Ultrastructural immunocytochemical localization of neurophysin and vasopressin in the median eminence and posterior pituitary of the guinea pig

Summary With the use of tissue prepared by freeze-substitution and the unlabelled antibody enzyme technique, neurophysin and vasopressin were localized at the ultrastructural level in the posterior pituitary and median eminence of the guinea pig. In the posterior pituitary neurophysin was found in the large neurosecretory granules (1300–1500 Å) of axons, Herring bodies, and nerve terminals. In some of these axons immunoreactive neurophysin was found outside of granules in the axoplasm. By light microscopy neurophysin was found in both the zona interna and zona externa of the median eminence; this was confirmed by electron microscopy. In the zona interna as in the posterior pituitary, neurophysin was localized both inside and outside the large neurosecretory granules. In the zona externa, immunoreactive deposit was primarily located in granules with a diameter of 900–1100 Å in nerve terminals abutting on the primary portal plexus. The distribution of vasopressin paralleled that of neurophysin except that the hormone was rarely extragranular. These results demonstrate for the first time that both neurophysin and vasopressin are present in granules of axons that are in contact with the hypophysial portal vasculature.The authors wish to thank Dr. Alan Robinson for the gifts of antiserum to bovine neurophysin I and for purified bovine neurophysin I; Dr. Ludwig Sternberger for the peroxidase-anti-peroxidase complex; and Dr. Robert Utiger for antiserum to lysine vasopressinSupported in part by U.S. Public Health Service grant RR-00167 to the Wisconsin Regional Primate Research Center from the National Institutes of Health. Primate Center publication No. 14-017.Recipient of NIH, NINDS Teacher-Investigator Award NS-1108.     

THE DISTRIBUTION OF NEUROPHYSINS AND POSTERIOR PITUITARY HORMONES IN THE PORCINE NEUROHYPOPHYSEAL SYSTEM

Specific, homologous porcine neurophysin I and II radioimmunoassays were established together with specific oxytocin and vasopressin radioimmunoassays. The levels of each of these proteins and peptides were measured in acid extracts of individual paraventricular nuclei, supraoptic nuclei, neurohypophyseal stalks and posterior pituitary lobes of 12 pigs in order to quantitate the neurophysin-hormone relationships in the porcine neurohypophyseal system. Neurophysin III was found to be immunologically identical to neurophysin I. Neurophysin measurements by radioimmunoassay were quantitatively validated by scanning densitometry of polyacrylamide gels stained with 0.5% amido schwarz. In the hypothalamic nuclei vasopressin was in 3–4 M excess of oxytocin but in the neurohypophyseal stalk and posterior pituitary lobe the hormones were equimolar suggesting that the rate of formation of vasopressin differs from that of oxytocin. Neurophysin I immunoreactivity was present in a 3:1 molar ratio with neurophysin II throughout the porcine neurohypophyseal system. In posterior pituitary lobes total neurophysins were equimolar to total hormone concentrations. The specific activity (pmol/mg extracted protein) of oxytocin increased 1800 times, vasopressin 560 times and neurophysins about 360 times from the paraventricular nucleus to the posterior pituitary lobe. In the hypothalamic nuclei relationships between immunoreactive neurophysin I and vasopressin, and between neurophysin II and oxytocin were highly significant. In the posterior pituitary lobe each immunoreactive neurophysin level correlated with both hormone levels. Quantification of densitometric scans of stained polyacrylamide gels from neurophypophyseal extracts and immunoreactivity patterns of neurophysins in eluates of sliced, duplicate gels indicated that neurophysin III decreased distally within the neurohypophyseal tract while neurophysin I increased. The results demonstrated that vasopressin was associated with porcine neurophysin I. However, oxytocin may be associated with both immunoreactive neurophysin I and neurophysin II in the porcine neurohypophyseal system if a 1:1 molar ratio of neurophysin to hormone is to be maintained. Neurophysin III contributed to the stoichiometry of this relationship.     

Neurosecretion in the external and internal zone of the median eminence of the cat and dog

Summary The immunoglobulin-enzyme bridge technique, in association with rabbit antiporcine neurophysin-II has been applied to the median eminence of the dog and cat in order to study the distribution of neurophysin-like antigens throughout this area of the brain and correlate the findings with the corresponding distribution of neurosecretory material (NSM) as revealed by the crotonaldehyde fuchsin stain.Neurophysin and NSM were both present in the hypothalamo-supraoptico-neurohypophysial system—the pathway taken by oxytocin, vasopressin and neurophysin from the hypothalamus to the posterior pituitary lobe.Whereas the tuberoinfundibular tract of the median eminence was almost devoid of NSM, the presence of neurophysin-like material was clearly evident using immunoperoxidase histochemistry. The significance of a protein in the external median eminence possessing determinants cross-reactive against anti-neurophysin serum is discussed.This work was financed by a grant from the Medical Research Council of New Zealand.     

The distribution of vasopressin-, oxytocin-, and neurophysin-producing neurons in the guinea pig brain

Summary The location, cytology and projections of vasopressin-, oxytocin-, and neurophysin-producing neurons in the guinea pig were investigated using specific antisera against vasopressin, oxytocin or neurophysin in the unlabeled antibody enzyme immunoperoxidase method. Light microscopic examination of the neurons of the supraoptic and paraventricular nuclei shows that hormone is transported not only in axons, but also in processes having the characteristics of dendrites. Neurons were found to contain only vasopressin or oxytocin; all neurons containing neurophysin appear to contain either vasopressin or oxytocin. In the neural lobe, vasopressin and oxytocin terminals are intermingled. In the median eminence, vasopressin and oxytocin fibers are intermingled in the internal zone. In a caudal portion of the median eminence, a number of vasopressin and neurophysin (but few oxytocin) axons enter the external zone from the internal zone, and surround portal capillaries. In the supraoptic nucleus, vasopressin neurons outnumber oxytocin neurons with a ratio of at least 5:1. The paraventricular nucleus is separated into two distinct groups of neurons, a lateral group consisting of only vasopressin neurons, and a medial group consisting of only oxytocin neurons. In addition to axons passing to the neurohypophysis, a number of axons appear to interconnect the supraoptic and paraventricular nuclei.Supported by the Deutsche Forschungsgemeinschaft (SFB 51, C/21 and C/27), (We 608/3)Acknowledgements. The authors are greatly indebted to Mmes. R. Köpp-Eckmann, B. Reijerman, A. Scheiber, I. Wild and Mr. U. Schrell for technical assistance, to Mmes. P. Campbell and U. Wolf for editorial assistance, and to Dr. R.R. Dries and Ferring Pharmaceuticals, Kiel, for the generous provision of high quality peptides     

Maturation of the hypothalamo-neurohypophysial system

Summary Sections of the hypothalamus, median eminence and pituitary from fetal and neonatal rats were examined with the immunoperoxidase staining technique and light microscopy. Purified antisera raised against vasopressin and oxytocin, and antisera cross-reactive with rat neurophysin were used to localize these antigens in the hypothalamo-neurohypophysial system (HNS). Neurophysin was detected throughout the HNS of the 18-day fetal rat. Vasopressin was present in the hypothalamus and pituitary of the 19-day fetus, and in the median eminence of the 4-day neonate. Oxytocin was not detected in the pituitary until 1–2 days after birth, in the hypothalamus after 4 days, and in the median eminence after 8 days. During the first days after birth the supraoptic nucleus was more mature than the paraventricular nucleus. The HNS did not approach maturity until at least 7 days after birth. The relative maturity of the supraoptic nucleus compared with the paraventricular nucleus, and the detection of vasopressin before oxytocin are evidence for the one-neuron-one-hormone theory. The data do not exclude the possibility that the fetal hypothalamo-neurohypophysial system, and perhaps the fetal hormone, vasotocin, affect the initiation and course of parturition.This work was financed by the Medical Research Council of New Zealand     

Structure of the tail fin in teleosts

The influence of adrenalectomy and administration of hypertonic saline on the amount of vasopressin, oxytocin, and neurophysin contained in the median eminence and the neural lobe of rats was studied by means of the following methods: (i) morphometric and microphotometric analyses of aldehyde fuchsin-stained histological sections of the neurohypophysis; (ii) immunohistochemical demonstration of vasopressin, oxytocin, and neurophysin in the neurohypophysis, and (iii) radioimmunological measurement of vasopressin and oxytocin in extracts of the median eminence and the neural lobe. Adrenalectomy increases the amount of vasopressin and neurophysin in the external layer of the median eminence but does not change the content of oxytocin. It has no influence on the amount of vasopressin, oxytocin, and neurophysin demonstrable in the inner layer of the median eminence and in the neural lobe two weeks after the operation. Hypertonic saline markedly diminishes the vasopressin, oxytocin, and neurophysin content of the inner layer of the median eminence and the neural lobe but reduces only slightly, if at all, the amount of vasopressin and neurophysin in the outer layer of the median eminence. The findings support the concept that osmotic stress reduces only the vasopressin and oxytocin content of the hypothalamus-neural lobe system and has no or only little influence on the vasopressin content of the outer layer of the median eminence.     

Different populations of granules and their distribution in the hypothalamo-neurohypophysial tract of the rat under various experimental conditions

In connection with previous studies (Krisch, 1974), a morphologic analysis of the median eminence of male, female, pregnant, and thirsting rats has been made, in order to evaluate: 1. whether the median eminence is an additional area where oxytocin and vasopressin may be released from neurosecretory axons in the zona interna; 2. whether it is possible to distinguish different populations of axons according to the mean diameters of their granules in the external zone of the median eminence. In the zona interna of the male and the female control animals, the oxytocin- and vasopressin-containing granules have already reached their final diameters, i.e. their size remains constant until they reach and are stored in the neurohypophysis. However, the relative proportions of these populations in the fibre layer of the median eminence indicate certain differences between male and female animals, concerning the turnover of oxytocin vasopressin. In pregnant, as well as in thirsting animals, there is an apparent decrease of the mean granule diameters in oxytocin- and vasopressin-containing axons. In particular in the latter group, ballooned axons in the fibre layer of the median eminence, containing disintegrating granules, tubular structures filled with electron dense material, and fine-flocculent material in the axoplasm, together with the morphometric results, support our concept that in case of highly increased peripheral demand the contents of the mature, large granules are released into the axoplasm of the fibre layer in the median eminence. In addition, axons containing granules smaller than 115 nm may be classified into four populations with mean diameters of 84 nm, 96 nm, 103 nm, and 100n. The same populations are found in the palisade zone of the median eminence as well as in the pericapillary zone, but in contrast to the zona interna there is a fifth population of axons containing very small granules with a mean diameter of about 70-75 nm. The relative proportions of these five populations differ between the zona palisadica and the pericapillary zone on the one hand and between the experimental groups on the other hand. Some morphological findings concerning the tanycytes are discussed with respect to previously published data.     

Relationship of the central ACTH-immunoreactive opiocortin system to the median eminence and the pituitary gland of the rat

Summary By immunocytochemical methods, the present study describes ACTH-immunoreactive fibers in the pituitary stalk and neural lobe. This opiocortin-hypothalamo-neural lobe projection arises in a bed nucleus of perikarya in the basal hypothalamus, follows supraoptico-hypophyseal fibers in the zona interna of the median eminence, and distributes throughout the neural lobe. No ACTH-immunoreactive fibers project to the zona externa; some are present in the subependymal layer and at the lateral margins of the median eminence. Further studies must identify the role of these fibers in posterior lobe function. It remains also to be determined whether this system terminates upon primary pituitary portal capillaries and delivers opiocortin neuropeptides to the adenohypophysis.Supported by NIH Grants HD-07962, NS-15345 and AM-22029The skillful technical assistance of Donna Wilson, Nancy Dembs and Jay Hocton is thankfully acknowledged     

Isolation of a third bovine neurophysin

1. A third native hormone-binding protein, neurophysin-C, has been isolated from acetone-desiccated bovine pituitary posterior lobes. 2. This protein was detected in lysates of neurosecretory granules isolated from bovine pituitary posterior lobes. 3. The molecular weight appears to be close to 10000. 4. Neurophysin-C is similar in amino acid composition to neurophysin-I and -II; it contains a single residue of tyrosine and of methionine. The N-terminal amino acid in all three neurophysins is alanine. 5. Neurophysin-C accounts for approximately 15% of the total hormone-binding protein present in the pituitary posterior lobes. 6. The new neurophysin forms complexes with oxytocin as well as with [8-arginine]-vasopressin. The complex with vasopressin has been crystallized. 7. Bioassay of the pressor and oxytocic activities of the protein-hormone complexes shows that neurophysin-C binds one molecule of either vasopressin or oxytocin.     

Neuropeptides in the median eminence: Their sources and destinations

By radioimmunoassay and immunocytochemical techniques, 14 neuropeptides have been measured and localized in the rat median eminence. Neuropeptides with inhibitory or stimulatory effects on the anterior pituitary hormones as well as posterior pituitary hormones are present in the median eminence in the highest concentrations of the central nervous system. All these peptides (LH-RH, TRH, somatostatin, CRF, vasopressin, oxytocin) are of preoptic or hypothalamic origin and they are transported to the median eminence by loop-like fiber systems through the lateral retrochiasmatic area. Within the median eminence, the pericapillary space constitutes the main common pathway. Three major transport routes—axons, vessels, liquor spaces—are separated from each others by only basement membranes, which allow free communications downwards to the pituitary but also backwards to the central nervous system.     

Distribution pattern in the human pituitary and hypothalamus of a new neuropeptide: The C-terminal glycoprotein-fragment of human pro-pressophysin (CPP)

Summary The distribution pattern of CPP-containing neurons and fibers in the human pituitary and hypothalamus was studied with a specific antiserum to human CPP and the unlabeled antibody technique. Immunoreactive CPP was found in the magnocellular neurons of the supraoptic nucleus (SON), the paraventricular nucleus (PVN) and in neurons scattered in the supraoptic hypophyseal tract. CPP-containing parvocellular neurons were found in the suprachiasmatic nucleus (SCN). The CPP-containing fibers from the magnocellular neurons formed a tract coursing through the median eminence and the pituitary stalk to the posterior lobe of the hypophysis. In contrast, no such fibers from the SCN projected to SON, PVN and the median eminence. This pattern is identical to that of vasopressin and its associated neurophysin-containing neurons and fibers and strongly supports the concept that CPP is a part of the common precursor for vasopressin and neurophysin II. The biological importance of human CPP other than being a precursor fragment remains to be elucidated.To whom requests for reprints should be addressed     

Vasopressinergic and oxytocinergic pathways in the central nervous system

Recent data obtained by immunohistochemical and other anatomical tracing methods indicate that oxytocin and vasopressin pathways are much more complex and extensive than previously recognized. In addition to the classic magnocellular neurons that project from the supraoptic and paraventricular (PVN) nuclei to the posterior pituitary gland, generally smaller neurons in various parts of the PVN send vasopressin fibers to the portal capillary bed in the median eminence, or send oxytocin or vasopressin projections to other brain and spinal cord sites. In addition, vasopressin neurons are also found in the suprachiasmatic nucleus and may contribute to extrahypothalamic projection areas. Many of these axonal projections appear to form synapses with other neurons in forebrain, hindbrain, and spinal cord regions, which suggests roles for these peptides in neuronal communication. In brain stem and spinal cord, terminal fields include both parasympathetic and sympathetic regulatory centers. Oxytocin terminals are also found on large intracerebral arteries where the peptide may regulate cerebral blood flow.     

Biosynthesis and axonal transport of rat neurohypophysial proteins and peptides

35S-cysteine injected adjacent to the supraoptic nucleus (SON) of the rat is rapidly incorporated into proteins. These 35S-cysteine-labeled proteins in the SON (1-24 h after injection) were separated by polyacrylamide gel electrophoresis, and the distribution of radioactive proteins on the gels was analyzed. 1 h after injection, about 73% of the radioactivity appeared in two peaks (both about 20,000 mol wt). With time, these peaks (putative precursors of neurophysin) decreased, as a 12,000 mol wt peak (containing two distinct neurophysins) increased in radioactivity. Both the 20,000- and 12,000-mol wt proteins are transported into the axonal (median eminence) and nerve terminal (posterior pituitary) regions of the rat hypothalamo-neurohypophysial system. Conversion of the larger precursor protein to the smaller neurophysin appears to occur, in large part, intra-axonally during axonal transport. Six distinct 35S-cysteine-labeled peptides (less than 2500 mol wt), in addition to arginine vasopressin and oxytocin, are also synthesized in the SON and transported to the posterior pituitary where they are released together with labeled neurophysin by potassium depolarization in the presence of extracellular calcium. These data provide support for the hypothesis that the neurohypophysial peptides (vasopressin and oxytocin) and neurophysins are derived from the post- translational clevage of protein precursors synthesized in the SON, and that the conversion process can occur in the neurosecretory granule during axonal transport.     

The topography of mesotocin and vasotocin systems in the brain of the domestic mallard and japanese quail: Immunocytochemical identification

Summary The neurosecretory systems producing mesotocin (MT) and vasotocin (VT) (the avian homologues of oxytocin and vasopressin, respectively) were characterized in the brains of the domestic mallard and Japanese quail by means of indirect immunofluorescence techniques using specific antisera. In the anterior preoptic region, including the organum vasculosum of the lamina terminalis, and at different levels of the supraoptic and paraventricular nuclei, separate mesotocin- and vasotocin-producing neurons were identified. Mesotocinergic and vasotocinergic neurons were also located in the tuberomammillary area, among the ectomammillary tract fibers. The supraoptico-neurohypophysial tract, formed by vasotocin- and mesotocincontaining axons, enters the internal zone of the median eminence and ends in the posterior lobe of the pituitary. The external zone of the rostral median eminence appears to contain vasotocin and mesotocin fibers, which terminate in close contact with the capillaries of the hypophysial portal system.With contributions by Dr. B. Kerdelhué, Laboratoire des Hormones Polypeptidiques du CNRS, 91190 Gif-sur-Yvette, France     

Localization of hormone secreting pathways in the brain by immunohistochemistry and light microscopy: a review.

Immunocytochemical techniques are now being used to localize hypothalamic neurosecretory hormones and related peptides in the mammalian brain. The data are probably incomplete, due primarily to false negative results. A number of previous assumptions concerning these pathways have been confirmed while other unexpected results were obtained. As expected, vasopressin and oxytocin and their associated proteins, neurophysins, were found in the magnocellular cell bodies of the hypothalamus and in their axonal projections to the neural lobe of the pituitary. Gonadotropin-releasing hormone (Gn-RH), somatostatin, and thyrotropin-releasing hormone (TRH) were located in what appears to be parvicellular nerve terminals on portal capillaries. Gn-RH has been found in perikarya in the arcuate nucleus, which is considered a source of fibers to the portal capillary bed. An extensive network of cell bodies and fibers in the preoptic area was also found to contain Gn-RH, and others in the periventricular nucleus in the anterior hypothalamus reacted with antiserum to somatostatin. Unexpected was considerable evidence that vasopressin is secreted directly into hypophyseal portal blood. This hormone and its neurophysin were also found in parvicellular neurons in the suprachiasmatic nucleus of rodents. All the hormones were found in fibers in the organum vasculosum of the lamina terminalis and in the posterior pituitary gland.     

Potentiation of vasopressin analgesia in rats treated neonatally with monosodium glutamate

The analgesic response elicited by central administration of arginine vasopressin (AVP) appears to be dependent upon the integrity of the hypothalamic paraventricular nucleus (PVN), since lesions placed in the PVN eliminate AVP analgesia. A projection to the zona externa of the median eminence constitutes one of the VP-containing efferents of the PVN. Neonatal treatment with monosodium glutamate (MSG) destroys perikarya of the arcuate nucleus and median eminence. The present study examined whether AVP analgesia was affected in the MSG-treated rat and whether these alterations were accompanied by specific changes in VP immunoreactivity in the zona externa of the median eminence. Female rats, neonatally treated with either MSG or a saline control, were tested as adults on the tail-flick test following intracerebroventricular injections of 0, 75, 150 and 500 ng doses of AVP. After testing, selected animals were prepared for AVP and oxytocin immunocytochemistry of the median eminence. Significant potentiations in the magnitude of AVP analgesia were observed in MSG-treated rats. AVP and oxytocin immunoreactivity in the zona interna and oxytocin immunoreactivity in the zona externa of the median eminence were similar in MSG-treated and control rats. In contrast, AVP immunoreactivity in the zona externa of the median eminence was markedly reduced in the MSG-treated rat. These data suggest that VP analgesia may normally be inhibited by those medial-basal hypothalamic neurons affected by neonatal MSG treatment.     

Hypothalamo-neurohypophysial system in hagfish: localization of neurophysin-, arg-vasopressin- and oxytocin-like immunoreactivity

Serial paraffin sections of hagfish brain (Eptatretus burgeri) were immunocytochemically examined for porcine neurophysin (NEU), arg-vasopressin (AVP) and oxytocin (OT). A big number of NEU-immunoreactive nerve cells were visible in the frontal part of the ventromedial hypothalamus in close vicinity of the third ventricle. These cells were bipolar in most cases and sometimes additionally stained for AVP but not for OT. NEU-reactive neurons extended axons to the median eminence and the posterior lobe of the pituitary which also contained AVP or OT-like immunoreactivity. Immunostaining outside the hypothalamus could not be observed. It is likely that there exists a mammalian-like oxytocinergic and a vasopressinergic hypothalamo-neurohypophysial system in hagfish, despite the large differences in levels of organization.     

Correlations between brain catecholamines,neurosecretion, and serum corticoid levels in osmotically stressed mallard ducks (Anas platyrhynchos)

The effects of depleting brain catecholamines with a combined treatment of reserpine and alpha-methyl-p-tyrosine on serum corticosterone levels and release of immunoreactive neurophysin from the median eminence, in osmotically stressed and unstressed mallard ducks, were studied. Corticoid levels in salt loaded birds were more than three times that of unstressed birds. The combined treatment of reserpine and alpha-methyl-p-tyrosine significantly decreased the concentration of brain monoamines in all experimental groups and raised serum corticoid levels in non-stressed birds to the same level found in the osmotically stressed animals. Immunoreactive neurophysin in the zona externa of the median eminence was depleted in all birds subjected to either osmotic stress and/or reserpine treatment but not in unstressed control birds. These preliminary data indicate that catecholamines may exert an inhibitory influence on both ACTH release from the anterior pituitary and neurophysin from the median eminence and that these two events may in some way be interrelated in the duck.     

Immunocytochemical study of the GABAergic innervation of the mouse pituitary by use of antibodies against gamma-aminobutyric acid (GABA)

Summary The GABAergic innervation of the mouse pituitary, including the median eminence, was studied at light microscopic and ultrastructural levels by use of a pre-embedding immunocytochemical technique with antibodies directed against GABA. In the median eminence, a high density of GABA-immunoreactive fibers was found in the external layer where the GABAergic varicosities were frequently observed surrounding the blood vessels of the primary capillary plexus. In the internal and subependymal layers, only few fibers were immunoreactive. The intense labeling of the external layer was observed in the entire rostro-caudal extent of the median eminence. In the pituitary proper, a dense network of GABA-immunoreactive fibers was revealed throughout the neural and intermediate lobes, entering via the hypophyseal stalk. The anterior and tuberal lobes were devoid of any immunoreactivity. The GABA-immunoreactive terminals were characterized in the median eminence, and in the intermediate and posterior lobes at the electron-microscopic level. They contained small clear vesicles, occasionally associated with dense-core vesicles or neurosecretory granules. In the intermediate lobe they were seen to be in contact with the glandular cells. In the posterior lobe and in the median eminence, GABA-immunoreactive terminals were frequently located in the vicinity of blood vessels. These results further support the concept of a role of GABA in the regulation of hypophyseal functions, via the portal blood for the anterior lobe, directly on the cells in the intermediate lobe, and via axo-axonic mechanisms in the median eminence and posterior lobe.     

Subcellular organization of neurophysins, oxytocin, [8-lysine]-vasopressin and adenosine triphosphatase in porcine posterior pituitary lobes

Posterior pituitary lobes from young pigs were fractionated by differential and sucrose-density-gradient centrifugation. The distributions of oxytocin and [8-lysine]-vasopressin were measured by bioassay and the distributions of neurophysin-I and -II by radioimmunoassays specific for each of these two proteins. Most of the hormone and neurophysin applied to the density gradient was localized in particles with the density expected of neurosecretory granules. However, the neurosecretory granules were separated into two bands (D and E). A close statistical correlation between the distributions of [8-lysine]-vasopressin and neurophysin-I, and of oxytocin and neurophysin-II on the gradients, suggested that in vivo porcine neurophysin-I binds [8-lysine]-vasopressin within one population of granules and porcine neurophysin-II binds oxytocin within another type of granule. However, there was no significant separation of oxytocin and vasopressin in fractions D and E. The molar ratios of hormones and neurophysins indicated that there was insufficient of either neurophysin to bind the [8-lysine]-vasopressin in the granule fractions or in the whole gland. Polyacrylamide-gel electrophoresis showed that only bands corresponding in mobility to porcine neurophysins-I, -II and -III were present in large amounts in the whole gland and in the granule fractions. The component with the mobility of neurophysin-III was, however, relatively enriched in whole young glands and granule fractions compared with adult gland extracts. It is suggested that the vasopressin that cannot be assigned to neurophysin-I may occur in (a) vesicles containing vasopressin but no neurophysin, (b) vesicles containing vasopressin and a protein that cannot be quantified by the radioimmunoassays used, such as porcine neurophysin-III, or (c) normal vasopressin–neurophysin granules which have accumulated extra vasopressin. Band E of the gradient was rich in adenosine triphosphatase activity, whereas band D possessed very little of this enzyme.