells and growth hormone cells immunostained with anti-S 100 protein serum in the pituitary glands of the goat. Cell Tissue Res 237:7-14 Shirasawa N, Mitui J, Mera F, Ishikawa H (1988) Comparative immunohistochemical study of the mammalian pituitary cells stained with antisera against S-100 protein and non-neuronal enolase. Biomed Res 9:477-487 Stoeckel ME, Schmitt G, Porte A (1981) Fine structure and cytochemistry of the mammalian pars intermedia. In: Peptides of pars intermedia. Ciba Foundation Symposium 81. Pitman, London, pp 101-127 Svalander C (1974) Ultrastructure of the fetal rat adenohypophysis. Acta Endocrinol (Copenh) 76 [Suppl 188]:1-133 Takahashi K, Isobe T, Ohtsuki Y, Akagi T, Sonobe H, Okuyama T (1984) Immunohistochemical study on the distribution of :~ and /~ subunits of S-100 protein in human neoplasm and normal tissues. Virchow Arch [B] 45 : 385-396 Velasco ME, Roessmann U, Gambetti P (1982) The presence of glial fibriliary acidic protein in the human pituitary gland. J Neuropathol Exp Neurol 41:150-163 Yoshida Y (1966) Electron microscopy of the anterior pituitary gland under normal and different experimental conditions. Meth Achievm Exp Pathol 1:439-454 Yoshimura F, Nogami H (1981) Fine structural criteria for identifying rat corticotrophs. Cell Tissue Res 219:221 228 Yoshimura F, Soji T, Sato S, Yokoyama M (1977) Development and differentiation of rat pituitary follicular cells under normal and some experimental conditions with special references to an interpretation of renewal cell system. Endocrinol Jpn 24: 435-449 Yoshimura F, Nogami H, Shirasawa N, Yashiro T (1981) A whole range of fine structural criteria for immunohistochemically identified LH cells in rats. Cell Tissue Res 217:1-10
Cell&Tissue Research 9 Springer-Verlag 1992
Immunohistochemical localization of keratin in bull, goat, and sheep anterior pituitary glands Takaki Shimada Department of Anatomy,Jikei UniversitySchool of Medicine, 3-25-8 Nishi-Shinbashi, Minato-ku,Tokyo, 105 Japan Received January 31, 1991 / Accepted September 10, 1991
Summary. Immunohistochemical localization of keratin, an intermediate filament protein, was studied in bull, goat, and sheep anterior pituitary glands, i.e., in animals of the order Artiodactyla. Strong immunoreactivity was detected in the cells of the marginal layer of bull and goat, as well as in cysts or large follicles in the anterior lobe of all 3 species. In addition, a number of stellateshape cells were immunoreactive for keratin and were distributed throughout the anterior lobe. The localization of keratin-positive cells in light-microscopic preparations correlated precisely with the localization of folliculo-stellate cells in adjacent ultrathin sections. In ultrastructural studies, many slender and elliptical membranous components which were different from smooth endoplasmic reticulum were observed in the cytoplasm of the some keratin-positive cells. Some of the folliculostellate cells in the 3 species were also immunoreactive for the c~ subunit of S-100 protein, which exists in some epithelial cells. On the other hand, immunolocalization of glial fibrillary acidic protein, a glial cell marker, could not be demonstrated in the anterior pituitary glands of the 3 species studied. These results suggest that keratinpositive folliculo-stellate cells express epithelial-like characteristics. Key words: Pituitary gland, pars anterior (distalis)- Folliculo-stellate cells - Keratin S 100 protein - Cow, goat, sheep (Artiodactyla)
Folliculo-stellate (FS) cells of the vertebrate pituitary gland have been studied both by ultrastructural and immunohistochemical methods. Ultrastructurally, for the most part, FS cells lack secretory granules and tend to be clustered. They form follicles which bear microvilli and/or cilia on their apical cell surfaces. Different hypotheses have been proposed regarding the functional roles of FS cells, such as ACTH secretory
cells (Farquhar 1957), supportive cells (Salazar 1968), dedifferentiated glandular cells (Horvath et al. 1974), phagocytic cells (Dingemans and Feltkamp 1972; Fukuda 1973), and stem cells (Yoshimura et al. 1977). However, their origin and functions have not been fully clarified. By immunohistochemistry, FS cells exhibit immunoreactivity for several different proteins. In particular, the S 100 protein has been a useful marker in the adult rat (Cocchia and Miani 1980; Nakajima et al. 1980; Shirasawa et al. 1983) and in the human (Nakazato et al. 1982; Lauriola et al. 1984; Takahashi et al. 1984; Girod et al. 1985). S 100 protein, however, is not a universal marker for FS cells in all species of animals (Shirasawa et al. 1984, 1988). In the goat, pituitary growth hormone (GH) cells were intensely immunostained with S 100 protein antiserum (Shirasawa et al. 1984); however, FS cells were immunoreactive for the S 100e component (Shirasawa 1986), which is the subunit of S 100 protein (Isobe et al. 1983). Furthermore, FS cells in the human pituitary were found to be immunoreactive for both S 100 protein and glial fibrillary acidic protein (GFAP) (Velasco et al. 1982; Morris and Hitchcock 1985). GFAP, which is contained in the human FS cells, is one of the intermediate filament proteins. Intermediate filaments are well-known as 10-rim filaments and as proteins forming a part of the cytoskeleton. Biochemical and immunological analyses have enabled intermediate filaments to be classified into 5 main groups (Lazarides 1980). Keratin is another type of intermediate filament protein. Keratin proteins have been identified in many diverse types of epithelia (Schlegel et al. 1980). In the study reported here, immunoreactive keratin was detected in bull, goat and sheep pituitary glands by immunohistochemistry. Pituitary keratin-containing cells were compared with pituitary hormone-containing cells, and with cells containing S 100~ protein and GFAP, the markers of the FS cells. Furthermore, the ultrastructure of the keratin-positive cells was observed using the superimposition technique.
252
Materials and methods
Tissues and superimposition technique Adult male bull (n=5), goat (n=3), and sheep (n=3) pituitary glands were fixed in sublimate-formalin (saturated sublimate solution:formalin, 7:3) for 16 h at 4~ C. After dehydration in an ethanol series, they were embedded in Paraplast, and sectioned at a thickness of 3-4 pm. In addition, frozen sections of bull pituitary gland were prepared as follows. They were dipped into OCT compound (Tissue Tek, Miles Scientific, Naperville, USA), and frozen in acetone cooled with dry ice. After sectioning using a cryostat, (5-6 gm in thickness), they were fixed for 10 min in methanol at - 2 0 ~ and then were fixed additionally in acetone at - 2 0 ~ for 30 sec. Both the Paraplast and cryostat sections were subsequently used for immunohistochemical studies. For electron-microscopic studies, small pieces of the pituitary gland from each animal were fixed in 2% glutaraldehyde, 0.1 M phosphate buffer, pH 7.4, at 4~ C for 3 h, and then postfixed with 1% osmic acid in Millonig solution (Millonig 1962), pH 7.4, at 4~ C for 2 h. After dehydration in an ethanol series, the pituitary glands were embedded in Epon-Araldite. Ultrathin sections were prepared for electron-microscopic observations, with adjacent thick sections cut for light-microscopic immunostaining, according to the superimposition technique described by Yoshimura and Nogami (1981). This procedure is a slight modification of the method of Beauvillain et al. (1975). Epon-Araldite thick sections were treated with a 14% methanol solution of sodium methoxide, and then were washed with methanol. Subsequently, the sections were immersed in a 1% aqueous solution of hydrogen peroxide for bleaching. The ultrathin sections were stained with uranyl acetate and lead citrate, and observed with a JEOL 100B electron microscope.
Immunohistochemical studies Immunohistochemical staining was carried out by the avidin-biotin-peroxidase complex method (Hsu et al. 1981). The primary antibodies used were anti-human skin keratin (Bio-Science Products, AG, Switzerland), a monoclonal antibody against human GFAP (Amersham International plc, England), and a polyclonal antibody against cow GFAP (Dakopatts, Denmark). The specificity of the anti-human skin keratin antibody was confirmed by the absorption test described below. In addition, other primary antibodies used were: (1) polyclonal antibodies against ovine GH, ovine prolactin, porcine ACTH 1-39, rat LHfl, and rat TSHfl antisera; and (2) a monoclonal antibody against the ~ subunit of S-100 protein (Japan Immunoresearch Lab. Co., Japan). Their specificities have been described previously (Yoshimura et al. 1981 ; Shirasawa and Yoshimura 1982; Shirasawa 1986).
Absorption tests Absorption tests were carried out on serial Paraplast sections of normal pituitary glands from the 3 species to check the specificity
Fig. l a, b. Adjacent Paraplast sections of bull anterior pituitary gland immunostained using anti-keratin antibody (a), and preabsorbed antibody (b). Immunoreactivity was lost by preabsorption. C Cyst. x 200
of the human skin keratin antibody. The antibody (1:200) was incubated with 10 ~tg of purified human skin keratin protein (Sigma, St. Louis, Mo., USA) at 4~ C for 3 days. As shown in Fig. 1, sections stained with the preabsorbed antibody did not demonstrate immunoreactivity. The antibody to keratin preincubated with 10 gg of bovine S 100 protein (Wako Chemical Industries Ltd., Osaka, Japan) and the antibodies to GFAP preincubated with 10 gg of purified human skin keratin protein did not lose their immunoreactivity.
Results
Ligh t-microscopic findings A n u m b e r o f cells reacted with the a n t i b o d y to keratin in pituitary glands o f bull and goat (Figs. 2, 3, respectively). These keratin-positive cells were stellate in shape with elongated cytoplasmic processes, and had a tendency to f o r m clusters. T h e y were distributed evenly t h r o u g h o u t the anterior lobe. I m m u n o r e a c t i v i t y was noted only in the c y t o p l a s m and the staining intensity was variable. In the anterior lobe o f sheep, similar stellate-shaped cells i m m u n o r e a c t i v e for keratin were also observed (Fig. 4). However, these cells were distributed singly t h r o u g h o u t the gland and there was little tendency to f o r m clusters. In pituitary glands o f bull and goat, the marginal layer o f the pars anterior was also i m m u n o reactive for keratin (Fig. 5) and consisted o f a layer o f cells which varied in thickness. Some ovine marginal cells possessed i m m u n o r e a c t i v i t y for keratin as well (Fig. 6). I m m u n o s t a i n i n g o f these marginal layer cells was stronger t h a n that observed in the stellate cells o f the bull and g o a t anterior lobe. Cysts or large follicles, which often contained colloid, were frequently observed in the anterior lobe in all o f the animals studied. Cells lining the cysts were strongly immunoreactive for keratin (Figs. 5, 7). The i m m u n o r e activity was observed in the cytoplasm only and the staining intensity was similar to that f o u n d in the marginal cells. The distribution patterns, cell size, and shapes o f keratin-positive cells were c o m p a r e d by i m m u n o h i s t o c h e m istry with that o f the pituitary h o r m o n e - p r o d u c i n g cells, G H , P R L , L H , TSH, and A C T H on adjacent sections (Figs. 8-12). G H , L H , and T S H cells were oval in shape, whereas keratin-positive cells were stellate-shaped (Figs. 8, 10, 11). P R L and A C T H cells were p o l y g o n a l ; however, their patterns o f distribution were different f r o m that o f keratin-positive cells (Figs. 9, 12). Cells i m m u n o r e a c t i v e with anti-S 100c~ a n t i b o d y were scattered t h r o u g h o u t the anterior pituitary glands o f the 3 animals studied. In the bull anterior pituitary gland, stellate-shaped cells and endothelial cells o f b l o o d vessels showed immunoreactivity, while in the g o a t and sheep anterior pituitary glands, immunopositive cells were stellate-shaped with elongated cytoplasmic processes (Figs. 13 b, 14b). I m m u n o r e a c t i o n p r o d u c t s were f o u n d in their c y t o p l a s m and nucleus. Their distribution was not u n i f o r m and they had a tendency to accumulate in the peripheral area o f the sheep anterior lobe, and in the area beneath the marginal layer o f the bull anteri-
253
Figs. 2--7. Paraplast sections immunostained for keratin. A number of stellate-shaped cells (Figs. 2, 3, 4), marginal cells (Figs. 5, 6) and cysts (Figs. 5, 7) exhibit immunoreactivity. ML Marginal layer', C cyst; I intermediate lobe; A anterior lobe. Figs. 2, 5 bull; Fig. 3 goat; Figs. 4, 6, 7 sheep. Figs. 2, 3, 4 x 320; Fig. 5 x 160; Figs. 6, 7 (counter-stained with hematoxylin) • 320
or lobe. There were few S 100~-positive stellate-shaped cells in the other regions of the gland. Some, but not all, of the marginal cells also contained S 100~ protein in anterior pituitary glands of bull and goat. Adjacent sections immunostained with anti-S 100~ and anti-keratin antibodies revealed that some ceils contained both S 100e and keratin proteins (Figs. 13, 14), while others were immunostained for only one of the proteins. There were no immunoreactive G F A P cells detected in anterior pituitary glands of the 3 species by the use of monoclonal or polyclonal anti-GFAP antibodies (Fig. 15a). However, in intermediate lobes of bull and goat, some stellate-shaped cells exhibited immunostainability for G F A P (Fig. 15b). The staining intensity was the same with either the monoclonal or polyclonal antibodies.
Electron-microscopicfindings of keratin-positive cells Ultrastructural characteristics of cells immunoreactive for keratin were observed using the superimposition technique (Figs. 16, 17). The ultrastructural counterparts
of the keratin-positive cells were agranular and irregular in shape, with elongated cytoplasmic processes and were identified as folliculo-stellate cells. The cytoplasmic processes contacted the parenchymal basement membrane that separated the glandular epithelium from the connective tissue. Keratin-positive cells were often clustered and formed follicles. In the bull anterior lobe, the follicular lumen was filled with colloid (Fig. 16) - an observation which was less frequently made in the anterior lobe of the goat. These cells had many microvilli which were closely interdigitated with each other. Intercellular junctions were well-developed between neighboring agranular cells (Figs. 18, 19, 20 a), and they were also observed between keratin-positive cells and granular cells (Fig. 19). Filamentous components were observed in the cytoplasm of the keratin-positive cells (Fig. 18). At the apical side of the FS cells, densely stained membranes were present in the adjacent FS cells. They were interdigitated with each other, and often presented round profiles with included cytoplasmic components (Figs. 18, 19). Some cells contained numerous slender and elliptical membranous components (Figs. 19, 20b). They were
254
Figs. 8--12. Adjacent Paraplast sections of bull anterior pituitary gland immunostained for keratin
(Figs. 8-12a), GH (Fig. 8b), prolactin (Fig. 9b), LH (Fig. 10b), TSH (Fig. l l b ) , and ACTH (Fig. 12b). Immunoreactive cells for keratin are different from hormone-producing cells (counter-stained with hematoxylin), x 500
255
Fig. 13a, b. Adjacent Paraplast sections of goat pituitary immunostained for keratin (a), and for S 100e (b). Some cells contain both proteins (arrow). A cyst (C) contains keratin, but not S 100~. x 500 Fig. 14a, b. Adjacent Paraplast sections on sheep pituitary gland immunostained for keratin (a) and S 100c~(b). Some cells are immunoreactive for both proteins (arrow). x 500 Fig. 15a, b. Frozen section of bull pituitary gland immunostained with monoclonal antibody to GFAP. Immunoreactivity detected in intermediate lobe (a, /), but not in anterior lobe (a, A). Stellate-shaped immunoreactive cells are seen in intermediate lobe (b). a x100;b x200
surrounded by unit membranes and contained electron dense material. The membranes of these components resembled the densely stained membranes observed at the apical side of the FS cell, but their ultrastructural characteristics were different from smooth endoplasmic reticulure (SER). Some FS cells either contained a few of these membranous components, or lacked them entirely. However, immunostainability to keratin had no relation to the existence of the above-mentioned special membranous components. The marginal layer cells of the bull pituitary gland, which demonstrated strong immunoreactivity for keratin, contained bundles of filaments (Fig. 21).
Discussion
The review of Moll et al. (1982) indicates that 19 different cytokeratin polypeptides with molecular weights ranging from 40 000-70 000 daltons have been identified in humans. Keratin, which is one of the intermediate filament proteins, has been detected immunohistochemically, especially in the human pituitary gland. Asa and
her colleagues showed that keratin was identified in the cells lining cystic structures in the pars intermedia (Asa et al. 1981), and in squamous cell nests in the pars tuberalis (Asa et al. 1983). H6fler et al. (1984a) reported that G H cells and A C T H cells were strongly positive for keratin, while P R L cells were weakly immunoreactive, and FS cells were unreactive. Ironside et al. (1987), using a polyclonal antibody to keratin, failed to detect keratin immunoreactivity in hormone-producing cells, but found immunostainability only in the follicular structures of the pars intermedia. In the present studies using anterior pituitary glands of bull, goat, and sheep, immunoreactive keratin was detected in the marginal layer cells and cystic structures. The immunoreactivity of cystic structures is in accordance with the findings in the human pituitary gland of Asa et al. (1981 and 1983) and of Ironside et al. (1987). Interestingly, those investigators used different monoclonal and polyclonal antibodies to keratin, which caused some disharmonious results of the immunohistochemical studies. Because keratin consists of several different types of polypeptide, immunoreactivities are different when using various antibodies (Moll et al. 1982). The antibody used in this study
256
Fig. 16. Cluster of bull FS cells in which one cell corresponds to an immunostained cell (arrow). Inset shows adjacent section immunostained with antibody to keratin. CF Colloidal follicle. x4050. Inset: x660 Fig. 17. Ultrastructure of goat FS ceils. Two immunoreactive cells correspond to cells present on electron micrograph demarcated by arrow and arrowhead. FS cells are clustered and display elongated cytoplasmic processes. Inset shows adjacent section immunostained with antibodies to keratin. • Inset: •
exhibited immunoreactivities in various epithelia such as stratified squamous epithelia, sweat glands, the urinary tract, small intestine, and bile duct making this antibody useful for detecting keratin in various epithelial cells and tissues by immunohistochemistry (not shown). G F A P , which is another type o f intermediate filament protein, also has been detected in the h u m a n pituitary
gland (Velasco et al. 1982; H6fler et al. 1984b; Morris and Hitchcock 1985). H6fler et al. (1984b) reported that G F A P and S 100 protein were detected in FS cells of the h u m a n anterior pituitary gland, and that 50% of the S 100-reactive FS cells reacted with the antibody to G F A P . In the present study, there were no immunoreacrive G F A P cells in the anterior pituitary glands of all
257
Fig. 18. Fine structure of bull FS cells illustrated with filaments (arrow) in cytoplasm, x 30000 Fig. 19. Fine structure of bull FS cells illustrating intercellular junctions (arrow) between neighboring FS (FS) and glandular (G) cells. Whorls of dense membrane with included cytoplasmic material appear at apical side of FS cells (arrowhead). Slender and elliptical membranous components containing electron-dense material are seen in FS cell cytoplasm (inset). x 15000. Inset: • 99000 3 species. The present result is in accordance with that of Stoeckel et al. (1981) who observed a strong reaction of pars intermedia stellate cells in the rat with antiserum to G F A P , while t h e reaction was negative in the FS cells of the remainder of the adenohypophysis. S 100 protein is one of the useful markers of FS cells in rat pituitary glands. On the contrary in the goat, FS
cells did not react with the antibody to S 100 protein, while G H cells were intensely and T S H and P R L cells faintly immunostained (Shirasawa et al. 1984). G H cells of sheep also contained S 100 protein (Shirasawa 1986). However, FS cells of goat and sheep pituitary glands reacted with an antibody to S 100c~ protein (Shirasawa 1986). S 100 protein is composed of c~ and ~ subunits
258
Fig. 20a, b. Electron-micrographs of sheep FS cells (FS) in which intercellular junctions are welldeveloped (a, arrow). Special membranous components containing electron-dense material which differ from the smooth endoplasmic reticulum (SER) are visible (b, arrowhead), a x 7500; b x 40500
(Isobe et al. 1983). In the present observations, the distribution of S 100~-reactive cells was not uniform in goat and sheep anterior pituitary glands. Some FS cells reacted with both antibodies to S 100c~ and keratin; however, keratin-positive cells did not always exhibit S 100c~ immunoreactivity. In goat and sheep pituitary glands,
S 100~ identified a smaller population of FS cells than keratin. There are several hypotheses about the origin of FS cells in the anterior pituitary gland. S 100 protein was first thought to be unique to the nervous tissues. It has been suggested that FS cells are of neuroectodermal ori-
259
Fig. 21. Bundles of filaments (arrowhead) observed in cytoplasm of bull marginal cells. intercellular junctions (arrow) are well-developed between neighboring cells. • 17730
gin and glial-like, because of their immunoreactivities for S 100 protein (Cocchia and Miani 1980; N a k a j i m a et al. 1980) and for G F A P (Velasco et al. 1982; Morris and Hitchcock 1985). In the fetal rat pituitary gland, follicles originate f r o m the forming epithelial cords (Yoshida 1966; Svalander 1974). Recently, it has been demonstrated that S 100 protein also exists in non-nervous tissues. According to H a i m o t o et al. (1987), S 100~ protein is contained in various epithelial cells of nonnervous tissues such as the sweat gland, m a m m a r y gland ducts, gall bladder, small intestine, and urinary bladder. Keratin also exists in the above-mentioned epithelial cells (Moll et al. 1982) that contain the S 100c~ protein. In the present observations, immunohistochemical characteristics of FS cells are in h a r m o n y with those of epithelial cells. The anterior pituitary FS cells of animals in the order Artiodactyla exhibit immunoreactive keratin and S 100c~, while there is no immunostainability to G F A P . It is suggested from this work that keratin-positive marginal cells, cystic structures, and FS cells express characteristics of epithelial cells more intensely than those of glial cells. The ultrastructural characteristics of the slender and elliptical m e m b r a n o u s components resemble those of the densely-stained m e m b r a n e which appears at the apical side of the FS cells. Thus, the elliptical profiles appear to be a part of the apical membrane. Forte et al. (1977) observed the dense m e m b r a n e in oxyntic cells, stimulated with histamine, at the early phase of the return
to resting state. It is possible that the dense m e m b r a n e and the elliptical m e m b r a n o u s profiles observed in this study are related to the m e m b r a n e uptake observed in oxyntic cells by Forte et al. (1977).
Acknowledgements. The author thanks Professor H. Ishikawa, Dr. H. Nogami (The Jikei University School of Medicine, Department of Anatomy), Dr. F. Nakamura (Hokkaido University), and Dr. N. Shirasawa (Shinsyu University), for their guidance, and Dr. D.C. Herbert (University of Texas Health Science Center at San Antonio, Department Cellular and Structural Biology), for his careful reading of the manuscript.
References Asa SL, Kovacs K, Bilbao JM, Penz G (1981) Immunohistochemical localization of keratin in craniopharyngiomas and squamous cell nests of the human pituitary. Acta Neuropathol (Berl) 54:257 260 Asa SL, Kovacs K, Bilbao JM (1983) The pars tuberalis of the human pituitary: a histologic, immunohistochemical, ultrastructural and immunoelectron microscopic analysis. Virchows Arch [A] 399: 49-59 Beauvillain JC, Tramu G, Dubois MP (1975) Characterization by different techniques of adrenocorticotropin and gonadotropin producing cells in lerot pituitary (Eliomys quercinus): a superimposition technique and an immunocytochemical technique. Cell Tissue Res 158 : 301-317 Cocchia D, Miani N (1980) Immunocytochemical localization of the brain-specific S-100 protein in the pituitary gland of adult rat. J Neurocytol 9:771 782
260 Dingemans KP, Feltkamp CA (1972) Nongranulated cells in the mouse adenohypophysis. Z Zellforsch 124:387 405 Farquhar MG (1957) "Corticotrophs" of the rat adenohypophysis as revealed by electron microscopy. Anat Rec 127:291 Forte TM, Machen TE, Forte GJ (1977) Ultrastructural changes in oxyntic cells associated with secretory function: a membranerecycling hypothesis. Gastroenterology 73 : 941-955 Fukuda T (1973) Agranular stellate cells (so-called follicular cells) in human fetal and adult adenohypophysis and in pituitary adenomas. Virchows Arch [A] 359:19-30 Girod C, Trouillas J, Dubois MP (1985) Immunocytochemical localization of S-100 protein in stellate cells (folliculostellate cells) of the anterior lobe of the normal human pituitary. Cell Tissue Res 241 : 505-511 Haimoto H, Hosoda S, Kato K (1987) Differential distribution of immunoreactive S100-c~and Sl00-fl proteins in normal nonnervous human tissues. Lab Invest 57:489-498 H6fler H, Denk H, Walter GF (1984a) Immunohistochemicaldemonstration of cytokeratins in endocrine cells of the human pituitary gland and in pituitary adenomas. Virchows Arch [A] 404:359 368 H6fler H, Walter GF, Denk H (1984b) Immunohistochemistry of folliculo-stellate cells in normal human adenohypophyses and in pituitary adenomas. Acta Neuropathol (Berl) 65:35-40 Horvath E, Kovacs K, Penz G, Ezrin C (1974) Origin, possible function and fate of "follicular cells" in the anterior lobe of the human pituitary: An electron microscopic study. Am J Pathol 77:199 212 Hsu SM, Rain L, Fanger H (1981) Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase technique : A comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29 : 577 580 Ironside JW, Royds JA, Jefferson AA, Timperley WR (1987) Immunolocalisation of cytokeratins in the normal and neoplastic human pituitary gland. J Neurol Neurosurg Psychiatry 50: 5765 Isobe T, Ishioka N, Masuda T, Takahashi Y, Ganno S, Okuyama T (1983) A rapid separation of S100 subunits by high performance liquid chromatography: the subunit compositions of S100 proteins. Bioehem Int 6 :419-426 Lauriola L, Cocchia D, Sentinelli S, Maggiano N, Maira G, Michetti F (1984) Immunohistochemiealdetection of folliculo-stellate cells in human pituitary adenomas. Virchows Arch [B] 47:189-197 Lazarides E (1980) Intermediate filaments as mechanical integrators of cellular space. Nature 283:249-256 Millonig G (1962) Further observation on a phosphate buffer for osmium solutions in fixation, in: Electron Microscopy, Fifth International Congress for Electron Microscopy, vol 2. Academic Press, New York, p 8 Moll R, Franke WW, Schiller DL, Greiger B, Krepler R (1982) The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell 31:11-24 Morris C, Hitchcock E (1985) Immunocytochemistry of folliculostellate cells of normal and neoplastic human pituitary gland. J Clin Pathol 38:481 488 Nakajima T, Yamaguchi H, Takahashi K (1980) S 100 protein
in folliculostellate cells of the rat pituitary anterior lobe. Brain Res 191:523-531 Nakazato Y, Yamaguchi H, Takahashi K, Mori T (1982) Immunohistochemical localization of nervous tissue proteins S-100, GFAP and astroprotein in extraneural tissue, In: Peters H (ed) Protides of biological fluids. Pergamon Press, Oxford, pp 47 50 Salazar H (1968) Ultrastructural evidence for the existence of a non-secretory, sustentacular cell in the human adenohypophysis. Anat Rec 160:419-420 Schlegel R, Banks-Schlegel S, Pinkus GS (1980) Immunohistochemical localization of keratin in normal human tissues. Lab Invest 42:91-96 Shirasawa N (1986) Differentiation of pituitary folliculo-stellate cells. In: Yoshimura F, Gorbman A (eds) Pars distalis of the pituitary gland: structure, function and regulation. Elsevier, Amsterdam, pp 175 181 Shirasawa N, Yoshimura F (1982) Immunohistochemical and electron microscopical studies of mitotic adenohypophysial ceils in different ages of rats. Anat Embryol 165:51-61 Shirasawa N, Kihara H, Yamaguchi S, Yoshimura F (1983) Pituitary folliculo-stellate cells immunostained with S-100 protein antiserum in postnatal, castrated and thyroidectomized rat. Cell Tissue Res 231:235-249 Shirasawa N, Yamaguchi S, Yoshimura F (1984) Granulated folliculo-stellate cells and growth hormone cells immunostained with anti-S 100 protein serum in the pituitary glands of the goat. Cell Tissue Res 237:7-14 Shirasawa N, Mitui J, Mera F, Ishikawa H (1988) Comparative immunohistochemical study of the mammalian pituitary cells stained with antisera against S-100 protein and non-neuronal enolase. Biomed Res 9:477-487 Stoeckel ME, Schmitt G, Porte A (1981) Fine structure and cytochemistry of the mammalian pars intermedia. In: Peptides of pars intermedia. Ciba Foundation Symposium 81. Pitman, London, pp 101-127 Svalander C (1974) Ultrastructure of the fetal rat adenohypophysis. Acta Endocrinol (Copenh) 76 [Suppl 188]:1-133 Takahashi K, Isobe T, Ohtsuki Y, Akagi T, Sonobe H, Okuyama T (1984) Immunohistochemical study on the distribution of :~ and /~ subunits of S-100 protein in human neoplasm and normal tissues. Virchow Arch [B] 45 : 385-396 Velasco ME, Roessmann U, Gambetti P (1982) The presence of glial fibriliary acidic protein in the human pituitary gland. J Neuropathol Exp Neurol 41:150-163 Yoshida Y (1966) Electron microscopy of the anterior pituitary gland under normal and different experimental conditions. Meth Achievm Exp Pathol 1:439-454 Yoshimura F, Nogami H (1981) Fine structural criteria for identifying rat corticotrophs. Cell Tissue Res 219:221 228 Yoshimura F, Soji T, Sato S, Yokoyama M (1977) Development and differentiation of rat pituitary follicular cells under normal and some experimental conditions with special references to an interpretation of renewal cell system. Endocrinol Jpn 24: 435-449 Yoshimura F, Nogami H, Shirasawa N, Yashiro T (1981) A whole range of fine structural criteria for immunohistochemically identified LH cells in rats. Cell Tissue Res 217:1-10
