THE MAMMALIAN HARDERIAN GLAND: A COMPARISON OF THE NORMAL AND DIABETIC CHINESE HAMSTER BASED ON FINE STRUCTURE AND ENDOCRINE EVIDENCE A Thesis Presented to the Faculty of the Department of Biology East Carolina University In Partial Fulfillment of the Requirements for the Degree Master of Science in Biology by TODD LOWELL SOBOL July 1983 f. T. ÏOTN»» LIBîSAStY THE MAMMALIAN HARDERIAN GLAND: A COMPARISON OF THE NORMAL AND DIABETIC CHINESE HAMSTER BASED ON FINE STRUCTURE AND ENDOCRINE EVIDENCE by Todd Lowell Sobol Approved By: THESIS DIRECTOR Gerhard Vi. Kalnius, Ph.D. COMMITTEE ÎÎEMBER Carlton Heckrotte, Ph.D. COMMITTEE ÎEMBER Evelyn M. McNeill, Ph.D. COI-nUTTEE MEMBER ~-J Everett C. Simpson,^Ph.D. CHAIRMAN OF THE DEPARTMENT OF BIOLOQJ ^ ^ Charles E. Bland, Ph.D. DEAN OF THE GRADUATE SCHOOL 674315 ABSTRACT Todd L. Sobol. The Mammalian Harderian Gland: A Comparison of the Normal and Diabetic Chinese Hamster Based on Fine Structure and Endocrine Evidence. (Under the direction of Dr. Gerhard Kalmus) Department of Biology, East Carolina University. July 1983. The Harderian gland is located in the orbit, just behind the eyeball, in animals possessing a third eyelid. The Harderian gland is species-specific in both structure and function. In hamsters, the gland is composed almost exclusively of secretory tubules and alveoli and has therefore been classified as compound tubuloalveolar. Both endocrine and exocrine functions have been proposed for the Harderian gland in rodents. Among these suggestions are: (1) that the Harderian gland serves as a source of lubrication for the eye through lipid secretion; (2) that it may influence pineal and/or reproductive function through actions as an extraretinal photoreceptor; and, (3) that it may serve a communicative function through pheromone secretion. The Chinese hamster (Cricetulus griseusi spontaneously develops diabetes and has been used to study the pineal-gonadal axis. Infertility is a complication of the diabetic state in this animal. The alterations in Harderian gland fine structure in the diabetic Chinese hamsters as well as the relationship of the Harderian gland to the retinal-pineal-gonadal chain are described in this paper. In the present study, the diabetic Chinese hamster Harderian gland weight was decreased in comparison to controls. The Harderian glands of diabetic animals were found to have a decrease in alveolar size, the number of cells per alveolus, cell radius and the number of type II cells per alveolus. No type II cells were observed in three diabetic animals. Structural changes seen in the Harderian glands of diabetic animals may indicate possible disturbances in secretory state. Concentric lamellae vary in extent and configuration as well as location and association with other organelles within the secretory cells of the Harderian gland. This is suggestive of two possible functions for concentric lamellae; (1) that the material within the structures may be lipids and proteins necessary for membrane formation; and/or, (2) that the structures may have secretory capacity. Diabetic animals v;ere observed of have increased relative concentric lamellae area to total cytoplasm as compared with controls. Cells with cleft structures were observed in the Harderian glands of diabetic animals to a far greater extent than controls. Cleft structures appear to be related to the formation of vacuoles within the cytoplasm of type II cells. The relationship between structural changes seen in the Harderian gland of the diabetic Chinese hamsters and endocrine gland weights has been described in this paper. Harderian gland weight and thyroid gland weight were lower in diabetics versus controls. Pineal gland v/eight, adrenal weight and gonadal weight were higher in diabetics versus controls. The total pituitary weights were similar in both diabetics and controls while the posterior pituitary weight was lower in diabetics versus controls and the anterior pituitary v/eight vías higher in diabetics versus controls. DEDICATION It gives me great pride to dedicate my work to my parents, thank them for their dedication to me throughout the years. ACKNOWLEDGMENTS I am deeply grateful to Dr. Gerhard Kalmus not only for his help and guidance but also for his friendship. His commitment toward achieving excellence in his profession is a quality to emulate. I want to thank Dr. Evelyn McNeill for her advice and instruction. I would also like to acknov/ledge her contribution of the Chinese hamster Harderian gland tissue and endocrine data. I want to thank the Upjohn Company for making the Chinese hamsters available. I sincerely appreciate the review and recommendations made by Dr. Everett Simpson and Dr. Carlton Heckrotte. I want to express my gratitude to the faculty and staff of the Department of Biology and acknowledge the funding of this thesis through a grant from the Department of Biology. I especially want to thank Laddie Crisp for his expert technical assistance in the microscopy and photo labs. Finally, due credit must also be extended to Dr. Albert Warshauer for allowing me "priority” usage of his Compucorp 600 Series word processor. TABLE OF CONTENTS PAGE LIST OF TABLES x LIST OF GRAPHS xi LIST IIOIF FIGURES xiiLIST OF PLATES xiiiLIST OF.GArBoBRsEsVIATIONS ixINTRODUCTION AND REVIEW OF LITERATUREI.IntroductionA. Discovered by Harder in 1694 1B. Definition 2C. Location 2D. Occurrence in mammals 2II.PhylogenyA. Variability in Mammals 4B. Migration vs Two Gland Theory 4C. Evolution — Homologous Way vs Analogous Way 5Anatomy A. Excretory Duct 7 B. Lobes in rodents 8 IV.Histology A. Architecture (vs Lacrimal) 10 B. Alveoli 1. Secretory Cells 11 2. Myoepithelial Cells 13 3. Cell Proliferation 13 4. Histochemical Properties 14 5. Sexual Dimorphism 15 a. Cell Types 15 b. Porphyrin 16 c. Ultrastructure 16 d. Manipulation Experiments 17 1. Castration 2. Androgens 3. Blinding 4. Pinealectomy 5. Ovariectomy e. Mast Cells 19 f. Concentrations of Metals 21 C. Innervation 22 V.Secretory Products A. Secretory Mechanism 24 B. Lipid 25 C. Porphyrin 1. Localization 26 2. Biosynthesis 26 3. Action 27 4. Content and Concentrations 28 5. Manipulation Experiments 30 D. Melatonin 1. Localization 35 2. Biosynthesis 35 3. Circadian Rhythm 36 4. Action 36 5. Manipulation Experiments 38 E. Taurine 39 VI.Function A. Endocrine and Exocrine Evidence 4l 1. Manipulation Experiments 2. Cytology B. Lubrication 42 C. Antibiotic Effect 42 D. Pheromones 42 E. Thermoregulation 43 F. Retinal-Pineal-Gonadal axis 1. Estradiol Target 44 2. Extraretinal Photoreceptor 44 3. Pineal Melatonin Rhythm 46 4. Porphyrin Production 46 5. Manipulation Experiments 47 VII. Diabetic Profile 48 MATERIALS AND METHODS I. Tissue Preparation 50 II. Observations Utilizing Light Microscopy 52 III. Morphometric Analysis of Electron Microscopy 52 IV. Statistics 53 RESULTS I. Characteristics at Sacrifice 58 II. Light Microscopy Observations 59 III. Electron Microscopy Morphometric Observations 59 A. Electron Microscopy Histological Observations 60 B. Concentric Lamellae 61 C. Cleft Structures 62 D. Mast Cells 63 DISCUSSION 122 I, Gross Anatomy 122 II. Secretory State 126 III. Type II Cells 126 IV. Concentric Lamellae 127 V. Cleft Structures 129 IV. Mast Cells 130 SUMMARY 131 APPENDIX A Earle's BSS-man Balanced Salt Solution 132 APPENDIX B Progressive Method for Hematoxylin and Eosin Staining...133 APPENDIX C Ehrlich Hematoxylin 134 APPENDIX D Araldite 6005 135 LITERATURE CITED 136 LIST OF TABLES TABLE PAGE 1. Summary of the occurrence of the mammalian Harderian Gland 3 2. Changes in concentrations of various metals in the Harderian glands of hamsters in response to pinealectomy castration or blinding as a function of sex 21 3. Diabetic Animal Characteristics at Sacrifice 64 4. Control Animal Characteristics at Sacrifice 65 5. Summary of Diabetic and Control Animal Characterisitcs at Sacrifice 68 6. Diabetic Animal Light Microscopy Observations 69 7. Control Animal Light Microscopy Observations 70 8. Summary of Diabetic and Control Animal Light Microscopy Observations 71 9. Diabetic and Control Animal Electron Microscopy Observations Type I Cells 72 10. Diabetic and Control Animal Electron Microscopy Observations Type II Cells 73 11. Diabetic and Control Animal Electron Microscopy Observations Cells with Cleft Structures 74 12. Diabetic Animal AH20-21 Electron Microscopy Observations 75 13. Summary of Diabetic and Control Animal Electron Microscopy Observations 76 LIST OF GRAPHS GRAPH PAGE 1. Blood Glucose Levels: Diabetic and Control Hamsters 78 2. Harderian Weight: Diabetic and Control Hamsters 78 3. Adrenal Gland Weight: Diabetic and Control Hamsters 79 4. Thyroid Gland Weight: Diabetic and Control Hamsters 79 5. Anatomical Changes Seen in Diabetic Hamsters Expressed as a Percent of the Control Values 80 6. Alveolar Size; Diabetic and Control Hamsters 8l 7. Number of Cells per Alveolus: Diabetic and Control Hamsters..8l 8. Cell Radius: Diabetic and Control Hamsters 82 9. Number of Type II Cells per Alveolus: Diabetic and Control Hamsters 82 10. Structural Decrease Seen in Diabetic Hamsters Expressed as a Percent of the Control Values 83 11. Relative Vacuole Area of Type I Cells, Type II Cells and Cells with Cleft Structrues 84 12. Relative Mitochondria Area of Type I Cells, Type II Cells and Cells with Cleft Structures 85 13. Relative Concentric Lamellae Area of Type I Cells, Type II Cells and Cells with Cleft Structures 86 14. Linear Regression of Relative Mtochondria Area Versus Relative Vacuole Area 87 LIST OF FIGURES FIGURE PAG] 1. Schematic representation of the eye and its accessory glands in mammals 3 2. The course of phylogenetic migration 5 3. Two possibilities for phylogenetic development of the Harderian and lacrimal glands 6 4. Diagram of the Harderian gland of a gerbil seen from the nasal side 9 5. Schematic illustration of a simple branched gland and a compound tubuloacinar gland 10 6. Diagrams showing the secretory process of lipid droplets 24 7. The synthetic pathway of porphyrins 28 8. Porphyrin in the Harderian glands of intact female mice during the estrous cycle 29 9. Relative Harderian gland weight and average porphyrin accretions of the female golden hamster over the estrous cycle 30 10. Influence of the extirpation of adrenal and/or ovary, sex hormone replacement and of pregnancy on the Harderian gland and porphyrin content in female mice 32 11. Influence of the extirpation of adrenal and/or testis and of sex hormone replacement on the weight of the Harderian gland and on the content of porphyrins in male mice 33 12. Taurine concentrations in the Harderian glands and retinas of mice during postnatal development 40 LIST OF PLATES PLATE PAGE 1. Secretory Alveoli at the Light Microscopic Level 54 2. Grid Placed Over an Electron Micrograph Demonstrating Morphometric Counting Procedure 56 3. Light Micrograph showing Compound Tubuloalveolar Structure 88 4. Secretory Cells of Diabetic Animal AH20-21 90 5. Secretory Cells of Diabetic Animal AC19-24 92 6. High Magnification Micrograph of a Secretory Cell of Diabetic Animal AC19-24 94 7. Secrtory Alveolus at the TEM Level 96 8. Type I and Type II Cells 98 9. Cells with Cleft Structures in Control AVI3-63 100 10. Cells with Cleft Structures in Diabetic AH20-21 102 11. Cells with Cleft Structures in Diabetic KI01-30 104 12. Sequence of the Association of Concentric Lamellae with Mitochondria and Vacuoles 106 13. Concentric Lamellae of Varying Configurations 108 14. Cleft Structures in Association with Vacuoles 110 15. Exocytosis at the Lumen by a Secretory Cell 112 16. Porphyrin Pigment Concretion 114 17. Exocytotic Vesicles at the Basal End of a Secretory Cell 116 18. Cytoplasmic Extension of a Myoepitelial Cell 118 19. Fenestrated Capillary in the Interstitial Area 120 LIST OF ABBREVIATIONS Â angstrum ACTH adrenocorticotropin ALA synthase (T-aminolevulinate synthase aîiL acetylraethoxytryptophol CAMP cyclic adenosine monophosphate cm centimeter db autosomal recessive mutation diabetes dl deciliter FSH follicle stimulating hormone s gram GnRH gonadotropin releasing hormone HIOMT hydroxyindole-o-methyltransferase kV kilovolt LD 12 hours light-12 hours dark LH luteinizing hormone LL constant light LM light microscopy Î-ÎA methoxyindoleacetic acid mg milligram MT methoxytryptamine MTL 5-methoxytryptophol MI'i methoxytryptophan nmole nanomole 0s04 osmium tetroxide PAN perchloric acid-naphthoquinone PAS periodic acid-Schiff SEM standard error of the mean STH somatotrophin TEM transmission electron microscopy TSH thyroid stimulating hormone ug microgram Mm micrometer (micron) X magnification tines THE MAMMALIAN HARDERIAN GLAND INTRODUCTION It is generally accepted that the Harderian gland was first described in the orbit of Cervus and Dama, i.e., deer and fallow deer, by Harder (1694). He coined the name "Glándula nova lachrymalis", because he believed that he had found a new gland near the nictitans gland which was then called Glándula lachrymalis. At that time, the lacrimal gland itself was called Glándula innomonata. After its discovery by Harder, many workers believed they had found the same gland in several animals including mammals, birds, reptiles and amphibians. It was not until the late 19th century that the distinction between the Harderian and the nictitans gland became recognized and histological studies on these glands began. In I89O, Peters divided the ocular glands associated with the third eyelid into two types, based on the width of the lumina. Lowenthal (1896) proposed the following criteria to distinguish between these two glands: 1) the Harderian gland manifests a single duct while the nictitans gland manifests several ducts; 2) the Harderian gland is well separated from the cartilage of the third eyelid while the nictitans is in contact with it; and, 3) the glandular lumina of the Harderian gland are wider than those of the nictitans gland. Based on these criteria, the Harderian gland and the nictitans gland are now called Glándula palpebrae tertiae profunda and superficialis, respectively, in veterinary science. Today, Lowenthal's criteria are not satisfactory to define the Harderian gland. Of the criteria, only the last is significant 2 while the others reflect secondary features of the Harderian gland in sone manmals (Sakai, 1981). DEFINITION - It is the histological rather than the gross anatomical features that characterize the mammalian Harderian gland. Sakai (1981) offers the following histological definition; "The mammalian Harderian gland [consists of] those ocular glands that have tubuloalveolar endpieces (tubular alveoli) and secrete lipid by a merocrine mechanism." LOCATION - The locations of the gland and the duct cannot be the criteria for characterizing the Harderian gland because of their variability among mammals. The Harderian gland is generally associated with the third eyelid (the nictitans or nictitating membrane) in the inner canthus of the eye as opposed to the lacrimal gland which pours its secretions near the outer canthus (Fig. 1). However, a gland associated with the third eyelid is not always the Harderian gland, and the Harderian gland is not always associated with the nictitating membrane (Sakai, I98I). OCCURRENCE - The Harderian gland is found within the orbit of all reptiles and birds, some amphibians and those mammals possessing nictitating membranes. Table 1 presents a summary of the occurrence of the Harderian gland in mammals. The Harderian gland is species-specific in both structure and function (Thiessen and Kittrell, 198O). This paper will profile the rodent Hcirderian gland using the hamster model v;henever possible. 3 Figure 1. Schematic representation of the eye and its accessory glands in mammals. The Harderian gland (H) is associated with the third eyelid while the lacrimal gland (L) pours its secretion near the outer canthus.® Table 1. Summary of the occurrence of the mammalian Harderian gland. Monotremata; ? Cetacea: + Maruspialia: + Carnivora: Insectívora: -H-/- (Fissipedia) - Dermoptera: p (Pinnipedia) + Chiroptera: - Tubulidentata: p Primates: - Proboscidea: P Edentata: + Hyracoidea: P Pholidota: p Siren ia: + Lagomorpha: -H- Perissodactyla: - Roden tia: 4f Artiodactyla: + ++ indicates the ensured presence (examined both histologically and histochemically); + probable presence (examined either histologically or histochemically); - absence at least in some genera; ? not yet reported.“ *From T. Sakai. 1981. The Mammalian Harderian Gland: Morphology, Biochemistry, Function and Phylogeny. Arch. Hist. Jap. iü: 299-333. 4 PHYLOGENY VARIABILITY IN MAMMALS - The higher up the evolutionary scale one goes, the less sebaceous-like are the Harderian glands and the more they resemble true lacrimal glands. From the standpoint of evolution, the Harderian glands and lacrimal glands appeared when the vertebrates left the water (Weaker, 1981). The mammalian Harderian gland appears to have evolved from the lacrimal gland of primitive ancestral mammals from which all eutherias and marsupials have descended. The gland occurs in primitive as well as highly developed mammals and may have presumably been lost secondarily in chiroptera, primates, terrestrial carnivores and perissodactyla (Sakai, 1981). MIGRATION VS TWO GLAND THEORY - The location of ocular glands is quite variable among vertebrates. Comparative anatomical studies have proposed two theories to explain the variations in location. The migration theory regards all accessory glands of the eye as the lacrimal gland and denies the existence of the Harderian gland. Figure 2 shows the proposed course of phylogenetic migration of the ocular gland. 5 Figure 2. The course of phylogenetic migration, a. Position in amphibians; b. position in reptiles and birds; c. position in man. From T. Sakai. 1981. The Mammalian Harderian Gland: Morphology, Biochemistry, Function and Phylogeny. Arch. Histol. Jap. M.: 299-333. Though all ocular glands in amphibians have a similar histological architecture, the migration theory is unacceptable since it fails to explain the occurrence of two histologically distinct glands in reptiles, birds and mammals. A second theory, the two gland theory, is based on the existence of two ocular glands, i.e., the lacrimal gland in the outer canthus and the Harderian gland in the inner canthus. This proposal is generally acceptable based on histological descriptions of the lacrimal and Harderian glands in addition to other ocular glands. The nictitans gland and Krause's gland are disseminated lacrimal glands vihile Meibom's and Moll's glands are modified dermal glands found in mammals (Sakai, 1981). HOMOLOGOUS WAY VS ANALOGOUS WAY - Assuming the two glands theory correctly describes the phylogeny of the ocular glands, the existence of the two histologically distinct glands in various vertebrates can be explained as having evolved in either a homologous vjay or in an analogous way (Fig. 3). The analogous way seems to have the greatest support. Mo similarity in the histological structure of the Harderian gland could be found among the various classes of vertebrates. The developmental process of the Harderian gland is similar to that of the lacrimal gland. Serous glands similar to the lacrimal gland sometimes occur in association with the duct of the Harderian gland or within the gland. The mammalian Harderian gland secretes lipid and differs from the Harderian gland found in lower classes. Thus it is reasonable to infer that the Harderian gland evolved from the lacrimal gland (Sakai, I98I). HG UG Va Vb ?-> Vc J'- (3) Figure 3. Two possibilities for phylogenetic development of the Harderian (HG) and lacrimal gland (LG), a. The homologous way; b. The analogous way. Va, Vb, Vc are various vertebrates. From T. Sakai. 1981. The Mammalian Harderian Gland: Morphology, Biochemistry, Function and Phylogeny. Arch. Histol. Jap. M.: 299-333- 7 GROSS ANATOMY EXCRETORY DUCT - The excretory duct of the Harderian gland varies in number. Location of their opening in relation to the nictitating membrane suggests a functional variation or an independence of the gland from the third eyelid. Since the ductal opening is thought to represent the embryonic origin of a gland and since serous acini are associated vjith the excretory duct of the Harderian gland, a close phylogenetic relationship is suggested between the Harderian gland and the lacrimal gland (Sakai, 1981). The excretory duct of the Harderian gland is usually single in number though variability exists. The deer possesses a single-trunk excretory duct which bifurcates to make two openings. VIhales, which have no third eyelid, have multiple excretory ducts (Sakai, 1981). Several large excretory ducts were observed near the medial border of the gland in mice (Strum and Shear, 1982). Precise location of the opening varies from species to species and seems to be independent of the third eyelid. Ducts open on the inner (temporal) side of the third eyelid in rabbits and pigs. In many rodents and in hedgehogs, the duct opens on the outer (nasal) side of the third eyelid suggesting a function other than lubrication for the eye or third eyelid (Sakai, 1981). Structure of the excretory duct has not been well documented. Very small glandular lobules are associated with and around the excretory ducts in the rabbit Harderian gland. In gerbils, many serous acini are noted in association with and around the excretory duct. The excretory duct branches extensively within the epithelium (Sakai, 1981). The interlobular duct system which empties into one major excretory duct in the armadillo 8 has lobules of mucus and lipid-secreting acini in close association (Weaker, 1981). Mice do not appear to have such glandular lobules in close association (Strum and Shear, 1982). Serous glands which are histologically similar to lacrimal glands are associated with the Harderian gland excretory duct. This suggests that the Harderian and lacrimal glands have the same phylogenetic origin (Sakai, 1981). LOBES IN RODENTS - The Harderian gland is well developed in rodents (hamsters, rats, mice, gerbils and guinea-pigs). It is located around the posterior half of the eyeball and occupies a considerable part of the orbit (Fig. 4). The surface of the gland is smooth and macroscopically, no adventitia are present. This is attributed to a well-developed orbital venous sinus which surrounds the orbital contents including the Harderian gland. Endothelium of the orbital venous sinus covers the surface of the gland and other orbital contents and serves as the main venous drain (Sakai, 1981). The gland is divided into lobules separated from each other by a small amount of connective tissue containing blood vessels and nerve fibers (Hoffman, 1971; Bucana and Nadakavukaren, 1972; Hoffman and Jones, 1981). Loose connective tissue connects the gland to the base of the nictitating membrane and suspends the gland in the orbital sinus by a glandular pedicle. The single excretory duct originates at the hilus through the loose pedicular connective tissue and opens on the external nasal side of the third eyelid (Sakai, 198I). The Harderian gland of the female golden hamster is speckled with reddish-brown or dark amber pigment granules due to porphyrin pigments (Clabough and Norvell, 1974) while the male gland is pale yellow in color 9 (Hoffman and Jones, 198I). In mice, the Harderian gland is speckled with dark brown pigments due to the existence of porphyrin. The gland in gerbils is dark grey due to a large number of melanocytes in the interstices (Sakai, 1981). Figure 4. Diagram of the Harderian gland of a gerbil seen from the nasal side. The asterisk shows the association with a rudimental nictitating membrane. From T. Sakai. 1981. The Mammalian Harderian Gland: Morphology, Biochemistry, Function and Phylogeny. Arch. Histol. Jap. M.: 299-333. 10 HISTOLOGY ARCHITECTURE - The glandular architecture of the Harderian gland has been classified by some authors as compound tubuloalveolar (tubuloacinar) (Norvell and Clabough, 1972; Bucana and Madakavukaren, 1972, 1973; Weaker, 1981) while others have considered the gland to be simple branched-tubuloalveolar (Sakai, 1981) (Fig. 5). Compound tubuloalveolar exocrine glands generally have irregularly branched tubules with numerous saccular outgrowths from the wall and at the blind ends (Plate 3). In simple branched-tubuloalveolar glands, the duct does not branch. Several acini may be arranged along a duct (Bloom and Fawcett, 1968). Secretory Secretory tubule acinus Figure 5. (A) A schematic illustration of a simple branched gland. (B) A compound tubuloacinar gland. From Vi.J. Krause and V.J. Cutts. 1981. Concise Text of Histology. 1st ed. Williams and VJilkins, Baltimore. 32-33. 11 Sakai and Yohro (I98I), studying the gerbil Harderian gland, followed the continuity of tubuloalveoli in serial celloidln sections and examined their arrangement within the gland. The gland contains several lobes with each lobe having both rather straight central and very tortuous peripheral tubuloalveoli. The central tubuloalveoli reach the hilus where the single excretory duct begins. The peripheral tubuloalveoli originate from the central ones and appear to be forked and twisted. VJeaker (1981) described the duct system in the nine-banded armadillo. The acini drain into an intralobular duct system lined with squamous to low cuboidal epithelial cells. These ducts emptied into interlobular ducts composed of simple columnar epithelium with occasional basal cells. The interlobular ducts form a single large excretory duct terminating at the base of the nictitating membrane. The duct, located deep in the parenchyma of the gland, coursed the entire length of the gland, receiving tributaries from the more peripheral areas of the organ. Strum and Shear (1982) observed a few small ducts in the mouse. These ducts, lined by cuboidal epithelial cells, were thought to be initial ducts to carry the secretory material from the more distal (alveolar) portions of the gland. Several large ducts were observed near the medial border of the gland. They were lined by stratified epithelium and emptied into the concave surface of the nictitating membrane. ALVEOLI SECRETORY CELLS - Each alveolus contains one or two types of secretory cells. Among the rodents, two types of glandular cells have been found in male golden hamsters, rats, house mice and field mice. Only one 12 type appears in female golden hamsters, gerbils, rabbits and guinea pigs (Bucana and Nadakavukaren, 1972; Sakai, I98I). The alveolus is formed by a single layer of secretory cells. The simple columnar cells usually have a single basally located nucleus and numerous vacuoles in the cytoplasm. The apices of secretory cells are often domed into the alveolar lumina (Bucana and Nadakavukaren, 1972; Payne et al.. 1977). The appearance of secretory cells exhibit little variability from species to species (Sakai, 1981). Based on the overall electron density, Bucana and Nadakavukaren (1972) referred to two types of cells in the secretory epithelium of the Harderian gland of the golden hamster present in both male and females as "light” and "dark." "The density difference between these two cell types is due to an abundance of ribosomes in the cytoplasm of the dark cells. In addition, dark cells have more numerous and larger vacuoles and increased numbers of mitochondria in comparison to the lighter cells. Each secretory cell has a basally located nucleus with two or more nucleoli. Golgi apparatus, scattered rough endoplasmic reticulum, and a prominent smooth endoplasmic reticulum are also found in both cell types. In all cases, vacuoles are bounded by a single membrane." There is no apparent difference in the smooth endoplasmic reticulum of dark and light cells in the male. In females, the cytoplasm of dark cells is characterized by the presence of dilated smooth endoplasmic reticulum and vacuoles that are generally larger than those found in light cells. The light cells from the females have numerous membrane structures arranged in concentric lamellar formations or in the form of Golgi with proliferation of cisternae. The membranes are generally smooth and are frequently associated with vacuoles. 13 The secretory cells of the male are characterized by the presence of membrane-bound clusters of cylindrical tubules that are randomly distributed throughout the cytoplasm. The diameter of these tubules is approximately 400 There is considerable variation in the number of tubules in a cluster. Each tubule appears to be composed of microtubular subunits. Clusters of tubules are present in both light and dark secretory cells and may be observed in close proximity to vacuoles, nuclei and mitochondria. Isolated tubules can also be found in the cytoplasm. Neither isolated tubules nor clusters of tubules are found in the female. Microvilli are found at the apices of the secretory cells. There is interdigitation between cells. The interdigitations may be very pronounced and appear as enlarged microvilli of the lateral sides of, as well as between, the secretory cells and cytoplasmic extensions of the myoepithelial cells (Bucana and Madakavukaren, 1972). MYOEPITHELIAL CELLS - Myoepithelial cells lie between the basement membrane and the secretory epithelium of the gland. The nuclei of the myoepithelial cells lie parallel to and close to the basement membrane. Cytoplasmic extensions, fine filaments in the cytoplasm and a long or pyknotic nucleus differentiate myoepithelial cells from the secretory epithelium. Pinocytotic vesicles are frequently found in these cells (Bucana and Madakavukaren, 1972). CELL PROLIFERATION - Several studies have noted that the Harderian gland has a relatively low rate of mitiotic activity and maintenance of gland weight (Hoffman and Jones, 1981; Hoffman and Alieva, 1982). Bucana and Madakavukaren (1972) failed to find mitotic activity after 3 weeks of age. In no case following steroid or protein hormone manipulation has any significant change in proliferation been noted (Hoffman and Alieva, 1982). 14 Harderian glands are selectively stimulated by cholera toxin showing an intense proliferation in alveolar cells. Of a variety of tissues exposed to cholera toxin only the Harderian glands and uteri exhibited obvious effects, i.e., increased weight and cellular proliferation. Further, there seems to be little correlation between porphyrins within the Harderian glands and proliferative activity (Hoffman and Alieva, 1982). Cell proliferation does not appear to require the participation of cAMP. DNA synthesis is increased without involvement of cAMP. "Increased mitotic activity is not due to stress, or to adrenocorticotropin (ACTH) or other hormones from the pituitary or its target tissues per se, nor is it a consequence of a release from an earlier mitotic blockade. V/ithout data to the contrary, one could speculate that the Harderian glands may well contain high concentrations of GM1 gangliosides on the cell surfaces. Studies have shown that exogenously applied GM1 ganglioside is functionally integrated into the surface membrane of intact cells and the suggestion has been made that GM1 ganglioside may possibly serve as the natural choleragen receptor. Further observations of GM1 gangliosides suggest a role in or an association with mitotic activity" (Hoffman and Alieva, 1982). HISTOCHEMICAL PROPERTIES - The cytoplasm of the secretory cells contains numerous acidophilic granules when stained with haematoxylin and eosin. The cells did not stain with either periodic acid-Schiff (PAS) or mucicarmine. In frozen sections stained with osmium tetroxide (0s04), the granules were colorless while they stained blue with the perchloric acid-naphthoquinone (PAN) technique. These results indicate the presence of cholesterol and the absence of unsaturated fats. Biochemical assay of the Harderian glands from male and female armadillos indicate the presence of 39.9nig^ (mg per 100ml of homogenate) and 58.3mgí total cholesterol. 15 respectively (Weaker, I98I). SEXUAL DIMORPHISM - The Harder!an gland of the golden hamster exhibits a sexual dimorphism in both structure and secretory activity. The sexual dimorphism of the golden hamster Harderian gland is characterized by: (1) cell types (two in the male, one in the female); (2) ultrastructural features (polytubular complexes present in the male but absent in the female); (3) storage of porphyrin (up to 1000 times greater in the female than in the male); (4) the number of mast cells present in the interstitial tissue and connective tissue capsule of the gland (an average of 40 times greater in the female); and, (5) concentrations of metals (Hoffman and Jones, 1981; Payne al., 1982). Cell Types - Females are characterized by a single cell type, i.e., type I, which are darker, possess grossly visible spots of porphyrin pigment, and synthesize and store large quantities of porphyrins. Males possess two cell types, i.e., type I and II, which are pale yellow, possess little porphyrin and tend to be heavier than females (Payne et al., 1977;Lin and Nadakavukaren, 1979; Sun and Nadakavukaren, I98O; Hoffman and Jones, 1981). Ultrastructural studies by Bucana and Nadakavukaren (1972) showed the presence of membrane-bounded clusters of cylindrical tubules in the acinar cells of the gland from adult male golden hamsters. In female, golden hamsters, membranous structures are arranged in concentric lamellae or in the form of Golgi with proliferated cisternae. Whether two distinct cell types aire formed or whether types I and II represent different stages in a holocrine secretory cycle is unclear. Payne (1977) described prepubertal male golden hamster Harderian glands as resembling the female pattern including high porphyrin content. This confirmed an earlier study by Bucana and Nadakavukaren (1973) of the 16 postnatal development of the golden hamster Harderian gland. Based on size distribution of vacuoles, two types of alveolar cells can be recognized in male hamsters 4 weeks and older. Differentiation of the alveolar cells into light and dark types takes place during the third week of postnatal development. Porphyrin - The Harderian glands of 2 week old animals of both sexes contain a black-brown pigment. The appearance of pigment is accompanied by a proliferation of membranes. In the female the amount of pigment increases with increasing age until 4 weeks of age and then remains at a relatively stable level. In the male, no pigment was found after 4 weeks of age. The synthesis or storage of porphyrin in the Harderian glands of adult male hamsters appears to be inhibited by increased levels of gonadotropins (Hoffman, 1971). Thus it may be suggested that the decrease in porphyrin may be due to increasing levels of testosterone and/or gonadotropins as the male sexually matures (Bucana and Nadakavukaren, 1973). Ultrastructure - Membrane-bound "juxtanuclear structures” are present in the alveolar cells of both sexes until 3 weeks of age. These structures apparently serve different functions in males and females. In the young male, the cylindrical tubules appear to be assembled in the juxtanuclear structures. The disappearance of pigment granules from the alveolar cells of maturing male hamsters coincides with the appearance of tubular clusters in the cytoplasm. This, along with the observation that tubules are not found in the alveolar cells of females of any age (Bucana and Nadakavukaren, 1972), suggests that these tubules may contain an enzyme or enzyme system necessary for the breakdown of pigment molecules. In the young female hamsters the juxtanuclear structures may serve as the site of membrane formation. An increase in electron-dense material in the 17 juxtanuclear structures may be lipids and proteins necessary for membrane formation and corresponds with the appearance of membranes. Pigment formation in both sexes appears to coincide with an increase in cytoplasmic membranes. Manipulation Experiments - Manipulation of adult hamsters has shown that type II cells in the male can be converted into type I cells and that type I cells in the female can be converted to type II cells. Hoffman (1971) described the histology of the conversion of the male type Harderian gland into the female type. In male hamsters, type II cells appear to be transformed to type I cells by accumulating more lipid, thus giving rise to larger droplets with less uniformity of size. Using blinded, sham-pinealectomized females, Clabough and Norvell (1974) suggested that the apparent stages in the conversion of type I to type II cells indicated a transformation of type I to type II cells. Castration of hamsters results in conversion of the male to the female type of Harderian gland within 20-30 days. Males experience a significant reduction in the weight of the gland, disappearance of type II cells and the presence of solid porphyrin accretions (Payne et al.. 1977). Hoffman (1971) found that the administration of androgens, i.e., testosterone propionate, to castrated golden hamsters maintains the presence of type II cells within the Harderian gland and suppresses the production of porphyrin. Payne and co-workers (1977) tested five naturally occurring androgens and found that all maintained normal frequencies of type II cells and all except dehydroepiandrosterone prevented deposition of porphyrin. Dehydroepiandrosterone was unable to suppress increases in the weight of the body and the pituitary gland after castration and could not maintain the weight of the seminal vesicles or sexual behavior. 18 Testosterone, dihydrotestosterone, and androstenedione maintained glandular weight and sexual behavior. Clabough and Morvell (1973) treated female golden hamsters with testosterone propionate and found that the Harderian gland tended to exhibit the male type. Thus androgens appear to be responsible for maintaining the activity and morphology of the Harderian gland (Payne et. aj^., 1977). Some authors have suggested that steroidal effects are indirect and that gonadotrophins influence the gland more directly (Hoffman, 1971; Clabough and Norvell, 1973). The experiments of Shirama and co-workers (1981) (described on page 31) indicate that in mice, progesterone stimulates porphyrin levels and that both testosterone and some adrenal substance, possibly an androgen, reduce Harderian gland porphyrin. These results seem to show that these steroids act directly on the mouse Harderian gland, since it is thought that progesterone is unable to affect plasma gonadotropins and prolactin levels in oophorectomized animals. Clabough and Norvell (1974) concluded that in both male and female hamsters, the pineal gland in the blinded animal is responsible for the presence of type II cells and for the decrease in the formation and storage of porphyrin granules in the Harderian gland. The effect of castration is greatly reduced or abolished if the animals are also blinded. However, conversion to the female type will still occur in blinded animals if the Harderian glands of both orbits are unilaterally enucleated. In the blind-castrated animals, the pineal gland seems to be responsible for preventing the conversion into the female type. Responses to blinding were not observed if the animal had also been pinealectomized. Blinding leads to elevated pineal antigonadotropic activity. Thus it was suggested that following castration, transformation of Harderian glands of the male type 19 into the female type is probably a consequence of elevated levels of gonadotrophic hormone. Blinding, by enhancing pineal antigonadotropic activity, prevents this transformation (Clabough and Norvell, 1973; 1974). After eleven weeks, female golden hamsters that were blinded and sham-pinealectomized had significantly reduced uterine weights and had Harderian glands with diminished porphyrin and various concentrations of type II cells like those seen in the male. Neither loss of uterine weight nor alteration toward the male type of Harderian gland occurred in female hamsters that were blinded and pinealectomized (Clabough and Norvell, 1974). It is difficult to reconcile the known levels of luteinizing hormone (LH) in the plasma with porphyrin production. The sexual dimorphism in the hamster may be related to the LH surge in the female since it appears likely that porphyrin formation is dependent on elevated levels of gonadotrophic hormones in the golden hamster (Clabough and Norvell, 1973; 1974). However, in the female, when ovariectomy results in a continually rising level of LH, the production of porphyrin does not exceed the level found during the oestrous cycle (Hoffman, 1971; Payne et. al.; 1977). Mast Cells - "The Harderian gland of the female golden hamster contains an average of forty times more mast cells in the interstitial tissue than the male gland. A similar sex difference occurs in the connective tissue capsule of the gland. Mast cell numbers in the male Harderian gland increase greatly between two and five months after castration, reaching levels comparable to those found in the female. This post-castrational rise in mast cell numbers is prevented by androgen administration. Furthermore, if mast cell numbers are allowed to increase in castrates, subsequent androgen administration will reverse the rise. It 20 appears that the numerical difference is not matched by ultrastructural differences; mast cells in male, female and castrated male glands were essentially similar in appearance" (Payne et. , I982). The correlation of numbers of mast cells with hormonal status has been reported in several tissues. Mast cells are said to fluctuate during the cycle and to increase during pregnancy and lactation. The ability of androgens to prevent the rise in mast cell numbers after castration, or to suppress it once it has begun, indicates a powerful relationship between mast cell numbers and hormonal status (Payne et. al., 1982). "As regcirds to the functional significance of this sex difference and hormonal control, it is difficult to evaluate because there is little conclusive evidence on the functions of either the Harderian gland or mast cells. It has been postulated that heparin may have a lipid-clearing action and that mast cell numbers may rise in response to increased lipid deposition. Mast cells may play some role in iron metabolism and haemopoiesis." This may be of significance, since the female and castrated male Harderian glands actively manufacture porphyrins (compounds which, when complexed with iron, form haem). Anticoagulants, such as heparin are useful in female and castrated male glands where porphyrins are laid down as large intraluminal accretions which might temporarily occlude blood vessels at the site of formation of the accretion or during its passage to the main secretory duct. The current lack of evidence correlating mast cell numbers and porphyrin content does not support such an interpretation (Payne et. al., 1982). 21 Concentrations of Metals - A study of the concentrations of metals in the Harderian glands showed that glands from male golden hamster have lower concentration of Na, Mn and Ca and higher concentrations of Fe and Mo than female glands. No sex differences were evident in Cu, Mg, P or Zn (Table 2). Table 2. Changes in concentrations of various metals in the Harderian glands of hamsters in response to pinealectomy, castration or blinding as a function of sex.* Element Na Ca Cu Fe Mg Mn1 Mo P Zn Rx M F M F M F M F M F M F M F M F M F Pinx _ i 1 —» —? —? —* —? Castration T - T — T — —? Î 1 - —* — -* Blinding T - Î 1 - Î i i Male vs Female ± ± + ± ± M = Male; F = Female. Î, increase (from normal); i, decrease; —no change. +, higher (than female); —, lower; ±, no difference. *From R.A. Hoffman and J.W. Jones. 1981. Concentrations of Metals in the Harderian Gland of Male and Female Hamsters. Comp. Biochem. Physiol. PtA. Í2.: 153-156. Since pinealectomy has little effect on metalic concentrations in the Harderian gland, it can be concluded that the effects of blinding are via mechanisms exclusive of the pineal (Hoffman and Jones, I98I). This conclusion is contrary to that described earlier in this paper (Hilton, 1971; Clabough and Norvell, 1973; 1974). Blinding decreases Zn in both 22 sexes and Mn in the female and increases Ca in the male and Cu and Mo in the female. Gonadectomy increases concentrations of Wa, Cu and Mn in the male and Fe in the female and decreases Mo in the male. It is suggested that iron is in higher concentrations in the male and it may be possible that androgens increase and estrogens decrease Harderian gland iron (Hoffman and Jones, 1981). On the basis of limited data, Hoffman and Jones (I98I) suggested that Mo is associated with the cell type II and reduced porphyrin synthesis in the male Harderian gland. In view of the yellow color of the male gland, high concentrations of flavoprotein might be suspected. These concentrations of metals would reflect the relative abundance of the cell type II and parallel the high concentrations of Mo in the male. Manganese is associated with cell type I and porphyrin synthesis in both males and females. Ii/hether Mn is an integral part of a metalloenzyme or acts in a metal activating enzyme system is unknown. Its concentration within the gland appears to be under the same regulatory influences which modify the cellular composition and Mo concentration (Hoffman and Jones, 1981). INNERVATION - Both adrenergic and cholinergic nerve terminals have been demonstrated in the Hairderian gland. Cholinergic terminals have small clear vesicles and are associated with alveolar endpieces. Adrenergic terminals having characteristic dense-cored vesicles are associated with blood vessels (Norvell and Clabough, 1972; Sakai, 1981). Nerve endings are also present in apposition to both the cytoplasmic extensions of the myoepithelial cells and the secretory cells. The nerve endings adjacent to the myoepithelial cells have myoneural junctions. The innervation of the Harderian gland was observed to be similar in both male and female hamsters 23 (Bucana and Nadavukaren, 1972). Observations of the presence of nerve fibers associated with blood vessels of the interlobular connective tissue support the conclusion that the secretory activity of the gland is influenced by blood flow to the gland (Bucana and Nadakavukaren, 1972). The process of secretion may be affected directly by nervous stimulation of the myoepithelial cells. Rats injected with the neurotransmitter acetylcholine will produce, in a matter of minutes, a copious secretion of "bloody tears" from the Harderian gland. Myoepithelial cells may respond to acetylcholine by contraction, thus squeezing out the contents from the secretory cells in a manner similar to the contractile response of the myoepithelial cells of the mammary glands to oxytocin called "milk let down." The role of norepinephrine in regulating secretion by the gland is not understood (Bucana and Nadakavukaren, 1972; Strum and Shear, 1982). "The presence of nerve endings in apposition to the myoepithelial cells and the myoneural junctional folds indicates that the neurotransmitter passes a myoneural junction instead of diffusing through the connective tissue. This may also explain the fast response of the Harderian gland to the injection of acetylcholine" (Bucana and Nadakavukaren, 1972) 24 SECRETORY PRODUCTS SECRETORY MECHANISM - A unit membrane surrounds cytoplasmic secretory vacuoles. Lipid droplets in the alveolar lunina of the Harderian gland are not surrounded by a unit membrane. Thus a merocrine type of secretion, i.e., exocytosis to release secretory vacuoles is indicated (Fig. 6) (Bucana and Nadakavukaren, 1972). Merocrine secretion refers to release of a product through the plasma membrane without loss of apical cytoplasm (Krause and Cutts, 1981). Figure 6. Diagrams showing the secreting process of lipid droplets (broken circles) by merocrine (a) and apocrine mechanisms (b). From T. Sakai. 1981. The Mammalian Harderian Gland: Morphology, Biochemistry, Function and Phylogeny. Arch. Histol. Jap. ÜiL: 299-333- The Harderian gland is the sole example in which oily secretory vacuoles are secreted by exocytosis. In these glands, lipid vacuoles are secreted by the holocrine (in sebaceous glands) or the apocrine mechanism (in mammary glands) (Sakai, 1981). 25 The chemical nature of the Harderian material is species specific and has not been completely characterized in any species. A major portion is lipid-like in composition and is associated with porphyrin pigments, melatonin, taurine and other components (Thiessen and Kittrell, 1980). The cells also have the capacity to produce proteins, as indicated by the presence of rough endoplasmic reticulum, a Golgi complex and secretory granules in the cytoplasm (Weaker, 1981). LIPID - The main components of the lipid are either glycerol ether diesters (1-alky1-2, 3-diacylglycerol) or wax esters (Sakai, 1981). This lipid secretion lubricates the eye and may also contain compounds and enz3rmes having antibacterial properties (Strum and Shear, 1982). Gas-liquid chromatography of the fatty acid composition of lipids showed significant sexual differences in the Heirderian glands of golden hamsters. "The major fatty acids of the female gland lipids were C16:0 and Cl8:1 which accounted for more than 70^ of the total fatty acids. Most of the remaining fatty acids of the female gland were also long-chain types, e.g. Cl 8:0, Cl 8:2 and C20:0. On the other hand, the male glands had a greater proportion of fatty acids with chain lengths ranging from CIO to C15 than those in the female glands. The proportion of C16:0 and C18:1 fatty acids was considerably lower in the male glands than in the female glands. Fatty acids with chain lengths ranging from CIO to Cl5 accounted for about 50% of the total fatty acids in the male glands, while the remainder consisted of Cl5 to C20 types" (Lin and Nadakavukaren, 1981). Both male and female Harderian glands showed similar compositions of neutral lipids and phospholipids. Cholesterol, fatty acids, alkyldiacylglycerol, phosphotidyl choline, phosphotidyl ethanolamine and an unidentified phospholipid v/ere separated by thin-layer chromatography. It 26 is noteworthy that triglycerides were not detected in the hamster Harderian glands. The absence of triglycerides could indicate that the major lipids are secretory products of the gland rather than lipid storage products, since lipids are generally stored as triglycerides. PORPHYRIN Localization - Porphyrin in the Harderian gland was first identified by Derrien and Turchini (1924). Porphyrin pigment concentrations have been found in the Harderian gland of some rodents, i.e., rats, mice and female hamsters, while they are absent in most other mammals. In some mammals the Harderian gland, while containing no pigment accretions, may still secrete a small amount of porphyrin (Sakai, 198I) (Plate 1). Paule and co-workers (1955) identified reddish-brown or dark amber pigment granules found in the lumina and acini of female hamsters as being pigment granules of porphyrin. Porphyrin of Harderian glands exhibit a red fluorescence in ultraviolet light. In female hamsters, a fast-fading red fluorescence is exhibited by the alveolar cells and a persistent red fluorescence by the luminal pigment granules (Clabough and Norvell, 1973). Biosynthesis - It has been shown that the secretory epithelium of the Harderian gland is capable of porphyrin biosynthesis. Type I cells are responsible for the secretion of lipid droplets, probably containing a big porportion of unsaturated fats, with which the porphyrins may be associated (Shirama et al.. 1981). Glycine is a major precursor in the biosynthesis of porphyrins (Fig. 7). The porphyrins are constructed from four molecules of the monopyrrole derivative porphobilinogen. Porphyrin biosynthesis is 27 regulated by the concentration of heme protein product, such as hemoglobin, which can serve as a feedback inhibitor of early steps in porphyrin synthesis (Lehninger, 1982) .Aaminolevulinate synthase (ALA synthase), which is the rate-limiting enzyme and catalyzes the first step, is inhibited by the presence of protohemin IX (Sakai, 1981) (Fig. 7). Chromatography of both free porphyrins and methyl esters in the hamster Harderian gland shows tetracarboxylic coproporphyrin and dicarboxylic protoporphyrin as the principal components. Tricarboxylic harderoporphyrin was present in trace amounts while no significant quantities of uroporphyrin or hepta-, hexa- or pentacarboxylic porphyrins could be found (Payne, 1977). Porphyrin in the Harderian gland of rats is composed of protoporphyrin IX (64$), harderoporphyrin (29$) and coproporphyrin III (9$) (Sakai, 1981). In mice, only protoporphyrin is present (Strum and Shear, 1982). Action - Porphyrins readily chelate with metal ions forming such compounds as heme, chlorophyll, cytochromes and others. The metal ion lies at the center of the porphyrin nucleus with four of its ligand sites occupied by the pyrrole nitrogens (Hoffman and Jones, 1981). Porphyrin in the Harderian gland is mostly metal-free and therefore characteristically red-fluorescent. The accumulation of metal-free porphryin is attributed to a lack of metal chelating enzyme (Sakai, 1981). Though porphyrins can play important roles in the form of metal chelates and as catalytic cofactors, the significance of porphyrin in the Harderian gland is unknown. VIetterberg and co-workers (1970) expressed the view that porphyrins in the Harderian gland serve as the converter of ultraviolet frequencies that are below perceptual threshold to frequencies within a visual range 28 succ^nyl Co A * glycine (¿-AlASAse) i-omlnolevullnate x2 porphobilinogen x4 V J uroporphyrinogen m horderoporphyrlnogen- horoeroporphyrln protoporphyrin IX + Fe^* globln ? orotoheme IX - - I - -protohemln IX \ hemoglobin Figure 7. The synthetic pathway of porphyrins. The reactions indicated by continuous arrows occur in the Harderian gland; those by broken arrows do not. (T-ALAase is c/'-aminolevulinate synthase. From T. Sakai. 1981. The Mammalian Harderian Gland: Morphology, Biochemistry, Function and Phylogeny. Arch. Histol. Jap. M.: 299-333. Content and Concentration - In hamsters, the porphyrin content in the Harderian gland of females may be as great as a hundred times higher than in males (Hoffman, 1971). The level in female mice was 20 or more times that in males (Shirama et al. 198I). Very low porphyrin content in the male hamster may be due to very low ALA synthase activity while the opposite case may be observed in the female (Lin and Hadakavukaren, I982). 29 Shirana and co-workers (1981) reported that in intact female mice, both porphyrin content and concentration in the Harderian gland varied v/ith the stage of the estrous cycle. The level was lowest during metestrus and highest during diestrus (Fig. 8). SA £ s C. cl oestrus oestrus oestrus Figure 8. Content and concentration of porphyrin in the Harderian glands of intact female mice during the estrous cycle are expressed as ug pair Harderian glands (solid line) and as ug lOOmg wet wt tissue (broken line). Values are also shown in intact male mice as ug pair glands (hatched bar) and as ug/IOOmg wet wt tissue (solid bar). The closed circle for each group represents the weight of Harderian gland. From K. Shirama, F. Furuya, Y. Takeo, K. Shimizu and K. Maekawa. 1981. Influences of Some Endocrine Glands and of Hormone Replacement on the Porphyrins of the Harderian Glands of Mice. J. Endocr. 21: 305-311» 30 Payne and co-workers (1979) found both porphyrin content and the ALA synthase activity level fluctuate over the estrous cycle shov/ing highest levels in the estrus period in the female hamster (Fig. 9). Thus the hamster cycle differs from that of mice. Figure 9. (A) Relative Harderian gland weight. (B) Average porphyrin accretions per unit area within the Harderian gland of the female golden hamster over the estrous cycle. From A.P. Payne, J. McGadey, M.R. Moore and G.G. Thompson. 1979. Changes in Harderian Gland Activity in the Female Golden Hamster During the Oestrous Cycle, Pregnancy and Lactation. Biochem. J. 178; 597-604. In pregnant hamsters on days 4-16, the porphyrin concentration of the Harderian gland was higher than in intact estrous hamsters (Payne et al., 1979). In contrast, the glandular porphyrin in pregnant mice on day 15 was similar to that in intact estrous mice (Shirama et al., I98I). Manipulation Experiments - In rats, dietary pantothenic acid deficiency results in a high activity of ALA synthase and excessive 31 accumulation of porphyrin leading to hypertrophy of the Hcirderian gland. Under this condition, the excessive secretory products of the Harderian gland are accumulated on whiskers and furs, and look like red tears or a bloody substance (Sakai, 198I). Figure 10 shows the effect of adrenalectomy, ovariectomy, hormone replacement and pregnancy on Harderian gland porphyrins in female mice. Ovariectomy slightly increased porphyrin concentration, but not its content. In adrenalectomized females, the normal estrous cycle occurred 33-52 days before death. Both porphyrin content and concentration of the glands from these animals (estrous phase) were significantly higher that those of unoperated estrous mice. However, in ovariectomized-adrenalectomized mice, the porphyrin level was not significantly different when compared with that in intact control mice (Shirama e¿ al., 198I). Administration of esterons to the ovariectomized-adrenalectomized mice significantly increased porphyrin content but not its concentration. Esterons also prevented the reduction of gland v/eight normally caused by these operations. Administration of progesterone stimulated an approximately twofold increase in porphyrins which was similar to levels in adrenalectomized-estrous mice. On the other hand, implantation of either testosterone or testosterone plus progesterone into ovariectomized-adrenalectomized females caused a marked decrease in porphyrins. The glands in these groups resembled those of males structurally. In pregnant mice, the levels were similar to those in intact estrous mice (Shirama et. al. , 1 98I ) . 32 Figure 10. Influence of the extirpation of adrenal (ADX) and/or ovary (OVX), sex hormone replacement (Prog=progesterone; T=testosterone) and of pregnancy on the weight of the Harderian gland (mg=open bars) and on the content of porphyrins in female mice. Porphyrins are expressed as ;ag/pair glands (hatched bars) and pg/100mg wet v/t tissue (solid bars). From K. Shirama, T. Furuya, Y. Takeo, K. Shimizu and K. Maekawa. 198I. Influences of Some Endocrine Glands and of Hormone Replacement on the Porphyrins of the Harderian Glands of ULce. J. Endocr. 91: 305-311. Figure 11 shows the influence of the testes, adrenals and hormone replacement on the porphyrin levels in the Harderian gland in male mice. Castration caused a significant increase in the porphyrins producing about three times the levels in intact males. Although adrenalectomy alone did not cause any rise, in castrated-adrenalectomized males, the porphyrins were as high as in females. The histological appearance of the Harderian glands in this group was consistent with the porphyrin concentration. The subcutaneous implantation of testosterone into the 33 castrated-adrenalectomized mice for 30 days completely prevented the rise in porphyrins, while progesterone implantation caused a further rise in contrast with the effect of testosterone. The subcutaneous administration of testosterone and progesterone into those operated mice decreased the concentration of porphyrins in the gland. Finally, subcutaneous injections of testosterone into the progesterone-treated males resulted in a dose-dependent reduction of the gland porphyrins in castrated-adrenalectomized mice fShirama et al.. 198I). O CL Figure 11. Influence of the extirpation of adrenal (ADX) and/or testis (Cast) and of sex hormone replacement (T=testosterone; Prog=progesterone) on the weight of the Harderian gland (mg=open bars) and on the content of porphyrins in male mice. Porphyrins are expressed as ;jg/pair glands (hatched bars) and jag/IOOmg wet wt tissue (solid bars). From K. Shirama, T. Furuya, Y. Takeo, K. Shimizu and K. Kaekawa. 1981. Influences of Some Endocrine Glands and of Hormone Replacement on the Porphyrins of the Harderian Glands of Mice. J. Endocr. 9I: 305-311. 34 Castration of hamsters results in conversion of the male to the female type of Harderian gland within 20-30 days. Males experience a significant reduction in the weight of the gland, disappearance of type II cells and the presence in the gland of solid porphyrin accretions. Porphyrin content shows a progressive increase from about 4 nmol/g tissue in intact male hamsters to about 2500 nmol/g tissue in the castrated hamsters for 8 months. Almost all the alteration in the total porphyrin content is due to changes in the concentration of dicarboxylic protoporphyrin. The level of tetracarboxylic coproporphyrin first increases after castration and then declines. This suggests the possibility of some temporary instability in porphyrin production in the newly synthesizing gland. The increase in the total amount of porphyrin with time after castration may be caused by a direct increase in the capacity of the gland to synthesize porphyrin or by normally present small amounts of porphyrin not being eliminated, which would result in the formation of solid intraluminal accretions over time ( Payne et. al.., 1977 ) • Uroporphyrinogen-III-cosynthase activity was lower in the Harderian gland of the male hamster than the female. Uroporphyrinogen-III-cosynthase activity doubled in the castrated male suggesting the influence of testosterone on enzyme activity (Lin and Nadavukaren, 1982). Administration of androgens, i.e., testosterone propionate, to castrated hamsters maintains the presence of type II cells within the Harderian gland and suppresses the production of porphyrin (Hoffman, 1971). Thus a role for steroid hormones is suggested in the regulation of porphyrin biosynthesis in the hamster Harderian gland. 35 MELATONIN Localization - The possible endocrine capabilities of the Harderian gland may be the result of indoleamines, particularly the 5-methoxyindoles. Among the 5-niethoxyindoles, melatonin, N-acetyl-5-methoxytryptamine, was isolated first and has been extensively studied. 5-methoxytryptophol (MTL) has also been isolated, and like melatonin, is known to be present in the general circulation. Melatonin and MTL play an important role in numerous physiological processes and are capable of modifying sexual development and reproduction in birds and mammals (Pevet et al.. 198O; Reiter et al.. 198I) In the Harderian gland, melatonin was first detected at approximately 10 days after birth. It progressively increased v/ith age and reached the adult levels between 25-30 days of postnatal life (Bubenik et al., 1978). Using a specific fluorescence labelled antibody technique, Vivien-Roels and co-workers (I98I) found melatonin to be localized in the secretory cells of the alveoli and mostly restricted to the cytoplasm surrounding the nucleus. Differences in the number of reacting cells appear from one alveolus to another; in some alveoli, most of the cells are fluorescent, while in other alveoli no reacting cells were observed. Biosynthesis - All 5-methoxyindoles are synthesized by the enzyme hydroxyindole-o-raethyltransferase (HIOMT). Although probable, it is not known whether the Harderian gland is capable of synthesizing all the 5-methoxyindoles found in the pineal (Pevet et al.., 198O; Reiter et al.. 1981). "In the pineal, serotonin can be acetylated to N-acetyl serotonin, oxidized to 5-hydroxyindole-3-acetic acid, or metabolized to 5-hydroxytryptophol. The 5-hydroxyindoles, i.e. , 5-hydroxytryptophan. 36 5-hydroxytryptamine, 5-hydroxyindole-3-acetic acid, 5-hydroxytryptophol and N-acetylserotonin, can all be methylated by HIOMT. Thus at least five different methylated products can be formed in the pineal" (Pevet et al.. 1980). Circadian Rhvthmicitv - Under natural conditions, both melatonin biosynthesis and rhythmicity differ in the pineal and Harderian glands and the retina. Most striking is the fact that the Harderian glands synthesize 10-50 times more of practically all the methoxyindoles when compared with the pineal and the retina. This holds for the whole year for melatonin, MTL, acetylmethoxytryptophol (ailL) and methoxyindoleacetic acid (MA). However, in wintertime methoxytryptophan (MVJ) and methoxytryptamine (MT) synthesis are very high in the Harderian gland and the retina whereas these compounds are replaced by MA, their probable endproduct, in both spring and summer. It appears that for each compound, a circadian rhythm may at times be present and at other times absent during any particular season of the year. V/hen circadian rhythms are present, they are usually different in the different organs (Balemans et. al., 1983). The serotonin concentration in the Harderian gland corresponds exactly, but in an inverse way, with that of the HIOMT activities (Pevet et al.. 1980). "The primary control of melatonin synthesis within the pineal seems to be prevailing photoperiodic conditions. Low levels of melatonin production are associated with the period of light while large increases in pineal melatonin have been examined during the period of darkness" (Reiter ei. , 1981). Action - It has been suggested that melatonin may act directly on: (1) the organs by which it is produced, such as the pineal, retina, Harderian gland or intestine; (2) the hypothalamo-hypophysio-gonadal axis; 37 and, (3) the gonads. Melatonin synthesized in the Harderian gland may act intracellularly or extracellularly as a substance inhibiting or facilitating the visual process by modifying the porphyrin content of the Harderian gland (Wetterberg e£. ai., 1970; Pevet et. al., I98O). Balemans and co-workers (1983) concluded that MIL, aML and MT are probably as significant as melatonin is. They further agree with the suggestion that the methoxyindole compounds may constitute a "family” just like the steroid families. One may wonder whether melatonin shows any direct endocrine activity. On correlating the organ weights one may wonder whether melatonin would not act on the pineal gland itself, a supposition which is supported by the observations of Wurtman and co-workers (1964) that the pineal and of Bubenik and co-workers (1978) that both the retina and the Harderian gland take up melatonin from the circulation. Pineal fluctuations in melatonin and MT content appear almost insignificant in relation to the almost constantly enormous amounts present in the Harderian gland of the golden hamster. Balemans and co-workers (1983) suggest that both melatonin and MT, if released in the circulation, act on the pineal causing the release of some other, as yet unknown, factor influencing directly the reproductive endocrine system. Both MT and MTL are antigonadotropic. Concentrations of melatonin, MTL and MA are lowest in March when the light period is longer than in December and it is not as cold as in December, while aML pineal concentration is then highest. The opposite can be observed for the Harderian gland. This suggests that external factors in the formation of the circannual rhythm are involved and that also another, as yet unknown, circannual clock controlling rhythmicity may be present (Balemans et al.. 1983). 38 "Another possible explanation of the regulation of gonadal activity may be that more than one methoxyindole should necessarily act together. This may explain why, for example, melatonin shows different effects when injected at different times of the day or probably also in the year" (Balemans ^ al,., 1983). Manipulation Experiments - Exposure of rats to constant light decreased melatonin content while porphyrin content is also decreased. Administration of melatonin increases porphyrin content and concentration in the Harderian gland of rats. Reproduction is curtailed in rodents under long dark cycles, when melatonin levels are high, and is active during long light cycles, when the level is low due to the retinal-pineal-gonadal axis (Shirama, 1978; Weaker, 1981). Gonadectomy influences photoreceptor damage which is caused by continual light exposure. High circulating levels of gonadotrophins, vihich follow gonadectomy, may suppress retinal melatonin production. There is some evidence that these hormones also hamper pineal melatonin synthesis (Reiter et. al., 1983). Peak Harderian gland and retina melatonin levels were augmented by pinealectomy. Thus something of pineal origin may normally curtail melatonin levels within the Harderian gland (Reiter et al., 1983). Thus via the 5-methoxyindoles, parallel but different information such as light and temperature could be integrated and could be of great importance for the adaptation of the endocrine reproductive axis to modifications in the environment (Panke et al.. 1979). Harderian gland rhythm was similar to that observed in the retina indicating that these rhythms may be generated by a similar signal (Reiter et al.. 1983). 39 TAURINE - The presence of taurine, an amino sulfonic acid, has been identified by amino acid analysis in the Harderian gland of the mouse. Radioactivity in the gland after intraperitoneal injection of labeled taurine suggests that the gland can accumulate taurine from the plasma. It is not known whether the Harderian gland is capable of synthesizing taurine and whether this amino acid plays a role in the function of the gland. Taurine has been shown to be synthesized from cysteine in vivo in a number of tissues in various species. However, in the cat, dietary taurine is required to maintain function of the retina (Hilton ^ al.. , 1 981 ) . High concentrations of taurine in the retina and pineal gland prompted measurement in the Harderian gland. The retina of the mouse was found to have concentrations of taurine two to three fold higher than in the Harderian gland until the time of eye opening, after which concentrations in the retina markedly increased to greater than 160 nmoles per mg dry retina and gradually decreased in the Harderian gland to 31-1 nmoles per mg dry weight of gland (Fig. 12) (Hilton et. al., 1981). The decrease 4 days postpartum in the Harderian gland suggest that lower levels of taurine in Harderian glands of adult mice are associated with the attainment of sexual maturity. No differences in the concentrations of taurine in male and female Harderian glands were noted. This lack of sex-related differences in taurine concentrations of the gland suggests that endogenous testicular hormones do not influence the concentrations of taurine in the adult Harderian glands (Hilton et al.. 1981). This is in contrast to the observations of sexual dimorphism by Lin and Nadakavukaren (1979), Sun and Nadavukaren (1980), and Norvell and Clabough (1972). 40 »m m Mit Figure 12. Taurine concentrations in the Harderian glands and retinas of nice during postnatal development. The number by each point represents the number of animals used to obtain each value. From F.K. Hilton, G.H. Raque and M.A. Hilton. 1981. Changes in Concentrations of Taurine in Murine Harderian Glands and Retinas During Postnatal Development. J. Exp. Zool. 216: 493-495. 41 FUNCTION The function of the Harderian gland has remained an enigma since its discovery. Many of its multiple functions appear to be species-specific. Neither the phylogenetic pattern of its appearance nor the ecological or demographic circumstances of its use suggest an invariant function. It is found in diurnal and nocturnal species, in species that emphasize olfactory and visual communication systems, in terresterial and aquatic species, and in those species that possess a variety of demographic and social traits (Theissen and Kittrell, 1980). ENDOCRINE AND EXOCRINE EVIDENCE - Both endocrine and exocrine functions have been suggested for the Harderian gland. The Harderian gland appears to be regulated by hormones and may synthesize certain steroids (endocrine function) as well as synthesizing and secreting lipid and porphyrins in to a lumen (exocrine function) (Strum and Shear, 1982). The Harderian gland of rats atrophied following either thyroidectomy or hypophysectomy. Neonatal injections of thyroxine into rats caused porphyrins to appear in the Harderian gland sooner than normal vihereas neonatal hydrocortisone treatment retarded the appearance of porphyrins. Evidence of a role in steroid synthesis comes from a report that the Harderian gland in rats selectively accumulates a precursor of pheromone steroids (Johnson et al.. 1979; Strum and Shear, 1982). The cytology of the Harderian gland lends further evidence for endocrine capabilities. The abundance of smooth endoplasmic reticulum in the secretory epithelial cells is characteristic of steroid secreting endocrine cells. It should be noted that the cells involved in cholesterol 42 and lipid synthesis may also have large amounts of smooth endoplasmic reticulum (Strum and Shear, 1982). There appears to be active exchange between the blood vascular system and the acinar cells. Although microvilli are present on all free surfaces of the cell, they are more numerous and longer on the surfaces adjacent to fenestrated capillaries. Fenestrated capillaries are characteristically associated with endocrine glands. Thus rapid exchange, i.e., uptake and release, may be taking place between the blood vascular system and cells of the Harderian gland (Weaker, 1981). LUBRICATION - The association of the Harderian gland with the third eyelid suggests that its main function may be to lubricate the eye. Aquatic mammals have a well-developed Harderian gland while their lacrimal gland is rudimental or absent (Sakai, I98I). The oily secretion of the Harderian gland may protect the corneal surface of the eyes from sea water. Because of its oily nature, the secretion can serve as a lubricant and also prevent the overflow of tears (Bucana and Nadakavukaren, 1972). ANTIBIOTIC EFFECT - Unsaturated glycerine ethers present in the secretion of the Harderian gland of the rabbit may have an antibiotic effect (Bucana and Nadakavukaren, 1972). In some avian species, the Harderian gland may be a component of the peripheral lymphoid system. The exposed surface of this gland appears to have a bearing on the immune response (Schramm, 198O). In the chicken, the Harderian gland probably plays a role in the immune response, since large numbers of cells resembling plasma cells and antibody producing cells have been detected in this gland (Lin and Nadakavukaren, 1979). PHEROMONES - Female hamsters exhibit cyclic and seasonal variations in body weight and productivity of pheromones. Payne (1977) suggested that 43 pheromones in the Harderian material could serve a communicative function to give information on sex and hormonal status of the individuals. It is known that male hamsters show more aggressive behavior towards other males than towards females. Payne (1977) devised an experiment to test the possibility that diverse Harderian gland extracts could influence the aggressive response of a male towards an opponent male. The opponent male was daubed around the eyes and snout with an homogenate of female Harderian gland material. The results suggest that female Harderian gland homogenate can inhibit aggressive responses toward a male opponent, even if other cues remain masculine. THERMOREGULATION - In the gerbil, the Harderian gland seems to have at least two functions, chemocommunication and thermoregulation. The spread of Harderian lipids on the fur as a result of autogrooming can act as a mechanical insulation against environmental changes in temperature and dampness. Lipids may also help conserve internally-generated heat and prevent loss of body water. The effects of fur lipids could be multiple and depend on a variety of internal and external parameters (Thiessen and Kittrell, 1980). Gerbils respond to Harderianectomy with decreased fur lipids, while rats, hamsters and mice did not decrease fur lipids. This suggests that the gerbil*s response is unique among rodents. The use of insulating lipids may be related to the gerbil*s ecology in the cold deserts of Mongolia and Northeast China. Chemoattraction and fur-lipid spread occur simultaneously during an autogroom. The Harderian material initiates mutual investigations between the groomer and conspecifics. Thus, short-term evaporative cooling with saliva and long-term thermoregulation by the application of fur lipids are 44 accomplished. The functions of communication and thermoregulation appear to be inexorably linked (Thiessen and Kittrell, 1980). RETINAL-PINEAL-GONADAL AXIS Estradiol Target - VJeaker and co-workers (1983) established the Harderian gland as a target organ for estrogen in the female armadillo. Nuclear uptake and retention of tritiated estradiol was observed in the duct of the gland, the periductal mucous cells of the mucous-secreting lobules, and the fibroblasts in the interstitial connective tissue. In contrast to the mucous-secreting cells, label was not observed in the lipid-secreting cells. Extraretinal Photoreceptor - Johnson and co-workers (1979) demonstrated that the Harderian gland was acting as an extraretinal photoreceptor in adult male and female rats. Using hypophysectomized (HYPEX) rats, it was demonstrated that an enlargement of the Harderian gland with exposure to high-intensity illumination was due directly to a radiant energy-dependent mechanism and was not due to endocrine etiology, i.e., pituitary hormone mediation. Exposure of adult female rats to constant light (LL) eventually induces persistent estrus which is characterized by constant vaginal cornification and cystic follicular ovaries devoid of corpora lútea. Rats exposed to LL from the day of birth are less likely to go into persistent estrus compared to rats kept in LL beginning at the 70th day of age. Exposure of rats to LL during the eye-closed period delays the onset of persistent estrus caused by LL (Shirama, 1978). In the developing rat Harderian gland, the lobules of the gland were 45 widely separated by connective tissue in newborn rats, and the glandular epithelium was thin, consisting of only a single layer of low cells. At days 15 and 70, the size of the lobules increased progressively. The effects of LL were not recognizable in the structure of the gland, though a rapid increase in the weight of LL rat Harderian glands was observed at 10 to 20 days of age (Shirama, 1978). Wetterberg and co-workers (1970) implicated the Harderian gland as a photoreceptor for neonatal rats. In adult rats, however, it does not seem to have photoreceptive capability, though the glands regress after exposure to LL for prolonged periods. Daily treatment with a Harderian gland homogenate, made from glands taken from 12h light-12h dark (LD) exposed pups, given during the 1st through the 15th days of life reduced the delay in persistent estrus caused by neonatal light exposure. Harderianectomy significantly increased ovarian and uterine weights. Thus, Shirama (1978) suggests that exposure of neonatal female rats to LL inhibits persistent estrus inductive effect of LL and lowers the antigonadal effect of the Harderian gland. Exposure of rats to constant light for a prolonged period results in decreased porphyrin content. However, administration of melatonin increased porphyrin content and concentration in the gland. VJetterberg and co-v;orkers (1970) expressed the view that porphyrins in the Harderian gland serve as the converter of ultraviolet frequencies that are below perceptual threshold to frequencies within a visual range. Reduced melatonin levels may act as a substance inhibiting the visualization process by lowering porphyrin content. Thus, injections of Harderian gland homogenate from LD pups may have supplemented the LL-induced melationin deficiency. However, since neonatally administered hematoporphyrin rats could not alter the 46 susceptibility to LL, further study is needed to reach a definite conclusion (Shirama, 1978). The hamster shows seasonal rhythmicity in reproduction. The testis of the male hamster show atrophy in autumn and winter, a recrudescence in the beginning of spring and activity at the end of spring and in summer. These fluctuations in testicular size and function are initiated by short and long photoperiods, respectively (Stetson and Tate-Ostroff, 198I). The synthesis of melatonin and MTL is highest when the gonads are growing in weight and lowest when they regress (Balemans et al.. I983). Pineal Melatonin Rhvthm - The Harderian gland may act as an extraretinal photoreceptor effecting the pineal serotonin rhythm in neonatal rats (VJetterberg et al., 1970). The gland may have a role in mediating changes in pineal melatonin metabolism in hamsters. Peak melatonin levels were significantly shifted or decreased following removal of the Harderian gland. The presence of a melatonin circadian rhythm, as found in the pineal, would be suggestive of a Harderian gland-reproductive system interaction (Pevet et. al., 198O). Porphyrin Production - Porphyrin production in the female hamster Harderian gland was found to change over the oestrous cycle, pregnancy and lactation ( Payne et. al.., 1979; Shirama et al.. 1981). '•Clinical studies have generally shown that porphyrin production in normal women rises during pregnancy. Of particular interest is the frequency with which attack of acute porphyria could be precipitated with either endogenously produced or exogenously administered progesterone. The relation of human porphyria to known changes in sex steroids and adrenal steroids is similar to the relationship of Harderian gland porphyrins with these steroids. Thus the Harderian gland could prove a useful system to 47 study the role of various substances in the regulation of raannalian porphyrin biosynthesis" (Shirama et. al. , I98I ) . Manipulation Experiments - Manipulation experiments already described in this paper suggest there may be a retinal-pineal-gonadal chain. The effect of castration is greatly reduced if the animal is also blinded. In male hamsters, castration and blinding results in minimal pigment formation. However, castration, blinding plus pinealectomy in combination result in pigment formation similar to that in the female (Clabough and Norvell, 1973). Adult female hamsters blinded for 11 v/eeks exhibit a diminished amount of porphyrin pigment. Blinding combined with pinealectomy prevents this change (Clabough and Norvell, 1974). 48 DIABETIC PROFILE The diabetic Chinese hamster (Cricetulus griseus) has been shown to be an adequate model for human diabetes. Extensive investigation with the diabetic Chinese hamster has shown biochemical and physiological changes similar to those in man. Diabetic hamsters have less insulin than nondiabetics. Plasma insulin in severe diabetics fails to increase with glucose load (Gerritsen and Dulin, 1967). Organ systems of the diabetic hamster have shovm numerous pathologies. Beta cell degranulation and/or destruction similar to that of diabetic man has been exhibited (Luse et. al. . 1967). Marked bladder distension in ketonuric hamsters is similar to the asymptomatic diabetic syndrome of man (Gerritsen et. ^1.» 1974). The development of glomerulopathy in diabetic hamsters is suggestive of similar changes in diabetic man (Shirai e¿ i 1967). Gastric dilation and significantly delayed gastric emptying occurs in diabetic hamsters (Diani et. ai.., 1979). In the testes, reduced thickness of the germinal epithelium and widening of the lumina of the seminiferous tubules have been displayed. The mucopolysaccharide content of the diabetic hamster skin was shown to be altered in comparison to controls (Sirek and Sirek, 1967). In a study of the thoracic aorta, the internal elastic lamina was fragmented and calcium deposits were found in the media. Proliferation of smooth muscle was also observed (McCombs et al. . 1974). Vascular lesions have been displayed in the retina and brain. Glycogen accumulation in the outer nuclear layer of the retina was reported (Soret et al.. 1974). Central and peripheral nervous systems of the diabetic hamster have also shown various 49 abnormalities. In brain neuronal processes, these aberrations consisted of megamitochondria in dendrites, dense fibrils in the axoplasm and degenerate axons, dendrites and myelin (Luse et al.. 1970). "The Syrian or golden hamster ÍMesooricetus auratus) has often been used to study the mammalian pineal-gonadal axis because its gonads are large and are very sensitive to certain experimental manipulations. The Chinese hamster (Cricetulus sriseus) shares these characteristics and, in addition, spontaneously develops diabetes. Infertility is a complication of the diabetic state in this animal and in the diabetic human; therefore, the Chinese hamster may prove to be a useful model for studying reproductive physiology particularly as it relates to diabetes" (McNeill and Smith, 1982). Since the diabetic mutant mouse is infertile, observed differences between the diabetic mouse and its littermate controls may implicate in some way either the diabetic condition, infertility, or both (McNeill, 1978). Clabough and Norvell (1973) suggested that the Harderian gland may be a link in the retinal-pineal-gonadal chain as described earlier in this paper. Gross morphology of the Harderian gland of the diabetic Chinese hamster indicates alterations of this gland in the diabetic state. Therefore, the present study was begun in search of fine structure alterations 50 MATERIALS AND METHODS TISSUE PREPARATION - Tissues used in this study were collected from twelve adult female Chinese hamsters (. Cricetulus griseus) obtained from the Upjohn Company (Kalamazoo, Michigan), Six spontaneously diabetic hamsters were matched for age with six nondiabetic controls. The diabetic condition ranged from 14 to 26 months (415-771 days) in duration (Table 3)• The animals were housed under diurnal artificial lighting (lights on from 0600-1800 hours) and provided viater and Purina Mouse Breeder Chow a¿ libitum. In late September, all animals were weighed prior to being sacrificed under Nembutal anaesthesia (4mg/100g body weight). Blood was drawn from the heart for use in determing the blood glucose level and the vascular system was washed with Earle's balanced salt solution (Appendix A). The animals were fixed by perfusion with 2.5% glutaraldehyde in 0.075M cacodylate buffer. The endocrine glands, i.e., the Harderian gland, pineal, posterior and anterior pituitary, adrenal,thyroid, and the gonads were disected with the aid of a dissecting scope. Extraneous tissue was deleted. Wet weights for each gland and the gonads were obtained on a Mettler H54/H542 balance. One Harderian gland from each animal was divided randomly and part of the tissue was prepared for light microscopy (LH) while the remainder was prepared for transmission electron microscopy (TEM). Tissue for use in the LM study was cut into approximately 1.5-2.0 ram segments while that for use in the TEM study 51 was diced into approximately 1.0 mm cubes with alcohol-cleaned razor blades. The tissues were immersed in cold 2.5p glutaraldehyde in 0.075M cacodylate buffer for two hours. Tissue for use in the LH study was prepared using standard procedures for embedding in paraffin. Dehydration v/as accomplished through a graded ethanol series by placing the tissue in each of the following solutions for 15 minutes: water, 35^, 50^, 70%, 95%, and two times in absolute ethanol. After overnight clearing, the tissue was infiltrated with hot paraffin and placed into plastic holders. The tissue was oriented in these holders which were allowed to cool in cold running tap water. Six pm sections were cut utilizing a Leitz rotary microtome. The sections were mounted on glass slides lightly coated with albumin using distilled water to spread out the tissue. The slides were left on a 46 C slide warmer overnight. A progressive method (Appendix B) was used for staining with Ehrlich Hematoxylin (Appendix C) and Eosin. Tissue for use in the TEM study was washed in 0.1M cacodylate buffer for 15 minutes. The tissue was post-fixed in 4^ osmium tetroxide for 1 hour and washed in 0.1M cacodylate buffer for 15 minutes. Dehydration was accomplished through a graded ethanol series by placing the tissue in the following solutions for 1 hour each: 30^, 50%,, 70%, 95%, 95%, and two times in absolute ethanol. The tissue was placed in propylene oxide, embedded in Araldite 6005 (Appendix D), and oriented in beam capsules. After incubation and cooling, the tissue blocks were removed from the beam capsules. Blocks were trimmed with a single edge razor blade to remove excess resin. Thin sections (60-90 Â) were cut on a Reichert 52 0mU2 ultramicrotome with a DuPont diamond knife and collected on 200 mesh copper grids. Sections were stained for 5 minutes with uranyl acetate and 5 minutes in lead citrate. OBSERVATIONS UTILIZING LIGHT MICROSCOPY - The total number of sections per block usually ranged from seventy to eighty. Sections were randomly selected to be photographed and sampled utilizing a Zeiss VIL standard research grade microscope. Photographs were taken at magnifications of 4X, 10X, 40X, 63X and 100X with a Nikon AMF Hicroflex camera using SO115 Technical Pan film (Eastman Kodak). HC110 developer at dilution F was used. Prints were processed on Kodabromide F-5 paper with Dektol developer. Observations were made at a magnification of 100X using a Plan 100/1.25 oil immersion objective and an ocular micrometer. Twelve to 26 alveoli were randomly selected from each section. Each alveolus vms selected on the criteria that is was cut in cross section and had a concentric lumen. Alveolar size was determined by a measurement of the diameter beginning at the basal end of one cell and traversing the lumen to the basal end of an opposing cell (Plate 1). The cell radius was determined to be the average distance from the apical to the basal surfaces of the cells comprising the alveolus. The number of cells per alveolus was counted. In addition, the number of type II cells per alveolus was recorded. Nuclear position and number within those cells with nuclei was noted. MORPHOMETRIC ANALYSIS OF ELECTRON MICROGRAPHS - Tissue was examined with the Hitachi HS-8 transmission electron microscope at an accelerating voltage of 50kV. Photographs were taken at a magnification of 4,200X on Kodalith LR film (Eastman Kodak). Prints 53 were processed on Kodabromide (8 inch by 10 inch) F-5 paper at a final magnification of 10,987X. The cells comprising the alveoli were divided into three sample groups on the basis of the size and shape of the vacuoles within their cells. The sample groups were: (1) type I cells, (2) type II cells and (3) cells with cleft structures. An acetate grid with a 1 cm square lattice was superimposed over each micrograph selected at random from the sample groups of each animal studied (Plate 2). The object of this method was to calculate the ratio of the particular area in question, to the total area of the cytoplasm (Weibel, 1979). The areas of three organelles were selected for observation as follows: (1) relative vacuole area to total cytoplasm, (2) relative mitochondria area to total cytoplasm and (3) relative concentric lamellae area to total cytoplasm. The points on the grid falling on the particular organelle were counted using a manual counter. STATISTICS - Statistics on data were carried out using a Hewlett-Packard System 45-B Desktop Computer. A General Statistics Program was used to evaluate data while a Graphic Analysis Program was used for assistance in drawing graphs. Data is presented as mean ± standard error of the mean (SEM). The Student-t distribution using a 95% confidence interval was used to determine statistical significance (Freund, 1979). PLATE 1 1. Secretory alveoli cut in cross-section. The overlay demonstrates the procedure used to measure alveolar diameter and cell radius. The number of cells per alveolus as well as the number of type II cells per alveolus were counted. This micrograph was taken of a diabetic animal with no type II cells. L, lumen; N, nucleus. x195 2. Snail porphyrin pigment accretion (arrowhead) within the lumen of a secretory alveolus. N, nucleus; arrow, myoepithelial cell nucleus. x282 3. Large porphyrin pigment concretion (arrowhead) within a duct. x145 4. Secretory alveoli with type II cells (T2) and type I cells (T1). Mote the capillaries seen between alveoli. arrow, myoepithelial cell nucleus. xl87 PLATE 2 An acetate grid with a 1 cm square lattice superimposed over an electron micrograph used in the morphometic analysis. Points on the grid falling on the particular organelle being observed were counted using a manual counter. L, lumen; M, mitochondria; ME, myoepithelial cell; MV, microvilli; T, tubular cluster; V, vacuole; arrow, concentric lamellae. x11,299 V Í4 í M'lfc fJ^gK. fe' 58 RESULTS CHARACTERISTICS AT SACRIFICE - The characteristics of the individual diabetic and control animals at the time of sacrifice are shown in Tables 3 and 4. Each endocrine gland exhibited a wide range of values and there were no correlations between the gland weights. The range and median values for total body vieight were similar in both diabetics and controls. Blood glucose measurements ranged from 93 mg/dl to 167 ng/dl in controls and 362 mg/dl to 545 mg/dl in diabetics. Mondiabetic control hamsters exhibited a negative test for urine glucose, i.e., Diastix, and a negative urine ketone test, i.e., Ketostix. All diabetic animals displayed 4+ Diastix values and small to negative Ketostix values vihich were in accordance with the criteria for diabetic Chinese hamsters described by Sirek and Sirek (1967). Marked bladder distention vías observed in diabetic animals AC19-24, AC21-01 , AH19-37 and AH20-21. An excessive amount of fluid was seen in the colon of the diabetic hamster AH20-21. Mean values of the gross anatomy characteristics at sacrifice are summarized in Table 5. Control values were greater than those of diabetic animals for the Harderian gland, posterior pituitary and thyroid gland. Mean weights were nearly equal for the total pituitary. Control values were less than those of diabetic animals for total body weight, blood sugar, pineal weight, anterior pituitary weight, adrenal weight and gonadal weight. A statistically significant difference v-ias 59 observed in the blood sugar level, Harderian gland v/eight, adrenal weight and thyroid weight (Graphs 1, 2, 3» 4). A summary of the gross anatomical changes seen in diabetics expressed as a percent of the control values is presented in Graph 5. LIGHT MICROSCOPY OBSERVATIONS - The light microscopy observations of the Harderian glands of the individual diabetic and control animals are shown in Tables 6 and 7. It should be noted that in three of the diabetic animals, i.e., AC19-24, AH20-21 and KI01-30, no type II cells Viere observed. The mean values of the light microscopy observations are summarized in Table 8. Diabetic animals viere found to have a statistically significant decrease in alveolar size, the number of cells per alveolus, cell radius and the number of type II cells per alveolus (Graphs 6, 7, 8, 9). A summary of the structural decrease seen in diabetics at the light microscopic level expressed as a percent of the control values is presented in Graph 10. Observations of nuclear position in those cells whose nuclei were discernible revealed a greater number of nuclei basally located in the control animals than in the diabetics (Plate 1). In the case of diabetic animal AH20-21, almost all nuclei were centrally located within the cells. Controls showed a larger number of binucleated cells as compared with the diabetics. ELECTRON MICROSCOPY MORPHOMETRIC OBSERVATIONS - The electron microscopy observations of type I cells, type II cells and cells with cleft structures for the individual diabetic and control animals are shown in Tables 9, 10 and 11 respectively. It should be noted that the secretory cells of the diabetic animal AH20-21 had very few vacuoles 60 and a dense cytoplasm. Thus, this animal did not fit the pattern for either type I or type II cells and had to be classified separately (Table 12). The mean values of the electron microscopy observations are summarized in Table 13. Type I cells in diabetic animals were found to have increased relative vacuole area to total cytoplasm, relative mitochondria area to total cytoplasm and relative concentric lamellae area to total cytoplasm as compared to control values. Type II cells in diabetic animals were reported to have decreased relative vacuole area with increased relative mitochondria area and relative concentric lamellae area as compared to control values. Cells with cleft structures in diabetic animals were reported to have decreased relative vacuole area with increased relative cleft structure area, relative vacuole plus cleft structure area, relative mitochondria area and relative concentric lamellae area to total cytoplasm. The relative vacuole area, mitochondria area and concentric lamellae area for each cell type are compared in Graphs 11, 12 and 13, respectively. ELECTRON MICROSCOPY HISTOLOGICAL OBSERVATIONS - The Harder!an glands of diabetic animals ranged from having very few vacuoles in animal AH20-21 (Plate 4) to extensively vacuolated tissue with many small round vacuoles and thin strips of cleft structures in animal AC19-24 (Plate 5). In both diabetics and controls, concentric vacuoles were observed to be either bound with a unit membrane or having a punctulate appearance with no limiting membrane. Coalescing vacuoles were frequently noted in both diabetics and controls. In control animals, membranes seemed to be in a more dynamic state as evidenced by frequent observations of diffuse membranes. 61 The perinuclear and basal regions of the secretory cells were filled with well-developed networks of tubular profiles of smooth endoplasmic reticulum. The rough endoplasmic reticulum was poorly developed and only a few flattened Golgi complexes were found. These features were consistent in both diabetics and controls. The mitochondria in diabetic animal AH20-21 were small and predominately spherical (Plate 4). The mitochondria of other diabetic animals were similar to those of controls ranging from spherical or ovoid to filiform in shape and tended to be scattered throughout the cytoplasm. In some cells, increased numbers of mitochondria v;ere observed in the perinuclear cytoplasm in both diabetics and controls. In diabetic animals AC19-24 and KI01-30, mitochondria in light cells had a normal appearance while those in dark cells were swollen with reduced cristae (Plate 6). Small mitochondria were seen in control animals AVI3-63 and AVI4-11. These two control animals also showed some evidence of disruption of the mitochondrial membranes and swollen mitochondria. It should be noted that these two controls were the only control animals in which cleft structures were observed. Control animals had more extensive and more numerous interdigitation of cytoplasmic processes than diabetics. The luminal surfaces of many secretory cells of the controls had areas with a smooth appearance while diabetics generally had uniform projections of microvilli surrounding the lumen. Membrane "ghosts" were found in the lumen of control animals to a far greater extent than in diabetics (Plates 7,8). CONCENTRIC LAMELLAE - Membranous lamellar sturctures were observed dispersed throughout the cytoplasm of secretory cells of the 62 Harderian glands in both control and diabetic animals. In each cell type, i.e., type I , type II and cells with cleft structures, the relative area of concentric lamellar structures in diabetics exceeded that of controls. The relative area of these structures for control type I and type II cells was nearly the same while in diabetics, the relative area in type I cells was greater than that of type II cells. Cells with cleft structures had a much higher relative area of concentric lamellae than either type I or type II cells for both controls and diabetics. The membranes are frequently associated with vacuoles. Some of the lamellar structures were observed in close association to or continuous with mitochondria. Others were observed near smooth and rough endoplasmic reticulum (Plates 11, 12). The membranes vary in extent and configuration. Plate 13 shows several concentric spirals. Mote the dilations at the periphery and subsurface cisternae. Single membrane bound concentric structures were frequently encountered. CLEFT STRUCTURES - Cleft structures were seen in the secretory cells of control animals AVI3-63 and AVI4-11 (Plate 9) and in diabetic animals AC19-24, AC21-01, AH20-21 and KI01-30 (Plates 10, 11). Cleft structures had a flattened plate-like shape with curved or semicircular configurations. These structures were frequently observed to be in multiple stacks and anastomosed with each other or bifurcated. Individual clefts appeared in various widths. Each cleft was composed of a light, narrow space and a unit membrane was frequently seen lining one side of the space (Plate 14). The relative concentric lamellae area to total cytoplasm of those 63 cells with cleft structures in both diabetics and controls exceeded those values for type I and type II cells as Table 13 indicates. For the controls, the relative vacuole plus cleft structure area to total cytoplasm was less than those values for type I and type II cells while relative mitochondria area was similar. For the diabetics, the relative vacuole plus cleft structure area was less than that for type I cells and greater than that of type II cells. The relative mitochondria area showed the opposite relationship. MAST CELLS - Immature mast cells were frequently seen in the interstitial tissue of the Harderian gland of diabetic animals. This was indicated by a relatively conspicuous Golgi complex and few granules. Other mast cells were seen to have predominately small, dark homogeneous granules. In control animals, the cytoplasm showed numerous granules of varying size and density from small, dark homogeneous granules to highly pronounced stippled granules. Areas of coalescence of the stippled granules were frequently observed. TABLE 3 DIABETIC ANIMAL CHARACTERISTICS AT SACRIFICE Subline Age Duration of Duration of Diastix Ketostix Body Blood (Days) Diabetic State Ketonurla Weight Sugar (Days) (Days) (Grams) (mg/dl) AC19-24 825 771 174 •*4 small 31 545 AC21-01 484 415 187 +4 small 27 472 AH19-37 515 466 143 +4 negative 32 505 AH20-21 474 424 227 +4 negative 38 545 BBOI-16 609 568 187 +4 negative 39 374 KlOl-30 454 427 181 +4 small 30 362 TABLE 3 (continued) DIABETIC ANIMAL CHARACTERISTICS AT SACRIFICE Subline Hazderian Pineal Posterior Anterior Ibtal Adrenal Thyroid Gonadal Weight Weight Pituitary Pituitary Pituitary Weight Weight Weight (ing) (ing) Weight Weight Weight (ng) (mg) (mg) (mg) (ing) (ing) AC19-24 66,81 0.23 0.31 2.65 2,96 7.08 ____ 3.^ AG21-01 78.24 0.33 2,28 2.61 7.67 2.93 6.43 AH19-37 91.87 0.59 0.52 3.80 4.32 7.40 4,34 10.50 AH20-21 78.88 0.95 0.70 3.48 4.18 7.18 5.32 7.31 BBOl-16 64,42 0,16 0,11 2.74 2.85 6,80 3.95 6.94 KIOl-30 82.84 0.12 0,08 2.77 2.85 6,26 3.75 9.80 TABLE 4 CONTROL ANIMAL CHARACTERISTICS AT SACRIFICE Subline Age DLastix Ketostlx Body Blood (Days) Weight Sugar (grams) (mg/di) AVO9-7O 84l negative 38 93 AVI3-6O 497 — negative 32 167 AV13-63 502 — negative 27 98 AV14-11 515 — negative 29 128 AV14-19 486 — negative 27 103 AVI3-76 819 — negative 29 154 TABLE 4 (continued) CONTROL ANIMAL CHARACTERISTICS AT SACRIFICE Subline HaiJderian Pineal Posterior Anterior Total Adrenal Thyroid Gonadal Weight Weight Pituitary Pituitary Pituitary Weight Weight Weight (ms) (mg) Weight Weight Weight (mg) (mg) (mg) (mg) (mg) (mg) AVO9-7O 77.3^ 0,18 0.19 1,67 1,86 6.98 6.83 6,62 AVI3-6O 96.79 0.37 0,26 2,82 3.08 6,67 6,44 5.85 AV1>63 80,84 0,81 0,92 3.74 4,66 5.69 6.57 4,74 AV14-11 98.39 0,54 1.19 4,25 6.63 5.14 8.97 AV14-19 92,87 0,15 0.14 2.00 2,l4 4,78 5.15 6.23 AVI3-76 101.17 0,18 0,18 2.67 2.85 3.5^ 4,26 5.90 TABLE 5 SUMMARY OF DIABETIC AND CONTROL ANIMAL CHARACTERISTICS AT SACRIFICE?- Ilabetic (n) Control (n) Body Weight (grams) 31.17+1.5^ (6) 30,33+1.71 (6) Blood Sugar (mg/dl) ^67,17+33.32 (6)* 123.83+12,71 (6) Harder!an Weight (mg) 77.18+4,17 (6)^HH^ 91.23+4,02 (6) Pineal Weight (mg) 0.4l+p,l6 (5) 0,37±P,11 (6) Posterior Pituitary 0,34+p,10 (6) 0,48+0,19 (6) Weight (mg) Anterior Pituitary 2,95±p,23 (6) 2,86+0,40 (6) Weight (mg) Total Pituitary 3.30+0,31 (6) 3.34±0,58 (6) Weight (mg) Adrenal Weight (mg) 7,07+0,20 {6)**** 5.72+0,55 (6) Thyroid Weight (mg) 4,06^0,39 (5)^H^ 5,73+0.42 (6) Gonadal Weight (mg) 7,^10+1.04 (6) 6,39+0.58 (6) Eîach Value represents mean + SEM ?Statistically different from control values iP<0,00l) ??Statistically different from control values (P<0,01^ ???Statistically different from control values iP<0,02) +«+*Statistically different from control values (P<0,03) TABLE 6 DIABmC ANIMAL LIGHT MICROSCOPY OBSERVATIONS^ Subline Alveolar Number Cell Niomber of Size of Cells Radius Type II Cells (urn) Per Alveolus (urn) Per Alveolus AG19-24 5^.90±1.35 11.55±P.44 I7.25±p.55 0,00 AC21-01 52,9640,90 ir»96+p,21 16,13+0,48 1.13 AHL9-37 48,C%P,75 11,60+0,12 14,68+0,37 1,40 AH20-21 46,2^1,16 12,00+0,47 14,67+0,52 0,00 BBOl-16 46,30+1,16 10,93+0.27 13,60+0,42 1.13 KlOl-30 54,52+1,23 11.80+0,24 15.%P.75 0,00 ^Each value represents mean + SiM TABLE 7 GONTBOL ANIMAL LIGHT MICROSCOPY OBSERVATIONS^ Subline Alveolar Number Cell Number of Size of Cells Radius Type II Cells (urn) Per Alveolus (urn) Per Alveolus AVO9-7O 58.32+2,15 17,08^3.94 18,00+0.74 0.96 • AVI3-6O 6i.05±3.97 15.73+1.15 16,09+0,87 2.27 AVI3-63 73.50±3.^8 13.80+0,78 22.20+0.93 1.00 AV14-11 62.73±2.^8 16.55±P»97 18.25+0,70 1.85 AV14-19 67,48+2.14 18,24+0,78 21.76+0,63 2.72 AVI3-76 58,17+1.61 16,92+0,72 17.5^.68 5.04 ^Each value represents mean + SM TABLE 8 SUI'IMARY OP DIABETIC AND CONTROL ANIMAL LIGHT MICROSCOPY OBSERVATIONS^ Diabetic (n) Control (n) Alveolar Size (urn) 50.49+1.67 (6)* 63.54+2.43 (6) Number of Cells Per Alveolus 11.64+0.16 (6)* 16.39+0.62 (6) Cell Radius (urn) 15.36+0.53 (6)^ 18.97+1.00 (6) Number of Type II Cells 0.61+0.28 (6)*** 2.31tp.62 (6) Per Alveolus Each value represents mean + SM *Statlstlcally different from control values ÍP **Statistically different from control values {p<<00,t00005l5J ***Statistlcally different from control values (P<0t02) TABLE 9 DIABETIC AND CONTBOL ANIMAL ELECTRON MICROSCOPY OBSERVATIONS OF TYPE I CELLS^ Diabetic Relative Vacuole Area Relative Mitochondria Area Relative Concentric Lamellae Area Subllne To Total Cytoplasm To Total Cytoplasm To Total Cytoplasm AC19-24 0,422+0,04 0,078+0,00 0,087+0,01 AC21-01 0,49040,07 0,109+0,02 0.035+0,01 BÏÏ01-16 0.313±P.03 0.095+0,02 0,061+0,01 KIOI-3O 0,362+0,00 0,14040,01 0,046+0,01 Control Subline AVO9-7O 0,381+0,06 0,04ii^,01 0,027+0,01 AVI3-6O 0.371+0,02 0,06040,00 0,010+0,00 AVI3-63 0.353±P*02 0,108+0,01 0,028+0,00 AV14-11 0,36540,07 0,106+0,01 0,028+0,00 AV14-19 0,260+0,02 0.123+0,00 0,065+0,01 AVI3-76 0,42940,04 0,093+0.02 0,049+0,01 ^Each value represents mean + SEl^I TABLE 10 DIABETIC AND CONTROL ANIMAL ELECTRON MICROSCOPY OBSERVATIONS OF TYPE II CELLS^ nabetic Relative Vacuole Area Relative Mitochondria Area Relative Concentric Lamellae Area Subline To Total Cytoplasm To Total Cytoplasm To Total Cytoplasm AG21-01 0.393±P.17 0,113+0,02 0,04l+p,02 BBOI-16 0,222+0,02 0ol65+p,0l 0,057^,00 Control Subline AVO9-7O 0,516+0,04 0,02^.01 0.014+0,00 AVI3-6O 0.339±p.03 0.093+0.03 0,024+0.00 AVI3-63 0.375+0.00 0.098+PO00 0,045+0,00 AV14-19 0.332+0.04 0.089+0.00 0,067+0,01 AVI3-76 0.257+0.01 0,140+0.02 0.037+0,00 ^Each value represents mean + SEI'I TABLE 11 DIABETIC AND CONTHDL ANIMAL ELECTRON MICROSCOPY OBSERVATIONS OF CELLS WITH CLEFT STRUCTURES^ Diabetic Relative Relative Relative Relative Relative Subline Vacuole Cleft Structure Vacuole Plus Mitochondria Concentric Area To Total Area To Total Cleft Structure Area To 'Total Lamellae Cytoplasm Cytoplasm Area To Total Cytoplasm Area To Total Cytoplasm Cytoplasm AC19'-24 0,169+0,04 0,186+0,04 0.355 0.173+0,11 0,221+0,10 AC21-01 0,178+0,00 0,178+0,00 0.356 0,156+0,00 0,044+0,00 AH20-21 0,225+0.16 0,310+0,17 0.535 0,112+0,00 0,128+00O3 KIOl-30 0,180+0,00 0,062+0,00 0,242 0.073+0,00 0,146+0,00 Control Subline AVI3-63 0.211+0,00 0,053+0,00 0,264 0,094^,00 0,053+0.00 AV14-11 0,272+0,00 0,108+0,04 O.38O 0,065+0,01 0,056+0,00 ^Each value represents mean + SHI TABLE 12 ELECTRON MICROSCOPY OBSERVATIONS OP THE SECRETORY CELLS OF DIABETIC ANIMAL AH20-21^ Relative Vacuole Area Relative Mitochondria Area Relative Concentric Lamellae Area To Total Cytoplasm To Total Cytoplasm To Total Cytoplasm 0,058+0.03 0.104+0,02 0.125+0,02 ^Eiach value represents mean + SEM TABLE 13 SUMMARY OF DIABETIC AND CONTROL ANIMAL ELECTRON MICROSCOPY OBSERVATIONS^ Type I Cells Diabetic (n) Control (n) Relative Vacuole Area 0.39740.04 (4) 0.360+0,02 (6) To Total Cytoplasm Relative Mitochondria Area 0.106+0.01 (4) 0,089+0,00 (6) To Total Cytoplasm Relative Concentric Lamellae Area 0,057+0.01 (4) 0,035+0,00 (6) To Total Cytoplasm Tÿpe II Cells Relative Vacuole Area 0,308+0.09 (2) 0,364+0.04 (5) To Total Cytoplasm Relative Mitochondria Area 0,139+0.03 (2) 0,089+0,02 (5) To Total Cytoplasm Relative Concentric Lamellae Area 0.049+0,01 (2) 0,037+0.01 (5) To Total Cytoplasm ^Each value represents mean + SE4 TABLE 13 (continued) SUI-IMARY OF DIABETIC AND CONTROL ANIMAL ÍLBCTRON MICROSCOPY OBSERVATIONS^ Cells With Cleft Structures Diabetic (n) Control (n) Relative Vacuole Area 0,188+0,01 W 0 • i0 (2) To Total Cytoplasm Relative Cleft Structure Area 0,l84+p,05 W 0 •01 •0 (2) To Total Cytoplasm Relative Vacuole Plus Cleft Structure Area 0,372+0,06 (4) 0,322+0,06 (2) To Total Cytoplasm Relative Mitochondria Area 0,129+0,02 (4) 0,080+0,01 (2) To Total Cytoplasm Relative Concentric Lamellae Area 0,135lP,04 (^) 0,05540,00 (2) To Total Cytoplasm ^Each Value represents mean + SM GRAPH 1 BLOOD GLUCOSE LEVELS / DIABETIC AND CONTROL HAMSTERS GRAPH 2 HARDERIAN WEIGHT DIABETIC AND CONTROL HAMSTERS GRAPH 3 ADRENAL GLAND WEIGHT DIABETIC AND CONTROL HAMSTERS AOREHAL WQGHT (mg) oiAaenc CONTROL GRAPH 4 THYROH) GLAND WEIGHT DIABETIC AND CONTROL HAMSTERS / GRAPH 5 ANATOMICAL CHANGES SEEN IN DIABETIC HAMSTERS EXPRESSED AS A PERCENT OF THE CONTROL VALUES LEGEND: BOO WT - BODY WEIGHT HARO - HARDERIAN GLANO WEIGHT RNEAL - PINEAL GLANO WEIGHT PO PIT - POSTERIOR PriUlTARY WEIGHT AN PIT - ANTERIOR PITUITARY WEIGHT TO PIT - TOTAL PITUITARY WEIGHT AOREN - ADRENAL GLANO WEIGHT THYR - THYROID GLAND WEIGHT GONAD - GONAOAL WBGHT GRAPH 6 ALVEOLAR SIZE DIABETIC AND CONTROL HAMSTERS ALVEOLAR 9ZE (urn) OlABCnC CONTROL GRAPH 7 NUMBER OF CELLS PER ALVEOLUS DIABETIC AND CONTROL HAMSTERS CELL NUMBER nABETIC CONTROL GRAPH 8 CELL RADIUS DIABETIC AND CONTROL HAMSTERS CELL RAOaJS (urn) NUMBER OF TYPE H CELLS PER ALVEOLUS DIABETIC AND CONTROL HAMSTERS CELL NUMBER OMBCTIC CONTROL GRAPH 10 STRUCTURAL DECREASE SEEN IN DIABETIC HAMSTERS EXPRESSED AS A PERCENT OF THE CONTROL VALUES PERCENTAGE LEGEND: HARO WT - HAROERIAN GLAND WBGHT ALV SIZE - ALVEOLAR SIZE # OF CELLS ~ NUMBER OF CELLS PER ALVEOLUS CELL RAO - CELL RADIUS # TYPE II - NUMBER OF TYPE II CELLS PER ALVEOLUS GRAPH 11 RELATIVE VACUOLE AREA OF TYPE I CELLS, TYPE n CELLS AND CELLS WITH CLEFT STRUCTURES .45 TYPE I TYPE II VAC CLEFT VAC+CLEFT LEX»ID: TYPE I - REUTIVE VACUOLE AREA IN TYPE I CELLS TYPE II - REUTIVE VACUOLE AREA IN TYPE II CELLS VAC - REUTIVE VACUOLE AREA IN CELLS WITH CLEFT STRUCTURES CLEFT - REUTIVE CLEFT STRUCTURE AREA IN CELLS WITH CLEFT STRUCT. VAC+CLEFT - REUTIVE VACUOLE AREA PLUS CLEFT STRUCTURE AREA IN CELLS WITH CLEFT STRUCTURES GRAPH 12 RELATIVE MITOCHONDRIA AREA OF TYPE I CELLS, TYPE n CELLS AND CELLS WITH CLEFT STURCTURES LEGEND: TYPE I - RELATIVE MITXJCHONDRIA AREA !N TYPE I CELLS TYPE II - REUTIVE MITOCHONDRIA AREA IN TYPE II CELLS CLEFT STRUC - RELATIVE MITOCHONDRIA AREA IN CELLS WITH CLEFT STRUCTURES GRAPH 13 RELATIVE CONCENTRIC LAMELLAE AREA OF TYPE I. TYPE n CELLS AND CELLS WITH CLEFT STRUCTURES LEGEND; TYPE I - RELATIVE CONCENRTTC LAMELLAE AREA IN TYPE I CELLS TYPE II - REUTTVE CONCENTRIC LAMELLAE AREA IN TYPE II CELLS CLEFT STRU - RELATIVE CONCENTRIC LAMELLAE AREA IN CELLS WITH CLEFT STRUCTRUES GRAPH U LINEAR REGRESSION OF RELATIVE MITOCHONDRIA AREA VERSUS RELATIVE VACUOLE AREA REUTIVE MITOCHONDRIA AREA TO TOTAL CYTOPIASM RELATIVE VACUOLE AREA TO TOTAL CYTOPLASM LEGEND: REPRESENTS CONTROL VALUES (CORRELATION COEFFICIENT ~-0.608) REPRESENTS DIABETIC VALUES (CORRELATION COEFFICIENT-0.417) PLATE 3 Light micrograph verifying the compound tubuloalveolar structure of the Chinese hamster Harderian gland. A, alveolus; L, lumen; ST, secretory tubule. x239 PLATE 4 Secretory cells of diabetic animal AH20-21. The cell at the upper right of the micrograph is a cell with cleft structures (C). Note the abundance of small mitochondria (M), polyribosomes (P) and rough endoplasmic reticulum (R). Also note few vacuoles (V) in the cytoplasm. JC, junctional complex; L, lumen; N, nucleus; arrow, concentric lamellae. x9,8l8 PLATE 5 Secretory cells of diabetic animal AC19-24. The cell in the upper right of the micrograph is a cell with cleft structures (C). Note the abundance of small vacuoles (V), tiny cleft structures, and swollen mitochondria (M). IN, interstitial area; ME, myoepithelial cell; N, nucleus; arrow, concentric lamellae. x10,327 PLATE 6 High magnification of a secretory cell of diabetic animal AC19-24. Note the abundance of swollen mitochondria (M) with disrupted cristae (CR). Also note the abundance of small concentric lamellae (arrow) and polyribosomes (P). V, vacuole. x62,4l0 PLATE 7 Secretory alveolus at the electron microscopic level. F, fenestrated capillary; GH, membrane ghost; JC, junctional complex; L, lumen; M, mitochondria; MC, mast cell; ME, myoepithelial cell; V, vacuole; arrow, concentric lamellae. x10,489 PLATE 8 Type I (Tl) and type II (T2) cells. Note the vacuole in the process of exocytosis at the upper left of the micrograph. Also note extensive interdigitation (I) and smooth endoplasmic reticulum (S). G, Golgi complex; L, lumen; ME, myoepithelial cell; M, nucleus; arrow, concentric lamellae. x8,876 PLATE 9 Cells with cleft structures (C) in control animal AVI3-63. Note the abundance of filiform mitochondria (M) which appear to be disrupted, IN, interstitial area; L, lumen; MV, microvilli; N, nucleus; NÜ, nucleolus; R, rough endoplasmic reticulum; V, vacuole; arrow, concentric lamellae. x9,425 PLATE 10 Cells with cleft structures (C) in diabetic animal AH20-21. Note the abundance of rough endoplasmic reticulum (R) and relatively few, small mitochondria (M). ME, myoepithelial cell; N, nucleus; V, vacuole; arrow, concentric lamellae. x9,623 PLATE 11 Cell with cleft structures in diabetic animal KI01-30. Note the association of concentric lamellae (arrow) with mitochondria (M) and vacuoles (V). Also note the association of cleft structures (C) with vacuoles. JC, junctional complex; L, lumen; liV, microvilli. x19,888 PLATE 12 Proposed sequence of the association of concentric lamellae with mitochondria and vacuoles. 1. Elongated mitochondrion (M) curved in a semicircle, arrow, concentric lamellae. x26,250 2. Mitochondrion (M) whose membrane is continuous with a concentric lamellar structure (arrow). x23f088 3. On the far left, a mitochondrion (M) whose membrane is continuous with a concentric lamellar structure (arrow) with a vacuole (V) forming within the semicircle. On the far right, a concentric lamellar structure surrounding a mitochondrion and a vacuole. In the center, a vacuole with possible remnants of a mitochondrion surrounded by a closely associated concentric lamellar structure. Also note the close association of cleft structures (C) with vacuoles. x35,121 PLATE 13 Concentric lamellae of varying configurations. 1. Concentric membranous structure (arrow) with two layers. M, mitochondria; V, vacuole. x79,428 2. Concentric membranous structure (arrow) with multiple layers. x177,133 PLATE 14 Cleft structures (C) in association with vacuoles (V). P, polyribosomes; R, rough endoplasmic reticulum; S, smooth endoplasmic reticulum; arrow, concentric lamellae. x24,511 PLATE 15 Exocytosis at the lumen by a secretory cell. Mote the abundance of smooth endoplasmic reticulum (S) and polyribosomes. Vacuoles were observed coalescing near the lumen. Vacuoles were often observed to have a well defined limiting membrane (VI) while many had a diffuse membrane (V2). L, lumen; M, mitochondria; MV, microvilli; arrow, concentric lamellae. x55,015 PLATE 16 Porphyrin pigment concretion (arrowhead) in the lumen (L) of secretory alveolus. I, interdigitation; M, mitochondria; N, nucleus; V, vacuole; arrow, concentric lamellae. x15,278 PLATE 17 Exocytotic vesicles (EX) at the basal end of a secretory cell (SC) indicating a possible endocrine secretory mechanism. CM, cell membrane; IN, interstitial area; ME, myoepithelial cell. xl88,235 PLATE 18 Cytoplasmic extension of a myoepithelial cell (ME) with numerous endocytotic vesicles (EN). CM, cell membrane; IN, interstitial area; SC, secretory cell. x74,734 PLATE 19 Fenestrated capillary (F) in the interstitial area accompanied by a mast cell (MC) and myoepithelial cells (ME). SC, secretory cell. x6l,043 122 DISCUSSION The results of this study may be categorized in the following manner: (1) the gross anatomical differences and their importance as related to the Harderian gland; (2) the possible disturbance of the secretory state as indicated by structural changes seen in the diabetic animals; (3) the possible significance of reduced numbers of type II cells in the Harderian gland in the diabetic animals; (4) the presumed importance of concentric lamellae in the cytoplasm of secretory cells; (5) an examination of cells with cleft structures; and, (6) an analysis of possible mast cell differences. GROSS ANATOMY - Johnson and Sidman (1979) gathered endocrine and reproductive data suggesting that abnormal hypothalamic function may be the immediate cause of some of the physiologic abnormalities observed in diabetic mice. Mice homozygous for the autosomal recessive mutation diabetes (db) have physiological abnormalities that include hyperglycemia, hyperinsulinemia, obesity, thermoregulatory disturbances and sterility of both sexes. Hypoplastic vaginal epithelia, uteri and ovaries can respond comparably to control tissues upon hormonal stimulation. Gonadotropin release from the pituitary gland appeared to be depressed in female mutant mice but responded normally to exogenous gonadotropin releasing hormone (GnRH) in both sexes. Hypothalamic GnRH content was greater than normal in adult female mutant mice suggesting that in the females, GnRH release may be inadequate with secondary blunting of pituitary function. 123 Boas and Bates (1954) reported that hypophysectomy decreased the v/eights of Harderian glands. The specific pituitary tropic hormone(s) responsible for maintenance of the Harderian glands has (have) not been determined. Boas and Bates (1954) and Smelser (1943) suggested that thyrotrophin or somatotrophin, or both, as well as thyroxine, may be required. Hoffman (1971) suggested that in castrated male golden hamsters, increased secretion of gonadotrophins, rather than reduced male hormone secretion, i.e., primarily testosterone, leads to structural and functional modifications of the Harderian glands toward the female type. Hypophysectomized animals had significantly reduced porphyrins in the Harderian glands (Hoffman, 1971). In the present study, the posterior pituitary weight is reduced while the anterior pituitary weight is greater in the diabetic versus the control. The total pituitary weight is nearly the same in the diabetic and control. It should be noted that the weight difference in the subdivisions of the pituitary may not have been recognized in many previous literature studies where only the total pituitary was weighed. The posterior pituitary (neurohypophysis), a neurohemal organ, is responsible for releasing vasopressin (ADH, antidiuretic hormone) and oxytocin. The anterior pituitary is responsible for secreting somatotrophin (STH, growth hormone), adrenocorticotrophin (ACTH), thyrotrophin (TSH), and gonadotrophins, i.e., luteinizing hormone (LH), follicle-stimulating hormone (FSH) and prolactin (lactogenic hormone, luteotrophin). Somatotrophin (STH), when administered to mammals: (1) tends to produce hyperglycemia, thus aggravating the diabetic state (”diabetogenic” effect); (2) inhibits the action of insulin 124 (”anti-insulin" effect); (3) increases muscle glycogen when given to hypophysectomized subjects ("glycostatic" effect); and, (4) produces permanent diabetes mellitus in certain species when given over prolonged periods possibly by exhaustion of the B-cells in the pancreatic islets. Ilhereas STH alone has little effect on such target organs as the adrenal cortex, thyroid and gonads, it markedly enhances the effectiveness of the trophic hormones, i.e., ACTH, TSH, FSH, and LH, specific for these organs (Turner and Bagnara, 1976)- The adrenal weight v/as found to be significantly greater in the diabetic versus control in the present study. Turner and Bagnara (1976) describe the functional status of the adrenal cortex as regulated by ACTH. Hypophysectomy results in shrinkage of the cortex while the medulla is not influenced. Permanent (metacorticoid) diabetes may be produced in normal rats, rabbits and guinea pigs by administering such adrenal cortical hormones as cortisol or cortisone. These hormones may permanently damage the B-cells of the pancreas. The 11-oxygenated steroids of the cortex probably influence blood glucose in two ways. Firstly, they inhibit the incorporation of amino acids into protein and stimulate protein mobilization, thus augmenting the supply of gluconeogenic materials. Secondly, these steroids may retard the utilization of glucose by peripheral tissues. The adrenal medulla secretes epinephrine and norepinephrine. Stimulation of a particular area of the hypothalamus causes mainly epinephrine to appear in the adrenal venous blood. Stimulation of another area promotes the appearance of norepinephrine in the venous effluent (Turner and Bagnara, 1976). In the present study, the thyroid gland was found to weigh 125 significantly less in the diabetic versus the control. The atrophy of the Harderian gland, its increased relative vacuole area in type I cells and cells with cleft structures, and the decrease in relative vacuole area in type II cells may be influenced by decreased secretions from the thyroid. Boas and Bates (1954) reported that thyroidectomy in the rat causes Harderian gland regression but not if the animals were treated with thyroxine. In the female, the thyroid gland also affects the porphyrin content. Wetterberg and co-workers (1970) noted that treatment of neonatal animals with thyroxine accelerates development of the Harderian gland. In the rat, growth of the Harderian glands after eye opening is correlated ontogenetically with an increase in thyroxine secretion (Gray et al.. 1982). Jostes (1976) treated intact rats and guinea pigs with thyroid stimulating hormone (TSH). The type II cells of the rat showed an increase in vacuoles and hypertrophy of the Harderian gland with widening of tubuli and acini was reported in guinea pigs. Gonadal weight v/as found to be increased in the diabetic versus the control in this study. Estrogen has a curative effect when administered to alloxan induced diabetic rats. This hormone causes an increase in the number of pancreatic islets and B-cells. This stimulating action of estrogen on the pancreas is especially pronounced when it acts in the presence of insulin (Turner and Bagnara, 1976). This raises the question of whether cessation of ovarian function affects maturity onset diabetes. Morphologic abnormalities in the retina, pancreas and kidney of the diabetic Chinese hamster have shown a relationship to length and/or severity of diabetes (Soret et. al., 1974). It has been suggested that 126 the incidence of pathologies in the small Intestine of the diabetic hamster may be related to severity of the disease (Diani et al.. 1979). The manifestation of neuropathology has been correlated with duration of diabetes in the Chinese hamster (Schlaepfer et al.. 1974). Further studies of the histological differences between the endocrine glands in the diabetic and control animals may provide evidence for the interpretations suggested by this study. A histochemical study of porphyrin content would also be of interest. SECRETORY STATE - A possible disturbance in the secretory mechanism in the alveolar cells of the diabetic animals was suggested by a linear regression plotting relative mitochondria area to total cytoplasm versus relative vacuole area to total cytoplasm (Graph 14). The negative or downward slope of the graph for control values indicates a relationship that would be expected for actively secreting glandular tissue, i.e., more actively secreting cells have reduced vacuole area and increased mitochondria area and vice versa. Values for diabetic animals do not fall in close association to the control regression line. This indicates a possible disturbance in the secretory mechanism in the alveolar cells of the diabetic animals. The Karderian gland of diabetic animal AH20-21 is probably the least active of the Harderian glands observed due to very few vacuoles and centrally located nuclei. TYPE II CELLS - Manipulation experiments using adult golden hamsters have shown that type II cells in the male can be converted into type I cells and that type I cells in the female can be converted to type II cells (Hoffman, 1971; Clabough and Norvell, 1974). In male hamsters, type II cells appear to be transformed to type I cells by 127 accumulating more lipid, thus giving rise to larger lipid droplets with less uniform size (Hoffman, 1971). Blinding induces a highly functional pineal gland, which appears to produce gonadal atrophy or abnormalities by inhibiting or modifying gonadotrophin release. One might postulate, therefore, that the male type Harderian gland is maintained whenever the pineal gland is hyperfunctional. However, in blinded and pinealectomized male golden hamsters, where there is also no conversion of male type of histology to the female type, a pathway exclusive of the pineal gland prevents the conversion, lihether the pathway nay be a direct effect via neural mechanisms or indirect via hormonal imbalance cannot be concluded (Hoffman, 1971). From the data presented in this study, there appears to be little correlation between the presence and/or relative abundance of type II cells and the pineal gland weight. In the present study, the diabetic animals had significantly reduced numbers of type II cells. Mo type II cells were seen in three of the six diabetic hamsters observed in this study. CONCENTRIC LAMELLAE - The significance of the membranous lamellar structures in unknown. These structures vary in extent and configuration as well as location and association with other organelles within the secretory cells. This may be suggestive of variations in function. Bucana and Nadakavukaren (1972) suggested that membranous "juxtanuclear structures” may serve as a site of membrane formation. An increase in electron-dense material in the "juxtanuclear structures” of Harderian glands in postnatal female hamsters may be lipids and proteins necessary for membrane formation and corresponds with the appearance of membranes 128 The structures observed in the present study are comparable to those reported in the pineal of the normal and diabetic Chinese hamster (McNeill and Smith, 1982), diabetic mutant mouse and absent in littermate controls (McNeill, 1978), mole (Pevet, 1974, 1976), mole-rat (Pevet et al.. 1976), hedgehog (Pevet, 1976) and noctule bat (Pevet et. al.. 1977). "Since the diabetic mutant mouse is infertile, the presence of these membranes in the diabetic mouse and not in the littermate controls led to the conclusion that the membranes were implicated in some way in either the diabetic condition, infertility or possibly both" (McNeill and Smith, 1982). McNeill and Smith (1982) discussed the analogies betv/een the membranous structures seen in the pineal and annulate lamellae. The function of annulate lamellae is unknown though their morphology is suggestive of secretory capacity. Pineal activity is known to be correlated with gonadal suppression. The presence of these membranous structures in the pinealocytes of diabetic mutant mice and similar membranes in nondiabetic animals support an association of the structures with pineal function vihich is antigonadotropic and, therefore, related to infertility (McNeill, 1978). The endocrine capabilities of the Harderian gland may be a result of indoleamines and it is probable that the Harderian gland is capable of synthesizing all the 5-methoxyindoles found in the pineal (Pevet e¿ al., 198O; Reiter et. al., I98I). Balemans and co-workers (1983) suggest that both melatonin and methoxytryptamine (MT), if released in the circulation by the Harderian gland, act on the pineal causing the release of some other, as yet unknown, factor influencing directly the reproductive endocrine system. Both MT and 5-methoxytryptophol (MTL) are knov/n to be antigonadotropic in the 129 golden hamster. CLEFT STRUCTURE - Kaiho and Ichikavia (I982) coined the term "cleft structure" to describe a peculiar structural feature of the alveolar cells of the Harderian gland. The cleft structures observed were similar in appearance to the "crystalline structure" found in the epithelial cells of human gall bladders with cholesterosis or "cholesterol ester cleft" in lacunal cells of snake skin. An electron opaque material with a laminated, myelin-like appearance was detected in the secretory granules of type II alveolar cells of rat and mouse Harderian glands which were processed with routine preparatory procedures. This material was considered to be phospholipids which are more or less contained in Harderian gland secretion (Watanabe, I98O). Kaiho and Ichikawa (I982) reported that the material in the cleft structures of the Harderian gland of the gerbil was not stained by conventional procedures but appeared to be electron-opaque and myelin-like in structure after treatment v/ith ferrocyanide-reduced osmium tetroxide. It was concluded that fully saturated phospholipids were contained in the material of the cleft structure and are probably different in chemical nature from those reported previously. The electron-opaque materials were detected not only in the cleft structrues but also in the peripheral region inside the limiting membrane of secretory droplets in some occasions. However, these materials are rarely found in mature secretory droplets. The suggestion was made that the secretory droplets may increase in size by wrapping the cleft structures around droplets. The possibility exists that the electron-opaque materials are processed and change their chemical nature during incorporation into granular contents of maturing 130 secretory droplets. Thus, the abundance of cleft structures in the diabetic animals may in some way be related to the reduction in type II cells. If cleft structures represent an immature stage in the development of mature secretory granules, then the present study may reflect a decrease in secretory activity. The decrease in relative mitochondria area to total cytoplasm as well as the decrease in relative vacuole area to total cytoplasm are in agreement with this suggestion as is the gross decrease in Harderian gland weight in the diabetic animal. The increase in relative concentric lamellae area may imply that the concentric lamellae are acting as a storage site for lipids and proteins which may be contributed for membrane formation. The diabetic animals show a proportionally greater increase in concentric lamellae over type I and type II cells than the controls. This may imply a connection with the possible secretory capacity of these structures which may contribute to the diabetic state. MAST CELLS - Payne and co-workers (I982) suggested that a powerful relationship may exist between mast cell numbers and hormonal status in the Harderian gland. The correlation of numbers of mast cells with hormonal status has been reported in several tissues. Mast cells are said to fluctuate during the female cycle and to increase during pregnancy and lactation. Hoffman (1971) reported that thjnr’oxine administration was followed by increased abundance of mast cells. Further study attempting to quantify differences in mast cell numbers and determine whether changes are due to migration of mature cells or cell formation is warranted. 131 SUMMARY 1. The total body weights were similar in both diabetics and controls. 2. The blood glucose levels were significantly higher in diabetics versus controls. 3. Harderian gland weight and thyroid gland weight viere significantly lower in diabetics versus controls. 4. Pineal gland weight, adrenal weight and gonadal weight were higher in diabetics versus controls. 5. The total pituitary weights were similar in both diabetics and controls while the posterior pituitary weight was lower in diabetics versus controls and the anterior pituitary was higher in diabetics versus controls. 6. The Harderian glands of diabetic animals were found to have a significant decrease in alveolar size, the number of cells per alveolus, cell radius and the number of type II cells per alveolus. No type II cells were observed in three diabetic animals. 7. Structural changes seen in the Harderian glands of diabetic animals may indicate possible disturbances of the secretory state. 8. Concentric lamellae vary in extent and configuration as well as location and association with other organelles within the secretory cells of the Harderian gland. Diabetic animals v/ere observed to have increased relative concentric lamellae area to total cytoplasm as compared with controls. 9. Cells with cleft structures were observed in the Harderian glands of diabetic animals to a far greater extent than controls. 10. Immature mast cells were frequently seen in the interstitial tissue of the Harderian glands of diabetic animals. 132 APPENDIX A Earle’s Balanced Salt Solution Constitutents Grams of Constitutents per Liter NaCl 6.80 CaClj 2H;i.O 0.27 KCl 0.40 NaHCO 2.20 MgSOi^THaO 0.20 Glucose 1 .00 Water (ml) 1000 Mean Freezing Point -0.59 Sodium nitrite may be added to the perfusion solutions as a vasodilator. From a stock solution, add 0.01 ml per 100 ml of salt solution. Procedure: The basic salt solution with a pH checked to 7.3 or 7.4, is perfused until the liver blanches partially. This indicates that the blood is being washed out and the fixative may be turned on. The procedure is carried out at room temperature, and the solutions are gravity-fed, at a pressure of about 140 cm water. 133 APPENDIX B Progressive Staining Method For Paraffin Sections Using Hematoxylin and Eosin Fluid Time Toluene 2 minutes Absolute Alcohol 2 minutes 95Í Alcohol 2 minutes 70^ Alcohol 2 minutes Running Tap Water 5 minutes Ehrlich Hematoxylin 3 minutes Running Tap Viater 5 minutes Eosin 25 seconds 70^ Alcohol 1 dip 95% Alcohol 3 dips Absolute Alcohol 3 minutes 1:1 Toluene:Absolute Alcohol 2 minutes Toluene 3 minutes Toluene 2 minutes 134 APPENDIX C Ehrlich Hematoxylin Constitutent Quantity Hematoxylin 2 grams Ammonia alum (NH^Al(SOi),)^ IZHjlO) 3 grams Alcohol, methyl or ethyl 100.0 ml Glycerin 100.0 ml Distilled water 100.0 ml Glacial Acetic Acid 100.0 ml Ripens in 6-8 weeks or may be ripened for immediate use with 0.24 grams sodium iodate. From G. Humason. 1979. Animal Tissue Techniques. 4th ed W.H. Freeman and Co.: San Francisco. 112. 135 APPENDIX D Araldite 6005 Constitutent Quantity Araldite 6005 20 ml Dodecenyl Succinic Anhydride (DDSA) 20 ml Dibutyl Phthalate (DBF) 3.2 ml Benzyldimethylamine (BDMA) 0.8 ml 136 LITERATURE CITED Balemans, M.G.M., Pevet, P. and Vanbenthem, J. 1983. Day/Night Rhythmioity in the Methylating Capacities for Different 5-Hydroxyindoles in the Pineal, the Retina and the Harderian Gland of the Golden Hamster During the Annual Seasons. J. Neural Trans. 5^.: 53-72. Bloom, W. and Fawcett, D.¥. 1968. A Textbook of Histology. 9th ed. Saunders, Philadelphia. 97-110. Boas, N.F. and Bates, R.W. 1954. Role of the Thyroid and Anterior Pituitary Glands in the Maintenance of the Harderian Gland of the Rat. Endocrinology. 55.: 601-612. Bubenik, G.A., Purtill, R.A., Brown, G.A., and Grota, L.J. 1978. Melatonin in the Retina and the Harderian Gland; Ontogeny, Diurnal Variations and Melatonin Treatment. Exp. Eye Res. 21.’ 323-333. Bucana, C.D. and Nadakavukaren, M.J. 1972. Fine Structure of the Hamster Harderian Gland. Z. Zellforsch. 120: 178-187. Bucana, C.D. and Nadakavukaren, M.J. 1972. Innervation of the Hamster Harderian Gland. Science. 175; 205-206. Bucana, C.D. and Nadakavukaren, M.J. 1973. Ultrastructural Investigation of the Postnatal Development of the Hamster Harderian Gland: II. Male and Female. Z. Zellforsch. 142: 1-12. Clabough, J.W. and Norvell, J.E. 1973. Effects of Castration, Blinding and the Pineal Gland on the Harderian Glands of the Male Golden Hamster. Neuroendocrinology. 12.: 344-353. Clabough, J.W. and Norvell, J.E. 1974. Pineal Influence on the Harderian Glands of Female Golden Hamsters. Anat. Rec. 178: 119-126. Derrien, E. and Turchini, J. 1924. Sur L’accumulation D'une Porphyrine Dans La Glande De Harder Des Rongeurs Du Genre Mus Et Sur Son Mode D'excretion. C.R. Soc. Biol. 51: 637-639. Diani, A.R., Grogan, D.M., Yates, M.E., Risinger, D.L. and Gerritsen, G.C. 1979. Radiologie Abnormalities and Autonomie Neuropathy in the Digestive Tract of the Ketonuric Diabetic Chinese Hamster. Diabetologia. H: 33-40. Freund, J.E. 1979. Modem Elementary Statistics. 5th ed. Prentice-Hall, Inc.: Englewood Cliffs, N.J. 247-254, 338-341. Gerritsen, G.C. and Dulin, W.E. 1967. Characteriaztion of Diabetes in Chinese Hamsters. Diabetologia. 3.: 74-84. Gerritsen, G.C., Johnson, M.A., Soret, M.G. and Schultz, J.R. 1974. Epidemology of Chinese Hamsters and Preliminary Evidence for generic heterogeneity of diabetes. Diabetologia. IQ.: 581-586 . 137 Gray, L.E., Kavlock, R.J., Chernoff, N., Ferrell, J., McLanb, J. and Ostby, J. 1982. Prenatal Exposure to the Herbicide 2,4-Dichlorophenyl-p-Nitrophenyl Ether Destroys the Rodent Harderian Gland. Science. 215 ; 293-294. Harder, J.J. 1694. Glandule nova lachrymalis una cum ductu excretorio in cervis and damis. Acta Erudirorium Lipsiae. 49-52. Hilton, F.K., Raque, G.H. and Hilton, M.A. 198I. Changes in Concentrations of Taurine in Murine Harderian Glands and Retinas During Postnatal Development. J. Exp. Zool. 216: 493-495. Hoffman, R.A. 1971. Influence of Some Endocrine Glands, Hormones and Blinding on the Histology and Porphyrins of the Harderian Glands of the Golden Hamsters. Am. J. Anat. 122; 463-478. Hoffman, R.A. and Alieva, J.J. 1982. Cholera Toxin: Mitogen for Harderian Glands of Golden Hamsters. Comp. Biochem. Physiol. 232.: 383-388. Hoffman, R.A. and Jones, J.W. I98I. Concentrations of Metals in the Harderian Gland of Male and Female Hamsters. Comp. Biochem. Physiol. PtA. il: 153-156. Humason, G. 1979. Animal Tissue Techniques. 4th ed. W.H. Freeman and Co., San Francisco. 112. Johnson, D.D., Rudeen, P. K.and O'Steen, W. K. 1979. Photically Induced Experimental Exophthalmos: Role of the Harderian and Pituitary Glands. Invest. Ophthalmol, and Vis. Science, li.: 1280-1285. Johnson, L.M. and Sidman, R.L. 1979. A Reproductive Endocrine Profile in the Diabetes (db) Mutant Mouse. Biol. Repro. 22.: 552-559. Jostes, V.H. 1976. Immunohistochemical Investigations on the Harderian Gland of Rat and Guinea Pig Under the Influence of TSH and Thyrohormones. Acta Histochera. 57: 127-156. Kaiho, M. and Ichikawa, A. 1982. An Application of the Ferrocyanide-Reduced Osmium Tetroxide for the Ultrastructural Analysis of Gerbil Harderian Gland. J. Electron Microscopy 21 : 410-414. Krause, W.J. and Cutts, V.J. 1981. Concise Text of Histology. 1st ed. Williams and Viilkins, Baltimore. 32-33. Lehninger, A.L. 1982. Principles of Biochemistry. Srd ed. VJorth Publishers, New York. 625-627. Lin, U. and Nadakavukaren, H.J. 1979. The Androgenic Effect on the Fine Structure of the Harderian Gland in the Male Hamster. Cell Tissue Res. 198: 119-127. 138 Lin, W. and Nadakavukaren, M.J. 1981. Harderian Gland Lipids of Hale and Female Golden Hamsters. Comp. Biochem. Physiol. PtB. 70 ; 627-630. Lin, W. and Nadakavukaren, M.J. 1981. Aminolevulinate Synthase Activity in the Harderian Gland of Male and Female Golden Hamsters. Comp. Biochem. Physiol. PtB. 71 ; 539-540. Loewenthal, N. 1896. Drusenstudien. I. Die Harder’sche Druse. Int. Monatschr. Anat. Physiol. 13.: 27-65. Luse, S.A., Caramia, F., Gerritsen, G.C. and Dulin, W.E. 1967. Spontaneous diabetes mellitus in the Chinese hamster: An Electron ILLcroscopic study of the Islets of Langerhans. Diabetologia. 3: 97-108. Luse, S.A., Gerritsen, G.C., and Dulin, W.E. 1970. Cerebral Abnormalities in Diabetes Mellitus: An Ultrastructural Study of the Brain in Early Onset Diabetes in the Chinese Hamster. Diabetologia. £: 192-204. McCombs, H.L., Gerritsen, G.C., Dulin, W.E. and Chobanian, A.V. 1974. Morphologic Changes in the Aorta of the Diabetic Chinese Hamster. Diabetologia. IQ.: 601-610. McNeill, M.E. 1978. Membranous Structures in Pinealocytes of the Infertile Diabetic Mutant Mouse (C57Bl/Ks-db/db). J. Neural Trans. 12.: 207-221 . McNeill, M.E. and Smith, C. 1982. Membranous Structures in Pinealocytes of the Chinese Hamster (Cricetulus griseusi: Are They Annulate Lamellae? In The Pineal and Its Hormones. Alan R. Liss, Inc.: 11-19. Norvell, J.E. and Clabough, J.W. 1972. Adrenergic and Cholinergic Innervation of the Hamster Harderian Gland. Science. 178: 1102-1103. Panke, E.S., Reiter, R.J. and Rollag, M.D. 1979. Effect of Removal of the Harderian Glands on Pineal Melatonin Concentrations in the Syrian Hamster. Experientia. 35.^ 1405. Paule, W.J., Hayes, E.R. and Marks, B.H. 1955. The Harderian Gland of the Syrian Hamster. Anat. Rec. 121 : 349-350. Payne, A.P. 1977. Pheromonal Effects of Harderian Gland Homogenates on Aggressive Behavior in the Hamster. J. Endocrinology. 23: 191-192. Payne, A.P., McGadey, J., Moore, M.R. and Thompson, G.G. 1977. Androgenic Control of the Harderian Gland in the Male Golden Hamster. J. Endocr. 25.: 73-82. 139 Payne, A.P., McGadey, J., Moore, M.R. and Thompson, G.G. 1979. Changes in Harderian Gland Activity in the Female Golden Hamster during the Oestrous Cycle, Pregnancy and Lactation. Biochem. J. 173; 597-604. Payne, A.P., McGadey, J., Johnson, H.S., Moore, M.R. and Thompson, G.G. 1982. Mast Cells in the Hamster Harderian Gland: Sex Differences, Hormonal Control and Relationship to Porphyrin. J. Anat. 135 : 451-462. Pevet, P. 1974. The Pineal Gland of the Mole (Taloa eurooaea L.). I. The Fine Structure of the Pinealocytes. Cell Tissue Res. 153; 277-292. Pevet, P. 1976. Correlations Betv/een Pineal Gland and Sexual Cycle. An Electron Microscopical and Histochemical Investigation on the Pineal Gland of the Hedgehog, Mole, Mole-rat and VJhite Rat. Thesis 1. Pevet, P., Kappers, J. Ariens, llevo, E. 1976. The Pineal Gland of the Mole-Rat fSoalax ehrenbermi. Nehring). Cell Tissue Res. 174: 1-24. Pevet, P. , Kappers, J. Ariens, Voûte, A.M. 1977. The Pineal Gland of Nocturnal Mammals. I. The Pinealocytes of the Bat (Nvtalus noctula. Schreber). J. Neural Transm. 40; 47-68. Pevet, P., Balemans, M.G.M., Legerstee, W.C. and Viven-Roels, B. 1980. Circadian Rhymicity of the Activity of Hydroxyindole-O-Methyl Transferase (HIOMT) in the Formation of Melatonin and 5-Hethoxytryptophol in the Pineal, Retina, and Harderian Gland of the Golden Hamster. J. Neural Transmission. US.: 229-245. Pevet, P., Balemans, M.G.M., Legerstee, W.C. and Viven-Roels, B. 1980. Circadian Rhymicity in the Activity of HIOMT in the Formation of Melatonin and of 5-Methoxytryptophol in the Retina, Harderian Gland and Pineal of the Male Hamster.(Abstract). Int. Symposium on Melatonin. September 28-30. Reiter, R.J., Richardson, B.A. and Hurlbot, E.C. 1981. Pineal, Retinal and Harderian Gland Melatonin in a Diurnal Species, the Richardson’s Ground Squirrel. Meurosci. Lett. 22.’ 285-289. Reiter, R.J., Richardson, B.A., Matthews, S.A., Lane, S.J. and Ferguson, B.N. 1983. Rhythms in Immunoreactive Melatonin in the Retina and Harderian Gland of Rats: Persistence After Pinealectoray. Life Science. 32.: 1229-1236. Sakai, T. 1981. The Mammalian Harderian Gland: Morphology, Biochemistry Function and Phylogeny. Arch. Histol. Jap. M.: 299-333. Sakai, T. and Yohro, T. 198I. A Histological Study of the Harderian Gland of Mongolian Gerbils, Meriones meridianus. Anat. Rec. 200 : 259-270. 140 Schlaepfer, W.VÍ., Dulin, W.E. and Gerritsen, G.C. 1974. Segmentai Demyelination in the Distal Peripheral Nerves of Chronically Diabetic Chinese Hamsters. Diabetologia. IQ.: 541-547. Schramm, U. 1980. Lymphoid Cells in the Harderian Glands of Birds. Cell and Tissue Res. 205: 85-94. Shirai, T., Welsh, G.W., and Sims, E.A.H. 1967. Diabetes Mellitus in the Chinese Hamster II: The Evolution of Renal Glomerulopathy. Diabetologia. 3.: 266-286. Shirama, K. 1978. Induction of Persistent Estrus by Constant Light: Effects of Neonatal Constant Light and Harderan Function. Neuroendocrinology. 2Jj 193-203. Shirama, K., Furuya, T., Takeo, Y., Shimizu, K. and Maekawa, K. 1981. Influences of Some Endocrine Glands and of Hormone Replacement on the Porphyrins of the Harderian Glands of Mice. J. Endocr. 01 : 305-311. Sirek, O.V. and Sirek, A. 1967* The Colony of Chinese Hamsters of the C.H. Best Institute: A Review of Experimental VJork. Diabetologia. 3.: 65-73. Smelser, G.K. 1943. Changes Induced In the Harderian Gland of the Guinea Pig by the Injection of Hypophyseal Extracts. Anat. Rec. 35 : 41-57. Soret, M.C., Dulin, W.E., Mathew, J. and Gerritsen, G.C. 1974. Morphologic Abnormalities Observed in Retina, Pancreas, and Kidneys of Diabetic Chinese Hamsters. Diabetologia. IQ.: 567-573. Stetson, H.H. and Tate-Ostroff, B. 1981. Hormonal Regulation of the Annual Reproductive Cycle of the Golden Hamster. Gen. and Comp. Endo. its.: 329-344. Strum, J.M. and Charles, R.S. 1982. Harderian Glands in Mice: Fluorescence, Peroxidase Activity and Fine Structure. Tisue and Cell. liL: 135-148. Sun, C.Y. and Nadakavukaren, M.J. I98O. Effect of Testosterone on the Female Hamster Harderian Gland Pigmentation and Ultrastructure. Cell and Tissue Res. 2Q7: 511-517. Thiessen, D.D. and Kittrell, E.M. 198O. The Harderian Gland and Thermoregulation in the Gerbil (Meriones unguiculatus). Physiol, and Behav. ¿1: 417-424. Turner, C.D. and Bagnara, J.T. 1976. General Endocrinology. 6th ed. W.B. Saunders, Philadelphia. 141 Vivien-Roels, B., Pevet, P., Dubois, H.P., Arendt, J. and Brown G.M. 1981. Immunohistochemical Evidence for the Presence of Melatonin in the Pineal Gland, Retina and Harderian Gland. Cell and Tissue Res. 217; 105-116. Watanabe, M. 1980. An Autoradiographic, Biochemical and Morphological Study of the Harderian Gland of the Mouse. J.Morphol. 162: 344-365. Weaker, F.J. 1981. Light Microscopic and Ultrastructural Features of the Harderian Gland of the Nine-Banded Armadillo. J. Anat. 122; 49-65. VJeaker, F.J., Villareal, C. and Sheridan, P.J. 1983. Cellular Localization of Estradiol in the Harderian Gland of the Female Armadillo. Zoology. 225: 5-8. Weibel, E.R. 1979. Stereological Methods. Vol.1. Academic Press, New York. Wetterberg, L., Geller, E. and Yuwiler, A. 1970. Harderian Gland; An Extraretinal Photoreceptor Influencing the Pineal Gland in Neonatal Rats? Science. 167: 884-885 . Wetterberg, L., Yuwiler, A., Geller, E. and Schapiro, S. 1970. Harderian Gland: Development and Influence of Early Hormonal Treatment On Porphyrin Content. Science. 168; 996-998. Wolly, G.W. and Worley, J. 1954. Sexual Dimorphism in the Harderian Gland of the Hamster (Cricetus auratus). Anat. Rec. 118; 416-417. VJurtman, R.J., Axelrod, J. and Potter, L.T. 1964. The Uptake of H-Melatonin in Endocrine and Nervous Tissues and the Effects of Constant Light Exposure. J. Pharmac. Exp. Ther. 142; 314-318. Zar, J.H. 1974. Biostatistical Analysis. Prentice-Hall, Inc., Englewood Cliffs, N.J. 121-126, 488-489.