Increased hypothalamic glutamate receptors induced by water deprivation.

To examine the role of receptor changes in the adaptive response to physiological stimulation, the density and distribution of excitatory amino acid receptors within the hypothalamus and other brain regions were examined in rats deprived of water for 2 days. Membrane binding assay revealed an increase in glutamate receptor density and a small shift in the affinity of glutamate for the receptor. Regional analysis of these changes by receptor autoradiography specific for NMDA, non-NMDA or metabotropic glutamate receptor binding indicated that NMDA and metabotropic receptor densities are increased in the brain. Regional increases were found principally for the NMDA receptor binding within the supraoptic nucleus, anterior hypothalamus, caudate-putamen and globus pallidus with no significant changes in 24 other brain regions. No significant changes were found in any brain regions for AMPA receptors. Metabotropic and kainate receptors tended to parallel the NMDA receptor changes, although few regions reached statistical significance. These changes indicate that brain regions associated with water balance regulation show selective adaptive increases in NMDA receptors during water deprivation which may facilitate prolonged activation of these cells.

Vasopressin mRNA changes during kindling: the effects of kindling site and stage.

Because of the many anatomical and functional links to the limbic system, the neuroendocrine system is often affected by limbic disturbances. Limbic seizures in humans and animals alter neuroendocrine function and hormone levels. We have shown that in an animal model for partial seizures, the amygdala kindled rat, plasma vasopressin levels are elevated and a sustained increase in vasopressin (VP) mRNA follows stage 5 kindled seizures. In the present experiments we sought to determine when during the course of amygdala kindling the VP mRNA increase occurs and whether specific anatomical pathways mediate this increase. Animals kindled to early seizure stages (stages 1, 2 or 3) had no consistent increase in VP mRNA in the supraoptic nucleus (SON) while animals kindled to generalized seizures, stages 4 or 5, invariably had increased VP mRNA relative to controls. Electrical kindling to stage 5 seizures from two other brain sites, the dorsal hippocampus and the anterior olfactory nucleus, consistently resulted in a significant increase in VP mRNA one week after completing kindling. In all experiments the increase in VP mRNA in the SON showed no differences related to the side or proximity of the electrodes used for kindling. Measures of water balance did not change following kindling. These results indicate that kindled seizure generalization is a prerequisite for the long-term increase in VP mRNA. Furthermore, the VP mRNA increase appears to involve polysynaptic pathways accessible from different limbic kindling sites. These studies support the hypothesis that changes in mRNA regulation may contribute to the neuroendocrine pathophysiology accompanying limbic seizures.

Kindled seizures induce a long-term increase in vasopressin mRNA.

Neuroendocrine disturbances are among the significant problems associated with animal and human seizures. To investigate the mechanisms for these disturbances, we examined changes in the expression of vasopressin (VP) mRNA in the hypothalamic magnocellular neuroendocrine cells of rats after amygdala kindled seizures, a model for temporal lobe epilepsy. A prominent increase in VP mRNA was found in the supraoptic nucleus of kindled animals by one week after the last seizure which persisted for at least 4 months. The increase occurred bilaterally in the SON and remained unchanged despite the absence of further stimulation, seizures or change in body fluid homeostasis. Since the VP mRNA change after kindling correlated with the duration of afterdischarge but not the number of amygdala stimuli the change appears to be an effect of the seizure. This chronic increase in VP mRNA appears to reflect a change in neuroendocrine gene expression and may identify an important new mechanism of plasticity that contributes to the neuroendocrine disturbances accompanying epilepsy.

Glutamate receptors in the rat hypothalamus and pituitary.

Although several recent anatomical and physiological studies indicate that glutamate receptors are likely to play a role in the regulation of various hypothalamic functions, no attempt has yet been made to specifically characterize glutamate receptor densities, subtypes, or localization in the hypothalamus. To provide this basic information, we have characterized and mapped the binding of [3H]glutamate to N-methyl-D-aspartate (NMDA), non-NMDA, and metabotropic glutamate receptors throughout the diencephalon. Membrane binding assays revealed a [3H]glutamate binding density of 2.6 pmol/mg protein, approximately one third of the hippocampal density. Binding of subtype-specific agonists and antagonists was complex, but clearly indicated that each major glutamate subtype is present in all hypothalamic and preoptic regions in the following approximate relative densities: NMDA > metabotropic Glu receptor > kainate > or = alpha-amino-3-hydroxy-5- methyl-4-isoxazolepropionic acid. Receptor autoradiography confirmed the widespread presence of all major glutamate receptor subtypes with roughly the following relative regional densities: ventromedial, dorsomedial > paraventricular, anterior hypothalamic, supraoptic > arcuate, suprachiasmatic, lateral hypothalamic > preoptic area >> pituitary neural lobe, white matter > pituitary anterior lobe (negligible). Subtype expression varied regionally, with rostral hypothalamic and preoptic regions having proportionally higher levels of non-NMDA vs. NMDA binding. High densities of glutamate receptors in ventromedial and medial hypothalamic regions suggest a prominent role in neuroendocrine and autonomic regulation.

Quantitative mapping of glutamate presynaptic terminals in the supraoptic nucleus and surrounding hypothalamus.

Although the hypothalamus is generally regarded to have low levels of glutamate receptors, anatomical and physiological studies have provided consistent evidence implicating glutamate as a potential transmitter for the control of neuroendocrine cell activity. To clarify the extent of the contribution of synapses utilizing glutamate for control of vasopressin/oxytocin neuroendocrine cells, we mapped the density and location of glutamate immunoreactive terminals in the supraoptic nucleus and surrounding hypothalamus. Colloidal gold particle densities in presynaptic terminals were measured from electron micrographs of: (1) the magnocellular neuroendocrine cell perikarya (main body of the supraoptic nucleus), (2) the dendritic field of the magnocellular neuroendocrine cells (ventral dendritic neuropil) and (3) the hypothalamic perinuclear zone dorsal to the supraoptic nucleus. In addition, serial sections were stained, alternatively, for glutamate or GABA to determine glutamate staining in GABA cells. Terminals with high glutamate immunoreactivity were clearly distinguished from the glutamate precursor staining found in GABA terminals and were abundant at all rostral-caudal levels within each region. The number of glutamate terminals identified in each region was similar but represented a very high proportion of all terminals in the ventral dendritic neuropil (38%) vs. the main body of the supraoptic nucleus and the perinuclear zone (20-22%). The regional variation in the relative proportion of glutamate terminals was determined largely by differences in the number of non-glutamate terminals within each region. Glutamate and GABA terminals together accounted for over two-thirds of the innervation of vasopressin/oxytocin neuroendocrine cells. No systematic relationship was observed between excitatory and inhibitory inputs on the same cell. These results suggest that glutamate is the predominant excitatory transmitter used for control of vasopressin/oxytocin cells. The relative contribution of glutamate neurotransmission to a particular region will depend, in part, on the number and type of competing non-glutamate terminals.

Ultrastructural distribution of glutamate immunoreactivity within neurosecretory endings and pituicytes of the rat neurohypophysis.

An ultrastructural analysis of post-embedding glutamate immunocytochemistry within the neural lobe of the pituitary was used to explore the possible role of glutamate within the magnocellular neuroendocrine cells. Relative densities of a colloidal gold marker associated with various cellular and subcellular compartments of the neural lobe were quantified by computer analysis of electron micrographs. Robust glutamate immunoreactivity was observed in both pituicytes (cytoplasm, mitochondria and nucleus) and neurosecretory endings. Within the neurosecretory endings, glutamate staining was specifically localized to the microvesicles with no overlap into the neurosecretory granule population. Stimulation of the vasopressin/oxytocin neurosecretory system by water deprivation increased glutamate content in pituicytes and mitochondria within neurosecretory endings but had little influence on microvesicle glutamate content. The results are consistent with the existence of multiple functional pools of immunoreactive glutamate in both pituicytes and neurosecretory endings. Microvesicles within the neurosecretory endings exhibit many properties of secretory vesicles, appear to be functionally independent of the neurosecretory granules, and have sufficient glutamate immunoreactivity to suggest that this amino acid may be compartmentalized for release in the neural lobe.

Muscarinic cholinergic control of vasopressin secretion from the acute hypothalamoneurohypophysial explant.

Much of the afferent input thought to modulate vasopressin release from the magnocellular neuroendocrine cells of the supraoptic nucleus terminates in the region dorsal to the supraoptic nucleus. Cholinergic cells within this region may participate in the local processing of these afferent signals via synapses onto muscarinic cholinergic receptors. To investigate the role of these local synapses in vasopressin secretion, we characterized the muscarinic cholinergic influence on vasopressin secretion from the acute hypothalamoneurohypophysial explant in vitro. Acetylcholine induced a small dose-related secretion of vasopressin which could be totally blocked by atropine but not the nicotinic cholinergic antagonist, hexamethonium. Nicotine failed to release vasopressin from the explant, whereas alpha-bungarotoxin elicited a hypothalamic release of vasopressin which was atropine insensitive. Thus, local muscarinic receptors in the hypothalamus appear to participate in the control of neurohypophysial vasopressin secretion. The small magnitude of effect, however, is consistent with an indirect modulatory role rather than a major driving force for activation of the magnocellular neurons.

Vasopressin mRNA expression in individual magnocellular neuroendocrine cells of the supraoptic and paraventricular nucleus in response to water deprivation.

Vasopressin neuroendocrine function involves the regulation of both secretion and synthesis from magnocellular neuroendocrine cells but the coordination of these two processes is poorly understood. To explore the temporal relationship between physiological stimulation and vasopressin mRNA levels we measured vasopressin mRNA content within individual magnocellular neurons of the supraoptic and paraventricular nucleus during the course of water deprivation. Analysis of autoradiographic silver grain densities from in situ hybridization of an [125I]dCTP-labeled oligonucleotide specific for vasopressin mRNA revealed a wide variety of resting vasopressin mRNA levels and differential responses to water deprivation in the magnocellular neuroendocrine cells. During water deprivation, the vasopressin mRNA content of the paraventricular nucleus increases rapidly and with shorter latency and greater incremental response than the supraoptic nucleus. Double-labeling experiments with combined in situ hybridization and immunocytochemistry identified a population of vasopressin immunoreactive cells which maintain very low basal levels of vasopressin mRNA. The location of these cells correlates with the location of increased silver grain densities during water deprivation. One subset of vasopressin magnocellular neurons failed to show high levels of vasopressin mRNA, indicating that all cells are not equally responsive to water deprivation. These patterns of vasopressin mRNA expression suggest the presence of functional subpopulations of vasopressin neuroendocrine cells which may reflect stimulus-specific patterns of afferent input to the supraoptic and paraventricular nucleus.

Amygdala kindling elevates plasma vasopressin.

Acute and chronic effects of epilepsy on endocrine function are known to occur in humans with partial seizures of limbic origin and in animals with limbic kindled seizures. The amygdala, a component of the limbic system, has dense hypothalamic connections and amygdala stimulation in monkeys and cats result in vasopressin release. In the present study we sought to determine if amygdala stimulation in the rats results in an immediate acute release of vasopressin and to determine if acute or chronic changes occur in vasopressin release in the fully kindled animal. Plasma vasopressin, osmolality and hematocrit were measured in blood samples drawn from rats with implanted venous catheters before and after stimulation and at different stages of kindling. Low-frequency (15 Hz) electrical stimulation of the amygdala was followed by an immediate, 3-fold increase in plasma vasopressin concentration. Moreover, although the 60 Hz kindling stimulus did not result in a significant immediate rise in plasma vasopressin prior to kindling, after kindling to stage 5 seizures the 60 Hz kindling stimulus resulted in seizures and a significant immediate rise in plasma vasopressin. In addition, we found that kindling was followed by a significant, though modest, rise in the resting plasma vasopressin without an accompanying change in osmolality or hematocrit. We conclude that kindling results in a persistent alteration in the vasopressinergic neuroendocrine system.

Vasopressin and oxytocin regulation of cyclic AMP accumulation in rat hypothalamo-neurohypophysial explants in vitro.

Addition of vasopressin to hypothalamo-neurohypophysial explants in vitro increased cyclic AMP accumulation whereas exogenous oxytocin decreased cyclic AMP. An opposite response pattern was observed in the neural lobe of the pituitary where vasopressin decreased and oxytocin increased cyclic AMP accumulation. Forskolin elicited a 3-fold greater increase in cyclic AMP in the neural lobe than in the supraoptic nucleus and enhanced the sensitivity of the tissues to both vasopressin and oxytocin. The ability of both vasopressin and oxytocin to modulate local cyclic AMP metabolism suggests the possibility of internal feedback within the hypothalamo-neurohypophysial system.

Kindling in spontaneous hypertensive rats.

Vasopressin is a neurohormone and neuromodulator with many effects on behavior. Rats lacking vasopressin have been found to develop kindled seizures more slowly with amygdala stimulation. In the present study the spontaneous hypertensive (SH) rat and rats from the parent strain, the Wistar-Kyoto (WKY) rat received amygdala and pyriform kindling. The SH rat has been reported to have increased plasma vasopressin and increased brain vasopressin release. Plasma vasopressin, osmolality and hematocrit were also measured in blood samples obtained through chronic, indwelling catheters implanted in SH, WKY normal and Sprague-Dawley rats. SH rats were found to kindle with fewer afterdischarges than WKY normal rats with both amygdala and pyriform cortex stimulation. The total afterdischarge duration required to reach each kindling stage was significantly shorter in the SH rat. Plasma osmolality and vasopressin were significantly higher in the SH rats compared to WKY normal rats and Sprague-Dawley rats. These findings provide additional evidence that vasopressin may influence the establishment of enduring behaviors such as kindled seizures.

Metabolic mapping of functional activity in rat brain and pituitary after water deprivation.

Glucose utilization in rat brain and pituitary was measured in control and water-deprived rats by autoradiographic assessment of the metabolic trapping of radioactivity from [1-14C]glucose. Two days of water deprivation resulted in significant increases in hematocrit, plasma osmolality and vasopressin levels, indicating a functional activation of magnocellular vasopressin neurons. The uptake and retention of radioactivity from [1-14C]glucose in the dehydrated rats, compared to controls, was 103% greater in the magnocellular portion of the paraventricular nucleus and 74% greater in the supraoptic nucleus. Water deprivation also resulted in significant increases in glucose utilization (30-40%) in the lateral and anterior hypothalamic areas, somatosensory cortex and cingulate cortex. No change in glucose utilization after 2 days of water deprivation was apparent in the parvocellular paraventricular nucleus, periventricular nucleus of the hypothalamus, corpus callosum, organum vasculosum of the lamina terminalis (OVLT) or the subfornical organ (SFO). In the pituitary, glucose utilization was increased in the neural lobe but was unchanged in the anterior and intermediate lobes after water deprivation. Under the conditions of the present study, no increase in metabolic activity was apparent in 2 brain regions thought to be possible sources of osmoreception, the OVLT and SFO. These results do not support, but do not exclude, functional involvement of the OVLT and SFO in regulating the activity of magnocellular neurons of the paraventricular nucleus and supraoptic nucleus during chronic water deprivation.

Light and electron microscopic localization of glutamate immunoreactivity in the supraoptic nucleus of the rat hypothalamus.

The distribution of glutamate immunoreactivity was mapped within the supraoptic nucleus of the rat hypothalamus utilizing a specific anti-glutamate antibody. Magnocellular neuroendocrine cells of the supraoptic nucleus showed intense immunoreactivity for glutamate which varied with the conditions of fixation. Within the perikarya, reaction product was found associated with the endoplasmic reticulum but not the mitochondria, Golgi, dense bodies or neurosecretory granules. A relatively high density of glutamate-immunoreactive terminals was found in the supraoptic nucleus. These terminals were less affected by fixation condition and were generally found contacting large, glutamate-immunoreactive processes within the ventral dendritic neuropil of the supraoptic nucleus. The pattern and characteristics of glutamate immunoreactivity in the supraoptic nucleus suggested the presence of two distinct glutamate pools. The magnocellular neuroendocrine cells may contain a large, labile metabolic pool of glutamate. These cells, in turn, appear to receive glutamate synaptic input from a more stable pool consistent with suggestions that glutamate may be used as a transmitter within this system.

Local synaptic organization of cholinergic neurons in the basolateral hypothalamus.

A monoclonal antibody to choline acetyltransferase (ChAT) was utilized for immunocytochemical identification of cholinergic neurons in the basolateral hypothalamus. Light and electron microscopic examination revealed a network of cell bodies, dendrites, and axonal processes dorsolateral to the supraoptic nucleus. Within this region the cells immunoreactive for ChAT receive numerous unlabeled terminals which contact dendrites, cell soma, axons and occasional somatic spines. In a few cases, small ChAT-immunoreactive terminals were observed contacting a cholinergic cell soma or large dendrite. Many ChAT-immunoreactive fibers were directed toward the supraoptic nucleus forming a dense local network but very few of these fibers penetrated deeper than approximately 20 micron into the supraoptic nucleus. A total of 63 ChAT-immunoreactive terminals were mapped within the basal hypothalamus, of which the vast majority contacted unlabeled dendrites immediately dorsolateral to the supraoptic nucleus. Labeled terminals were rare or nonexistent in the medial portions of the hypothalamus or deep within the supraoptic nucleus. This pattern of ChAT terminal densities correlates with the distribution of binding for the muscarinic cholinergic probe, [3H]quinuclidinylbenzilate, but not the binding of the putative nicotinic cholinergic probe, [125I]alpha-bungarotoxin, which is high within the supraoptic nucleus. Thus, the cholinergic neurons of the basal hypothalamus appear to form a network of intrinsic connections which probably represent input to muscarinic cholinergic receptors. No evidence was found to suggest that cholinergic presynaptic terminals were colocalized with the alpha-bungarotoxin binding protein within the supraoptic nucleus.

Accumulation of circulating endogenous and exogenous immunoglobulins by hypothalamic magnocellular neurons.

Rat monoclonal antibodies, used in immunocytochemistry of normal rat brain, result in a granular reaction product within neurons innervating areas lacking a blood-brain barrier. Immunocytochemical characterization shows that the staining is independent of the primary antibody and exclusively dependent on the presence of anti-rat immunoglobulin. This granular staining could be selectively eliminated by pre-adsorption of the anti-rat immunoglobulin with purified rat immunoglobulin or disruption of microtubule retrograde transport systems by intraventricular injection of colchicine. A dependence on retrograde transport and complete independence from local synthesis was further substantiated by the rapid uptake and accumulation of intravenously administered rabbit or rat [125I]immunoglobulins by the supraoptic-neurohypophysial system. Immunoelectron microscopy was used to identify the endogenous rat immunoglobulin within lysosome-like organelles in the cytoplasm of magnocellular neuroendocrine cells. The uptake and incorporation of plasma macromolecules into the lysosomal system of magnocellular and other neurons projecting to regions with a weak blood-brain barrier may represent a novel mode of blood-central nervous system interactions.

Characteristics and distribution of high- and low-affinity alpha bungarotoxin binding sites in the rat hypothalamus.

When binding of 125I-alpha bungarotoxin (125I-alpha BTX) to hypothalamic membranes is observed over a wide range of concentrations, 3 binding sites can be identified, with estimated equilibrium dissociation constants (Kds) of 4.1 X 10(-11) M, 6.2 X 10(-10) M, and 9.1 X 10(-7) M for high-, low-, and very-low-affinity interactions, respectively. The densities of the high- and low-affinity sites were similar at 14-21 fmol/mg protein, whereas the very-low-affinity site had approximately 1000 X greater capacity. Association and dissociation kinetics predicted a biphasic binding reaction, with association rate constants of 1.38 X 10(8) M-1 min-1 and 7.53 X 10(7) M-1 min-1 and dissociation rate constants of 5.23 X 10(-3) min-1 and 1.80 X 10(-3) min-1. The presence of Na+ inhibited the binding of 125I-alpha BTX with a half-maximally effective concentration of 22 mM. This decrease in binding was associated with the observation of a single binding site with a Kd of 4.3 X 10(-10) M and a density of 12.1 fmol/mg protein. In competition binding experiments, alpha BTX, curare, nicotine, and quinacrine were the most potent competitors. Acetylcholine competed with 125I-alpha BTX binding at 2 sites with estimated affinities of 3.6 X 10(-8) and 7.4 X 10(-5) M. In the rostral hypothalamus, high-affinity binding of 125I-alpha BTX was localized to the region of the supraoptic nucleus, paraventricular nucleus, suprachiasmatic nucleus, and the nucleus circularis complex. Within magnocellular regions, binding was closely associated with neurophysin-immunoreactive neurons and processes, while in the region of the suprachiasmatic nucleus, the binding was in a perinuclear region surrounding parvocellular neurophysin-immunoreactive neurons.

Differential distribution of muscarinic cholinergic and putative nicotinic cholinergic receptors within the hypothalamo-neurohypophysial system of the rat.

Binding of the muscarinic cholinergic receptor probe [3H]quinuclidinylbenzilate ([3H]QNB) and the putative nicotinic receptor probe [125I]alpha-bungarotoxin ([125I]alpha BTX) to vasopressin (VP) and oxytocin (OT) neuroendocrine cells was investigated with a combination of quantitative receptor binding, autoradiography and immunocytochemistry. A single high-affinity site was labelled by [3H]QNB in the hypothalamus and pituitary (KD = 0.76-1.44 X 10(-10) M) with a mean hypothalamic density of 213 fmol/mg protein compared with only 56 fmol/mg protein in the pituitary. Analysis of autoradiographic silver grains from [3H]QNB binding revealed a relative absence of binding associated with magnocellular VP and OT cell groups in the hypothalamus. The median eminence and neural lobe of the pituitary contained low levels of [3H] QNB binding, which, however, were the highest within the hypothalamo-neurohypophysial system. The ligand [125I]alpha BTX binds with both a high and low affinity to sites within the hypothalamus and pituitary (high-affinity KD = 0.77-1.03 X 10(-10) M). In the hypothalamus the density of high-affinity binding sites (25 fmol/mg protein) is approximately 2.5 times greater than in the pituitary. In contrast to [3H]QNB, high-affinity binding of [125I]alpha BTX was found to be highly concentrated within the supraoptic nucleus, nucleus circularis, and the magnocellular areas of the paraventricular nucleus. Autoradiographic silver grains were distributed over both VP and OT immunoreactive neurons and processes. Binding within the neural lobe was very low. These data suggest that the cholinergic regulation of VP and OT release may occur via nicotinic cholinergic receptors at the level of the magnocellular cell bodies and predominantly via muscarinic cholinergic receptors within the neural lobe.

Supraoptic neurons in the rat hypothalamo-neurohypophysial explant: double-labeling with Lucifer Yellow injection and immunocytochemical identification of vasopressin- and neurophysin-containing neuroendocrine cells.

By combining intracellular electrophysiology with double-labeled intracellular dye-marking and immunocytochemical identification of the same magnocellular neuroendocrine cell, we studied supraoptic neurons in the rat hypothalamo-neurohypophysial explant in vitro. This report examines neurophysiological and light microscopical features of vasopressin- and neurophysin-immunoreactive, pituitary-projecting supraoptic neurons.

Complications of metrizamide myelography.

Adverse neurobehavioral reactions have not been emphasized as a complication of metrizamide myelography. We encountered six such reactions in approximately 250 metrizamide myelograms. All reactions followed either cervical myelography or panmyelography via lumbar puncture. We also treated a single case of tonic-clonic seizure after intracranial spill of metrizamide in a patient without a history of seizure disorder, and a case of myoclonus following a thoracic metrizamide myelogram that showed a highgrade block. Metrizamide should not be used if an intrathecal block is suspected, or if the location to be studied makes intracranial spill difficult to avoid.