Distribution of beta-endorphin and related peptides in the hypothalamus and pituitary.

β-Endorphin and its precursors, β-lipotropin and pro-opiocortin, in the pituitary and hypothalamus of the rat, were studied by means of radioimmunoassays with antisera to β-endorphin, β-melanocyte stimulating hormone and adrenocorticotropin. Gel filtration chromatography in combination with radioimmunoassay revealed the existence of β-endorphin, β-lipotropin and pro-opiocortin in all the tissues studied, with the highest concentration in the posterior-intermediate lobe of the pituitary gland. Great variation was observed in the relative amounts of these three β-endorphin-like immuno-reactive peptides in different regions. The ratio of β-endorphin to β-lipotropin is highest in the hypothalamus, is less in the posterior intermediate pituitary lobe and lowest in the anterior pituitary lobe. A β-melanocyte-stimulating hormone-like immunoreactive peptide, which is larger than porcine β-melanocyte-stimlating hormone and smaller than β-lipotropin, was detected in the posterior-intermediate lobe but not in the anterior lobe of the pituitary.The results suggest that the pro-opiocortin may act mainly as a precursor of adrenocorticotropin in the anterior pituitary and that β-endorphin may act mainly in the intermediate lobe of the pituitary where there is conversion of β-lipotropin to β-endorphin and the β-melanocyte-stimulating hormone-like immunoreactive peptide.     

Estrogen induction of galanin synthesis in the rat anterior pituitary gland demonstrated by in situ hybridization and immunohistochemistry

The distribution of galanin messenger (mRNA) and galanin-like immunoreactivity (Gal-LI) in the anterior and posterior pituitaries of control and estrogen-implanted female rats was determined by in situ hybridization and immunohistochemical methods. In control ovariectomized animals galanin mRNA was undetectable in the posterior, intermediate and anterior pituitary. However, 4 days after implantation with 10 mg of diethylstilbestrol, galanin mRNA was clearly present in the anterior pituitary, but not in the posterior or intermediate lobe. By immunohistochemistry Gal-LI was readily visualized and detected in the posterior lobe, but clearly undetectable in cells of the intermediate and anterior pituitary lobes of control animals. After estrogen administration numerous cells exhibiting intense Gal-LI were evident in the anterior lobe, while Gal-LI remained unchanged in the intermediate and posterior lobes. These results indicate that in control animals galanin is stored, but not synthesized, in the posterior pituitary and that after estrogen administration galanin production is substantially increased in the anterior pituitary. We conclude that the expression of galanin in the anterior pituitary is regulated by estrogen and suggest that galanin may be a pituitary hormone.     

Angiotensin converting enzyme immunohistochemistry in rat brain and pituitary gland: correlation of isozyme type with cellular localization

We have localized angiotensin converting enzyme in rat brain and pituitary gland immunohistochemically with an anti-rat lung angiotensin converting enzyme monoclonal antibody. The distribution of immunoreactive angiotensin converting enzyme is identical with that of binding sites for the angiotensin converting enzyme inhibitor, [3H]captopril. Most intense staining is in the choroid plexus and subfornical organ, with intermediate values in the caudate-putamen, globus pallidus, entopeduncular nucleus, pars reticulata of the substantia nigra, posterior pituitary and anterior pituitary. Lower levels are observed in the supraoptic and paraventricular nuclei of the hypothalamus. Within the basal ganglia angiotensin converting enzyme immunoreactivity is distributed throughout the neuropil; no cell bodies are stained, even after colchicine treatment. The punctate pattern of immunoreactivity in the anterior pituitary corresponds to the distribution of endothelial cells. The posterior pituitary is stained diffusely. Angiotensin converting enzyme is increased by 45% in the posterior lobe after pituitary stalk section, demonstrating that this diffuse staining is associated with pituicytes. Antibody specificity was demonstrated by the immunoaffinity purification of angiotensin converting enzyme to homogeneity from crude tissue extracts using anti-angiotensin converting enzyme antibody and protein A-sepharose. The apparent molecular weight by sodium dodecyl sulfate polyacrylamide gel electrophoresis of lung, choroid plexus and anterior pituitary angiotensin converting enzyme is 175,000. In the substantia nigra and caudate putamen, where angiotensin converting enzyme is localized to neuronal as opposed to epithelial cells, the molecular weight is 165,000. The pituicyte angiotensin converting enzyme of the posterior pituitary is 170,000 daltons.     

Effect of stress and dexamethasone on immunoreactive β-endorphin levels in rat hypothalamus and pineal

In response to mild stress the levels of immunoreactive β-endorphin in rat anterior pituitary, hypothalamus and pineal fell within 10 minutes from 210 to 129 pmol/lobe, 1.47 to 0.89 pmol/mg protein and 2.53 to 0.41 pmol/gland, respectively. No alterations were found to take place in β-endorphin levels in posterior pituitary or plasma. Dexamethasone pre-treatment given 18 h prior to stress resulted in significantly greater reduction of β-endorphin levels in hypothalamus and pineal than stress alone—hypothalamic levels fell to 0.73 pmol/mg protein and pineal to 0.07 pmol/gland. Plasma β-endorphin levels in dexamethasone pretreated stressed rats were significantly lower than in intact rats (42 fmol/ml vs. 98 fmol/ml). The almost complete disappearance of β-endorphin from the pineal in response to stress and dexamethasone suggests that pineal does not itself synthesize the hormone but only utilizes and/or stores it. Gel filtration analysis of the β-endorphin im-munoreactivity in tissue extracts and plasma showed that anterior pituitary and plasma contain three immunoreactive components, eluting like β-endorphin,βP-Epotropin and pro-opiocortin, whereas only β-endorphin-like material was detected in posterior pituitary, hypothalamus and pineal.     

Distribution of laminin in the murine pituitary

The distribution and cellular localization of the glycoprotein laminin were investigated by light and electron microscopic immunocytochemistry in the adult murine pituitary gland. Immunoblots confirmed that laminin was the only protein in the pituitary gland of the adult male mouse to react with antilaminin serum. Laminin immunoreactivity was demonstrated at the light microscopic level simultaneously with that of β-follicle stimulating hormone (β-FSH) and β-luteinizing hormone (β-LH). In addition to its distribution is basal laminae, laminin immunoreactivity was coincidently expressed in gonadotrophs with the immunoreactivities of β-FSH and β-LH. Electron microscopic immunocytochemistry was employed on aldehyde-fixed sections embedded in L.R. White. Sites of binding of primary antisera to laminin were identified with affinity-purified secondary antisera directly coupled to 20 nm particles of colloidal gold. Three antisera recognizing laminin were compared and found to result in an identical pattern of immunoreactivity. Laminin was found extracellulary only in formed basal laminae in all three lobes of the pituitary and was not found in extracellular matrices of connective tissue. Laminin immunoreactivity was also found intracellularly in gonadotrophs but in none of the other endocrine or non-endocrine cells of the anterior lobe. Within gonadotrophs, only secretory granules were labeled. The majority, but not all, secretory granules were labeled in each of the gonadotrophs examined, and the proportion of granules labeled with laminin could not be increased by doubling the concentration of anti-laminin serum. Laminin immunocreactivity segregated with the subset of secretory granules containing β-FSH. In contrast, laminin immunoreactivity was absent in the smaller subset of secretory granules that contain serotonin. No secretory granules were labeled within the endocrine cells of the intermediate lobe, nor within secretory granules of neural elements in the posterior lobe.     

Serological studies on rabbit antibodies to the rabbit anterior pituitary

Rabbits immunized with suspensions or extracts of rabbit anterior pituitary in Freund adjuvant may develop specific antibodies to components of the rabbit pituitary. Immunofluorescent staining with such antisera occurred in isolated cells of the anterior pituitary. These correspond to cells stained with acid fuchsin, i.e. acidophils or alpha cells. Some of the pituitary antisera fix complement with pituitary extracts. A tanned-cell haemagglutination test using pituitary extracts as coating antigen yielded positive reactions with some of the pituitary antisera. The rabbit antisera appeared to be specific for the anterior pituitary within the limits of the rabbit organs tested. Hog, guinea-pig, dog and beef pituitaries share the antigen, but monkey and human pituitaries fail to react.

Immunofluorescent staining revealed that antisera reacted with the pituitary of the antibody producing rabbit, i.e. the sera contain pituitary autoantibodies. No direct or indirect evidence for pathological changes in the autoantibody producing animals could be found.

     

Is motilin a cerebellar peptide in the rat?

Summary Motilin was demonstrated by the immunoperoxidase technique in endocrine cells of the gastrointestinal tract using several specific antisera. Motilin-like immunoreactivity could only be demonstrated with one of these antisera and was observed in Purkinje cells and dendrites of the cerebellum, in pyramidal cells and dendrites of the cerebral cortex and in dendrites of the CA3 field of the hippocampus of the rat.Very low motilin-like immunoreactivity was found in cerebellum as well as in cerebral cortex using radioimmunoassay. However, using reverse phase liquid chromatography combined with UV-detection and radioimmunoassay, no peak of a peptide corresponding to synthetic motilin was detectable in rat cerebellar extracts, in contrast to findings in rat duodenum. The results do not suggest that motilin is an intrinsic neuroactive substance of the cerebellum.     

Effect of tibolone administration on central and peripheral levels of allopregnanolone and beta-endorphin in female rats

OBJECTIVE: We evaluated the effects of tibolone oral administration on neuroendocrine function by investigating the modulation exerted by tibolone administration on allopregnanolone and central and peripheral beta-endorphin (beta-EP) levels in ovariectomized rats. DESIGN: Female Wistar rats (N = 64) were included: 48 rats were ovariectomized, 8 cycling rats were included as controls, and 8 cycling rats were treated with placebo. The ovariectomized animals were divided into six groups: untreated rats and those that received 14-day oral treatment with either placebo, estradiol valerate (E2V) 0.05 mg/kg/d, or tibolone (0.1, 0.5, or 2 mg/kg/d. beta-EP levels were assessed in the frontal lobe, parietal lobe, hippocampus, hypothalamus, anterior pituitary, neurointermediate pituitary, and plasma, whereas allopregnanolone levels were measured in the frontal lobe, parietal lobe, hippocampus, hypothalamus, anterior pituitary, adrenal glands, and serum. RESULTS: The administration of tibolone (0.5 and 2 mg/kg/d) in ovariectomized rats induces a significant increase of allopregnanolone in the frontal lobe, parietal lobe, hippocampus, hypothalamus, whereas in serum a significant increase of allopregnanolone occurs only with the dose of 2 mg/kg/d, a significant decrease in allopregnanolone levels occurs in the adrenal glands. No changes occurred in the anterior pituitary. Tibolone doses of 0.5 and 2 mg/kg/d induced a significant increase in beta-EP content in the frontal lobe, hypothalamus, and neurointermediate lobe; and, at doses of 2 mg/kg/d, in the parietal lobe, anterior pituitary, and plasma, without changes in the hippocampus. Compared with E2V, 0.5 mg/kg/d tibolone showed a similar effect on allopregnanolone and beta-EP in most brain regions. CONCLUSIONS: Tibolone administration affects beta-EP and allopregnanolone levels, playing a role as a neuroendocrine modulator.     

Deprivation of REM sleep in the rat and the opioid peptides beta-endorphin and dynorphin

Effects of 'rapid eye movement' sleep deprivation (REMd) on two opioid peptides, beta-endorphin and dynorphin, were studied in rats. Both peptides were measured by radioimmunoassay techniques. The level of beta-endorphin was estimated in the hypothalamus, in the anterior lobe of the pituitary and in the blood. The amount of dynorphin was estimated in the hypothalamus. REMd was induced for 72 h and achieved by two different methods, the platform technique and the pendulum technique. Three control groups were additionally run. As a consequence of REMd, an increase in beta-endorphin level was discovered in the blood plasma, while a small decrease was found in the hypothalamus. No changes could be detected for beta-endorphin levels in the pituitary or for hypothalamic dynorphin concentration. The deprivation effects are interpreted as belonging to a group of changes, all of which point to a small increase in tonic arousal as a result of REMd.     

Galanin coexists with vasopressin in the normal rat hypothalamus and galanin's synthesis is increased in the Brattleboro (diabetes insipidus) rat

Galanin is a peptide containing 29 amino acid residues, that is present in the median eminence, in the magnocellular neurons of the supraoptic (SON) and paraventricular nuclei (PVN) of the rat hypothalamus and in the posterior pituitary. We report here that: (1) immunoreactivity for galanin (GAL) and vasopressin coexist in the SON of normal rats, (2) levels of mRNA encoding preprogalanin are markedly elevated in the PVN and SON of Brattleboro (diabetes insipidus) rats, as determined by in situ hybridization histochemistry but (3) levels of GAL-like immunoreactivity (GAL-LI) are significantly reduced in the posterior pituitary of these rats, as determined by radioimmunoassay. We suggest that production and possibly secretion of the peptide GAL may be increased in the Brattleboro rat.     

Distribution and co-existence of Met-enkephalin-like and mesotocin-like immunoreactivity in the neural lobe of the pituitary of the frog

Content and distribution of Met-enkephalin (Met-ENK)-like immunoreactivity in the hypothalamo-hypophysial system of bovine, rat and frog were examined using specific radioimmunoassay and immunohistochemistry. Ultrastructural localization and co-existence of Met-ENK-, mesotocin (MT)- and vasotocin (VT)-like immunoreactivity in the neural lobe of the frog pituitary was examined by a method combining pre-embedding peroxidase-antiperoxidase immunostaining for Met-ENK with post-embedding immunocolloidal gold staining for MT or VT. The highest concentrations of immunoassayable Met-ENK were present in the neural lobe of the pituitary of the frog. In addition to nerve fibers showing only MT-like or Met-ENK-like immunoreactivity, nerve fibers containing neurosecretory granules showing both MT- and Met-ENK-like immunoreactivities were very rich. But VT-like and Met-ENK-like immunoreactivity was confirmed separately in different axon terminals.     

Organisation of basement membrane components in the human adult and fetal pituitary gland and in pituitary adenomas

 Cell–matrix interactions undoubtedly have a role in the development and maintenance of the complex nonrandom structure of the human pituitary gland. We have extended previous studies by documenting the patterns of immunoreactivity for type IV collagen, laminin and fibronectin in the fetal gland, comparing these with the adult patterns. In both we have examined the differences between the anterior lobe and intermediate zone in an attempt to elucidate the apparent differences in functional response between corticotrophs in the two areas. We have also examined expression of these proteins in a series of pituitary adenomas. Finally, we have immunolocalised β₄ integrin, a component of the α₆β₄ laminin receptor, in the adult gland and in adenomas. In the anterior lobe of the adult gland, type IV collagen and laminin were present in both epithelial and vascular basement membrane. Fibronectin was related to the basement membrane but showed a less continuous distribution. β₄ Integrin was expressed on the basal aspects of pituitary cells, in association with laminin, suggesting that this did identify the α₆β₄ laminin receptor. In addition, immunoreactivity was present on the lateral margins of some pituitary cells, which might indicate a role in cell–cell adhesion. None of the proteins showed specific association with any particular cell type, suggesting that these specific interactions do not regulate differentiation. This pattern of expression had developed in the fetal gland by the second trimester, with expression relating to vessels preceding that in epithelial basement membrane. Type IV collagen, laminin and fibronectin were also expressed in epithelial and vascular basement membrane in the intermediate zone of the adult gland, and around Rathke’s cleft in the fetal gland. However, the organisation differed, with larger groups of cells enclosed within a single basement membrane. Possible vascular connections demonstrated between the posterior lobe and the intermediate zone would permit access of posterior lobe hormones to this zone. Our data confirmed disruption of expression in pituitary adenomas, type IV collagen, laminin and β₄ integrin having a mainly perivascular distribution, with more variable immunoreactivity for fibronectin. Received: 17 February 1997 / Accepted: 17 May 1997     

Nerve growth factor-like activity in the rat pituitary intermediate lobe

Nerve growth factor (NGF)-like immunoreactivity was demonstrated in the rat pituitary intermediate lobe by indirect immunofluorescence method using antisera specific to beta-NGF isolated from adult male mouse submaxillary salivary gland. Co-culture of frozen or fresh intermediate lobes with newborn rat superior cervical ganglion resulted in marked fiber growth from the ganglion, which was totally inhibited by NGF antiserum, suggesting the presence in situ and secretion in vitro of biologically active pituitary NGF. Pituitary stalk transection caused decrease in both the NGF immunoreactivity and biological activity. These findings suggest that pituitary NGF level is under neural regulation.     

Comparative localization of corticotropin and corticotropin releasing factor-like peptides in the brain and hypophysis of a primitive vertebrate, the sturgeonAcipenser ruthenus L.

Summary The sturgeon is a primitive actinopterigian fish that, unlike modern teleosts, possess a portal vascular system that connects a true median eminence with the anterior pituitary as in mammals. The occurrence and localization of corticotropin and corticotropin releasing factor-like immunoreactivities were examined in the brain of the sturgeon (Acipenser ruthenus L.) by immunocytochemistry with antisera raised against synthetic non-conjugated human corticotropin, and rat/human corticotropin releasing factor. In the hypothalamus, corticotropin-immunoreactive parvicellular perikarya were found in the infundibular nucleus and in dendritic projections to the infundibular recess. In addition, ependymofugal corticotropin-immunoreactive fibres were found to terminate in the ventral hypothalamus. Corticotropin releasing factor-immunoreactive neurons were found in the rostral portion of the ventral hypothalamus (tuberal nucleus), and in the vicinity of the rostral aspect of the lateral recess. These cells projected to the dorsal hypothalamus, the ventral hypothalmus, the median eminence, the anterior and posterior telencephalon, the tegmentum mesencephali, and the pars nervosa of the pituitary. An affinitypurified UI antiserum failed to stain the sturgeon hypothalamus. Corticotrophs in the rostral pars distalis of the pituitary were also corticotropin-immunoreactive. In the neurointermediate lobe, only about 50% of cells of the pars intermedia appeared to be corticotropin-positive, the rest appeared unstained. These results suggest that the presence of corticotropin-like and corticotropin releasing factor-like peptides in the brain is a relatively early event in vertebrate evolution, already occurring in Chondrostean/Actinopterigian fishes, as exemplified byA. ruthenus.The close spatial relationship between corticotropin releasing factor immunoreactivity and corticotropin immunoreactivity in the ventral hypothalamus ofA. ruthenus supports a possible interaction between the two systems in that area of the sturgeon brain. The pars intermedia might be an important site for corticotropin synthesis, even though the possibility cannot be excluded that the antiserum was recognizing the proopiomelanocortin molecule. The occurrence of corticotropin releasing factor immunoreactivity in the region of median eminence/pars intermedia of the sturgeon suggests that the sturgeon corticotropin releasing factor might regulate the adenohypophyseal release of proopiomelanocortin products in the same manner as in other vertebrates. The presence of extrahypothalamic corticotropin releasing factor-immunoreactive projections suggests further neuromolulatory functions for this peptide inA. ruthenus.     

Demonstration of a pancreatic proelastase 2-alpha 1-protease inhibitor complex in normal human plasma

A peak of immunoreactive pancreatic elastase 2 with a molecular weight consistent with that of a complex of elastase 2 and alpha 1-protease inhibitor (also referred to as alpha 1-antitrypsin) can be detected by radioimmunoassay in normal human serum or plasma (Geokas et al., J. Biol. Chem. 252:61-67, 1977). This material has been purified by gel filtration on Sephadex G-200 and by ion-exchange chromatography on DEAE-cellulose. The alpha 1-protease inhibitor-bound immunoreactive elastase 2 has been dissociated by incubation with hydroxylamine, and the resulting immunoreactive product isolated by gel filtration on Sephadex G-100. The dissociated immunoreactive elastase 2 was shown by affinity chromatography on turkey egg white inhibitor-bound agarose, before and after activation by bovine trypsin, to consist only of proelastase 2. A second peak of immunoreactive material associated with the high molecular weight fraction of plasma has been shown to result from a specific interaction of the 125I-labeled phenylmethanesulfonyl-elastase 2 employed as tracer in the radioimmunoassay with alpha 2-macroglobulin, resulting in apparent immunoreactivity. These results demonstrate that all of the detectable immunoreactive pancreatic elastase 2 in normal human plasma is proelastase 2 bound to alpha 1-protease inhibitor.     

Heterologous radioimmunoassay of atrial natriuretic polypeptide in dog and rabbit plasma

Atrial natriuretic peptide (ANP) was measured in plasma of dogs and rabbits by radioimmunoassay (RIA) using a commercially available anti alpha-ANP serum and compared to our measurements of ANP in rats and humans. Plasma concentration of ANP in dog coronary sinus (234.9 +/- 41.0 pg/ml) was significantly greater than in systemic arterial blood (81.2 +/- 8.4 pg/ml). Gel filtration of dog coronary sinus plasma resulted in an ANP peak with the elution volume (Ve) of synthetic atriopeptin III (AIII) and a minor peak eluting with the void volume (Vo). Rabbit systemic arterial plasma ANP was 53.3 +/- 4.3 pg/ml and yielded one peak, with a Ve of AIII. Ion exchange chromatography of dog and rabbit atrial extracts (AE) resulted in a major ANP region which resembled AIII. Gel filtration of AE showed larger molecular species as well as AIII. Dilutions of dog and rabbit plasma and AE were parallel with the AIII standard in radioimmunoassay.     

Distribution of neuropeptide Y-like immunoreactivity in the rat central nervous system--I. Radioimmunoassay and chromatographic characterisation

The distribution of neuropeptide Y-like immunoreactivity in the rat brain was investigated by means of immunochemical techniques. In the first part of the study (present paper) neuropeptide Y radioimmunoassays were characterised and the chromatographic properties and regional distribution of neuropeptide Y-like immunoreactivity was investigated. The second part of the study (accompanying paper) involved immunohistochemical techniques. Extracts from several regions of rat brain were found to contain immunoreactivity that behaved like synthetic porcine neuropeptide Y in three test systems: dilution in the radioimmunoassay (test of antigenic properties), gel chromatography (molecular weight), reverse phase high performance liquid chromatography (solubility properties). Experiments were conducted to optimise the extraction of neuropeptide Y. Boiling 0.1 M sodium phosphate buffer, pH 7.4, extracted at least two times as much immunoreactivity from whole brain pieces as other buffers. The nature of the extracted immunoreactivity was confirmed using chromatography. Experiments (using added iodinated or unlabelled neuropeptide Y standards) demonstrated that the differences between extraction media could not be explained by differential recovery of the peptide, although differences in recovery between media existed. Tissue sample weight was found to influence neuropeptide Y recovery. Evidence that rat neuropeptide Y-like immunoreactivity was not identical to the porcine peptide was obtained from experiments which demonstrated an early eluting peak of immunoreactivity in addition to the main peak on high performance liquid chromatograms. This material could be generated by oxidation of extracted rat neuropeptide Y, suggesting the presence in the rat peptide of a methionine residue. Some evidence of high molecular weight neuropeptide Y precursors was obtained from chromatography of hypothalamus extracts. Bovine pancreatic polypeptide-like material represented less than 1% of the amounts of neuropeptide Y in the brain. The distribution of neuropeptide Y-like immunoreactivity was non-uniform in the rat brain with highest concentrations observed in the hypothalamus, amygdaloid complex and periaqueductal central gray matter. Other regions of forebrain contained moderate to high concentrations including olfactory tubercle, striatum, nucleus accumbens, neocortex and hippocampus. Negligible amounts were detected in the cerebellum. In spinal cord immunoreactivity was concentrated in the dorsal horn, although measurable amounts were found in the ventral horn. The neurointermediate but not anterior lobe of the pituitary contained neuropeptide Y.     

Galanin lacks binding sites in the porcine pituitary and has no detectable effect on oxytocin and vasopressin release from rat neurosecretory endings.

High concentrations of immunoreactive galanin-like material in rat hypothalamus, median eminence and neurohypophysis have been reported in the literature suggesting a regulatory role of galanin on hormone release from the anterior and posterior lobe of the pituitary. We studied binding of iodinated galanin to crude membrane preparations from porcine anterior hypothalamus, anterior and neurointermediate lobe of the hypophysis. In contrast to the hypothalamus where specific binding of 125I-galanin was found, there was no displaceable galanin binding in membranes of the anterior or neurointermediate lobe of porcine pituitaries. Effects of galanin on oxytocin and vasopressin release were investigated using isolated neurosecretory endings from rat neurohypophyses. Galanin had no detectable effect on the release of oxytocin or vasopressin.     

Immunocytochemical localization of the brain-specific S-100 protein in the pituitary gland of adult rat

Summary The brain-specific S-100 protein was localized at the electron microscopic level in the anterior and posterior pituitary gland of adult rat by indirect immunoperoxidase histology. The protein was found in the stellate cells of the pars distalis and tuberalis, in the marginal cells that line the hypophyseal cleft and in the glia-like cells, the pituicytes, of the neural lobe. The pituicytes, the stellate cells and the marginal cells have in common at least two properties: they all express a brain-specific marker and they are satellite cells to the secretory axons in the neural lobe and of the secretory cells in the adenohypophysis. These properties suggest that the S-100 cells in the pituitary gland are neuroectodermal in origin, possibly glial in nature.     

Immunohistochemical localization of neuron-specific enolase in the human hypophysis and pituitary adenomas

Neuron-specific enolase (NSE) was localized, using the immunoperoxidase technique, in the cytoplasm of the five adenohypophyseal hormone-secreting cell types, and in nerve fibers of the pars nervosa of the human pituitary. Crooke's hyaline material was negative. Neuron-specific enolase was found in all pituitary adenoma types; there was no correlation between degree of granularity or differentiation of tumor cells and intensity of NSE immunopositivity. One hypothalamic hamartoma was positive for NSE; a craniopharyngioma and a neurohypophyseal granular cell tumor were not. Neuron-specific enolase was present in peptide hormone-producing endocrine cells outside the pituitary and in their tumors; the majority of other tumors were negative for NSE, although one breast carcinoma, one ovarian cystadenocarcinoma, and one lymphoma were positive for NSE. In control studies, absorption of NSE antisera with growth hormone abolished immunoreactivity; there was no immunologic cross-reaction demonstrable by radioimmunoassay.