Cellular correlates of compensatory axonal sprouting in the magnocellular neurosecretory system of the rat by John Andrew Watt A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biological Sciences Montana State University © Copyright by John Andrew Watt (1993) Abstract: Neuronal and non-neuronal correlates of compensatory sprouting in the neural lobe (NL) were investigated using a unilateral hypothalamic knife-cut of the hypothalamoneurohypophysial tract to partially denervate the rat NL. Metabolic, morphometric and immunocytochemical methods were employed to investigate the cellular activity which accompanies the magnocellular neurosecretory system's response to partial denervation. Analysis of the metabolic effects of partial denervation indicate a transient urine hypoosmolality within the first 24 hours post-surgery (PS) followed by a chronic urine hyperosmolality which persists throughout the 3 month investigative period. Daily water intake (drinking) and urine excretion volumes demonstrate drop below control levels immediately following surgery which persists throughout the investigative period. Morphometric measurement of the magnocellular oxytocinergic and vasopressinergic neurons in the contralateral supraoptic nucleus (SON) demonstrated an enlargement of both cell populations within 10 days PS which remained significantly above control levels for at least 30 days PS. These data also suggest the oxytocinergic cell population may enlarge more vigorously or at an earlier time than their vasopressinergic counterparts. Cell counts of glial cell nuclei within 1 micron coronal sections of the NL revealed an increase in the number of pituicyte nuclei by 2 days PS which remained above age-matched control values for at least 30 days PS. The number of microglial cell nuclei numbers increased by 2 days PS but returned to normal by 10 days PS. Morphometric analysis of the NL vascular plexis revealed a strong temporal correlation between the enlargement of the NL vascular plexis and the process of compensatory collateral sprouting of neurosecretory axons in the NL. Immunocytochemical analysis of bromodeoxyuridine (Brdu) substitution suggests that the microglial but not the pituicyte populations within the NL are undergoing hyperplasia between 1 and 6 days following partial denervation. Analysis of nerve growth factor receptor (NGFR) expression in the NL indicates a dramatic increase in the proportional area immunoreactive for NGFR between 5 and 10 days PS, which then returns to normal levels. Immunocytochemical comparison between NGFR, iC3b compliment receptor and leukocyte common antigen (LCA) immunoreactivity suggest that a subpopulation of microglial cells may up-regulate their expression and/or increase the numbers of cells expressing the NGF receptor during a period when axonal degeneration and phagocytosis of neuronal debris is occurring. These data describe the varieties of neuronal and non-neuronal activities which underlie the compensatory axonal sprouting of intact magnocellular neurons within the partially denervated NL. CELLULAR CORRELATES OF COMPENSATORY AXONAL SPROUTING IN THE MAGNOCELLULAR NEUROSECRETORY SYSTEM OF THE RAT by John Andrew Watt A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biological Sciences MONTANA STATE UNIVERSITY Bozeman, Montana January 1993 APPROVAL of a thesis submitted byJohn Andrew Watt This thesis has been read by each member of the thesis committee and has been found to be satisfactory regarding content, English usage, format, citations, bibliographic style, and consistency, and is ready for submission to the College of Graduate Studies. (Z,/rn Chairperson, Graduate Committee D Approved for the Major Department 14 Towogry Date Head, Biology Department Approved for the College of Graduate Studies Graduate Dean ill STATEMENT OF PERMISSION TO USE In presenting this thesis in partial fulfillment of the requirements for a doctoral degree at Montana State University, I agree that the Library shall make it available to borrowers under rules of the Library. I further agree that copying of this thesis is allowable only for scholarly purposes, consistent with "fair use" as prescribed in the U.S. Copyright Law. thesis Requests for extensive copying or reproduction of this should be referred to University Microfilms International, 300 North Zeeb Road, Ann Arbor, Michigan 48106, to whom I have granted "the exclusive right to reproduce and distribute copies of the dissertation in and from microfilm and the right to reproduce by abstract in any format." iv TABLE OF CONTENTS vo -J in u H INTRODUCTION..............................................................I Statement of Purpose........................... .......... Plasticity in the Mammalian Central Nervous System...... The Magnocellular Neurosecretory System.................. The Fine Structure of the Neural Lobe.................... Development of the Vascular Plexus of the Neurohypophysis Neuronal Reorganization and Recovery of Function.................... 12 Sprouting in the Magnocellular and Parvocellular Neurosecretory Systems............................................ . •14 Nerve Growth Factor and its Receptors.............................. ..19 Justification, Rational a.nd Specific Objectives..................... 23 Role of Glia in Axonal Sprouting..................................... 23 Neuronal Activation in Axonal Sprouting.............................. 26 Specific Objectives............................................... 28 Non-neuronal Response to Injury................................... 28 Neuronal Activity................................................. 29 METHODS.................................................................. 30 Animals............................................................... 3^ Surgery............................................................... 30 Tissue Preparation................................................... 3I Tissue Microtomy..................................................... 33 Immunohistochemistry.................................................. 33 Image Analysis and Morphometric Measurements.........................38 192-IgG Proportional Area Analysis................................ 38 SON Cellular Morphometry.......................................... 39 Glial Morphometry................................................. 40 Metabolic -Analysis........... 41 Statistical Analysis................................................. 41 RESULTS.................................................................. 42 Post-Mortem Observations Within the Hypothalamus.................... 42 Lesion-Induced Changes in Urine Osmolality, Excretion and Daily Water Intake................................... 44 Hypertrophy of SON Somata and Nuclei following Partial Denervation................................................ 46 Morphometric Assessment of the Neural Lobe Vascular Plexus.................................................... 53 Low Affinity (p75) Nerve Growth Factor Receptor Immunoreactivity in the N L ......................................... 59 Morphometric Assessment of Glial Cell Numbers........................66 DISCUSSION............................................................... 73 Summary.................................................. 90 REFERENCES CITED........................................................ 92 V LIST OF TABLES Table Page 1. Size Distribution Bins of Vascular Morphometry...... ............. 55 2. Results of Glial Cell Counts - Lesion Group Means........... ..... 67 3. Results of Glial Population/Proliferation Estimates............... 71 -I vi LIST OF FIGURES Figure Page 1. Oxytocinergic Perivascular Plexis Within the Hypothalamus......................................... 43 2. Effects of Unilateral Hypothalamic Lesion on Urine Osmolality........... 45 3. Effects of Unilateral Hypothalamic Lesion on Daily Drinking Volume....................................... 46 4. Effects of Unilateral Hypothalamic Lesion on Daily Urine Excretion Volume............................... 47 5. Photomicrograph Demonstrating Silver-Enhanced Oxytocinergic SON Neurons....................................... 48 6. Hypertrophy of Vasopressinergic Neurons......................... 7. .49 Hypertrophy of Nuclei in Vasopressinergic Neurons.................. 50 8. Hypertrophy of Oxytocinergic Neurons................................ 51 9. 10. Hypertrophy of Nuclei in Oxytocinergic Neurons..................... 52 Effects of Lesion on Blood Vessel Density in the N L ....................................................... 54 11. Photomicrograph of NL Vasculature.................................. 56 12. Size Distribution of PS 2 Neural Lobe Vasculature..................................................... 57 13. Size Distribution of PS 90 Neural Lobe Vasculature...................................................... 58 14. Mean Proportional Area of NL Vasculature........................... 59 15. Mean Proportional Area of NL Capillary Plexus..................... 60 16. 192-IgG Immunoreactivity in the N L ................................. 61 17. Photomicrograph of NGFR immunoreactivity patterns in the NL at 5 days P S .................................. 62 18. Photomicrograph of NGFR Immunoreactivity Patterns in the NL at 90 days P S .................... ............ 64 19. Photomicrograph of Microglial Cells in the NL at 5 Days P S ........ 65 20. Pituicyte Density in the N L ........................................ 68 21. Pituicyte Nuclei Cross-Sectional Area.............................. 69 22. 23. Microglia Density in the NL Following Surgery....... .70 Confocal Photomicrographs of GFAP/Brdu and GSA Lectin/Brdu Labeling in the NL at 5 Days P S ................. 72 vii ABSTRACT Neuronal and non-neuronal correlates of compensatory sprouting in the neural lobe (NL) were investigated using a unilateral hypothalamic knife-cut of the hypothalamoneurohypophysial tract to partially denervate the rat N L . Metabolic, morphometric and immunocytochemical methods were employed to investigate the cellular activity which accompanies the magnocellular neurosecretory system's response to partial denervation. Analysis of the metabolic effects of partial denervation indicate a transient urine hypoosmolality within the first 24 hours post-surgery (PS) followed by a chronic urine hyperosmolality which persists throughout the 3 month investigative period. Daily water intake (drinking) and urine excretion volumes demonstrate drop below control levels immediately following surgery which persists throughout the investigative period. Morphometric measurement of the magnocellular oxytocinergic and vasopressinergic neurons in the contralateral supraoptic nucleus (SON) demonstrated an enlargement of both cell populations within 10 days PS which remained significantly above control levels for at least 30 days PS. These data also suggest the oxytocinergic cell population may enlarge more vigorously or at an earlier time than their vasopressinergic counterparts. Cell counts of glial cell nuclei within I micron coronal sections of the NL revealed an increase in the number of pituicyte nuclei by 2 days PS which remained above age-matched control values for at least 30 days PS. The number of microglial cell nuclei numbers increased by 2 days PS but returned to normal by 10 days PS. Morphometric analysis of the NL vascular plexis revealed a strong temporal correlation between the enlargement of the NL vascular plexis and the process of compensatory collateral sprouting of neurosecretory axons in the NL. Immunocytochemical analysis of bromodeoxyuridine (Brdu) substitution suggests that the microglial but not the pituicyte populations within the NL are undergoing hyperplasia between I and 6 days following partial denervation. Analysis of nerve growth factor receptor (NGFR) expression in the NL indicates a dramatic increase in the proportional area immunoreactive for NGFR between 5 and 10 days PS, which then returns to normal levels. Immunocytochemical comparison between NGFR, iC3b compliment receptor and leukocyte common antigen (LCA) immunoreactivity suggest that a subpopulation of microglial cells may up-regulate their expression and/or increase the numbers of cells expressing the NGF receptor during a period when axonal degeneration and phagocytosis of neuronal debris is occurring. These data describe the varieties of neuronal and non-neuronal activities which underlie the compensatory axonal sprouting of intact magnocellular neurons within the partially denervated N L . I INTRODUCTION Statement of Purpose These studies were designed to further elucidate the cellular correlates of the compensatory collateral sprouting which occurs within the magnocellular neurosecretory system (MNS) of the rat (Watt and Paden, 1991). The objectives of this research were to investigate the response of the MNS to partial denervation and to explore the non­ neuronal activity associated with the initiation and maintenance of compensatory collateral sprouting of intact MNS axons following partial denervation of the neural lobe (NL). The goals were five fold. I) Determine if the observed increase in relative area of partially denervated NL occupied by glia is due to cellular hypertrophy or hyperplasia of resident pituicytes and/or perivascular cells. 2) Investigate the possibility of a differential response in cellular activation/hypertrophy between oxytocin and vasopressin-immunoreactive cells in the contralateral supraoptic nucleus (SON). 3) Investigate the vascular response to partial denervation of the NL and its possible correlations with degeneration and sprouting of neurosecretory axons. 4) Evaluate the neurosecretory activity of vasopressinergic (VP) neurons by measuring alterations in drinking volume, urine excretion volume and urine osmolality. 5) Investigate the expression and possible up- regulation of nerve growth factor receptors (NGFR) on glial cells in the NL. The first objective was accomplished in part by investigating the proliferative activity of pituicytes and/or perivascular cells. Glial involvement in central nervous system (CNS) disorders has been repeatedly demonstrated to occur in Alzheimer's disease (Mandybur and Chuirazzi, 1990), experimental allergic encephalomyelitis (Cammer et al., 1990) and muscular sclerosis (Lee et al., 1990). The data obtained in the present investigation may provide clues regarding the mechanisms 2 regulating the glial response to denervation and ultimately to improve our understanding of the glial role in CNS repair. The second goal was to investigate the role of neuronal activity in the sprouting response by comparing the response of oxytocin vs vasopressin MNS neurons to partial denervation of the N L . This was accomplished by comparing the extent and time course of cellular hypertrophy during the compensatory sprouting of vasopressin magnocellular neurons. The third goal was to investigate the response of the NL vascular plexus following partial denervation. As will be described below, the growth of the NL capillary plexus during axonal development and regenerative sprouting of severed neurosecretory axons appears to be required for successful axonal and terminal maturation. Therefore, a morphometric analysis of NL vasculature was undertaken to determine if growth of the NL capillary plexus correlates with axonal sprouting. The metabolic effects of partial denervation of the antidiuretic system, the MNS, were explored as a means of gaining an indirect estimate of circulating plasma VP levels and the relationship of plasma VP levels to the induction of the compensatory sprouting response of vasopressinergic axons. Finally, alterations in the expression of nerve growth factor receptors were measured immunocytochemically as a means of determining the contribution of perivascular microglial cells to the post-denervation response. A brief historical review of the literature related to neural regeneration within the mammalian central nervous system will be presented next. This section will be followed by a detailed description of the structure and development of the neurosecretory system at both the light and ultrastructural level. This is intended to familiarize the reader with the close association of the components of the MNS and the interactions which underlie its development, function and inherent plasticity. This will be followed by a brief review of the different forms of injury-induced reorganization known to occur in the neurosecretory system and the contribution of nerve growth factor and 3 its receptors to CNS and neurosecretory function and response to injury. Finally, justification, rationale and specific aims will be presented. Plasticity in the Mammalian CNS Plasticity is a neurobiological concept which generally refers to the capacity of the central nervous system to modify existing synaptic contacts in response to injury, hormonal perturbations or environmental stimuli (Cotman et al., 1984). The principal evidence for plasticity comes from investigations of injury-induced responses (Cotman et al., 1984). Several forms of injury-induced axonal growth have been observed within the brain; these differ in several ways. First, regeneration generally describes the regrowth of a single fiber from the proximal end of the severed axon and is most frequently observed following injury to peripheral nerves. In the brain regenerative sprouting refers to the outgrowth of multiple collateral fibers from the proximal end of the severed axon. Second, the distance over which regrowth may occur is generally more extensive in regenerating axons than in sprouting axons (Cotman et al., 1984). Sprouts may arise from collaterals of intact fibers sharing the partially denervated terminal field (compensatory collateral sprouting), or from sustaining collaterals of injured axons proximal to the site of injury (regenerative sprouting; Kiernan, 1971). Injury-induced sprouting, often termed reactive synaptogenesis, may occur in several different forms, e.g., sprouts arising from nodes of Ranvier (nodal sprouting), from preexisting axon collaterals (collateral sprouting), or from the terminal bouton of intact axons innervating a partially denervated terminal field (terminal sprouting; Cotman et al., 1984). Although sprouting in any form can be extensive, it is generally more limited in the distance over which it may occur (Cotman et al., 1984), and tends to occur more rapidly (Murray, 1986) than regeneration. Cellular events accompanying the sprouting process have also been observed. For example, cellular hypertrophy following injury has been 4 well documented. Following hypophysectomy both somata and nuclei of SON neurons hypertrophy paralleling the formation of a neural lobe-like structure in the proximal infundibulum (Moll and DeWeid, 1962; Raisman, 1973). Likewise, somata and cell nuclei of SON and paraventricular (PVN) neurons hypertrophy during compensatory sprouting of intact neurosecretory fibers in the NL (Watt and Paden, 1991). Such a response may be attributed to a heightened level of activity in the sprouting neurons (Moll and De Weld, 1962). Following unilateral ablation of the left striatum and overlying cortex Pearson et al., (1987a) described, an increase in cell size in the contralateral pars compacta and ipsilateral pars reticulata of the substantia nigra, and the contralateral globus pallidus. These investigators also demonstrated hypertrophy of cholinergic cells of the rat basal forebrain following damage to the contralateral cortex. This enlargement was apparent by 7 days post­ surgery and persisted for at least 300 days PS (Pearson et al., 1984). The degree of hypertrophy observed was related to the age of the animal (Pearson et al., 1985) and the degree of damage sustained (Pearson et al., 1987b). They went on to conclude that the hypertrophy represented a cellular response accompanying distal axonal sprouting (Pearson et al., 1987b). Cotman et al. (1977) have reported that increase in cell size is one of the earliest manifestations of the sprouting response. In several cases a correlation has been demonstrated between cellular hypertrophy and sprouting of intact efferents of the enlarged cells contralateral to the site of injury (Goldschmidt and Steward, 1980; Hendrickson, 1982; Headon, 1985; Pearson, 1987a), in a manner similar to that previously reported by the author (Watt and Paden, 1991). Thus, cellular hypertrophy in response to direct axonal injury or in response to partial denervation of a terminal field shared with other lesioned neurons is a well described cellular event. 5 The Maanocellular Neurosecretory System The magnocellular neurosecretory system (MNS) is composed of the large oxytocin and vasopressin-producing neurons of the paraventricular, supraoptic and accessory nuclei. The axons which originate from these neurons project through the ventromedial hypothalamus, median eminence and infundibular stalk to the neural lobe where they form synaptoid contacts with the outer basement membrane of the capillary plexus (Silverman, 1983; Ju et al., 1986). The supraoptic nucleus is comprised of a compact mass of magnocellular neurons lying just anterolateral to the optic chiasm and optic tract (termed the principle division)(Ju et al., 1986), and a loose aggregation of magnocellular neurons positioned along the dorsolateral aspect of the optic tract and ventral surface of the brain at the level of the arcuate nucleus (termed the retrochiasmatic division)(Rhodes et al., 1981). There exists a partial segregation of vasopressinergic and oxytocinergic (OT) neurons within the principle division of the SON. The VP neurons are located predominately in the posteroventral region while the OT neurons are found predominately within the anterodorsal region (Vandesande and Dierickx, 1977; Rhodes et al., 1981; Swanson and Sawchenko, 1983). The paraventricular nucleus, once thought to be a uniform aggregation of magnocellular neurons (Christ, 1969) is now known to be composed of 3 topographically distinct magnocellular and 5 parvocellular subdivisions (Armstrong et al., 1980; Swanson and Sawchenko, 1983; Silverman, 1983). The three magnocellular subgroups are described as the anterior, the medial and the posterior magnocellular subdivisions. The five parvocellular subdivisions are comprised of the anterior, medial, posterior, lateral and periventricular cell groups (Swanson and Sawchenko, 1983). A topographically distinct arrangement of VP and OT neurons also exists in the PVN. The anterior and medial magnocellular subdivisions are comprised almost exclusively of OT neurons. In the 6 posterior magnocellular subdivision the VP neurons are located predominately within the posterodorsalateral region and the OT producing neurons are found predominately in the anteroventromedial region (Swanson and Sawchenko, 1983). In addition to vasopressin and oxytocin, at least 30 other neuropeptides have been immunocytochemically localized within the MNS including somatostatin, dynorphin, enkephalin, gastrin, and luteinizing hormone releasing hormone (Silverman, 1983; Swanson and Sawchenko, 1983; Ju, et al., 1986). Retrograde transport studies utilizing horseradish peroxidase (HRP) injections into the neural lobe have further delineated the cell groups contributing terminals to the NL (Armstrong et al., 1980; Kelly and Swanson, 1980; Ju et al., 1986). Almost all of the cells of the SON are labeled following these injections. SON axons were seen rising in a fan-like array from the dorsal aspect of the principle division of the SON before coursing ventromedialIy toward the internal lamina of the median eminence (ME) and infundibular stalk. Axons arising from the more lateral regions of this division first course anterolaterally while the medially placed cells send axons anteromedially. Both groups then streamed caudalIy over the optic tract to join the fibers arising from the dorsal SON. Those axons originating from the retrochiasmatic division of the SON were seen to emerge from the dorsal aspect and stream medially, dorsolaterally and anterodorsally. However, only the anterodorsal axons were traceable for any distance and were seen to join their principle division counterparts along the dorsal surface of the optic tract (Ju et al., 1986). The cell groups most heavily labeled within the PVN following retrograde transport of HRP from the NL are predominately concentrated within the medial and posterior magnocellular subnuclei (Silverman, 1983), corresponding to the PVM and PVL of Armstrong et al. (1980). However, scattered large and small neurons within several parvoeellular 7 subnuclei are also labeled, with the periventricular division being the most substantial contributor (Swanson and Kuypers, 1980; Ju et al., 1986). The axons arising from the magnocellular divisions of the PVN project ventroposterolaterally from the PVN to course around the dorsal and ventral aspects of the fornix then sweep posteroventromedially to occupy the lateral horns of the zona interna of the median eminence on their way to the NL (Ju et al., 1986; Swanson and Sawchenko, 1983). The magnocellular projections arising from the PVN and SON remain topographically segregated throughout their course through the hypothalamo-neurohypophysial tract. The SON efferents occupy the central two thirds of the zona interna, decussate within the median eminence, then continue through the infundibular stalk to form a widespread terminal distribution throughout the entire N L , with a higher concentration in the central core. The PVN fibers, on the other hand,, maintain an ipsilateral distribution throughout their course through the lateral zona interna and infundibular stalk, and terminate primarily in the periphery of the NL (Silverman, 1983; Alonso and Assenmacher, 1981). The functional correlates, if any, of this anatomical separation remain unknown (Silverman, 1983). Silverman et al. (1980) have also reported reciprocal connections between oxytocinergic neurons of the PVN nuclei and between the ipsilateral PVN and SON nuclei. The regions receiving afferents from the contralateral PVN appear to be the PVL and PVM subnuclei (Silverman, 1983). Anatomical and electrophysiological evidence suggests the existence of bilateral polysynaptic (Takano et al., 1990) neural connections between the SON nuclei which may facilitate the sychronous firing of oxytocin and vasopressin neurons. The Fine Structure of the Neural Lobe The NL is traditionally described at the fine structural level as comprised solely of pituicytes, perivascular microglial cells, fenestrated vascular endothelia, the endothelial and parenchymal 8 basement membranes which delimit an extensive perivascular space, and the nerve fibers and terminals of neurosecretory axons. Within the confines of the perivascular space occasional fibroblasts and collagen fibrils may also be observed (unpublished observations). Axons arising from dopaminergic (Baumgarten et al., 1972; Saavedra et al., 1985) epinephrinergic and norepinephrinergic (Saavedra et al., 1985) and serotoninergic neurons (Saland et al., 1987) have also been reported in the N L . However, the oxytocin and vasopressin neurosecretory axons comprise the vast majority of fibers pervading the N L . In the adult NL the neurosecretory axons are typically surrounded by the processes of pituicytes or completely engulfed within the pituicyte cell body. Rarely does one observe a "naked" axon within the N L . Likewise, the nerve terminal, when not in direct contact with the basement membrane, is ensheathed by pituicyte or parenchymal microglial processes. There is no agreement yet on whether pituicytes occur in different forms or whether differing morphologies represent the same cell type in different functional states (Fugita and Hartman, 1961; Dellman and Owsely, 1969); Zambrano and De Robertis, 1968; Kruslovic and Brucknerr' 1969). Dellman (1966) has described "dark" or fibrous pituicytes surrounding thin caliber axon cylinders in the proximal NL and a second "protoplasmic" pituicyte surounding axonal swellings and terminals within the body of the N L . The "dark" pituicyte is characterized ultrastructurally by an electron dense cytoplasm, little perinuclear cytoplasm, thin processes, and an electron dense nucleus with abundant heterochromatin particularly clumped at the inner surface of the nuclear envelope. In contrast, the protoplasmic pituicyte has a distinctly thicker perinuclear cytoplasmic region, an electron lucent cytoplasm, thick processes, and a large round nucleus filled with euchromatin with only small amounts of clumped heterocromatin attached to the nuclear envelope. It is possible that these dark pituicytes represent a parenchymal microglial cell (see below) while the "protoplasmic 9 pituicytes" represent the only pituicyte population endogenous to the NL. This distinction may have important functional implications in the partially denervated N L . A third cellular element endogenous to the N L , located within the perivascular space, is the perivascular microglial cell (OliviereSangiacomo, 1972). These cells are characterized by an elongate or oval nucleus displaying condensed heterochromatin clumped around the nuclear envelope, an electron dense cytoplasm, extensive granular endoplasmic reticulum, particularly in the perinuclear region but also extending into the long thin fibrous processes, lysosomes, numerous Golgi complexes and many free polyribosomes scattered throughout the cytoplasm (Oliviere-Sangiacomo, 1972; Watt and Paden, 1991). The long processes extend throughout the perivascular space and are often observed to surround or partially envelop neurosecretory axons which have penetrated the outer basement membrane of the perivascular space (OliviereSangiacomo, 1972). Perivascular cells often contain dense inclusions and phagocytosed debris arising from the degeneration of neurosecretory axons in the normal animal (Oliviere-Sangiacomo, 1972) and following unilateral destruction of the PVN (Zambrano and De Robertis, 1968) or SON (Watt and Paden, 1991). Development of the Vascular Plexus of the Neurohvnonhvsis As a neuroheamal organ, the neurohypophysis is characterized by an extensive penetrating vascular plexus. This plexus is composed in part by the border vessels located between the dorsal surface of the NL and the intermediate lobe (IL) and the superficial vessels which surround the remaining lateral and ventral external surface of the gland (D'Amore and Thompson, 1987; Eurenius, 1977). The internal vascular plexus is thought to arise through an extensive invasive sprouting from the external vascular network during development of the N L . This development occurs in three principle 10 stages. The first stage lasts until fetal day 17, the second from fetal day 18 until the IOth day following birth, and the third from day 10 through day 30 (Galabov and Scheibler, 1978; Galabov and Scheibler, 1983). By 30 days of age the NL capillary plexus is thought to be fully mature (Galabov and Scheibler, 1983; D 'amore and Thompson, 1987). Stage I is characterized by the development of the superficial and border plexuses surrounding the NL anlage, followed closely by the ingrowth of vascular sprouts at embryonic day 14 which arise from this peripheral network (Galabov and Scheibler, 1983). This vessel growth follows a distinct periperal to central pattern of maturation. This pattern is opposite that of the maturation of neurosecretory axons which first penetrate the internal core of the NL then sprout toward the periphery (Galabov and Scheibler, 1978). During the early stages of vessel ingrowth, the perivascular space is typically very narrow and closely applied to the vessel wall. Ultrastructural analysis of the NL anlage at the later stages of the first developmental period show the NL to be packed with immature pituicytes. No morphological indications of functional activity, such as neurosecretory contacts or the presence of granules within the axons or pituicytes, are apparent by fetal day 17. Thus, in the first stage of NL development, the vascular development is clearly more advanced than that of the neurosecretory axons or pituicytes (Galabov and Scheibler, 1983). The second stage of NL vascular development is of particular interest as the first indications of functional activity become apparent (Galabov and Scheibler, 1978; Galabov and Scheibler, 1983). The inward penetration of vascular sprouts continues concurrent with the first appearance of endothelial fenestrae. Pituicytes show signs of continued maturation and functional activity and the perivascular cells have increased in size and numbers. The perivascular space, delimited by the parenchymal and endothelial basement membranes, widens, branchs extensively, and extends deep into the NL parenchyma. The neurosecretory fibers in the central core of the 11 NL contain numerous granules and make contact with the outer basement membrane of the perivascular space (Galabov and Scheibler, 1983). During the ensuing days of fetal development, until the day of birth, the density of the vascular plexus'increases considerably. From the day of birth to the IOth day in the rat the vascular network continues to develop. Some immature capillaries are still observed within the central region of the NL but are few in number. There is an increased incidence of endothelial fenestrae and there are more fibrillar structures apparent within the perivascular space. At the beginning of this stage the pituicytes cover a large portion of the surface area of the perivascular basement membrane. However, as the second stage proceeds, these contact zones diminish as more neurosecretory axons establish terminal contact with the perivascular basement membrane (Galabov and Scheibler, 1983). From 10-30 days of age the capillary network gradually attains the appearance of the mature animal. The perivascular cells appear to be highly active, as indicated by increased Golgi apparati, and the general appearance of the perivascular space corresponds to that seen in the adult animal (Galabov and Scheibler, 1983). The precise interrelationship between the development of the capillaries, pituicytes and neurosecretory axons remains unclear; however, it is of some importance to reiterate that the development of the capillary plexus precedes the ingrowth of the neurosecretory axons. Likewise, the appearance of endothelial fenestrae and widening of the perivascular space does not occur until contact is made between the neurosecretory axons and the basement membrane of the capillary network. These observations may indicate an inductive influence of the axons upon the final stage of maturation of the capillary plexus. Whether the initial growth of the capillary network has an inductive influence on the subsequent axonal growth also remains unresolved. 12 Neuronal Reorganization and Recovery of Function If regeneration and sprouting represent forms of anatomical recovery from injury, the question arises whether or not such changes coincide with some degree of functional recovery. This is particularly relevant to the MNS since the regeneration of the proximal infundibulum, following hypophysectomy, leads to complete recovery of fluid balance. Although numerous studies have described some form of axonal reorganization following axotomy (eg., regeneration, sprouting of intact collaterals or sprouting of injured axons), relatively few studies have demonstrated a correlation between axonal growth and recovery of function (Finger and Almli, 1985). In fact, the associations made between neural regrowth and functional recovery have been so sporadic that several opposing hypotheses have arisen regarding the underlying function of the reorganizational process. Some investigators maintain that the reorganizational process is strictly governed by the need of the system to achieve a level of function close to that maintained prior to the injury. Others suggest that the neural changes associated with injury are related to developmental processes and may result from an alteration in the balance between growth stimulating and growth suppressing factors (Finger and Almli, 1985). Other investigators argue that any recovery of function associated with the reorganizational process is strictly coincidental. Support for the developmental hypothesis comes from studies of age-related factors and sprouting (Kawamoto and Kawashima, 1985a, 1985b, 1987). These studies have demonstrated that the initiation of the sprouting response is most rapid in the developing brain, decreases in the mature brain and is slower still in the aged brain (Cotman et a l ., 1984; Finger and Almli, 1985). If injury induced reorganization is related to a developmental process, it follows that this reorganization would be most likely to occur or to occur more rapidly during periods of rapid brain growth (Gall et al., 1981; Tsukahara et al., 1981; Cotman and Lynch, 1984; 13 Tsukahara, 1985). This does not, however, indicate that neural reorganization cannot take place in adult tissue. As mentioned above, there is ample evidence suggesting that mature systems do retain this ability but generally to a more limited extent (Steward et al., 1976; Cotman et al., 1977, 1984; Kawamoto et al.,1985; Needles et al., 1986). There are conditions under which sprouting in a previously denervated zone may contribute to recovery of function. Loecshe and Stewart (1975) demonstrated that unilateral damage to the entorhinal cortex resulted in initial impairment of performance of a trained task in rats; subsequently, task performance improved. Histologic examination demonstrated that sprouting had occurred within the contralateral entorhinal cortex and that the time course of sprouting was correlated with that of behavioral recovery. Wuttke et al. (1977) reported functional recovery paralleling axonal regrowth in the anterior hypothalamus. Serotonergic input to the hypothalamus was interrupted using 5,7-dihydroxytryptamine (DHT) chemical axotomy. Initially there was a marked reduction in plasma luteinizing hormone (LH) levels; however, two months following surgery the LH levels returned to normal values. Axonal regeneration was evidenced by an increase in 3H-SHT uptake and release in the anterior hypothalamus which paralleled the recovery of LH levels. Functional recovery following unilateral chemical axotomy of the descending serotonergic bulbospinal afferents was described by Nygen et al. (1974). Using the hindlimb extensor reflex as a measure of recovery, these investigators demonstrated that the initial 5-HT hypersensitivity observed following chemical axotomy disappeared within three months following surgery. This recovery was shown to closely parallel the time course of sprouting. The prolific regrowth observed following damage to the axons of the magnocellular neurosecretory system has been repeatedly correlated with functional recovery (Raisman, 1977; Kawamoto, 1985; Kawamoto and 14 Kawaahima, 1987). Hypophysectomy, the removal of the pituitary gland and distal infundibulum, is followed by the reestablishment of neurosecretory contacts with the capillaries of the mantel plexus in the zona externa of the median eminence (Raisman, 1977) and within the pseudo-neural lobe formed proximal to the stalk section (Kawamoto and Kawashima, 1987). Coincident with the development of these neurosecretory contacts, the polyuria and polydipsia associated with hypophysectomy disappear. The functional recovery achieved is such that the animal not only maintains homeostatic hydration levels but is able to resist moderate dehydration stress (Moll and DeWeid, 1962). Polenov (1974) found that salt loading hypophysectomized rats with a 1% NaCL drinking solution results in a substantial decrease in the amount of neurosecretory materials present within the proximal infundibulum. These animals demonstrate a direct correlation between alleviation of diabetes insipidus and release of secretory products from the reorganized neurosecretory endings. The functional competence achieved by the reorganized stalk was such that animals were able to withstand a twenty day salt loading regimen without deleterious effect. Sprouting in the Maanocellular and Parvocellular Neurosecretory Systems The magnocellular neurosecretory system has been the subject of extensive investigations of axonal sprouting in a variety of experimental preparations. Interruption of the hypothalamo- neurohypophysial tract either by hypophysectomy, neurolobectomy, infundibular stalk transection or hypothalamic lesion will result in morphological reorganization of the neurosecretory fibers proximal to the site of lesion in a variety of species. This effect was first described by Stutinsky in the rat (1951, 1952) and eel (1957). These initial reports were followed by further studies in the rat (Billenstein and Leveque, 1955; Moll, 1957; Moll and De Weid, 1962; Dellman and Owsley, 1969; Raisman, 1973; Polenov et al., 1974; Polenov et al., 1981; 15 Yulis and Rodriguez, 1982; Kawamoto and Kawashima, 1986, 1987), mouse (Kawamoto, 1985; Kawamoto and Kawashima, 1985), goat (Beck, et al., 1969), ferret (Adams, Daniel and Prichard, 1969), rabbit (Gaupp and Spatz, 1955), frog (Rodriguez and Dellman, 1970), Rhesus monkey (Antunes et al., 1978), and human (Beck and Daniel, 1959). Following hypophysectomy the magnocellular neurosecretory axons which are normally confined to the internal zone of the median eminence grow ventrally into the external zone. Here they establish functionally competent contacts with fenestrated capillaries. These "synaptoid contacts" differ morphologically from axoaxonic, axodendritic or axosomatic synapses. The contact occurs between the terminal ending of the axon and the outer of two basement membranes enclosing the perivascular space. In an inactive state, these terminals are separated from the basement membrane by the active inclusion of a pituicyte foot process. Upon metabolic demand, the pituicyte withdraws and contact is reestablished. Hence these synaptoid contacts are dynamic and display an inherent functional plasticity in both the SON nuclei and neurohypophysis (Theodosis and Poulain, 1987). Ultrastructural examination found these neurosecretory contacts to be identical to those of the intact neural lobe (Raisman, 1977). Various investigators found that the neurosecretory axons reorganized into a miniature neural lobe­ like structure at the proximal end of the transected infundibulum (Billenstein and Leveque, 1955; Moll and DeWeid, 1962; Kiernan, 1971; Raisman, 1973; Polenov et al., 1974; Kawamoto and Kawashima, 1987). Formation of new neuroheamal contacts in both the ME and proximal infundibular stalk is accompanied by a general hypertrophy of the capillary mantel plexus (Raisman, 1973). A similar process has been described in ferrets (Adams et al., 1969) and mice (Kawamoto and Kawashima, 1985). The axonal growth observed in ferrets represents the most robust regenerative activity of neurosecretory axons yet observed. Adams et al., (1969) severed the infundibular stalk in ferrets but 16 left the cut ends juxtaposedi A Gomori stain for neurosecretory material was then used to follow the Course of regeneration from two weeks to twelve months. By two weeks the proximal portion of the lower infundibular stalk displayed clear indications of regenerating fibers having penetrated the scar tissue barrier. One to three months following surgery, the regenerating axons had grown as far as the caudal end of the infundibular stalk. By five months the regenerating fibers had invaded the neural lobe and in two animals sacrificed at one year post-surgery the entire neural lobe had been reinnervated. Several animals also developed a neural lobe-like structure just distal to the site of lesion. This growth is similar to the pseudo-neural lobe-like structure formed in the transected infundibular stalk following hypophysectomy. Regeneration of severed MNS axons was also reported in rats following replacement of the neural lobe immediately after its removal, by suction, from the sella turcica. However, successful reinnervation of the neural lobe was only observed in two cases where the proximal infundibular stalk had not degenerated (Kiernan, 1971). Silverman and Zimmerman (1982) have demonstrated a sprouting response within the parvocellular neurosecretory system following unilateral electrolytic lesion of the paraventricular nucleus in adrenalectomized rats. As described above, the parvocellular neurosecretory axons projecting from each PVN normally remain ipsilateral within their terminal field in the external zone of the median eminence (Alonso and Assenmacher, 1981; Silverman and Zimmerman, 1982; Vandesande et al., 1977; Antunes et al., 1977). Approximately 21 days following unilateral destruction of the PVN, however, both HRP injections into the intact contralateral PVN and immunocytochemical localization of vasopressin-neurophysin within the median eminence revealed a bilateral innervation of the external zone. The authors concluded that these previously undetected fibers arose through a 17 compensatory sprouting response of the undamaged terminals of the contralateral PVN efferents (Silverman and Zimmerman, 1982). Regenerative sprouting of neurosecretory axons has also been demonstrated following interruption of the hypothalamo-neurohypophysial (HNT) tract within the hypothalamus (Danilova and Polenov, 1977; Antunes et al., 1979; Dellmann et al., 1985, 1986, 1987a, 1987b, 1988, 1989). However, growth occurs to a more limited extent than observed following interruption of the infundibular stalk (Danilova and Polenov, 1977; Antunes et al., 1970; Dellman et al., 1987a, 1988). In the intact animal the neurosecretory fibers comprising the HNT tract normally do not form connections with the capillaries, venules or arterioles that traverse the hypothalamus (Antunes et al., 1979; Dellman et al., 1987a). However, following transection of the HNT neurosecretory axons, the proximal segments of these fibers will form new neuroheamal contacts with newly formed vessels surrounding the lesion tract (Danilova and Polenov, 1977; Antunes et al., 1979; Dellman et al., 1987a) and with resident vessels of the hypothalamus riot previously innervated by neurosecretory fibers (Antunes et al., 1979). Antunes et al. (1979) have reported such regeneration arising from both paraventricular and supraoptic neurosecretory axons, including oxytocinneurophysin and vasopressin-neurophysin positive fibers. Exuberant growth of neurosecretory axons has also been observed in the ventral leptomeniges and surrounding pial vessels following disruption of the ventral hypothalamic surface (Danilova and Polenov, 1977; Antunes et al., 1979; Dellman et al., 1988). However, Antunes et al. (1979) reported observing hypothalamic sprouting no earlier than 8 weeks postsurgery (PS) while Dellman et al. (1987a) reported ultrastructural indications (e.g. the presense of morphologically defined neurosecretory granules with axon terminals) of perivascular sprouts surrounding hypothalamic lesion tracts and within the ventral leptomeninges as early as 5 days PS (Dellman et al., 1988). These fiber plexes were observed 18 within connective tissue spaces seemingly devoid of vessels, within arachnoid trabeculae and perivascular spaces, as well as surrounding the leptomeningial (pial) blood vessels (Dellman et al., 1988). Unfortunately, Dellman's group did not discriminate between the two classes of neurophysins, precluding any direct comparison of sprouting efficacy of fibers containing vasopressin vs those containing oxytocin under these conditions. Unsuccessful heterotypic tissue grafts placed within the hypothalamus are characterized by the absence of capillaries (Dellman et al., 1987b), underlying the essential nature of rapid revascularization for graft survival (Weigand and Gash, 1988). When NL allografts were positioned within the HNT, neurosecretory axons were observed invading the graft site within five days (Dellman et al., 1987b). Variations in the degree to which a graft was innervated were common and were correlated with the vascular density of the graft. In other words, graft sites containing the highest densities of vessels contained the highest numbers of neurosecretory fibers (Dellman et al., 1987b). During the initial stages of axonal invasion of the graft sites both terminal and preterminal neurosecretory fibers were observed. Preterminal fibers were typically found deeply invaginated into the perikaryon of pituicytes or completely enwrapped by pituicyte or "microglial-Iike" cell processes (Dellman et al., 1987b). Terminal fibers were less intimately associated with the pituicytes and were often observed abutting perivascular basement membrane (Wiegand and Gash, 1988) much as would be seen in the normal N L . However, as time passed, the axon-pituicyte relationship became more and more extensive (Dellman et al., 1987b). Stereotaxic placement of intrahypothalamic allographs of sciatic nerve, optic nerve and hypodermal connective tissue within the HNT result in varying degrees of penetration by neurosecretory axons (Dellman et al., 1985a, 1985b, 1986a, 1987). The relative density of 19 invading neurosecretory axons appears to be dependent upon the degree of vascularization. Hence, NL allografts contain the highest number of axonal profiles while optic nerve allografts contain the least (Dellman et a l ., 1987b). However, within optic nerve allografts the highest density of neurosecretory fibers were always found in the immediate vicinity of blood vessels (Dellman et al., 1989). These observations indicate that the vasculature must play a prominent role in the attraction and maintenance of the neurosecretory axonal ingrowth. Although these data do not exclude the possible, and likely, contribution of pituicytes to the development and growth of neurosecretory axons in the normal animal, they do indicate the underlying importance of blood vessels and the various components of the perivascular network to a regenerative and/or compensatory sprouting response within the compromised neurosecretory system. Nerve Growth Factor and its Receptors Neuronal differentiation and survival in the developing and adult CNS is mediated by a variety of neurotrophic factors most notable of which is nerve growth factor (NGF). NGF is known to promote the survival and differentiation of neural crest-derived sensory neurons, sympathetic neurons and magnocellular cholinergic neurons of the rat basal forebrain (Williams et al., 1986; Taniuchi et al., 1986) and striatum (Martinez et al., 1985; Mobely et al., 1985). The highest levels of NGF protein and its message are found in those brain regions receiving the most dense cholinergic innervation: the cortex and hippocampus (Springer, 1988). Direct evidence indicating the contribution of NGF to survival and function of basal forebrain cholinergic neurons has emerged from studies of hippocampal plasticity. Unilateral lesion of the fimbria-fornix results in loss of innervation arising from the magnocellular cholinergic neurons of the basal forebrain. Concurrent with the degeneration of the cholinergic 20 terminals is a substantial rise in levels of nerve growth factor (NGF) protein, but not mRNA, within the denervated hippocampus within 10 days of axotomy (Larkfors et al, 1987; Korshing et al., 1986). Since only the protein and not the message increased, it has ,been suggested that this increase is due to the loss of cholinergic efferents which retrogradely transported the NGF protein away from the hippocampus. With the subsequent collateral sprouting of cholinergic fibers within the denervated field, the NGF protein levels return to normal (Milner and Loy, 1980). As a result of the axotomy the cholinergic neurons of the basal forebrain degenerate. However, chronic ventricular infusion of NGF (Hefti, 1986; Williams et al., 1986) or transplantation of NGF secreting tissues (Springer et al., 1988) will result in the survival of the axotomized neurons. These data indicate the reliance of specific central neurons in the adult brain on NGF for survival and plasticity. Although the mechanism underlying NGF activity within a target tissue remains enigmatic, it is known that the ability of a target tissue to utilize NGF is reliant upon the presence of high affinity nerve growth factor receptor (NGFR)(Springer, 1988). of the receptor exist: At least two forms a high affinity form which is absolutely required for the internalization and utilization of NGF by the receiving cell, and a low affinity form which binds NGF but does not internalize the resulting NGF-NGFR complex, a step that is essential for NGF activity (Bernd and Greene, 1984; Hosang and Shooter, 1987; Johnson et al., 1988). Autoradiographic mapping of I125 NGF binding has demonstrated high affinity NGF receptors in the basal forebrain (Bernd et al., 1988), the caudate putamen, and most sympathetic and primary sensory neurons (Richardson et al., 1986). Immunocytochemical analysis and immunoprecipitation assays have demonstrated NGF receptors in the olfactory tubercle, medial septal nucleus (Taniuchi et al., 1986), the vertical and horizontal limbs of the diagonal band of Broca, the basal nucleus of the forebrain (Kiss et al., 1988), Purkinje cells of the 21 cerebellum (Taniuchi et al., 1886; Pioro and Cuello, 1988), brainstem, cortex, hippocampus (Taniuchi et al., 1986), suprachiasmatic nucleus (Sofroniew et al., 1989), and posterior pituitary (Yan and Johnson, 1990) of the rat. Hence, there is a growing body of evidence implicating NGF activity and NGFR expression throughout the adult CNS. A growing number of non-neuronal tissues and cell types, not believed to be NGF responsive, have been shown to express the low affinity form of NGF receptor and/or the NGF protein during development, following injury, or in cell culture. These include Schwann cells (Taniuchi et al., 1986; Johnson et al., 1988; Assouline and Pantazis, 1989; Matsuoka et al., 1991), cultured astrocytes (Furakawa et al., 1986; Yoshida and Gage, 1991), fibroblasts (Furukawa et al., 1985; Yoshida and Gage, 1991), and activated microglia (Yoshida and Gage, 1991) . Fibroblasts will secrete NGF but have not been shown to express NGF receptor in rat or human (Assouline and Pantazis, 1989). CNS astrocytes also express NGF message (Lu et al., 1991) and protein, but not the NGF receptor, (Assouline and Panatazis, 1989) in culture (Yoshida and Gage, 1991), during development and following axotomy of the optic nerve of the adult rat (Lu et al., 1991). In vitro studies have demonstrated the expression of NGF to be highest during stages of rapid glial growth (Yoshida and Gage, 1991). Non-neuronal cells express only the low affinity form of .the NGF receptor (Johnson et al., 1988) and are not capable of internalizing or utilizing the NGF protein. The expression of NGF receptors and secretion, of NGF protein by Schwann cells is developmentally regulated (Johnson et al., 1988). development. Binding of NGF occurred from days E6-E16 of After this stage the receptor expression diminished to undetectable levels (Johnson et al., 1988). However, following peripheral axotomy a reexpression of low affinity NGF receptor by the ensheathing Schwann cells (10-50 fold) appeared within 3 days and peaked at 5-7 days (Taniuchi et al., 1986; Johnson et al., 1988). The 22 reexpression was confined to the region distal to the lesion. No reexpression was detected proximal to the lesion where axon-Schwann cell contact remained, indicating a contact mediated suppression of NGF receptor expression by Schwann cells (Johnson et al., 1988). The presentation of bound NGF on the surface of the Schwann cell (presumably secreted by the same Schwann cell) may play a role in the cell's contribution to axonal regeneration following axotomy by providing an NGF-Iined path for the regenerating axon to follow (Johnson et al., 1988). Once the regenerating axon reached its NGF secreting target tissue, NGF secretion and expression of NGF receptors were shut off. Within the posterior pituitary of the neonatal rat, NGF receptor expression (detected immunocytochemically) continues to intensify in expression throughout the first year of life (Yan et al., 1990). Following transection of the pituitary stalk, the NGF receptor immunoreactivity rose dramatically within 3 days and peaked at approximately 15 days following stalk transection (Yan et al., 1990). This correlates closely to the time when axon degeneration was occurring and phagocytosis of the axonal debris was at a maximum. Dehydration, lactation, castration in male rats, pregnancy and colchicine injection all produced alterations in glial and/or neurosecretory activity in the neural lobe. However, none of these manipulations resulted in an increase in NGF receptor immunoreactivity. These data would suggest that the increased glial activity associated with phagocytosis of the axonal debris was responsible for the increase in NGFR immunoreactivity. The cell type within the neural lobe thought to be responsible for phagocytosis is the perivascular microglial cell (Zambrano and Robertis, 1979; Watt and Paden, 1991). In summary, the occurrence of NGF and NGF receptors has been repeatedly demonstrated in both neuronal and non^neuronal cells of the peripheral and central nervous systems. Within the adult CNS the role of NGF and NGF receptors remains unresolved. Indications of up 23 regulation of NGF protein and receptor following axotomy reflect a potential for CNS repair similar, but more limited in extent, to that observed in the peripheral nervous system. Justification, Rational and Specific Objectives Role of Glia in Axonal Sprouting. The proliferative astroglial response to penetrating injury often leads to inflammation and the development of scar tissue which may act as a barrier to axonal regeneration. However, there is accumulating evidence to suggest that astroglia and microglia may also play a supportive role in axonal sprouting, typically under conditions in which inflammation, tissue insult and disruption of the blood-brain barrier do not occur (Gall et al., 1979; Gage et al., 1988). The principle advantage of our model of collateral sprouting (Watt and Paden, 1991) is that the use of a unilateral hypothalamic knife cut, instead of an infundibular stalk section, hypophysectomy or neural lobectomy, permits the sprouting response of uninjured neurons to be studied in the intact NL without the complications arising from hemorrhage or scar formation. The NL in turn offers a highly simplified tissue, of CNS origin but lying outside the blood-brain barrier, in which to investigate problems such as the role of glial and neuronal activation in the sprouting response of central neurons. At least two distinct populations of glial cells reside within the N L , pituicytes and perivascular cells. Immunocytochemical application of an antibody directed against glial fibrillary acidic protein (GFAP), a known astrocyte-specific marker, indicates pituicytes are GFAPpositive and thus an astroglial-like cell (Velasko et al., 1979; Suess and Pliska, 1981; Salm et al., 1982). Pituicytes are characterized by highly asymmetric and extensive processes in close juxtaposition to neurosecretory axons and have a more velate appearance compared to astrocytes (Salm et al., 1982). Although the functional role of 24 pituicytes remains unresolved, they are known to ensheathe axonal endings within the NL when these axons are not in contact with the external basement membrane of the perivascular space. However, upon stimulation of the anti-diuretic system the pituicytes will rapidly withdraw from the axon terminals allowing terminal/basement membrane contact for neurosecretion. Additionally, pituicytes will cover the basement membrane left exposed upon axon terminal retraction (Wittkowski and Brinkman, 1974). We believe the intimate relationship between pituicytes and axons may be crucial in the sprouting response. A similar contiguous relationship between pituicytes and axons has been reported in the developing NL (Galabov and Scheibler, 1978; Dellman and Sikora, 1981) and during regenerative sprouting in the proximal infundibulum (Kiernan, 1971). NL perivascular cells are thought to be related to CNS microglia (Fugita and Hartman, 1961; Zambrano and de Robertis, 1968). Microglia, in turn are thought to be of mesenchymal origin and thus are not related to neuronal or neuroglial elements which arise from the germinal matrix (Dickson and Mattiace, 1989). Antibodies which recognize mesodermally- derived elements such as macrophages will also recognize microglia (Perry et al., 1985; Guilian et al., 1976; Esiri and McGee, 1986; Hickey and Kimura, 1986). Likewise antibodies directed against major histocompatibility complex (MHC) class II, commonly detected on immune cells, will also recognize microglia (Dickson and Matiace, 1989). However, microglia can be distinguished from tissue macrophages and blood monocytes on the basis of histochemical, morphologic and mitogenic analyses (Giulian et al., 1986). Our data indicate a functional relationship between perivascular cells of the NL and activated macrophages. We have observed that perivascular cells are solely responsible for phagocytosis of degenerating neurosecretory axons following knife cut (Watt and Paden, 1991). This activity is apparent at 5 days following knife cut and is characterized by an apparent 25 increase in the number of perivascular cells within the N L , an increase in the number of electron dense cytoplasmic lysosomal granules, highly dilated rough endoplasmic reticulum (RER) filled with a dense amorphous material and the presence in some perivascular cells of phagocytic debris (Watt and Paden, 1991). Likewise, Zambrano and De Robertis (1968) have shown perivascular cells to be exclusively involved in the phagocytosis of degenerating neurosecretory axons within the NL following bilateral electrolytic destruction of the PVN. Neuroglia and microglia may play an important role in neural repair following injury. Hypertrophy of astrocytes occurs surrounding sites of brain injury and neuronal death (Jacobson, 1978) as does proliferation of astrocytes in proximity to the site of an injury (Cavanagh, 1970; Latov et al., 1979). Microglia proliferate at a very high rate within 2-5 days following axotomy of the facial nerve (Kreutzberg and Barron, 1978) and within the outer molecular layer of the hippocampus following partial deafferentation (Gall et al, 1979). In this instance the proliferation of microglial cells preceded the increase in numbers of astrocytes. This sequential pattern is supported by evidence suggesting microglial cells secrete factors involved in the proliferation of astrocytes (Giulian and Baker, 1985). Furthermore, production of "glia-promoting factors" can be elicited by brain injury (Giulian et al, 1976). The possibility remains that the perivascular cells of the neurosecretory system are in fact a form of microglial cell. If so, it would seem likely that these cells would express immunologic cell surface markers as do microglia and macrophages (Perry et al, 1985; Esiri and McGee, 1986; Hikey and Kimura, 1988; Dickson and Mattiace, 1989). Likewise, immunopositive staining of pituicytes with GFAP indicates that these cells are at least related to, and possibly share a common ontogeny with, brain astrocytes. If so, the relationship between these cell types in the neural lobe may closely parallel that observed between the microglia and astrocytes of the brain, particularly 26 in response to brain injury or neuronal death. My working hypothesis is that the degenerative debris resulting from axonal degradation stimulates the activation, proliferation and phagocytic response of perivascular cells. This phase is followed by the activation, hypertrophy and possible proliferation of resident pituicytes, possibly as a result of one or more factors released by the activated perivascular cell population. The focus of this investigation was on the hypertrophy and hyperplasia of the glial population. Neuronal Activation in Collateral Sprouting. Cellular hypertrophy has been repeatedly demonstrated to correlate with the onset of compensatory sprouting of intact afferents. to a partially denervated site in a variety of CNS systems (Goldschmitt and Steward, 1980; Headon et al., 1985; Hendrickson and Dineen, 1982; Pearson, et al., 1987a; Pearson, et al., 1987b). Likewise, hypertrophy of SON neurons has been demonstrated to occur concomitant with the reformation of neurosecretory nerve terminals in the proximal infundibulum following hypophysectomy (Raisman, 1987). ' We have demonstrated that hypertrophy of magnocellular neurons in the contralateral SON coincided with the increase in axon number observed at 30 and 90 days following partial denervation of the NL (Watt and Paden, 1991). These observations are consistent with the conclusion that compensatory sprouting of intact neurosecretory efferents was occurring at this time. Cellular hypertrophy in the magnocellular neurosecretory system has also been associated with heightened secretory activity induced by dehydration or salt loading (Enestrom, 1967; Morris and Dyball, 1974; Peterson, 1966). However, these acute effects occur rapidly, and the somata return to normal size soon after cessation of the osmotic stimulation (Enestrom, 1967; Morris and Dyball, 1974). contrast, the prolonged delay observed in our study between partial denervation of the NL and the increase in cell size suggests the In 27 cellular enlargement is not the result of increased vasopressin synthesis or secretion. We believe the increase in SON and PVN cell size observed may reflect the increased cellular metabolic activity required to maintain an enlarged terminal arbor. Furthermore, this enhanced activational state may be a critical factor in determining the extent of the regenerative response. Herman and Gash (1987) demonstrated that vasopressinergic, vs oxytocinergic, cellular activity following NL ablation could be selectively inhibited via chronic vasopressin infusion. Results of their study suggest inhibition of vasopressinergic cell activity inhibits axonal regeneration and ultimately results in cell death, with no effect on the activational state, cell survival, or regenerative capacity of the axotomized oxytocinergic neurons. These results indicate the critical importance of heightened cellular activity to regenerative capacity (although, the specific form of activity, e.g., gene expression, synthetic, or electrophysiological, remains unclear). An advantage of the MNS as a model system for the study of sprouting is the possibility of selectively activating and suppressing vasopressinergic or oxytocinergic cell activity using different hormones, (e.g., vasopressin and oxytocin), and different physiological states (e.g., pregnancy, lactation, dehydration). Other investigators have demonstrated a difference in the temporal pattern of vasopressin and oxytocin immunoreactivity in the median eminence following hypophysectomy relative to the onset of sprouting of vasopressin axons as a function of age in both rats (Ibata et al., 1983; Kawamoto and Kawashima, 1987) and mice (Kawamoto, 1985). These findings suggest that there may well be differences in the ability of oxytocin vs vasopressin cells to undergo compensatory sprouting. The basis of such differences and their relation to cell activity remain unknown. I proposed to explore this problem by determining the temporal pattern of oxytocin vs vasopressin neuronal hypertrophy in the contralateral SON and its 28 possible correlation with the onset of compensatory sprouting by oxytocin and vasopressin axons in the partially denervated N L . Further exploration of any differences in vasopressin vs oxytocin neuronal activation and sprouting may provide new information regarding factors that affect the neuronal propensity for compensatory sprouting. Specific Objectives. A novel model of compensatory collateral sprouting of intact magnocellular vasopressinergic efferents following partial denervation of the rat NL has been reported (Watt and Paden, 1991). In addition to the sprouting response, further analysis revealed an increase in the relative area of the NL occupied by glial cells. This increase corresponded to the period of degeneration, activation of the vasopressinergic neurons and onset of sprouting. My objectives are divided into two principle areas of investigation: I) to further investigate the role of non-neuronal elements in the facilitation, initiation, and/or maintenance of the compensatory sprouting response and, 2) to measure the neuronal response to partial denervation and possible correlations with the sprouting response. These data may provide clues regarding the mechanisms regulating the glial and neuronal response to denervation and ultimately to improve our understanding of the glial role in central nervous system (CNS) repair. The specific aims of the proposed research and hypotheses to be tested are as follows: Non-Neuronal Contribution to MNS Response to Injury I) To determine if the observed increase in relative area of partially denervated NL occupied by glia is due to cellular hypertrophy of resident pituicytes and/or perivascular cells or to a proliferative response resulting in an increase in the absolute number of cells in either glial population. 2) To determine the vascular response to partial denervation of the NL and its possible correlations with degeneration and sprouting of neurosecretory axons. 29 3) To investigate the expression and possible upregulation of NGF receptors on glial cells in the N L . Neuronal Activity I) To investigate the possible occurrence of regenerative sprouting of neurosecretory axons within the hypothalamus. 2) To investigate the possible alterations in drinking volume, urine excretion volume and urine osmolality as an indirect means of determining whether the sprouting response may result from a lesioninduced deficit in plasma vasopressin levels. 3) To determine if a differential response in cellular activation/hypertrophy exists between oxytocin- and vasopressinimmunoreactive cells in the SON. 30 METHODS Animals Male Holtzman rats (bred from stocks originally obtained from Charles River) were 35 to 40 days of age at the time stereotaxic surgery was performed. The sprouting efficacy of transected neurosecretory axons has been shown to decrease with age (Kawamoto and Kawashima, 1985, 1987) and this reduced propensity has been observed as early as 23 days post-parturition (Kawamoto and Kawashima, 1987). Therefore, young rats between the ages of 35-40 days post-parturition were utilized in an effort to minimize any age-related impairment of the neurosecretory system response to partial denervation. Animals were either individually or group housed under a 12L:12D light cycle and had ad lib access to lab chow and tap water throughout the investigations. Surgery Prior to surgery each animal received Sodium-Pentobarbital, (50mg/kg intraperitoneal (IP)) to induce a surgical plane of anesthesia. Animals received methoxyflurane to prolong or deepen the plane of anesthesia, as necessary. Each animal received a preanesthetic injection of atropine sulfate, 0.5mg subcutaneous (SC), to reduce mucous secretions. Following anesthesia, the heads were shaved, washed with 100% ethanol (ETOH) and positioned in the stereotaxic apparatus with the incisor bar set at 3 mm below the horizontal position. Incisions were then made in a caudal to rostral direction and the dorsal surface of the cranium exposed, cleared of fascia and dried. The tip of the stereotaxic knife blade (see appendix for materials and design), was positioned at bregma and the coordinates recorded. coordinates, AP -3mm to +5mm, ML -0.5mm, The lesion (Paxinos and Watson, 1986) were then marked on the surface of the cranium with bregma serving as stereotaxic zero. A burr hole was then made through the skull using a 31 Dremel moto-flex drill and a #3 carbide burr, with care taken to ensure that the hole had clean, smooth edges to reduce the possibility of scalp irritation after surgery. The dura mater was incised using a sterile, 27 gauge hypodermic needle. Any excessive bleeding was interrupted by a.pplying direct pressure with a Kimwipe pad directly to the wound site. Once the incision was completed and cleared of debris, the stereotaxic knife blade was then repositioned at -2mm A P , and lowered through the brain to the ventral surface of the cranial cavity. When the knife blade contacted the skull surface, the blade flexed slightly. The blade was then retracted Imm and the knife slowly brought forward to the rostral extent of the lesion. The blade was then retracted and any blood and/or debris was cleared from the site. first filling the burr hole with Gelfoam. The wound was closed by The surface of the cranium was then gently wiped clean and nitrofurazone topical antibiotic applied liberally, and the scalp sutured. The animal was then removed from the stereotaxic apparatus and replaced in his home cage for recovery. All sham controls were prepared in a like manner except that the knife blade was lowered only 4 mm from the dorsal surface of the skull. This lesion depth transected the cortex but did not impact the hypothalamus, thus sparing the PVN and SON nuclei from damage. Tissue Preparation Tissue was collected from groups of experimental animals sacrificed between 12 hours and 150 days PS as well as from age-matched sham-operated and intact controls. Animals to be used for morphometric analysis of immunocytochemically-labeled vasopressin and oxytocin neurons of the SON were perfused intracardially with 4% paraformaldehyde in .IM phosphate buffer under deep ether anesthesia, then decapitated and the brain and NL removed. The brain and attached NL were then submerged and postfixed in the perfusion medium overnight, then dehydrated, cleared in xylene and embedded in paraffin. Animals to be 32 used for Bromodeoxyuridine (Brdu) substitution analysis were first placed under light ether anesthesia and then given a I ml IP injection of 20.0 mg/ml Brdu in .007M NaOH vehicle. Animals were allowed to survive for three hours following injection and then were perfused intracardially with saline solution for three minutes under deep ether anesthesia. The brain and attached NL were removed and immersion fixed with Carnoys' acidified alcohol fixative overnight. The tissue was then dehydrated for 3 hours in absolute ethanol, cleared for three hours in xylene and embedded in paraffin. Animals involved in the capillary morphometry and glial population estimates were anesthetized with ether and then perfused intracardially with 4% paraformaldehyde plus 2% glutaraldehyde in .IM phosphate buffer. The NL was then removed and postfixed overnight in the perfusion medium, then dehydrated, cleared in xylene and embedded in paraffin. The NL was postfixed 4 hours in 2% osmium tetroxide, dehydrated in ethanol and propylene oxide and then embedded with Spurr's epoxy resins in flatmount Beem capsule molds with the rostral end of the NL positioned at the tip of the Beem capsule for orientation. Tissues to be utilized for immunohistochemical analysis of the p75 low affinity nerve growth factor receptor (NGFR) were first perfused intracardially under deep ether anesthesia with cold saline for two minutes and then perfused for 20 minutes with Nakane's periodate-lysineparaformaldehyde (PLP) fixative. paraformaldehyde, The PLP fixative containing 2% .075M lysine and .OlM sodium periodate in .IM PO4 was prepared immediately before use. Following perfusion, the brain and NL were removed intact and postfixed overnight in PLP, cryoprotected in 30% sucrose/PBS for 24-48 hours at 4°C and then snap frozen in molds filled with OCT freezing compound using isopentane chilled to -160 degrees centigrade in liquid nitrogen. The tissues were then stored at -70 degrees centigrade until sectioned. Once sectioned and mounted on gelatin-coated slides, the sections were stored at room temperature. 33 Tissue Microtomy Paraffin microtomy was performed on a Leitz Wetzlab microtome. For the investigation of SON neuronal hypertrophy ten micron serial sections were collected from each specimen block, float-mounted singly on gelatin coated slides and dried overnight on a slide warmer prior to use. Those sections utilized for Brdu analysis were sectioned at six microns and mounted in a similar manner. NGFR analysis were All fresh frozen tissues for sectioned at ten microns using a Reichert-Jung Frigocut 2800 cryostat and thaw mounted on gelatin-coated slides. For capillary morphometry and glial nuclear counts one micron semithin sections of Spurr's embedded NL were sectioned using a Reichert Nr 318 435/E ultramicrotome using glass knives and collected from each of four subject animals per lesion group and three animals per agematched control group. Sections were mounted on gelatin-coated microscope slides by placing each section on a drop of distilled water and then drying briefly on a slide warming tray. All sections on a given slide were then coated with a liberal volume of 1% basic toluidine blue and warmed for approximately 5-7 minutes before being rinsed in distilled water. The staining intensity was then adjusted by brief immersion of the slide in 95% and 100% ethanol and coverslipping with Permount. Immunohistochemistrv In a previous investigation we were able to demonstrate hypertrophy of SON and PVN neurons and their respective nuclei which coincided with the collateral sprouting response of vasopressinergic neurosecretory axons (Watt and Paden, 1991). However, the Nissl stain utilized to visualize these neurons did not allow for a direct comparison between the vasopressin- and oxytocin-immunoreactive neurons within either nuclei. Therefore, the possibility of a differential response between vasopressin and oxytocin immunoreactive neurons within 34 the SON was investigated by selectively immuneIabeling each of the two neuronal populations for subsequent morphometric analysis. Both peroxidase and silver-intensified immunogold preparations were utilized in an effort to produce the most sensitive and accurate estimation of neuronal size, as described in detail below. Immunohistochemical application of anti-vasopressin-neurophysin (gift of Dr. Alan Robinson) and anti-synthetic oxytocin (Accurate) was performed on paraffin-embedded coronal sections from each of three experimental groups (6-8 animals/group) and four intact control groups (4 animals/group). The intact groups were included to control for any age-related increase in neuronal and/or nuclear size. The number of sections prepared for each animal was solely dependent on the number required to measure a minimum of 25 cell bodies and their respective nuclei in each animal. and nuclei measured. Most animals, however, had 35 to 50 cell bodies Sections were cleared of all traces of paraffin and then rehydrated through decreasing concentrations of ethanol leading to phosphate buffered saline (PBS). All sections were then incubated in either rabbit anti-vasopressin-neurophysin (1:1000) or rabbit anti­ synthetic oxytocin (1:2000) in blocking serum composed of bovine serum albumin and normal goat serum, pH 7.4, overnight at room temperature in a plastic hydration chamber. Sections were then rinsed in PBS for two successive ten minute increments and exposed to the secondary antibody, biotinylated goat anti-rabbit IgG (1:200)(Cappel) for one hour at room temperature in a hydration chamber. Sections were again rinsed twice in PBS then incubated in avidin-biotin complex (ABC)(Vector) for one hour as before. Finally, all sections were rinsed in PBS and then developed in 3,3' diaminobenzidine (DAB)(Sigma) solution containing 300 jul of I mg/ml glucose oxidase and 600 jul of 1% cobalt chloride per 100 ml of DAB solution. Silver-enhanced immunohistochemistry was performed as follows. Sections were cleared of paraffin in xylene and then rehydrated through 35 successive decreasing concentrations of ethanol to tris-phosphate buffered saline (TPBS) pH 7.4. All sections were then incubated for one hour in TPBS blocking serum (GSTPBS) containing 3.0% normal goat serum. The GSTPBS was then wicked away using Kimwipes and the primary antisera, rabbit anti vasopressin-neurophysin (1:40,000)(gift of Dr. Alan . Robinson), was applied and left overnight at room temperature in a plastic hydration chamber. All sections were then washed for 60 minutes in GSTPBS and the secondary antibody, goat anti-rabbit IgG conjugated to a I nm gold particle (BioCell), was applied and incubated at room temperature for 60 minutes. All sections were then washed in GSTPBS for 60 minutes and then washed for 10 additional minutes in sterile water (pyrogen-free for injection)(Kendall McGraw). Sections were washed in .2M citrate buffer, pH 3.85, for 10 minutes to slow the rate of silver deposition. Sections were then placed in a silver development solution which was prepared by combining 15 mis of 5.6% hydroquinone in sterile water, 15 ml of 0.7% silver lactate in sterile water, 10 ml of 2.OM sodium citrate buffer (pH 8.5) and 60 ml sterile water. Sections were then developed under a sodium vapor lamp for 8-16 minutes or until the reaction was judged to be complete. The development reaction was then halted by replacing the silver solution with a 5% acetic acid stop bath for one minute, and then fixing the sections in Kodak fixer for 10 minutes. All sections were then washed in a continuous flow of tap water for 30 minutes, counterstained in a 1% working solution of cresyl violet (Aldrich) and coverslipped in Permount. Low affinity (p75) nerve growth factor receptor (NGFR) expression in the NL was measured by applying the mouse monoclonal antibody 192-IgG (gift of Dr. E.M. Johnson, Jr.) to tissue sections selected from lesion and sham operated animals sacrificed at post-surgery days 0, 2, 5, 7, 10, 35 and 90. A minimum of 8 sections of NL were selected from each of 4-6 animals within each group. Tissue sections were selected from approximately the same depths through the NL across all animals in all 36 groups to ensure that a reasonably equivalent comparison was made between animals and between experimental groups. The large number of sections stained and the nature of the analysis (see below) required that sections be prepared in batches of 18. In order to ensure that variations in staining intensity not bias the analysis, each batch contained sections selected from animals representing both lesion and sham preparations from several time points. Sections were washed for 30 minutes in Tris-buffered saline and then incubated overnight with primary antibody 192-IgG (1:100) in TBS blocking serum (TBS/NHS), pH 7.6, containing 1% normal horse serum. All incubations were conducted at room temperature in a plastic hydration chamber. Following the primary antibody, the sections were washed for a minimum of two 10 minute intervals in TBS and then incubated with the secondary antibody, biotinylated horse anti-mouse (1:500)(Vector) in TBS/NHS, pH 7.6. Sections were then washed in TBS for two 10 minute intervals and incubated in ABC for 60 minutes. Again, sections were washed in TBS for two successive 10 minute intervals and then developed in DAB in darkness for 60 minutes. All sections were then washed for a minimum of ten minutes in TBS, dehydrated through increasing concentrations of ethanol, cleared in xylene and coverslipped in Permount. Immunocytochemical analysis of Brdu substitution was performed to determine if the apparent increase in the number of pituicyte nuclei was due to cell mitosis. Cell counts indicated that this increase occurred as early as 24 hours following the knife cut. Therefore, animals injected with Brdu were sacrificed at .5, I, 2, 3, 4, 6, 7, 10 and 30 days PS. Sham controls were sacrificed at I, 2 and 10 days PS. As before, tissue sections were selected from approximately the same depths through the NL across all animals in all groups to ensure that a reasonably equivalent comparison was made between animals and between experimental groups. Selected sections were cleared of all traces of paraffin in xylene and then rehydrated through decreasing concentrations 37 of ethanol leading to PBS. Because the anti-Brdu antibody will recognize only single stranded Brdu substituted DNA, denaturation was performed by a 60 minute incubation in ZN HCL. The primary antisera, mouse anti Brdu (1:100)(Becton-Dickenson), diluted in blocking serum (PBS containing 1% bovine serum albumin (BSA) and 1% normal goat serum) was then applied and left overnight. All incubations were performed at room temperature in a plastic hydration chamber. Sections were then washed in two successive 10 minute PBS rinses and then exposed in succession to the secondary antibody, biotinylated goat anti-mouse (1:200) in blocking serum for 60 mins., two 10 minute PBS rinses, ABC for 60 minutes, two 10 minute PBS rinses, then .025% DAB containing .03% glucose oxidase solution (Sigma) with .03% cobalt enhancement for 15-20 minutes in darkness. Each section was then counterstained in Gill's Hematoxylin, dehydrated through a series of increasing concentrations of ethanol, cleared in xylene and coverslipped in Permount. Once having determined that Brdu substitution had occurred, it was necessary to identify the proliferative cell type(s ). This was accomplished by utilizing both dual-label immunofluorescence and dual­ label peroxidase immunocytochemistry. Immunofluorescent preparations were subsequently analyzed utilizing confocal microscopy. For dual fluorescence immunocytochemistry, selected sections were cleared of all traces of paraffin in xylene and then rehydrated through decreasing concentrations of ethanol leading to PBS. Sections were then incubated in 2N HCL for 60 minutes, followed by three 10 minute PBS washes. The primary antisera, both mouse anti-Brdu (1:100)(Becton- Dickenson), and rabbit anti-glial fibrillary acidic protein (GFAP)(1:250)(Dako) were diluted in blocking buffer (containing 1% (BSA) and 1% normal mouse serum), was then applied and left overnight. All incubations were performed at room temperature in a plastic hydration chamber. Sections were then washed in two successive 10 minute PBS rinses and exposed in succession to: biotinylated goat anti-mouse 38 antibodies (b-GAM) for 60 minutes, two 10 minute PBS washes, an avidinphycoerythrin (1:200)(Avidin-PE)(Accurate-Phycoprobe) and fluorescein isothyocyanate-goat anti-rabbit (1:500)(FITC-GAR)(Hyclone) solution for 60 minutes, and two 10 minute PBS washes. Slides were then coverslipped in PBS:glycerol with Img/ml O-phenylenediamine. For dual label peroxidase immunocytochemistry the sections were cleared, rehydrated and denatured as described above and then exposed to the following sequence: anti-Brdu (1:100) overnight at room temp., 2x10 min PBS washes, b-GAM (1:200) for 60 minutes at room temp., 2x10 min PBS washes, ABC for 60 minutes, room temp., 2x10 min PBS washes, development in DAB with cobalt enhancement for 15-20 minutes, PBS wash for a minimum of 30 minutes, the second primary antibody (e.g. anti GFAP overnight at room temp.), 2x10 min PBS washes, secondary antibody b-GAR (1:200) for 60 minutes at room temp., 2x10 min PBS washes, ABC for 60 minutes at room temp., 2x10 min PBS washes, DAB without cobalt enhancement for a 15-30 min development period, followed by a PBS wash, dehydration, clearing and coverslipping in Permount. This paradigm resulted in a highly compartmentalized and very black Brdu nuclear staining against a ruddy brown cytoplasmic GFAP staining, allowing the two antibody staining patterns to be easily discerned and compared. Image Analysis and Morphometric Measurements 192-IaG Proportional Area Analysis Quantification of 192-IgG immunopositive proportional area was performed on an MCID image analysis system (Imaging Research Inc.). Microscope illumination was stabilized with a SOLA CVS transformer to prevent fluctuations in illumination intensities from biasing measurements of the 192-IgG immunopositive area. Caution was taken when reaccessing the MCID program that illumination intensities and microscope apperature settings were precisely the same as for all previous sessions. Calibrations were checked prior to each sampling 39 session against a known sample. Images were projected from an Olympus BH-2 light microscope using a IOx objective through a Sierra Scientific CCD high resolution camera for relay to the MCID program and displayed on an Electrohome monitor for image editing and analysis. The IOX objective was the lowest magnification objective that could be utilized without compromising the clarity and resolution of the immunostained region under analysis. However, this magnification required that each coronal section of the NL be divided into multiple sample frames for analysis. These data were then compiled to determine the proportional area of NL occupied by 192-IgG immunopositive cells within a given animal and within a given experimental group. All data were downloaded directly into a spreadsheet program (Quatro Pro) for statistical analysis. SON Cellular Morphometry Morphometric measures of somata and their respective nuclear were collected from the SON contralateral to the knife cut. The purpose of these measures was to investigate the possibility of a differential response between the vasopressin-neurophysin and oxytocin immunoreactive neuronal populations. Previous measures had shown a close correlation between somata and cell nuclear hypertrophy (Watt and Paden, 1992). Therefore, nuclear measures were again collected as a means of providing evidence of enhanced neuronal activity and verification of the somata measures. The criteria for cell selection was that a distinct cell boundary, a clearly defined nuclear envelope and a prominent nucleolus all be present. Measures were collected by first projecting a microscopic image of the cells of the contralateral SON onto a ZIDAS digitizing pad via an Olympus drawing tube. First, the periphery of the cell body was traced using a stylus interfaced with the digitizing pad. The nucleus was then measured and the cross sectional area of each measure was then 40 automatically printed. All measures collected were then loaded into a spreadsheet program (Quatro Pro) for analysis. Mean cross sectional area of cell bodies and nuclei were then calculated for each animal and each experimental time point. Glial Morphometry Morphometric measures of pituicyte nuclei numbers and cross sectional area were collected from one micron thick sections of Spurr'sembedded N L . Pituicyte nuclei have previously been reported to undergo a mitotic and hypertrophic response during periods of increased neurosecretory activity induced by dehydration (Krsulovic and Bruckner, 1969). The purpose of these measures was to determine if pituicyte nuclei may also enlarge during degeneration or axonal sprouting in the partially denervated N L . Such a response would indicate heightened pituicyte activity and provide indirect evidence of pituicyte involvement in the response to partial denervation. Measures were collected by first projecting a microscopic image of a coronal section of the NL onto a ZIDAS digitizing pad via a drawing tube. Sample fields were selected in each of three NL coronal sections per animal by carefully aligning a grid containing five evenly spaced squares along the long axis of the tissue section. The grid was placed on the ZIDAS and observed overlying the section via the drawing tube. Each square corresponded to one .02 mm2 sample field as viewed through a IOOX oil objective. At IOX each respective sample field was centered within the viewing field and the IOOX objective then engaged. A blank paper was then placed on the ZIDAS overlying the grid and each nuclear profile counted was checked off with either a black (pituicyte) or red (perivascular cell) felt tip pen, thereby ensuring that no nuclei would be counted more than once. This also served to provide a permanent record of all cell counts. The cross sectional area of each NL coronal section sampled was then measured and the mean number of nuclei per unit 41 area then determined for each section, animal and experimental group. Measures of pituicyte nuclear cross sectional area were collected by tracing the boundary of each pituicyte nuclear envelope within the sample field that was clearly defined and contained a prominent nucleolus. All measures were then loaded into a spreadsheet (Quatro Pro) for analysis. The luminal surface of every.blood vessel displaying a clearly defined endothelial wall within each of the five sample fields was traced in a like manner. Again, these data were then loaded into a spreadsheet for analysis. Metabolic Analysis A total of 12 lesion animals and 6 sham control animals were individually housed in Nalgene metabolic cages throughout the 100 day experimental period. Daily drinking and urine excretion volumes were determined directly from graduated drinking and urine collection tubes, respectively. Urine samples were collected at a consistent time every day (10:00 am) by pulling off a 2.0 ml sample from a consistent sample depth and the sample was briefly centrifuged at low speed to remove any precipitation or food contaminants prior to analysis. A 10.0 ul sample was then applied to a vapor pressure osmometer (Wescor) to determine the relative urine osmolality for each animal every day. Baseline data on urine excretion volume, daily drinking volume and urine osmolality were established over a seven day period prior to surgery. This period ensured the subject animal's adjustment to the metabolic cage as well as the daily sample collection and cage cleaning routine. Statistical Analysis One way analysis of variance was performed using the MSUSTAT AVlW program (developed by Dr. R.D. Lund). ANOVA was followed by testing the significance of differences in group means using the MSUSTAT program COMPARE and the Neuman-Keuls and Least Significance Difference test. > 42 RESULTS Post-Mortem Observations within the Hypothalamus The efficacy of the hypothalamic knife cut was determined by examining brain sections stained with cresyl violet which were selected from the most rostral to the most caudal aspect of the lesion tract in every animal. Only those animals which sustained a complete unilateral transection of the hypothalamo-neurohypophysial tract were included in any subsequent investigation. The knife cut was often accompanied by a marked edema and cerebral hemorrhage resulting in varying degrees of cavitation surrounding the knife path. This cavitation extended from the dorsal surface to varying depths of the brain in most lesion animals, particularly at early post-lesion times, and to a lesser extent in sham controls. The cavitation typically compressed the ipsilateral PVN, resulting in complete neuronal death. PVN neurons spared. Rarely were any ipsilateral No apparent displacement of the contralateral PVN or ipsilateral SON was ever observed. Within the cavity of the lesion tract was an extensive invasion of macrophages, monocytes and other blood born elements which persisted throughout the first 10 days following surgery. Within the parenchyma surrounding the knife cut an extensive gliosis was consistently observed which persisted for the first two weeks post-lesion. A marked hypertrophy of astrocytes was apparent throughout the ipsilateral hemisphere and particularly in close proximity to the lesion tract. No such astrocyte hypertrophy was notable in the contralateral hemisphere except within the corpus callosum. Neovascular growth was observed surrounding the knife tract in all animals as early as 10 days following the knife cut with new vessels developing in close juxtaposition to the lesion wall. These vessels ranged in size from small arterioles to vessels with diameters greater than any observed within the hypothalamus of the normal animal. In a 43 serendipitous observation it was noted that by 90 days following the knife cut much of the neovasculature surrounding the lesion tract was selectively immunostained, using silver-enhanced immunogold labeling, with antibodies against the neuropeptide oxytocin-neurophysin. However, when adjacent sections were prepared in an identical fashion with antibodies against vasopressin-neurophysin, no immunostaining was ever observed (Fig. I). This observation was made in animals of both the 30 and 90 day PS groups across literally dozens of like preparations. Immunostaining was never observed in any tissues prior to 30 days PS, Figure I. Oxytocinergic Perivascular Plexis Within the Hypothalamus. A. Silver-enhanced immunogold localization of oxytocin-neurophysin immunoreactive plexis surrounding the neovasculature in the 90 day animal (solid arrows). B. An adjacent section stained with vasopressinneurophysin in an identical fashion (solid arrows). A linear scar forms the boundary of the lesion tract (asterisk). Magnification, 200X. 44 nor were individual axons or fiber bundles observed coursing toward these apparent neurosecretory plexuses. However, fibers were observed projecting between individual plexuses in close juxtaposition to the cavity wall. Perivascular plexuses were observed in animals regardless of the extent of neuronal loss in the ipsilateral SON. The possibility of non-specific staining seems remote given that the internal surface of the lesion cavity and tears within the tissue show no immunoreactivity. Likewise, many vessel profiles were observed throughout the brain which showed negative immunoreactivity. The specificity of staining for oxytocin vs vasopressinneurophysin peptides, the appearance of immunopositive plexuses at only 30 and 90 days PS and the absence of non-specific staining suggest that the plexuses form through extensive outgrowth of hypothalamic neurosecretory axons. Lesion-Induced Changes in Urine Osmolality. Excretion, and Daily Water Intake. The principle objective of this experiment was to determine if the stimulus for the compensatory axonal sprouting observed was due to a response by the neurosecretory system to recover from a functional deficit in circulating plasma VP levels or to a growth response due in part to lesion-induced stimuli. The first step in exploring these possibilities was to determine if, in fact, the antidiuretic system had been compromised by the lesion. The possibility of lesion-induced changes in circulating plasma VP levels was explored indirectly by measuring urine osmolality as well as daily drinking and urine excretion volumes. Daily measurements of urine osmolality revealed an initial transient hypoosmolality, followed within 48 hours by a chronic and significant (P<.05, t-test, t=l.76, n=79 ) hyperosmolality which persisted throughout the 90 day post-surgical survival period (Fig. 2). The occurrence of the transient hypoosmolality would suggest a substantial decrease in plasma VP levels 45 immediately following surgery which occurred in spite of a concomitant decrease in both water intake and urine excretion volumes. The transient rebound in urine osmolality which occurred from day 2 to day 5 PS may reflect either a compensatory over-responsiveness resulting in heightened secretion of plasma vasopressin or release of VP stores from degenerating terminals. The urine osmolality then stabilized at a level significantly above that of sham control values for the remainder of the experimental period. A chronic and significant (P<.05, t-test, t=1.78, n=79 ) decrease in both daily water intake (Fig. 3) and urine excretion volumes (Fig. 4) URINE OSMOLALITY 2750 ♦ LESION D SHAM CONTROL 2500 M 2250 2000 1750 1500 1250 0 5§ 10 15 20 25 30 35 40 45 50 55 60 65 70 75 50 55 90 I DAYS PO ST-LESION Figure 2. Effects of unilateral hypothalamic lesion on urine osmolality. Urine osmolality decreased markedly within the first 24 nours foiiowing surgery and then immediately increased to a signifacantly (P<.05) hyperosmotic state which persisted throughout the experimental period. 46 began immediately following surgery and persisted throughout the postsurgical period, paralleling the increased urine osmolality. These data suggest that the lesion did not result in a functional deficit in circulating VP levels. Hypertrophy of SON Somata and Nuclei Following Partial Denervation. Cellular hypertrophy within the magnocellular neurosecretory system occurs following experimental induction of heightened metabolic activity via salt loading (Paterson and Leblond, 1977) or dehydration and lactation (Watson, 1965) and is an accepted indicator of heightened cellular activity (Vogels et al., 1990). DAILY DRINKING VOLUME Figure 3. Effects of unilateral hypothalamic lesion on daily drinking volume. Daily water intake was significantly (P<.OS) reduced within 24 hours of the surgical intervention and remained depressed throughout the experimental period. 47 Furthermore, cellular hypertrophy following partial denervation of a terminal field has been correlated with axonal sprouting contralateral to the site of injury (Goldschmidt and Steward, 1980). In a previous investigation (Watt and Paden, 1991) an enlargement of both magnocellular somata and their respective cell nuclei was demonstrated in the contralateral SON and PVN following partial denervation of the NL. Furthermore, these data indicated the hypertrophy occurred during the sprouting phase of the post-lesion response. However, the hrstochemical staining paradigm used did not distinguish between oxytocin and vasopressin magnocellular neurons, the two cell types in the nucleus (Fig. 5). URINE EXCRETION VOLUME 35 H ♦ LESION D SHAM CONTROL 25 - 15 10 - 0 5 1 10 15 20 25 30 35 40 45 50 55 60 65 70 75 SO S5 90 5 DAYS PO ST-LESICN 48 These distinct cell populations display differences in the temporal pattern of axonal outgrowth and penetration of the NL terminal field during development. It seemed reasonable then to explore the possibility of a differential hypertrophic response between the vasopressin and oxytocin immunoreactive neurons within the contralateral SON following partial denervation of the N L . A highly significant enlargement of vasopressin immunoreactive neurons (F=12.6, P<.00001, one way ANOVA, df=6/29) and nuclei (F=8.67, P<.00001, one way ANOVA, df=6/30) occurred following surgery when compared to controls. Figure 5. Photomicrograph Demonstrating Silver-Enhanced Oxytocinergic SON Neurons. Demonstration of the silver-enhanced immunogold labeling of oxytocinergic magnocellular neurons in the SON utilizing anti synthetic-oxytocin antibodies (solid arrowheads). The non-immunolabeled cells are vasopressinergic magnocellular neurons (arrows). Note the numerous immunolabeled fibers within and dorsal to the nucleus. (OC) Optic chiasm, (BV) blood vessel. Magnification, IOOX 49 By 10 days PS there was a 12% increase (one way ANOVA, Neuman-Keuls post hoc comparison, not statistically significant (NS)) in the cross sectional area of vasopressin-immunoreactive somata over age-matched control values (Fig. 6). By 30 days PS the difference had increased to 23.7% (P<.01, Neuman-Keuls) followed by a return to a 12% enlargement above age-matched intact control values by 90 days post lesion (NS). Within the control groups a continued growth of the vasopressin somata was evident throughout the experimental period. A significant 27% (P<.01, Neuman-Keuls) increase in somata size was apparent between the lesion day 0 and 90 day PS control. Hypertrophy of V a s o p r e s s i n N e ur o ns Figure 6. Hypertrophy of Vasopressinergic Neurons. The cross-sectional areas of at least 25 vasopressin magnocellular neurons were measured in the contralateral SON nucleus of each animal. Age-matched sham controls were included at 10, 30 and 90 days post-surgery. Somata were hypertrophic by 10 days, had reached significantly larger areas than controls by 30 days and remained enlarged throughout the post-surgical period. Note the continued enlargement of somata with age in control animals. 50 This continued cellular growth beyond the first 35 days post­ parturition demonstrates the importance of including age-matched controls in these experiments. The nuclei of these same VP-IR neurons reflected a similar, albeit not identical, pattern (Fig. 6). At 10 days PS there is a non-significant but readily apparent 12.7% decrease in the lesion group nuclear cross-sectional area when compared to age-matched control values. However, by 30 days PS a significant (P<.01, Neuman- Keuls) increase in cross sectional area became evident with nuclear area 20% above age-matched control values. By 90 days PS the nuclear enlargement was no longer significantly different from Hypertrophy of V a s o p r e s s i n Cell Nuclei Figure 7. Hypertrophy of Nuclei in Vasopressinergic Neurons. The cross-sectional areas of at least 25 vasopressin magnocellular cell nuclei were measured in the contralateral SON of each animal. Agematched sham controls were included at 10, 30 and 90 days post-surgery. A significant hypertrophy of somata occurred by 30 days (PC.01), which then returned to values no different from age-matched controls by 90 days PS. 51 age-matched control values. Thus, VP cellular nuclei demonstrate a more temporally restricted response than observed in VP somata. A continuing age related enlargement in nuclear cross sectional area was also apparent in the control groups reflected by a significant 28% (P<.01, Neuman-Keuls) increase between day 0 and 90 days PS. A highly significant enlargement of OT-immunoreactive somata (F=9.24, P<.00001, one way ANOVA, df=6/28) and cell nuclei (F=6.6, P < .0002, one way ANOVA, df= 6/29) was also observed (Fig. 7). By 10 days PS a robust enlargement of OT somata was evident with mean cell Hypertrophy 0 of Oxyto ci n 10 30 DAYS POST-SURGERY Neurons 90 Figure 8. Hypertrophy of Oxytocinergic Neurons. The cross-sectional areas of at least 25 oxcytocin magnocellular neurons were measured in the contralateral SON nucleus of each animal. Age-matched sham controls were included at 10, 30 and 90 days post-surgery. A significant hypertrophy of somata occurred by 30 days (P<.01), which then returned to values no different from age-matched controls by 90 days PS. 52 cross sectional area 24.7% above the age-matched control value. This difference was maintained throughout the experimental period, with experimental values 17% and 9.8% above age-matched control values at 30 and 90 days post-lesion, respectively. Average nuclear cross sectional areas were also markedly enlarged (NS) by 10 days PS reaching 19% above age-matched control levels (Fig. 8). This difference was maintained at 17.4% at 30 days PS but had disappeared by 90 days PS. Enlargement of somata and cell nuclei was also evident in the control tissues reflecting a continuing growth process in these cells throughout the experimental period. Hypertrophy of Oxytoci n Cell Nuclei Figure 9. Hypertrophy of Nuclei in Oxytocinergic Neurons. The crosssectional areas of at least 25 oxytocinergic magnocellular cell nuclei were measured in the contralateral SON of each animal. Age-matched sham controls were included at 10, 30 and 90 days post-surgery. A marked enlargement was apparent by 10 days PS persisting at 30 days, and disappearing by 90 days PS. 53 Although both the vasopressin and oxytocin immunoreactive neurons undergo a lesion-induced hypertrophy there are indications that the oxytocinergic neurons and their respective nuclei may respond more rapidly and/or more vigorously than their vasopressinergic counterparts. Morphometric Assessment of the Neural Lobe Vascular Plexus. Vascular endothelial cells show little proliferative activity in the mature animal (D'Amore and Thompson, 1987) indicating a low rate of collateral vessel growth under normal conditions. However, under certain conditions these normally quiescent cells may exibit considerable growth. Formation of the corpus luteum, chronically excersized muscle, psoriasis, atherosclerosis, arthritis, tumor vascularization and wound healing are all conditions which promote or require vascular growth. In the neurosecretory system an extensive collateral vessel ingrowth has been demonstrated in the relatively avascular proximal infundibulum following hypophysectomy (Raisman, 1973; Adams, et al, 1969; Beck, et al, 1969). Furthermore, the neovascular sprouting precedes that of the regenerative sprouting of the neurosecretory axons. It has been suggested that this vascular growth may act as a stimulus for the subsequent regenerative axonal sprouting (Raisman, 1973). Given this possible link between neovascular sprouting and neurosecretory axonal sprouting, a morphometric investigation of the vascular network within the partially denervated NL was undertaken in an effort to further elucidate the mechanisms involved in the initiation or maintanance of compensatory sprouting in the N L . Figure 10 illustrates the mean number of vessel profiles/mm2 within the NL of lesion and age-matched intact control animals. Although development of the NL vascular plexus is thought to reach maturity by 20 days of age in the mouse (Eurenius, 1977) and 30 days in the rat (Galabov and Schiebler, 1983) these data, indicate a continued increase in capillary profiles until at least 45 days of age (PS 10). 54 The number of vessel proflies/unit area then shows a downward trend reflecting an age-related reduction in vessel numbers or an increase in NL area continuing to at least 90 days PS. In contrast, the lesion groups show a slight increase in vessel density by 2 days PS which continues for at least 90 days reaching levels significantly higher (P<.01, Neuman-Keuls) than age-matched control values at 90 days (Fig. 10). It seems likely that these data reflect the maintenance of the early postnatal vessel density as opposed to a lesion-induced collateral sprouting of NL vessels. The difference observed in vessel densities between 90 day lesion (Fig. 11a) and age-matched controls Density of Vessel Profiles (PC.OI) IN T A C T 1200 N=4 LESIO N 5 5 N=4 960- LU CC < z> O 720- </) LU E O cr a. 480- 240- 0 0 2 5 10 DAYS POST-LESION 30 90 Figure 10. Effects of lesion on blood vessel density in the N L . A gradual increase in vessel density occurred throughout the post-surgical period. There was a concomitant age-related decline in the vessel density in control animals which resulted in a significant difference in vessel density by 90 days PS. 55 (Fig. 11b) was readily apparent. In order to further explore this possibility the size of the vessels within each experimental group was sorted by cross-sectional area into I of 10 bins (Table I). The rational for this approach was that an increase due to endothelial proliferation would be evident as a relative increase in the number of small vessels, i.e. capillaries, as opposed to the larger arterioles and venuoles. Table I. Size Distribution Bins of NL Vascular Morphometry Eouivalent Circle Vessel Area Range (jum2) Bin # .933-1.33 1.34-5.34 5.35-9.34 9.35-13.4 13.5-33.4 33.5-53.4 53.5-93.4 93.5-134 134-540 541-941 I 2 3 4 5 6 7 8 9 10 Mean (pm2) 1.13 3.33 7.33 11.35 23.37 43.38 73.38 113.39 340.00 741.00 diameter(pm) perimeter(pm) 3.77 6.47 9.60 11.94 17.14 23.35 30.37 37.75 65.36 96.50 1.20 2.06 3.05 3.80 5.45 7.43 9.67 12.02 20.81 30.72 Examination of the PS 2 lesion group indicates little difference in the number of small vessels, when compared to day 35 intact controls. However, there is a marked increase in the number of larger vessels (bins 6-9) which may be indicative of an acute vasodilatory response (Fig. 12). By PS 5 these differences in distribution of vessel size are still apparent although to a less pronounced degree, consistent with indications of an early, acute vasodilatory response. By 10 days PS a marked increase in the number a smaller vessels becomes apparent. By 30 days the continued increase in vessel numbers now extends to include all size distribution bins (data not shown). This trend is best illustrated at 90 days PS where increased vessels were observed in all size categories (Fig. 13). In summary, the increase in large diameter vessels at 2 and 5 days PS with no comparable increase in the number of small vessels indicates a vasodilatory 56 response. This is followed at 10 days PS by a slight increase in the number of small vessels which continues at 30 and 90 days PS. By 90 days PS however, a shift to increased numbers of large vessels is apparent as well. By 125 days of age (PS 90) a distinct pattern becomes evident. Vessel numbers drop off markedly from earlier age groups in bins 3-5 while remaining essentially equal in bins 7-10. These data indicate an age-related reduction in the number of small vessels with no equivalent reduction in the largest vessels measured (Fig. 11). These data indicate either growth of new collateral vessels or the maintenance of the early postnatal vascular plexus over a period when, in the normal animal, a decline in the number of small vessels is apparent. Figure 11. Photomicrograph of NL Vasculature. Higher density of blood vessels is maintained following unilateral knife cut. A. Neural lobe of the lesion animal at 90 days PS. Note the large number of small vessels perforating the N L . B. Neural lobe of age-matched intact control tissue demonstrating the relative absence of blood vessels. Magnification, 1000X. 57 The proportional area of the sample fields occupied by vessel lumen was calculated to determine the net effect of the change in vessel numbers on proportional vascular area (Fig. 14). These data indicate a significant (P<.01, Neuman-Keuls) increase in the proportional area of the NL occupied by vascular elements within 48 hours PS when compared to lesion day 0 controls. By 5 days PS the response had waned with no further differences between lesion and control animals observed. However, from 30-90 days PS an upward trend in vessel number/unit area became apparent in all size distribution bins (see Fig. 11). Why then would this increase not be evident in the proportional area analysis? SIZE DISTRIBUTION OF NEURAL LOBE VASCULATURE O i I 1 2 I 3 I 4 I 0 I 6 B IN 7 i 8 l 9 l I 1 0 # Figure 12. Size Distribution of PS 2 Neural Lobe Vasculature. Increased vessel profiles were observed in the largest bin size categories reflecting a vasodialation response 2 days following partial denervation. 58 As a means of reapproaching this apparent discrepency the proportional areas were again plotted but excluding bins 8, 9, and 10 (Fig. 15). The largest vessels, presumably arterioles and venuoles, by their very size could mask any increase in proportional area resulting from an increase in the size of the capillary plexus. These results demonstrate a pattern essentially identical to those displayed in Figure 10. The increase in proportional area observed at 2 days PS has returned to levels more representative of the vessel density observed at the same time, while by 90 days PS the proportional area of the lesion group is significantly higher (P<.05, Neuman-Keuls) than intact controls. Figure 13. Size Distribution of PS 90 Neural Lobe Vasculature. Increased vessel profiles were observed in all size categories at 90 days PS reflecting the absence of an age-related decline in vessel density apparent in control tissues. 59 These values now appear to more accurately reflect the observed increase in vessel density. Although these data do not necessarily reflect a growth response per se they do indicate that the principle alteration of the NL vasculature was in the capillary plexus as opposed to the larger arterioles and venuoles. Low Affinity (o75) Nerve Growth Factor Receptor Immunoreactivitv in N L . Monoclonal antibody (Mab) IgG-192 has been shown to bind specifically to low affinity nerve growth factor receptors (NGFR) in intact N L , infundibulum and median eminence but not anterior lobe. Proportional 1.00 < UJ Area of Capillary Plexus INTACT (P<.05) LESION N =4 0.80 GC < -J < O 0.60 N=3 U- CU O LU Z — O 0.40 E O CL O CC CL 0.20 0.00 0 Figure 14. Mean Proportional Area of NL Vasculature. An early, acute increase in the mean proportional area of NL vasculature is evident at 2 days PS (P<.01). No further differences were observed between lesion animals and controls. 60 An up-regulation of NGFR densities has also been demonstrated in neural lobe at 7 and 15 days following infundibular stalk section with the immunopositive cells found predominately, but not always, in the vicinity of capillaries (Yan and Johnson, 1990). However, the identity of the cells demonstrating this increased expression of NGF receptors has not yet been conclusively determined. Armed with these data, I sought to explore the possible upregulation of NGFR-immunoreactivity (NGFR-IR) in the NL and its possible correlation with degeneration and sprouting. Pr o p o rt i o n a l A re a of Total V a s c u la t u r e INTACT 0.60 oc < —J g (PC.01) N=4 < Ul LESION N=4 0.48 0.36 0.24 OC 2 O OC O- 0.12 0.00 5 10 DAYS POST-LESION Figure 15. Mean Proportional Area of NL Capillary Plexus. Proportional area analysis of the capillary plexus demonstrates no acute vasodilation at 2 days PS. An age-related decline in capillary area was observed in all control tissues. In contrast, the lesion animals show an increase in the proportional area occupied by capillaries which correlates closely with the increased density of the capillary plexus noted earlier. 61 Figure 16 illustrates the results of the proportional area analysis of NGFR-IR in the N L . A highly significant increase (F=9.8, P<.0002, one way Anova, df=11/59) in the proportional area of NL occupied by NGFR-IR elements occurred following partial denervation. Two days following surgery no noticable difference existed between lesion animals and age-matched sham controls. By 5 days PS a significant 70% (Pc.05, least significant difference test (LSD)) increase in the proportional area of NL occupied by NGFR-IR elements had occurred in lesion animals when compared to sham controls (Fig. 17). N G F R P ropo rtional A rea A nalysis 0.20 0.16 < UJ CE < -I < Z O t— CC O CL O 0.12 0.08 CE CL 0.04 0.00 2 5 7 10 35 DAYS POST DENERVATION 90 Figure 16. 192-IgG Immunoreactivity in the N L . A significant increase in the proportional area of NL immunoreactive for nerve growth factor receptor was evident within 5 days PS. This trend continued reaching peak values by 10 days PS then returned to normal levels by 30 days PS. An age-related increase in NGFR proportional area is evident as well. 62 Figure 17. Photomicrograph of NGFR immunoreactivity patterns in the NL at 5 days PS. A) Immunocytochemical application of 192-IgG in 5 day post-lesion animals demonstrates a highly immunoreactive staining pattern surrounding blood vessels and extending outward in long finger­ like projections (solid arrows), presumably within the perivascular space. Numerous 192-IgG negative glial cell nuclei are evident throughout the section (solid arrowhead). B) Age-matched sham control tissue demonstrating a reduced intensity 192-IgG immunoreactivity and proportional area. Magnification, 650X 63 This difference was also significant at 7 days PS (P<.01, LSD) and at 10 days following surgery. By 35 days following surgery the proportional area of NGFR-IR elements had fallen below those observed at 10 days PS and had returned to levels no longer different from age-matched control values. By 90 days PS NGFR-IR levels had increased over those observed at 35 days PS but were not significantly different from age-matched sham, control (Fig. 18). The observed increase in NGFR-IR which was apparent in. control tissue reflects a previously reported age-related increase in NGFR-IR in the NL of rats (Yan and Johnson, 1990). What is important to note here is the significant increase in NGFR-IR which occurs between 5 and 10 days following surgery, a time when perivascular cell activation and phagocytic activity is at a maximum (Watt and Paden, 1991). These data suggest that a heightened expression of the p75 low affinity NGF receptor coincides with these cellular events. The possible expression of NGFR antigens by microglia was further explored immunocytochemically utilizing known microglial markers Mab OX42 (iC3b receptor; Serotec) and OX-I (leucocyte common antigen; Serotec). OX-I antigens are thought to be upregulated in activated microglial cells (Perry and Gorden, 1987). staining patterns observed. Figure 19 illustrates the Mab OX-42 stains microglial cells in close proximity to blood vessels and putative parenchymal microglial cells in a highly specific and localized manner. Individual OX-42 positive microglial cells and their characteristic reticular processes could easily be discerned at the light level. The distribution of these immunoIabeled cells was essentially homogeneous with no apparent difference in the staining pattern or relative immunoreactivity between the parenchymal and perivascular compartments. A very similar pattern was characteristic of the OX-I immunostaining reflecting no differences in the distribution of labeled cells as a result of the activation of a subpopulation of microglial cells. 64 Figure 18. Photomicrograph of NGFR immunoreactivity patterns in the NL at 90 days PS. A. 90 days post-lesion 192-IgG staining intensity (solid arrowheads) is reduced from levels observed at 5 days PS and are no different from age matched sham controls. However the staining pattern and the proportional area of staining are similar to that observed at 10 days PS reflecting an age-related increase in nerve growth factor receptor expression. B. Age-matched sham control tissue. Magnification, 650X. 65 Figure 19. Photomicrograph of microglial cells in the NL at 5 days PS. A. OX-42 immunoreactive microglial cells at 5 days PS are distributed randomly throughout the tissue. Both putative parenchymal (solid arrows) and perivascular (solid arrowhead) microglia are apparent. B . OX-I microglial staining pattern at 5 days PS demonstrates a similar pattern to, but a higher staining intensity than, OX-42 labeled cells (solid arrows). C. NGFR immunoreactive microglia are located predominately surrounding and extending out from blood vessels as long velate projections (solid arrows). Magnification, 650X. 66 The NGFR staining pattern showed similarities as well as some differences compared to OX-42 and OX-I staining pattens in the ‘ distribution of the NGFR-IR cells. Distinctly labeled microglial cells appear to be located within the NL parenchyma. However, the majority of the NGFR-immunoreactive sites have a more reticular and somewhat velate appearance. This is particularly evident surrounding the blood vessels and the 192-IgG positive cells appear to be localized primarily within the perivascular space. This staining pattern becomes more evident with time and is particularly well illustrated at 90 days PS. The OX-42 and OX-I staining patterns show no maturational change. Morphometric Assessment of Glial Cell Numbers Astrocyte hypertrophy and proliferation have been reported to occur both within and surrounding sites of brain injury in a variety of experimental paradigms (Sjost^rand, 1966; Rose et al., 1976; Latov et al., 1979; Nathaniel and Nathaniel, 1981). Perivascular and parenchymal microglia have also been implicated in the extensive gliosis which accompanies brain injury (Streit and Kreutzberg, 1988; Streit et al., 1989). We too have reported an injury-related increase in the area of the NL occupied by glial elements which reached peak levels by five days post-surgery before gradually returning to normal a month later (Watt and Paden, 1991). However ,â– the specific glial cell populations contributing were not determined. Therefore an investigation was undertaken to determine whether glial proliferation and/or cellular hypertrophy might account for the observed increase in glial area. The initial experiment involved determining the relative density of pituicyte and microglial cells by counting glial nuclei in one micron coronal sections of toluidine blue stained plastic-embedded NL at varying times following knife cut (Table 2). As illustrated in Figure 20, a significant increase (F=IO.7, PC.OOOl, one way ANOVA, df=9/23) in pituicyte nuclear density was observed within 48 hours of denervation 67 which then remained above normal for 30 continuous days. By 2 days P S ’ pituicyte nuclei have increased a significant 39% above intact control values (P<.05, Neuman-Keuls). This increase was reduced slightly by 5 days PS reaching a value of 36% above normal (P<.05, Neuman-Keuls). The pituicyte numbers continue to decrease with no significant difference observed between PS 30 lesion animals and controls. Table 2. Exp. Group Results of Glial Cell Counts - Lesion Group Means* *. Pituicytes Lesion Animals: 418.0 ± 2 dps 309.0 ± 5 dps 10 dps 395.0 ± 30 dps 239.7 ± 91.6 ± 90 dps 150 dps 197.3 ± 11.5 26.1 21.2 23.8 18.4 15.7 NL Area mm Microglia 293.0 225.0 257.1 179.4 93.25 125.2 Intact Control Animals: 315.0 ± 23.7 293.0 0 dps 10 dps 210.0 ± 12.9 154.2 30 dps 190.2 ± 13.6 121.6 91.7 90 dps 102.7 ± 8.8 ± ± ± ± ± ± Pits/mm 18.2 24.9 18.0 8.9 16.2 6.7 .34 .26 .37 .26 .12 .39 1229.4 1188.0 1067.5 921.9 763.2 505.7 + 34.9 ± 9.5 ± 14.5 ± 12.6 .37 .23 .26 .41 851.3 913.4 731.6 250.4 ± ± ± ± ± ± ± ± ± ± Micro/mm 49.1 39.2 17.1 68.4 54.5 32.3 69.2 46.2 67.3 46.1 861.8 865.4 694.8 690.0 777.1 321.0 800.8 669.0 467.5 764.0 ± ± ± ± ± ± 38.0 130.4 13.0 40.9 57.3 46.0 ± 36.1 ± 37.8 ± 43.4 ±47.3 Four animals were included in all lesion groups. Three coronal sections were analyzed per animal. Three animals were included in all control groups except 0 dps which included four. *Mean ± sem Days post-surgery (dps) or equivvalent age for sham controls. We have previously reported shrinkage of the NL at 10 days following surgery with a return to normal size by 30 days PS (Watt and Paden, 1991). Such a reduction in NL cross-sectional area would almost certainly result in an apparent increase of cell density in the absence of glial death. However, the substantial elevation in pituicyte nuclei observed at 30 days PS provides strong evidence that the increase in cell density noted here reflects a true increase in relative pituicyte cell numbers. Further analysis of pituicyte contribution to the observed gliosis was investigated by measuring the cross-sectional area of pituicyte 68 nuclei. Having previously demonstrated a strong correlation between increased neuronal size and cell nuclear hypertrophy in the SON it seemed plausible that a similar response might be observed in the event of pituicyte cellular hypertrophy. Contrary to expectations, a significant 25% reduction in nuclear cross-sectional area was apparent by 2 days PS (F=9.11, P<.01, one way ANOVA, df=7/22) when compared to day O intact controls (Fig. 21). By 5 days PS the nuclear size had enlarged to levels slightly below, but not significantly different from, that of control values. No further differences were observed through the remainder of the investigative period. P it ui cy t e Density WMA LESION 1200 Lobe INTACT CNTL (P<.05) (Re.05) N -4 1500 5 5 iu cr < 3 O in Ne ur al N -4 - N -4 N -4 900- % LU 600- O => Q- 300" 0 5 10 30 DAYS POST-SURGERY 90 150 Figure 20. Pituicyte density in N L . Pituicyte density in the NL increases significantly 5 days following partial denervation of the NL (P<.05) when compared to day 0 control values. Pituicyte density remains high through the first 30 days PS then returns to normal levels. 69 The relative density of microglial cell nuclei was also examined in the same Spurr's embedded, coronal NL sections described above. These data indicate a moderately significant lesion-induced increase in microglial density (F=3, Pc.016, one way ANOVA, df=9/23)(Fig. 22). Two days following the lesion microglial nuclear counts revealed a non-significant 7% increase in cell density over lesion day O intact controls. These values continue to climb reaching a substantial 23% increase over lesion day O control values by 5 days PS. By 10 days PS microglial numbers have returned to values similar to age-matched controls. By 30 days PS microglial numbers in lesion brain have Nuclei C r o s s - S e c t i o n a l A r e a ### LESION INTACT CNTL 0.75 N -4 0.60 0.45 0.30 0.15 0.00 0 2 5 10 30 DAYS POST-SURGERY 90 150 Figure 21. Pituicyte nuclei cross-sectional area. Pituicyte nuclei cross-sectional area decreases significantly (P<.01) over the first 48 hours following surgery when compared to day 0 intact controls then returns to normal size by 5 days PS. These data may reflect the difference in size between immature daughter cells and their nuclei and the larger more mature pituicyte nuclei following mitotic activity. 70 remained essentially unchanged from values obtained at 10 days PS. By 90 days PS microglial numbers in control tissues have returned to values more typical of normal N L . Note that microglial numbers drop remarkably between 90 and 150 days PS reaching values significantly lower than observed at 2, 5 and 90 days PS. differences exist between groups. No other significant Having established that an increase in the number of pituicytes and microglia had occurred we next chose to determine whether or not this increase stemmed from cellular proliferation. Counts of Brdu immunopositive nuclei revealed that in the normal animal .4% of the NL cell population is actively synthesizing Mi cr ogl ia Density Figure 22. Microglial density in the neural lobe following surgery. An increase in microglial density occurs within the first 5 days PS then returns to normal values by PS 10. Note the age-related decrease in microglial density at PS 150. 71 DNA at any given time (Table 3). Within 24 hours of partial denervation this estimate increases to 1.2% of the cell population, an increase of 65%. By 2 days PS a significant 1.6% of the cell population is Brdu positive reflecting the peak level of mitotic activity observed in the NL (P<.05, Neuman-Keuls). A subsequent decline in Brdu positive nuclei occurs by 3 days PS dropping to .84% of the cell population. This trend continues with Brdu cell numbers returning to normal values by 6 days following surgery. Thus, the mitotic activity observed within the NL occurs within a period of time when degeneration and phagocytic activity within the NL are at peak levels. Table 3. Results of Glial Population/Proliferation Estimates. Exp.(n) Sham: I dps(7) Lesion: I dp s (6) 2 dp s (4) 3 dp s (3) 4 dps(5) 6 dps(2) 7 dps(4) Brdu cells\section %Brdu 2008+115 8.3+.69 0.4 2046+156 2089+71.8 1887+108.4 1829+86.2 2196+229 1983+137 24+1.7 33.4+4.2 16+3.7 14±3.8 9.2+2.1 4±.5 1.2 1.6 0.84 0.76 0.41 0.20 Estim. cell numbers\sect. % increase over control 65.5 76.1 51.8 40.7 9.7 — 2--- -— ;-- TT-,— =-7 Three .02mm" sample fields were measured/section. Six sections were measured/animal. ( ) denotes the number of animals in the indicated group, (dps) days post-surgery Immunocytochemical markers were next employed to demonstrate the specific cell type(s ) undergoing the mitotic activity. Dual peroxidase immunocytochemistry using anti-GFAP, an astrocyte-specific immunolabel, and anti-Brdu showed that in all sections and all animals examined the Brdu-positive nuclei did not co-localize with GFAP immunoreactivity. A second attempt was made by utilizing the superior resolution achievable with the confocal microscope and dual fluorescence immunocytochemistry. Again, the GFAP and Brdu labels were clearly not localized to the same cell type (Fig. 23a) indicating the pituicytes are not actively dividing. Next, a dual fluorescence label pairing anti-GSA Lectin, a 72 marker for microglia and macrophages, and anti-Brdu was also examined with confocal microscopy. These data showed a definite colocalization of both antibodies to the same cell type (Fig. 23b). However, these data cannot differentiate between blood-born monocytes, macrophages and brain-derived microglia, leaving any conclusions regarding the cell type(s) undergoing mitosis unresolved. Figure 23. Confocal Photomicrographs of GFAP/Brdu and GSA-Lectin/Brdu labels in the NL at 5 days PS. A. GFAP-immunoreactive pituicytes (arrows) at 5 days PS are clearly not Brdu immunopositive (solid arrows). B. GSA-Lectin immunopositve microglial cells (arrows) show intense Brdu label in the cell nuclei (solid arrowheads). Magnification, 450X. 73 DISCUSSION Observations of cavitation, astrocytic hypertrophy surrounding the wound, gliosis, the influx of massive numbers of monocytes and macrophages within the wound cavity and neovascular growth surrounding the knife track are consistent with the reports of others (Rio-Hortega, 1932; Danilova and Polenov, 1977; Antunes et al., 1979; Ling, 1981; Yoshida et al., 1986; Dellman et al., 1987a). In this investigation numerous newly formed vessels were apparent as early as 10 days PS, the earliest time investigated in hypothalamic tissue, indicating the rapid onset of this growth response. No regional heterogeneity in vessel growth was observed with new vessels forming at all levels of the lesion tract. The stimulus for the gliosis and neovascularization may be due in part to the mononuclear phagocytes observed within the wound cavity and penetrating the surrounding brain parenchyma (Banda et al., 1982; Knighton et al., 1983; Guilian, et al., 1988). Outside the CNS these cells are recognized as potent secretors of interleukin-I (!LI)(Guilian, et al., 1988) and have been demonstrated to induce rapid and extensive astrocytic hypertrophy and neovascular proliferation following injection into the rat brain parenchyma (Guilian, et al., 1988). Both types of mononuclear phagocytes present here, e.g., blood-born macrophage and brain microglia, have been shown to secrete IL-I (Guilian and Baker, 1985, 1986; Giulian, 1987) as well as other factors (Banda et al., 1982; Knighton et al., 1983) which act to condition the wound fluid and surrounding tissues resulting in the subsequent gliosis and neovascularization. The regenerative response by neurosecretory axons observed in the hypothalamus following hypothalamic knife cut is consistent with the reports of others (Danilova and Polenov, 1977; Antunes et al., 1979; Dellman et al., 1987a) but differs in several ways. Immunocytochemical indications of regeneration by neurosecretory axons were observed within the hypothalamus and ventral thalamus and were specifically localized to 74 vessels in close proximity to the lesion tract. Perivascular plexuses were located on both the medial and lateral boundaries of the knife track and extended dorsal to the paraventricular nucleus. In all cases complete loss of the ipsilateral PVN had occurred, while the ipsilateral SON had remained intact. A differential growth response was observed between the oxytocin and vasopressin producing neurons. These data clearly indicate a regenerative sprouting response of oxytocinergic neurons and the absence of growth stemming from vasopressinergic neurons. In our preparations the occurrence of these plexuses was never observed prior to 8 weeks following the knife cut. a much more rapid response. Others have reported Dellman et al (1987a) observed neurophysin- immunoreactive axons within the perivascular space of vessels surrounding a knife tract by 5 days after transection of the HNT tract in the rat. Antunes et al., (1979) reported the appearance of neurophysin-immunoreactive perivascular plexus in the rhesus monkey no earlier than 8 weeks following hypothalamic lesion. The magnitude of this response observed by Antunes (1979) was quite remarkable. The neurosecretory plexuses were found surrounding vessels throughout the hypothalamus as well as the ventral thalamic regions. Formation of the most dorsal plexuses would require growth over distances of several millimeters from the ipsilateral SON. Our observations of plexuses both proximal and distal to the ipsilateral SON raises questions regarding the origin of the axons involved. One cannot exclude the possibility that the plexuses located on the distal side of the lesion tract, particularly those found in the ventral thalamus, originated from the contralateral SON and/or PVN nuclei. However, the lack of clearly defined immunoreactive axons coursing toward the plexus from either nuclei precludes any attempt to determine the ^ource of origin of these presumptive neurosecretory terminals. Having established the remarkable propensity of the MNS to undergo compensatory axonal growth following injury, we next undertook an 75 investigation of the stimulus responsible for illiciting the growth response. Therefore, chronic measurements of urine osmolality, urine excretion volume, and water intake were undertaken with the expectation that any initial deficit in VP secretion following partial denervation of the NL might be reversed by subsequent compensatory sprouting of VP axons. However, the complete absence of any lesion-induced deficit, other than a very brief post-surgical depression in urine osmolality, makes it impossible to correlate the observed sprouting with functional recovery. The absence of a functional deficit in VP secretion was not the result of an incomplete unilateral lesion, since all lesions were confirmed post-mortem prior to data analysis. In addition, animals receiving bilateral hypothalamic lesions using the same stereotaxic coordinates exhibited lowered urine osmolality as well as heightened water intake and urine excretion characteristic of diabetes insipidus (unpublished observations). Rather it appears that the reserve capacity of the VP system is such that destruction of almost half of the magnocellular neurosecretory terminals is insufficient to compromise its normal function. This conclusion is consistent with the report of Jones and Pickering (1972) who demonstrated that the normal daily requirement of VP in water balanced rats comprises only 5% of the total neurohypophysial content. Thus, in retrospect, it is not surprising that unilateral transection of the hypothalamo-neurohypophysial tract failed to compromise normal VP secretion. No attempts were made to determine whether any functional deficit in VP release might become apparent following partial denervation of the NL if subjects were subjected to hyperosmotic stress or water deprivation. The decrease in urine excretion volumes and daily drinking volume was also an unexpected observation. One possible explanation for these events is that the lesion may have induced an increase in tonic VP secretion by disrupting the contralateral connection between the PVN and SON nuclei. The PVN nuclei are characterized by reciprocal synaptic 76 connections between oxytocinergic neurons of the PVL and PVM subdivisions (Silverman et al., 1980). The SON nuclei have been reported to have polysynaptic reciprocal connections as well (Takano et al., 1990). Although the precise function of these reciprocal connections remains enigmatic, they may be involved in coordinating secretory activity between their respective nuclei. Hence, their disruption could result in the disinhibition of VP secretion, resulting in reduced daily water intake and a concomitant decrease in urine excretion volume. The chronic urine hyperosmolality and decreased urine excretion volume observed following the lesion indicates that a functional deficit in circulating VP is not required to induce compensatory sprouting of magnocellular neurosecretory axons. Herman et al. (1986) have hypothesized that activation of damaged neurons may play a role in determining their regenerative success following injury. These authors have demonstrated that continuous infusion of VP will selectively impair vasopressinergic cell survival following neurohypophysectomy (Herman et al, 1987), an effect which may be attributed to negative feedback of peripheral VP on vasopressinergic cell activity (Herman et al, 1987; Hubert et al, 1989). Likewise, sprouting of intact parvocellular vasopressinergic axons into the zona externa of the median eminence following contralateral lesion of the PVN was only apparent when hyperactivity of these neurons was induced by adrenalectomy (Silverman and Zimmerman, 1982). The precise mechanisms underlying the VP hypersecretory activity observed in the present experiments remain unresolved. However, the reduced water intake may have led to increased neurosecretory activity by intact vasopressinergic neurons, as reflected in increased ^irine osmolality. Thus, increased neuronal activity may be associated with the sprouting response, whereas deficits in circulating hormone levels play no role in the compensatory sprouting response previously reported by Watt and Paden (1991). 77 Neuronal activity was further investigated by examining the neuronal hypertrophy which occurs in the SON following partial denervation of the NL (Watt and Paden, 1991). Each neuronal population was specifically immunolabeled allowing a direct and highly selective comparison to be made of the cross sectional area of oxytocin and vasopressin somata and their respective cell nuclei. The results confirm and extend our previously reported observations of cellular hypertrophy accompanying the compensatory sprouting response (Watt and Paden, 1991). A highly significant cellular enlargement was observed in both populations of neurons following surgery with each showing a similar temporal pattern of response. By 10 days PS both the oxytocinergic and vasopressinergic somata had enlarged above age-matched controls. This was an unexpected observation as previous investigations gave no indications of a cellular response prior to 30 days PS (Watt and Paden, 1991). the two groups. However, the dynamics of this response differed between The greater enlargement of the oxytocin vs vasopressin neurons suggests either an earlier onset of hypertrophy and/or sprouting of oxytocin neurons than VP neurons or that a heightened level of activity may be necessary to support the growth of new terminals within the hypothalamus as well as the N L . In a previous investigation (Watt and Paden, 1991) we had argued that the neuronal hypertrophy observed was a reflection of the heightened cellular activity necessary to support the ongoing axonal sprouting and continued maintenance of the enlarged terminal arbor (Watt and Paden, 1991). This interpretation was based on the tight correlation observed between the onset of cellular enlargement at 30 days PS and the robust sprouting event first detected at this time. Our interpretation was consistent with the reports of others demonstrating a correlation between the onset of sprouting of intact afferents and hypertrophy of the host neurons in other CNS systems (Goldschmidt and Steward, 1980; Headen et al., 1985; Hendrickson and Dineen, 1982; Pearson et al., 78 1987a, 1987b). Likewise, in the SON neuronal hypertrophy has been shown to occur concomitant with the reformation of neurosecretory axon terminals in the proximal infundibulum following hypophysectomy (Raisman, 1977). Hypertrophy of magnocellular neurons has also been associated with increased neurosecretory activity induced through osmotic stimulation via dehydration or salt loading (Enestrom, 1967; Morris and Dyball, 1974; Peterson, 1966; Reinhardt et al., 1969). For example, dehydration of 1.5, 3 and 5 days will result in an enlargement of 25, 28 and 86% in SON neuronal cross-sectional area, respectively (Reinhardt et al., 1969). With 12 days of dehydration SON cellular enlargement reached 180% of normal (Enestrom, 1967). However, these effects were notably transient with a rapid return to normal following cessation of the osmotic stress (Enestrom, 1967; Morris and Dyball, 74). Watt and Paden (1991) reported a delay of 30 days observed prior to the detection of a cellular hypertrophy which led to the conclusion that the hypertrophy was correlated solely with the sprouting response and not related to a heightened secretory activity. The present results necessitate a reevaluation of these conclusions. That heightened cellular activity can result in enlargement of the active cell appears indisputable. In this paradigm two factors may contribute to such a response; the osmotic stress induced by having effectively decreased the volume of daily drinking intake with the subsequent increase in secretory activity, and the heightened synthetic activity necessary to both provide the materials required for terminal growth and the continued maintenance of the enlarged terminal arbor. Although the time of initiation of the sprouting process has not yet been elucidated the duration of the growth has conclusively been shown to proceed for at least 90 days post-surgery (Watt and Paden, 1991). Likewise, the chronic increase in urine osmolality begins within 2 days and continues throughout the 90 day investigative period. Hence, either of these two responses could account for the cellular hypertrophy and 79 each probably contributes substantially to the overall effect. Determining the relative contribution of each would certainly be the logical next step in the pursuit of this investigative paradigm. Non-neuronal components of the magnocellualr neurosecretory system may also play a role in the initiation and maintanace of the compensatory sprouting response. Analysis of the vasculature in the partially denervated NL illustrated several distinct alterations resulting from the lesion. A significant increase in proportional area of the NL occupied by vessel lumen 48 hours following denervation indicates a vasodilatory response which then wanes by 5 days PS with no further change. This dilatory response presumably corresponds with increased blood flow in the NL at a time when hypersecretion is thought to occur. An increase in blood flow and vasodilation of NL vasculature has been reported to accompany moderate stimulation of hormone release (Sooriyamoorthy and Livingston, 1972, 1973). Under more chronic stimulation endothelial proliferation has also been reported (Paterson and Leblond, 1977), exemplifying the apparent malleability of the NL vasculature to enhanced metabolic demand. The return of luminal area to normal presumably reflects the stabilization of fluid balance in these animals although hypersecretory activity is maintained throughout the experimental period. An increase in the density of blood vessels also arises as a result of the lesion and correlates closely with the temporal pattern of axonal sprouting reported previously (Watt and Paden, 1991). The relative increase in vascular density, which presumably corresponds to an increase in perivascular basement membrane, may provides additional target area for the sprouting fibers. However, under these conditions the partial denervation would result in a surplus of target area. It seems more likely that the hypervascularization occurs either as a result of the heightened secretory activity which remains elevated throughout the 90 day post-surgical period or as a result of endothelial 80 stimulating factors arising from activated non-neuronal cells within the NL. A number of factors may be involved in the induction of collateral vessel growth. Hypoxia associated with areas of high metabolic activity stimulates neovascular growth (Knighton, et al., 1981). Hypoxia may in turn stimulate the release of angiogenic factors from macrophages (Knighton, et al., 1981). Macrophages are the major cell component of wound fluid (Greenburg and Hunt, 1978) and act to condition the fluid with secreted factors (Banda, et al., 1982; Knighton, et al., 1983) which in turn can induce a cascade of events leading to neovascularization of the area surrounding the wound site. Ameboid microglia, thought to arise from a monocyte line (Perry and Gordon, 1988), are considered brain macrophages and secrete a variety of stimulatory factors including IL-I (Fagen and Gage, 1990; Giulian et al., 1986) which may in turn stimulate the proliferation of fibroblasts (Greenburg and Hunt, 1978; Knighton, et al., 1983). The activated fibroblasts may then release angiogenic factors such as acidic and basic fibroblast growth factors (Y anagisawa-Miwa, et al., 1992) or secrete extracellular matrix components which may stimulate angiogenesis as well (Young and Herman, 1985). Substrates of fibronectin and interstitial collagen have been shown to induce a rapid growth response of cultured endothelial cells (Young and Herman, 1985). In addition, mechanical factors such as shear stress resulting from intraluminal turbulence or heightened intraluminal pressure (Yanagisawa-Miwa, et al., 1992) and vasodilation (D'amore and Thompson, 1987) may promote collateral vessel outgrowth. In addition to questions regarding the stimulatory factors involved in collateral vessel outgrowth the possible role of hypervascularity in the compensatory sprouting response must also be explored. During embryogenesis the NL anlage is penetrated first by blood vessels then by the neurosecretory fibers. Maturation of both 81 fibers and capillary plexus then proceeds together indicating a possible trophic interaction between vessels and neurosecretory fibers during development (Galabov and Scheibler, 1983). Likewise., following hypophysectomy the initial development of a capillary plexus within the relatively avascular stalk precedes the penetration of regenerating neurosecretory fibers (Billenstein and Leveque, 1955; Moll, 1958; Raisman, 1973), leading to the suggestion that capillary growth may act as a stimulus for the subsequent regenerative sprouting of neurosecretory fibers (Raisman, 1973). A similar regenerative pattern has been observed in the hypothalamus. Following transection of the hypothalamo-neurohypophysial tract the proximal segments of the transected neurosecretory fibers will form new neuroheamal contacts with the neovasculature surrounding the lesion tract (Danilova and Polenov, 1977; Antunes et al., 1979; Dellman et al., 1987a) and with resident vessels of the hypothalamus not previously innervated by neurosecretory fibers (Antunes et al., 1979). Considerable growth of neurosecretory axons has also been observed in the ventral leptomeninges and surrounding pial vessels following disruption of the ventral hypothalamic surface (Danilova and Polenov, 1977; Antunes et al., 1979; Dellman et al., 1988). Stereotaxic placement of lntrahypothalamlc allographs of sciatic nerve, optic nerve and hypodermal connective tissue within the HNT has also been demonstrated to result in varying degrees of penetration by neurosecretory axons (Dellman et al., 1985a, 1985b, 1986a, 1987). The relative density of invading neurosecretory axons appears to be dependent upon the degree of vascularization (Dellman et al., 1987b), with the highest density of neurosecretory fibers always found in the immediate vicinity of blood vessels (Dellman et al., 1989). These observations indicate that the vasculature must play a prominent role in the attraction and maintenance of the neurosecretory axonal ingrowth. Although the data from this study do not exclude the possible, and likely, contribution of pituicytes to the 82 development and growth of neurosecretory axons in the normal animal, they do indicate the underlying importance of blood vessels and the various components of the perivascular network to a regenerative and/or compensatory sprouting response within the compromised neurosecretory system. In addition to the vasculature the pituicyte, parenchymal microglia and perivascular cells also play a role in the response of t h e , MNS to partial denervation. Therefore a series of investigations were undertaken to describe specific attributes of the glial response to denervation. The first was to investigate the role of nerve growth factor receptor regulation in the MNS response. Increased levels of NGFR immunoreactivity have been reported during the massive degeneration that occurs in the posterior pituitary following infundibular stalk transection (Yan et al., 1990). However, that preparation does not allow for regeneration of neurosecretory axons. We chose to explore the expression of NGFR immunocytochemically in the partially denervated NL during the degenerative phase and the subsequent sprouting response as a means of further elucidating the role of NGFR expression and the contribution of microglial cell activity to the sprouting event. Results clearly demonstrate an increase in the proportional area of the NL occupied by NGFR-immunoreactive elements during the first 10 days of post-denervation response (Fig. 16). This is a time when degeneration of the severed axons and phagocytosis o f .the resulting axonal debris is at a maximum (Watt and Paden, 1991).. Yan et al. (1990) have demonstrated 192-IgG (NGFR) immunoreactivity on the surface of perivascular microglial cells and what appeared to be microglial cells within the parenchyma of the NL following stalk transection. These cells are thought to be solely responsible for phagocytosis of axonal debris in the partially denervated NL following bilateral electrolytic destruction of the paraventricular nuclei (Zambrano and de Robertis, 1988) or unilateral transection of the hypothalamo-neurohypophysial tract (Watt 83 and Paden, 1990) and during the turnover of neurosecretory axons in the intact NL (Pow et al., 1989). Pituicytes have not been conclusively demonstrated to express the NGF receptor (Yan et al., 1990). The return of NGFR proportional area to normal values by 30 days PS could indicate either a downregulation of NGF receptor expression by microglia or a reduction in the number of microglia expressing this receptor. This reduction in NGF receptor immunoreactivity was observed at a time when phagocytic activity was complete and a robust compensatory sprouting response was in progress (Watt and Paden, 1991). These observations indicate that the principle period of activation of the microglial cells within the NL is concurrent with phagocytic activity. The role of NGF receptor expression in the initiation or maintenance of the sprouting response remains undetermined. However, the activity of the NGFR expressing glia may activate the pituicytes (Giulian and Young, 1986) thus preparing the NL for the subsequent sprouting event. Comparisons of 192-IgG, OX-42 and OX-I staining patterns suggest some differences in the distribution of NGFR immunoreactive microglial cells when compared to the other microglial labels (Fig. 19). The NGFR staining pattern is reticular and velate in appearance and is localized primarily surrounding blood vessels and extending outward in long finger-like projections. Although the limiting basement membrane of the perivascular space cannot be readily discerned at the light level, the general pattern of NGFR staining suggests that the majority of NGFR immunoreactive cells are located within this compartment. Ultrastructural analysis will be required to confirm these observations. In contrast, the OX-42 and OX-I staining patterns are indicative of isolated, stellate cells scattered uniformly throughout the NL parenchyma and in close proximity to, but not encompassing, blood vessels. Nor are these cells joined in the long wispy pattern observed with NGFR staining. These observations indicate that the NGFR immunoreactive cells represent a subpopulation of NL microglial cells 84 comprised predominantly of the perivascular type. The increase in staining area which accompanies the post-lesion response may reflect the expression of NGF receptors on microglial cells that have entered a phagocytic state. Ultrastructural analysis of the degenerating NL demonstrates the presence of degenerating axons within the perivascular compartment at 10 days PS (unpublished observations). Alternatively, the perivascular microglial cells, having become activated by the degeneration or injury, may then migrate into the parenchymal compartment to engulf axonal debris (Perez-Figares et ail., 1986). Yan et al., (1990) reported the infrequent occurrence of NGFR immunoreactive microglia located directly within the parenchymal basement membrane following stalk transection. These immunoreactive cells were typically within necrotic areas indicative of phagocytic activity. The reduction in relative proportional area immunoreactive for NGF receptors by 30 days PS supports this hypothesis. The age-related increase in proportional area immunoreactive for NGFR is in agreement with earlier reports (Yan et al, 1990). The similarity in staining intensity and pattern observed between young and mature animals suggests a similar underlying mechanism involved in both the lesioned animal and in mature aging rats. Pow et al (1989) have reported the involvement o f ,perivascular microglial cells in the engulfment of otherwise healthy neurosecretory axons in the intact N L . These authors suggest that growing neurosecretory axons penetrate the parenchymal basement membrane and are then engulfed by the resident microglia. This activity may reflect a continuous remodeling of the neurosecretory terminal field. I have also observed an increased incidence of axonal engulfment by perivascular microglial cells in the 90 day PS animal, supporting the possibility of such remodeling (unpublished observations). The turnover of intact neurosecretory terminals may provide an explanation for the age-related increase in NL areas occupied by NGFR immunoreactive cells. 85 In summary, an increase in proportional area of NL immunoreactive for NGF receptor was observed during the first 10 days following partial denervation and as a result of normal aging. Comparison of staining patterns of known microglial markers with NGFR suggests a subpopulation of microglial cells may be involved in this response. This in turn may reflect the phagocytic activity ongoing during the period of heightened receptor immunoreactivity . The return of NGFR staining area to normal levels following this period suggests little direct contribution of NGFR expression to the sprouting response. CNS injury either as a result of peripheral axotomy or penetrating brain injury initiates a cellular response from the non-neuronal elements surrounding the sites of injury and/or distal degeneration. This response may include proliferation of glial cells (Gall et al., 1979: Schwartz et al., 1988; Steit and Kreutzberg 1987; Kreutzberg and Barron, 1978), invasion of blood born elements including macrophages and monocytes (Perry et al., 1987), or migration and hypertrophy of astrocytes (Gall et al., 1979). Increased glial proliferation has also been demonstrated in the SON and NL following chronic water deprivation or NaCL imbibement (Murray, 1968; Krsulovic and Bruckner, 1969; Paterson and Leblond, 1977), pituicytes have since been shown to be one of the proliferating cell types (Murugaiyan et al., 1992). In the present investigation the proliferative response of the glial population within the partially denervated NL was studied using counts of cell nuclei on semithin plastic sections, bromodeoxyuridine substitution and immunocytochemistry. Counts of pituicyte nuclei revealed a significant increase in the density of pituicytes within the NL from 2-30 days following surgery, thus supporting earlier findings of increased volume occupied by glial elements (Watt and Paden, 1991). However, the substantial increase in pituicyte density observed within 48 hours of deafferentation is surprising in light of previous reports of delayed activation of astrocytes within regions of axonal degeneration (Gall et 86 al., 1979; Fagen and Gage, 1990). In this paradigm all nuclei were counted in a given coronal section thus eliminating the possibility of sampling error as a result of a heterogeneous distribution of glia. One cannot exclude the possibility of shrinkage in these tissues, yet the cross-sectional areas Of the sections examined across experimental groups offer no indication of such an effect. In, those cases where an equivalent sample area was compiled between two age-matched groups e.g., PS 2 day vs lesion day 0 control and PS 30 day lesion vs PS 30 day control, the absolute numbers of glia within the lesion groups are substantially higher than the age-matched control (Table 2). These data support indications of an increase in pituicyte numbers, while arguing against the possibility of a shrinkage-related artifact. Morphometric analysis of pituicyte nuclear cross-sectional area was done to determine the contribution of pituicyte cellular hypertrophy to the post-lesion glial response. Neuronal hypertrophy within the magnocellular neurosecretory system has been demonstrated to occur as a result of heightened metabolic activity (Watson, 1965; Murray, 1968; Paterson and Leblond, 1977) and is an accepted indicator of heightened cellular activity (Vogels et al., 1990). The enlargement of magnocellular neuronal nuclei during heightened cellular activity has been demonstrated in this and a previous investigation (Watt and Paden, 1991). Therefore, the possibility of such a response by activated pituicytes was investigated. However, measures of pituicyte nuclei during the post-lesion period reflect no increase in nuclear size. On the contrary, a significant reduction in nuclear size was observed at 2 days PS. This reduction occurs at a time when cell numbers are rapidly increasing, and may provide further indirect evidence of pituicyte proliferation. Following cell division the nuclei of both parent and daughter cells may be reduced in size. The subsequent return to normal size over the next 3-8 days would indicate reduced pituicyte proliferation, with subsequent growth and maturation of the post-mitotic 87 cells. Hence, these data may also reflect the time course over which the proliferative event may occur. The absence of nuclear hypertrophy may indicate the absence of cellular hypertrophy coinciding with the post-lesion response. However, to my knowledge the incidence of glial nuclear enlargement coinciding with cellular enlargement has not previously been reported for any glial type. Therefore, these data do not necessarily reflect the absence of pituicyte cellular hypertrophy per se. The occurrence of astrocyte cellular hypertrophy within deafferented regions (Rose et al., 1976; Visayan et al., 1991) and surrounding wound sites (Author, unpublished observations) has been repeatedly observed. Thus, it remains a distinct possibility that such an event may be occurring in the partially denervated NL and warrants further investigation. Microglial nuclear counts were compiled in the same plastic coronal sections discribed above. These data reflect an insignificant increase in nuclear numbers between 2 and 5 days PS. By 10 days PS the numbers of microglial cells have returned to normal values where they remain for the rest of the investigative period. The transient increase in cell numbers observed here coincides with the period of degeneration and phagocytosis of axonal debris which occurs over the first 10 days following knife cut (Watt and Paden, 1991). These data agree very closely with those of Kreutzberg and Barron (1978) who reported a rapid proliferation of microglial cells in the facial nucleus 2-5 days following facial nerve axotomy. activity had ceased. Thereafter, all proliferative Proliferation of microglial cells has also been reported in sites of neuronal degeneration distant from the site of axotomy (Kreutzberg and Barron, 1978) and surrounding the site of transection in the optic nerve (Perry et al., 1987). The microglial response typically precedes astrocyte activation (Fagen and Gage, 1990) and may be responsible for the release of factors which stimulate astroglial activity (Giulian and Young, 1986). Ameboid microglia are 88 considered brain macrophages and are responsible for the removal of degenerating tissue in the CNS. The early, albeit limited, increase in microglial cells reported here correlates closely with the post-lesion degeneration observed in the N L . The return of microglial numbers to normal levels following this degenerative phase indicates an indirect role of microglia in the sprouting response. However, it must be emphasized that the insignificant increase in microglial density observed suggests a limited role at best. Although the glial nuclear counts provide strong indirect evidence of glial hyperplasia, further investigation of this response was undertaken using Brdu substitution to specifically label the cells undergoing mitosis. All animals were injected at approximately the same time of day (10-llam) and sacrificed 3 hours later. This provided control for circadian differences in the cell cycle from animal to animal and across groups. The normal cell cycle for astrocytes is 20 hours with an S phase of 9.4 hours; G2 less than 3 hours; G2+M approx. 5 hours, and G1 lasts 5 hours (Korr et a l ., 1979). Therefore, the Brdu substitution period of 3 hours utilized here must be interpreted as a snapshot of the ongoing proliferative activity and as such, a conservative estimate of the percent population of glial cells in a proliferative state. In the normal animal (35 days post-parturition) the percent population undergoing DNA synthesis in the NL was found to be 0.4% of the total glial population (Table 3). Korr et al., (1979) has reported a similar ratio of growing glial cells to all other glial cells of 0.4% in mouse brain. Within 24 hours of partial denervation we observed an increase to 1.2% of the total glial population representing a 65% increase over control values. At two days PS a peak value of 1.6% of the glial population was found in a growth state, a 76% increase over control values. The growth activity then begins a gradual decline over the next 4 days, reaching levels below normal (ns) by 7 days PS, the latest time point thus far examined. 89 Brdu immunopositive nuclei were characterized by a black DAB-CoCL2 precipitate localized within the nuclear envelope, thus allowing stained nuclei to be easily discerned from unlabeled, counterstained nuclei. The distribution of Brdu nuclei was essentially random throughout the sections with no apparent segregation within specific compartments e.g., perivascular space, surrounding blood vessels or within the parenchyma, making it difficult to ascertain the specific cell type(s ) that had been labeled based on nuclear morphology at the light level. Therefore, dual label fluorescence lmmunocytochemistry was utilized as a means of identifying the Brdu immunopositive cell species. These results indicate an absence of co-localization of Brdu nuclei with GFAP immunopositive cell bodies in all sections examined. However, the relatively thin 6 micron sections and the limited number of sections analyzed leave room for the possibility of pituicytes (GFAP+) which are Brdu positive but were not observed in the course of this investigation. Ultimately, confirmation of the cell types exhibiting Brdu substitution will require E.M. immunocytochemical localization or further dual label immunocytochemical analysis at the light level. Pituicyte numbers may increase as a result of migration of pituicytes and/or astrocytes from the infundibular stalk. Astrocyte migration into regions of denervation and axonal degeneration have been reported to occur within the partially denervated hippocampus (Gall et al., 1979). The pituicyte is a characteristically motile cell particularly during periods of heightened neurosecretory activity (Tweddle and Hatton, 1980; Hatton et al.,1984; Montagnese et al., 1988). Hence, the pituicyte displays all the attributes necessary for such a response. Dual fluorescence labeling of cells with Brdu and GSA-Lectin confirmed the presence of microglial hyperplasia in the N L . Some caution must be exercised in the interpretation of these data however, as GSA-Lectin will also stain invading monocytes and macrophages. Given 90 the extensive axonal degeneration and the lack of a blood brain barrier in the NL the presence of invasive blood-born cells cannot be discounted. However, invading leukocytes were never observed in the course of extensive analysis of high-clarity plastic semithin sections of NL or during exhaustive ultrastructural analysis completed in a previous investigation (Watt and Paden, 1991). Given that such an event is possible it must be a rare occurrence at best. Summary The compensatory collateral sprouting of intact neurosecretory fibers which follows partial denervation of the NL in the rat provides a useful model system to investigate the cellular correlates of both degenerative and sprouting events. Unilateral transection of the hypothalamo-neurohypophysial tract results in the loss of over 40% of the axons within the NL terminal field (Watt and Paden, 1991). A period of axonal degeneration and phagocytosis by resident microglial cells then occurs followed by the .compensatory sprouting of the remaining intact neurosecretory fibers within the terminal field between 10 and 90 days post-lesion (Watt and Paden, 1991). The period over which these events occur can be divided into two principle phases, a degenerative phase including approximately the first 10 days PS, and a sprouting phase which follows and includes the first 90 days PS. During the degenerative phase lesion-induced activity first becomes evident with the appearance of Brdu labeled glial nuclei within 24 hours PS, and the subsequent increase in pituicyte and microglial cell numbers counted in cross sections of NL within 48 hours PS. return to normal values by 10 days PS. Microglial cell numbers then However, the pituicyte numbers remain elevated at least 30 days PS, well into the sprouting phase. The degenerative phase is further characterized by the metabolic alterations in urine osmolality, urine excretion volumes and drinking volumes and the acute vasodilation of the NL vasculature within 48 hours PS. Between 5 and 10 days PS continued glial activity was indicated by the 91 increase in proportional area of NL immunoreactive for nerve growth factor receptors. However, b y 10 days PS, the approximate end of the degenerative phase, NGFR proportional area had returned to normal. By the end of the degenerative phase, as defined by the lack of axonal debris and phagocytic activity within the NL (unpublished observations), a marked hypertrophy of the contralateral intact neurosecretory neurons was observed. The neuronal enlargement, indicative of heightened cellular activity, continued through at least 30 days PS before returning to normal by 90 days PS. The pattern of neuronal hypertrophy correlated closely with the sprouting phase and may reflect the heightened cellular activity necessary to sustain the sprouting response. The occurrence of hypothalamic perivascular neurosecretory plexuses by 30 days PS provided further indications of cellular activity and highlight the propensity of the MNS for structural plasticity. Finally, an injury-related increase in the density of blood vessels in the NL was also observed between 30 and 90 days PS, coinciding with the sustained increase in pituicyte density through the sprouting phase. In conclusion, anatomical and metabolic aspects of the temporal pattern of cellular activity within the partially denervated NL have been described. These data offer indications of the cellular activity which may underlie the sprouting response but add little to our present understanding of the mechanisms responsible for initiating and maintaining such a response. However, the anatomical simplicity of the N L , the robust nature of the MNS sprouting response, and the anatomical data described herein provide an excellent model system for the continued investigation of factors involved in the CNS response to injury. 92 REFERENCES CITED Adams, J.H., Daniel, P.M. and M.M.L. Prichard (1969) Degeneration and regeneration of hypothalamic nerve fibers in the neurohypophysis after pituitary stalk section in the ferret. J. Comp. Neurol. 135:121-144. Adams, J.H., Daniel, P.M., and M.M.L. Prichard (1971) Changes in the hypothalamus associated with regeneration of the hypothalamoneurohypophysial tract after pituitary stalk section in the ferret. J. Comp. Neurol. 142:109-124. Aleta J.M., S.A. Lewis, N.J. Cowan and L.A. Greene (1988) Nerve growth factor regulates both the phosphorolation and steady-state levels of microtubule-associated protein 1.2 (MAP 1.2). J. of Cell Biol. 106:1573-1581. Alonso, G., and I . Assenmacher (1981) Radioautographic studies on the neurohypophysial projections of the supraoptic and paraventricular nuclei in the rat. Cell Tissue Res. 219;525-534. Alonso, Cl., Szafarczyk, A., and I. Assenmacher (1986) Radioautographic evidence that axons from the area of supraoptic nuclei in the rat project to extrahypothalamic brain regions. Neuroscience Letters 66:251-256. Antunes, J.L., Carmel, P.W., and E.A. Zimmerman (1977) Projections from the paraventricular nucleus to the zona externa of the median eminence of the rhesus monkey: An immunocytochemical study. Brain Res. 137:1-10. Antunes, J.L., Carmel, P.W., Zimmerman, E.A., and M. Ferrin (1978) Regeneration of the magnocellular system of the rhesus monkey following hypothalamic lesions. Ann. Neurol. 5:462-469. Armstrong, W.E., Warach, S., Hatton, G.I., and T.H. McNeill (1980) Subnuclei in the rat hypothalamic paraventricular nucleus: A cytoarchitectural, horseradish peroxidase and immunocytochemical analysis. Neuroscience 5.:1931-1958. Assouline, J.G. and N.J. Pantazis (1989) Detection of a nerve growth factor receptor on fetal human Schwaan cells in culture: Absenceof the receptor on fetal human astrocytes. Dev. Brain Res. 45:1-14. Banda, M. J., D . R. Knighton, T. K. Hunt and Z . Werb (1982) Isolation of a nonmitogenic angiogenesis factor from wound fluid. Proc. Natl. Acad. Sci. USA 79:7773-7777. Bandaranayake, R.C. (1976) Localization of functions in the magnocellular neurosecretory nuclei of the mammalian hypothalamus by autoradiography. Acta. Anat. 95:408-420. Barer, R., and K. Lederis (1966) Ultrastructure of the rat neurohypophysis with special reference to the release of hormones. Zellforschung. 75:201-239. Baumgarten, H.G., A. Bjorklund, A.F. Holstein and A. Norin (1972) Organization and ultrastructural identification of the catecholamine nerve terminals in the neural lobe and pars intermedia of the rat pituitary. Z . Zellforschung. 126:483-517. Z. 93 Beck, E., and P.M. Daniel (1959) Some changes in the hypothalamus and proximal pituitary stalk after stalk section. J. Physiol. Lond. 146:22-24. Beck, E., Daniel, P.M., and M.M.L. Prichard (1969) Regeneration fof hypothalamic nerve fibers in the goat. Neuroendo. 5:161-182. Berg, M.M, D.W. Sternberg, B.L. Hempstead and M.V. Chao (1991) The lowaffinity p75 nerve growth factor (NGF) receptor mediates NGF-induced tyrosine phosphorylation. Proc. Natl. Acad. Sci. USA 88:7106-7111. Bernd, P. and L.A. Greene (1984) Association of I125 nerve growth factor with PC12 pheochromocytoma cells. J. Biol. Chem. 259:15509-15516. Bernd, P., H.J. Martinez, C.F. Dreyfus and I.B. B1Iack (1988) Localization of high-affinity and low-affinity nerve growth factor receptors in cultured rat basal forebrain. Neurosci. 26(1):121-129. Billenstien, D.C., and T.F. Leveque (1955) The reorganization of the neurohypophysial stalk following hypophysectomy in the rat. Endocrinology 56:704-717. Bjorklund, A., Katzman, R., Stenevi, U., et al. (1971) Development and growth of axonal sprouts from noradrenaline and 5-hydroxytryptamine neurons in the rat spinal cord. Brain Res. 31:21-31. Bjorklund, A., and U. Stenevi (1979) Regeneration of monoaminergic and cholinergic neurons in the mammalian central nervous system. Physiol. Rev. 59(1):62-100. Brown, J.O., and G .P .McCouch (1947) Abotive regeneration of the transected spinal cord. J .Comp. Neurol. 87:131-137. Buck, C.R., H.J. Martinez, M.V. Chao and I.B. Black (1988) Differential expression of nerve growth factor receptor gene in multiple brain areas. Dev. Brain Res. 44:259-268. Cajal, S . Ramon y. (1928) Degeneration and regeneration of the nervous system. (ed. R.M. May) Hafner Pub. Co. NY, (1968). Cammer, W., Tansey, F.A. and C.F. Brosnan (1990) Reactive gliosis in the brains of Lewis rats with experimental allergic encephalomyelitis. J. of Neuroimmunology 27:111-120. Cavanagh, J.B. (1970) The proliferation of astrocytes around a needle wound in the rat brain. J. Anat. 106:471-487. Christ, J.F . (1969) Derivation and boundaries of the hypothalamus, with atlas of hypothalamic grisae. In The Hypothalamus, (eds., Haymaker, W., Anderson, E., and W.J.H. Nauta) pp. 13-60, Charles C. Thomas, Springfield, 111. Clark, W.E. Ie G. (1943) The problem of neuronal degeneration in the central nervous system II. The insertion of peripheral nerve stumps in the brain. J. Anat. 77:251-259. Clemente, C.D., and W.F. Windle (1954) Regeneration of severed nerve fibers in the spinal cord of the adult cat. J. Comp. Neurol. 101:691731. 94 Clemente, C.D. (1955) Structural regeneration in the mammalian central nervous system and the role of neuroglia and connective tissue. in Regeneration in the central nervous system, (ed. W.F. Windle) Springfield, 111. Thomas Press, pp. 147-161. Clemente, C.D. (1964) Regeneration in the vertebrate central nervous system. Int. Rev. Neurobiol. 6:258-301. Cotman, C.W., and G.S. Lynch (1976) Reactive synaptogenesis in the adult nervous system: The effects of partial deafferentation on new synapse formation. in Neuronal Recognition (ed. S. Barondes), Plenum Press, NY pp. 69-108. Cotman, C., Gentry, C., and O. Steward (1977) Synapse replacement in the dentate gyrus after unilateral entorhinal lesion: Electron microscopic analysis of the extent of replacement of synapses by the remaining entorhinal cortex. J. Neurocytology 6:455-464. CotmanfC.W., Nietro-Sampedro, M., and E.W. Harris (1981) Synapse replacement in the central nervous system of adult vertebrates. Physiol. Rev. 61(3):684-784. Cotman, C.W., M. Nietro-Sampedro (1984) Cell biology of synaptic plasticity. Science 225:1287-1294. Coulter, D.H., Elde, R.P., and S.L. Unverzagt (1981) CO-Localization of neurophysin- and enkephalin-Iike immunoreactivity in cat pituitary. Peptides 2.:51-55. Cowan, W.M. (1970) Anterograde and retrograde trans-neuronal degeneration in the central and peripheral nervous system. in Contemporary Research Methods in Neuroanatomv. W.J.H. Nauta and S.O.E. Ebbesson eds., Springer-Verlag, NY pg 217-251. D 'Amore, P.A. and R.W. Thompson (1987) Mechanisms of angiogenesis. Rev. Physiol. 49:453-464. Ann. Daniel, P.M. and M.M.L. Prichard (1975) Studies of the hypothalamus and pituitary gland. Acta. Endocrinol. Suppl. 201:21-37. Dellman, H-D., and P.A. Owsley (1969) Investigations of the hypothalamoneurohypophysial neurosecretory system of the grass frog (rana pipiens) after transection of the proximal neurohypophysis. Z . Zellforschung. 94:325-336. Dellman, H-D, and E.M. Rodriguez (1970) Herring bodies; an electron microscopic study of local degeneration and regeneration of neurosecretory axons. Z . Zellforschung. 111:293-315. Dellman, H-D (1973) Degeneration and regeneration of neurosecretory systems. Int. Rev. Cytol, 36:215-315. Dellman, H-D, Lue, L-F, and S.I. Beilin (1985) A neurophysin histochemical and fine structural study of grafts of sciatic nerve (SN), optic nerve (ON) and connective tissue into the rat hypothalamus. Int. Symp. Neural. Regener. Pacific Grove (Abstr.) pg.45. Dellman, H-D, Lue, L-F, and S.I. Beilin (1986) Peptidergic neurosecretory axons regenerate into sciatic nerve autographs in the rat hypothalamus. Neuroendocrinology 44:292-298. 95 Deliman, H-D., Lih-Fen Lue, and S. Beilin (1987a) Fine structural characteristics of neurophysin-positive perivascular plexus that develope in the rat hypothalamus following interruption of the hypothalamo-neurohypophysial tract. Cell Tissue Res. 247:137-143. Dellman, H.D., Lue, L-F., and S.T. Beilin (1987b) Neurosecretory axon regeneration into intrahypothalamic neural lobe allographs: Neurophysin immunohistochemistry and fine structure. Exp. Brain Res. 67:543-555. Dellman, H-D, Lue, L-F, Beilin, S.I. and M. Ovassat (1988) An immunohistochemical and fine structural analysis of peptidergic hypothalamic neurosecretory axon regeneration into the leptomeninges of the rat. Brain Res. 450:181-189. Dickson, D . and L.A. Mattiace (1989) Astrocytes and microglia in human brain share an epitope recognized by a B-lymphocyte-specific monoclonal antibody (LN-I) Am. J. Pathol. 135( U :135-147. Dierickx, K., Vandesande, F., and J. DeMay (1976) Identification in the external region of the rat median eminence of separate neurophysinvasopressin and neurophysin-oxytocin containing nerve fibers. Cell Tissue Res. 168:141-151. Enestrom, S . (1967) Nucleus supraopticus. A morphological and experimental study in the rat. Acta. Pathol. Microbiol. Scand. Suppl. 186:1-99. Esiri, M. and JO'D. McGee (1986) Monoclonal antibody to macrophages (EMB/ll) labels macrophages and microglial cells in human brain. J. Clin. Pathol. 39:615-621. Eurenius, L. (1977) An electron microscopic study of the differentiating capillaries of the mouse neurohypophysis. Embryol. 157:89-108. Anat. Fagen, A. and F.H. Gage (1990) Cholinergic sprouting in the hippocampus: A proposed role for TL-I. Exp. Neurol. 110:105-120. Finger, S., and C.R. Almli (1985) Brain damage and neuroplasticity: Mechanisms of recovery or development? Brain Res. Rev. 10:177-186. Folkman, J. (1986) How is blood vessel growth regulated in normal and neoplastic tissue? G.H.A. Clowes Memorial Lecture. Cancer Res. 46:467473. Fugisawa, I., K. Kikuchi, K. Nishimura, and K. Togashi (1987) Transection of the pituitary Stalk: development of an ectopic posterior lobe assessed with MR imaging. Radiology 165 ( 2 ) : 487-489. Fugita7 H. and J.F. Hartmann (1961) Electron microscopy of neurohypophysis in normal, adrenaline-treated and pilocarpine-treated rabbits. Z .Zellforschung. 54:734-763. Furikawa, Y., N. Tomioka, W. Sato, E . Satoyoshi, K. Hyashi and S. Furukawa (1989) Catacholaminesincrease nerve growth factor mRNA content in both mouse astroglial cells and fibroblast cells. FEB 247:463-467. Gage, F.H., Olejniczak, P., and D.M. Armstrong (1988) Astrocytes are important for sprouting in the septohippocample circuit. Exp. Neurol. 102:2-13. 96 Galabov7 P., and T.H. Scheibler (1978) The ultrastructure of the developing neural lobe. Cell Tissue Res. 189:313-329. Galabov7 P., and T.H. Scheibler (1983) Development of the capillary system in the neurohypophysis of the rat. Cell tissue Res. 228:685696. Gall7 C., G. Rose and G. Lynch (1979) Proliferative and migratory activity of glial cellsin teh partially deafferented hippocampus. Neuol. 183:539-550. Comp. Gall7 C., and G. Lynch (1981) The regulation of sprouting in adult hippocampus: Some insights from developmental studies. in LesionInduced Neuronal Plasticity in Sensory Motor Systems. (H. Flohr and W. Precht) pg 27-37. , Goldschmidt, R.B., and O. Steward (1980) Time course of increases in retrograde labeling and increases in cell size of entorhinal cortex neurons sprouting in response to unilateral entorhinal lesions. J. Comp. Neurol. 189:359-379. Greenburg7 G. R. and T. K. Hunt (1978) The proliferative response in vitro of vascular endothelial and smooth muscle cells exposed to wound fluids and macrophages. J. Cell Physiol. 97:353-360. Giulian7 D. and T.J. Baker (1985) Peptides released by ameboid microglia regulate astrocyte proliferation. J . Cell Biol. 101:2411-2415 Giulian7 D . and L.B. Lachman (1985) Interleukin-I stimulation of astroglial proliferation after brain infury. Science 228:497-499. Giulian7 D . and T.J. Baker (1986) Characterization of ameboid microgliaisolated from developing mammalian brain. J. Neurosci. 6:21632178. Giulian7 D . and D.G. Young (1986) Brain peptides and glial growth. II.Identification of cells that secrete glial promoting factors. J. of Cell Biol. 102:812-820. Giulian7 D . R. Allen, T.J. Baker and Y.J. Tomazawa (1986a) Brain peptides and glial growth. I . Glia-promoting factors as regulators of gliogenesisin the developing and injured central nervous system. Cell Biol. 102:803-811. Giulian7 D., T.J. Baker, L.N. Shin and L.B. Lachman (1986b) Interleukin1 of the central nervous system is produced by ameboid microglia. J. Exp. Med. 164:594-604. Goldschmidt, R.B. and O. Steward (1980) Time course of increase in retrograde labeling and increases in cell size of entorhinal cortex neurons sprouting in response to unilateral entorhinal lesions. J. Comp. Neurol. 189:359-379. Hatton7 G.I., and J.K. Walters (1973) Induced multiple nucleoli, nucleolar margination and cell size changes in supraoptic neurons during dehydration and rehydration in the rat. Brain Res. 55:137-154. Hatton7 G .I., Perlmutter7 L.S., Salm7 A.K., and C.D. Tweedle (1984) Dynamic neuronal-glial interactions in hypothalamus and pituitary: Implications for control of hormone synthesis and release. Peptides 5(11:121-138. 97 Headon, M.P., Sloper, J.J., Hiorns, R.W., and T.P.S. Powell (1985) Effects of monocular closure at different ages on deprived and undeprived cells in the primate lateral geniculate nucleus. Dev. Brain Res. 18:57-78. Hefti, F. (1986) Nerve growth factor promotes survival of septal cholinergic neurons after fimbrial transections. J. Neurosci. 6:21552162. Hempstead, B.L., D. Martin-Zanca, D.R. Kaplan, L.F. Parada and M.V. Chao (1991) High-affinity NGF binding requires coexpression of the trk proto­ oncogene and the low affinity NGF receptor. Nature 350:678-683. Hendrickson, A., and J.T. Dineen (1982) Hypertrophy of neurons in dorsal lateral geniculate nucleus following striate cortex lesions in infant monkeys. Neuroscience Letters 30:217-222. Herman, J.P., Marciano, F.F., Wiegand, S.J. and D.M. Gash (1987) Selective cell death of magnocellular vasopressinergic neurons in neurohypophesectomized rats following chronic administration of vasopressin. J. Neurosci. 2:2564-2575. Hickey, W.F. and H. Kimura (1987) Perivascular microglial cells of the CNS are bone marrow-derived and present antigen in Vivo. Science 239:290-292. Hjorth-Simonsen, A. (1972) Projection of the lateral part of the entorhinal area to the hyppocampus and fascia dentata. J. Comp. Neurol. 146:219-231. Hommes, O.R. and C.P. Leblond (19 ) Mitotic division of neuroglia in the normal adult rat. J. Comp. Neurol. 129:269-278. Hosang, M. and E.M. Shooter (1987) The internalization of nerve growth factor by high affinity receptors on pheochromocytoma PC12 cells EMBO 6:1197-1202. Ibata, Y., Fukui, K., Obata-Tsuto, H.L., Murabe, Y., and M. Kawata (1983) Changes in the hypothamao-hypophyseal neuron system following hypophysectomy. in Immunoperoxidase techniques, J. Histochem. and Cytochem. 29:577-580. Jacobson, M. (1978) Developmental Neurobiology. Plenum Press, NY, pgs 49-55. Johnson, E.M. Jr., M. Taniuchi and P.S. DiStefano (1988) Expression and Possible function of nerve growth factor receptors on Schwann cells. TINS 11(7):299-304. J u , G., Lue, S., and J. Tao (1986) Projections from the hypothalamus and its adjacent areas to the posterior pituitary in the rat. Neuroscience 19(3):803-828. Kao, C.C., Chang, L.W., and J.M.B. Bloodworth Jr. (1977) Axonal regeneration across transected mammalian spinal cords: An electron microscopic study of delayed microsurgical nerve grafting. Exp. Neurol. 54:591-615. Katzman, R., Bjorklund, A., Owman, C.H., Stenevi, U., and K.A. West (1971) Evidence for regenerative axon sprouting of central catacholamine neurons in the rat mesencephalon following electrolytic lesions. Brain Res. 25:579-596. C 98 Kawamoto, K. (1985) Immunocytochemical study of vasopressin and oxytocin in the neurosecretory system during reorganization of the neural lobe in mice. Zool. Sci. 2:371-380. Kawamoto, K., and S. Kawashima (1985a) Plasticity of vasopressin- and oxytocin-containing fibers in the median eminence of hypophysectomized young and old mice. Brain Res. 330:189-193. Kawamoto, K., and S. Kawashima (1985b) Changes of vasopressin- and oxytocin-immunoreactive neurons after hypophysectomy in young and old mice. Exp. Gerontol. 20:295-304. Kawamoto, K., and S. Kawashima (1986) Effects of glucocorticoids and vasopressin on the regeneration of neurohypophysial hormone-containing axons after hypophysectomy. Zool. Sci. 3.:723-726. Kawamoto, K., and S. Kawashima (1987) Regeneration of neurohypophysial hormone-producing neurons in hypophysectomized immature rats. Brain Res. 422:106-117. Kelly, J., and L.W. Swanson (1980) Additional forebrain regions projecting to the posterior pituitary: Preoptic region, bed nucleus of the stria terminalis, and zona incerta. Brain Res. 197:1-9. Kennedy, W., Navarro, X., and H. Kamel (1988) Reinnervation of sweat glands in the mouse: Axonal regeneration versus collateral sprouting. Muscle and Nerve 11:603-609. Kiernan, J.A. (1971) Pituicytes and the regenerative properties of neurosecretory and other axons in the rat. J . Anat. 109(I):97-114. Kiss, J. J. McGovern and A.J. Patel (1988) Immunohistochemical localization of cells containing nerve growth factor receptors in teh different regions of the adult rat forebrain. Neuroscience 27.(3):731748. Knighton, D.R., I. A. Silver and T. K. Hunt (1981) Regulation of wound healing angiogenesis-Effeet of oxygen gradients and inspired oxygen concentration. Surgery 90j_2J_:262-270. Knighton, D.R., T. K. Hunt, H. Schwenstuhl and B. J. Halliday (1983) Oxygen tension regulates the expression of angiogenesis factor by macrophages. Science 221:1283-1284. Korr, H., B. Schultz and W. Maurer (1979) Autoradiographic investigations of glial proliferation in the brain of adult mice. II. Cycle time and mode of proliferation of neuroglia and endothelial cells. J. Comp. Neurol. 160:477-490. Korshing, S., R. Heumann, H. Thoenen, and F . Hefti (1986) Cholinergic denervation of the rat hippocampus by fimbrial transection leads to a transient accumulation of nerve growth factor (NGF) without change in mRNA NGF content. Neurosci. Lett. ,6:175-180. Kreutzberg, G.W., and K.D. Barron (1978) 5 '-Nucleotidase of microglial cells in the facial nucleus during axonal reaction. J. of Neurocytol. 7:601-610. Kromer, L. (1987) Nerve growth factor treatment after brain injury prevents neuronal death. Science 235:352-355. 99 Krsulovic, J. and G. Bruckner (1969) Morphological characteristics of pituicytes in different functional stages. Light and electron microscopy of the neurohypophysis of the albino rat. Z . Zellforsch. 99:210-220. Kurosumi, K., Matsuzawa, T., and S. Shibasaki (1961) Electron microscope studies on the fine structures of the pars nervosa and pars intermedia, and their morphological interrelation in the normal rat hypophysis. Gen. Comp. Endo. 1:433-452. Larkfors, L.I., T. Stromberg, T. Ebendal, and L. Olsen (1987) Nerve growth factor protein level increases in the adult rat hippocampus after a specific cholinergic lesion. J. Neurosci. Res. 18:525-531. Latov, N., Nilaver, G., Zimmerman, E.A., Johnson, W.G., Silverman, A.J., Defendini, R. and L. Cote (1979) Fibrillary astrocytes proliferate in response to brain injury. A study combining immunoperoxidase technique, for glial fibrillary acidic protein and radioautography of tritiated thymidine. Dev. Biol. 72:381-384. Lee, S.C., Moore, G.R., Golenwsky, G. and C.S. Raine (1990) Multiple sclerosis: A role for astroglia in active demyelination suggested by class II MHC expression and ultrastructural study. J. Neuropathol. Exp. Neurol. 49:122-136. Ling,E.A. (1981) The origin and nature of microglia. In Advances in cell neurobiology, vol. 2, S. Fedoroff and L. Hertz, eds., pp. 33-82, Academic press, NY. Loeshe, J., and O. Steward (1977) Behavioral correlates of denervation and reinnervation of the hippdcample formation of the rat: Recovery of alternation performance following unilateral entorhinal cortex lesions. Brain Res. Bull. 2.:31-39. Lu B., M. Yokooyama, C.F. Dreyfus and I.B. Black (1991) NGF gene expression in actively growing brain glia. J. of Neurosci. 11(2):318386. Madri, J. A. and S. K. Williams (1983) Effects of monocular closure at different ages on deprived and undeprived cells in the primate lateral geniculate nucleus. Dev. Brain Res. 18:57-78. J. Cell Biol. 97:153165. Mallet, S and A.N. Barkely (1991) A new superfamily of cell surface proteins related to the nerve growth factor receptor. Immunology Today 12:412-416. Mandybur, T.I. and C.C. Chuirazzi (1990) Astrocytes and the plaques of Alzheimer's disease. Neurology, 40:635-639. Martinez, H.J., C.F. Dreyfus, G.M. Jonakait and I.B. Black (1985) Nerve growth factor promotes cholinergic development in brain striatal cultures. Proc. Natl. Acad. Sci. USA 82:7777-7781. Matsuoka, I., M. Meyer and Hans Thoenen (1991) Cell-type-specific regulation of nerve growth facto (NGF) synthesis in non-neuronal cells: Comparison of Schwann cells with other cell types. J. of Neurosci. 11(10):3165-3167. McCouch, G.P., Austin, G.M., Liu, C.N. and C.Y. Liu (1985) Sprouting as a cause of spasticity. J. Neurophysiol. 21:205-216. 100 Milner, T.A. and R. Loy (1980) A delayed sprouting response to partial hippocampal deafferentation: Time course of sympathetic ingrowth following fimbrial lesions. Brain Res. 197:391-399. Mobely, W.C., J.L. Rutkowski, G.I. Tennekoon, K. Buchananand M.V. Johnston (1985) Choline acetytransferase activity in striatum of neonatal rats increased by nerve growth factor. Science 229:284-287. Moll, J. (1957) Regeneration of the supraoptico-hypophysial and paraventriculo-hypophysial tracts in the hypophysectomized rat. Zellforschung. 46:686-709. Z. Moll, J., and D . De Weid (1962) Observations on the hypothalamoposthypophysial system of the posterior lobectomized rat. Gen. Comp. Endo. 2:215-228. Monroe, B. (1967) A comparative study of the ultrastructure of the median eminence, infundibular stem and neural lobe of the hypophysis of the rat. Z . Zellforschung. 76:405-432. Montagnese C., D.A. Poulain, J-D. Vincent and D.T. Theodosis (1988) Synaptic and neuronal-glial plasticity in the adult oxytocinergic system in response to physiological stimuli. Brain Res. Bull. 20:681-692. Morris, J.F. (1976a) Hormone storage in individual neurosecretory granules of the pituitary gland: A quantitative ultrastructural approach to hormone storage in the neural lobe. J. Endo. 68:209-224. Morris, J.F. (1976b) Distribution of neurosecretory granules among the anatomical compartments of the neurosecretory processes of the pituitary gland: A quantitative ultrastructural approach to hormone storage in the neural lobe. J. Endo. 68:225-234. Morris, J.F. and R.E.J. Dyball (1974) A quantitative study of the ultrastructural changes in the hypothalamo-neurohypophyseal system during and after experimentally induced hypersecretion. Cell Tissue Res. 149: 525-535. Murray, J.G., and J.W. Thompson (1957) The occurence and function of collateral sprouting in the sympathetic nervous system of the cat. J . Physiol. 135:133-162. Murray, M. (1968) Effects of dehydration on the rate of proliferation of hypothalamic neuroglia cells. Exp. Neurol. 20:460-468. Murray, M. (1986) Reactive synaptogenesis in the CNS. A comparison of regenerating and sprouting systems. Neurochemical Path. 5:205-220 Murugiyan, P, P.S. Klinkhachorm and A.K. Salm (1992) Dehydration-induced proliferation of supraoptic nucleus (SON) and posterior pituitary astrocytes revisited. Soc. Neurosci. Abstr. 37.11 Olivecrona, H. (1957) Paraventricular nucleus and pituitary gland. Acta Physiol, scand. 40 suppl. 136:1-185. Nathaniel, E.J.H. and D.R. Nathaniel (1981) The reactive astrocyte. Adv. Cell Neurobiol. 2:249-301. Oliviera-Sangiacomo, C. (1972) On the fine structure of the perivascular cells in the neural lobe of rats. Z. Zellforshung. 132:2534. 101 Needles, D.L., Nietro-sampedro, M., and C.W. Cotman (1986) Induction of a neurite-promoting factor in rat brain following injury or deafferentation. Neuroscience 18(3):517-526. Nordmann, J.J. (1977) Ultrastructural morphometry of the rat neurohypophysis. J. Anat. 123(1):213-218. Nygren, L-G., Fuxe, K., Jonsson, G., and L. Olsen (1974) Functional regeneration of 5-hydroxytryptamine nerve terminals in the rat spinal cord following 5,6-dihydroxytryptamine-induced degeneration. Brain R e s . 78:377-394. Olivecrona, H. (1957) Paraventricular nucleus and pituitary gland. Acta. Physiol, scand. 40 suppl. 136:1-185. Paterson, J.A. and C.P. Leblond (1977) Increased proliferation of neuroglia and endothelial cells in the supraoptic nucleus and hypophysial neural lobe of young rats drinking hypertonic sodium chloride solution. J. Comp. Neurol. 175:373-390. Paxinos, G. and C. Watson (1986) coordinates. Acad. Press. "NY. The rat brain in stereotaxic Pearson, R.C.A., Sofroniew, M.V., and T.P.S. Powell (1984) Hypertrophy of immunocytochemically identified cholinergic neurons of the basal nucleus of meynert following ablation of the contralateral cortex in the rat. Brain Res. 311:194-198. Pearson, R.C.A., Sofroniew, M.V., and T.P.S. Powell (1985) Hypertrophy of cholinergic neurons of the rat basal nucleus following section of the corpus callosum. Brain Res. 338:337-340. Pearson, R.C.A., Neal J.W., and T.P.S. Powell (1987a) Bilateral morphological changes in the substantia nigra of the rat following unilateral damage of the striatum. Brain Res. 400:127-132. Pearson, R.C.A., Neal, J.W., and T.P.S. Powell (1987b) increase in immunocytochemical staining of GABAergic axons in the superior colliculus and thalamus of the rat following damage to the ipsilateral striatum and frontal cortex. Brain Res. 412:352-356. Perez-Figaes, J.M., Fernandez-Llebrez, P., Perez, J. and F . Marin-Giron (1986) Presence of microglia-like cells in the neurohypophysis of normal and dehydrated rats. Acta Anat. 127:195-200. Perry, V.H. and S. Gordon (1987) Modulation of CD4 antigen on macrophages and microglia in rat brain. J. Exp. Med.166:1138-1143. Perry, V.H., M.C. Brown and S. Gordon (1985) The macrophage response to central and peripheral nerve injury. A possible role for macrophages in regeneration. J. Exp. Med. 165:1218-1223. Peterson, R.P. (1966) Magnocellular neurosecretory centers in the rat hypothalamus. J. Comp. Neurol. 128:181-190. Pioro, E.P. and A.C. Cuello (1988) Purkinje cells of adult rat cerebellum express nerve growth factor receptor immunoreactivty: Light microscopic observations. Brain Res. 455:182-186. Pittman, Q.J., Blume, H.W., and L.P. Renaud (1978) Electrophysiological indications that individual hypothalamic neurons innervate, both median eminence and neurohypophysis. Brain Res. 157:364-368. 102 X Polenov7 A.L., Ugrumov7 M.V . , Propp7 M.V., and M.A. Belenky (1974) The hypothalamo-hypophysial system of hypophysectomized rats. I. Ultrastructure of nerve fibers in "intact" and dehydrated animals. Cell Tissue Res. 155:541-554. Polenov7 A.L., Belenky, M.A., and Bogdanovic-Stosic. (1981) The hypothalamo-hypophysial system in hypophysectomized rats II. Structure and ultrastructure of the median eminece. Cell Tissue Res. 218:607622. Polverini7 P. J., R. C. Cotran7 M. A. Gimbrone7 Jr. and E. R. Unanue (1977) Activated macrophages induce vascular proliferation. Nature 269:804-806. Raisman7 G. (1969) Neuronal plasticity in the septal nuclei of the adult rat. Brain Res. 14:25-48. Raisman7 G. (1973) Electron microscopic studies of the development of new neurohaemal contacts in the median eminence of the rat after hypophysectomy. Brain Res. 55:245-261. Raisman7 G. (1977) An ultrastructural study of the effects of hypophysectomy on the supraoptic nucleus of the rat. J. Comp. Neurol. . 147:181-208. Rechardt7 L., and H. Hervonen (1982) Ultrastructural changes in the neurohypophysis of the aged male rat. Cell Tissue Res. 226:51-62. Reinhardt7 H.F., L.CH. Henning7 and H.P. Rohr (1969) MorphometrischUltrastrukturelle untersuchungen am nucleus supraopticus der ratte nach dehydration. Z.Zellforchung 102:172-181. Rhodes7 C.W., Morrell, J.I., and D.W. Pfaff (1981) Immunohistochemical analysis of magnocellular elements in rat hypothalamus: Distribution and numbers of cells containing neurophysin, oxytocin, and vasopressin. J. Comp. Neurol. 198:45-64. Richardson7 P.M., V.M.K. Verge Issa and R.J. Riopelle (1986) Distribution of neuronal receptorsfor nerve growth factor in the rat. J. of Neurosci. 6(8):2312-2321. Rio-Hotega7 P. (1932) Microglia. In Cytology and cellular pathology of the nervous system, vol. 2, W. Penfield ed. pp. 481-584, P.B. Hober7 NY. Rodriguez7 E.M., and H-D Dellman (1970) Ultrastructure and hormonal content of the proximal stump of the transected hypothalamo-hypophysial tract in the frog (rana pipians). Z . Zellforschung. 104:449-470. Rose7 G 7 G. Lynch and C.W. Cotman (1976) Hypertrophy and redistribution of astrocytes in the deafferented dentate gyrus. Brain Res. Bull. 1:8792. Rothballer7 A.B. and S.C. Skaryna (1960) Morphological effects of pituitary stalk section in the dog withparticular reference to neurosecretory material. Anat. Rec. 136:5-25. Saavedra7 J.M. (1985) Central and peripheral catecholamine innervation of the rat intermediate and posterior pituitary lobes. Neuroendocrinology 40:281-284. 103 Salaand7 L.C., J.A. Wallace, E. Reyes, A. Samora, D. Maez and F. Cumunas (1987) Effects of the serotonin-uptake inhibitor, fluoxetine, on immunoreactive serotonin innervation in the rat. pituitary gland. Brain Res. Bull. 19:261-267. Salm, A.K., 6.1. Hatton and G. Nilaver (1982) Immunoreactive glial fibrillary acidic protein in pituicytes of the rat neurohypophysis. Brain Res. 236:471-476. Schneider, G.E., and S.R. Jhavari (1974) Neuroanatomical correlates of spared or altered function after brain lesions in the newborn hamster, in Plasticity and Recovery of Function in the Central Nervous System. (eds., D.G.Stein, J.J. Rosen, and N . Butters), Academic Press, NY pp. 65-110. Scharrer, E.A. and G.J. Wittenstein (1952) The effect of the interuption of the hypothalamo-hypophysial neurosecretory pathway in the dog. Anat. Rec. 112:387-390. Silverman, A.J., Hoffman, D.L., and E.A. Zimmerman (1980) The descending afferent connections of the parvocellular nucleus of the hypothalamus (PVN). Brain Res. Bull. 6:47-61. Silverman, A.J., and E.A. Zimmerman (1982) Adrenalectomy increases sprouting in a peptidergic neurosecretory system. Neuroscience 7(11):2705-2714. Silverman, A.J. (1983) Magnocellular neurosecretory system. Neurosci. 6:357-380. Ann. Rev. Sofroniew, M.V. (1980) Projections from vasopressin, oxytocin, and neurophysin neurons to neural targets in the rat and human. J . Histochem. and Cytochem. 28(5):475-478. Sofroniew, M.V., O. Isacson and T.S. O'Brien (1989) Nerve growth factor receptor in the rat suprachiasmatic nucleus. Brain Res. 476:358-362. Springer, J.E. (1988) Nerve growth factor receptors in the central nervous system. Exp. Neurol. 102:354-365. Stenevi, U., Bjorklund, A., and R.Y. Moore (1972) Growth of intact central adrenergic axons in the denervated lateral geniculate body. Exp. Neurol. 35:290-299. Steward, O., Cotman, C., and G. Lynch (1976) A quantitative autoradiographic and electrophysiological study of the reinnervation of the dentate gyrus by the contralateral entorhinal cortex following ipsilateral entorhinal lesions. Brain Res. 114:181-200. Streit, W.J. and G.W. Kreutzberg (1987) Lectin binding by resting and reactive microglia. J. of Neurocytol. 16:249-260. Streit, W.J., M.B. Graeber and G.W. Kreutzberg (1989) Expression of Ia antigen perivascular and microglial cells after sublethal and lethal motor neuron injury. Exp. Neurol. 105:115-126. Suess, U. and V. Pliska (1981) Identification of the pituicytes as astroglial cells by indirect immunofluorescence-staining for the glial fibrillary acidic protein. Brain Res. 221:27- 104 Swanson, L.W., and H.G.J.M. Kuypers (1980) The paraventricular nucleus of the hypothalamus: Cytoarchetectonic subdivisions and the organization of projections to the pituitary, dorsal vagal complex and spinal cord as demonstrated by retrograde fluorescence double-labeling methods. J. Comp. Neurol. 194:555-570. Swanson, L.W., and P.E. Sawchenko (1983) Hypothalamic integration: Organization of the paraventricular and supraoptic nuclei. Ann. Rev. Neurosci. 6:269-324. Takano, S., H. Negoro, K. Honda and T. Higuchi (1990) Electrophysiological evidence for neural connections between the supraoptic nuclei. Neuroscience Letters 111:122-126. Taniuchi, M., J.B. Schweitezer and E.M. Johnson Jr. (1986) Nerve growth factor receptor molecules in rat brain. Proc. Natl. Acad. Sci. USA 83:1950-1954. Taniuchi, M., H.B. Clark and E.M. Johnson Jr. (1986a) Induction of nerve growth factor receptor in Schwaan cells after axotomy. Proc. Natl. Acad. Sci. 83:4094-4098. Theodosis, D.T., and D.A. Poulain (1987) Oxytocin-secreting neurones physiological model for structural plasticity in the adult mammalian brain. TINS 10(10):426-430. A Tsukahara, N., and Y. Fujito (1981) Neuronal plasticity in the newborn and adult feline red nucleus. in Lesion-induced neuronal elasticity in sensory motor systems. (H. Flohr and W. Precht) pg 65-74. Tuskahara, N. (1985) Synaptic plasticity in the red nucleus and its possible behavior correlates. In Synaptic Plasticity. C.W. Cotman, ed., p p . 201-229, The Guilford Press, New York. Tweedle C.D. and G.I. Hatton (1980) Glial cell enclosure of neurosecretory endings in the neurohypophysis of the rat. Brain Res. 192:555-559. Vandesande, F., Dierickx, K., and G. DeMay (1977) The origin of the vasopressinergic and oxytocinergic fibers of the external region of the median emihece of the rat hypophysis. Cell Tissue Res. 180:443-452. Velasko, M.E., Roessman, U. and P. Gambetti (1979) Immunohistochemical demonstration of glial fibrillary acidic (GFA) protein in human pituitary gland. J. Neuropath, exp. Neurol. 38: Vijayan, V., J.W. Gaddes, K.J. Anderson, H. Chang-Chui, W.G. Ellis and C.W. Cotman (1991)Astrocyte hypertrophy in the Alzheimers Disease Hippocampal Formation. Exp. Neurol. 112:72-79. Vogels, O.J.M., C.A.J. Broere and R. Niewenhuys (1990) Neuronal hypertrophy in the human supraoptic and paraventricular nucleus in aging and alzheimers disease. Neurosci. Letters, 109:62-67. Watson, W.E. (1965) An autoradiographic study of the incorporation of nucleic-acid precursors by neurones and glia during nerve regeneration. J. Physiol. Lond. 180:741-753. Watt, J.A. and C.M. Paden (1991) Unilateral Hypothalamic Lesion Results in Sprouting of Contralateral Magnocellular Neurosecretory Axons in the Rat Neural Lobe. Exp. Neurol. 111:9-24. 105 Weskamp, G. and L.F. Reichardt (1991) Evidence that biological activity of NGF is mediated through a novel subclass of high affinity receptors. Neuron 6:649-663. Wiegand, S.J. and D.M. Gash (1988) Characteristics of vasculature and neurovascular relations in intraventricular anterior hypothalamic transplants. Brain Res. Bull. 20:105-124. Williams, L.R., S. Varon, G.M. Peterson, K. Wicktorin, W. Fisher, A. Bjorklund and F.H. Gage (1986) Continuous infusion of nerve growth factor prevents basal forebrain neuronal death after fimbria fornix transection. Proc. Natl. Acad. Sci. USA 83:9231-9235. Windle, W.F., and W.W. Chambers (1950) Regeneration in the spinal cord of the cat and dog. J. Comp. Neurol. 93,:241-257. i Wittkowski, W. and H. Brinkman (1974) Changes of extent of neurovascular contacts and neuro-glial synaptoid contacts in the pituitary posterior lobe of dehydrated rats. Anat. Embryol. 146:157-165. Wuttke, W., and A. Bjorklund (1977) De- and regeneration of brain serotonin neurons following 5,7-DHT treatment: Effects of serum L H , FSH, and prolactin levels in male rats. Brain Res. 134:317-331. Yamanaka, C, T. Momoi, I. Fujisawaand K. Kikuchi (1990) Neurohypophysial function of an ectopic posterior lobe in patients with growth hormone deficiency. Acta. Endocrinol. 122,(5):664-670. Yan, Q., H.B. Clark and E.M. Johnson, Jr. (1990) Nerve growth factor receptor in the neural lobe of rat pituitary gland: immunohistochemical localization, biochemical characterization and regulation. Yanagisawa-Miwa, A., Y. Uchida, F. Nakamura, T. Tomaru, H. Kido, T. Kamijo, T. Sugimoto, K. Kaji, M. Utsuyama, C. Kurashima and H. Ito (1992) Salvage of infarcted myocardium by angiogenic action of basic fibroblast growth factor. Science 257:1401-1403. Yoshida, Y., F. Ikata, and M. Yamada (1986) Proliferation of reactive microvessels in experimental infarction: Chronological and three dimensional observation. IOlh International Congress of Neuropathology, Stocolm, A416. Yoshida, K. and F.H. Gage (1991) Fibroblast growth factors stimulate nerve growht factor synthesis and secretion by astrocytes. Brain Res. 538:118-126. Young, W. C. and I . M. Herman (1985) Increased proliferation of neuroglia and endothelial cells in the supraoptic nucleus and hypophysial neural lobe of young rats drinking hypertonic sodium chloride solution. J. Comp. Neurol. 175:373-390. Yulis, C.R., and E.M. Rodriguez (1982) Neurophysin pathways in the normal and hypophysectomized rat. Cell Tissue Res. 227:93-112. Vogels, O.J.M., C.A.J. Broere and R. Nieuwenhuys (1990) Neuronal hypertrophy in the human supraoptic and paraventricular nucleus in aging and alzheimer's disease. Neuroscience Letters 109:62-67. Zambrano D., and E . de Robertis (1968) The ultrastructural changes in the neurohypophysis after destruction of the paraventricular nuclei in normal and castrated rats. Z. Zellforschung. 88:496-510.