Book - Sex and internal secretions (1961) 3

From Embryology
Embryology - 6 Aug 2020    Facebook link Pinterest link Twitter link  Expand to Translate  
Google Translate - select your language from the list shown below (this will open a new external page)

العربية | català | 中文 | 中國傳統的 | français | Deutsche | עִברִית | हिंदी | bahasa Indonesia | italiano | 日本語 | 한국어 | မြန်မာ | Pilipino | Polskie | português | ਪੰਜਾਬੀ ਦੇ | Română | русский | Español | Swahili | Svensk | ไทย | Türkçe | اردو | ייִדיש | Tiếng Việt    These external translations are automated and may not be accurate. (More? About Translations)

A personal message from Dr Mark Hill (May 2020)  
Mark Hill.jpg
I have decided to take early retirement in September 2020. During the many years online I have received wonderful feedback from many readers, researchers and students interested in human embryology. I especially thank my research collaborators and contributors to the site. The good news is Embryology will remain online and I will continue my association with UNSW Australia. I look forward to updating and including the many exciting new discoveries in Embryology!
Young WC. Sex and internal secretions. (1961) 3rd Eda. Williams and Wilkins. Baltimore.
Section A Biologic Basis of Sex Cytologic and Genetic Basis of Sex | Role of Hormones in the Differentiation of Sex
Section B The Hypophysis and the Gonadotrophic Hormones in Relation to Reproduction Morphology of the Hypophysis Related to Its Function | Physiology of the Anterior Hypophysis in Relation to Reproduction
The Mammalian Testis | The Accessory Reproductive Glands of Mammals | The Mammalian Ovary | The Mammalian Female Reproductive Cycle and Its Controlling Mechanisms | Action of Estrogen and Progesterone on the Reproductive Tract of Lower Primates | The Mammary Gland and Lactation | Some Problems of the Metabolism and Mechanism of Action of Steroid Sex Hormones | Nutritional Effects on Endocrine Secretions
Section D Biology of Sperm and Ova, Fertilization, Implantation, the Placenta, and Pregnancy Biology of Spermatozoa | Biology of Eggs and Implantation | Histochemistry and Electron Microscopy of the Placenta | Gestation
Section E Physiology of Reproduction in Submammalian Vertebrates Endocrinology of Reproduction in Cold-blooded Vertebrates | Endocrinology of Reproduction in Birds
Section F Hormonal Regulation of Reproductive Behavior The Hormones and Mating Behavior | Gonadal Hormones and Social Behavior in Infrahuman Vertebrates | Gonadal Hormones and Parental Behavior in Birds and Infrahuman Mammals | Sex Hormones and Other Variables in Human Eroticism | The Ontogenesis of Sexual Behavior in Man | Cultural Determinants of Sexual Behavior
Historic Disclaimer - information about historic embryology pages 
Mark Hill.jpg
Pages where the terms "Historic" (textbooks, papers, people, recommendations) appear on this site, and sections within pages where this disclaimer appears, indicate that the content and scientific understanding are specific to the time of publication. This means that while some scientific descriptions are still accurate, the terminology and interpretation of the developmental mechanisms reflect the understanding at the time of original publication and those of the preceding periods, these terms, interpretations and recommendations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)

SECTION B The Hypophysis and the Gonadotrophic Hormones in Relation to Reproducetion

Morphology of the Hypophysis Related to its Function

Herbert D. Purves, M.Sc, M.B., Ch.B.

Director, Endocrinology Research Otago University, Dunedin, New Zealand


I. Introduction

The aim of this chapter is to present certain morphologic characteristics of the hypophysis. Emphasis is given to those aspects of the structure which are, in the light of modern knowledge, of importance in elucidating the functions of this organ. Although the study of morphologic features does not by itself provide clear answers to questions concerning the endocrine functions of the hypopliysis, it is evident that the morphologic peculiarities of the gland are related to its functions, and that the consideration of morphologic data in conjunction with the evidence derived from physiologic observations does assist in the construction of acceptable hypotheses concerning these functions.

II. Embryonic Development

The adenohypophysis is derived from Rathke's pouch, a process of epithelial tissue derived from the buccal ectoderm. The distal portion of this process forms a hollow structure which lies in close contact with the infundibulum. The infundibulum is an outgrowth of neural tissue from the floor of the third ventricle, and differentiates into the neurohypophysis.

The work of Burch (1946) suggested that contact of the epithelial element with the neural element was necessary for the differentiation of the adenohypophysis from the former. He found in the frog {Hyla regilla) that translocation of the infundibulum or of the epithelial anlage, at a stage in development before these elements made contact, prevented the differentiation of the epithelial element. The pars intermedia did not develop, nor did acidophil and basophil cells appear in the epithelial cell mass. These results must, however, be reconsidered in view of the later findings of Etkin (1958). Etkin succeeded in transplanting the epithelial anlage at the earliest possible stage, and found that this does not prevent the differentiation of a pars intermedia. The development of normal pigmentation showed that the pars intermedia so differentiated functioned in its transplanted position and, in the later tadpole stages, pigmentation was greater than normal, indicating hyperf unction of the pars intermedia such as develops in the differentiated pars intermedia which is subsequently transplanted in the frog. Etkin's experiments were done with the wood frog, Bana sylvatica, but it is unlikely that the differences are due to species differences. Etkin observed differentiation of a jmrs distalis, but no chromophil cells were present at the stages examined. These chromophil cells usually do not appear until a later stage in normal animals. It has, therefore, not yet been demonstrated that a full differentiation of a functional pars distalis can occur without continued contact with the infundibu

It should be noted that Etkin's results do not show that contact with neural tissue is unnecessary for the initiation of the differentiation of the adenohypophysis from its epithelial anlage, because the buccal ectoderm is in contact with neural tissue before the development of either Rathke's pouch or the infundibulum. They do show, however, that the development of the adenohypophysis can proceed without continued contact with the neural element.

According to Eakin and Bush (1957), the pars nervosa develops without any consistent departure from normality in frogs (Hyla regilla) adenohypophysectomized at the limb-bud stage. Regeneration of a pars anterior often occurred giving albino tadpoles which metamorphosed normally. In one case a black nonmetamorphosing tadpole was found to have an adenohypophyseal fragment composed exclusively of cells similar to those of the normal pars intermedia.

The first appearance of granulated cells in the developing adenohypophysis presents some features of endocrinologic interest. Jost and Danysz (1952) found glycoprotein granules in basophil cells in the rabbit fetus after 20 days gestation. In the rat (Jost and Tavernier, 1956) glycoprotein granules appear in the ventral portion of the pars anterior after 15 days and in the central portion after 17 days gestation. In both species the granules appear when the first evidence for gonadotrophin and thyrotrophin is observed.

In the human fetus basophil cell granules appear in the developing hypophysis much earlier than acidophil cell granules. Basophil cell granules are seen after 8 weeks gestation, whereas the first definite acidophil cell granules appear only at the 19- to 20week stage (Pearse, 1953; Romeis, 1940). Romeis identified these early-appearing basophil cells as /3-cells (purple ^-cells) in the specific sense in which he uses this term. Inasmuch as intermedin appears in the human fetal hypophysis about the same time as these basophil cells (Keene and Hewer, 1924) , it is possible that the ^-cells of Romeis in the human hypophysis are intermedin-secreting cells and correspond to the cells of the pars intermedia of other mammals. More convincing evidence in support

of this hypothesis is presented in the section on the human hypophysis.

III. Anatomic Subdivisions of the Hypophysis

Definition of Terms

The hypophysis is very variable in form in different vertebrates. Even among the mammals there is considerable variation. The comparative anatomy of the hypophysis has been reviewed by Romeis (1940), Green (1951), and Herlant (1954a). The nomenclature is in a state of considerable confusion and ill adapted to the needs of the comparative morphologist. Confusion arises from the great numbers of synonyms in use, the use as synonyms of terms that are not strictly synonymous, and the use of the same term for structures that are not strictly homologous.

Because of the importance of a rigid and consistent system of nomenclature, it seems advisable for the purpose of this chapter to adoi^t such a system and to use a single term for each structure, avoiding all use of synonyms except where it is necessary to quote from, or to refer in specific terms to, the publications of other authors. I shall make no attempt to list the various synonyms or near synonyms in use, or to eciuate my usage with the varying usages which appear in past and current literature.

A. Anatomic Divisions and Components

The mammalian hypophysis (and that of most terrestrial vertebrates) may be divided into three parts (Fig. 3.1). (1) The median eminence, forming the floor of the third ventricle of the brain and contiguous with the hypothalamus. (2) The hypophyseal stalk, an attenuated connection between the other two divisions. (3) The lobular hypophysis, an expanded structure enclosed in the pituitary fossa, the latter being a cavity or sac formed by dura and bone.

Each division consists of an adenohypophyseal component derived by way of Rathke's pouch from the buccal ectoderm and a neurohypophyseal component derived from the neural tissue in the floor of the third ventricle. These components will be named by the addition of prefixes to tlif names characterizing the three division-, thus:



neural component

pars enninens pars nervosa

(•,•) adeno component^^ (epit-heliGJ )

pars tuberalis pars intermedia pars anterior

Adeno-eminonce Neural eminence Adenostalk Neural stalk

Adenolobe Neural lobe

Fig. 3.1. Diagrams representing sagittal sections of a conventionalized mammalian hypophysis showing the divisions, components and parts. The adeno-eminence and adenostalk together form the pars tuberalis. The adenolobe is almost completely divided by the hypophyseal cleft into an anterior lobe and an intermediate lobe. In many mammals the anterior lobe is entirely composed of pars anterior tissue and the intermediate is entirely composed of pars intermedia tissue as shown here. For further details refer to the text.

It must be added here that the division between eminence and stalk is in some species indefinite and cannot always be made; the stalk is variable in extent, and in some species it does not exist as a junctional region between the other divisions. Clearly the division into eminence and stalk is of descriptive value only and has no functional significance.

In most mammals, and in most other terrestrial vertebrates with the exception of birds, Rathke's pouch adheres to the neural component during development. The cavity of Rathke's pouch persists in the adult as a flattened cavity — the hypophyseal cleft (residual lumen) — which allows an easy separation of the adenolobe into two parts (Fig. 3.2) . The portion of adenolobe which is adherent to the neural lobe is the intermediate lobe. The composite structure formed by the adherent neural and intermediate lobes is called the posterior lol)e. The remaining portion of the adenolobe is the anterior lobe. Hypophyseal hormones are commonly extracted from the separated anterior and posterior lobes. The anterior lobe hormones are somatotrophin, prolactin, corticotrophin, thyrotrophin, follicle-stimulating hormone (FSH), and luteinizing hormone (LH), and are the products of the pars anterior tissue which the anterior lobe contains. Posterior lobe hormones are intermedin, which derives from the pars intermedia tissue in the intermediate lobe, and the neurohypophyseal hormones, oxytocin and vasopressin.

In whales (Wislocki and Geiling, 1936) and porpoises (Wislocki, 1929) and in the armadillo (Oldham, 1938), manatee (Oldham, McCleery and Geiling, 1938) , elephant (Wislocki, 1939), pangolin (Herlant, 1954a), beaver (Kelsey, Sorensen, Hagen and Clausen, 1957) , and the whole class of birds (DeLawder, Tarr and Geiling, 1934; Rahn and Painter, 1941), the cavity of Rathke's pouch is obliterated during development so that there is no hypophyseal cleft (Fig. 3.3). Moreover, in all these animals there is no adherence of the adenolobe to the neural lobe, and these two structures can be easily separated from one another. This is not a separation into anterior lobe and posterior lobe, which terms are not applicable to this type of hypophysis.

Figs. 3.2, 3.3, 3.4, and 3.5. Diagrams of sagittal sections of the hypophyses of rat, cat, ox, and blue whale showing some of the variations encountered. The left side of the diagram is anterior, the right, posterior. (Modified from B. Romeis, in Handbnch der mikroskopischen Anatomie des Menschen, Vol. 6, Part 3, JuHus Springer, Berlin, 1940.)

Fig. 3.2. The rat hypophysis is of the conventional form but rotated to bring the anterior lobe below and the posterior lobe above the axis of the gland.

Fig. 3.3. The hypophysis of the blue whale. In this form the adenolobe is not divided by a cleft and is not in close contact with the neural lobe, being separated from it by a fold of connective tissue (D) continuous with the dura. The adenolobe is filled with pars distalis tissue which combines the functions of the pars anterior and the pars intermedia of the more usual mammalian form. In these respects the hypophyses of birds, porpoises, the armadillo, the elephant, the pangolin, and the beaver resemble the hypophysis of the whale.

Fig. 3.4. In the hypophysis of the ox the pars intermedia is less extensive than the intermediate lobe and the remainder of the intermediate lobe is occupied by typical pars anterior tissue. The segment of pars anterior tissue which is in the intermediate lobe and is separated from the bulk of the pars anterior by the hypophyseal cleft is known as the cone of Wulzen (Wulzen, 1914).

Fig. 3.5. In the cat hypophysis the neural lobe is deeply embedded in the adenolobe. Par? intermedia tissue occupies the entire intermediate lobe and extends into the anterior IoIkparticularly in the caudal region.

The relation between the adenolobe and neural lobe in man is unique and is described in the section on the human hypophysis.

B. Histologic Divisions of the Hypophysis: Functional Parts

For the purpose of correlating structure with function, it is necessary to separate the hypophysis into parts on the basis of internal structure or of function, a separation independent of variations in gross anatomy. It matters not in theory whether the separation is based on fine structure or on function, because one determines the other. In practice the exact delimitation of the functional partition is determined by methods which are histologic or cytologic. For the determination of the homology between the histologically distinct parts of the hypophysis in different species, function must be the determining criterion. Histologically distinctive parts in different species can be considered homologous and be given the same name only if they are equivalent in function. Nature has set a trap for those who are careless in this respect. The partition of the adenhypophysis into functional parts is variable.

In terrestrial vertebrates the neurohypophysis consists of two functionally distinct parts (Green, 1951). Although following Green in this partition of the neurohypophysis, I must, however, reject his terminology, which used the anatomic terms "median eminence" and "neural lobe" in a modified sense. I propose, therefore, to use for these parts the names "pars eminens and "pars nervosa"

The two parts of the neurohypophysis are: (1) Pars eininens — consists of the neural eminence and neural stalk and the continuation of this tissue in the neural lobe to the point where it becomes the pars nervosa. Characterized by a vascularization in common with the adenohypophysis. Its venous effluents form portal vessels which pass into and arc the main blood supply of the pars anterior or the pars distalis of the adenohy

pophysis. (2) Pars nervosa — that part of the neural lobe whose venous effluent passes to the systemic veins directly.

There are two other differences between the pars eminens and the pars nervosa which characterize them. First, the arteries and veins of the pars eminens are confined entirely to the surface, where they connect with a rich vascular network. The interior of the pars eminens is relatively sparsely supplied with capillary loops — simple hairpin loops in small animals, more complex vascular spikes in larger animals — which penetrate it from the surface network. In contrast, the arteries and veins of the pars nervosa ramify within its substance and connect with a capillary network of the usual form. Second, in histologic sections the pars eminens is characterized by a low content, the pars nervosa by a high content, of stainable neurohypophyseal secretion.

In most mammals, and in most other terrestrial vertebrates except birds, the parts of the adenohypophysis (usual form) are three. ( 1 ) Pars tuberalis — consists of the adeno-eminence and adenostalk. Continues distally with the pars anterior, from which it is distinguished by the absence of the chromophil cells which characterize the latter. (2) Pars intermedia — a part of the adenolobe adjacent to and adherent to the neural lobe. Histologically it is characterized by the presence of cells of uniform appearance. These contain basophil granules, in quantities that vary greatly in different species. In some species granules appear to be absent. Functionally the pars intermedia is ciiaracterized by the presence of intermedin and of no other known hormone. The pars intermedia in some species is co-extensive with the intermediate lobe. In some, e.g., the sheep and the cow (Fig. 3.4), it is less extensive than the intermediate lobe, and the remainder of this lobe consists of pars anterior tissue (Wulzen, 1914). Such a detached portion of the pars anterior is known as a cone of Wulzen. In some species, sucli as the cat (Fig. 3.5). the pars intermedia is more extensive than the intermediate lobe, and extends somewhat into the anterior lobe. (3) Pars anterior — this tissue occupies the part of the adenolobe which is not pars intermedia. It is characterized by the presence of a mixed cell population including several types of chromophil cells.

In man, whales, porpoises, armadillo, manatee, elephant, pangolin, beaver, and in the whole class of birds, there is no pars intermedia (Fig. 3.3). The adenolobe contains throughout a mixed population of cells resembling that of the pars anterior. This structure differs from the pars anterior in containing intermedin in addition to the anterior lobe hormones, and in all probability contains an additional cell type corresponding to the pars intermedia cells. This structure I will call pars distalis to indicate that it is not strictly homologous wth the pars anterior as previously defined (Geiling, 1935). In these species, therefore, the parts of the hypophysis are two: (1) Pars tuberalis — as in other species. (2) Pars distalis — the whole adenolobe containing a mixed cell population throughout its extent.

It may be noted here that, although in mammals the absence of a pars intermedia is correlated with the absence of the hypophyseal cleft, in lower vertebrates a pars intermedia may be present despite the absence of the cleft (Green, 1951).

C. Zones in the Pars Anterior and Pars Distalis

The different cell types in the pars anterior or pars distalis are not uniformly distributed. The inequalities in the distribution of the acidophil cells, although not more marked than those of other cell types, are more conscipuous because of the larger number of these cells and their conspicuous staining. Zones rich in acidophil cells are called acidophil zones, zones poor in acidophil cells are called basophil zones. The zona tuberalis in the pars anterior adjacent to the junction of the latter with the pars tuberalis is a basophil zone.

This zonation, seen in many pituitaries, aids in the distinction between specific cell types and functional variants. When two cell groups with different stainability are shown to have different distributions characteristic for each, the groups so differentiated may be accepted as types. When there is no difference in the distribution of cells with minor differences in coloration, etc., the possibility remains that these different appearances are those of a single cell type in different functional states.

IV. A General Statement of the Problems of the Special Cytology of the Hypophysis

Nine distinct hormones are secreted by the hypophysis. Some simplification of the problem of relating the secretion of each hormone to specific cytologic details results from the fact that 2 of these hormones are secreted from the neurohypophysis and 1 from the pars intermedia, leaving only 6 to be secreted from the pars anterior. A complete revolution in the physiology of neurohypophysis has recently resulted from the discovery of methods for staining neurosecretion in this structure. It is indeed a source of wonder and delight that so much new and fascinating knowledge should result from the application of a simple technical discovery. The last 20 years have also seen great advances in the cytology of the pars anterior. Indeed, the progress here can be claimed to be more extensive, if less spectacular, than in the neurohypophysis. For all this the author, constrained to review the comparative cytology of the pars anterior, approaches his task with foreboding. A peculiar difficulty arises here which is not encountered in the neurohypophysis. The staining reactions of the granules of functionally equivalent cells are not consistent from species to species. This variation is easily understandable. The hormones themselves vary in chemical composition from species to species, even to the extent of the hormone from one mammalian species being inactive in another species. The secretory granules of pars anterior cells contain the hormones in association with hormonally inactive proteins no less variable than the hormones themselves. Functionally equivalent cells in different species may therefore have no specific component common to both despite the fact that they serve an identical function in the two species.

In this situation generalizations concerning the cytology of the pars anterior are difficult and hazardous. The treatment of the subject given here seems to be consistent with the recorded observations in a numl)cr of species. Without doubt further investigation, especially in additional species, will reveal its inadequacies.

V. Secretory Granules

A. The Nature Of Secretory Granules

Many of the cells of the pars anterior contain characteristic secretion granules stainable by a variety of procedures which indicate that the granules are composed mainly of characteristic proteins. The secretion of the different hormones is intimately related to the formation and discharge of secretory granules in different cells, and the nature of the hormone secreted by any cell is intimately related to the chemical nature of the secretory granules in that cell. The special cytology of the hypophysis, which deals with the differences between cells with different functions, is practically confined to the study of the secretory granules, these being the only truly specific characters that can be made visible by current techniques.

Granules similar to those in the cells of the pars anterior are seen in other tissues where the cells secrete proteins which act as hormones or enzymes. Examples are the granules of the hormone-secreting a- and ^-cells of the pancreatic islets, and the zymogen granules of the enzyme-secreting acinar cells of the pancreatic parenchyma. The nature of zymogen granules of the pancreas and the relationship of the granule to the secretory process have been clarified by the epoch-making researches of Palade (1956 », and of Siekevitz and Palade (1958a, b).

As seen in electron micrographs, the zymogen granules in the guinea pig pancreas are spherical bodies about 600 m/x in diameter. At certain stages in their formation a smooth membrane becomes visible at the periphery. It may be assumed that zymogen granules are bound by a smooth-surfaced membrane throughout their intracellular existence, and that, when in the fully formed granule an enclosing membrane is not visible in electron micrographs, it is because it is closely applied to the homogeneous content of equally high density.

Secretion of zymogen into the ducts after feeding the previously fasted guinea pig results in the disappearance of membrane-enclosed granules from the apical pole of the cell and the appearance of the dense content of the granule without enclosing membranes in the lumen of the duct. Apparently the enclosing membranes are left behind at the cell border and therefore presumably coalesce with the cell membrane during the passage of the granule through this structure.

Restoration of the cell content of enzymes and enzyme precursors occurs rapidly after secretory discharge. These products are formed in the basal part of the cell in which the synthesizing system is located. This system is revealed in electron micrographs as a closely packed arrangement of rough-surfaced membranes forming part of the endoplasmic reticulum. In light microscopy it is revealed by the cytoplasmic basophilia due to the ribonucleic acid component of the system. It is especially to be noted that the formation of the secretory product is not accompanied by any formation of membrane-enclosed zymogen granules, but dense material in a more or less diffuse form appears in the cavities of the endoplasmic reticulum. The filling of relatively large vesicles in the Golgi zone with dense material and the accumulation of membrane-enclosed zymogen granules in the apical region of the cell occur relatively late in the secretory cycle and seem to result from intracellular transport and enclosure in membranes of products synthesized earlier in the basal region.

One implication of these observations is that a hormone could be formed and released by an endocrine cell without being stored in the form of granules. This is possible inasmuch as the granule does not play any part in the formation of the secretion.

Electron microscopy of the rat hypophysis indicates that the granules are similar in structure to zymogen granules of the pancreas. The enclosing membrane is easily visible in the basopiiil granules of the pars anterior and in the neurosecretory granules of tlic pars nervosa as an electron-dense nienibrane enclosing a less dense content. The acidophil granules of pars anterior cells, like the zymogen granules of the pancreas, seem to contain solid protein of a density that does not permit the demonstration of an enclosing membrane except in thoye graniilcs tliat are only ixartially filled (Farquhar and Wellings, 1957).

There is no necessity for secretory granules to contain a single substance. The zymogen granules of the pancreas contain a mixture of enzymes and enzyme precursors. It is certain that some granules in the hypophysis contain a mixture of materials, some hormonal and others not hormonal. The staining reactions of specific granules may therefore be due to a nonhormonal constituent. This seems to be true of acidophil granules in the pituitary, inasmuch as the characteristic staining is due to a protein which is highly insoluble over a wide range of pH, whereas the hormones characteristic of these cells, growth and lactogenic hormone, are more soluble and can be extracted by procedures which do not dissolve the characteristic protein. Similar considerations apply to the neurosecretion in the nerve fibers of the neural lobe, because the characteristic hormones are octapeptides, whereas the characteristic staining is due to a protein with high cystine content presumably identical with the van Dyke protein I van Dyke, Chow, Creep and Rothen, 1942) . The granules of basophil cells contain glycoprotein, and the characteristic staining reactions seem to be due to the glycoproteins because they are similar to the staining reaction seen in other sites containing glycoprotein. In the basophil cells, however, it is not known whether the glycoproteins responsible for the staining reactions are identical with the hormones or whether they exist in the vesicular glands in association with hormonally active molecules themselves not stainable or not made visible by staining reactions because of their low concentration. Parallelism between intensity of staining and hormone content, which has been used to support the claim that it is the hormone product itself of these cells which is stained, does not in fact contribute to the solution of this problem. Staining reactions demonstrate the numbers of granules present, and parallelism between staining and hormone content indicates that the hormone is likewise contained within the granule. Chemical investigation of the nature of the isolated hormones and of the protein materials responsible for the staining reactions is the only source of evidence which might show that the staining reactions are due to a product distinct from the hormonally active product. This may be regarded as established for acidophil cell granules, neurosecretory granules, and for the glycoprotein granules of the pars intermedia cells, but has not been established for glycoprotein-containing cells of the basophil cells of the pars anterior. In this site, therefore, it is still possible that the staining reactions arc due to the hormonally active product itself and not to any associated but hormonally inactive protein.

B. The Staining Reactions of Secretory Granules by Modern Methods

The recognition of a diversity of cell types in the pars distalis as reported by Schonemann (1892) dei)ends on the presence of distinctive proteins in different cells. Although at one time theories which made the different cell types stages of a secretory cycle had a certain popularity, we now know that the different cell types made visible by staining methods have different secretory functions. This was first demonstrated by the observations of Smith and Smith (1923a, b). This momentous experiment showed a diversity of endocrine function in the anterior lobe and a correlation between function and type of cell. It also destroyed the hypothesis that different cell types could be merely different phases of a secretory cycle. All these results followed from the demonstration that in the bovine hypophysis there is constantly present a zone situated anteromedially which is poor in acidophil cells and, therefore, is mainly comprised of basophil cells and chromophobes. The tissue from this zone implanted into tadpoles had a predominantly thyroid-stimulating and consequent metamorphosis-inducing effect, whereas tissue rich in acidophils promoted body growth and was poor in thyroid-stimulating action. It is clear, then, that staining reactions in pituitary cells may be related to the nature of the hormone produced, and we may, therefore, consider in what manner the staining reactions are linked to the specific function. Although it is reasonable to assume that differences in the product of different cells arise from variations in their enzvme content, the slight differences in the enzymes which would produce these effects are not, with present technicjues, demonstrable by staining reactions. It is therefore not possible at present to stain different cell types in the pituitary, if by this phrase we mean the demonstration of differences in the cytoplasmic make-up of cells with different secretory functions. Differences in staining, which are observed with more or less ease, are due to the staining reactions of the products of hormone synthesis within the cell ; they do not demonstrate differences in the synthesizing mechanisms directly. One consequence of this is that specific cell types are only differentiated w^hen there is, within their cytoplasm, an adequate amount of their specific product in the form of granules.

Specific staining of granules requires methods which stain granules while leaving other cytoplasmic components unstained. Simple basic dyes such as methylene blue are not useful for this purpose. The basophilia of the basophil granules is weak, the binding power is weaker than that of ribonucleic acid. The capacity for dye absorption is great at high pH, so that cells with heavy accumulations of basophil granules appear densely stained, and it is this density of staining in densely granulated cells, rather than any exceptional strength of binding, that justifies the use of the term basophil for a certain class of cells in the adenohypophysis (Peterson and Weiss, 1955). Simple acid dyes do stain certain secretory granules specifically and find a use as counterstains to other procedures.

Modern methods of staining allow clearcut differentiations of specific granules by methods that do not depend on the distinction between acidophilia and basophilia.

The results of histochemical procedures may be expected to show a more consistent relation to function than the results of empiric staining methods. The McManus (194G» periodic-acid Schiff (PAS) reaction, which under certain conditions is specific for i)rotein-carbohydrate complexes in sections of animal tissues, allows a partition of the granules into two classes, which may have a consistent functional significance. An intense red or magenta color produced by this procedure in certain granules indicates that they contain glycoprotein. These glycoprotein-containing granules are basophil granules, showing the same characteristic basophilia wherever they occur in all cells of all species. Their staining reactions to acid dyes are, however, variable. The PAS reaction can be combined with counterstains which demonstrate the granules composed of simple proteins. The latter granules are acidophil granules.

Inasmuch as both the acidophil class and basophil class are composite, additional methods must be applied to effect a further separation of each class. Methods of general usefulness for this purpose are: (1 1 Staining by a mixture of acid dyes in such methods as Heidenhain's azan or Crossmon's (1937) modification of Mallory's stain. Granules may be stained orange, red, or blue, or a purple color from a mixture of red and blue dye. The coloration in each case depends on acidophilia; it has no relation to basophilia. The varying colors depend on variations in the strength and the quality of the acidophilia. Acidophil granules, like other strongly acidophilic structures, are stained either orange or red. Basophil granules may be orange, red, purple, or blue. (2) The use of elastic tissue stains such as kresofuchsin, resorcin-fuchsin, or Gomori's (1950) aldehyde-fuchsin, of which only the latter is in present-day use as a stain for hypophyseal granules. With these stains it is possible to stain some basophil (glycoprotein) granules while leaving others unstained.

By combining the results of the above methods of staining, at least 10 distinctly different staining reactions have been observed in specific granulation in the adenohypophyses of vertebrates. To date, however, no more than 4, or possibly 5, distinct staining reactions have been observed in the pars anterior of any one species. There is no possibility, therefore, of any constant relationship between staining and function.

Before progressing to the subject of cell classes and cell types, the problem presented by the existence of cytoplasmic ribonucleic acid should be discussed. Ribonucleic acid occurs in the cytoplasm of acidophils and basophils. It is a component of the Palade granules (Palade, 1955) which are described in the section on electron microscopy of the adenohypophysis. These granules are quite distinct from the secretory granules.

At one time the basophilia of basophil granules was ascribed to ribonucleic acid. It is now clear both from electron microscopy and from specific staining responses that cells of the basophil class generally contain less cytoplasmic ribonucleic acid tlian do acidophil cells and that ribonucleic acid is not a constituent of basophil granules.

The cytoplasmic basophilia demonstrated by the red staining by pyronin in the pyronin-methyl-green stain shows an entirely different distribution from the basophilia revealed by neutral toluidine blue or by the Gram stain and it appears probable that this pyronin-staining is specific for cytojilasmic ribonucleic acid. Pyronin basophilia is observed in all classes of cells of the anterior lobe and is increased in conditions where rapid secretion rather than storage is thought to be taking place. It is notably increased in the large degranulated cells observed in the hypophysis of pregnant animals and in animals treated by estrogen (Herlant, 1943).

Treatment for a short time with a ribonuclease preparation will remove the pyronin basophilia without removing the Gram-positive reaction or the violet metachromasia with toluidine blue in the basophil granules (Foster and Wilson, 1951; Pearse, 1952). Prolonged exposure to bile salts (Foster and Wilson, 1951) or ribonuclease preparations (Pearse, 1952) will remove the basophilia of the basophil granules, as well as the cytoplasmic basophilia. This does not indicate that the basophilia of the basophil granules is due to ribonucleic acid. Pearse (1952) has demonstrated the absence of ribonucleic acid in specific basophil granules by means of the coupled tetrazonium reaction. Acetylation for 8 hours blocks the reaction in the basophil granules but does not block the reaction of nucleolic and cytoplasm.

VI. Cell Classes and Cell Types

A. Classification

Present knowledge has diminished somewhat the number of distinct adenohypophyseal hormones, and it seems possible that no more than six may have to be accommodated in the pars anterior. These are somatotrophin or the growth hormone, prolactin or the lactogenic hormone, corticotrophin, thyrotrophin, FSH, and LH. It seems probable from recent developments that six cell types may be present in the pars anterior. It is, therefore, profitable to discuss pituitary cytology and its relation to pituitary function on the assumption that each hormone may be the single product of a specific cell type. This may, or may not, be true; a system of classification must not be too rigid.

The differentiation of the cells of the anterior pituitary into acidophils, basophils, and chromophobes must be regarded as a differentiation of an unknown number of specific cell types into three classes.

The cells are classified by, and take their names from, the staining reactions of their specific granules. Even the smallest detectable amount of granulation serves for this purpose, provided there is no confusion with mitochondria or cytoplasmic basophilia.

It may be noted here that it is usual to say that cells are stained by a reagent when they contain a large number of uniformly distributed stained granules. This usage is dangerous unless the convention is adopted and rigidly adhered to that all staining reactions of cells, unless explicitly stated to be otherwise, refer to the staining of specific granulation. It is particularly important when recording the absence of staining to note and state if there are granules that are unstained by the reagent. An absence of staining when there is an absence of granules is not at all of the same significance. The failure to differentiate between the staining of granules and the staining of other cytoplasmic components is a common error.

Cell types of the acidophil class contain specific granules which have similar properties and, therefore, react alike to the cruder staining methods. In the same way the specific granulation of the cell types of the basophil class have certain properties in common which are to be ascribed to a similarity in the chemical nature of the secretions which enable them to be recognized as a distinctive class. Chromophobic cells whichlack specific granulations of cither the acidophil or basophil cell classes will include all cells in a resting or inactive phase which temporarily do not contain specific granulation, and will also include any specific chromophobe type of cell which does not contain stainable granules at any time.

The characteristic of the acidophil cells is that they contain specific granules which are of a solid protein nature of high insolubility, easily preserved by any method of fixation. These acidophil granules have a strong affinity for acid dyes of all sorts, their degree of acidophilia being somewhat less than that of the hemoglobin of the red blood corpuscles. The cells of the basophil class are characterized by the presence of specific granulation in the form of droplets of glycoprotein which are not highly refractile and whose contents are freely soluble at physiologic pH. Their retention in the cell depends on the impermeability of the membranes, and they are released or dissolved when the cytoplasmic integrity is destroyed by cytolytic agents. The glycoproteins of the basophil cells have affinities for acid dyes which vary among the specific types in any one species and also from species to species. For a rigid division of chromophil cells into the acidophil and basophil classes, appeal must be made to an investigation of the chemical nature of the granulation by the PAS reaction. The subsequent subdivision of these classes into specific cell types is at present only partially achieved by tinctorial methods, and then only in favorable cases. It depends on empiric staining procedures, the results of which are not consistent from species to species.

B. Favorable Species

The fii'st feature that would be noticed by anyone who applied standardized staining procedures to pituitaries of a number of manrmalian species would l)e the greatly differing results obtained. These differences depend not only on the relative proj^ortions of the different cell types, but also on the intensity of coloration and quality of the staining reaction. Thus some mammalian species provide pituitaries whose anterior lobes do not at first sight appear promising, because basophil cells occur only in small numbers and with weak staining reactions not conducive to their differentiation into specific types. Acidophil cells may seem to be all of one type and all give the same staining reactions, either because only one type is present, or because two types are present but with staining reactions too close to be separated by the standardized arbitrary procedui'e used. Other species provide pituitaries which, from the first inspection, show themselves favorable for detailed study in that a wider variety of different cell types with strong and distinctive staining reactions is revealed.

Examples of mammalian i)ituitaries with staining properties favorable to the differentiation of the specific types are those of the bat (Herlant, 1956a,), dog (Purves and Griesbach, 1957a), monkey (Dawson, 1954bj, and cat (Herlant and Racadot, 1957). In these species the presence of five distinctive cell types in the pars distalis can be inferred from the results of the staining procedures. In other species the number of cell types is presumably similar, but the staining reactions of the granules are less favorable for the purjioses of tinctorial differentiation.

C. Reactivity

In si)ecies whose pars anterior cells contain granules with favorable staining reactions, subsequent experimentation will show whether the cell types so revealed will favor the experimenter by marked variations in their appearances in correlation with changes in their secretory activity. In some sj^ecies the variations in appearance, in conditions producing widely different levels of secretion of specific hormones, are so small and inconspicuous that much time and careful measurement may be necessary to detect the morphologic equivalent of the secretory change. In other species the most striking alterations of cell size, contour, and granule content occur as accompaniments of changes of secretion rate, including the appcarance of new forms and substances which are rarely to be seen in the "normal" pituitary. Examples of such changes are the appearance of thyroidectomy cells and castration cells in the rat i)ituitary under conditions producing rapid secretion of thyrotrophin or gonadotrophins respectively.

Reactivity in this sense, by enabling the significance of tinctorial or morphologic differences to be tested, is of primary importance both in the verification of partitions based on tinctorial or morphologic differences and in the identification of cell types as secretors of specific hormones. Only those differences which are correlated with a difference in reactivity, and which can be related to the production of specific hormones, can be regarded as of specific significance. Variations in the size and shape of the cell or its nucleus, the amount of cytoplasmic basophilia, the form of the Golgi apparatus, and the amount of granulation may be related to variations in the state of activity in the single cell type and are, therefore, of functional significance but not of specific value.

D. Specificity of Granules

One implication of the method of classification formulated here is that all the secretory granules within any one cell will be of a single type. This seems to be true of the majority of mammals, judging from my own observations. Admittedly it is possible with some techniques to obtain different colors in granules in the one cell, just as it is possible to stain erythrocytes two different colors in the one blood vessel, but such color differences are not significant. Apart from this, there is the possibility, since secretory granules in general contain a mixture of proteins and peptides and not a single substance, that granules vary in quality, due to differing proportions of the several components at different times as a result of different rates of secretion or from other effects. If this is so, differing shades of color may be obtained with some procedures in granules containing the same hormone.

E. Changes in Cell Proportions with Alterations in Function

The modifications in secretory activity, which occur as the result of environmental influences, or after experimental intervention, produce in some species distinct differences in the appearances of stained sections of the pars anterior. These differences are often referred to as "changes in the ]iroportions of chromophil cells." Since in many instances such changes in the proportions of different cell types occur rapidly and without any proportionate number of mitoses, changes in the proportions of cells have, in the past, been ascribed to the transformation of cells of one type into cells of another type. It cannot be said as yet that transformation of one cell type to another does not occur, but, if it does, it must be limited. The impossibility of telling what appearance any given cell would have shown at some time other than the time at which it was taken from the animal precludes a direct examination of transformations. The zonation phenomenon, however, is of the greatest assistance in setting limits to transformations that may occur. The stability of the acidophil and basophil zones in certain mammals shows that basophil cells and the chromophobes of the basophil zone are not transformable into acidophils, nor are acidophils transformable into basophils. Moreover, the fact that the anterior lobe hormones have been shown in some species to be not uniformly distributed within the anterior lobe, but to have characteristic distributions, different for the different hormones, suggests that each hormone is the product of a specific cell type not transformable into cells with a different function. '

If, therefore, basophils cannot be transformed into acidophils or acidophils into l^asophils, if for the most part each hormone is produced by its own specific cell type which is not transformable into another type with a different function, how can the apparent difference in the proportions of acidoi^hils, basophils, and chromophobes be explained? Three effects are operative. First, in many species a large proportion of the cells of the pars anterior are classified as chromophobes. Most of these cells are cells in a temporarily inactive phase and contain either no specific granulation or granulation in amounts too small to give decided staining reactions. With an alteration in secretion rate large numbers of chromophobe cells may accumulate granules. The transformation, however, is only from a specific cell low in granule content to one with high granule content, and not the production of a specific cell type from an undifferentiated cell. Second, the size of individual cells greatly influences the proportionate count obtained from the examination of thin sections. The larger cells not only show a greater area of cross section in the sections in which they appear, but they also appear more frecjuently owing to their greater extension in the direction perpendicular to the plane of section. The third effect, which is also related to the size of the cell as well as to its granule content, is the ease of recognition. When the cytoplasm is scanty the true nature of the cell may be missed and it may be classed as a chromophobe, whereas with equal granule density but with an increased width of cytoplasm it would be classed as a chromophil. These three effects are responsible for most of the variations in cell proportions reported in experimental studies of the pars distalis.

It should be noted here that, although marked variations in apparent cell pro])ortions occur in some species, the hypophyses of other species are remarkably uniform in appearance despite extreme variation in secretory function. Thus the bovine hypophysis shows little change in response to thyroxine deficiency, thyroxine administration, castration, or administration of sex hormones. The tendency for experimental work to be concentrated on those species in which functional changes are accompanied by marked cytologic changes has obscured the fact that such obvious cytologic changes are not a necessary accomjianiment to changes in a secretion rate.

VII. The Acidophil Cell Class

A. Acroophil Granules

Acidojihil granules consist of a membrane (enclosing solid contents. The presence of the membrane has been demonstrated by electron microscopy (Farquhar and Wellings, 1957). The contents include an insoluble protein, phospholipid, and hormones.

In size acidophil granules range from 0.1 to about 1.0 jx. In species where two types of acidophil cells can be seen, it is usual for the granules in one type to have a larger average size than those in the other. In species with large granules, the fact that there is a variation of size in the granules within individual cells can be determined with the light microscope. The granule size varies with the species, being relatively^ large in man, dogs, and cats, but fine in guinea pigs, mice, and sheep.

The acidophil granules are highly refractile in the fresh state. In living cells, under dark-ground examination, they are seen to be in constant rapid motion and show a refractility and luminosity greater than that of any element in basophil or chromophil cells. Their high refractility is a consequence of their solid nature. In typical acidophil cells a high concentration of granules is present in the cytoplasm, and this confers on acidophil-containing areas an opacity which enables them to be distinguished in the fresh state from zones in which acidophil cells are absent (Smith and Smith, 1923b).

In rats, almost total degranulation of the acidophil cells follows complete thyroxine deficiency, and in such hypophyses the normal opaque appearance of the anterior lobe changes to the translucent semitransparent api)earance normally seen in the acitlophilfree areas of the bovine hypophysis (Purves and Griesbach, 1946).

Acidophil granules are quite different from mitochondria. Specific staining methods for the demonstration of mitochondria leave the granules unstained (Severinghaus, 1932; 1939). When what may be termed ordinary methods of fixation and staining are used, the staining reactions of mitochondria and acidophil granules are similar.

The acidophil granules, like mitochondria, are readily released when the cells are disrujited in water, saline solutions, or sucrose solutions. In suspension they have the appearance of spherical highly refractile bodies of varying size. They have a higher density than any other element, but sediment more slowly than nuclei and erythrocytes because of their smaller size. On prolonged centrifugation acidophil granules make their way through the layer of nuclei and erythrocytes and api)car on the bottom of the centrifuge tube. Tiiey differ from mitochondria in not swelling up and disintegrating in saline solutions or distilled water, and in not collai)sing on drying.

H. L1I>()1I) Co of Acidophil Granules

The acidophil granules isolated by centrifugation (Herlant, 1952a) are positively stained by the Baker acid hematein test for phospholipid. In sections the same test produces a dark coloration of acidophil cells which is attributable to, in the main, staining of the granules (Rennels, 1953). There is some conflict of opinion as to whether this staining indicates the presence of phospholipid or is the result of some property of the granule protein. Herlant (1952a) and Elftman (1956) consider the reaction not to indicate phospholipid, but the evidence on which this conclusion is based is contradictory to the observations of Racadot (1954) and Ortman (1956) . The value of the hematein staining as a specific stain for acidophil granules is diminished by the positive staining of mitochondria and of phospholipid and other materials in basophil cells.

C. Sulfhydryl and Disulfide content of Acidophil Granules

Ladman and Barrnett ( 1954) described the staining of acidophil cells in the rat by the Barrnett-Seligman technique which is l)resumed to be specific for sulfhydryl and disulfide groups. This staining in acidophil cells is obtained only after fixation in Zenker's fluid. After fixation in acid alcohol and a number of other fixatives the cells do not stain. This observation could be interpreted to indicate the presence in acidophil granules of a sulfur-rich protein soluble in acid alcohol and preserved only by fixation in Zenker's fluid. Because of the possibility of formation of mercury compounds by the action of the fixative, and because such compounds might also react with the reagent used to demonstrate sulfhydryl groups, these results should be interpreted with caution. An increased staining after Zenker fixation, in comparison with acid alcohol fixation, is common to many animal tissues (Barrnett, 1953; Barrnett and Seligman, 1954). It has not been shown that treatment with Zenker's fluid after previous fixation in acid alcohol does not induce this type of staining. Adams and Swettenham (1958) showed that staining of the acidophil granules in the human pars distalis by the Barrnett-Seligman reaction was not due to cystine, because it was not blocked by oxidation of the sulfide linkages to sulfonic acid groups by performic acid. It should also be noted that the acidophil granules did not react positively to other histochemical tests for cystine.

D. Hormone Content of Acidophil Cells

An indication of the probable nature of the hormonal secretions of cells of the acidophil class is obtainable from the assay of portions of the acidophil and basophil zones of the pars anterior. Experiments of this sort were first made by Smith and Smith (1923a, 1)), who found that the central basophil zone of the bovine pituitary stimulated metamorphosis in hypophysectomized tadpoles with accompanying stimulation of the thyroid gland; the outer acidophil-rich portions caused an unusual stimulation of growth without causing either metamorphosis or activation of the thyroid. These results indicate the probability that growth hormone is present in high concentration in parts of the gland which are rich in acidophil cells. Azimov and Altman (1938) and Friedman and Hall ( 1941 ) found a differential concentration of prolactin in the peripheral acidophilic zones of the bovine hypophysis.

The problem of the distribution of the corticotroi)hin in the predominantly acidojihilic and basophilic zones was apparently decided by the consistent results obtained by Smelser ( 1944) and Giroud and Martinet ( 1948) , who found that the adrenal weightincreasing action was predominantly situated in the basophil zone. Later discoveries concerning the unreliability of the adrenal weight increase as a measure of adrenocorticotrophic activity made these results of less significance than was formerly believed. Some observations suggested that the adrenocorticotrophic hormone may, in fact, be present in high concentration in the acidophil zone. Chiti and Zinolli (1952) found that both the acidophilic and basophilic zones of the pig hypophysis have an action on the guinea pig adrenal cortex, and that there is a difference in the response to the two zones. The response to the basophilic zone was more marked in the fasciculata of the adrenal cortex, whereas the tissue of the acidophilic zone caused an enlargement of the reticularis. Desclaux, Soulairac and Chaneac (1953) using beef pituitaries found that implants of the basophil zone placed in contact with the adrenal of the rat did not modify the lipid content of the adjacent adrenal cortex, whereas similar implants from the acidophilic zone produced a discharge of lipid followed by a re-accumulation. The response, therefore, to the local presence of tissue from the acidophilic zone was similar to that produced by injections of corticotrophin. Inasmuch as Halmi and Bogdanove (1951) have shown that rat pituitaries which have lost their acidophil granules after total thyroidectomy still contain normal amounts of corticotrophin, it would seem that corticotrophin is not a constituent of the acidophil granules in the rat.

My own assays of the basophil and acidophil zones of the pig pars anterior, using the ascorbic acid depletion assay, showed that both zones contained high concentrations of corticotrophin, but that there was a higher concentration in the basophil zone than in the acidophil zone. In this species, as in the rat and bovine, corticotrophin is not associated with acidophil granules, and although a different situation may be found in other species, I would ascribe Herlant's (1952b, 1953a ) results to the absorption of the hormone by the granules or other particles sedimenting with them. The results of implantation experiments are explicable by the fact that corticotrophin is present in acidophil and basophil zones in about the same amount; apparently the histologic response is modified by the presence of other materials, different for each zone.

Inasmuch as somatotrophin and prolactin are the two hormones which are in higher concentration in the acidophil zones of pig and beef anterior lobes, it is concluded that these hormones are in the acidophil cells. The two hormones are not necessarily produced in the same cell, because in many mammals two types of acidophil cells can be seen. Purves and Sirett (1959) have shown that prolactin and somatotrophin have different disti'ilnitions in the anterior lobe of the wallaby hyjiophysis, indicating a separate origin for each hormone.

E. Hormone Content of Acidophil Granules

Acidophil granules prepared by the differential centrifugation of the susi)ension prepared by disintegrating the anterior lobes of sheep, pig, and ])ovin(> hypophyses in hypertonic sucrose solution, were examined for their hormone content by Herlant (1952a, b; 1955). The easily sedimented granules contain prolactin and somatotrophin, whereas thyrotrophin and gonadotrophins remain in the supernatant, from which they can be thrown down only by centrifuging at much higher speeds. The granules also contain corticotrophin, but in view of the well known tendency of this hormone to become firmly bound to proteins, this finding is of doubtful significance. Only if it were shown that corticotrophin added to a suspension of granules did not become bound to them could the association of this hormone with the granules be regarded as indicative of its in vivo location.

The results of Brown and Hess (1957) are not in conflict with Herlant's findings, although they interpreted them as showing somatotrophin in the mitochondrial fraction. Having experimented with both sheep and beef pituitaries, I am of the opinion that Brown and Hess underestimated the rate of sedimentation of acidophil granules, and that there is no likelihood of sedimenting the mitochondria in this material and leaving the acidophil granules in the supernatant.

Combining the evidence obtained from assays of tissue and of separated granules, it can be concluded that cells of the acidophil cell class produce somatotrophin and prolactin, and that these hormones are stored in their granules.

F. Acidophil Cell Types

In some species an easy differentiation of the acidophil cells of the pars anterior into two si)ecific types is obtained by staining methods. The differentiation depends on the tendency of the granules in one cell type to be stained orange with Orange G, whereas the other type is stained a red color with azocarmine in the azan method, or with acid fuchsin in Crossmon's (1937) method, or with erythrosin in the Cleveland-Wolfe method used by Wolfe, Cleveland and Campbell (1933)^.

It will be convenient for the purposes of discussion to adopt the term carminophil" used by Friedgood and Dawson (1940) for red-stained cells, and the term orangeophil" used by Lacour (1950) for orange stained cells. These terms are expressive of the relative affinity of the two types of granules in a species in which such a differentiation is obtained. The term "orangeophil cell" means the cell whose granules show the greater tendency to orange staining; the term "carminophil cell" means the cell whose granules show the greater tendency to be stained red, whether by azocarmine, acid fuchsin, or erythrosin. The equivalence of the three methods was demonstrated in the dog by Hartmann, Fain and Wolfe (1946), and by Goldberg and Chaikof! (1952a). I find that Crossmon's method gives in the human pituitary an equivalent differentiation to that obtained by Romeis (1940) with a modified azan method (Fig. 3.6). The results are most convincing when a differentiation is obtained by Crossmon's method, because in this method the differentiation appears spontaneously on staining the slide in an acidified mixture of orange G and acid fuchsin. The ease with which the differentiation is achieved varies in different species. In the sheep the Crossmon method gives two shades of red-orange or orange-red that are distinguished from each other only with difficulty, whereas in the dog the colors are pure orange and pure red. The azan method permits of modifications both in the solutions and the sequence which allows it to be accommodated to the varying staining properties in different species. The azan method, therefore, allows differentiations to be obtained that are not possible by simpler methods. For this versatility a price must be paid in the form of a tendency to inconsistent and variable results.

An orangeophil-carminophil distinction has been reported in the hypophyses of the dog (Wolfe, Cleveland and Campbell, 1933; Hartmann, Fain and Wolfe, 1946) ; armadillo (Oldham, 1938) ; rabbit and cat (Dawson and Friedgood, 1937, 1938a; Dawson, 1939; Friedgood and Dawson, 1940); man (Romeis, 1940); bovine (Gilmore, Petersen and Rasmussen, 1941); opossum (Dawson, 1938); ferret (Dawson, 1946); monkey (Dawson, 1948) ; l)at (Herlant, 1956a) ; and wallaby (Ortman and Griesbach, 1958). In vertebrates other than mammals two types of acidophils have been recognized in the hypophyses of birds (Rahn, 1939; Rahn and Painter, 1941; Payne, 1942. 1943); snakes (Hartmann, 1944; Cieslak, 1945); newts (Copeland, 1943); fishes (Scruggs, 1939) ; and frogs (Green, 1951).

Fig. 3.6 (/(,;/ ) >( i i nm <.i i Ik p.-u- anterior of the dog showmti iwti i\ |i<- <>[ ,i< iiLiplnl cell as seen after Crossmon staining. (1) Orangeophils; {2) carminopliils. Crossmon, X 820.

Fig. 3.7 (right). The same section as Fig. 3.6 decolorized and restained by PAS. The orangeophils (1) are pale, whereas the carminophils (2) are distinctly stained. Other cellwhich are more intensely colored by PAS than by Crossmon, are basophils. PAS, X 810.

It is not certain that the two types of cells observed in some of the above investigations were both acidophils. Mikami (1957) considers the cells stained orange by the azan method in the rostral zone of the pars distalis of the fowl to be basophils because their granules are intensely stained by the PAS reaction. Of different significance are the definite but relatively weak colorations observed in the carminophil cells of the dog (Purves and Griesbach, 1957a), and frog (Ortman, 1956a). The amount of color formed must be considered in relation to the amount of protein present. In the fully granulated acidophil cell there is a large amount of protein. With close packing of the granules, as much as 50 per cent of the cytoplasmic space can be occupied by solid protein granules. The amount of specific protein in basophil cells is usually very much less. Any nonspecific staining of proteins, such as by a weak iodine solution, or by the methods used for staining paper electrophoresis strips, will show the high protein content in the fully granulated acidophil. Calculations based on the color produced by PAS in the mixed proteins of blood plasma (approximately 7 per cent solution of protein containing 1 per cent of sugar by weight) make it evident that ciuite a strong color would be produced in acidophil cells if there were but one reacting group per molecule of protein. It is only when the color is intense in relation to the amount of protein reacting that the presence of many reacting groups and, therefore, probably the presence of polysaccharide, may be inferred. The color produced in the granules of carminophil cells is not of this intensity.

In order to resolve the doubt concerning the significance of moderate PAS colorations in heavily granulated cells, Purves and Griesbach (1956) used a jireliminary buffer extraction before fixation and found that the carminophil granules in the dog were insoluble, and were still stainable by PAS after extraction at pH 7.5, a treatment that dissolved all the basophil granules. They, therefore, defined basophil granules as granules containing soluble glycoproteins. This definition enables one to bypass the question 'Ts the PAS reaction in carminophil granules sufficiently strong to indicate the presence of glycoprotein?", the answer to which is a matter of opinion, and to substitute for it a practical solubility test which is easy to carry out and interdict.

When a sensitive PAS sec[uence is used^ acidophil cells generally show a weak color which is easily masked by counterstaining (Fig. 3.7). The cytoplasm of chromophobes shows a similar weak coloration; against this background the acidophils may appear to be unstained. The granules of the carminophil cells of the sheep, dog, frog, wallaby (Ortman and Griesbach, 1958) , and cat (Herlant and Racadot, 1957), give reactions distinctly stronger than this. The observation of Pearse (1951) that the carminophil cells of the rabbit do not contain glycoprotein does not exclude the possibility of a weak reaction in this species, because he used the Hotchkiss (1948) modification of the PAS sequence, which suppresses weak reactions and is allegedly more specific for glycoproteins.

Herlant and Racadot inferred that the carminophil cell in the cat was a basophil, but because the color produced by the PAS reaction is weak compared to that in typical basophils, I would regard this as a case to be submitted to the solubility test before a definite decision is made.

G. Secretory Functions Of Acidophil Cell Types

In view of the evidence presented earlier, which indicates that acidophil cells secrete somatotrophin and prolactin, it can be ex])ected that in species in which two types of acidophil can be distinguished, one type will secrete somatotrophin, the other prolactin. The reactivities of the cell types in mammals, in which two types are distinguishable, show that this is indeed the case and the point is discussed at some length in the section on acidophil cells and somatotrophin secretion which follows.

In general one of the two specific cell types is reactive in relation to the reproductive cycle and shows marked fluctuations in activity, which can be correlated with the secretion of prolactin at times when its luteotrophic, mammotrophic, or lactogenic action is apparent. The second type is relatively stable in relation to the reproductive cycle, and is assumed to secrete somatotrophin. Evidence in support of this assumption is available in some species.

Mammals in which the cai-iniiiophil cells are active during pregnancy and lactation are cat (Dawson, 1946; Herlant and Racadot, 1957), rabbit (Pearse, 1951), monkey (Dawson, 1948), and bat (Herlant, 1956a).

In the rabbit the discharge of luteinizing hormone which causes ovulation occurs within 1 hour after coitus. Pearse (1951) found at this time a discharge of granules from basophil cells which presumably were secreting luteinizing hormone. For the first 3 to 4 hours after coitus the carminophil cells were accumulating granules ; thereafter there was a slow discharge of granulation during the next 10 hours. These changes in the carminophil cell are consistent with a luteotrophic function for the secretion of these cells.

The responses of the carminophil cell of the cat (Dawson, 1946) are similar to those of the rabbit. They are consistent with an accumulation of granules at estrus and for a time after mating, with a secretion with luteotrophic action immediately after ovulation, with a further accumulation of granules during the later stages of pregnancy when a luteotrophic action by the hypophysis is not necessary, and with a phase of strong secretory activity with a lactogenic action, beginning at parturition and continuing for the first 3 weeks of lactation. It should be noted here that Herlant and Racadot (1957), while confirming the luteotrophic action of carminophil cell secretion, consider that in the cat this cell secretes the luteinizing hormone. They relate lactogenesis to the secretion of a hormone, presumably prolactin, by chromophobe cells, which are rich in cytoplasmic ribonucleic acid. I am inclined to think, from the similarity in the responses of carminophil cells in cat and rabbit, that they serve the same function in both species, but admit the necessity for further investigation in the cat.

Because of its unusual breeding cycle, the bat (Herlant, 1956a) presents peculiar advantages for the correlation of specific cell types with specific secretion. In this species the carminophil cells show two phases of secretory activity during the breeding cycle. The first phase begins at the time of ovulation and terminates during the latter half of pregnancy; the second phase begins a little before parturition and continues throughout lactation. In animals which do not lactate these cells involute rapidly after parturition. The responses indicate secretion of prolactin with luteotrophic and lactogenic actions at appropriate times.

In two mammalian species the carminophil cell is the stable type, and the orangeophil the reactive type during the breeding cycle. These are: (1) Human hypophysis. The orangeophil cells are active during pregnancy and contain orangeophil granulation towards the end of pregnancy. These cells have been called pregnancy cells by Erdheim and Stumme (1909). Romeis (1940) calls the orangeophil cells seen in small numbers in the hypophyses of males and nonpregnant females "e"-cells, and the numerous orangeophil cells seen in late pregnancy "7/"-cells. Any differences between the two are presumably due to different states of activity. The granules of both show the same staining reactions, and only one type of orangeophil cell is seen in any one hypophysis. (2) The rat (Lacour, 1950; Dawson, 1954a). Purves and Griesbach (1952) inferred the separate existence of somatotroi)hin and prolactin-secreting cells from the differential effects of estrogen and a deficiency of thyroxine on the acidophil cell population, but could not achieve a tinctorial differentiation. The orangeophil staining achieved by Lacour was not reproducible. Dawson's (1954a) modification of the azan stain has usually given an orangeophil reaction in the active-looking acidophils of the pregnant rat pituitary, and in the similar cells appearing after estrogen administration but the staining is sometimes unsuccessful. Obviously the staining affinities of the two types of granulation in the rat are very similar.

In all these species except the rat, the reactive acidophil type, whether it be the carminophil cell in the cat, rabbit, monkey or bat, oi- the orangeophil cell in the human hypophysis (Floderus, 1949), has a different distribution from that of the nonreactive type, and the marked increase in the number of visible cells at certain times is due to the appearance of granules in previously nongranulated, and, therefore, chromophobic cells. In the rat the orangeophil staining appears to be the result of a change in the nature of the granules in a jiroportion of the acidophils, which in the normal nonpregnant animal are carminophil.

It appears, then, that in a nmiiber of mammals the secretion of somatotrophin and the secretion of prolactin are the functions of two distinctive acidophil types, not transformable into one another, and with variable staining reactions in different species. In most of the species studied thus far, the prolactin-secreting cell is the more carminophil type. In view of the variations in staining affinities which make the differentiation of two types by staining methods easy in some species and difficult in others, it is possible that in some species two types of acidophil occur whose granules stain alike, and the human hypophysis is presumably one in which the prolactinsecreting cell is the more orangeophil. The variability in the nature of the hormones in different species (the variability of somatotrophin is particularly well attested) makes this variability in the staining reactions of cells serving the same functions in different species credible.

Some recent observations indicate that more than two acidophil cell types may be present in some species. Herlant (personal communication) has found four distinct acidophil cell types in the pars anterior of the mole {Talpa europanaea), and Ortman and Griesbach (1958) have evidence for four acidophil cell types in the wallaby {Wallabia nifogrisea).

Purves and Sirett (1959) found prolactin concentrated in the rostral portion and somatotrophin concentrated in the caudal portion of the anterior lobe of the wallaby hypophysis. Somatotrophin, therefore, is secreted by orangeophil cells in the caudal zone of the pars anterior, and it is probable that the carminophil cells of the rostral zone are the source of the prolactin.

II. Acidophil Cells in Relation to Somatotrophin Secretion

The association of the acidophil cells with the production of growth hormone was first deduced from the observations of acidophil cell adenomas in the hypophyses of patients showing the symptoms of acromegaly or gigantism, and, as Gushing and Davidoff stated in 1927, no one today can have anv reasonable doubt that the sugstance which provokes the overgrowth is a product of the acidophil cells.

The absence of acidophil cells in the pars anterior of the dwarf mouse (Smith and MacDowell, 1930) is often cited as evidence for the origin of a growth-regulating hormone secreted by the acidophil cells. Smith and MacDowell's report indicated that there may be deficiencies in cell types other than the acidophils, but Ortman (1956b) finds that basophils, both /3-cells and 8-cells, are present and appear similar to those present in normal litter mates. It seems, therefore, that there is a specific deficiency in growth hormone secretion associated with a deficiency of acidophil cells in these animals. However, the dwarf mice also show evidence of deficient thyrotrophic, gonadotrophic, and adrenocorticotrophic secretion and growth may be stimulated in them by thyroxine administration (Nielson, 1952) as well as with growth hormone. Treatment with both hormones is necessarj^ to produce the appearance of full-grown mice. The lack of acidophils in the dwarf mouse hypophysis cannot be due to degranulation of these cells by thyroxine deficiency, because thyroxine administration does not cause regranulation.

Hewer (1943) described a case of human dwarfism in which the hypophysis was markedly deficient in acidophil cells.

It should be noted that the cessation of growth in the rat after hypophysectomy is due to the loss of thyrotrophin and consequently of the thyroid secretion as well as the loss of somatotrophin. This view is necessitated by the observations of Geschwind and Li (1952) who showed that either thyroxine or somatotroi)hin produces growth in hypophysectomized rats. Changes in the hypophysis resulting from thyroxine deficiency accompany the arrest of growth which occurs in totally thyroxine-deficient rats. Zeckwer, Davison, Keller and Livingood (1935) pointed out that the disappearance of acidophil cells occurs in the rat hypophysis after thyroidectomy and related this to the cessation of growth in such animals, because in partial thyroxine deficiency resulting from incomplete ablation of the thyroid, the acidophil cells were retained and growth continued. The relation of the acidophil cell degranulation to thyroxine supply was studied by Purves and Griesbach (1946) who showed that, although 2.5 /xg. of DL-thyroxine were required to prevent thyroidectomy changes in the basophils of the rat hypophysis, the acidophils were protected from degranulation by 0.5 ^g. per day and the effect of even smaller quantities in retaining some of the acidophil cells was observed. The rate of growth observed in these animals was related directly to the content of acidophil granules.

Degranulation of the acidophil cells occurs in rats treated with potent goitrogenic agents. As judged by the acidophil cell response, thiourea is much less effective than thiouracil in suppressing thyroid secretion because the administration of 0.25 per cent of thiourea in the drinking water does not cause the loss of acidophil cells.

Using the regranulation of the acidophil cells as a sensitive indicator of small amounts of thyroxine, Purves and Griesbach (1946) demonstrated that the administration of iodide in relatively high dosage (1 mg. per day) produced in thyroidectomized rats an extrathyroidal synthesis of material with thyroxine-like activity in amounts which were of physiologic significance. The partial regranulation of the acidophils on high iodide intakes in totally thyroidectomized animals is associated with continued growth at a subnormal rate. They concluded that from injections of 1.3 mg. of potassium iodide per day, approximately 0.12 /xg. of L-thyroxine might be produced. Hum, Goldberg and Chaikoff (1951) showed that injections of iodide (1 mg. per day or more) caused regranulation of acidophil cells in rats whose thyroid tissue had been destroyed by administration of radioactive iodine and considered that the effect was due to extrathyroidal synthesis of thyroxine.

Marine, Rosen and Spark, (1935) reported almost total loss of acidophils from the hypophyses of thyroidectomized rabbits, but overlooking changes in the basophil cells, they related the acidophil cell changes to the production of thyrotrophic hormone.

In the thyroxine-deficient dog, changes in the acidophil class of cell are apparently limited to the a-cell. Goldberg and Chaikoff (1952b) observed in dogs with complete thyroid destruction produced by radioactive iodine, a complete degranulation of the a- (orangeophil) cells whereas the e(carminophil) cells were well preserved and full of granules. The animals were adult and the effect on growth was not observed. The selective degranulation of the a-cells is, however, consistent with the view that these cells are concerned with somatotrophin secretion.

VIII. The Basophil Cell Class

A. Basophil Granules

In the light of modern knowledge basophil granules are defined as secretory granules containing soluble glycoproteins. The ]\IcjManus (1946) PAS reaction serves as a group reaction for the identification of all granules of the basophil class, as it imparts to them an intense red or magenta color indicative of a high content of protein-bound carbohydrate. Glycogen and certain lipoid inclusions are two other intracellular substances which can be colored intensely by the PAS reaction. In paraffin sections pretreated with diastase these substances will be absent. Insoluble substances of an unknown nature which give a color with the PAS reaction occur in some acidophil cells, and should be sought for in material submitted to extraction before fixation.

Suitable extraction methods are: (1) Freeze and thaw the tissue on the stage of a freezing microtome three times before fixation. Basophil granules will dissolve in the mixture of intracellular and extracellular fluids. (2) Place small pieces of tissue in cold alcohol or acetone for 30 minutes, then in buffer for 1 hour before fixation. This allows some control of the pH at least in the peripheral portions of the tissue. (3) Perfuse the animal by way of the aorta with buffer which is saturated with ether at 37°C., or to which desoxycholate has been added as a cytolyzing agent. This method allows the assay of the extracted tissue for content of unextractable hormones. In the other methods the granule contents, although in solution, still remain to a large extent inside the tissue.

It should be pointed out here that mechanical damage resulting from the handling of the delicate tissues of the hypophyses of smaller animals causes a loss of the granules from basophil cells. Such hypophyses are best fixed in situ.

Basophil granules show a high capacity, but only a weak binding power, for basic dyes (Peterson and Weiss, 1955) . This basophilia does not serve for the characterization of basophil granules because of the prevalence of cytoplasmic basophilia and the staining of acidoi)hil granules by basic dyes.

There has in the past been some confusion between the basophilia of basophil granules and that of ribonucleic acid. Dempsey and Wislocki (1945) pointed out that the staining of basophils with aniline blue was not concerned with cytoplasmic ribonucleic acid, and suggested that the aniline blue stained an acidophil substance which represented the true secretory granules of the cells. The nature of the specific granules in basophil cells was clarified by the researches of Herlant (1942), who observed a violet metachromatic staining of the basophil granules in the human pituitary by neutral solutions of toluidine blue. This metachromasia is observed only in undehydrated sections and demonstrates a specific type of basophilia quite distinct from that due to cytoplasmic ribonucleic acid. This type of basophilia, which is common to many sites containing polysaccharide material, was considered by Herlant to indicate that the specific granules of the basophil cells were composed of glycoprotein. After a short treatment with alcohol to cytolyze the cells, the glycoprotein material was found to be extractable by water, and the resultant extract was rich in FSH. Herlant demonstrated the glycoprotein nature of the basophil granules by the Bauer method for polysaccharide, and the observations were subsequently confirmed by the more satisfactory McManus (1946) method for glycoprotein (Herlant, 1949). Other observations confirming the glycoprotein nature of basophil gramdes have been reported by Catchpole (1947, 1949), Pearse (1948), and Purves and Gricsbach (1951a I.

The Herlant metachromasia is observed ill the basophil granules and indeed in all PAS stainable structures in rat, dog, and litiinan hypophyses. This metachromasia docs not serve to characterize basophil granules, inasmuch as the carminophil cells of the dog and of the wallaby are also metachromatically stained by toluidine blue.

Basophil granules vary considerably in size in different species. In the rat they are seen in electron micrographs as spherical bodies of a maximal size of 150 mix. The individual granules are not seen by the light microscope, and the flocculent "granules" seen in stained sections are the result of the uneven distribution of the granules and their aggregation during fixation and dehydration. In the human hypophysis the basophil granules are much larger, and are singly visible in lightly granulated cells in stained sections. They vary greatly in size and the largest among them are as large as, or larger than, acidophil granules.

In fresh tissue the basophil granules do not show the high refractility of the solid acidophil granules, and are therefore presumed to be vesicular, consisting of a membrane enclosing a solution of glycoprotein.

The staining procedures most generally used in the study of the cytology of the adenohyphophysis are various trichrome procedures using mixtures of acid dyes. As has been said before, by these methods basophil granules may be stained orange, red, purple, or blue. By such methods it is, of course, not in general possible to distinguish acidophil granules from basophil granules. This was not realized until recently because of a premature assumption that only two types of chromoj^hil cells are present in the pars anterior of mammals, and because of a widespread erroneous belief that all blue dyes are basic dyes. In the literature, the term "basophilic" is constantly applied to structures stained blue by aniline blue in variations of the trichrome staining procedures and is, therefore, used as a synonym for blue rather than as an indication of staining affinity.

It may be stated here, as regards the multiplicity of basophil cell tyi)es included in the basophil cell class, that the pars intermedia cells are basophils, and that there are in addition three specific types of basophil cells distinguishable in the pars anterior of certain favorable mammalian species. The immediately succeeding sections are concerned with the basophil granules and the basophil cell types of the pars anterior.

B. Hormone Content of Basophil Cells

Evidence of the hormonal content of basophil cells can be obtained by the assay of portions of tissue rich in basophil cells and comparison with the results obtained from adjacent portions rich in acidophil cells. Investigations of this type have been made by Smith and Smith (1923bj, Voitkevitch (1937a, bj,Herlant ( 1943), Smelser (1944), and Giroud and Martinet (1948). These investigators agree that tissue rich in basophil cells has a high content of thyrotrophin and gonadotrophin. It seems that the gonadotrophic effects include both follicle-stimulating and luteinizing actions. Corticotrophin in both pig and bovine anterior lobes is also somewhat more concentrated in the basophil zone. Because of the prevalence of chromophobes as well as basophils in the basophil zones, it cannot be inferred that all these hormones are in the basophil cells. Moreover, Smelser has shown that the ratio of the potencies of the two zones is different for thyrotrophin, corticotrophin, and gonadotrophin, so that these hormones seem to be stored in different cells.

Inasmuch as three distinctive types of basophil cell can be seen in the pars anterior of some other species, it is likely that a similar diversity occurs in the bovine hypophysis. This species is, however, not favorable for the discernment of specific basophil cell types.

C. Hormone Content of Basophil Granules

The basophil granules are sensitive structures, and lose their glycoprotein content when the cells in which they are contained are damaged, as by freezing or mechanical distortion. It appears that this is the result of the instability of the enclosing membrane to solutions of low osmolar content.

Thyrotrophin, FSH, and LH are, like the the glycoproteins of basophil granules, soluble in water at neutral pH, and are completely extractable from anterior lobes after acetone drying or mechanical disintegration. These hormones seem to be located in membrane-enclosed structures in living cells, because they can be centrifuged down from suspensions of anterior lobe tissue disintegrated in 0.88 M sucrose. McShan and Meyer (1952), and McShan, Rozich and Meyer (1953), using the anterior lobes of castrate rats, were able to recover as much as 75 per cent of the gonadotrophin in a small granule fraction sedimented by centrifuging at 20,400 X g for 1 hour. There was evidence that grinding increased the amount of hormone in solution, presumably by disruption of granules. The hormone was extracted from the granules by isotonic saline, and from this solution could no longer be sedimented. This behavior is accounted for by the enclosure of the hormone in a membrane with properties similar to that of the mitochondrial membrane and, therefore, stabilized by high sucrose concentrations.

Herlant (1952a) worked with the anterior lobes of sheep, pig, and beef hypophyses, and obtained a small granule similar to that obtained by McShan and his associates which contained gonadotrophic activity. He found that these small granules gave an intense coloration with the PAS reaction, and thus demonstrated that they were basophil granules. It does not seem that the thyrotrophin content of the small granule fraction has been tested except by Brown and Hess (1957) . Their results are not easily interpreted because they sometimes used frozen instead of fresh beef anterior lobes and their fractions are wrongly identified, but it does seem that thyrotrophin can be centrifuged down from material dispersed in sucrose solution (0.25 m in their experiments).

Although these observations are not by themselves conclusive, when considered in conjunction with the cytologic observations presented in the succeeding section, they provide confirmatory evidence for the view that thyrotrophin, FSH, and LH are made by basophil cells and are contained within their characteristic granulation. The behavior of corticotrophin in tissue dispersed in solution needs further investigation.

D. Does the Periodic Acid-Schiff Reaction Demonstrate the Hormone Content of Basophil Cells?

The fact that in the rat pars anterior the amount of PAS reacting material in thyrotrophs, FSH cells, and LH cells, parallels the content of thyrotrophin, FSH, and LH, respectively, raises the question as to whether or not the hormones themselves are responsible for the staining reactions. The parallelism between depth of staining and hormone content does not contribute to the solution of this problem, because this parallelism is a consecjuence of the fact that the hormone is stored in the specific granules and the number of granules present determines the depth of staining. This cjuestion, therefore, resolves itself into two parts. First, are the hormones glycoproteins, and second, are the hormones major constituents of the granule contents?

The reported sugar contents of the preparations of thyrotrophin, FSH, and LH produced by the fractionation of extracts from the anterior lobes of domestic animals do not establish the glycoprotein nature of the hormones, because it has never been shown that similar glycoprotein-containing fractions do not result from the fractionation of extracts of tissue which are free from these hormones. Taking advantage of the ease with which rat hypophyses can be obtained free from these particular hormones by pretreatment of the animal with thyroxine and estrogen, I have compared the extracts from hormone-free anterior lobes with those from normal anterior lobes. Similar amounts of protein with similar sugar contents were extracted from both types of gland. Fractionations by salt or alcohol which resulted in potent preparations of thyrotrophin, FSH, and LH, when applied to the extracts of the normal gland, gave a similar partition of the glycoproteins extracted from the hormone-free gland. It is clear that the preparations of the reputedly glycoprotein hormones obtained by fractionation of glandular extracts with varying concentrations of salts or alcohol are composed mainly of inert materials resulting from the fractionation of a mixture of plasma proteins and soluble cytoj^lasmic constituents.

In extracts from castrate rat anterior lobes in which the gonadotrophins are present at many times the normal level, the FSH fraction contains a small but definitely increased amount of PAS reacting material compared with the corresponding fraction from an equal weight of hormone-free anterior lobe. There is. therefore, in the l)as()phil granules a glycoprotein which accompanies the FSH during fractionation of the extract, and this glycoprotein is present in amounts which would contribute to the staining reactions of the granules. The potency of this material is close to that of the highly purified FSH preparations reported by Steelman, Kelly, Segaloff and Weber (1956).

Persons observing pars anterior cells under the microscope often form impressions which exaggerate the amount of specific granulation present in the pars anterior. They often also have exaggerated ideas about the potency of the separated hormones. Taking account of the small percentage of granulated thyrotrophs in the rat pars anterior and the small fraction of the cytoplasmic content of these cells which is in the form of granules, there must be only about 1 /xg. of specific thyrotroph granulation per mg. dry weight of anterior lobe. Yet the anterior lobe has a potency of 0.2 I.U. thyrotrophin per mg. dry tissue, equivalent to 10 /xg. of the potent preparation of thyrotrophin obtained by Condliffe and Bates (reported by Sober and Peterson, 1958).

It seems reasonable to assume that basophil granules, like other specific secretory granules, contain a mixture of substances, and it appears likely that, at least in the rat pars anterior, the glycoprotein hormones form a considerable proportion of the granule content and contribute significantly to its staining reactions.

E. Differential Staining of Basophil Cells by Resorcin-Fuchsin, Kresofuchsin, Aldehyde-Fuchsin, and Alcian Blue: ^Cells and 8-Cells

The three basoi)hil cell types in the pars anterior of some niainmals can be distinguished from each otlici' by diagnostic features which are independent of any specific diffei-ential staining of the granules, but e\-en in these species a differential staining of the sjiecific granulation is an important aid in the study of specific types. In some species the granules of different cell types react differently to acid dyes and appear in different colors after trichrome staining methods. The recognition of the presence of a multiplicity of basophil types was delayed not so much by the difficulty of distinguishing between them as by the difficulty of recognizing the basophil nature of the different types in the species in which the different types were highly distinctive. It is for this reason that the first clear recognition of two types of basophil cells came from the discovery of Romeis (1940) that kresofuchsin applied to sections of human pars distalis gave a differential staining of the granules of cell types that by the azan procedure were stained alike. There seem to be four types of basophil cells in the human pars distalis. Two of them which Romeis did not consider sufficiently distinctive to warrant separation were designated by him as "/3-cells." The other two types he designated as "y-cells" and "8-cells." Kresofuchsin stained the granules of both types of /?-cells and left the granules of y-cells and 8-cells unstained. The ^- and the 8-cells after azan staining were so much alike that a clear differentiation between them could be obtained only by the use of kresofuchsin, but the ycells were so distinctive that the use of kresofuchsin to distinguish them from ftcells was not necessary. In this way kresofuchsin acquired a reputation for distinguishing y8-cells from S-cells rather than for distinguishing /3-cells from y-cells and 8cells. The premature assumption that only two types of basophil cells would be present in the pars distalis of man and the pars anterior of other mammals led to the widespread use of the terms "^S" and "8" to designate cell types or cell groups in species other than man, and thus set the stage for an episode of nomenclatural confusion, the results of which have been exacerbated by a naive assumption that cell types designated by the same Greek letter by different authors in different species ought to have the same functions.

The history of the use of elastic tissue stains — kresofuchsin, resorcin-fuchsin, and aldehyde-fuchsin — is interesting and shows the difficulties that are experienced with the use of these materials. Erdheim and Stumme (1909) introduced the use of resorcinfuchsin as an elective stain for the granules of basophil cells in the human hypophysis, and it appears among the staining reactions listed l)v Bailev and Davidoff (1925) as characteristic of basophil cells. The term /?" was proposed by these authors to replace the term "basophil" which they thought was too specific in meaning to characterize a type of cell which could be stained specifically in several ways, some of which did not appear to involve basophilia.

Berblinger and Burgdorf (1935), in a study of connective tissue in the human hypophysis, used the commercially available kresofuchsin as a stain for elastic tissue, and counterstained with orange G — phosphomolybdic acid and aniline blue. They noticed in the pars distalis that, although kresofuchin stained the granules of some of the basophil cells, there were other cells which did not differ in any visible respect except that their granules were not stained by kresofuchsin and were stained only by the aniline blue. Thornton (personal communication) and I too have observed the same phenomenon with paraffin sections of formalin-fixed human hypophyses stained with aldehyde-fuchsin. According to the age of the aldehyde-fuchsin solution and the duration of staining, the granules of a varying proportion of the basophils are stained with aldehyde-fuchsin, whereas others remain unstained in an erratic and nonreproducible fashion that cannot possibly represent a distinction between different functional cell types. Rodriguez (1937) used Berblinger and Burgdorf's staining combination in a study of the cells of the anterior lobes of the hypophyses of a number of mammalian species. He found kresofuchsin-stainable granulated cells in bovine, horse, sheep, dog, monkey, and rabbit hypophyses. He did not find such cells in the guinea pig or rat, although there were in these species cells with a diffuse staining.

Using resorcin-fuchsin for the staining of basophil cells, von Soos (1934) apparently ol)tained results similar to those produced by kresofuchsin. He found cells with resorcin-funchsin stainable granules in nine mammalian species but not in birds. In most species the number of basophil cells revealed by resorcin-fuchsin was similar to the number shown by Mallory staining of adjacent sections.

At the time of the above-mentioned investigations, Romeis was using a combination of kre^ofuchsin and Heidenhain's azan which, although differing in details from the method of Berblinger and Burgdorf, should have given, as far as the basophils were concerned, equivalent results. His results are published in his magnificent monograph (Romeis, 1940) on the hypophysis. Unlike the results of Berblinger and Burgdorf, kresofuchsin in Romcis's hands gave a clear-cut discrimination between two distinctive groups of cells, those whose granules stained with kresofuchsin and those whose granules stained not with kresofuchsin but only with the aniline blue of the azan stain. During his investigation Romeis found that the kresofuchsin supplied by the manufacturers differed from that he had obtained earlier in that, while still staining elastic tissue, it was no longer useful for staining the hypophysis. He found that resorcin-fuchsin he made himself could be substituted for kresofuchsin with results which were equal in every respect.

Romeis examined the hypophyses of a number of mammalian species other than man with results that differed from those of Rodriguez and von Soos in some respects. He found cells with granules stainable by kresofuchsin or resorcin-fuchsin in some species in which these authors had not been able to demonstrate them. As has been stated earlier, Romeis called the cells whose granules stained with kresofuchsin ^-cells, and those whose granules did not stain with kresofuchsin, but were stained with aniline blue, 8-cells. He found both /?- and 8-cells in the bovine hypophysis and in the hypophyses of the dog, horse, cat, and rat. He found /3-cells alone in the bat and guinea pig; in the latter they were very scanty. He also noted that kresofuchsin stained strongly the pars intermedia cells in the guinea pig, mouse, and pig.

The results that Romeis obtained with resorcin-fuchsin have been duplicated by Ezrin, Swanson, Humphrey, Dawson and AVilson (1958). Despite many attempts, other investigators have not been able to repeat these results using Romcis's method. There is, however, no doubt about the correctness of his observations.

Gomori (1950) introduecul ahU^liydofuclisin as an elastic tissue stain, and Ilahni (1950, 1951a) used it to differentiate between ^- and 8-cells in the rat and mouse. It has since been used successfully by a number of authors and on a number of species. The question whether aldehyde-fuchsin gives results equivalent to those once given by kresofuchsin is one which is not easily answered. The results are variable, depending on the fixative and on the treatment of the slide before staining, so that the term "stainable by aldehyde-fuchsin" cannot be given any precise meaning. Moreover, there are in the pars anterior of many, if not all mammals, three types of basophil cells, and according to conditions, the aldehyde-fuchsin may stain the granules of one, two or all three types, or fail to stain any of them. I think, however, that it may be assumed, that with any particular fixative and prior treatment of the slide, the elements most readily stained by kresofuchsin or resorcinfuchsin would also be most readily stained by aldehyde-fuchsin.

Aldehyde-fuchsin suffers from the defect of variability in staining power from batch to batch, and it often happens, as Romeis found with kresofuchsin, that a batch of this stain will stain elastic fibers but fail to stain basophil granules. Gabe (1953) described the preparation of an aldehydefuchsin which can be kept as a dry powder. To obtain results equivalent to those given by Gomori's preparation it is necessary to dissolve this powder in a concentration of 0.5 per cent in 70 per cent alcohol acidified with hydrochloric acid. Gabe's own procedure does not give the same results.

Recently Herlant and Racadot (1957) have indicated that Alcian blue (Steedman, 1950) at pH 1 stains the basophil granules in the toad and cat hypophysis that are stained by aldehyde-fuchsin. I find that a 3 per cent solution of Alcian blue in 70 per cent alcohol acidified with hydrochloric acid stains rat and human hypophyses with results equivalent to those obtained with aldehyde-fuchsin except that elastic fibers are not stained. H this procedure proves equally useful in other species it may supplant aldehyde-fuchsin and thus eliminate the uncertainty attending the use of this at present indispensable stain.

F. The Nomenclature Of Basophil Cells

One difficulty in the use of the term "^" for a basophil cell type whose granules stain with aldehyde-fuchsin and "8" for a basophil cell type whose granules do not stain with aldehyde-fuchsin is the presumption that only two types of basophil cells are present. In addition there is the possibility that with a different staining technicjue the staining reactions may be reversed, so that one man's /3-cells could be another man's S-cells. When three easily distinguished cell types are present, it is necessary to show that the granules of only one type are stainable by aldehyde-fuchsin before the term /? may be allocated. The naming of the remaining two types will be arbitrary whether one is called 8 or not. The major objection to the use of the terms /3 and 8 according to rules based on staining procedures is the lack of agreement as to which rule and which staining procedure should be used. Herlant (1956a) suggested that the terms 13 and 8 should be transposed in species other than man so that cells whose granules are stained by aldehyde-fuchsin would be called 8-cells, and the cells whose granules are unstained would be called ^-cells. There are in addition a number of usages of the terms /S and 8 to denote cells distinguished by staining reactions other than aldehydefuchsin in its several variants. It seems therefore advisable to avoid altogether the application of these terms according to rule until there is agreement as to which rule is to be followed. The usages of Romeis in man, Halmi in the rat and mouse, and Goldberg and Chaikoff in the dog can best be regarded as arbitrary.

In the present account cells whose granules are stained by aldehyde-fuchsin after any specified staining procedure will be stated to be AF-positive and called AF cells after such procedure. Cells whose granules are not stained by aldehyde-fuchsin after such a procedure will be stated to be AF-negative and called non-AF cells. It is emphasized that cells which are AFpositive after one procedure may be AFnegative after a different procedure.

The varying degrees of acidophilia of basophil granules allow in some species a distinction between "purple" basophils and "blue" basophils (Purves and Griesbach, 1957b). The granules of purple basophils retain more or less of the red component of azan or Crossmon or other trichrome staining methods and appear in purplish shades after procedures that result in a blue coloration of the granules of blue basophils. Purple basophils can be distinguished from blue basophils by counterstaining PAS stained sections with phosphotungstic acid-orange G (Herlant, 1953). The purple basophils, being relatively acidophilic, retain the orange G and become brick red, whereas the blue basophils remain magenta.

In sections stained by aldehyde-fuchsin and counterstained by a trichrome method we may distinguish purple AF cells, blue AF cells, purple non-AF cells and blue nonAF cells. Inasmuch as in many species there are in the pars anterior two basophil types which can be distinguished from each other by certain diagnostic features, although their granules stain alike, it is necessary to introduce additional terms to enable each cell type to be given a distinctive name. Such terms can be derived from the distribution of the cells in the pars anterior — "peripheral" or "central," "rostral" or "caudal" — or from some other diagnostic feature^"pale" for lightly granulated cells, etc. Such terms have no precise significance; they are used merely to generate a sufficient number of terms to enable each distinguishable cell type to receive a distinctive name. Thus, after fixation in formol-sublimate there are in the rat pars anterior blue AF cells (/3-cells of Halmi, functionally thyrotrophs), and peripheral and central blue non-AF cells (functionally FSH cells and LH cells, respectively). The advantage of these terms is that they are, or should be, free from any implication that cells similarly designated in different species must have the same function.

G. Specific Basophil Cell Types

It is now established that in the pars anterior of certain mammals (the rat, Purves and Griesbach, 1951a, 1954; monkey, Dawson, 1954b; bat, Herlant, 1956a; and dog, Purves and Griesbach, 1957a) three types of basophils can be distinguished. In the rat and bat, species which are favorable for the identification of cell types, the three basophil cell types have been identified as thyrotrophs, FSH cells, and LH cells. It will be noted that it is possible to distinguish cell types by specific staining reactions or by other diagnostic features without identifying them. By "identification" is meant here the allocation of specific function or functions to a distinguishable cell type.

It is not yet certain that all vertebrates have separate FSH and LH or that the physiologic reciuirements of all species would require the two hormones, if present, to be secreted separately and by different cells. In some species there may be only one type of basophil cell with gonadotrophic functions. It is likely, too, that in some species the duality of the gonadotrophic basophils is concealed by the fact that the staining reactions of the two types may be so alike that a distinction by tinctorial methods cannot be easily or consistently obtained. The term "gonadotroph" can be used as an inclusive term to denote the two types of cells secreting gonadotrophins and is also applicable in cases where a distinction of two such cell types has not been made.

Thyrotrophs and gonadotrophs have been identified in the guinea pig (D'Angelo, 1955), Xenopus (Cordier, 1953; Saxen, Saxen, Toivonen and Salimiiki, 1957), two species of fish, Phoxinus phoxinus (Barrington and Matty, 1955) and Caecobarbus geerstii (Olivereau and Herlant, 1954) , and in the pars distalis of the fowl (Payne, 1944, 1949; Brown and Knigge, 1958; Mikami, 1957 ». From observations recorded by Her

Fig. 3.8. Diagram of a horizontal section through the rat hypophysis showing the distribution of three types of basophil cell. P.D., pars distalis; P. I., pars intermedia; P.N., pars nervosa; R.L., hypophyseal cleft or residual lumen ; P.G., folliclestimulating hormone cells (peripheral gonadotrophs); C.G., luteinizing hormone cells (central gonadotrophs) ; T.. thyrotrophs.

lant (1954a) it seems that a differentiation of two kinds of basophil cells presumably thyrotrophs and gonadotrophs is generally observable in amphibians and bony fishes.

H. The Three Functional Types of Basophil Cell in the Rat Hypophysis

The identification of three functional types of basophil cells was first achieved in the rat by means which did not depend on specific staining reactions, the shape of the cells, their distribution and other minor features (Fig. 3.8). The pars anterior of the rat hypophysis has certain advantages which helped these identifications. Its hormone content and secretory function have been extensively investigated. It contains an adequate number of w^ell granulated basophils. Its small size allows its cell population to be evaluated under the microscope without an inordinate expenditure of time and the use of tedious cell enumeration techniques.

Another favorable quality shown by the rat hypophysis is its reactivity. The hormone content can be varied over a wide range by experimental procedures, and variations in secretion rate are accompanied by striking changes in the size and granule content of basophil cells. Especially conspicuous are the hyalinized cells derived from basophil cells. Under conditions of rapid secretion there is an accumulation of a structureless protein solution called hyaline substance in certain small cytoplasmic vesicles. By distension and coalescence of these vesicles one or more large hyalinefilled spaces result.

Evidence suggesting that the basophil cell class might consist of two distinct functional subdivisions, one concerned with thyrotrophic activity and one with gonadoti'ophic, was obtained from the study of the changes occurring after thyroidectomy and castration. In both these conditions an increase in the number and activity of cells of the basophil class are observed and the initial cellular hypertrojihy is followed by hyalinization of a proportion of the basophills. The hyalinized cells appearing after thyroidectomy (Figs. 3.9 and 3.10) are called thyroidectomy cells, those appearing after castration (Figs. 3.11 and 3.12) castration or signet ring cells. In the rat there are a large number of consistent differences between thyroidectomy and castration cells. Morphologic differences were observed by Hohlweg and Junkmann (1933), Zeckwer, Davison, Keller and Livingood (1935); Zeckwer (1936, 1937, 1938a, b), and by Giiyer and Clans (1935, 1937). A full account of the thvroidectomv and castration

Fig. 3.9 {npixr left). Section of the rat pais anterior showing the eaily response of the thyrotrophs to thyroidectomy (6 days after thyroidectomy). The chister of large pale cells is formed bv thyrotrophs which have degranulated but have not vet begun to hyalinize. PAS, X 540.

Fig. 3.10 (dipper right). Thyroidrctomy cells in the rat pars anterior 66 days after thyroidectomy. Extensive accumulations (if li>;ihne substance distort the cells. Normal specific granulation is absent. The coarse, inteiisrly stained granules (T granules) are characteristic of prolonged thyroxine deficiency states. PAS, X 780.

Fig. 3.11 (loiver left). Section of the rat pars anterior showing changes in follicle-stimulating honuonc cells 10 wicks aft( i- laslration. The cells are numerous and large and a high proporI Kin lia\ r arciiiiiulaird hyahiir iiiaicrial to form "signet-ring" cells. There is an increase in the i-diitciit (if granule- wliicli fdini cdarse, darkly stained clumps. PAS, X 580.

Fig. 3.12 {lower right). Section of the rat pars anterior showing changes in luteinizing hormone cells 16 weeks after castration. A proportion of the cells have accumulated hyaline material and formed "signed-ring" cells. Both hyalinized and nonhyalinized cells show varying intensities of staining indicating considerable variation in granule content in different cells. There is often a higher content of granulation within the Golgi body than in the peripheral cytoplasm. The characteristic difference between the appearances of the granulation in follicle-stimulating hormone and luteinizing hormone cells is retained after castration. PAS, X 580.


Differences between Castration and Thyroidectomy Cells in the Rat

Castration Cells

Thyroidectomy Cells


Vesiculation gives rise to "signet

Vesiculation less regular

Schleidt (1914)

ring" cells

Vesicles are large, smooth, and well

Secretory accumulation less regular

Zeckwer et al. (1935)


Early stages do not resemble the

Early stages full of fine foamy vesicles

Zeckwer et al. (1935)

early stages of thyroidectomy cells

coalescing to form larger ones at 14th to 18th day

Not vesiciilated at 5 weeks

Large amounts of hyaline material at 5 weeks

Zeckwer (1938a, b).

Uniformly granulated

Poorly granulated

Zeckwer (1938a, b).

Suppressed by estrogen

Not affected by estrogen

Zeckwer (1938a, b).

Generally a large single vesicle

Usually multiple vesicles

Guyer and Claus (1935, 1937)

Cells usuall}' scattered

Cells commonly grouped

Guver and Claus (1935, 1937)

Nuclei not vesiculated

Nuclei commonly vesiculated

Guyer and Claus (1935, 1937)

Golgi body conspicuous and regular

Golgi body diffuse

Guyer and Claus (1935, 1937) Reese, Koneff and

Cytoplasm forms a definite band

Partitions between multiple vesicles

around the vesicle

are very fine and no band of cytoplasm at the periphery

Wainman (1943)

Final size not great

Final size very great

Reese, Koneff and Wainman (1943)

Shape compact, oval, or round

Shape irregular, polyhedral

Only a proportion of the cells hya

Practially all hyalinized

Reese, Koneff and


Wainman (1943)

Nuclei and nucleoli not enlarged

Nuclei and nucleoli enlarged and ve

Reese, Koneif and


Wainman (1943)

Mitochondria smaller

Mitochondria very large

Reese, Koneff and Wainman (1943)

cells was given by Reese, Koneff and Wainman (1943). The recorded differences between the two types of cells in the rat are summarized in Table 3.1.

Castration basophils, at first only weakly staining, become more strongly staining in the period from 2 weeks up to 4 months from operation. In long term experiments one of the striking differences between castration cells and thyroidectomy cells is the strong staining by aniline blue of the cytoplasmic granules of the castration cells whereas the thyroidectomy cells take so little stain that they have been considered by many observers to be chromophobes.

Zeckwer, Davison, Keller and Livingood (1935) were the first to trace the differences between these types of cells and to infer front them that these cells were the results of similar changes occurring in two distinct kinds of basophil cell. Guyer and Claus (1935; 1937), while observing differences in the cells, considered they were exjilicable as being different processes occurring in the same cell type. Severinghaus (1939) did not consider the changes produced by thyroidectomy different in any fundamental way from those produced by castration. He was no doubt misled by the fact that a certain number of "signet-ring" cells are always present (Wolfe, 1941, 1943) and arc therefore found in small numl)ers among the thyroidectomy cells in thyroidectomized animals.

The next important clue was pr()\i(l('(l by Reese, Koneff and Wainman ( 1943 ) who obsel•^•ed that two classes of basophil cells are normally present in rat hyi)ophyses, one oval or round in section (Figs. 3.13 and 3.14), the other polygonal and angular (Fig. 3.15). They noted that in the early response to thyroidectomy there is an increase in the iiiiiiibci' of cells which are irregular or i)olyh('(hal, whereas aftei- castration the in

Fig. 3.15 (upper Icjt). Section aldehyde-fuchsin (AF). The cells granules to be concentrated at the

1mi,. :;.1:] i '-/'/" - "<//'/ '• >< '-'idii of the mt pans aiUciiur showiug, iulliclf-.-liiuulating liormone cells stained by periodic acid-Schiff (PAS). The cells are ovoid in shape and contain clumps of densely stained granules. PAS, X 720.

Fig. 3.14 (lower left). Section of the rat pars anterior showing luteinizing hormone cells stained by periodic acid-Schiff (PAS). ■M The cells have rounded con tours and the

la basophil granules are more unil'orinly dis Wa perscd throughout the cytoplasiu than in the

^ other two basophil types. One cell near the

center of the field shows a prominent negative image of the Golgi body. The cytoplasm enclosed by the Golgi body has a higher ' granule content that the peripheral cyto plasm. PAS, X 720. of the rat pars anterior showing thyrotrophs stained by have angular contours and there is a tendency for the periphery. AF, X 720.

created number of basophils is due to an increase in conijiact cells which are oval or round.

The McManus (1946) PAS reaction for the demonstration of glycoproteins in histologic sections initiated researches based on the idea that this reaction might demonstrate the hormone content of basophil cells. Herlant (1949, 1951) and Pearse (1949, 1951 ) ascribed the reaction given by basophil granules to the presence of gonadotrophin. Catchpole (1949) found in castration cells and in certain basophil cells of normal animals glycoproteins which could have been gonadotrophins. He looked for but did not find glycoprotein with the solubility of thyrotrophin in thyroidectomy cells. He considered that the glycoprotein staining did not parallel the staining with aniline blue.

Purves and Griesbach (1951a) set out to test the hypothesis that the PAS reaction could. be used to demonstrate the hormones of basophil cells by examining rat hypophyses in which thyrotrophin or gonadotrophin or both were absent. Such hypophyses were obtained by treatment of intact rats with thyroxine or estrogen or both thyroxine and estrogen. In hypophyses that contained only thyrotrophin the glycoprotein reaction and the aniline blue-stained granules were confined to cells that were polyhedral in shai:)e and were distributed throughout the interior of the pars anterior. In hypophyses containing only gonadotrophins the glycoprotein reaction and the aniline blue-stained grnnules were confined to cell^ that were ovoid or spherical. The most conspicuous staining was in cells in the periphery of the pars anterior, especially on the inferior surface and at the anterior margin. In hypophyses which contained neither thyrotrophin nor gonadotrophin basophil cells could not be demonstrated by either the glycoprotein reaction or aniline blue.

These results showed that the different hormones were stored in the granules of different cells and that the hormone content was proportional to the granule content. Any staining method which demonstrated the granules would give the same result and, had the Dempsey and Wislocki (1945) suggestion that aniline blue stained the hormones been followed up, results similar to those obtained by the PAS reaction would have been obtained. The PAS reaction provided a more specific staining for basophil granules; its more important role, however, was the psychologic stimulus it gave to investigations relating staining to hormone content because there was a chemical reason for supposing that this reaction would demonstrate the content of thyrotrophin and gonadotrophins.

Purves and Griesbacli < 1951a) gave the name "thyrotrophs" to the cells whose granules contain thyrotrophin and the name "gonadotrophs" to the cells whose granules contain gonadotrophins.

Halmi (1951a, b) applied Gomori's (1950) aldehyde-fuchsin to the staining of the i)ars anterior and obtained the distinction between yS-cells and 8-cells that had previously been obtained by Romeis (1940). It was soon established that the ^-cells of the rat pars anterior are thyrotrophs (Purves and (h-iesbach, 1951b; Halmi, 1952a).

Restaining iJroccdincs demonstrate that tbe same cytoi)lasmic granules are revealed by aldehyde-fuchsin, ])hosphotungstic acidaniline l)lue, and the histochemical reaction for glycoprotein. The specific staining of the thyrotrophs is therefore a specific staining of thyrotrophin-containing granules. In rats wliich have been treated with thyroxine the granulation is reduced to such a low level that the thyrotrophs cannot be demonstrated by any of the staining i)roce(hn'cs (Purves and Griesbach, 1951c; Halmi, 19511), 1952b). In rats which have been thyroidectomized there is a rapid discharge of thyrotrophin which reduces the granule content in the cells to a low level. There is simultaneously a marked increase in the number of functioning thyrotrophs and an increase in the size of the individual cells. The total hormone content of the gland is affected in opposite directions by these two effects, and may not be greatly different from normal despite a low concentration of the hormone in the cell cytoplasm. Under these conditions staining by aldehydefuchsin is not obtained although the cytoplasm and the hyaline substance are still stainable by other methods. The thyrotrophs, therefore, after thyroidectomy are easily confused with 8-cells although they are still cleary differentiated from 8-cells by their characteristic shape (Purves and Griesbach, 1951b).

The specific staining of thyrotrophin-containing granules by aldehyde-fuchsin constitutes a notable advance in the study of the functional aspects of pituitary morphology. The differentiation of thyrotrophs from other cells by Halmi's (1951a) method clarifies not only the study of thyrotrophic function, but also that of gonadotrophic functions in the rat pituitary because the two types of basophils concerned with the secretion of the gonadotrophic hormones appear as 8-cells which can be studied without confusion with thyrotrophs.

It was soon apparent (Purves and Griesbach, 1952; Si{)erstein, Nichols, Griesbach and Chaikoft', 1954) that the gonadotrophs in the rat pars anterior are of two types. The peripheral gonadotrophs near the surface of the anterior lobe and especially concentrated on the inferior surface and at the anterior border show a consistently different a])pearance from the i)aler rounded cells whichare scattered throughout the interior and arc the only basophils adjacent to the l)ai's intci'niedia (Fig. 3.9).

This suggests the hypothesis that one form secretes FSH and the other LH. The central gonadotrophs are the more activelooking cells after castration and during pregnancy. There was therefore a conflict l)etween the at one time current view that after castration the secretion is predominantly follicle-stimulating and the probability that the hormone secretion during pregnancy is predominantly luteinizing. Af cer a year of investigation it was realized that the solution lay in getting rid of one of the hormones.

Treatment of the adult female rat with testosterone propionate had been shown to increase the gonadotrophin content of the hypophysis and to produce a gland which contains FSH without any admixture of LH (Laqueur and Fluhmann, 1942; Greep and Chester Jones, 1950).

Purves and Griesbach ( 1954) found that in female rats treated with 250 fxg. of testosterone daily for 1 week there was a large increase in the numbers of gonadotrophs of both types which could be stained by the glycoprotein reaction. After 3 weeks treatment, the pale central gonadotrophs were found diminished in size and showed signs of diminishing activity. After 4 weeks treatment, the central gonadotrophs had practically disappeared whereas large numbers of the peripheral type with a strong cytoplasmic glycoprotein storage remained. These peripheral gonadotrophs did not seem to be actively secreting because the cells themselves were small.

These observations seem to indicate with certainty that the peripheral type of gonadotroph is exclusively the site of formation and storage of FSH, whereas the central type is probably concerned with the formation and storage of LH.

It has not yet been possible to obtain rat hypophyses containing LH without FSH. In glands containing large amounts of LH there are always large numbers of richly granulated central gonadotrophs and there is no reason to think that the granules of these cells contain any other hormone.

The cells that secrete and store within their granules FSH are called FSH cells; the cells that secrete and store within their granules LH are called LH cells.

As yet, no consistent staining difference between the FSH granules and the LH granules has been obtained. The Wilson and Ezrin (1954) technique differentiates between thyrotrophs and gonadotrophs, the former being PAS-red whereas the latter tend to be PAS-purple. The granules of the thyrotrophs have little affinity for acid dyes and do not take up either the orange G or methyl blue applied as counterstains to sections which have first been stained by the PAS reaction. Their color is therefore a magenta red. Cells with a strong affinity for acid dyes take up orange G and hold it firmly. Strongly acidophilic basophils (amphophils) of this kind are found among the /3-cells of the human pars distalis; they appear brick-red after the Wilson and Ezrin staining method, the color resulting from the addition of orange G to the PAS color. These cells are also called PAS-red by Wilson and Ezrin. The difference in the two shades of red is visible in Figures 1 and 2 of Wilson and Ezrin's (1954) paper. In basophil cells with an intermediate degree of acidophilia the orange G which is taken up first is displaced more or less rapidly by methyl blue in the next step of the procedure. These cells therefore appear purple. The gonadotrophs of the rat have an intermediate degree of acidophilia and may be stained PAS-purple, but the colors obtained are not independent of the amount of granulation present. In this technique there is a tendency for strongly granulated cells to retain the orange G longer and to appear PAS-red whereas in lightly granulated cells the orange G is more rapidly replaced by methyl blue and the cells appear PAS-purple. The color differentiations obtained by Rennels (1957) and by Hildebrand, Rennels and Finerty (1957) resulted from this effect. In the normal rat the FSH cells are more densely granulated than the LH cells and can be stained PASred while the LH cells are stained PAS-purple, but when densely granulated LH cells appear in long term castrates they too are PAS-red. Ten days after castration when both FSH cells and LH cells are only lightly granulated both stain PAS-purple. I have not found any color differentiation between FSH cells and LH cells containing a similar amount of granulation. The conclusions of Hildebrand, Rennels and Finerty concerning the distribution of FSH cells and LH cells are therefore not likely to be correct.

Despite the absence of specific staining, the granules of FSH cells and LH cells can be shown to be different by solubility tests. I have made a considerable number of tests of this kind. The best results have been ol)tained by perfusion of long term castrate rats with an ether-saturated isotonic acetate buffer at pH 4.5 or 5. Prussian blue has been added to the perfusion medium so that the completeness of the perfusion can be checked. Under these conditions the glycoprotein of the granules of FSH cells was almost entirely extracted whereas that of the granules of LH cells was to a large extent retained. Assays showed that the buffer extraction removed the FSH and some of the LH; the extracted glands contained LH without FSH. Barrnett, Ladman, McAllaster and Siperstein (1956) obtained a similar distinction between the granules of FSH and LH cells by immersing hypophyses in 2.5 per cent trichloracetic acid. This treatment removed most of the glycoprotein from FSH cells and preserved the glycoprotein of LH cells intact. Assays showed a loss of FSH and a preservation of LH.

I interpret these observations as indicating that basophil granules, like other secretion granules, contain a mixture of proteins and that the glycoproteins of LH-cell granules are less soluble than those of FSHcell granules. The loss of thyrotrophin and FSH from whole rat hypophyses immersed in 2.5 per cent trichloracetic acid must be ascribed to inactivation rather than to complete extraction, but it may well be that the preservation of LH under these conditions is related to its insolubility in 2.5 per cent trichloracetic acid. It should not be inferred however that all glycoproteins either insoluble in, or rendered insoluble by trichloracetic acid are LH. The conclusion of Barrnett, Ladman, McAllaster and Siperstein that FSH often occurs in conjunction with LH should be regarded as unproven.

I. The Pars Anterior of the Bat With Special Reference to two Types of Gonadotrophs

The pars anterior of the hypoi:)hysis of the bat {Myotis '>nyotis) has been studied by Hcrlant (1956a). Five types of cells are distinguished by staining reactions. Two are acidophils and three basophils. Aided by the favorable staining properties of the cells and by the unusual nature of the reproductive cycle, Hcrlant has been able to mak(> identifications of FSH cells and LH ('(■Us in the bat that carry more conviction than the id(nitifications made in tiie rat.

The distinctive staining of the granules of FSH cells and LH cells in the bat makes it possible to assert that the two types are quite separate, and do not change from one type to another at different times.

The two acidophil types in the bat are: (1) Orangeophils. These are relatively stable throughout the reproductive cycle and are presumed to secrete somatotrophin. (2) Carminoi)hils. These are concentrated in the anteromedian zone. They are in evidence at the time of ovulation, become less prominent towards the end of pregnancy and show intense activity accompanied by a high content of cytoplasmic ribonucleic acid in animals that lactate after parturition. They are considered to secrete prolactin.

The three basophil types are: (1) Irregularly shaped basophils scattered through the par anterior, staining blue by trichrome methods and with a marked affinity for aldehyde-fuchsin. This type which is relatively stable throughout the reproductive cycle is provisionally identified as a thyrotroph. Its response to disturbances of thyroid function has not been tested.

(2) Basophils concentrated in the anteromedian zone, staining blue by trichrome methods and with a slight aflfinity for aldehyde-fuchsin. These elements are large and well granulated at the time of estrus in the autumn and continue to show an active appearance throughout the period of hibernation when ripe follicles are in the ovaries. They involute after ovulation in the spring and remain involuted during pregnancy. These cells are identified by Herlant as FSH cells.

(3) Basophils in the i)ostcrior two-thirds of the pars anterior which show a marked tendency to retain the red component of trichrome staining methods and are therefore violet or puri:)le but not blue after such procedures. They are brick-red in sections stained by PAS and counterstained with orange G, whereas the other two basophil types which do not stain with orange G are magenta. These cells are greatly hypertrophied through i)regnancy but undergo |)rompt involution after parturition regardless of whether lactation follows. Herlant identifies these cells as LH cells and this identification is in accord with the marked activity of LH cells during pregnancy in the rat. By immersing bat hypophyses in 10 per cent trichloracetic acid for 12 hours before fixation Herlant found that the granules of FSH cells were soluble, those of the LH cells insoluble in this medium.

IX. Corticotrophin : the Problem of Its Origin

A. Basophil Cells and Corticotrophin

Many attempts have been made to relate corticotrophin secretion to one or other of the specific cell types demonstrable by staining reactions, but the results have generally been inconclusive. The distribution within the anterior lobe of species which show distinct zoning is instructive. Smelser (1944) showed that corticotrophin was more concentrated in the basophil zone of the bovine pars anterior than in the acidophil zone. In the pig pars anterior, in which the zones are much more distinct, I have found that corticotrophin is present in portions of the anteromedial basophil zone that are apparently free from acidophil cells, the concentration there being 2 to 4 times that in the posterolateral acido})hil zone. The assays were made by the ascorbic acid depletion method. Giroud and Martinet (1948) found that the adrenal weightincreasing action is predominantly situated in the basophil zone. Rochefort and Saffran (1957) found 4 to 13 times as much corticotrophin in the anteromedial basophil zone of pig and beef hypophyses as in the posterolateral parts of the acidophil zone. Certainly in both cow and pig the corticotrophin is not in acidophil cells. In the cow it may be in either basophils or chromophobes, but in the pig it seems certain that it must be in some type of basophil cell, because there are few chromophobes in the pig pars anterior and a large part of the basophil zone is free of them ancl consists of basophil cells only.

Halmi and Bogdanove (1951) and Hess, Slade, Amnions and Hendrix (1955) found that the loss of acidophil granules in the thyroxine-deficient rat was not accompanied by any change in the content of corticotrophin in the hypophysis, nor was there any change in corticotrophin output. This indicates that the hormone is not in the acidophil granules in this species and is probably not formed by acidophil cells.

One indication of the possible nature of the granules to be expected in corticotrophin-secreting cells may be made from the the chemical nature of their hormonal product. This polypeptide has some relationship to intermedin, and the amino-acid sequence — methionine, glutamic acid, histidine, phenylalanine, arginine, tryptophan, glycine — is common to both (Landgrebe and Mitchell, 1958) . The presence of this structure in corticotrophin may account for its having some intrinsic melanin-dispersing properties. The intermedin-secreting cells of the pars intermedia may in some species have enough specific granulation to distinguish them from chromophobes, but when granules are present they are basophil granules containing glycoprotein and are stain:ible with aldehyde-fuchsin. We might expect, therefore, that if corticotrophinsecreting cells contain granules they will be basophil granules, although it is not necessary that they should contain such granules.

The origin of corticotrophin from basophil cells has gained a certain degree of acceptance based on the association of basophil cell adenomas with the Gushing syndrome in man. The evidence now is that these basophil cell changes are the result of the action of excess adrenal steroids and are not the cause of the syndrome even in those cases which seem to be due to excessive secretion of corticotrophin. This aspect of the subject is dealt with more fully in the later section on Gushing's syndrome.

The attempt by Marshall (1951) to demonstrate the presence of corticotrophin in basophil cells by the use of labeled antibody must be discussed, because it has been regarded by some as a conclusive demonstration of the relation of this hormone to basophil cells. The antibody labeled with a fluorescent grouping demonstrated the presence in basophil cells of an antigen which had been contained in the crude corticotrophin preparation used. There is, however, nothing to associate the antigen so demonstrated with the corticotrophin. Such protein-containing preparations of corticotrophin seem to contain many more molecules of protein than of corticotrophin, and the antigen could therefore have been one of the glycoprotein hormones known to be antigenic, or some nonhormonal constituent. The absence of evidence of antigenicity in corticotrophin would make the method inapplicable for this hormone unless it could be shown that the hormone w^as associated in the gland with a specific protein. There is no evidence so far that this is the case, because under certain circumstances the hormone is dialyzable from crude pituitary extracts (Tyslowitz, 1943; Geschwind, Hess, Condliffe, Evans and Simpson, 1950j , and in the extract it seems to be distributed indiscriminately among most of the proteins present (Astwood, Raben and Payne, 1952).

In the rat three types of basophil cells have been identified, and the question arises whether one of these secretes corticotrophin in addition to its specific glycoprotein hormone. The hypothesis that thyrotrophs may secrete corticotrophin has attracted some attention and has been subjected to specific investigation. Halmi and Bogdanove (1951) concluded that corticotrophin was not produced by thyrotrophs in the rat. In the pig hypophysis I have found, as Smelser (1944) found in the bovine hypophysis, that the distribution of the hormones between the basophil and acidophil zones is quite different for corticotrophin and thyrotrophin, the latter being almost exclusively in the basophil zone. This seems to make it certain that the same cell does not produce these two hormones.

The problem of the origin of corticotrophin thus seems to resolve into a choice between two alternatives: either corticotrophin is made by gonadotrophs, or it is made by an aditional specific cell type, the "corticotroph." Inasmuch as five specific cell types have been identified in the pars anterior of the rat and the bat, the search for an additional cell type which could be the corticotroph may be referred to as the search for a sixth cell type. Investigations aimed at the solution of this problem usually take the form of examining hypophyses in which corticotrophin secretion is proceeding at an abnormally rapid rate, and looking for cellular responses either in sections stained by methods which reveal acidophils and basophils or in sec

tions stained by unconventional methods which might stain the granules of a sixth cell type not revealed by conventional methods.

B. Hypophyseal Responses to Adrenal Ablation

Many reports of the cytologic changes observed after bilateral adrenalectomy have been published. Some observers have not remarked on any extensive alteration. Thus, Nicholson (1936) did not observe any change in the pars anterior of dogs after bilateral adrenalectomy, and Koneff, Holmes and Reese (1941) found that in rats to which sodium chloride was administered after adrenalectomy, cytologic changes were minimal. Houssay (1952) also remarked that in strains of rats which survive in good condition after adrenalectomy there is little hypophyseal disturbance compared with those which have severe insufficiency symptoms.

It therefore seems that the increased secretion rate or the depletion of the adrenocorticotrophin content of the hypophysis which follows adrenalectomy need not be accompanied by any marked change in the appearance after staining by the customary methods.

Certain changes in the acidophil and basophil cells which are observed in certain strains of rats after adrenalectomy, if additional salt is not administered, must be related to the extensive metabolic disturbances produced by the adrenal insufficiency. These changes have been described by Reese, Koneff and Akimoto (1939) . Changes in the acidophil cells consist of degranulation with a diminution in the size and number of cells. The Golgi body in these regressing acidophils is transformed from the usual acidophil form which envelops half the circumference of the nucleus into a concentrated spherical or oval body. These changes indicate decreased activity. Cells of the basophil class also show extensive degranulativc changes with reduction in size and number, but some basophil cells with enlarged Golgi bodies are seen, especially at short intervals after adrenalectomy. Tuchmann-Duplessis (1953) also found that, although most of the basophil cells undergo degenerative changes after adrenalectomy, there remain always a few active normal-looking cells. Colombo (1948, 1949) reported in rats, after bilateral or unilateral adrenalectomy, a change in basophil cells similar to that occurring after castration and not accompanied by any change in acidophil cells.

Griesbach (personal communication) observed that the LH cells were enlarged after adrenalectomy in the rat. There were also changes in thyrotrophs similar to those produced by exposure to cold. The changes in the thyrotrophs were prevented by thyroxine administration. Knigge (1957) found the thyrotrophs diminished in numbers after adrenalectomy. The gonadotrophs (8-cells of Halmi) were unchanged in numbers but were hypertrophied, and some of them were hyalinized 8 weeks after adrenalectomy. Knigge suggested that the 8-cells were the source of corticotrophin.

Rokhlina ( 1940) tested the effects of combining adrenalectomy with castration in rabbits and rats. In some rats thus treated no castration changes appeared in the basophil cells, whereas in others the transformation was delayed. Adrenalectomy performed on the 30th day after castration had some effect on the further develojimcnt of castration changes.

Brokaw, Briseno-Castrcjon and Finerty (1950) performed unilateral adrenalectomy on rats. By this means a relative adrenal insufficiency should be produced without the extensive metabolic disturbances accompanying complete adrenal deficiency. They observed a temporary increase in the pituitary acidophil cell percentage; the normal pituitary cytology was regained in approximately 50 days.

Herrick and Finerty (1940) found in adrenalectomized fowls an alteration in basophils in the pars distalis, which they correlated with the regression of the testes. The basophils showed progressive vesiculation, the vesicles being filled with hyaline material. They considered these basophils to be in a degenerating state. IMikami (1957) observed a degeneration of both thyrotrophs and gonadotrophs in the fowl after adrenalectomy. In addition there was a degranulation and remarkable enlargement of a third basophil type in the rostral zone of the pars distalis. These cells, designated by Mikami "V cells," were considered to secrete corticotrophin.

In the dog, Mikami (1956) found that the ^cells of Goldberg and Chaikoff (1952a) increased in number and enlarged in size after adrenalectomy. These were the only cells showing signs of increased activity after adrenalectomy. The ^-cells of Goldberg and Chaikoff are the cells termed "pale" cells by Purves and Griesbach (1957a) and are basophil cells with a low content of granules.

C. Hypophyseal Responses to Stress

An increased secretion of corticotrophin is produced in response to many different kinds of stress. Inasmuch as adaptation to various kinds of stress wall involve changes in a number of hormones, the hypophyseal responses to stress will be less suitable for the study of corticotrophin production than the specific disturbances produced by adrenal insufficiency. Thus, the response to cold exposure involves increased secretion of thyrotrophin as well as corticotrophin, and the increased activity in the basophil cells of the rat exposed to cold have been related thyrotro{)hin production rather than corticotrophin (Brolin, 1945; Allara, 1953; ]McXary, 1957). In rats exposed to cold, Griesbach, Hornabrook and Purves (1956) observed partial degranulation of thyrotrophs and early hyalinization giving rise to cells with a resemblance to Crooke's cells. This appearance is related to alteration in thyrotrophin secretion and can be prevented by thyroxine injections, but not by replacement with cortisone.

An increased granule content in basophil cells in the rat during inanition has been related to an increased storage of gonadotrophin (Pearse and Rinaldini, 1950). The increase in stainable basophils observed in the hypophyses of animals subjected to various forms of stress may, therefore, be related to changes in the content of the glycoprotein hormones which are the specific secretory products of these cells. Herlant (1936a, b) observed that the increase in stainable basophils in the rat hypophysis after ligation of the ureters or after the injection of hydrochloric acid, was accompanied by an increase in the gonadotrophic potency of the pituitary tissue.

Finerty and Binhammer (1952» and Finerty, Hess and Binhammer (1952) studied the early responses of the hypophysis to the acute stress of severe burns in the rat. No changes were observed in the differential cell counts nor in the degree of specific granulation of the cells stained by the azan method. There was, however, an increase in the number of acidophil cells per field in sections stained by the acid hematein method of Rennels (1951). Timmer and Finerty (1956) supplied the explanation of this discrepancy. The results of azan staining were expressed by differential cell counts and did not show any change. The results of the acid hematein staining were expressed as cells per field, and the increase in this number was the result of a shrinkage in the gland which occurred after scalding. Thus, although there were no more acid hematein stained cells in the hypophysis, they were closer together after scalding.

Knigge (1955) found, in thyroidectomized animals in which acidophil granules were absent, that there was no acid hematein staining before or after scalding.

D. The Sixth Cell Type

The pars anterior of the rat is particularly reactive to changes in the rate of hormone secretion, and the absence of any striking change in the stainable cells after adrenal ablation suggests that corticotrophin may be secreted by chromophobes in this species, or that corticotrophin is secreted by a different mechanism from that responsible for the secretion of other anterior lobe hormones. The discovery by Farquhar (1957) of a sixth cell type in the pars anterior of the rat may provide an explanation of these peculiarities. These cells, which can be clearly seen only by electron microscopy, are very different from any of the other five types. The cytoplasm is relatively empty and contains few formed elements (mitochondria, endoplasmic reticulum). Secretory granules are absent.

A distinctive feature of this additional cell type is its location throughout tiie anterior lobe in groups around follicles or ductules which contain colloid of low density. Some follicles are large and undoubtedly are analogous to colloid cysts, but small follicles which probably eould not be distinguished by light microscopy are much more numerous. The cells lining the follicles have angular contours, and the nucleus is eccentrically located. The amount of colloid seems to vary with the amount of corticotrophin in the gland, being increased after cortisone injection and decreased after partial adrenalectomy. No marked response of the cells to the stimulation of partial adrenalectomy was observed. The identification of this sixth cell type as a corticotroph can only be regarded as tentative. Farquhar suggested that corticotrophin may be stored in the form of colloid and released by some mechanism similar to that by which thyroid hormone is mobilized from the colloid of the thyroid follicles. This would explain the difficulties in relating corticotrophin secretion or storage to changes in the granulated cells observed by the light microscope.

My own observations show that this cell often has an extremely irregular shape with long and tenuous cytoplasmic projections extending between adjacent granulated cells or deeply indenting their cytoplasm. The cross section of the enclosed space is often elongated, suggesting that this cavity often takes the form of a cleft rather than a follicle or tubular ductule.

X. The Pars Intermedia and Intermedin Secretion

Intermedin is an adenohypophyseal hormone, distinct from the other adenohypophyseal hormones. This is evident from the fact that it is formed in a different site. In addition to causing dispersion of pigment in mclanophores, intermedin acting over a long time also stimulates the formation or deposition of melanin. Intermedin is secreted by a specific cell type in the adenohyjiophysis. The recognition of this specific tyjie is simple in most vertebrates because the cells are concentrated into a zone, the pars intermedia.

Evidence of the specific hormone content of the pars intermedia was first obtained by Smith and Smith (1923b) in tests of the different regions of the beef hypophysis, and has been confirmed by Lewis, Lee and Astwood (1937), and r.iroiid and :\Iartinet (1948).

Intermedin seems to be located in a small granule fraction in sucrose solution homogenates of pars intermedia tissue, because it sediments with the microsomal fraction on centrifugation (Jeener and Brachet, cited by Herlant, 1952a). In this respect it behaves like the hormones secreted by basophil cells in the pars anterior.

Evidence of specific secretion from pars intermedia tissue is provided by the interesting studies of Allen (1930) and Etkin (1941) in transplantation experiments in tadpoles. Their results indicate that severance of the stalk connection with the hypothalamus removes an inhibiting influence from the pars intermedia cells, and that the resultant hypertrophy and hyperplasia is accompanied by oversecretion of intermedin which produces a permanent blackening of the animals. Etkin (1958) found that the capacity of the pars intermedia to differentiate and secrete was not prevented by transplantation of the epithelial primordium of the pituitary to the tail bud of the wood frog where it developed away from any neural tissue. It was, however, in contact with the neural epithelium before transplantation and before Rathke's pouch had begun to form. Copeland (1943) observed in Triturus viridescens that the differentiation of the pars intermedia occurs immediately before the first metamorphosis, and at this time the characteristic pigmentation of the red eft stage begins to appear.

The pars intermedia, unlike the pars nervosa, does not undergo degeneration after stalk section in mammals (Rasmussen and Gardner, 1940; Brooks, 1938; Stutinsky, Bonvallet and Dell, 1950; Barrnett and Greep, 1951). Indeed, a hypertrophy of the pars intermedia is usually observed with signs of cellular activation. Fisher (1937) demonstrated that intermedin is still present in the cat hypophysis after the pars nervosa has degenerated after stalk section.

The intensity of specific staining of pars intermedia cells varies with the species. This variation is the result of variation in the cjuantity of specific granulation. The granulation when present appears to contain glycoprotein, because it gives a positive PAS reaction and is stainable by aldehyde-fuchsin without prior oxidation. The glycoprotein character of the intermedia cell granulation has been demonstrated in the bat by Herlant (1956a) and in the frog by Ortman (1954, 1956c). In both species the granules are stained by aldehyde-fuchsin. I have observed the same staining reactions in the granules of intermedia cells in the cat, dog, sheep, deer, and rat (Fig. 3.16). In the rat the glycoprotein reaction and the aldehydefuchsin staining are negative after extraction by perfusion with saline saturated with ether or by immersion in a neutral buffer after acetone treatment. Intermedia cells are, therefore, typical basophil cells contain cells by aldehyde-fuchsin (AF). The colloid of the hypophyseal cleft is not stained. PX, pars nervosa; PI, pars intermedia; HC, hypophyseal cleft; PA, pars anterior. Formol sublimate fixation, AF X 43.

Fig. 3.17 (light). Section of the rat hypophysis showing the alteration of the staining properties of the pars intermedia by fixation in Helly's fluid. Except for a few coarse granules of unknown nature, the pars intermedia cells are almost unstained in comparison with Figure 3.16. Key to lettering as in Figure 3.16. Helly, AF, X 150.

ing granules with a content of soluble glycoproteins. In the rat the granules are much more strongly stained by aldehyde-fuchsin than by PAS, and show an additional difference of behavior from the basophil granules of pars anterior cells in that their staining by aldehyde-fuchsin is not enhanced by prior oxidation with acid permanganate (Halmi and Davies, 1953).

The granules of intermedia cells stain a blue or purple color by trichrome staining methods (Romeis, 1940). In the cat, in which the staining reactions of intermedia cells are strong due to a high content of granulation the appearance of intermedia cells in stained sections is similar to that of typical pars anterior basophils.

The relation between the specific granules and the hormone secretion of the pars intermedia cells is still obscure. The hormone can be obtained as a peptide, but it may be that a combination of this peptide with the glycoprotein is the form in which the hormone is first produced and stored. Ortman (1954, 1956c) found in the frog that the glycoprotein granules of the pars intermedia cells were depleted during dark adaptation, but did not find any accompanying reduction in hormone content.

Not all hypophyses have a pars intermedia, although so far as is known all contain intermedin. In birds as a class, and in certain mammals, porpoise, whale, armadillo, manatee, elephant, pangolin, beaver, and man, the pars intermedia is absent. The problem of the cellular origin of the intermedin in hypophyses of this type is of great interest because of indications that one of the basophil cell types of the human pars distalis is the intermedin secretor of this species. In the porpoise and whale intermedin is present in the tissue of the adenolobe (Oldham, Last and Gelling, 1940), and although the specific cells which produce the hormone have not been identified, it must be assumed that they are present, scattered throughout the pars distalis. A similar distribution of intermedin was shown in the beaver by Kelsey, Sorenson, Hagen and Clausen (1957). In birds the intermedin is found only in the rostral portion of the adcnolol)c (De Lawder, Tarr and Gelling, 1934; Mialhe-Voloss and Benoit, 1954). In hen and duck hypophyses I have found aldehyde-fuchsin positive basophil cells in the rostral zone of the pars distalis, but the correlation with the distribution of the hormone is obscured by the presence of other aldehyde-fuchsin positive cells in the caudal zone. In the white-crowned sparrow [Zonotrichia leucophrys gambellii) aldehydefuchsin positive basophils are found only in the rostral zone. Traces of thyrotrophin are present; the concentration is the same in both zones. It may be that the aldehydefuchsin positive cells of the rostral zone are the intermedin secretors.

Much more definite information about the human hypophysis is available from the studies of ]\Iorris, Russell, Lanclgrebe and Mitchell (1956). These investigators correlated the hormone content of small portions of tissue from the anterior lobe, neural lobe, and areas of basophil cell invasion in the neural lobe, with the numbers of basophil cells present. The selection of appropriate parts was made by the use of an elegant method in which the presence of basophil cells could be ascertained by examining a pinhead-sized fragment. The results showed that intermedin was present in high concentration in the anterior lobe, whereas it was almost entirely absent from neural lobe tissue. Neural lobe tissue showing invasion by basophil cells was, however, equivalent in hormone content to samples from the anterior lobe. The invasion of the neural lobe by basophilic cells, which is a phenomenon peculiar to the human hypophysis, has permitted here the demonstration that cells containing glycoprotein secrete intermedin. For the identification of the specific type of cell responsible for intermedin secretion in the human pituitary, we can take advantage of the fact that although a number of basophil cell types can be seen in the pars distalis, the cells that invade the n(>ural lobe are of a single tyj^e. Thej^ are composed entirely of a variety of Romeis' (1940) /?-cells which stain with aldehydefuchsin and take a red oi' pui|)l(' shade from the trichrome counter stain. These cells will be designated as "purple ^-cells." Morris, Russell, Landgrebe, and Mitchell (1956) referred to these invading cells as 8-cells; they are, however, quite distinctly characterized as /^-cells by Romeis. The S-cells of Romeis do not invade the neural lobe.

The assays of Morris and his associates were concerned with intermedin and corticortrophin, and show that the invading basophil cells were not responsible for corticotrophin production. Although no information is available concerning the other hormones, it is virtually certain that the cells which secrete intermedin in the human will, like the cells secreting intermedin in other vertebrates, be responsible for this secretion only. In man this intermedin secretor is heavily granulated and stains strongly by PAS or by trichrome methods, as does the pars intermedia cell of the cat. It is particularly liable to retain the red stain of the trichrome method, a tendency which is greatly strengthened after fixation in Helly's fluid. Herlant (personal communication) has found that the staining of ^-cells in the human pars distalis by aldehyde-fuchsin is greatly diminished or even fails entirely after fixation in Helly's fluid. This behavior of the intermedin secretors of the liuman pars distalis is also found in the cells of the pars intermedia of the rat, which stain intensely with aldehyde-fuchsin after formol-sublimate fixation but stain very weakly after Helly fixation (Fig. 3.17). It has not been determined whether this behavior is a general one for intermedinsecreting cells.

In the human pars distalis the intermedin-secreting basophils are the most conspicuous of the basophil cells. Their prevalence is consistent with the high intermedin content of the human hypophysis (Landgrebe and Mitchell, 1958). Their failure to respond in sympathy with disturbances of gonadotrophin and thyrotrophin secretion, which has apparently set the human pituitary apart from those of experimental animals, need no longer be a stumbling block now that the nature of these conspicuous basophilic cells is recognized. Obviously the source of thyrotrophin and gonadotrophin must be sought in glycoprotein-containing cells other than the intermedin secretors.

It is possible that some species which have a discrete pars intermedia may also have some dispersed pars intermedia cells in the pars anterior. Landgrebe, Ketterer and Waring (1955) have referred to such an invasion of the pars anterior in some breeds of pigs.

XI. The Pars Tuberalis

The pars tuberalis consists of a layer of adenohypophyseal cells which surrounds the neural stalk and covers the surface of the median eminence of the tuber cinereum. It is composed almost entirely of cells of a single type which, from the absence of specific granules of the kind seen in anterior lobe cells, are classed as chromophobes. There are, in addition, small numbers of typical basophil cells which, according to Romeis (1940), do not differ in any respect from the basophil cells of the anterior lobe. Acidophil cells apparently identical with the cells in the anterior lobe are extremely rare. In the larger animals, the pars tuberalis cells are arranged in cords and follicular structures ; the latter enclose small amounts of colloid. No specific hormone has been demonstrated in the pars tuberalis other than small amounts of activities probably due to contamination with adjacent anterior lobe or pars intermedia tissue. The function of the pars tuberalis is at present unknown. It may be that it is the source of some trophic influence which is carried to the anterior lobe by the hypophyseal portal system.

XII. Cytologic Changes Accompanying Secretory Responses Concerned with Reproductive Function

A. Sexual Maturation in the Rat

Sexual maturation in the male rat is accompanied by a gradually increasing content of glycoprotein in the gonadotrophs. In the female, however, the reaction is entirely different. Maturation is accompanied by a considerable degranulation of gonadotrophs. This degranulation occurred between the ages of 35 and 42 days in the observations of Siperstein, Nichols, Griesbach and Chaikoff (1954), and, in a number of their rats degranulation of the peripheral gonadotrophs which are thought to be follicle stimulating in activity, was found to have preceded the degranulation of the central gonadotrophs. This observation is significant in view of the fact that the peripheral gonadotrophs ordinarily are more strongly granulated than the central ones. The observation suggests at least that a discharge of stored FSH precedes the discharge of LH on the occasion of the first ovulation.

B. Sexual Maturation in Other Animals

Characteristic globular basophil cells first appear in the red eft stage of Triturus viridescens simultaneously with the differentiation of the male and female gonads and are presumed to be gonadotrophic (Copeland, 1943). In the opossum (Didelphys virginiana) some interesting observations have been made by Wheeler (1943). A vesiculation of the basophil cells and an accumulation of hyaline material occurs in adult animals after castration, producing cells similar to the castration cells of the rat. Wheeler observed that at 100 days of age there occurs a vesiculation of basophils similar to that produced in adult animals by castration. The assumption is that during sexual maturation the hypophysis is called on to secrete gonadotrophic hormones in large amounts and that before full differentiation of the gonads, a temporary endocrine situation exists similar to that produced by castration.

C. Seasonal Breeding

In seasonally breeding animals an initiation of the breeding phase has been observed to be accompanied by basophil changes. In the wild cotton-tail rabbit Elder and Finerty (1943) found an increase in gonadotrophic potency in the male hypophysis in the spring, the maximal level being six times that of the level during the winter. This change was accompanied by an increase in the percentage of basophil cells from 4.4 to 13.8. In Necturus, Aplington (1942) related the seasonal activity of the testes to an increased activity in the number of granular basophil cells. A similar increased basophil granule content in Anolis rarolinensis during the spring was observed by Poris (1941).

D. Induced Sexual Maturation in the Female Rat

An important series of observations luis l)ecii made concerning the eft'ect of estrogen injections on immature female rats approaching the age of sexual maturation. In the immature female rat gonadotrophin secretion is inhibited by minute amounts of estrogen and as long as this relationship exists, no high estrogen levels can be naturally produced in the animal. With the approach of maturation this inhibiting action of estrogen on gonadotrophin must diminish. Indeed it is found that a condition is reached as the time of the first ovulation approaches when estrogen, instead of inhibiting gonadotrophin secretion, triggers the sudden release of these substances. The physiologic observations of Hohlweg and Chamorro (1937) indicate that a release of gonadotrophic hormones occurred between the second and fourth day after the injection of estrogen into the immature female rat approaching the expected time of first ovulation. The effect of the release of gonadotrophic hormones may be shown by the l)roduction of ovulation and corpora lutea or by follicular enlargement only, the actual response being different in different strains of rats. The liberation of the gonadotrophins after estrogen administration results in a marked drop in the hypophyseal gonadotrophin content which occurrs apjn'oximately between 3 and 4 days after estrogen administration (Bradbury, 1947). Purves and Griesbach (unpublished) have studied the responses of the gonadotrophs of the immature female rat after single estrogen injections and found that regardless of dose in the range from 1 to 100 jug. of estradiol benzoate, there was no change in the hypophysis in the first 2 days. As previously mentioned, in these immature female rats both FSH and LH cells are very numerous and the content of glycoprotein is high. Four days after estrogen administration, the gonadotrophs of the pars antei'ioi- ai'e almost free of glyco]irotein and can he recogniz(>d only with difficulty as large pale cells. It is important to note that at this adolescent stage there is no differential effect of estrogen on peripheral and central gonadotroi)hs or on the secretion of FSH 01' LH. The effect is rather that of the triggered discharge of the total gonadotroi)hin content of the hypophysis.

At the time of glycoprotein discharge from tlic liypophysis of innnature female rats after estrogen administration, large numbers of mitoses are observed in the chromophobes and granulated acidophils, an effect which is probably due to the action of the injected estrogen on these cells. The number of mitotic figures induced by this single injection of estrogen in the adolescent female rat is much greater than that observed after similar doses in adult animals and indicates that a proliferation of cells of the acidophil class occurs in the female rat hypophysis at the time of maturation. Baker and Everett ( 1947) found by accurate measurements of mitotic index that stimulation of mitoses in acidophil cells after estrogen administration is greater in innnature female rats than in the mature animal.

E. The Active Breeding Phase in the Female

The relation between hypophysis and gonad during the breeding phase is quite different from that before maturation. In the rat, with its short estrous cycle, there is a recurrent discharge of gonadotrophins at each ovulation and this maintains the gondaotrophic hormone content at a low level. In accordance with this the gonadotrophs whether studied by the glycoprotein staining reaction or by the Mallory staining reaction are inconspicuous because of the low content of specific granulation. There is during the diestrum an accumulation of specific granulation (Catchpole, 1949; Purves and Griesbach, 1951a). Before the thyrotrophs and gonadotrophs were distinguished, there were observations showing a cyclic change in basophil cells during the estrous cycle. Those by Wolfe and Cleveland (1933a) and Wolfe (1935) revealed a variation of basophil cells in their rats which is in exact agreement with the variations of the specific gonadotrophs (Table 3.2).

The basophil cells observed by Wolfe, which are described as being large, oval, finely granulated, and containing a negative image of the Golgi body in most of the cells, correspond in all details with the LH cells (central gonadotrophs) observed by Purves and Griesbach (1952).

Cyclic changes in the hypophysis of the dog were described by Wolfe, Cleveland and Campbell (1933) who distinguished in the pars anterior 4 types of cells, 3 of which contained specific granules. Goldberg and Chaikoff ( 1952a » distinguished 6 cell types of which 4 had specific granules. Type I of Wolfe, Cleveland and Campbell corresponds to the a-cell; their type II, which is selectively stained by azocarmine (Hartmann, Fain and Wolfe, 1946) corresponds to the e-cell of Goldberg and Chaikoff. Type III corresponds in the main to the 8-cell, but includes small numbers of cells with a peripheral accumulation of granules which are probably /3-cells. The 8-cells (type III) increased in numbers up to 10 per cent at the proestrum and w^ere then well filled with fine purplish stained granules. At the time of estrus the number of 8-cells was at a maximum (12.6 per cent) but at the same time they showed extensive degranulation. The number of 8-cells recognized during the lutein phase of the estrous cycle and during pseudopregnancy was low (2 to 4 per cent), and during the anestrum it was 5 per cent. The variation in number of 8-cells and of the granules in these cells corresponds to what would be expected from a cell type producing FSH, if it is right to assume that this hormone is not secreted during pseudopregnancy. Wolfe, Cleveland and Campbell also observed changes in e-cells (type II) which, from a level of 5.8 per cent during the anestrum, rose to 7 per cent at estrus and fell during the later lutein phase of the cycle to 2.5 per cent. At this time they also showed considerable degranulation. Present interpretation of these changes would associate the changes in the €-cells with the secretion of lactogenic hormone.

TABLE 3.2 (After J. M. Wolfe, Anat. Rec, 61, 321-330, 1935.)

Basophil Cell Percentages

Sexual State




Immature (17 days)

Immature (27 to 35 days) .

Mature (pre-estrus)

Mature (estrus)


7.4 2.8 0.9 0.6 1.3

0.7 0.9 1.9 4.4 3.5 2.7


8.6 4.8 5 2

Mature (metestrum)

Mature (diestrum)

4.0 3.9

Cyclic changes in cells of acidophil and basophil classes have also been described in the sow (Cleveland and Wolfe, 1933), rabbit (Wolfe, Phelps and Cleveland, 1934) and guinea pig (Chadwick, 1936). These observations are, however, not easily correlated with hormonal functions because the differentiation of specific cell types was not achieved.

F. Pseudopregnancy and Pregnancy

During pseudopregnancy and pregnancy five distinctive cell types are recognizable in the rat hypophysis as follows. (1) Acidophil cells, carminophil by Dawson's (1954a) method, in close relation to blood vessels and connective tissue septa and strongly granulated. (2) Acidophil cells, orangeophil by Dawson's (1954a) method, in the interior of the cell cords and large, active, and only lightly granulated. (3) Thyrotrophs which are more prominent and active in this phase than during the estrous cycle. (4) FSH cells which are strongly granulated and at least as prominent as those in the male hypophysis and in the immature female. (5) LH cells which are large, round, lightly granulated cells with large and prominent negative images of the Golgi body.

In animals treated with thyroxine the thyrotrophs are inhibited and no longer seen. Their activity in the untreated animal is probably an indication of an increased demand for thyroxine during this phase of the reproductive cycle. The strong secretory activity which is indicated in the second type of acidophil cell is considered to be related to the secretion of lactogenic hormone, whereas the activity of the LH cells is considered to be related to a high level of secretion of LH. The prominence of the FSH cells indicates a retention of FSH. The increased prominence of the FSH cells does not necessarily indicate an increased secretion of FSH inasmuch as other conditions which interrupt the estrous cycle permit the accumulation of glycoprotein in these cells. The increased glycoprotein storage correlates with the finding of elevated gonadotrophin levels in the rat hypophysis during pregnancy (Evans and Simpson, 1929; Zeiner, 1952).

The carminoi)hil coll in the rabbit, cat and monkey, shows during the reproductive

cycle marked fluctuations in activity which can be correlated with the secretion of prolactin at times when its luteotrophic or lactogenic action is manifest (Dawson and Friedgood, 1937, 1938b; Dawson, 1939, 1948; Friedgood and Dawson, 1940).


In the human hypophysis the predominant change in pregnancy is the activation of large numbers of cells which are chromophobic in the nonpregnant state. These pregnancy cells were first described by Erdheim and Stumme (1909). They lie in the lateral regions of the pars distalis and from the seventh month contain fine acidophil granules. These cells have been the source of some confusion and Rasmussen (1933) , finding normal proportions of acidophils, basophils, and chromophobes in the early months of pregnancy, denied the existence of a specific pregnancy cell although he recognized the enlargement of the chromophobes during this condition. Romeis (1940) confirmed the observations reported by Erdheim and Stumme and described the pregnancy cells as large cells with large nuclei. In the second half of pregnancy they show an accumulation of granules which Romeis called 7^-granules. Romeis considered the r;-granules to be a specific type of granule not present in the nonpregnant hypophysis. The granulated pregnancy cell closely resembles the e-cell as seen in the nonpregnant human hypophysis although there are some points of difference. The distribution of the granules is similar and the granules, like those in e-cells, are orangeophil. a-Cells, whose granules are carminophil, retain their normal appearance in the pregnancy hypophysis. It is probable that pregnancy cells are e-cells in an altered functional state. Floderus (1949) examined the distribution of pregnancy cells in the lunnan hypophysis and found that they are infi'cHinent in the upper posterior part of the pars distalis and are sometimes entirely lacking in this region. They are most abundant in the lower lateral parts of the gland. The pregnancy cells are often centrally located in the cell cords with other types, usually ordinary acidophils, located periph



erally. Cell cords composed mainly of pregnancy cells occur and isolated cords of this nature may be conspicuous in certain regions. The chromophobes of the vascular zone of the anterior lobe near its attachment to the stalk ("zona tuberalis") are not affected by pregnancy changes. The pregnancy cells are, therefore, not derived from chromophobes in general but from chromophobic cells having a specific distribution.

There are two important differences between human pregnancy cells and the carminophil cells of the rabbit, cat, and monkey. First, the human pregnancy cell granules are orangeophil rather than carminophil. The distinctive color difference between a-cells and pregnancy cells is, therefore, in the human hypophysis the reverse of that seen in the other species. Second, the distribution is different from that seen in the monkey where carminophil cells are especially concentrated in the zona tuberalis. Despite these differences, the pregnancy cells appear to be functionally analogous to the carminophil cells and are presumably the source of prolactin secretion.


The characteristic feature of the pars anterior at and shortly after parturition is the increased amount of acidophil granulation present. Kirkraan (1937) found in the guinea pig that the acidophils increase in numbers towards the end of pregnancy and attain a maximum soon after parturition. Wolfe and Cleveland (1933b) reported a similar increase in granulation of acidophils in the rat hypophysis towards the end of pregnancy. Everett and Baker (1945) found that after parturition in the rat, the acidophil cells increased by almost 100 per cent in the first three days of the lactation period. This increase in the number of acidophil cells visualized was accomplished without an increased number of mitoses or any increase in the size of the gland and was, therefore, due to the regranulation of acidophil cells which had been degranulated during pregnancy. Hunt (1949), however, found that mitoses were present in some but not all rat hypophyses after parturition. Even allowing for this

latter observation it seems that the increased number of acidophil cells at the time of parturition can be accounted for by the accumulation of secretory granules in the cells of the acidophil class. Purves and Griesbach (unpublished) have found that this increase in granulation occurs in the acidophil cells which are remote from the blood vessels and connective tissue septa and which are considered to be related to prolactin secretion. It has already been noted that in the cat a granulation of carminophil cells occurs which is at its maximum at the time of parturition (Dawson, 1946) , these cells also being considered the specific secretors of prolactin. In accordance with this view, Hurst and Turner (1942) reported that in the rat, rabbit, mouse, guinea pig, and cat the prolactin content of the hypophysis was at its highest level during the first few days after parturition.

In connection with the question whether there is a separate lactogenic factor secreted during lactation, different from the luteotrophic and mammogenic factor secreted during pregnancy as Turner (1939) postulated, cytologic observations indicate an activity in a single specific cell type associated with both mammary growth during pregnancy and lactogenesis after parturition. There is, during early lactation, a phase of secretory activity in these specialized acidophil cells which is at its maximum about the 16th day of lactation in the cat (Dawson, 1946). Thereafter the reaction wanes in a manner which suggests that a continuation of lactation is not dependent on the continued secretion of this factor. It is, therefore, probable that the continuation of an established lactation in those animals in which lactation is of considerable duration is not dependent on the continued secretion of prolactin. In conformity with this view, prolactin has been found to stimulate metabolic changes in slices of mammary gland tissue from rats on days 1 to 4 of lactation (Folley, 1952). On the other hand, purified prolactin preparations have not shown any galactopoietic effect in the cow during the declining phase of lactation (Folley and Young, 1938).

XIII. The Human Hypophysis


In the human hypophysis (Fig. 3.18j the adenolobe is closely adherent to the neural lobe. The hypophyseal cleft persists in the adult only in its distal portion. The remainder of the cleft is obliterated by fusion or is broken up into a number of scattered colloid-filled cysts.

No pars intermedia is present, consequently the pars distalis adheres to the neural lobe. This feature, which is present also in the higher apes, is not found in lower mammals. In the latter there is either a pars intermedia which is adherent to the neural lobe and separates it from the pars anterior, or there is a pars distalis which is not adherent to the neural lobe. A full description of the structural features present in the

Fig. 3.18. Diagram of a .sagittal section through the liuman hyi)oi)hysis. The anterior direction is to the left of the diagram. The adenolobe is occupied by pars distalis tissue (P.d.) containing a mixed cell population throughout and there is no pars intermedia. Colloid cysts (C.c.) occurring in the region adjacent to the neural lobe are probably remnants of the hypophyseal cleft. The adenolobe is adherent to the neural lobe. The neural eminence, the neural stalk, and the prolongation of the neural stalk within the neural lobe are here collectively called the pars eminens (P.e.). The pars nervosa (P.n.) shows an invasion by basophil cells (B) which migrate into the pars nervosa from the pars distalis. The extent of this invasion increases with increasing age and shows great variation in different individuals. The pars tuberalis is in(lic,il( d InP.t.

zone of contact between adenolobe and neuI'al lobe in the human hypophysis is given Ijy Romeis (1940), who reviews the earlier literature on the subject.

An invasion of the pars nervosa by basophil cells derived from the pars distalis is a unique feature of the human hypophysis. In lower mammals possessing a pars distalis such invasion is not possible because of lack of contact between the two parts. The cells of the pars intermedia in the more usual form of mammalian hypophysis do not show this invasion of the pars nervosa, although an irregularity of the plane of contact indicates a tendency for mutual interpenetration of the two tissues. As stated earlier, the cells invading the pars nervosa of the human hypophysis are intermedinsecreting cells.


The treatise of Romeis (1940) constitutes a landmark in the study of the human hypophysis, as indeed for the mammalian liyi)ophysis in general. Romeis' findings in the human hypophysis have, however, despite their completeness and their excellent presentation, not had the influence on the development of this subject that might reasonably have been expected. The availability of fresh pituitary tissue from surgical hypophysectomy has changed this situation. U'hen appropriate fixing and staining techniques are used, results equivalent to those described by Romeis can be consistently obtained, and even those workers who, through the use of inadequate techniques, have deprived themselves of the opportunity of observing the cell types described by Romeis, must accept the conclusion that such cell types are present and can be revealed by appropriate techniques. It is therefore advisable that those who would study the human hypophysis should identify the cells they see with those so clearly delineated by Romeis rather than to embark on schemes of classification which Romeis' results show to be inadequate, or to use the terminology of Romeis in applications other than those which he adopted.

The f3-, 8-, and y-celLs of Romeis are basopiiil cells because their granules give a



positive PAS reaction for glycoprotein (Herlant, 1958). Romeis noticed that some of the /?-cells retained azocarraine during the azan staining of the sections previously stained with resorcin-fuchsin, whereas others retained only a blue counterstain from the aniline blue. The researches of Griesbach (unpublished) have convinced me that these staining reactions indicate two distinct types of cell whose granules are stained by resorcin-fuchsin, i.e., two types of ^cells. After Helly fixation the difference between the two types is enhanced. Also after Helly fixation the staining of the granules of one type of ^-cell by aldehydefuchsin is greatly weakened, an effect which allows the red counterstain by azan in this type to be more clearly seen. The y8-cells are therefore seen to comprise purple AF cells and blue AF cells. These we may call in the human pars distalis "purple /3" and blue ^." The 8-cells are blue non-AF after Helly fixation, and the y-cclls are usually also blue non-AF.

The blue /?-cells are more variable in size and shape than the purple /?-cells and are more often found in multinucleated form than the latter. Their distribution in the gland is quite different from that of the purple /?-cells, so that they cannot be variants of a single cell type. Moreover, the staining reactions of the granules, except to resorcin-fuchsin or aldehyde-fuchsin after certain fixatives, are quite distinctive. There are indeed more dissimilarities between purple /3-granules and blue /3-granules (Fig. 3.19) than there are between the latter and the 8-granules.

Reasons have been given in the section on the pars intermedia and intermedin secretion for regarding the purple /?-cells as intermedin-secreting cells corresponding to the pars intermedia cells of the usual mammalian hypophysis. The blue /?-cells, the 8cells, and the y-cells should therefore be homologous with the three basophil cell types in the pars anterior of such mammals as the rat, bat, and dog.


The y-cells of Romeis are distinctive in appearance, containing fine glycoprotein

granules which are stained only feebly by acid dyes (Figs. 3.20 and 3.21). In addition they often contain droplets of glycoprotein, intensely stained by the PAS reaction. These cells have been termed "vesiculate chromophobes" (Pearse, 1952) . They are not likely to be confused with normally granulated jScells or 8-cells. The partition of the basophil cells by differential staining of the specific granules is therefore a partition of the purple ^-cells, blue ^-cells, and 8-cells (Fig. 3.22). Some of the methods used produce a differentiation between the purple ^-cells and the two types of blue cells. Because it was assumed that only two types of cell were involved, the distinction between purple and blue cells has been assumed to be a distinction between /S- and 8-cells. Purves and Griesbach (1957b) observing purple and blue cells in Crossmon (1937) stained sections of human pars distalis wrongly assumed this to be a distinction between 13- and 8-cells. Herlant (1953b; 1954b) differentiated the purple /?-cells from the other types by counterstaining PAS stained sections with orange G. The orange G stained the strongly acidophilic granules of the purple /3-cells and combined with the PAS color to produce a brick-red shade whereas the other basophil granules were magenta. Wilson and Ezrin (1954) used a method substantially the same as Herlant's PAS orange method with the addition of methyl blue. The methyl blue stained only the magenta granules of Herlant leaving the granules of the purple ^-cells brick-red. The purple cells of Purves and Griesbach, the brick-red cells of Herlant, and the PASred cells of Wilson and Ezrin are obviously the same cells and belong to the group of ^-cells. It is only when these staining methods are applied as counterstains to sections in which /?-cells have been electively stained by aldehyde-fuchsin, as in the method of Griesbach, that the dual nature of the /3cells is revealed.

Adams and Swettenham (1958) applied aldehj^de-fuchsin or Alcian blue to sections oxidized with performic acid and followed this with PAS staining. Their R cells which were red and not stained by aldehydefuchsin are the purple ^-cells, their S cells which were stained by aldehyde-fuchsin or




Fig. 3.19 (upper left). Section of the human pars distalis showing weak staining of the granules of purple ;3-cells by aldehyde-fuchsin after fixation in Helly's fluid. Much of the density of the purple /3-cells in the photomicrograph is due to the reddish tint given to them by the azan counterstain. A single blue jS-cell (B) is strongly stained by the aldehyde-fuchsin (AF). In the section it appeared an indigo color from superimposition of a blue color from the aniline blue of the counterstain. AF, azan, X 900.

Fig. 3.20 (upper right). Section of the human pars distalis showing 7-cells. The cells here show elongated forms. The granulation was stained a weak purplish shade. AF, azan, X 900.

Fig. 3.21 (lower left). Section of the human pars distalis showing 7-cells stained by periodic acid-Schiff (PAS). The conspicuous PAS +ve droplets which are unusuallj^ numerous in this field are not peculiar to the 7-cell since they may be found in other cell types, both basophil and acidophil. The faint grey granulation distributed throughout the cytoplasm is considered to be the specific 7-granulation. PAS, X 900.

Fig. 3.22 (lower right). Section of the human pars distalis fixed in Helly's fluid and stained with aldehyde-fuchsin (AF), counterstained with azan. Acidophils, blue /3-cells and purple /3-cells which were deeply stained in different colors in the section, appear black and cannot be clearly distinguished from one another in the photomicrograph. The numerous large rounded pale cells with grey granulation are 5-cells. The granules were stained a clear blue in the original. AF, azan, X 900.

Alei;iii blue comprise the blue /i- and 8-cclls. Herlant (1956b) obtained the same distribution of the aldehyde-fuchsin stain in sections oxidized with permanganate. It is to be noted that the effect of oxidation is to inhibit the staining of purple /?-cells by aldehyde-fuchsin and to render the 8-cells

stainal)lc; the hhic /i-cells are stained with or without oxidation. The effects of jiermanganate oxidation on the aldehyde-fuchsin staining of the human pars distalis are the same as those observed in the pars anterior and pars intermedia of the rat (Halmi and Davies, 1953). The staining of the pars



intermedia cells is weakened and the S-cells of the pars anterior become stainable whereas the ^-cells remain stainable. The difference between the species is that the intermedin-secreting cells in the human hypophysis are scattered throughout the pars distalis and not segregated in the pars intermedia.

Hellweg (1951) obtained by silver impregnation a specific staining of the 8-cells in the human pars distalis. This staining is similar to that observed by Knigge (1957) in the rat, in which a specific staining of Halmi's S-cells (gonadotrophs) was obtained by silver impregnation. Ezrin, Swanson, Humphrey, Dawson and Wilson (1958) obtained specific staining of 8-cells by their iron-PAS method, the 8-cell granules being stained by dialyzed iron which is subsequently converted to Prussian blue. A valuable feature of the iron-PAS method is that it is applicable to postmortem material fixed in formol-saline.


1. The S-Cell

Neither Hellweg (1951) nor Ezrin, Swanson, Humphrey, Dawson and Wilson (1958) found 8-cells in children before the age of puberty, and the latter investigators did not find them in the hypophyses of pregnant women. This finding can be correlated with the reduction in gonadotrophin content during pregnancy (Herlant, 1943). It is therefore certain that the 8-cells are the source of one or more of the hypophyseal gonadotrophins. Ezrin and his associates found that the number of 8-cells decreases with the duration of the terminal illness, being 8.5 per cent in patients dying in 24 hours and 1.8 per cent in those dying after 2 weeks or longer. This is in accordance with other evidence suggesting a reduction in gonadotrojihin during chronic illness.

2. The Blue /3-Cell

There is a considerable increase in the number of hypertrojihied, lightly granulated basophil cells in myxedema and cretinism (Herlant, 1954b, 1958). The staining reaction of these cells to Herlant's (1953b)

PAS-orange method indicated that they are blue basophils, and inasmuch as Russell (1957j found them stainable by aldehydefuchsin it is probable that they are blue /?-cells and not 8-cells as Herlant supposed. The blue /3-cells are therefore probably thyrotrophs. It should be noted that the staining reactions of the blue /?- cells are the same as those of the thyrotrophs of the pars anterior of the rat, dog, bat, cat, and guinea pig.

3. The y-Cell

y-Cells are well developed and fully functional in appearance in children and maintain much the same appearance in women up to the menopause. They do not seem to be altered in appearance during pregnancy (Herlant, 1958). They are quite distinct from pregnancy cells which in the latter half of pregnancy contain acidophil granules. The number of y-cells does not show any correlation with the duration of the terminal illness (Ezrin, Swanson, Humphrey, Dawson and Wilson, 1958). Herlant's observation that the y-cells are inactive in aged subjects does not correlate with any known change in hormone secretion. It is possible that these cells secrete corticotrophin but more investigation is necessary. Crooke and Russell (1935) noticed in Addison's disease a variable proportion of very large "chromophobes." Large chromophobes ("hypertrophic amphophils") have also been observed by Mellgren (1945, 1948) in this disease. The large chromophobes or hypertrophic amphophils of the above authors are presumably the y-cells of Romeis (1940), or more precisely the remains of the y-cells, inasmuch as these cells are susceptible to postmortem autolysis and are not well preserved in autopsy material. Griesbach (private communication) has observed a large number of large y-cells in the pars distalis of a patient who had been adrenalectomized for the treatment of a carcinoma some weeks before death.


Russfield (1957) divides the cells of the humaii pars distalis into acidophils, basophils, amphophils, hypertrophic ampho



phils, and chromophobes. The acidophils of Russfield are the a-cells of Romeis, the basophils are the conspicuous and strongly granulated ^-cells of Romeis. Russfield erroneously considers these "normal" basophils to be the 8-cells of Romeis. The amphophils consist of lightly granulated cells — 8-, y-, and c-cells — but may include «and ^-cells if these are lightly granulated. Hypertrophic amphophils are in the main y-cells whose cytoplasm and granules have been lost by postmortem autolysis, to which these cells are especially sensitive.

From the study of hypophyses of patients with endocrine disturbances (Burt


Fig. 3.23 {upper). Sod ion of tlio pars distalis of a patient who suffered from the Gushing syndrome. Many typical Grooke's cells are present. The hyahnized zones appear homogeneous at this magnification. The remaining granules are stained in llic manner typical of puri)le /i-cell graTuilation. AF, azan, X 900.

Fig. 3.24 (lower). ^wVww of a liuiuaii liypojihysis showing a basophil adenoma composed of cells with the staining leactions of purple ^-cells. In material fixed in formalin the purple ^-cells are strongly stained by aldehydo-fuchsui (.\1). From a specimen sui)iilicd 1)V Professor Dorothv Ru.ssell. AF, X5.

and Velardo, 1954) and of hypophyseal tumors (Russfield, Reiner and Klaus, 1956) , Russfield concludes that amphophils are capable of producing all the anterior lobe hormones, but there is no implication that they produce all these hormones simultaneously. This seems to mean no more than that large amounts of hormone may be secreted by lightly granulated cells. Russfield's results are of importance in directing attention to the fact that more information of endocrinologic significance can be obtained from the study of lightly granulated cells than by the enumeration of typical" acidophils and basophils; they do not conflict with the view that the cells producing different hormones are characterized by different types of granules, whose specific character can be distinguished by appropriate staining methods.


The purple /?- or intermedin-secreting cells are only lightly granulated in infants. The amount of granulation in these cells as measured by the intensity of the PAS reaction increases with age in a smooth continuous fashion which is neither accelerated nor retarded by puberty or the menopause (Herlant and Lison, 1951). The cells are not affected by pregnancy and the progressive increase in granule content follows the same course in both sexes.

The functional state of the purple /3-cells a])pears to be determined by the level of circulating corticosteroids and has not been shown to be related to any other factor. Although the physiologic significance of the responses of the purple j8-cells to variations in the corticosteroid level is concealed in the mystery which envelops the whole subject of the function of intermedin secretion in mammals, the responses themselves are definite, striking, and consistent. High levels of corticosteroids stimulate, low levels depress the cytologic activity of these cells.

The characteristic changes in the Gushing syndrome are a degranulation and hyalinization of the purpl(> /3-cells. The changes were first dcscnhcd by Crooke (1935), after whom the liyaHnized cells were named (Fig. 3.23). In my own observations I have found that it is only the jiurple ^-cells

which are affected by hyalinization. Moreover, the basophil cells which invade the neural lobe and are all purple y8-cells are sometimes hyalinized in the same way as the cells in the pars distalis. Hyalinization does not alter the staining reactions of the granules that remain in the hyalinized cells. It is notable, however, that the purple /3-cells which have invaded the neural lobe are normally less active and smaller in size than those that remained in the pars distalis. This difference in activity persists in the Gushing syndrome and, in consequence, the purple /3-cells in the neural lobe show a lesser degree of hyalinization than those in the pars distalis ; they may indeed not show any hyalinization in hypophyses in which many of the cells in the pars distalis are affected.

Degranulation of Crooke's cells may be complete and the cytoplasm may be coml)letely hyalinized. More characteristic is a partial degranulation having an annular distribution. The periphery of the cell may be granulated and the granules be preserved around the nucleus and Golgi body, or degranulation may occur around the nucleus and Golgi body with preservation of granules at the periphery. Sometimes a degranulated zone occurs with preservation of granules both at the periphery and in proximity to the nucleus and Golgi body. Hyalinization occurs in the degranulated zone and differs from the mature hyalinization of thyrotrophs and most gonadotrophs in the rat, in that the hyaline area is not sharply limited. The surface of the hyalinized zone shows a fine vesicular structure which gives to the hyalinized zone a faintly ground-glass appearance. Probably the hyalinization is similar to that of the filigree cells described by Farquhar and Rinehart (1954) in the rat.

The hyaline substance is not colored by the PAS reagent and has little affinity for basic dyes.

Crooke's cells are in general large cells with vesicular nonpyknotic nuclei and easily visible Golgi bodies. Both McLetchie (1942a, b) and Mellgren (1945) agree with Crooke that these appearances indicate increased secretory activity.

In the original description of the Gushing

syndrome the presence of a basophil adenoma in the hypophysis was noticed in three cases. Gushing emphasized the role of these adenomas as a primary factor in the causation of the syndrome, but considered it possible that the symptoms were due to stimulation of adrenocortical secretion by the basophil adenoma (Hubble, 1949) . However, such basophil adenomas are not found in all cases of the syndrome and are, moreover, not infrequently found in hypophyses of persons in the older age groups who have not exhibited any specific endocrine disturbances during life.

The cells of basophil adenomas found in association with Grooko's cells contain basophil granules with the same staining reactions as those of purple ^-cells (Fig. 3.24). The cells of the adenomas are usually not hyalinized. In this respect they are similar to basophil adenomas induced in the rat hypophysis by thyroxine deficiency or by castration. The association of basophil adenomas with hyalinization indicates a long continued strong stimulation of the affected cells (Purves, 1956).

G. crooke's cell changes produced by


Laqueur (1950, 1951) and Thornton (1956) have reported Grooke's cell changes in patients treated with cortisone. Golden, Bondy and Sheldon (1950), Dreyfus and Zara ( 1951 ), and Thornton (1956) have also reported similar changes after treatment with corticotrophin. The corticotrophin presumably acts by stimulating the adrenal cortex, the secretion from which is the effective agent in producing the hyalinization. From Thornton's observations it appears that hyalinization is produced in a few days by effective doses of corticosteroids and regresses equally rapidly after cessation of treatment. Grooke's cells therefore indicate a high level of corticosteroids in the circulation in the last few days before death. The erroneous assumption that the basophil cells which were hyalinized in the Gushing syndrome were pars anterior cells caused some investigators to postulate that the hyalinization was the expression of an increased secretion of thyrotrojihin or gonad



otrophin resulting from alterations of thyroid or gonadal function produced by high levels of costicosteroids. The fact that these cells are not hyalinized in myxedema or after castration shows that this is not so. In the present state of knowledge it seems probable that Crooke's cells are actively secreting intermedin, the effects of which are antagonized by corticosteroids. Only occasionally is the Gushing syndrome associated with hyperpigmentation (Edmunds, McKeown and Coleman, 1958). It can be affirmed that Crooke's cell changes have nothing to do with an increased secretion of corticotrophin because they are produced by conditions which cause suppression of corticotrophin secretion.

h. changes in the purple /3-cells in Addison's disease

Reports on the human pars distalis in Addison's disease indicate a diminution in the number of basophil cells or at least a scarcity of well granulated basophils (Kraus, 1923, 1926, 1927; Berblinger, 1932; Crooke and Russell, 1935). It is the purple /?-cells which disappear, apparently passing into an inactive state in which they lose their granules. Blue basophils remain apparently in a normal state and it is these cells which have been referred to as CrookeRussell cells. This interpretation is in conformity with the staining reactions of Crooke-Russell cells reported by Russell (1956) and by Wilson and Ezrin (1954). Presumably both blue /3- and 8-cells are present ; the material at my disposal has not been suitable for differential staining of these cell types.

In view of the inactivity of the purple /?-cells in Addison's disease, the hyperpigmentation in this condition must be ascribed to the intrinsic melanocyte stimulating activity of the corticotrophin which is being secreted in excessive amounts. Perhaps it is the effects of this side reaction of corticotrophin which cause suppression of int(n'mcdin secretion in this condition.


The pharyngeal hypophysis is a collection of cells found in the submucosa of the posterior pharyngeal wall and is a remnant of

the epithelial stalk of the hypophysis which connects the buccal ectoderm to Rathke's pouch at an early stage of embryonic development. The pharyngeal hypophysis is found only in man and seems to be constantly present (Romeis, 1940) . It is a mass of cells 3 to 4 mm. in length. The cells are mainly small with scanty cytoplasm, but larger chromophobes and occasional acidophils are present. Basophil cells are extremely rare.

It has been suggested that the pharyngeal hypophysis can take over some of the function of the sellar hypophysis when the latter is destroyed by disease processes (Tonnis, Miiller, Ostwald and Brilmayer, 1954; Miiller, 1958). This cannot be regarded as established for the residual function may be traceable to pars distalis cells that have escaped destruction.

The pharyngeal hypophysis is sometimes the site of adenoma formation (Miiller, 1958). Erdheim (1926) described a case of acromegaly in which was found an unaltered sellar hypophysis and an acidophil adenoma derived from the pharyngeal hypophysis.

XIV. Electron Microscopy of the Adenohypophysis

Many cytologic structures are so small that the details of their structure cannot be made out by means of the light microscope, the resolving power of which is limited to about 300 nifi. The limit of resolution of the electron microscope is about 1/1000 that of the light microscope. Because of the low electron density of organic materials and the consequent lack of contrast in thin sections of tissues, the available resolution of the electron microscope for cytologic detail is limited to about 1/50 of that of the light microscope, i.e., about 6 ni/x. This is a big advance, comparable to that produced by the change from the simple hand lens to the compound microscope.

The most fertile application of electron microscopy in the field of cytology has been the examination of ultrathin sections of tissues fixed in solutions containing osmic acid. Osmic acid has been for many years considered the best fixative for the preservation of fine structure in material exam



ined by light microscopy. Its use is unfortunately incompatible with most of the staining techniques on which light microscopists have come to rely. For the fixation of tissues for light microscopy, use has been made of a number of fixatives, which permit subsequent staining by various methods, but which do not satisfactorily preserve fine structure even at the level visible by light microscopy. These fixatives cause an extensive redistribution of the proteins of cells during fixation as is shown by an increase in opacity and light scattering power during fixation.

No exact correspondence can be expected between the appearances of structures seen in electron micrographs of osmic acid fixed tissues and of structures rendered visible by staining in tissues fixed by other methods.

Cytoplasmic structures visible in electron micrographs of adenohypophyseal cells (Figs. 3.25-3.32)1 include the endoplasmic reticulum, the components of the Golgi region, the Palade granules, mitochondria, secretion granules, and lipid droplets. Thus far electron microscopic observations of the pituitary have revealed nothing new with respect to the finer structures of these parts (Palade, 1953), and nothing that is suggestive of specific hypophyseal function. On the other hand, only a first step has been taken, but it is expected that further studies will lead to clarification of the many problems to which allusion has been made.


The endoplasmic reticulum is a cavitary system consisting of tubes, vesicles, and flattened sacs interconnected by narrower

^The electron micrographs of the cells found in the pars anterior and pars intermedia of the rat were contributed by Dr. Marilyn G. Farquhar who also supplied the descriptions. The electron micrographs, which were specially prepared for this publication are, as a result of recent technical advances, of higher quality than those appearing in the original publications. All were prepared from tissues fixed in osmium tetroxide buffered with acetate-veronal buffer to pH 7.4 and embedded in n-butyl methacrylate. Sections of 20 to 50 m/x were prepared with a Porter-Blum microtome (Servall) and examined and photographed in an RCA EMU-2 electron microscope. Further technical methods are detailed elsewhere (Farquhar, 1956).

channels to form a complicated network (Palade, 1956). The system is enclosed by a continuous membrane which is probably similar to and derived from the cell membrane (Howatson and Ham, 1957). The endoplasmic reticulum varies considerably in extent and form in different cells. When it is extensive the cross-sections of its membrane-enclosed cavities may occupy the greater part of the cytoplasm. The membranes are not revealed by any staining methods used in light microscopy and the cavity usually contains no stainable content. In consequence light microscopy gives little indication of the existence of the endoplasmic reticulum. The larger vesicles which form part of the endoplasmic reticulum of the basophil cells of the rat pars anterior are, however, visible in paraffin sections examined by light microscopy and confer on the cytoplasm a foamy appearance which was noted by Reese, Koneff and Wainman in 1943. Under conditions of rapid secretion, hyaline substance accumulates in these vesicles which become much more easily visible by reason of their enlargement and the presence of a stainable content. That hyalinization is an accumulation of a stainable material in cytoplasmic vesicles which are always present but normally empty was first stated by Reese, Koneff and Wainman and confirmed by the electron microscope studies of Farquhar and Rinehart (1954a, b). Electron microscopy did not therefore provide the first evidence for the existence of the endoplasmic reticulum, but it showed for the first time its full extent, its continuity, and its existence in all types of cells.


The Golgi region or zone has long been a focal point of discussion by investigators of pituitary morphology and function (Severinghaus, 1932, 1933, 1939). In suitable preparations studied by< light microscopy it is often seen as a conspicuous feature of granulated cells in the pars anterior, this specialized region of the cytoplasm being made evident by the absence from it of the granulation which is distributed throughout the remainder of the cytoplasm. In sections which are sufficiently thin, an unstained zone is seen producing an appearance which



' A




Fi(i.3.25. l^lcriron microjir.'il'ii sliowin^ a jionioii ot an aci(loi>iul irom rhc a(lonohy]loplly^ of a rat. Part of the nucleus (AO is present above and a segment of the cell membrane (cw) crosses the field below. A few dense ovoid secretory granules {gr) are scattereti throughout the field. Several mitochorndia (w) are also seen. They show double-layered limiting membranes and internal crests {cr) or "cristae mitochondriales" of Palade (1952) which are characteristic features of all mitochondria.

The endoplasmic reticulum (er) is seen in the lower portion of the cell. This cytoplasmic component occurs here in the form of parallel rows of long, membrane-limited sacs. When reconstructed in three dimensions the sacs are continuous with one another at the ends, and the membranes thus enclose a broad flat cavity. These appearances represent just one organizational variant of this highly complicated system (Palade and Porter, 1954).

A number of tiny dense particles {hji) (ca. lOOA) are distributed throughout the cytoplasm, but are porticularly concentrated along the membrane surfaces of the endoplasmic reticulum. These particles are generally considered to be the site of localization of cytoplasmic ribonucleoprotein (Palade, 1955).

Components of the Golgi apparatus can also ])<> identified in tlic cytoplasm. In electron micrographs the Golgi "complex" may be resohed into 3 coniiionents: relatively emptyajipearing vacuoles (t'oc) of varying sizes (Dalton and Felix. 1954; Sjiistrand and Hanzon, 1954; Farquhar, 1956); paired membranes (Gm) (ca. 7 m^ each) (Palade, 1952) ; small granules or vesicles (Gu) (ca. 40 m^) (Palade and Porter, 1954). X 31,500.











Fig. 3.26. Electron micrograph showing cells from the anterior pituitary of a young adult male rat. Three acidophils of the type which are thought to be responsible for the production of growth hormone (Hedinger and Farquhar, 1957; Farciuhar and Rinehart, 1954a) occupy most of the field. Their nuclei (AO are indicated.

This type of acidophil is characteristically rounded or ovoid in shape, and the cells typically are arranged in groups, as shown here. In electron micrographs the most distinctive feature is their content of variable numbers of dense, ovoid secretory granules (grr) of a characteristic size (ca. 350 ni/u maximal diameter). Large numbers of secretory granules are present in this field. The cell membranes {cm) can be clearly seen separating the cytoplasm of one cell from that of another. Mitochondria (m), endoplasmic reticulum (er), and Golgi material (G) may also be distinguished. X 10,300.

is termed "the negative image of the Golgi body." The negative image of the Golgi body or zone is especially conspicuous in the basophil cells of the rat pars anterior. In acidophil cells the Golgi body is usually smaller and the negative image is not seen unless the sections are thin, i.e., 2 to 3|U.. When acidophil cells are stimulated their Golgi zones become enlarged and the nega

tive image may be seen more easily. This happens in the rat hypophysis after estrogen treatment or at times when rapid secretion of the lactogenic hormone is occurring physiologically. The Golgi region has the appearance of a spheroidal shell enclosing an area of cytoplasm. In the basophil cells of the rat pituitary, the cytoplasm enclosed by the Golgi region is more deeply stained than






1* - 7-^l»,*!»^» « .







Fig. 3.27. Electron micrograph (if , I .-((imn iinn, il,, nm i lor ])ituitary of :i imini.-il adult female rat showing an acidojihil iti ihe i\ pu uJneh j> ihuught to be re.^pun.^iMc for the production of mammotrophic hormone (Hedinger and Farquhar, 1957; Farquhar and Rinehart, 1954a). The nucleus (A^^) and the cell membrane (cm) of the mammotroph are shown.

These cells are typically found alone, rather than in groups, in the normal, nonlactating animal. Their most distinctive feature in electron micrographs is their cytoplasmic content of very large, dense secretory granules (gr) with a maximal diameter of 600 to 900 m/i. In this cell they are predominantly found grouped to the left of the nucleus (N). The granules appear very dense and do not show evidence of internal structure. Their appearance is in contrast to that of the mitochondria (m) which are usually more elongated, much less dense, and show clear internal structure which is difficult to see in detail at this relatively low magnification.

Tubular and cisternal (elongated) profiles of the endoplasmic reticulum (er) as well as vacuoles of the Golgi complex (G) may also be identified in the cytoplasm. The two areas marked A represent segments of the cytoplasm of two adjacent acidophils of the type which are presumed to be responsible for the production of growth hormone. The smaller size of the secretory granules distinguishes these cells from the mammotrophic acidophil.

A portion of the nucleus of a thyrotroph (T) is seen to the right. Cytoplasmic processes of this cell extend out from the nucleus to encircle partially the mammotroph. The cell may be identified as a thyrotroph on the basis of its angular shape and content of very small secretory granules.

In the lactating animal acidophils of this type with large secretory granules are very numerous and can be seen in virtually every field (Hedinger and Farquhar, 1957). X 11,700.




Fig. 3.28. Electron micrograph showing a th.yrotroph from the anterior pituitary of a young adult male rat. The nucleus (A') and cell membranes (cm) are indicated. The irregular contour characteristic of thyrotrophs is illustrated in the angular shape of this cell.

In electron micrographs thyrotrophs can be distinguished by virtue of the size of their secretory granules which are smaller (maximal diameter ca. 100 iu/m) than those in any other type of anterior lobe cell (Farquhar and Rinehart, 1954b). In this cell the secretory granules are found, for the most part, lined up along the cell membrane.

As seen in this cell, the endoplasmic reticulum (er) of thyrotrophs is generally present in the form of small vesicular profiles with occasional elongated (cisternal) profiles. The mitochondria (?n) usually occur in the form of short rods and show a background matrix which is much less dense than the internal matrix of gonadotroph mitochondria (see Figs. 3.29 and 3.30). A group of vacuoles of the Golgi complex (G) can also be distinguished in the cytoplasm of the thyrotroph.

The thyrotroph is virtually surrounded by acidophils of the growth hormone type. The cytoplasm of these cells is labeled A. The size of their secretory granules (maximal diameter ca. 350 m/i) can be contrasted with the smaller thyrotrophic granules. X 10.000.

the rest of the cytoplasm. The PAS reaction shows that this is due to a greater concentration of glycoprotein granules in the enclosed cytoplasm. The appearance of the darkly stained cytoplasm within the negative image of the Goki body has not always

been interpreted correctly and some investigators have mistaken it for an early stage of hyalinization.

Severinghaus (1932, 1933) reported that in the rat the Golgi apparatus of the cells of the acidophil class has a different form









" ■ ,*/





    1. i


'♦• * '

• ' 1

¥■ **



. '•"•T



  • •*'


    1. .f..

v*v /;•*••.-,

.'/.€. "/:- *•


- -^.'K- ..

<s»" ;

^ , .,-...- '







l^li -."*#'i§fc'^^5#ff.

Fig. 3.29. Electron microgmph allowing a \li>- Lugr guu:uloii(ii>li iium iln ,1111(1101 pituitary of a young adult male rat. The nucleus (iV), a nucleolus {nc), and the cell membrane {cm) are shown.

This cell can be identified as a gonadotroph by virtue of its rounded contours and content of secretory granules {gr) with a maximal diameter of approximately 150 m^. The secretory granules of gonadotrophs are intermediate in size between the large secretory granules of acidophils and the small secretorj^ granules of thyrotrophs.

A spherical chain of small vacuoles {vac) circumscribes the Golgi apparatus which is located above the nucleus. Elements of the Golgi complex outline a cytoplasmic area nearlv as large as the nucleus.

Mitochondria (m) which have been sectioned in various planes are also visil)l(> in the cytoplasm. The mitochondria of gonadotrophs are generally more elongated and show a denser internal background matrix than other types of adenohypophyseal cells.

The endoplasmic reticulum {er) is seen here in the form of numerous vesicles which vaiy greatly in size. Some are relatively small and are of a size approaching that of the secretory granules. Others are rather large, for they measure several microns across at their greatest width. It can be seen that the intermmi of the vesicles appears homogeneously grey, and has a background density greater than that of the siuTOunding cytoplasmic matrix.

Gonadotrophs with this appearance have been associated with the secretion of folliclestimulating hormone (Farquhar and Rinehart, 1954a: Farduhar and Rinehart, 1955). They differ from the luteinizing hormone-gonadotroph (see Fig. 3.30) in possessing somewhat paler nuclei, more evenly distributed granules, and prominent vesicles of the endoplasmic reticulum with the homogeneous grey internum. X 6500.








  • ^y^

^ \A


•'.cm V*

^t ^ •' * -u


Fk;. 3,30. Election micioamphs illustiatiug another gonadotroph I'lom the anterior pituitary of a young adult male rat. Like the cell in Figure 3.29, this cell can be identified as a gonadotroph on the basis of its rounded contours, the size of its secretory granules (maximal diameter ca. 150 m^), and elongated mitochondria {m) with a dense internal matrix. The area occupied by the Golgi complex is seen directly to the right of the nucleus. It is outlined by a number of relatively empty-appearing vacuoles {vac).

This gonadotroph differs from the follicle-stimulating hormone gonadotroph illustrated in Figure 3.29 in several respects: the nucleus (iV) is more dense and shows a deep infolding, the secretory granules {gr) are aggregated into clumps, and no large vesicles of the endoplasmic reticulum are present. In addition, there are a number of relatively open areas visible in the cytoplasm (arrows) which are occupied only by a sparse, flocculent precipitate. The endoplasmic reticulum {ex) is seen here in the form of tiny tubular profiles. Gonadotrophs with these features have been associated with the secretion of luteinizing hormone or interstitial cell-stimulating hormone (Farquhar and Rinehart, 1954a; Farquhar and Rinehart, 1955).

A portion of an acidophil (^-1) with larger secretory granules is present above the gonadotroph. X 11,700.






t \ '

^ cm



bp cm ••

--mv % : *

5 ■

Fig. 3.31. Electron iiiici(j<ii,iiili illu-tiaiiii<> im.hious of sc\cial ( ( IN wlml; uii. 1,,, mixsuggested to be concerned with the formation of adrenocorticotrophic liormone (1^'arquhar, 1957). Two large nuclei (iV) and a segment of a third nucleus are shown.

Such cells are typically found in groups and are arranged around large follicles or smaller ductiles. Some of the follicles are quite large, measuring several microns across and are undoubtedly analogous to so-called "colloid cysts" sometimes seen in the anterior lobe by light microscopy. Other follicles, such as the one illustrated here, are quite small and would probably escape detection by light microscopy.

The cells which line the follicles or ductiles typically show tiny cytoplasmic projections or microvilli {mv) which pro.ject into the follicular lumina. In this field portions of three cells abut on the follicle and form microvilli which pro,)ect into the lumen.

The follicular cells characteristically do not contain secretory granules. Furthermore, in the normal animal their cytoplasm appears relatively empty, for organized cytoplasmic structures are sparse. Only a few mitochondria {m), occasionally tubular profiles of the endoplasmic reticulum (er), and basophilic particles {hp) are encountered.

In terms of their somewhat monotonous regularity, these cells resemble more closely the cells from the intermediate lobe than they do any other type of anterior lobe cell.

Because of the response of these cells to alterations in adrenal activity and their lack of response to other experimental procedures, it has been tentatively suggested (Farquhar, 1957) that these cells are responsible for the formation and/or transport of corticotronhin. X 10,800.









'"" *" 54r .- ^' ^"








Fig. 3.32. Election micrograph showmg a field of cells from the pars intermedia of the rat pituitary. The plane of section passes through the nuclei (A^) of four of the cells, whereas only the cytoplasm of several other cells is transected. The cell membranes (cm) stand out prominently.

These cells are seen to fit together like a mosaic, and they all show a similar appearance. Their cytoplasm contains relatively few secretory granules (gr) and mitochondria (m). The endoplasmic reticulum (er) is present in the form of tiny vesicular profiles with occasional tubules. Elements of the Golgi complex (G) are visible in several areas. In these cells the dense, paired Golgi membranes and granules predominate over Golgi vacuoles, which are not seen at all in this field. X 9100.

from that of the cells of the basophil class. In the acidophil cells the Golgi apparatus forms a net-like cap extending over one pole of the nucleus, whereas in the basophil cells it has the appearance of a spheroidal body lying in the cytoplasm at some distance from the nucleus. These characteristic differences were confirmed by Foster (1947) using the Sudan black staining method.

Among the chromophobes some cells have Golgi bodies similar to those of the granulated acidophils, whereas in others they are similar to those of the granulated basophils. This suggests that the chromophobic cells are not undifferentiated, rather that they consist in part at least of temporarily nonfunctioning but specifically differentiated cells.


The characteristic appearance of the Golgi body of the "normal" acidophil cell of the rat hypophysis is not retained through all the phases of activation which may be experienced by this cell. Wolfe and Brown (1942j found that in the hypophysis of the rat treated with estrogen the Golgi body of the acidophil cells enlarged and became more or less spheroidal in shape so that acidophil cells could no longer be distinguished from basophil cells on the basis of the Golgi body appearance alone. The change of the acidophil type of Golgi body from the paranuclear net to a rounded body is not confined to conditions of strong activation but is also seen after castration when the acidophil cells are undergoing a reduction in number and a shrinkage in cell and nuclei size.

In mammals other than the rat distinct differences in the form of the Golgi body in different cell types have not been described, although there are distinct differences in the relative size of the Golgi bodies in different cell types. In the lizard, Anolis carolinensis, (Poris, 1941) the Golgi bodies in the acidophil and basophil cell classes are different, but in this species, unlike the rat, the type of Golgi in the acidophil cells is the more compact one.

Electron microscopic studies of thin sections reveal that the Golgi zone is a system of smooth membranes enclosing spaces which appear to communicate with the endoplasmic reticulum. In Palade's (1955) account of the fine structure of nerve cells, the components of the Golgi region were termed "agranular reticulum." There is an association of three distinct structures: (1) A system of double membranes formed by the apposition of flattened vesicular structures. (2) A system of microvesicles with a dense content. (3) Large empty membrane-enclosed spaces ("vacuoles") resulting from the section of a hollow structure with a watery content.

The relationship of the three comi)onents to one another and to the classical Golgi apparatus or network is not entirely clear but it is likely that the "vacuoles" aw crosssections of an anastomosing canalicular system (Dalton and Felix, 1953; Lacy, 1954) which is responsible for the apiiearance of

the negative image of the Golgi zone. The demonstration of a network by osmium impregnation (Nassonov, 1923) is in the main the result of precipitation of osmium outside the canalicular space particularly at the site of the system of double membranes (Dalton and Felix, 1956; Gatenby and Lufty, 1956). Electron micrographic studies have clarified to some degree the relationship between the Golgi zone and secretory activity. The Golgi zone is not a region in which synthesis of proteins or peptides occurs. The Golgi complex appears to be concerned rather with the segregation or removal of water from maturing secretory products (Dalton and Felix, 1956) and the enclosure of the mature product in membranes to form secretory granules (Farquhar and Wellings, 1957) . The establishment of other relationships awaits other efforts. Particularly promising are the techniques developed by Peterson (1957) for identifying in stained sections the same fields seen in electron micrographs of the adjacent section.


Palade granules are dense bodies 10 to 15 m/x in diameter. They occur to some extent either singly or in clusters in the cytoplasm but are for the most part attached to the membrane of the endoplasmic reticulum forming in those regions where they occur a "rough-surfaced" membrane (Palade, 1955, 1956). Although much too small to be visible as granules under the light microscope, their existence, although not their form, can be made evident by specific staining. The granules contain ribonucleic acid and are therefore responsible for cytoplasmic basophilia. They figure in the literature of light microscopy as a substance, cytoplasmic ribonucleic acid or Nissl substance, and not as a structure. In certain secreting cells in which rough-surfaced membranes occur as closely i)acked, flattened sacs they are responsible for the appearance termed "ergastojihism" by Garnier (Pahide. 1955) .


Secretory granules arc splici-oidal niembi-ane-encloscd Ixxlics with a dense homolicncDUs coiitciit. The granules in dillVrent



cell types differ in size and in the density of their content. The specific differences in the chemical nature of the contents of secretory granules in different cells cannot yet be made apparent by electron microscopy. The electron microscope emphasizes the essential similarity of the specific granules of acidophil and basophil cells and indeed of secretory granules generally whether in endocrine or exocrine cells.

Lipid droplets are seen in electron micrographs as dense bodies, their density being no doubt the result of precipitation of osmium by the reducing action of the lipid. They are irregular in outline, probably as a result of solution of the lipid and distortion during dehydration and embedding. The lipid droplets do not have an enclosing membrane.

Farquhar and Rinehart have identified in electron micrographs the five granulated cell types which can be distinguished by light microscopy in the pars anterior of the rat. All five contain the same cytoplasmic structures; there is no distinctive structural element peculiar to any of the types. There are differences in the shape of the cells, in the extent and form of the endoplasmic reticulum, and in the size, density, and distribution of the secretory granules which enable the different types to be distinguished from one another.

The secretory granules of the basophil cells are less than 200 mfx in size. Although such granules can be seen individually under good optical conditions with the light microscope, the coarsely granular or flocculent appearance of the cytoplasm of basophil cells in stained sections is produced by the uneven distribution of the fine granules which are not individually resolved. The "basophil granules" of the basophil cells of the rat pars anterior as seen under the light microscope therefore bear the same relationship to the secretory granules as the Nissl granules of nerve cells bear to the Palade granules ; they are in fact clusters of granules made visible by the staining of their specific content.


In addition to the granulated cells the pars anterior of the rat contains cells with

a distinctive structural feature, namely, microvilli projecting from the free surface of the cell into a space which is enclosed by contiguous cells of the same type. These cells are not accessible to study by light microscopy and their appearance is so different from that of known endocrine cells that their function must be admitted to be problematical.

XV. The Neurohypophysis and Neurohypophyseal Secretion


Neurohypophyseal tissue contains connective tissue, nonmedullated axons, and interspersed neuroglial cells. In the neural eminence and neural stalk the axons run more or less parallel to one another and, although some of the axons terminate in these regions, most of them pass into the neural lobe where they end in the pars nervosa. The terminal enlargement which forms the neural lobe results in part from the branching of the axons in the terminal part of their course, in part from the increased number of glial cells in this region of the neurohypophysis.

In fish and amphibia the axons in the neurohypophysis arise from cells in the preoptic nucleus in the hypothalamus and form the preopticohypophyseal tract in the intrahypothalamic part of their course. In reptiles, birds, and mammals the homologue of the preoptic nucleus is separated into two parts, the supraoptic nucleus and the paraventricular nucleus. The axons of the neurons in the paraventricular nucleus pass towards the supraoptic nucleus and after running through or closely alongside this structure join the supraopticohypophyseal tract (Laqueur in discussion, Scharrer and Scharrer, 1954) . Lesions which are produced experimentally to destroy the supraoptic nuclei or interrupt the supraopticohypophyseal tract also interrupt the paraventriculohypophyseal fibers and produce a total neurohypophyseal deficiency. It therefore appeared at one time that neurohypophyseal activity in mammals was entirely dependent on the supraoptic nuclei alone, but it is now clear that the paraven



tricular nuclei have the same function as the supraoptic nuclei, being spatially separated parts of the same functional unit.

The organ responsible for neurohypophyseal secretion is thus more extensive than the neurohypophysis. It includes in addition the supraoptic and paraventricular nuclei and their tracts. Functionally we may distinguish four separate regions in this hypothalamo-neurohypophyseal system. The first region contains the supraoptic and paraventricular nuclei in which the neurohypophyseal secretion is elaborated. The second contains the paraventriculo- and supraoptico-hypophyseal tracts which serve two functions, the conduction of nerve impulses and the physical transport of neurohypophyseal secretion into the neurohypophysis. The third region is the tissue of the neural eminence, neural stalk, and in some species a portion of the neural lobe, which are in this account referred to collectively as the pars eminens. In this tissue some of the axons terminate, usually in relation to blood vessels (Scharrer and Scharrer, 1954; Scharrer, 1954) and mediate a secretory activity which could by virtue of the vascular link between this region and the pars anterior or pars distalis, modify the function of the adenohypophysis. In some amphibia a very considerable proportion of the preopticoneurohypophyseal fibers end in this region (Dawson, 1957). In this region there are also nerve fibers from the general area of the medial forebrain bundle and fibers from the lateral tuberal area (Green, 1951, 1956; Bargmann, 1954). The fourth division of the hypothalamo-neurohy])ophyseal system is the pars nervosa which in mammals contains most of the nerve fiber terminations of the neuroiiypophyseal tract and most of the stored secretion. It is also the major site of the release of the characteristic hormones which pass from it by way of systemic veins directly into the systemic circulation without contact with the pars anterior or pars distalis.

The early investigators assumed not unnaturally that the pars nervosa was the site of elal)oration of the hormones which were secreted from it. It is obvious that there must be something unique about the nerve

fibers or the glial cells, or else some unique structural element must be present to account for the secretory function. Herring (1908) considered that the cells present were ordinary glial cells and did not note any peculiarity of the axons other than that they were of larger caliber than nonmedullated fibers elsewhere. He, however, did discover some peculiar masses of protein without any definite structure. These bodies were only found in well fixed material and although they did not look to be likely sites for the formation of hormones attention was focused on them by the fact that they were peculiar to this gland.

Bucy (1930) found that these Herring bodies were in fact large end bulbs terminating some of the fibers of the neurohypophyseal tract. In these end bulbs the fibers were wound about like a ball of twine.

Bodian (1951) confirmed this explanation of Herring bodies and considered that they are terminal malformations, negligible in number compared with normal endings.

Bucy (1930) and Weaver and Bucy (1940) claimed that the neurohypophysis contained cells which were peculiar in their morphology and staining reactions and were found nowhere else in the nervous system. For these cells the term "pituicyte" was proposed. For a time the pituicytes contended with the Herring bodies and the nerve fibers for the responsibility for the production of the neurohypophyseal hormones. As objects to be studied they were certainly more promising in appearance than the nerve fibers, but it cannot be said that the studies of pituicytes have contributed to the elucidation of either neurohypophyseal or pituicyte function. The status of the pituicyte as a cell peculiar to the neurohypophysis is at the present time doubtful. Some investigators (Romeis, 1940; Slopcr. 19571 have stressed the great diversity of c(^ll types which can be found in the neurohypophysis, but this does not disi)ose of the i)ituicyte of Bucy. Leveque and Scharrer (1953) considered the existence of a cell specific to the neurohypophysis un|)i'o\-cn.

ilild ( I95(). in discussion I considcri'd tliat there are no consistent morphologic difTercnces between the glia of the neurohy


pophysis and astrocytic glia in any other parts of the central nervous system. Certainly the pituicytes do not show any of the features of secreting cells. Recent developments have made it clear that pituicytes are in no way responsible for the production or storage of neurohypophyseal secretion and any role that may be ascribed to them in facilitating the release of the hormones into the circulation is purely speculative. It seems unnecessary to assume that their function is different from that of similar glial cells elsewhere.

The hormone activity of the neurohypophysis is accounted for by two octapeptides, oxytocin and vasopressin. Both hormones contain cystine. There is evidence that in the glands they are combined with a specific protein referred to as the van Dyke protein (van Dyke, Chow, Greep and Rothen, 1942) . This protein has a molecular weight of approximately 30,000 and a strikingly high content of cystine (approximately 16.0 per cent; the cystine content of insulin is 12.0 per cent).

The morphologic studies of the pars nervosa which followed Bucy's designation of l)ituicytes as specific cells peculiar to the neurohypophysis did not contribute to the elucidation of neurohypophyseal function. The true explanation of the function of the neurohypophysis resulted from studies on neurosecretion which were proceeding contemporaneously with the pituicyte investigations but were not, before 1949, considered to be related directly to neurohypophyseal secretion.

"Neurosecretion" is a general term referring to the presence within certain neurons of accumulations of products, usually proteins, which are not common to the majority of neurones. The first description of 8 secretion-containing nerve cells was that by Speidel (1919) who described such cells in the spinal cord of the skate. The further development of the subject is amply reviewed by Scharrer and Scharrer (1940, 1945, 1954). Neurosecretory accumulations in tb.e hypothalamus were first described in the bony fish (Phoxiims laevis L.), and further papers by the Scharrers extended their occurrence to bony fishes generally, amphibians, reptiles, mammals, and man.

The term "neurosecretion" was first applied by Scharrer in vertebrates, but it is a general term including a wide variety of secretory phenomena in vertebrate and invertebrate neurones. With the investigations on neurosecretory material in invertebrates we are not here concerned, and further discussion will be limited to the neurosecretory material of certain neurones of the vertebrate hypothalamus.

Neurosecretion was first investigated by nonspecific methods of staining, and the evidence for the presence of a secretory product in neurones was the accumulation of protein material in large droplets and vesicles often so large as to cause distortion of the cell within which they were enclosed. Investigations by Scharrer and Scharrer (1940), disclosed a wide occurrence of such gross accumulations of an unusual material in neurones of the preoptic or supraoptic and paraventricular nuclei of many vertebrates, but the absence of any known function for this material was a retarding factor in the development of the subject. Moreover, man}'- species did not appear to have these neurosecretory droplets and in species where they were found they were often not observed in young animals but only in those of older ages. These observations seemed to make it unlikely that any essential function was mediated by this material. Palay (1943) was able to trace a transport of neurosecretory material along the axons of the preopticohypophyseal tract in fishes, but with the staining methods then in use it could not be shown that this was a general phenomenon in all vertebrates. This position has been entirely changed by the development of more specific methods for staining neurosecretion which permit its identification, even in traces, by staining methods which demonstrate the specific quality of the protein instead of the earlier method of recognition which depended on the physical form assumed by gross collections of the material.

Bargmann (1949) found that Gomori's (1941) method for staining the insulin-containing granules of the ^-cells of the pancreatic islets also stained intensely the accumulations of neurosecretory protein in the vertebrate hypothalamus. The intro



duction of this clirome-alum hematoxylin staining method revolutionized both the technical side of investigations of neurosecretion and the interpretation of its significance. With this method it became clear that the grosser collections of neurosecretion detected by earlier methods, which occur only in an erratic fashion, constituted but a small part of the hypothalamic neurosecretion which is found in neurones of every vertebrate hypothalamus (Bargmann and Hild 1949; Hild, 1950; 1951a, b; Bargmann, Hild, Ortman and Schiebler, 1950). The neurosecretory material could be seen throughout the course of the supraopticoand paraventriculo-hypophyseal tracts and their continuations in the neural eminence and stalk into the pars nervosa (Fig. 3.33). The greater part of the stainable neuro

FlG.3.33(w/^/n, ). (■(.rnii;,! ,M.-n.,n ..I il.c I,n|m,|,|,ysis of a normal rat stained with dilute uldehydcfuchsin (AF) after permanganate oxidation. The pans nervosa contains a large amount of darkly stained neurosecretion. P.N. pars nervosa; P.I., pars intermedia; PA., pars anterior. KMnOi , AF, X50.

Fig. 3.34 (loirer). Coronal section of the liypojiliysis of a rat which was deprived of drinking water for 5 daj^s. An extreme reduction in the content of neurosecretion is apparent in comparison with Figure 3.33. Key to lettering as in Figure 3.33. KMnO, , AF, X 50.

secretion in mammals is found in the nerve fiber terminals in the pars nervosa; the amount in other parts of the hypothalamoneurohypophyseal system in some species and especially in j^oung animals may be insignificant by comparison. The distribution corresponds to that of the neurohypophyseal hormones.

In addition to demonstrating the exact distribution of neurosecretion, specific staining made it possible to demonstrate that the material formed in the perikaryon is transported, probably by axoplasmic flow, towards the pars nervosa. After stalk section the stainable neurosecretion accumulates in the proximal stump of the transected axons <Hild, 1951b; Hild and Zetler, 1953; Scharrer and Scharrer, 1954).

Depletion of stainable neurosecretion in the pars nervosa occurs in animals subjected to dehydration (Hild and Zetler, 1953; Leveque and Scharrer, 1953). The loss of stainable material is almost complete in animals subjected to prolonged water deprivation (Fig. 3.34) and parallels the loss of hormone content. It is clear that dehydration is a powerful stimulus to the release of stainable neurosecretion from the pars nervosa and that the material released is either itself hormonally active or is accompanied by the hormones. Moreover, prolonged dehydration causes a depletion of both stainable neurosecretion and of hormone activity in the neurones and axons of the supraoptic and paraventricular nuclei in the hypothalamus. This shows that the hormone activity accompanies the stainable neurosecretion in its passage along the axons, the transport of both being accelerated by the stimulus of dehydration.

From these observations it is inferred that the neurohypophyseal hormones are produced in the perikarya of the neurones of the supraoptic and paraventricular nuclei and are transported along the axons to the fiber terminals from which they are_ released to be carried away by the adjacent blood vessels. In their formation, transport, and release the hormones are accompanied by or form part of a specific protein, the stainable neurosecretory substance. Observations of the type presented above have been repeated in a number of different spe



cies of vertebrates with results which are always concordant with this concept of the function of hypothalamic neurosecretion. The subject has been reviewed by Bargmann (1954), Scharrer and Scharrer (1954), Palay (1953), Hild (1956), and Sloper (1957). It is now generally accepted that the neurohypophyseal hormones have their origin in the neurosecretory cells of the hypotholamus.


Specific histochemical tests indicate that the neurosecretion in the hypothalamus and neurohypophysis is a protein with a high cystine content (Barrnett, 1954; Adams and Sloper, 1955). The test used by Adams and Sloper, performic acid oxidation followed by Alcian blue, depends on the oxidation of the sulfur groups of cystine to sulfonic acid groups by performic acid and subsecjuent binding of the basic dye Alcian blue by the strong acid groups so formed. Dawson's (1953) method depends on the same principles but involves the use of permanganate for the oxidation and aldehyde-fuchsin as a basic dye. It is probable that Gomori's (1941) chrome alum-hematoxylin method also demonstrates sulfonic acid groups formed by permanganate oxidation of cystine, because it as well as Dawson's method stains the insulin-containing granules of pancreatic islet ^-cells. There is little doubt that the secretion is a complex of the normally active octapeptides with the van Dyke protein which is demonstrated by these staining methods (Smith, 1951 ; Sloper, 1957). The amount of material stained is much greater than the content of specific octapeptides and corresponds with estimates of the amount of van Dyke protein present (Albers and Brightman, 1959). The stainable material is not a glyco-lipo-protein complex as suggested by Schiebler (1952) because tests for lipid and carbohydrate are weak, variable in different species, and often negative (Sloper, 1955; Howe and Pearse, 1956).

It is often stated that the stainable component of neurosecretion is an inert carrier substance which can be totally extracted from the tissue by organic solvents leaving the hormone activity behind. This is not correct. As Sloper (1955) demonstrated, neither the stainable material nor the hormone activity can be extracted by organic solvents, but both remain water soluble after treatment with organic solvents and neither can be demonstrated by staining unless a chemical fixation is used before exposure of the tissue to water. It would appear that about 10 per cent of the cystine content of the van Dyke protein octapeptide complex is in the octapeptide component and that the octapeptides, if they are retained by fixation, contribute to depth of staining of the neurosecretory material by the staining methods now in use.

C. The Physical Form of Neurosecretion in the Neurohypophysis: the Secretory Granule

Electron microscopy is required to reveal the form in which neurosecretion occurs in the mammalian neurohypophysis. Electron microscopic studies of the mammalian neurohypophysis have been published by Green and van Breemen (1955), Palay (1955, 1957) and Bargmann and Knoop (1957).

The relationships of the various structures in the felted mass of nerve fibers and cytoplasmic processes of pituicytes in the pars nervosa are especially difficult to visualize and this difficulty is responsible for some differences in the interpretation of the structure by the different authors. Our primary interest is with the form in which neurosecretory substance occurs. There is general agreement that it is visible in electron micrographs as numerous spherical bodies 100 to 180 m/x. in diameter. These l)odies have the typical structure of secretory granules as seen in the exocrine cells of the pancreas or the endocrine cells of the adenohypophysis. A dense limiting membrane is visible enclosing a homogeneous content. Identification of the granules with the stainable neurosecretion follows from the appearance of the Herring bodies. These bodies in the rat appear to be simple dilatations of the axones and do not have the end bulb structure described by Bucy (19301 and Bodian (1951) for the Herring bodies in other mammals. In these bulbous structures in the rat the secretory granules are so closely packed as to leave little room for any other structure.

I II. 3.35. Electronmicrograph of neurohyp(M'li\ -1- "i i ii. ni w \i mi .Mcii luiy material inside vesicular membrane.-, i.t.. liioplLi- ol m uiu^^ucietuiy .-ub.-iancc, ur neurosecretory granules. Electronmicrograph by S. L. Palay, National Institutes of Health, Bethesda. X 37,100.

Palay finds the neurosecretory granules in the neurones of the supraoptic nucleus and in the axones throughout their course. There is a concentration of the granules in the terminal part of the axons (Fig. 3.35) . The terminations are blunt or bulbous and end on a basement membrane separated by connective tissue from the basement membrane and endothelium of the adjacent capillaries. There do not appear to be any secretory granules in the pituicytes. Secretory granules are not found outside the axones in tissue spaces or in the lumen of the capillaries. It appears that only the soluble content of the granule is liberated during secretion ; the complete granules are not extruded.

D. The Hormone Content of the Neurosecretory Granules

From hypophyseal tissue homogenized in sucrose solution oxytocic and vasopressor activity can be deposited by centrifugation (Pardoe and Weatherall, 1955) and from this behavior it is inferred that the hormones are contained in the granules and that in sucrose solution the granules are preserved with their hormone content intact. Pardoe and Weatherall obtained evidence that the oxytocin and vasopressor activities were in particles of different size or of different fragility because a partial separation was obtained by differential centrifugation. This behavior of these hormones in tissue suspensions in sucrose solutions is paralleled by that of the gonadotrophic hormones, as found hv McShan and Mever (1952).

The stainable content of the neurosecretory granules is released by agents which destroy the membrane. It is removed in tissue which has been perfused by buffers saturated with ether.

E. The Distribution of Oxytocin and Vasopressin

The ratio of vasopressin to oxytocin varies markedly among different species of mammals and also is variable in different parts of the hypothalamo-neurohypophyseal system in any one species (Adamson, Engel, van Dyke, Schmidt-Neilsen and SchmidtNeilsen, 1956) . In the camel the ratio vasois 2.6, the ratio in the paraventricular nupressin: oxytocin in the supraotic nucleus cleus is 0.26. The ratios of vasopressin to oxytocin in the supraoptic and paraventricular nuclei of the dog are approximately 10 times those in the camel.

A simple explanation of these variations would be provided by the hypothesis that the two hormones are formed in different cells. One advantage of such an hypothesis is that it provides the only simple explanation of the ability to secrete the hormones separately (Smith, 1951). The fact that two types of neurosecretory cells have not been demonstrated in the supraoptic and paraventricular nuclei by staining reactions is no objection to this hypothesis; the staining reactions at present used to demonstrate neurosecretion would not be expected to differentiate cells whose secretions are so closely related chemically.

F. Inferences Concerning Rate of Secretion from Cytologic and Histochemical Studies

The staining of neurosecretion in sections allows an estimate of the hormone content and a demonstration of changes in hormone content of different parts of the hypothalamo-neurohypophyseal system to be made, but does not demonstrate the rate of formation or release of neurosecretion. The demonstration by Sloper (1957) that S^^labeled cystine is incorporated into the neurosecretion suggests a method by which the rate of turnover as distinct from the content may be investigated.

The neurosecretion must be formed by the ribonucleic acid-containing material in the perikaryon which in nerve cells is called Nissl substance. The marked accumulation of neurosecretion in some neurones of old animals of certain species which eventually fills the axon, the dendrites, and a large part of the perikaryon, is apparently the result of an aging process which makes it difficult for such cells to release their secretion. That there is a stagnation in such cells in indicated by the fact that the amount of Nissl substance in such cells is much reduced (Hild, 1956).

It is the cells which do not show such marked accumulations of neurosecretion, except at their axone terminals, and have large amounts of Nissl substance in their perikarya, that are presumably responsible for most of the secretory function of the system. There does not seem, therefore, to be any relation between the accumulations of neurosecretion in the cells of the hypothalamus and the secretory function; in particular there is no reason to think that the accumulation of neurosecretion in the hypothalamus is related more directly to alterations in adenohypophyseal than to a disturbance of neurohypophyseal function.

Pepler and Pearse (1957) observed hypertrophy of the cells of the supraoptic and paraventricular nuclei in rats given 2.5 per cent NaCl in the drinking water and also in lactating rats. In both groups there was an increase in the amount of specific acetylcholine esterase in the hypertrophied cells. From these observations it may be inferred that the reduction both of hormone content and stainable neurosecretion observed by Rennels (1958) in the neurohypophyses of lactating rats was accompanied by an increased production of neurosecretion and was therefore the result of an increased rate of discharge of this material. It would appear that the demonstration of the content of specific enzymes by methods which are capable of giving a cytologic localization may prove to be a valuable method of estimating the rate of secretion by endocrine cells. Such methods are badly needed to complement the information obtained from the staining of specific granules which indicates only the hormone content.

XVI. References

Adams. C. W. M., .'^nd Sloper, J. C. 1955. The liypothalamic elaboration of posterior pituitary principles in man, the rat and dog : histochemical evidence derived from a performic acidAlcian bhie reaction for cystine. J. Endocrinol., 13, 221-228.

Ad.^ms, C. W. M., and Swettenh.^m, K. V. 1958. The histochemical identification of two types of basophil cell in the normal human adenohypophysis. J. Path. & Bact., 75, 95-103.

Adamson, K., Engel, S. L., v.^n Dyke, H. B., Schmidt-Nielsen, B., and Schmidt-Nielsen, K. 1956. The distribution of oxytocin and vasopressin (antidiuretic hormone) in the neurohvpophvsis of the camel. Endocrinology, 58, 272-278.



Albers, R. W., .\-nd Brightmax, M. W. 1959. A major component of neurohypophysial tissue associated with antidiuretic activity. J. Neurochem., 3, 269-276.

ALL.ARA, E. 1953. Richerche sui processi secretori nelle ghiandole endocrine. IV. Modificazioni indotte dalla bassa temperatura sulla struttura dell'ipofisi di Mus rattus albinus. Boll. Soc. ital. biol. sper., 28, 1989-1991.

Allen, B. M. 1930. Source of the pigmentary hormone of amphibian hypophysis. Froc. Soc. Exper. Biol. & Med., 27, 504-505.

Aplington, H. W., Jr. 1942. Correlative cyclical changes in the hypophysis and gonads of Necturus maculosus rafinesque. Am. J. Anat., 70, 201-249.

AsTwooD, E. B., Raben, M. S., and Payne, R. W. 1952. Recent Progr. Hormone Res., 7, 1-57.

AziMOV, G. I., AND Altman, A. D. 1938. Various parts of the anterior lobe of the pituitary body and the physiology of lactation. Compt. rend. Acad. sc. U.S.S.R., 20, 621-625.

B.-iiLEY, P., AND Davidoff, L. M. 1925. Concerning the microscopic structure of the hypophysis cerebri in acromegaly. Am. J. Path., 1, 185207.

Baker, B. L., and Everett, N. B. 1947. The effect of diethylstilbestrol on the anterior hypophysis of thvroidectomized rats. Endocrinology, 41, 144-157.

Bargmann, W. 1949. Uber die neurosekretorische Verknijpfung \'on Hypothalamus und Neurohypophyse. Ztschr. Zellforsch., 34, 610-634.

Bargmann, W. 1954. Das Zwischenhmihypophysensystem. Berlin: Springer-Verlag.

B.argmann, W., and Hild, W. 1949. Uber die Morphologic der neurosekretorischen Verkniipfung von Hypothalamus und Neurohypophyse. Acta Anat., 8, 264-280.

Bargmann, W., Hild, W.. Ortman, R., and SchiebLER, T. H. 1950. Morphologische und experimentelle Untersuchungen Uber das hypothalamischhypophysare Svstem. Acta neuroveg., 1, 233-275.

Bargmann, W., and Knoop, A. 1957. Electronenmikroskoi)ische Beobachtungen an der Neurohypophyse. Ztschr. Zellforsch., 46, 242-251.

B.\RRiNGT0N, E. J. W., and M.^TTY, A. J. 1955. The identification of thyrotrophin-secreting cells in the pituitary gland of the minnow (Phoxinus phoxinna). Quart. J. Microscop. Sc-., 96, 193201.

Barrnett, R. J. 1953. The hi.*tochemical distribution of protein-bound sulfhydryl gioups. J. Nat. Cancer Inst.. 53, 905-926.

Barrnett, R. J. 1954. Histochemical denion.slration of di.sulfi(l(> grou])s in the neurohyixjpliy.-^is under normal and experimental conditions. Endocrinology, 55, 484-501.

Barrnett, R. j., and Creep, R. O. 1951. The pituitary gonadotropic activity of stalk-sectioned male rats. Endocrinology, 49, 337-348.

Barrnett, R. J., Ladman, E. J., McAlla.ster. N. J., -AND Siper.stei.\, E. R. 1956. The localization

of glycoprotein hormones in the anterior pituitary glands of rats investigated by differential protein solubilities, histological stains and bio-as.says. Endocrinology, 59, 398-418.

Barrnett, R. J., and Seligman, A. M. 1954. Histochemical demonstration of sulfhydryl and disulfide groups of protein. J. Nat. Cancer Inst., 54, 769-803.

Berblinger, W. 1932. Pathologie and pathologische Morphologic der Hypophyse des Menschen. In Handhuch der inneren Sekretion, M. Hirsch, Ed. Vol I, pp. 910-1097.

Berblinger, W., and Burgdorf, A. L. 1935. Neue Farbemethode zur Darstellung der Gewebebestandteile der Hypophyse des Menschen. Endokrinologie, 15, 381-388.

BoDiAN, D. 1951. Nerve endings, neurosecretory substance, and lobular organization of the neurohypophysis. Bull. Johns Hopkins Hosp., 89, 354-376. "

Bradbury, J. T. 1947. Ovarian influence on the response of the anterior pituitary to estrogens. Endocrinology, 41, 501-513.

Brokaw, R., Briseno-Castrejon, B., and Finerty, J. C. 1950. Quantitative studies of cell types in the rat hypophysis following prolonged periods of unilateral adrenalectomy. Texas Rep. Biol. & Med., 8, 312-319.

Brolin, S. E. 1945. A study of the structural and hormonal reactions of the pituitary body of rats exposed to cold. Acta Anat., Suppl. 3, 1165.

Brooks, C. McC. 1938. A study of the mechanisms whereby coitus excites the ovulationproducing activity of the rabbit's pituitary. Am. J. Physiol., 121, 157-177.

Brown, L. H. V., .\nd Hess, M. 1957. Separation of hormone activities in the anterior pituitary bv ultracentrifugation. Am. J. Physiol., 188, 25-29.

Brown. L. T. and Knigge, K. M. 1958. Cytology of the pars distalis in the pituitary gland of the chicken. Anat. Rec, 130, 395-396.

Bucy, P. C. 1930. The pars nervosa of the bovine hypophysis. J. Comp. Neurol., 50, 505-519.

Bvrch.A.B. 1946. An expi rinHiilal study of the histological and ililTni im.ition of the epithelial hypophy.<is in //.///</ /< (jdla. Univ. California Publ. Zool., 51, 185-213.

Burt, A. S., and Velardo. J. T. 1954. Cytology of the human adenohypophysis as related to bioassays for tropic liormones. J. Clin. Endocrinol., 14, 979-996.

C.^tchpole, H. R. 1947. Cellular distribution of glycoprotein in the anterior pituitary gland. In Communications of 17th lutcniational Physiological Congress, \k 05. Lon(h)n: Oxford University Press.

C.^TCHPOLE, H. R. 1949. Distribution of glycoprotein hormones in the anterior pituitary gland of the rat. J. Endocrinol., 6, 218-225.

CiiADWicK, C. S. 1936. Cyclic morphologic variations in the anterior hypophysis of the guinea pig. Am. J. Anat., 60, 129-147.



Chiti, E., and Zinolli, J. 1952. Relationships between pituitaries and adrenals. Arch. Marcliano Pat. Clin., 7, 41-57.

CiESL.AK, E. S. 1945. Relations between the reproducti\-e cycle and the pituitary gland in the snake (Thamnophis radix). Phvsiol. ZooL, 18, 299-329.

Clevel.a.\d, R., .\nd Wolfe, J. M. 1933. Cyclic histological changes and variations in the anterior hypophysis of the sow (Sus scrofi). Am. J. Anat"., 53, 191-220.

Colombo, E. 1948. The effect of bilateral adrenalectomy on the adenohvpophysis. Prensa med. argent., 35, 1951-1955. "

Colombo, E. 1949. Cellular changes of the anterior pituitary gland after unilateral adrenalectomy. Prensa med. argent., 36, 1496-1498.

Copel.and, D. E. 1943. Cytology of the pituitary gland in developing and adult Triturus viridescens. J. MorphoL. 72, 379-409.

CoRDiER, R. 1953. L'hypophyse de Xenopus. Ann. Soc. Roy. Zool. Belg., 84, 5-16.

Crooke, a. C. 1935. A change in the basophil cells of the pituitary gland common to conditions which exliibit the syndrome attributed to basophil adenoma. J. Path. & Bact., 41, 339349.

Crooke, A. C, .'Vnd Russell, D. S. 1935. Pituitary gland in Addison's disease. J. Path. & Bact.. 40, 255-283.

Crossmon, G. C. 1937. A modification of Mallory's connecti\'e tissue stain with a discussion of the principles involved. Anat. Rec, 69, 33-38.

CusHiNc. H.. .^XD D.^viDOFF, L. M. 1927. The pathological findings in four autopsied cases of acromegaly with a discussion of their significance. Rockefeller Inst. Med. Res. Monogr., 22, 1-131.

D.^LTON, A. J., AND Felix, M. D. 1953. Studies on the Golgi substance of the epithelial cells of the epididymis and duodenum of the mouse. Am. J. Anat., 92, 277-305.

Daltox, a. J., AND Felix. M. D. 1954. Cytologic and cytochemical characteristics of the Golgi substance of epithelial cells of the epidymis — in situ, in homogenates, and after isolation. Am. J. Anat., 94, 171-208.

Daltox, A. J., and Felix, M. D. 1956. A comparative study of the Golgi complex. J. Biophvs. Biochem. Cytol., Suppl., 2, 79-83.

D'Angelo, S. a. 1955. Thyrotrophic hormone production and aldehyde-fuchsin staining basophils in the rodent adenohj'pophvsis. Anat. Rec, 121, 282.

Dawson, A. B. 1938. The epithelial components of the pituitary gland of the opossum. Anat. Rec, 72, 181-193.

Dawson, A. B. 1939. Differential staining of the anterior pituitary gland of the cat. Stain Technol., 14, 133-138.

Dawson, A. B. 1946. Some evidences of specific secretory activity of the anterior pituitary gland of the cat. Am. J. Anat., 78, 347-410.

Dawson, A. B. 1948. The relationship of pars tuberalis to pars distalis in the hypophysis of the rhesus monkey. Anat. Rec, 102, 103-122.

Dawson, A. B. 1953. Evidence for the termination of neurosecretory fibers within the pars intermedia of the hypophysis of the frog, Rana pipiens. Anat. Rec, 115, 63-69.

Dawson, A. B. 1954a. Differential staining of two types of acidophile in the anterior pituitary gland of the rat. Anat. Rec, 120, 810.

D.\wsoN, A. B. 1954b. The regional localization of five distinct morphological types of cells in the anterior pituitary gland of the rhesus monkey. Anat. Rec, 120, 810.

Dawson, A. B. 1957. Morphological evidence of a possible functional interrelationship between the median eminence and the pars distalis of the Anuran hypophysis. Anat. Rec, 128; 7789.

Daw.son, a. B., and Friedgood, H. B. 1937. The occurrence and distribution of a third type of granular cell in the anterior pituitary of the cat. Anat. Rec, Suppl. 1, 70, 129.

Dawson, A. B., and Friedgood, H. B. 1938a. Differentiation of two classes of acidophiles in the anterior pituitary of the female rabbit and cat. Stain Technol., 13, 17-21.

Dawson, A. B., and Friedgood, H. B. 1938b. The occurrence of a second type of acidophile cell in the anterior pituitary of the female cat and its relation to sexual activity. Anat. Rec. Suppl., 70, 21.

DELA\yDER, A. M., T.arr, L., and Ceiling, E. M. K. 1934. The distribution in the chicken's hypophysis of the so-called posterior lobe principles. J. Pharmacol. & Exper. Therap., 51, 142-143.

Dempsey, E. W., and Wislocki, G. B. 1945. Histochemical reactions associated with basophilia and acidophilia in the placenta and pituitary gland. Am. J. Anat., 76, 277-301.

Desclaux, p., Soulairac, A., and Chaneac, H. 1953. Origine de I'hormone corticotrope a partir des cellules eosinophiles de I'hypophyse. Compt. rend. Soc bioL, 147, 44-45.

Dreyfus, G., and Zara, M. 1951. La cellule de Crooke: cellule reactionelle au course de I'hypercorticisme. Bull. Soc. Med Hop. Paris, 67, 15-16.

Eakin, R. M., .^nd Bush, F. E. 1957. Development of the amphibian pituitary with special reference to the neural lobe. Anat. Rec, 129, 279-295.

Edmunds, A. W. B., McKeown, K. C, and Coleman. P. N. 1958. A case of Cushing's 33^drome with pigmentation and severe hypokalaemic alkalosis. J. Clin. Path., 11, 237-243.

Elder, W. H., and Finerty, J. C. 1943. Gonadotrophic activity of the pituitary gland in relation to the seasonal sexual cycle of the cottontail rabbit (Sylvilagus floridanus mearnsi). Anat. Rec, 85, 1-16.

Elftman, H. 1956. Response of the anterior pi



tuitary lo dirhiomate oxidation. J. Histochem., 4,410-411.

Erdheim, J. 1926. Pathologic der Hypophysengeschwiilste. Ergebn. Patli., 21, 482-561.

Erdheim, J., and Stumme, E. 1909. Uber die Schwangerschaftsveranderung bei der hypophyse. Beitr. path. Anat., 46, 1-17.

Etkin, W. 1941. On the control of growth and activity of the pars intermedia of the pituitary by the hypothalamus in the tadpole. J. Exper. Zool., 86, 113-139.

Etkin, W. 1958. Independent differentiation in components of the pituitary complex in the wood frog. Proc. Soc. Exper. Biol. & Med., 97, 388-393

Ev.\NS, H. M., AND Simpson, M. E. 1929. The effect of pregnancy on the anterior hypophysis of the rat and cow as judged by the capacity of implants to produce precocious maturity. Am. J. Physiol., 89, 379-380.

Everett, N. B., and Baker, B. L. 1945. The distribution of cell types in anterior hypophysis during late pregnancy and lactation. Endocrinology, 37, 83-88.

Ezrin, C, Swanson, H. E., Humphrey, J. G., D.^wsoN, J. W., and Wilson, W. 1958. The delta cell of the human adenohypophysis : its response to acute and chronic illness. J. Clin. Endocrinol., 18, 971-936.

Farquhak, M. G. 1956. Preparation of ultrathin tissue sections for electron microscopy : review and compilation of procedures. Lab. Invest., 5, 317-337.

F.\rquhar, M. G. 1957. "Corticotrophs" of the rat adenohypophysis as revealed by electron microscopy. Anat. Rec, 127, 291.

Farquhar, M. G., and Rinehart, J. F. 1954a. Electron microscopic studies of the anterior pituitarv gland of castrate rats. Endocrinology, 54,51&^541.

F.ARQUH.AR, M. G., AND RiNEH.'VRT, J. F. 1954b.

Cytologic alterations in the anterior pituitary gland following thyroidectomy: an electron microscope study. Endocrinology, 55, 857-876.

FaRQUH.'VR, M. G., AND RiNEHART, J. F. 1955.

Further evidence for the existence of two types of gonadotrophs in the anterior pituitary of the rat. Anat. Rec, 121, 394.

F.\rqxjhar, M. G., .^nd Wellings, S. R. 1957. Electron microscopic evidence suggesting secretory granule formation within the Golgi apparatus. J. Biophys. Biochem. Cytol., 3, 319322.

Finerty, J. C, AND Binhammek, R. 1952. The effects of severe burns on ]>ituit;uy cytology and adrenocorticotrophic function. Anat. Rec, 112,329.

Finerty, J. C, Hess, M., and Biniiammer, R. 1952. I'ituitary cytological manifestations of heightened adrenocorticotrophic activitv. Anat. Rec, 114, 115-125.

FisPlER, C. 1937. The site of roiniation of I ho posterior lobe hormones. Endociiiiologv. 21, 19-29.

Floderus, S. 1949. Changes m ihc human liy

pophysis in connection with pregnancy. Acta. Anat., 8, 329-346.

Folley, S. J. 1952. Some aspects of the phj'siology of the anterior pituitary lactogenic hormone. Ciba Foundation Colloquia Endocrinol., 4, 381-401.

Folley, S. J., and Y'oung, F. G. 1938. The effect of anterior pituitary extracts on estabhshed lactation in the cow. Proc. Roy. Soc, London, ser. B, 126, 45.

Foster, C. L. 1947. Some observations upon the Golgi elements of the anterior pituitary cells of normal and stilbestrol-treated male rats, using the Sudan black technique. Quart. J. Microscop. Sc, 88/4, 409-117.

Foster, C. L., .■vnd Wilson, R. R. 1951. Grampositive cells of the human anterior pituitary. Nature, London, 167, 528-529.

Friedgood, H. B., and Dawson, A. B. 1940. Physiological significance and morphology of the carmine cell in the cat's anterior pituitary. Endocrinology, 26, 1022-1031.

Friedm.^n, M. H., and Hall, S. R. 1941. The site of elaboration of the pituitary gonadotrophic hormone and of prolactin. Endocrinologv, 29, 179-186.

Gatenby, J. B., AND LuFTY, R. G. 1956. The Golgi apparatus and the electron microscope. Nature, London, 177, 1027-1029.

Gabe, M. 1953. Sur ciueltiues applications de la coloration par la fuchsine-paraldehyde. Bull. Microscope. AppL, Paris, 3, 153-162.

Ceiling, E. M. K. 1935. The hypophysis of the finback (Balaenoptera physalis) and the sperm (Physeter megalocephalus) whale. Bull. Johns Hopkins Hosp., 57, 123-142.

Geschwind, I. I., Hess, G. P., Condliffe, P. G., Evans, H. M., and Simpson, M. E. 1950. Preparation of nonprotein fractions possessing adrenocorticotrophic activity from fresh sheep pituitary glands. Science, 112, 436-437.

Geschwind, 1. 1., and Li, C. H. 1952. Interaction of growth hormone with tliyroxiue and other hormones, as measured by the tibia test. J. Clin. Endocrinol., 12, 937."

Gilmore L. 0., Petersen, W. E., and Rasmussen, A. T. 1941. Some morphological and functional relationships of the bovine hypophysis. Univ. Minnesota Agric Exper. Sta., Tech. Bull. 145.

Giroud, A., AND M.\RTiNET, M. 1948. Localisations de certaines functions dan.s le lobe antericur de I'hypophyse. Ann. endocrinol., 9, 343-349.

Goldberg, R. C, and Chaikoff, I. L. 1952a. On the occurrence of six cell types in the dog anterior pituitary. Anat. Rec, 112, 265-274.

Goldberg, R. C, AND Chaikoff. 1. 1. 1952b. Myxedema in the radiothyroiderlomizcd dog. Endocrinology, 50, 115-i23.

Golden, A., Bondy, P. K., and Sheldon, W. H. 1950. Pituitary basophile hyperplasia and Crookc'.s hyaline changes in man after ACTH tiierapv. Proc Soc. Exper. Biol. & Med., 74, 455-458.

Gomori, G. 1941. Observations with differential



stains on human islets of Langerhans. Am. J. Path., 17, 395-406.

GoMORi, G. 1950. Aldehvde-fuchsin : a new stain for elastic tissue. Am. J. Clin. Path., 20, 665666.

Green, J. D. 1951. Comparative anatomy of hypophysis with special reference to its blood supply and innervation. Am. J. Anat., 88, 225-311.

Green, J. D. 1956. Neural pathways to the hypophysis in Hypothalamic-hypophysial Interrelationships, W. S. Fields, Ed., pp. 3-16. Springfield, 111.: Charles C Thomas.

Green, J. D., and v.^n Breemen, V. L. 1955. Electron microscopy of the pituitary and observations on neurosecretion. Am. J. Anat., 97, 177227.

Greep, R. O., .and Chester Jone.s, L 1950. Steroid control of pituitary function. Recent Progr. Hormone Res.. 5, 197-261.

Guyer, M. F., .\nd Claus, P. E. 1937. Vacuolation of the anterior pituitary gland following castration, implantation of cancer tissue, and thyroidectomj'. Anat. Rec, 67, 145-155.

Griesbach, W. E., Hornabrook, R. W., and Purves, H. D. 1956. Changes in the rat pituitary produced by exposure to cold. In C ommunications of XX th International Physiological Congress, p. 370. London: Oxford University Press.

Halmi. N. S. 1950. Two types of basophils in the anterior pituitary of the rat and their respective cvtophvsiological significance. Endocrinologj', 47, 289-299.

Halmi. N. S. 1951a. Differentiation of two types of basophils in the adenohypophysis of the rat and the mouse. Stain TechnoL, 27, 61-64.

Halmi, N. S. 1951b. Further observations on two tvpes of basophil cells in the anterior pituitary. Anat. Rec, 109, 300.

H.ALMi, N.S. 1952a. Two types of basophils in the rat pituitary: "thyrotrophs" and gonadotrophs" vs. beta and delta cells. Endocrinology. 50, 140-142.

Halmi, N. S. 1952b. The effects of graded doses of thyroxine on the anterior pituitary of hypothyroid male albino rats. Anat. Rec, 112, 1736".

Halmi, N. S., and Bogdanove. E. M. 1951. Effect of thyroidectomy on ACTH content of rat adenohvpophvsis. Proc Soc Exper. Biol. & Med., 77, 518-520.

Halmi, N. S., and Davies. J. 1953. Comparison of aldehyde-fuchsin staining, metachromasia and periodic acid-Schiff reactivity of various tissues. J. Histochem., 1, 447-459.

H.artm.\nn, J. F. 1944. Seasonal cytological changes in the anterior hypophysis of the garter snake. Am. J. Anat., 75, 121-149.

Hartmann, J. F., Fain, W. R., and Wolfe, J. M. 1946. A cytological study of the anterior hypophysis of the dog with particular reference to the presence of a fourth cell type. Anat. Rec, 95, 11-27.

Hedinger, C. E., and Farquhar, M. G. 1957. Elektronenmikroskopische LTntersuchungen von zwei Typen acidophilen Hypophysen\order

lappenzellen bei der Ratte. Schweiz. Ztschr. allg. Path., 20, 766-768.

Hellweg, G. 1951. tJber die Silberimpragnierbaren Zellen der menschlichen Hvpophyse. Ztschr. Zellforsch., 36, 349-360.

Herlant, M. 1936a. Influence de I'uremie experimentale sur I'hypophvse du rat. Compt. rend. Soc biol., 123, 538-540.

Herlant, M. 1936b. Influence de la basophilie experimentale sur le pouvoir gonadotrope de I'hypophyse du rat. Compt. rend. Soc. biol., 123, 1267-1270.

Herlant, M. 1942. Determinisme de la basophilie des cellules cyanophiles du lobe anterieur de I'hypophyse. Extrait du Bulletin de I'Acadamie royale de Belgique (Classe des Sciences) Seance de 4 juillet 1942, Nos. 7-9.

Herlant, M. 1943. Recherches sur la localisation histologique des hormones gonadotropes femelle au niveau de I'hypophyse anterieure. Arch. Biol.. 54, 225-357.

Herlant, M. 1949. Study of the pituitary body with the periodic acid-Schiff reaction. Nature, London, 164, 703-704.

Herlant, M. 1951. Etude des cellules basophiles de I'hypophyse par la methode histophotometrique. Ann. Endocrinol., 12, 80-88.

Herlant, M. 1952a. Separation des activites du lobe anterieur de I'hypophyse par la methode des centrifugations differentielles. Ann. endocrinol., 13,611-623.

Herlant, M. 1952b. Les granulations acidophiles de I'hypophyse et leur signification physiologique Compt. rend. Soc. biol., 146, 1280-1882.

Herlant, M. 1953a. La corticotrophine est bien elaboree par les cellules acidophiles de I'hypophyse. Ann. endocrinoL, 14, 64-70.

Herlant, M. 1953b. Localisation des hormones gonadotropes au niveau des cellules du lobe anterieur de I'hypophyse. Extract de la 2" Reunion des Endocrinologistes de la Langue jrangaise, pp. 81-100. Paris: Masson & Cie.

Herlant, M. 1954a. Anatomie et physiologie comparees de I'hypophyse dans la serie de vertebres. Bull. Soc." Zool." France, 79, 256-281.

Herlant, M. 1954b. Nouvelles recherches sur les cellules basophiles de I'hypophyse chez I'homme. Compt. rend. A. Anat., 40, 443-447.

Herl.\nt, M. 1955. Mise en evidence de la somatotrophine dans les suspensions de granulations acidophiles de I'hypophyse par son influence sur des cellules a du pancreas. Compt. rend. Acad. Sc, 240, 1479-1481.

Herlant, M. 1956a. Correlations hypophysogenitales chez la femelle de la Chauve-Souris, Myotis myotis (Borkhausen). Arch. Biol., 67, 89180.

Herlant, M. 1956b. Les cellules basophiles du lobe anterieur de I'hvpophvse chez I'homme. Arch. Biol., 67, 539-553.

Herlant, M. 1958. La dualite des cellules gonadotropes au niveau du lobe anterieur de I'hvpophvse. Bull. Soc Rov. Beige Gynec et Obst.. 28, 69-81.

Herlant, M.. and Lison, L. 1951. Etude des eel



lules basoplules de rhypophyse par la methode histophotometrique. Ann. endoc-rinol., 12, BOSS.

Herl.\nt, M., and Rac.-vdot, S. 1957. Le lobe anterieur de I'hypophyse de la chatte au coins de la gestation et de la lactation. Arch. Biol., 58, 217-24S.

Herrick. E. H., .\nd Finerty, J. C. 1940. The effect of adrenalectomy on the anterior pituitaries of fowls. Endocrinology, 27, 279-282.

Herring, P. T. 1908. The histological appearances of the mammalian pituitary body. Quart. J. Exper. Physiol., 1, 121-159.

Hess, M., Sl.^de, I. H., Ammons, J. C, and Hendrix, V. 1955. Pituitary ACTH content of thyroidectomized-stressed rats. Anat. Rec, 121, 309.

Hewer, T. F. 1943. Ateleiotic dwarfism with normal sexual function: a result of hypopituitarism. J. Endocrinol., 3, 397-400.

HiLD, W. 1950. Zur Frage der Neurosekretion im Zwischenhirn dcr Schleie {Tinea vulgaris) und ihrer JiczK hunger zur Neurohypophyse. Ztschr. ZelUorsrh, 35, 33-46.

HiLD, W. 1951a. Vergleichende Untersuchungen iiber Neurosekretion im Zwischenhirn von Amphibien und Reptilien. Ztschr. Anat., 115, 459-479.

HiLD. W. 1951b. Das Verhalten des neurosekretorischen Systems nach Hypophysenstieldurchschneidung und die physiologische Bedeutung des Neurosekrets. Acta m luoxcg., 3, 81-91.

HiLD, W. 1956. Neuroscriviion m the central nervous system. In 11 ypntlKihinnc-hypophysial Interrelationships, W. S. Fields, Ed., pp. 17-30. Springfield, 111.: Charles C Thomas.

HiLD, W., and Zetler, G. 1953. Experimenteller Beweis fiir die Entstehung der sog. Hypophysenhinterlappenwirkstoffe in Hypothalamus. Arch. ges. Physiol., 257, 169-201.

Hildebrand, L. E., Rennels, E. G., and Finerty, J. C. 1957. Gonadotrophic cells of the rat hyl)ophysis and their relation to hormone production, "ztschr. Zellforsch., 46, 400-411.

Hohlweg, W., and Chamorro, a. 1937. Uber die luteinisierende Wirkung des Follikelhormons (lurch Beeinflussung der luteogeneu Hypophysenvorderlappensekretion. Klin. Wchnschr., 16, 196-197.

Hohlweg, W., and Junkmann, K. 1933. Uber die Beziehungen zwischen Hypophysenvorderlappen und Schilddriise. Arch. ges. Physiol., 232, 148-158.

HoTCHKiss, R. D. 1948. Microchemical reaction resulting in staining of polysaccharide structiu'cs in fixed tissue preparations. Arch. Biochem., 16, 131-141.

Houssay, B. a. 1952. Structural changes in the anterior pituitary with special reference to the adrenal cortex. Ciba Foundation Collo(|uia Endocrinol., 4, 33-35 (in discussion).

Howe, A., and Pearse, A. G. E. 1956. A histochemical investigation of noiu-oseerctory substance in the rat. J. Histochem., 4, 561-569.

Howatson, a. F.. and Ham, A. W. 1957. The fine

structure of cells. Canad. J. Biochem. Phvsiol., 35, 549-564.

Hubble, D. 1949. Gushing 's syndrome and thymic carcinoma. Quart. J. Med.," 18, 133-147.

Hum, R. F., Goldberg, R. C., and Chaikoff, I. L. 1951. Effect of e.xcess iodide upon anterior pituitary cytology of the completely thyroidectomized rat and its bearing on the ciuestion of extrathyroidal thyroxine synthesis. Endocrinology, 49, 21-24.

Hunt, T. E. 1949. Mitotic activity in the hypophysis of the rat during pregnancy and lactation. Anat. Rec, 105, 361-373.

Hurst, V., .\nd Turner, C. W. 1942. Lactogenic hormone content of anterior pituitary gland of albino mouse as compared to other species. Endocrinology, 31, 334-339.

JosT, A., AND Danysz, P. 1952. Recherches histochemiciues (reaction de MacManus) et experimentales (decapitation) sur la physiologic de rhypophyse foetale du lapin. Compt. rend. Soc. biol., 146, 1062-1066.

JosT, A., AND Tavernier, R. 1956. Application de la methode a I'acide periodique-Schiff (MacManus) a I'etude de I'hypophyse foetale du rat. Compt. rend. Acad. Sc, 243, 1353-1356.

Kahn, K. 1935. Action de I'hyperinsulinisation chronique sur I'histostructure de la pars anterieure de I'hypophyse du cobaye. Bull. Histol. appliq. a Physiol., 12, 300-317."

Keene, M. F. L., and Hewer, E. E. 1924. Glandular activity in the human foetus. Lancet, 2, 111-112.

Kelsey, F. 0., Sorenson, L., H.\gen, A. A., and Clausen, H. J. 1957. Distribution of oxytocin, vasopressin, and intermedin in the hypophysis of the beaver. Proc. Soc. Exper. Biol. & Med., 95, 703-704.

Kirkman, H. 1937. A cytological studj^ of the anterior hypophysis of the guinea pig and a statistical analvsis of its cell types. Am. J. Anat., 61, 233-287."

Knigge, K. M. 1955. Cytological changes in the hypophyses of intact and thyroidectomized rats subjected to stress. Anat. Rec, 122, 295-315.

Knigge, K. M. 1957. Cytology and hormone content of rat pituitary glands following adrenalectomy. Proc Soc. Exper. Biol. & Med., 94, 640643.

KoNEFF, A. A., Holmes, R. 0., and Reese, J. D. 1941. Prevention of adrenalectomy changes in the anterior pituitary of the rat by sodium chloride administration. Anat. Rec, 79, 275289.

Kraus, E. J. 1923. Zur Pathologie der basol)hilen Zellcn dcr Arch path. Anat., 247, 421-447.

Kraus, E. J. 1926. Die Hypophyse. Handb. Spez. path. Anat. Histol., 8, 810-950.

Kiuus, E. J. 1927. Zur Pathologic des Morbus Addisoni. Beitr. path. Anat., 78, 283-296.

Lacour, F. 1950. Recherches sin- la relation entre les cellules hypophysaires a granulations oraiigees (cellules e de Romeis) et les phe



nomenes de lactation. Compt. rend. Soc. biol., 144, 248-249. L.\cY, D. 1954. The Golgi apparatus and neutral red granules in living and frozen-dried endocrine cells of the pancreas of the mouse. J. Roy. Microscop. Soc, 74, 1-9.

L.ADMAN, A. J., .^ND B.\RRNETT, R. J. 1954. HlstO chemical demonstration of protein-bound sulfydryl and disulfide groups in cells of the anterior pituitary. Endocrinology. 54, 355-360.


1955. Hormones of the posterior pituitary. In Ch. IX. The Hormones. G. Pincus and K. V. Thimann, Eds., New York: Academic Press, Inc.

Landgrebe, F. W., and Mitchell, G. M. 1958. The neural intermediate lobe hormone ("B"). Proc. Roy. Soc. Med., 51, 560-561.

Laqueur, G. L. 1950. Cytological changes in human hypophyses after cortisone and ACTH treatment. Science, 112, 429-430.

L.AQUEUR, G. L. 1951. The human hypophysis in diseases of the adrenal cortex. Stanford Med. Bull., 9, 75-87.

L.^QUEUR, G. L., AND Fluhmann, C. F. 1942. Action of testosterone on the female rat hypoi)liysis. Endocrinology, 31, 300-302.

Leveque, T. F., and Scharrer, E. 1953. Pituicytes and the origin of the antidiuretic hormone. Endocrinology, 52, 436-446.

Lewis, D., Lee, F. C., and Astwood, E. B. 1937. Some observations on intermedin. Bull. Jolm Hopkins Hosp., 61, 198-209.

Marine, D., Rosen, S. H., and Sp.^rk, C. 1935. Effect of iodine and desiccated thyroid on anterior pituitary of goitrous and thyroidectomized rabbits. Proc. Soc. Exper. Biol. & Med., 32, 803-810.

Marshall, J. M. 1951. Localization of adrenocorticotrophic hormone by histochemical and immunochemical methods. J. Exper. Med., 94, 21-30.

McLetchie, N. G. B. 1942a. On staining the basophil cells of the himian hypophysis with special reference to the abnormal basophil cells of Cushing's syndrome. J. Endocrinol., 3, 323331.

McLetchie, N. G. B. 1942b. The pituitary basophilism syndrome of Harvey Gushing. J. Endocrinol., 3^ 332-346.

McManus, J. F. A. 1946. Histological demonstration of mucin after periodic acid. Nature, London, 158, 202.

McN.ARY, W. F. 1957. Progressive cytological changes in the hypophysis associated with endocrine interaction following exposure to cold. Anat. Rec, 128, 233-253.

McShan, W. H., and Meyer, R. K. 1952. Gonadotrophic activity of granule fractions obtained from anterior pituitary glands of castrate rats. Endocrinology, 50, 294-303., W. H., Rozich, R., and Meyer, R. K. 1953. Biochemical properties of fractions obtained from rat anterior pituitary glands by dif

ferential centrifugation. Endocrinologv. 52, 215-222.

Mellgren, J. 1945. The anterior pituitary in hyperfunction of the adrenal cortex; an anatomical study with special reference to Syndroma Morgagni and notes on prostatic hypertrophy. Acta path, et microbiol. scandinav., Suppl. 60, 22, 1-177.

Mellgren, J. 1948. Experimental investigation into genesis of adrenogenital syndromes by analysis of morphology of pituitary in hyperand hypofunction of adrenal cortex in rats. Acta. path, et microbiol. scandinav., 25, 284307.

MiALHE-VoLOSS, C., AND Benoit, J. 1954. L'intermedine dans I'hypophyse et I'hypothalamus du canard. Compt. rend. Soc. biol., 148, 56-59.

MiKAMi, S. 1956. Cytological changes in the anterior pituitary of the dog after adrenalectomy. J. Fac. Agric. Iwate Univ., 3, 62-68.

MiK.\Mi, S. 1957. The cytological significance of regional patterns in the adenohypophysis of the fowl. J. Fac. Agric. Iwate Univ., 3, 473-545.

Morris, C. J. O. R., Russell, D. S., Landgrebe, F. W., and Mitchell, G. M. 1956. The melanophore-expanding and corticotrophic activity of human pituitarv tissue. J. Endocrinol., 14, 263-267.

MiJLLER, W. 1958. On the pharyngeal hypophysis. In Endocrine Aspects oj Breast Cancer, pp. 106110. Edinburgh and London: E. & S. Livingstone, Ltd.

N.\ssoNov, D. N. 1923. Uber das Golgische Binnennetz und reine Beziehungen zu der Sekretion. Untersuchungen iiber einige Amphibiendrusen. Arch. mikr. Anat., 97, 136.

Nicholson, W. M. 1936. Observations on the pathological changes in suprarenalectomized dogs, with particular reference to the anterior lobe of the hvpophvsis. Bull. Johns Hopkins Hosp.. 58, 405-417."

N1EL.S0N, E. L. 1952. Studies on hereditary dwarfism in mice. XIII. Effect of the growth hormone and thyroxine on the growth of bones in mice with hereditary pituitary dwarfism. Acta path, et microbiol. scandinav., 30, 10-20.

Oldham, F.K. 1938. The pharmacology and anatomv of the hvpophvsis of the armadillo. Anat. Rec, 72, 265-291.

Oldh.^m, F. K., Last, J. H., and Geiling, E. M. K. 1940. Distribution of melanophore-dispersing hormone in anterior lobe of cetaceans and armadillo. Proc Soc Exper. Biol. & Med., 43, 407-410., F. K., McCleery, D. P., and Geiling, E. M. K. 1938. A note on the histology and pharmacology of the hypophysis of the manatee (Trichechus iminguis). Anat. Rec, 71, 27-32.

Olivereau, M., and Herlant, M. 1954. Etude histologique de I'hypophyse de Caecoharbus geertsii blgr. Bull. Acad. Roy. Belg., 40, 50-57.

Ortman, R. 1954. Cytochemical .study of the physiological activities in the pars intermedi.i of Rana pipiens. Anat. Rec, 119, 1-8.



Ortman, R. 1956a. Two types of acidophils in the anterior pituitarj^ of the frog and their relation to the periodic acid-Schiff reaction. J. Histochem., 4, 471-475.

Ortman, R. 1956b. A study of some cytochemical reactions and of the hormone content of the adenohypophysis in normal and in genetic dwarf mice. J. Morphol., 99, 417-431.

Ortman, R. 1956c. A study of the effect of several experimental conditions on the intermedin content and cytochemical reactions of the intermediate lobe of the frog {Rami pipiens). Acta endocrinol., 23, 437-447.

Ortman, R., and Griesbach, W. E. 1958. The cytology of the pars distalis of the wallaby pituitary. Austrahan J. Exper. Biol., 36, 609-618.

Palade, G. E. 1952. The fine structure of mitochondria. Anat. Rec, 114, 427-451.

Palade, G. E. 1953. An electron microscope study of the mitochondrial structure. J. Histochem., 1, 18^211.

Pal.ade, G. E. 1955. A small particulate component of the cytoplasm. J. Biophys. Biochem. CytoL, 1, 59-68.

PAL.-iDE, G. E. 1956. Intracisternal granules in the exocrine cells of the pancreas. J. Biophys. Biochem. CytoL, 2, 417-422.

Pal.\de, G. E., and Porter, K. R. 1954. Studies on the endoplasmic reticulum. I. Its demonstration in cells in situ. J. Exper. Med., 95, 285-289.

Pal.ay, S.L. 1945. Neurosecretion. VII.Thepreopticohvpophysial pathway in fislie.s. J. Comp. Neurol.. 82, 129-143.

Pal.ay, S. L. 1953. Neurosecretory phenomena in the hypothalamohypophysial system of man and monkey. Am. J. Anat., 93, 107-127.

Palay, S. L. 1955. An electron microscope study of the neurohypophysis in normal, hydrated, and dehydrated rats." Anat. Rec, 121, 348.

Palay. S. L. 1957. The fine structure of the neurohypophysis. Progr. NeurobioL, 2, 31-44.

Pardoe, a. v., and Weatherall, M. 1955. The intracellular localization of oxytocic and vasopressor substances in the pituitary glands of rats. J. Physiol., 127, 201-212.

Payne, F. 1942. The cytology of the anterior pituitary of the fowl. Biol. Bull., 82, 79-111.

Payne, F. 1943. The cytology of the anterior pituitary of broody fowls. Anat. Rec, 86, 1-13.

Payne, F. 1944. Anterior pituitary-thyroid relationship in the fowl. Anat. Rec, 88, 337-350.

Payne, F. 1949. Changes in the endocrine glands of the fowl with age. J. Gerontol., 4, 193-199.

Pearse, a. G. E. 1948. Cytochemistry of the gonadotropic hormones. Nature, London, 162, 651.

Pearse, A. G. E. 1949. The cytochemical drmon.stration of konadotropic liomuine in tin' liuman antc-rior hvpophvsis. J. Path. iV: liacl ., 61, 195-201.

Pearse, A. G. E. 1951. The application of cytochemistry to the localization of gonadotropliin in the pituitary. J. Flndocrinol., 7, xl\iii-i.

Peahse, A.G. E. i952. Oli.^^ervations on the local

ization, nature, and chemical constitution of some components of the anterior hvpophvsis. J. Path. & Bact., 64, 791-809.

Pearse, A. G. E. 1953. Cytological and cytochemical investigations on the foetal and adult hypophysis in various physiological and pathological states. J. Path. & Bact., 65, 355-370.

Pearse, A. G. E., and Rin.^ldini, L. M. 1950. Histochemical determination of gonadotrophin in the rat hypophysis. Brit. J. Exper. Path., 31, 540-544.

Pepler, W. J., AND Pearse, A. G. E. 1957. The histochemistry of the esterases of rat brain, with special reference to those of the hypothalamic nuclei. J. Neurochem., 1, 193-202.

Peterson, R. R. 1957. An electron microscopic study of pituitary basophiles in the guinea pig. Anat. Rec, 127, 346.

Peterson, R. R., and Weiss, J. 1955. Staining of the adenohypophyses with acid and basic dyes. Endocrinology, 57, 96-108.

PoRis, E. G. 1941. Studies on the endocrines of reptiles. II. Variation in the histology of the hypophysis of Anolis caroli7iensis, with a note on the Golgi configuration in cells of the pars anterior and pars intermedia. Anat. Rec, 80, 99-121.

PuRVES, H. D. 1956. Hyalinization and basophil adenomata in the pituitarv gland. Proc Roy. Soc Med., 49, 1014-1016.

PuRVES, H. D., AND Griesbach. W.E. 1946. Acidophil degranulation in relation to goitrogenic agents and extrathyroidal thyroxine synthesis. Brit. J. Exper. Path., 27, 170".

PuRVES, H. D., AND Griesbach, W. E. 1951a. The site of thyrotrophin and gonadotrophin production in the rat pituitary studied by McManusHotchkiss staining for glycoprotein. Endocrinology, 49, 244-264.

PuRVES, H. D., AND Griesb.ach, W. E. 1951b. Specific staining of the thyrotrophic cells of the rat pituitarv by the Gomori stain. Endocrinologv. 49, 427-428.

PuRVES. H. D., AND Griesbach, W. E. 1951c The significance of the Gomori staining of the basophils of the rat ])ituitary. Endocrinology, 49, 652-662.

PuRVES, H.D., andGriesb.uh.W.E. 1952. Functional differentiation in the acidophil cells and the gonadotrophic basophil cells of the rat pituitarv. Proc. Univ. Otago Med. Sch., 30, 27.

PuRVES, H. D., .AND Griesb.ach, W. E. 1954. The site of follicle stimulating and luteinizing hormone production in the rat i)ituitary. Endocrinology, 55, 785-793.

PuKVEs, H. D., and Griesbach, W. E. 1957a. A study on the cytology of the adenohypophysis of the dog. J. Endocrinol., 14, 361-370.

PuRVKS. H. D., and Griesb.\ch. W. E. 19571). Pituitary cytology in relation to thyrotrophic liormono secretion. Ciba Fouiulation Collo<|uia Endocrinol., 10, 51-57.

PuRVES, H. D., AND Sirktt. N. E. 1959. A study of tlio hormone contents of the rostral and

caudal zones of the pars anterior of the wallabj' pituitary. Australian J. Exper. Biol., 37, 271278.

Racadot, J. 1954. Mise en evidence de lipides complexes au niveau de I'hypophyse apres inclusion a la paraffine. Compt. rend. Soc. biol., 148, 167-170.

Rahn, H. 1939. The development of the chick pituitary with special reference to the cellular differentiation of the pars buccalis. J. Morphol., 64, 483-517.

Rahx, H., .and P.\i.\ter, B. T. 1941. A comparative histology of the bird pituitary. Anat. Rec, 79,297-312."

Rasmussen, a. T. 1933. The percentage of the different types of cells in the anterior lobe of the hypophysis in the adult human female. Am. J. Path., 9, 459-471.

Rasmussen, A. T., and G.ardner, W. J. 1940. Effects of hypophysial stalk resection on the hypophysis and hypothalamus of man. Endocrinology, 27, 219-226.

Rennels, E. G. 1951. Use of acid hematein for staining acidophils of the rat hypophysis. Anat. Rec, 3, 652-662.

Rennels, E. G. 1953. Localization of phospholipids in the rat hypophj'sis. Anat. Rec, 115, 659-671.

Rennels, E. G. 1957. Two tinctorial types of gonadotrophic cells in the rat hypophysis. Ztschr. Zellforsch., 45, 464-471.

Rennels, E. G. 1958. Effects of lactation on the neurohypophysis of the rat. Texas Rep. Biol. & Med., 16, 219-231.

Reese, J. B., Koneff, A. A., and Akimoto, M. B. 1939. Anterior pituitary changes following adrenalectomy in the rat. Anat. Rec, 75, 373-403.

Reese, J. D., Koneff, A. A., and Wainaun, P. 1943. Cytological differences between castration and thyroidectomy basophils in the rat hypophysis. In Essays in Biology, pp. 473-482. Berkeley: University of California.

Rochefort, G. J., and S.affran, M. 1957. Distribution of adrenocorticotrophic hormone in the pituitary gland. Canad. J. Biochem. Physiol., 35, 471-480.

Rodriquez, H. 1937. Systematische Untersuchungen am Hirnanhang einiger Siiugetiere iiber die Brauchbarkeit der Farbung nach BerblingerBurdorf zur Darstellung der Epithelien der Adenohypophyse. Endokrinologie, 19, 151-160.

RoKHLiNA, M. L. 1940. Effect of gonadepinephrectomy on the structure of the anterior lobe of the hypophysis. Compt. rend. Acad. Sc U.S.S.R., 27, 504-507.

RoMEis, B. 1940. Die Hypophyse. In Handbuch der mikroskopischen Anatoinie des Menschen, von Mollendorff, Ed. Vol. 6, Part 3. Berlin: Julius Springer.

Russell, D. S. 1956. Proc Roy. Soc. Med., 49, 1018-1019 (m discu,s.sion).

Russell, D. 1957. The pituitary gland (hypophysis). In Pathology, W. A. D. Anderson, Ed. St. Louis: C. V. Mosby Company.

Russfield, a. B. 1957. The endocrine system. In Analydcal Pathology, Ch. 5. New York: McGraw-Hill Book Company.

Russfield, A. B., Reiner, L., and Klaus, H. 1956. The endocrine significance of hypophyseal tumors in man. Am. J. Path., 32, 1055-1075.

S.AXEX, L., S.axen, E., Toivonen, S., and Salimaki, K. 1957. Quantitative inve.stigation on the anterior pituitary-thyroid mechanisms during frog metamorphosis. Endocrinology, 61, 35-44.

ScH.ARRER, E. 1954. Neurosecretion and anterior pituitary in the dog. Experientia, 10, 264-266.

ScHARRER, E., AND ScHARRER, B. 1940. Secretoiy cells within the hypothalamus. A. Res. Nerv. & Ment. Dis., Proc, 20, 170-194.

ScH.ARRER, E., AND ScHARRER, B. 1945. Neurosecretion. Physiol. Rev., 25, 171-181.

ScH.\RRER, E., AND ScH.'VRRER, B. 1954. Homioues produced by neurosecretory cells. Recent Progr. Hormone Res., 10, 183-240.

ScHiEBLER, T. H. 1952. Zur Histochemie des neurosekretonischen Hypothalamisch - Neurohypophysiiren Systems (II Teil) Acta Anat., 15, 593-416.

ScHLEiDT, J. 1914. Uber die hypophyse bei feminierte Mannchon and maskulierten Weibchen. Zentralbl. Physiol., 27, 1170.

ScHONE.MANN, A. 1892. Hvpophvsis und Thyreoida. Arch. path. Anat.. 129, 310-336.

Scruggs, W. M. 1939. The epithelial components of the teleost pituitary gland as identified by a standardized method of selective staining. J. Morphol., 65, 187-213.

Severinghaus, a. E. 1932. Anterior pituitary changes referable to the reproductive hormones, and the influence of the thyroid and the adrenals on genital function. In Sex and Internal Secretions, E. Allen, Ed., p. 805. Baltimore: The Williams & Wilkins Company.

Severinghaus, A. E. 1933. A cytological study of the anterior pituitary of the rat with special reference to the Golgi apparatus and to cell relationship. Anat. Rec, 57, 149-175.

Severinghaus, A. E. 1939. Anterior hypophyseal cytology in relation to the reproductive hormones. In Sex and Internal Secretions, 2nd ed., Allen, Danforth and Doisey, Eds., p. 1045-1087. Baltimore: The Williams & Wilkins Company.

SiEKEViTZ, P., AND P.\L.\DE, G. E. 1958a. A cytochemical study on the pancreas of the guinea pig. I. Isolation and enzymatic activities of cell fractions. J. Biophys. Biochem. Cytol., 4, 203-218.

SiEKEViTZ, p., .\ND Palade, G. E. 1958b. A cytochemical study of the guinea pig. II. Functional variations in the enzymatic activity of microsomes. J. Biophys. Biochem Cytol., 4, 309318.

Siperstein, E., Nichols, G. W., Jr., Griesbach, W.

E., AND Chaikoff, I. L. 1954. Cytological

changes in the rat anterior pituitary from birth

to maturity. Anat. Rec, 118, 593-620.

SJO.STRAND, F. S., AND H.ANZON, V. 1954. Ultra



structure of Golgi apparatus of exocrine cells of mouse pancreas. Exper. Cell Res., 7, 415-429.

Smelser, G. K. 1944. Differential concentration of hormones in the central and peripheral zones of the bovine anterior pituitary gland. Endocrinology, 34, 39-43.

Smith, P. E., and MacDowell, E. C. 1930. An hereditary anterior-pituitary deficiency in the mouse. Anat. Rec, 46, 249-257.

Smith, P. E., .\nd Smith, I. P. 1923a. The response of the hypophysectomized tadpole to the intraperitoneal injection of the various lobes and the colloid of the bovine hypophysis. Anat. Rec, Suppl., 25, 150-151.

Smith, P. E., and Smith, I. P. 1923b. The topographical separation in the bovine anterior hypophysis of the principle reacting with the endocrine system from that controlling general body growth with suggestions as to the cell types elaborating these encretions. Anat. Rec, Suppl., 25, 150-151.

Smith, S. W. 1951. The correspondence between hypothalamic neurosecretory material and neurohypophysial material in vertebrates. Am. J. Anat., 89^ 195-231.

Sloper, J. C. 1955. Hypothalamic neurosecretion in the dog and cat, with particular reference to the identification of neurosecretory material with posterior lobe hormone. J. Anat., 89, 301316.

Sloper, J. C. 1957. Diabetes insipidus: the contribution of histochemical and isotope technicjues to the investigation of its pathogenesis. Proc Roy. Soc Med., 51, 335-339.

Sober, H. A., and Peterson, E. A. 1958. Protein chromatography on ion exchange cellulose. Fed. Proc, 17, 1116-1126.

Speidel, C. C. 1919. Gland cells of internal secretion in the spinal cord of the skates. Carnegie Inst., Washington, No. 281, Dept. Marine Biol., 13, 1-31.

Steedman, H. F. 1950. Alcian Blue 8 GS: a new stain for mucin. Quart. J. Microsc. Sc, 91, 477479.

Steelman, S. L., Kelly, T. L., Seg.\loff, A., and Weber, G.F. 1956. Isolation of an apparently homogeneous follicle stimulating hormone. Endocrinology, 59, 256.

Stutinsky, F., Bonvallet, M., and Dell, P. 1949. Les modifications hypophysaires au cours du diabete expermentale chez le chicn. Ann. endocrinol., 10, 505-517.

Thornton, K. R. 1956. The action of ACTH and cortisone on the anterior pituitary. Proc Roy. Soc Med., 49, 1016-1018.

Timmer, R. F., and Finerty, J. C. 1956. Cytological and weight changes in the pituitary gland of the severely stressed rat. Proc Soc. Exper. Biol., 91, 420-423.

Tonnis, W., MtJLLER, W., Oswald, F., and Brilmayer, H. 1954. Kann die Rachendach-hypophyse eine vikariicrcnde Function Ausiiben. Klin. Wchnschr., 32, 912-914.

Tuchmann-Duplessis, H. 1953. Structural changes in the anterior pituitary with special reference

to the adrenal cortex. Ciba Foundation Colloquia Endocrinol., 4, 33-53 (in discussion).

Turner, C. W. 1939. The mammary glands. In Sex and Internal Secretions, 2nd ed., Allen, Danforth and Doisy, Eds., Ch. II; pp. 740-803. Baltimore: The Williams & Wilkins Company.

Tyslowitz, R. 1943. Corticotrophin obtained by ultrafiltration of pituitary extracts. Science, 98, 225-226.

VAN Dyke, H. B., Chow, B. F., Greep, R. O., and RoTHEN, A. 1942. The isolation of a protein from the pars neuralis of the ox pituitary with constant oxytocic, pressor and diuresis-inhibiting activities. J. Pharmacol. & Exper. Therap., 74, 190-209.

VoiTKEViTCH, A. A. 1937a. The role of the anterior lobe of the hypophysis in the growth and differentiation of anura. Bull. Biol, et Med. exper. U.S.S.R., 3, 255-257.

VoiTKEviCH, A. A. 1937b. The role of the anterior lobe of the hypophysis in the growth and differentiation of urodela. Bull. Biol, et Med. exper. U.S.S.R., 3, 258-259.

voN Soos, J. 1934. Vergleichende histologische. Untersuchungen liber die Topographie und die Bedeutung der basophilen Zellen der Hypophyse. Frankfurt. Ztschr. Path., 47, 82-96.

Weaver, T. A., and Bucy, P. C. 1940. The anatomical relationships of the hypophysial stem and the median eminence. Endocrinology, 27, 227-235.

Wheeler, R. S. 1943. Normal development of the pituitary in the opossum and its responses to hormonal treatments. J. MorphoL, 73, 43-87.

WiL.soN, W. D., AND EzRiN, C. 1954. Three types of chromophil cells of the adenohypophysis. Am. J. Path., 30,891-899.

WisLocKi, G.B. 1929. The hypophysis of the porpoise {Tursiops truncates). Arch. Surg., 18, 1403-1412.

WisLOCKi, G. B. 1939. Note on the hypophysis of an adult Indian elephant. Anat. Rec, 74, 321328.

Wi.sLocKi, G. B., AND Ceiling, E. M. K. 1936. The anatomv of the hypophysis of whales. Anat. Rec, 66, 17-41.

Wolfe, J. M. 1935. The normal level of the various cell ty])es in the anterior pituitaries of mature and immature female rats and further observations of cvclic changes. Anat. Rec, 61, 321-330.

Wolfe, J. M. 1941. I'lftVcts of testosterone prol)ionate on the structure of the anterior pituitaries of old male lats. Endocrinology. 29, 969974.

Wolfe, J. M. 1943. The effects of atlvancing age on the structure of the anterior hypo{)hysis and ovaries of female rats. Am. J. Anat., 72, 361383.

Wolfe, J. M., and Brown, A. D. 1942. Action of diethylstilbestrol on cytological cli^n.icf eristics of anterior pituitaries of female iai>. tuti'ilier with certain observations on the effect (if castration. Endocrinology, 31, 467-478.

Wolfe, J. M., and Cleveland, R. 1933a. Cyclic histological variations in the anterior hypoph.vsis of the albino rat. Anat. Rec, 55, 233-249.

Wolfe, J. M.. and Clevel.'\nd, R. 1933b. Pregnancy changes in anterior hypophysis of albino rat. Anat. Rec. 56, 33-45.

Wolfe, J. M., Cleveland, R., and Campbell, M. 1933. Cyclic histological variations in the anterior hypophysis of the dog. Ztschr. Zellforsch., i7, 420-452.

Wolfe, J. M., Phelps, D., and Cleveland, R. 1934. The anterior hypophysis of the rabbit during oestrus and pseudopregnancy. Am. J. Anat., 55, 363-405.

WuLZEN, R. 1914. The morphology and histology of a certain structure connected with the pars intermedia of the pituitarv bodv of the ox. Anat. Rec, 8, 403-415.

Zeckwer, L T. 1936. Thyrotropic effects of pituitaries from cretin rats. Am. J. Physiol., 117, 518-524.

Zeckwer, I. T. 1937. Morphological changes in the pituitaries of rats resulting from combined thyroidectomv and gonadectomv. Am. J. Path., 13, 985-992.

Zeckwer, I. T. 1938a. Differences between castration cells and thyroidectomy cells of the pituitary of the rat in response to the administration of estrone and thvroid extract. Am. J. Path., 14, 773-782.

Zeckwer, I. T. 1938b. The effect of combined thyroidectomy and gonadectomy on compensatory adrenal hypertrophy in unilaterally adrenalectomized rats. Am. J. Phvsiol., 123, 266271.

Zeckwer, I. T., Davison, L. W., Keller, T. B., AND LiviNGooD, S. C. 1935. The pituitary in experimental cretinism: structural changes in the pituitaries of thyroidectomized rats. Am. J. Med. Sc, 190, 145-157.

Zeiner, F. M. 1952. Pituitary gonadotrophic fluctuation during pregnancv in the rat. Anat. Rec, 113,255-267.

Historic Disclaimer - information about historic embryology pages 
Mark Hill.jpg
Pages where the terms "Historic" (textbooks, papers, people, recommendations) appear on this site, and sections within pages where this disclaimer appears, indicate that the content and scientific understanding are specific to the time of publication. This means that while some scientific descriptions are still accurate, the terminology and interpretation of the developmental mechanisms reflect the understanding at the time of original publication and those of the preceding periods, these terms, interpretations and recommendations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)
Young WC. Sex and internal secretions. (1961) 3rd Eda. Williams and Wilkins. Baltimore.
Section A Biologic Basis of Sex Cytologic and Genetic Basis of Sex | Role of Hormones in the Differentiation of Sex
Section B The Hypophysis and the Gonadotrophic Hormones in Relation to Reproduction Morphology of the Hypophysis Related to Its Function | Physiology of the Anterior Hypophysis in Relation to Reproduction
The Mammalian Testis | The Accessory Reproductive Glands of Mammals | The Mammalian Ovary | The Mammalian Female Reproductive Cycle and Its Controlling Mechanisms | Action of Estrogen and Progesterone on the Reproductive Tract of Lower Primates | The Mammary Gland and Lactation | Some Problems of the Metabolism and Mechanism of Action of Steroid Sex Hormones | Nutritional Effects on Endocrine Secretions
Section D Biology of Sperm and Ova, Fertilization, Implantation, the Placenta, and Pregnancy Biology of Spermatozoa | Biology of Eggs and Implantation | Histochemistry and Electron Microscopy of the Placenta | Gestation
Section E Physiology of Reproduction in Submammalian Vertebrates Endocrinology of Reproduction in Cold-blooded Vertebrates | Endocrinology of Reproduction in Birds
Section F Hormonal Regulation of Reproductive Behavior The Hormones and Mating Behavior | Gonadal Hormones and Social Behavior in Infrahuman Vertebrates | Gonadal Hormones and Parental Behavior in Birds and Infrahuman Mammals | Sex Hormones and Other Variables in Human Eroticism | The Ontogenesis of Sexual Behavior in Man | Cultural Determinants of Sexual Behavior

Reference: Young WC. Sex and internal secretions. (1961) 3rd Eda. Williams and Wilkins. Baltimore.

Cite this page: Hill, M.A. (2020, August 6) Embryology Book - Sex and internal secretions (1961) 3. Retrieved from

What Links Here?
© Dr Mark Hill 2020, UNSW Embryology ISBN: 978 0 7334 2609 4 - UNSW CRICOS Provider Code No. 00098G