Paul C. Goldsmith

Immortalization of hypothalamic GnRH by genetically targeted tumorigenesis

By genetically targeting tumorigenesis to specific hypothalamic neurons in transgenic mice using the promoter region of the gonadotropin-releasing hormone (GnRH) gene to express the SV40 T-antigen oncogene, we have produced neuronal tumors and developed clonal, differentiated, neurosecretory cell lines. These cells extend neurites, express the endogenous mouse GnRH mRNA, release GnRH in response to depolarization, have regulatable fast Na+ channels found in neurons, and express neuronal, but not glial, cell markers. These immortalized cells will provide an invaluable model system for study of hypothalamic neurosecretory neurons that regulate reproduction. Significantly, their derivation demonstrates the feasibility of immortalizing differentiated neurons by targeting tumorigenesis in transgenic mice to specific neurons of the CNS.

ULTRASTRUCTURAL CHANGES IN IMPOTENT PENILE TISSUE : A COMPARISON OF 65 PATIENTS

To determine whether impotence is caused by specific and consistent changes in erectile tissue, we compared the ultrastructure of the corpora cavernosa in 6 controls with that in 59 patients undergoing implantation of a penile prosthesis. The impotent patients were divided into groups based on a medical history of hypertension (10), pelvic surgery (9), alcoholic smokers (8), hypertensive alcoholics (3), hypertensive alcoholic smokers (3), smokers (3), diabetics (8), diabetic smokers (3), Peyronies disease (3), spinal cord injury (3) and isolated causes (6). Our data demonstrate that different behavioral and/or medical conditions produce similar degenerative tissue responses. There is no single or specific cause of impotence that is manifest by consistent changes in erectile tissue.

Ultrastructural localization of luteinizing hormone-releasing hormone in the median eminence of the rat

The nature of cells in the hypothalamus that produce hypothalamic hypophysiotropic hormones remains unsettled. To investigate this problem electron microscopic immunocytochemistry was performed on thin sections of the median eminence of proestrous rats using antibodies to synthetic luteinizing hormone-releasing hormone (LHRH) and the peroxidase-antiperoxidase (PAP) technique. PAP complexes indicating the presence of LHRH were found over small, homogeneously dense neurosecretory granules 75-90 nm in diameter that occurred in clusters within neurons. Positive staining was more prevalent in the palisade zone in the anterior median eminence than at more posterior levels. LHRH positive granules were not obsereved in neuronal processes in the neurohemal contact zone, but appeared closer to the pericapillary space in the posterior median eminence than at more anterior levels. Occasional groups of LHRH positive granules were also found in the internal layer and hypendymal zone. No staining was observed in tanycytes or glial elements. These results support the hypothesis that hypothalamic hypophysiotropic hormones are produced by neurons and are stored in granules of homogeneous size and density in nerve processes located in the median eminence in the proximity of the hypophysial portal plexus.

Activin and inhibin in the human adrenal gland. Regulation and differential effects in fetal and adult cells.

Recent experimental data have revealed that activins and inhibins exert pivotal effects on development. As part of our studies on growth and differentiation of the human fetal adrenal gland, we examined the subunit localization, as well as the mitogenic and steroidogenic actions of activin and inhibin in human fetal and adult adrenals. All three activin and inhibin subunit proteins (alpha, beta A, and beta B) were detected in the fetal and adult adrenal cortex. Immunoreactive activin-A dimer was demonstrated in midgestation fetal and neonatal adrenals. ACTH1-24-stimulated fetal adrenal cell expression of alpha and beta A subunit messenger RNA. In addition, ACTH elicited a rise in levels of immunoreactive alpha subunit secreted by fetal and adult adrenal cells. Human recombinant activin-A inhibited mitogenesis and enhanced ACTH-stimulated cortisol secretion by cultured fetal zone cells, but not definitive zone or adult adrenal cells. Recombinant inhibin-A had no apparent mitogenic or steroidogenic effects. Thus, activin selectively suppressed fetal zone proliferation and enhanced the ACTH-induced shift in the cortisol/dehydroepiandrosterone sulfate ratio of fetal zone steroid production. These data indicate that activin-A may be an autocrine or paracrine factor regulated by ACTH, involved in modulating growth and differentiated function of the human fetal adrenal gland.

Neuroregulatory and neuroendocrine GnRH pathways in the hypothalamus and forebrain of the baboon

The distribution of neurons containing gonadotropin-releasing hormone (GnRH) in the baboon hypothalamus and forebrain was studied immunocytochemically by light and electron microscopy. GnRH was present in the perikarya, axonal and dendritic processes of immunoreactive neurons. Three populations of GnRH neurons could be distinguished. Most of the GnRH neurons which are assumed to directly influence the anterior pituitary were in the medial basal hypothalamus. Other cells that projected to the median eminence were found scattered throughout the hypothalamus. A second, larger population of neurons apparently was not involved with control of the anterior pituitary. These neurons were generally found within afferent and efferent pathways of the hypothalamus and forebrain, and may receive external information affecting reproduction. A few neurons projecting to the median eminence were also observed sending collaterals to other brain areas. Thus, in addition to their neuroendocrine role, these cells possibly have neuroregulatory functions. The inference is made that these bifunctional neurons, together with the widely observed GnRH-GnRH cellular interactions may help to synchronize ovulation and sexual behavior.

Location of the neuroendocrine gonadotropin-releasing hormone neurons in the monkey hypothalamus by retrograde tracing and immunostaining*,**.

In order to localize neuroendocrine gonadotropin‐releasing hormone (GnRH) neurons in the monkey hypothalamus, four juvenile cynomolgus macaques (one female, three males) were each given two or three microinjections (0.2 to 0.3 μl per site) of the retrograde tracer wheat germ agglutinin‐apoHorseradish peroxidase‐10 nm colloidal gold into the superficial, median eminence region of the infundibular stalk. Five to 15 days following surgery, the brains were fixed by perfusion and vibratomed at 40 μm in the frontal plane. Every 12th section was immunostained with rabbit anti‐GnRH using the peroxidase anti‐peroxidase technique with diaminobenzidine as the chromogen. Neuroendocrine GnRH neurons were easily identified in tissue sections as brown, immunostained cell bodies containing more than three distinct, dark blue, tracer‐filled lysosomes. Neuronal counts from each complete series of sections were compiled by anatomical region, and the percentages of GnRH and neuroendocrine GnRH neurons determined.

Glutamate and Gabaergic Neurointeractions in the Monkey Hypothalamus: A Quantitative Immunomorphological Study

Glutamate (Glu) and gamma-aminobutyric acid (GABA) are the most abundant excitatory and inhibitory neurotransmitters in the mammalian hypothalamus. Glu and GABA-containing neurons have both been shown to synapse with gonadotropin-releasing hormone (GnRH) and other neuroendocrine systems in the hypothalamus of several species. Since their direct interactions could play a pivotal role in governing neuroendocrine function, we performed double-label immunostaining for Glu and for glutamic acid decarboxylase (GAD) as a marker for GABAergic neurons in hypothalamic sections from adult female cynomolgus monkeys. Ultrastructural analysis of 785 Glu-immunoreactive (-ir) and GAD-ir elements in the medial septum (MS), arcuate nucleus-ventral hypothalamic tract (VHT1), supraoptic nucleus (SON), paraventricular nucleus (PVN), and median eminence (ME) revealed that 63% were Glu-ir, 28% were GAD-ir, and 9% were Glu + GAD-ir. In addition, we observed surprisingly consistent labeling of 2-4% somata (SOM), 65-80% dendrites (DEN), and 15-30% axons and terminals (AXO) in all of these areas. Characterization of 177 interactions (36% synapses, 64% contacts) by pre/post-transmitter content indicated that 29% contained Glu/GAD, 15% Glu/Glu, and 15% Glu/Glu + GAD, while 16% were unlabeled/Glu, 9% were unlabeled/GAD, and 16% expressed other transmitter combinations. Regional analysis of these interactions showed that 43% occurred in the MS, 22% in VHT1, 14% in SON, 9% in PVN, and 12% in the ME. AXO/DEN interactions made up 51% of all labeled interactions characterized, and were comprised 29% of Glu/GAD, 22% of Glu/Glu, and 18% of the Glu/Glu, and 18% of the Glu/Glu + GAD type. AXO/DEN synapses were more prevalent than contacts in all areas except the PVN and of course the ME, where anatomical synapses do not occur. AXO/SOM interactions represented approximately 15% of all those identified, and were predominantly unlabeled/Glu (71%) and unlabeled/GAD (18%) synapses. Almost all (95%) AXO/SOM synapses and 75% of the contacts occurred in the MS. DEN/DEN interactions, 28% of the total, were composed 50% of Glu/GAD, 12% of Glu/Glu, and 18% of the Glu/Glu+GAD type. The relatively few DEN/DEN synapses all appeared in the MS, whereas much more abundant DEN/DEN contacts were more widely distributed. DEN/SOM interactions, 6% of the total, appeared only as contacts, with the majority (60%) again located in the MS. In addition, the MS contained 48% of all asymmetrical synapses (vs. 35% in VHT1 and 17% in SON), 62% of all symmetrical synapses (vs. 19% in VHT1 and 14% in SON), and 35% of all contacts (vs. 21% in VHT1 and 12% in SON) identified.(ABSTRACT TRUNCATED AT 400 WORDS)

Expression of Estrogen and Progesterone Receptors in Glutamate and GABA Neurons of the Pubertal Female Monkey Hypothalamus

We have previously reported direct glutamate (Glu) synapses upon GnRH-containing neurons in the primate hypothalamus, and extensive interactions between Glu and aminobutyric acid (GABA) neurons in areas associated with reproductive function. Both Glu and GABA are known to affect peripubertal GnRH neurohormone release, but their relative roles remain unclear. In a developmental survey, estrogen receptors (ER) and progesterone receptors (PR) were virtually undetectable after immunostaining the hypothalamus of prepubertal monkeys, but were clearly evident in neurons of adults. We hypothesized, therefore, that Glu and GABA neurons which develop ER or PR expression during puberty may participate in reactivation of the hypothalamic-pituitary-gonadal axis. To identify those neurons in midpubertal female cynomolgus monkeys, we performed immunofluorescence staining for ER or for PR in separate sets of hypothalamic sections, and then immunostained for Glu or for glutamate decarboxylase (GAD, to identify GABA neurons) using a contrasting fluorophore. ER and PR were localized in the cytoplasm and nuclei of Glu and GAD neurons in nine hypothalamic and related brain regions. Quantitation revealed intranuclear ER in an average of 80% of the Glu neurons in all regions analyzed, and an average of 84% of the GAD neurons in all regions except the supraoptic nucleus (28%). Intranuclear PR expression was more variable, occurring in an average of 93% of the Glu neurons in seven regions, but in only 41% in the medial preoptic area, and 0% in the arcuate-periventicular zone. In addition, while intranuclear PR was seen in 96% of the GAD neurons in the septum, it appeared in 67% of the GAD neurons in the paraventricular nucleus, 47% in the medial preoptic area, 40% in the periventricular zone, and was absent from neurons in the supraoptic nucleus and mammillary bodies. In summary, certain subpopulations of Glu and GABA neurons in principal hypothalamic regions of the female monkey express ER and PR at midpuberty. Taken together with previous findings, these results suggest that Glu and GABA neurons which become sensitive to steroid hormones may help regulate GnRH neurohormone release and promote the onset of puberty. Since neuronal expression of ER or PR connotes sensitivity to gonadal feedback, and intranuclear translocation signals transcriptional activation, these results provide insights into the specific neuronal events involved in the peripubertal transition in primates.

Interactions between vasopressin- and gonadotropin-releasing-hormone-containing neuroendocrine neurons in the monkey supraoptic nucleus

Vasopressin (VP) is a hypophysiotropic hormone which is also implicated in the control of gonadotropin-releasing hormone (GnRH) secretion. We examined whether VP- and GnRH-immunoreactive (-IR) elements interact directly in the supraoptic nucleus (SON) of cynomolgus monkeys. Neuroendocrine (NEU) neurons in 4 juveniles were retrogradely labeled from the median eminence with wheat germ agglutinin apohorseradish peroxidase conjugated to gold before aldehyde perfusion. Frontal vibratome sections were immunostained for GnRH with peroxidase-antiperoxidase (PAP) and for VP with 5- or 15-nm gold. Many of the GnRH-IR and more than half the VP-IR cell bodies in the SON were NEU. VP-IR elements formed axodendritic and axosomatic symmetrical synapses with one another. In addition, VP-IR boutons also synapsed with NEU GnRH-IR neurons. Although GnRH axon terminals and dendrites contacted VP-IR dendrites and NEU cell bodies, we were unable to find convincing examples of GnRH/VP synapses through serial sections, perhaps due to the use of PAP-diaminobenzidine as the GnRH (afferent) immunolabel. In summary, our study demonstrates anatomical synapses between VP-IR and other VP and GnRH-IR neurons in the SON, in which postsynaptic VP or GnRH cell bodies were NEU. On the other hand, reciprocal GnRH/VP contacts but no true synapses were seen. However, the results suggest coordinated roles for VP and GnRH in NEU control of gonadotropin secretion. Whether VP itself and/or coexistent neuroeffectors act directly on NEU GnRH secretion remains to be determined. As such, VP neurons could help coordinate suppression of gonadotropins and augmentation of glucocorticoids during the stress response in primates.

Exogenous Glutamate Enhances Glutamate Receptor Subunit Expression during Selective Neuronal Injury in the Ventral Arcuate Nucleus of Postnatal Mice

Administration of high doses of glutamate (Glu) leads to selective neurodegeneration in discrete brain regions near circumventriclular organs of the early postnatal mouse. The arcuate nucleus-median eminence complex (ARC-ME) appears to be the most Glu-sensitive of these brain regions, perhaps because of the intimate relationships between its neurons and specialized astroglial tanycytes. To investigate the mechanism of Glu-induced neuronal loss, we administered graded doses of the sodium salt of glutamate (MSG) to postnatal mice, measured their plasma Glu concentrations, and performed microscopic analyses of the ARC-ME region 5 h after treatment. Nursing, 7-day-old mouse pups (CD1, Charles River, Hollister, Calif.) were injected subcutaneously with single doses of 0.1–0.5 or 1.0–4.0 mg of MSG per g BW, or with water vehicle alone. Mice were decapitated 5 h later and the brains immediately fixed by immersion in buffered aldehydes. Frontal vibratome tissue sections at comparable levels of the ARC-ME were examined by light microscopy. A dose of 4.0 mg MSG/g BW caused neurodegeneration throughout the ARC region, while 1.0 mg/g MSG resulted in less extensive damage. Injection of 0.2 mg MSG/g BW, which raised plasma Glu concentrations 17-fold after 15 min, was the minimum dose tested at which nuclear and cytoplasmic changes were observed in a small group of subependymal neurons near the lateral recesses of the third ventricle. Higher doses of 0.3–0.5 mg MSG caused injury to additional neurons situated farther laterally, but damage remained confined to the ventral region of the ARC nucleus. Ultrastructural examination showed some subependymal neurons with pyknotic nuclei, reduced cytoplasmic volume, and swollen subcellular organelles, while others had fragmented and condensed nuclear material. Immunostaining for tyrosine hydroxylase indicated that dopamine neurons were spared at the threshold dose, but suffered damage after higher doses of MSG. Immunostaining for Glu receptor subtypes revealed that 0.2 mg MSG/g BW enhanced neuronal expression of NMDAR1 and of GluR2/4, and that higher doses of MSG preferentially increased NMDAR1 expression in injured neurons. These results extend previous reports of Glu sensitivity in the ARC-ME region of 7-day postnatal mice. A dose of 0.2 mg MSG/g BW s.c. causes clear but discrete injury to specific subependymal neurons of undetermined phenotype near the base of the third ventricle. Slightly higher doses of MSG evoke damage of additional neurons confined to the ventral region of the ARC traversed by tanycytes. These same greater amounts of MSG promote dose-related increase in the expression of NMDAR1 more than of GluR2/4 in injured ARC neurons, suggesting that elevated Glu receptor levels may contribute to or be related to neuronal cell death. Taken together with previous findings, the data suggest that Glu responsitivity in the ARC-ME of the postnatal mouse may result from transient developmental conditions involving the numerical ratios and juxtaposition between tanycytes and neurons, expression of Glu receptors, and perhaps other ontogenetic factors which may not persist in the mature adult.