Chris W. Pool

Immunofluorescence of vasopressin and oxytocin in the rat hypothalamo-neurohypophyseal system

The present paper deals with the development of an immunofluorescence procedure that allows specific localization of vasopressin and oxytocin in the hypothalamo-neurohypophyseal system (HNS) of the rat. Antibodies against arginine-vasopressin (AVP), lysine-vasopressin (LVP) and oxytocin were raised by injecting these hormones that were covalently bound to thyroglobulin into rabbits. The vasopressin-immunized rabbits showed periods of diabetes insipidus, while histology of the HNS revealed an intact neurosecretory system with signs of increased endogenous hormone synthesis in the supraoptic nucleus and increased release in the neurohypophysis of some rabbits. The daily water intake of the oxytocinimmunized rabbits was similar to that of control rabbits. The development of antibodies against vasopressin as measured in the immunofluorescence procedure showed a course that was quite different from the curve of the titer as determined by radioimmunoassay (RIA). Also the specificity of the antibodies used in the immunofluorescence procedure was found to be quite different from their specificity in a RIA system. Potency and specificity of the antibodies have to be studied therefore within the immunofluorescence procedure itself. Using freshly frozen acetone-postfixed hypothalami or pituitaries, no sharp localization of immunofluorescence could be obtained in the HNS. Therefore prefixation was performed. Both, the type and the duration of prefixation revealed quite different results regarding the immunofluorescence in the neurosecretory cell bodies in the hypothalamus and of their endings in the neurohypophysis. The best immunofluorescence results were obtained using 6 hours glyoxal-prefixation for the hypothalamus and 24 hours formalin-prefixation for the pituitary. The cross-reaction of the antibodies for oxytocin or vasopressin was tested on synthetic hormones that were bound to CNBr-activated agarose beads and mounted on glass slides. All anti-plasmas showed cross-reaction on beads containing the heterologous antigen. The plasmas were purified by incubation with beads containing the heterologous hormone until the crossreacting component had been removed. Using purified antibodies, the distribution of oxytocin and vasopressin cells within the HNS was investigated. More oxytocin containing cells were localized in the rostral part and more vasopressin in the caudal part of both, the supraoptic (SON) and paraventricular nucleus (PVN). Comparable percentages of oxytocin and vasopressin containing cells were found in the SON and PVN. The absolute amount of oxytocin containing cells was 2.5 times more in the SON than in the PVN, which seems to contradict the “classical” view that the PVN predominantly or entirely synthetizes oxytocin. In addition, fluorescence was found using antibodies against vasopressin in the suprachiasmatic nucleus in Wistar rats and heterozygous Brattleboro rats, but not in this nucleus of homozygous Brattleboros.

Differences in colocalization between Fos and PHI, GRP, VIP and VP in neurons of the rat suprachiasmatic nucleus after a light stimulus during the phase delay versus the phase advance period of the night

Two groups of four rats each received a 15‐minute light stimulus during the first part of the night (ZT14) and the second part (ZT19), respectively. After 45–60 minutes, the animals were killed by perfusion fixation. Adjacent Vibratome sections through the suprachiasmatic nucleus (SCN) were double‐immunostained for the presence of peptide histidine isoleucine (PHI), gastrin releasing peptide (GRP) or vasoactive intestinal peptide (VIP) with Fos by using fluorophore‐conjugated secondary antibodies. A few sections were triple‐immunostained for PHI, GRP or VIP with vasopressin (VP) and Fos. Sections were analyzed with a confocal laser scanning microscope. It turned out that the ZT19 light stimulus induced 4.2 times more nuclear profiles in the SCN immunoreactive for Fos than the light stimulus given at ZT14. The SCN of control animals did not show any Fos immunoreactivity. After the ZT14 light stimulus, approximately 33% of the Fos profiles showed colocalization with a perikaryal profile immunoreactive for PHI, GRP or VIP, whereas at ZT19, this percentage had doubled to approximately 65%. After the light stimulus at ZT14, the relatively low Fos induction was numerically and proportionally most prominent in the PHI‐immunoreactive perikarya. As compared with ZT14, the increase of Fos after the ZT19 light stimulus was most pronounced in the GRP‐immunoreactive perikarya (21×) followed by VIP (15×) and PHI (5×) This outcome suggests that at least three different cell groups characterized by, respectively, PHI alone, GRP, and VIP fully or partly colocalized with PHI, play a prominent role during light‐induced phase shifts: the PHI neurons during light‐induced phase delays, the GRP and VIP/(PHI) neurons during light‐induced phase advances.

Can specificity ever be proved in immunocytochemical staining.

Any investigator working with immunocytochemical (ICC) localization techniques can appreciate the statement of Petrusz et al. (1) that specificity is the most difficult criterion to fulfill. According to these authors, the only direct way to establish specificity is absorption of the antibody by an antigen of high purity. Ever since we compared systematically the ICC localization data in tissue with those obtained in model systems we have, however, noticed that the absorption technique alone is absolutely insufficient to prove the specificity for the homologous antigen. Since overestimation of the value of the absorption test used in this way is found throughout the ICC localization literature, we thought it worthwhile to react and present some data that illustrate our point of view, and to suggest possible alternative specificity tests. We agree with Petrusz et al. (1) that radioimmunoassay (RIA) cannot help us to characterize the antibodies for ICC procedures. The difference between the specificity revealed by these two techniques may be at least partly due to the much higher dilution in which the antibodies are used in the RIA as was illustrated by immunoelectronmicroscopical data showing dilution dependent specificity (5). Specificity has thus to be studied in the same antibody dilution as is used for the ICC localization technique. The dissimilarity between RIA and ICC holds not only true for the specificity but also for the potency of the antiserum. During the course of antibody development against different hormones in rabbits, we frequently observed that the titre in the RIA system was rising or remained elevated (e.g., 1:10”) while the fluorescence obtained with the same antibody in tissue or on model systems gradually decreased to useless values (cf. e.g., 4). As Petrusz et al. (1) already stated, the proof that positive staining results exclusively from immunochemical binding of the antibodies in the primary antiserum to the tissue presents relatively little difficulty. Although the increasing dilution of the primary antiserum as proposed by these authors may be of some value, possible disturbing factors causing nonimmunochemical reactions will also be diluted in this way. It seems, therefore, preferable to test the pre-immune serum of the same rabbit in the same dilution to prove the (non)immunocytochemical nature ofthe reaction. Notjust the plasma from a single rabbit since each animal may have its own characteristics in this respect. For the same reason it seems even more advisable to use plasma that has been absorbed to the antigen as a check for the antibody antigen reaction. This plasma contains all the factors that influence the ICC reaction with the exception of the antibodies that will be studied further. We disagree fundamentally with Petrusz et al. (1) with respect to the value of the absorption test alone to prove specificity. If no ICC staining is obtained anymore after absorption of the antibody by the antigen, it proves no more than that all the antibodies were bound to the added antigen. In the first place it does not exclude staining caused by unwanted or unexpected antibodies which were raised because of impurities in the injected antigen which is also used for the absorption test. Most antigens have to be purified from biological material and will thus never be absolutely pure. In the second place it does not exclude cross reaction. This phenomenon is even not mentioned by Petrusz et al. (1) although it is one of the main problems of the ICC localization techniques, even when, theoretically speaking, pure antigen could be used. The data we obtained with antibodies that were raised to synthetic argininevasopressin (AVP) can serve as an illustration of this point. These antibodies showed a good positive reaction in rat brain and pituitary sections. However, in homozygous Brattleboro rats that lack AVP by a genetical defect, a bright immunofluorescence was also found using these antibodies. This fluorescence appeared to be due to cross reaction with the structurally closely related hormone oxytocin. Yet, solid phase absorption to AVP of this antibody against AVP prevented any ICC staining and would thus be “specific” according to Petrusz’ criteria. Specific localization of AVP was only possible after removal of the cross reacting antibodies by solid phase

Enkephalin immunoreactivity in synaptoid elements on glial cells in the rat neural lobe

Opioid peptides were localized in fibres of the rat neural lobe using various immunocytochemical methods at the light- and electron-microscopical level. Leu-enkephalin immunoreactivity was present in beaded fibres distributed throughout the neural lobe. These fibres surround the neurohypophyseal glial cells (pituicytes) and make synaptoid contacts upon their soma and processes. The reaction product was localized both in dense-core vesicles of about 100 nm in diameter and diffusely spread over the cytoplasm. No arguments in support of the co-existence of enkephalins and the neurohypophyseal hormones vasopressin and oxytocin in the same terminal were found. It is suggested that pituicytes might mediate the inhibitory effect of opiod peptides on vasopressin and oxytocin release from the neural lobe.

Adeno-associated viral vector-mediated gene transfer of brain-derived neurotrophic factor reverses atrophy of rubrospinal neurons following both acute and chronic spinal cord injury.

Rubrospinal neurons (RSNs) undergo marked atrophy after cervical axotomy. This progressive atrophy may impair the regenerative capacity of RSNs in response to repair strategies that are targeted to promote rubrospinal tract regeneration. Here, we investigated whether we could achieve long-term rescue of RSNs from lesion-induced atrophy by adeno-associated viral (AAV) vector-mediated gene transfer of brain-derived neurotrophic factor (BDNF). We show for the first time that AAV vectors can be used for the persistent transduction of highly atrophic neurons in the red nucleus (RN) for up to 18 months after injury. Furthermore, BDNF gene transfer into the RN following spinal axotomy resulted in counteraction of atrophy in both the acute and chronic stage after injury. These novel findings demonstrate that a gene therapeutic approach can be used to reverse atrophy of lesioned CNS neurons for an extended period of time.

Activation of Vasopressin Neurons in Aging and Alzheimer's Disease

The supraoptic (SON) and paraventricular nuclei (PVN) of the human hypothalamus are production sites of vasopressin (AVP) and oxytocin (OXT). Although the hypothalamus is affected in Alzheimers disease (AD), previous work has not only shown that in these two nuclei no neurons are lost, neither during aging nor in AD, but that the number of AVP‐expressing neurons and their nucleolar size had even increased with age. These observations indicated that the peptide synthesis of the AVP neurons was activated in the oldest age‐groups. Recently published, qualitative observations, using the area of the Golgi Apparatus (GA) as a sensitive parameter for neurosecretory activity, confirmed the activation of SON and PVN neurons with age in human; however, in this report the neurons were not identified according to their neuropeptide content. In the present quantitative study we determined whether the AVP neurons were indeed activated as a result of the aging process in controls and AD patients. We applied a polyclonal antiserum directed against the medial cisternae of the GA on formalin‐fixed, paraffin‐embedded tissue sections taken from the dorsolateral SON (dl‐SON) of 10 controls and 10 AD patients, and performed our measurements in this area that is known to be predominantly occupied (90–95%) by AVP neurons. In addition, the sparse OXT cells present in the area of study, were excluded from the measurements on the basis of alternative sections stained for OXT. In the dl‐SON, the area occupied by the GA and the cellular profile area per patient were quantified by means of image analysis. The results show a significant increase in GA area with age in controls and in AD, demonstrating an activation of the AVP neurons in the dl‐SON of the human hypothalamus in these two conditions. No changes were observed in the cellular profile areas with age, neither in the controls nor in AD, suggesting that the GA area is a much more sensitive parameter for monitoring activity changes in post‐mortem material than neuronal size. It is proposed that this activation of AVP cells with age, which has been suggested to be a compensatory response to the age‐related loss of AVP receptors in the kidney, might be the basis of the stability of these neurons in aging and AD.

Quantification of antiserum reactivity in immunocytochemistry. Two new methods for measuring peroxidase activity on antigen-coupled beads incubated according to an immunocytoperoxidase method.

Antigens covalently coupled to agarose beads provide a matrix for an economical, sensitive, and quantitative immunocytochemical detection of antiserum bindings potencies. Despite some very powerful features (e.g., the ability to control the outcome of a solid phase adsorption on the same matrix), the use of this technique is not very widespread when compared with the other enzyme-linked immunosorbent assay (ELISA) techniques. The main reason for this is the necessity for rather laborious measurements of the immunocytochemical tracer on individual beads. A description of two new methods for the batch measurement of the peroxidase activity on immunoperoxidase incubated antigen-coupled beads is presented. The first method involves the measurement of the diaminobenzidine (DAB) extinction from a large number of beads with a scanning microspectrophotometer. In the second method, during the peroxidase reaction, the beads are incubated with o-phenyldiamine (OPD), which is soluble both in the reduced and oxidized form, whereby absorbance measurements of the supernatant of the beads in a normal spectrophotometer are possible. The sensitivity and the quantitative relation between bound first antibody and absorbance are compared for both methods. From the two immunoperoxidase procedures used (the three step peroxidase-antiperoxidase and the two-step peroxidase conjugate procedure) only the latter met the conditions for a quantitative (first) antibody assay.

In situ hybridization for vasopressin mRNA in the human supraoptic and paraventricular nucleus; quantitative aspects of formalin-fixed paraffin-embedded tissue sections as compared to cryostat sections

In order to study the suitability of formalin-fixed paraffin-embedded brain tissue for vasopressin (AVP)-mRNA detection, we used symmetric halves of 5 human hypothalami. In every case, one half was formalin fixed for 10-35 days and paraffin embedded while the other half was frozen rapidly. Following in situ hybridization (ISH) histochemistry on systematically obtained sections of the supraoptic (SON) and paraventricular nucleus (PVN) of both halves, total amounts of AVP-mRNA in these nuclei were estimated using densitometry of film autoradiographs. Total amounts of radioactivity were found to vary considerably between patients and amounted to 1297 +/- 302 arbitrary units (AU) (PVN) (mean +/- SEM) and 2539 +/- 346 (SON) for the cryostat sections and 868 +/- 94 (PVN) and 1259 +/- 126 (SON) for the paraffin tissue. Variations introduced by the method itself yielded a coefficient of variation of only 0.19. Furthermore, a non-significant negative trend with postmortem delay was found in cryostat tissue, but not in paraffin sections. No effect of fixation time was observed in the paraffin tissue. Both ways of tissue treatment have specific advantages and disadvantages that may be different for other probes or other brain areas. For ISH of a highly abundant mRNA like AVP in a very heterogeneous brain area such as the human hypothalamus, formalin-fixed paraffin-embedded tissue sections can be used for quantitative analysis of entire brain nuclei because of the small variation in this tissue, the remarkably good signal recovery (some 75% as compared to cryostat sections) and its practical advantages with regards to anatomical orientation, storage and sampling of the tissue.

Press-blotting on gelatin-coated nitrocellulose membranes: A method for sensitive quantitative immunodetection of peptides after gel isoelectric focusing

A method is presented for the fixation of peptides in nitrocellulose membranes after isoelectric focusing on thin polyacrylamide gels. Focusing gels are covered with gelatin-coated nitrocellulose membrane. Using glutaraldehyde, focused peptides are covalently fixed onto this membrane. Fixed peptides are stained using the peroxidase-anti-peroxidase method and the immunoreaction is quantified by rendering the membrane transparent and measuring the optical density of the precipitated chromogen in each band. The effect of pore size and gelatin content of the membrane, glutaraldehyde concentration and fixation time on fixation efficiency and immunostaining has been investigated. Gelatin coating considerably increases the efficiency of glutaraldehyde fixation of peptides and greatly enhances antibody-binding. Consequently, sensitive quantitative immunodetection is possible and, depending on the antiserum, peptides are readily detected in quantities down to 10 pg.

The monoclonal antibody Alz-50, used to reveal cytoskeletal changes in Alzheimer's disease, also reacts with a large subpopulation of somatostatin neurons in the normal human hypothalamus and adjoining areas

The monoclonal antibody Alz-50 is directed against Alzheimers disease-related modified tau proteins and reveals cytoskeletal changes, i.e. neurofibrillary tangles and dystrophic neurites. The present study shows that, in the hypothalamus of non-demented control subjects, this same antibody gives a distinctive staining pattern of a subpopulation of somatostatin neurons and beaded fibres. Furthermore, Alz-50 occasionally recognizes somatostatin-containing cell bodies and dystrophic neurite-like fibers in the (neuritic) senile plaques of AD patients. These observations have direct consequences for the interpretation of Alz-50 staining in diagnostic usage and for the assessment of Alzheimers disease-like changes induced by beta-amyloid in experimental animal brains. On dot spotting, Alz-50 was found to bind to a number of fragments from the somatostatin precursor, of which somatostatin 15-28 stained best. Preadsorption of Alz-50 by somatostatin 15-28, as well as other specificity tests, failed, however, to provide any clue to the nature of the unknown compound(s) stained in the control hypothalamus.