Transition of immunocytochemistry from an anatomical to a biochemical science: evolution of the concept of neurotypy

Expression of neurofilament proteins by Purkinje cells: ultrastructural immunolocalization with monoclonal antibodies

Purkinje cell bodies in rodent cerebellum have been shown to express neurofilament protein epitopes but neurofilaments are rarely seen in these perikarya by classical morphological approaches. In an attempt to solve this enigma the ultrastructural distribution of two neurofilament epitopes was studied by immunoelectron microscopy with two monoclonal antibodies (Mabs) of divergent specificity: one, Mab 04-7 recognized a phosphorylated epitope, the other, Mab 02-135 a non-phosphorylated epitope. Longitudinal filamentous elements were heavily labeled in basket cell axons and afferent nerve fibers with both Mabs. While Mab 04-7 was unreactive with Purkinje cells, the immunoperoxidase reaction product with Mab 02-135 was distributed in the form of patches with no filamentous substructure throughout the cytoplasm of these cells. The data complement the results of other immunocytochemical studies showing the presence of all 3 neurofilament constituent proteins in Purkinje cell bodies, and lead to the conclusion that in these perikarya the majority of neurofilament proteins are not assembled in the form of neurofilaments.

Immunocytochemically-stained vasopressin binding sites in rat brain. Ventricular application of vasopressin/Accurel in the Brattleboro rat

An immunocytochemical procedure was developed to localize binding sites for vasopressin (VP) in the brain of Brattleboro (di/di) rats after 2 weeks of continuous ventricular administration of the peptide. Accurel-polypropylene tubing loaded with 0.15, 1.5 or 15 micrograms vasopressin was implanted into the lateral ventricle. Subsequently, bound VP was detected immunocytochemically in 2 distinct patterns: in perineuronal structures and dots between cells, in the lateral septum (dorsorostral part), striatum, cingulate cortex, granular cells of the dentate gyrus of the hippocampus, pyramidal cells of CA1 and CA3 hippocampal areas and around cerebellar Purkinje cells. The high dose (15 micrograms) loaded implants revealed the most intense staining; in the cytoplasm of neuronal cell bodies in the lateral and medial septum, striatum, cingulate cortex, bed nucleus of the stria terminalis, organum vasculosum of the laminae terminalis and locus coeruleus. The most intense staining in cell bodies was observed in brains which had low-loaded implants (0.15-1.5 microgram). A variety of controls, proved that no aspecific uptake was involved in the present procedure. The distribution of VP binding sites was only partly coincident with known sites of VP fiber innervation, and largely agrees with data obtained by autoradiographic techniques for [3H]VP binding. The present immunocytochemical technique gave a higher resolution than the currently used autoradiographic techniques. The differences in pattern and intensity of staining due to increasing the dosage rate of the in vivo vasopressin treatment, might mean that the current procedure retains preferentially either low or high affinity populations of binding sites depending on the implanted dose.

An immunohistochemical study of human central and peripheral nervous system tumors, using monoclonal antibodies against neurofilaments and glial filaments

Monoclonal antibodies that recognize either neurofilaments or glial filaments were used with the peroxidase-antiperoxidase (PAP) method to retrospectively study 100 tumors of the central and peripheral nervous systems in paraffin-embedded sections. Only neoplasms of putative neuronal origin or with presumed neuronal differentiation (paraganglioma, ganglioglioma, ganglioneuroblastoma, ganglioneuroma, neuroblastoma, ovarian teratoma and pheochromocytoma) contained tumor cells with immunoreactive neurofilament, but such cells were more common in the more differentiated or benign neoplasms in this category. Glial filament immunoreactivity was observed in tumor cells of glial origin and in tumor cells with foci of glial differentiation arising within the central nervous system, consistent with findings from previous studies using anti-glial-filament antisera. With the exception of a benign cystic teratoma, no glial filament immunoreactivity was observed outside the central nervous system. Some immunoreactive neurofilaments, but not glial filaments, were arranged in presumably abnormal balls, cords, or clumps within tumor cells, possibly reflecting cytoskeletal alterations related to neoplastic transformation. These findings indicate that monoclonal antibodies against intermediate filament proteins such as neurofilaments and glial filaments retain their specificity and sensitivity when employed in paraffin sections in conjunction with the peroxidase-antiperoxidase method. They suggest that such reagents are useful probes for the evaluation of the histogenesis or degree of differentiation in human nervous system tumors. Finally, they permit the speculation that the analysis of the intermediate filaments of tumor cells, as contrasted with those in normal cells, may provide new insights into the biology of neoplasms.

Neurohypophysial peptides and the hippocampus. I. Vasopressin immunoreactivity in the rat hippocampus

Immunocytochemical studies have demonstrated that nerve fibres containing immunoreactive vasopressin project to many areas of the central nervous system. In the present investigation, the presence of immunoreactive arginine vasopressin (IR-AVP) in the hippocampus of Wistar rats was confirmed by radioimmunoassay. The vasopressin content of the dorsal hippocampus was 30.3 +/- 7.3 pg IR-AVP/mg soluble protein (mean +/- SEM, n = 9) and that of the ventral hippocampus was 81.4 +/- 8.3 pg IR-AVP/mg soluble protein (n = 9), while tissue from the cerebral cortex contained no detectable vasopressin. That the immunoreactivity was due to vasopressin was confirmed by its absence in hippocampal or cortical tissue from homozygous Brattleboro rats, which are genetically unable to synthesize vasopressin.

Monoclonal antibodies distinguish phosphorylated and nonphosphorylated forms of neurofilaments in situ

The immunocytochemical staining patterns of 37 neuron-specific monoclonal antibodies previously described fell into four groups: (i) anti-synapse-associated, (ii) anti-neurofibrillar, (iii) anti-perikaryonal-neurofibrillar, and (iv) a single antibody reactive with a widely distributed epitope that covered the patterns of groups ii and iii. Antibodies of groups ii, iii, and iv were shown to be specific to neurofilament triplet subunits, even though there was little overlap in staining patterns between groups ii and iii. We examined nine of these antibodies as to their ability to distinguish functional states of neurofilaments dependent upon phosphorylation. Upon digestion with phosphatase, electroblot staining of neurofilament components was abolished with the five antibodies from group ii, enhanced with the three antibodies from group iii, and unaffected with antibody iv. Immunocytochemical staining of Bouin-fixed paraffin sections of rat brain was unaffected by phosphatase pretreatment. With antibodies of group ii, digestion with trypsin also left staining unaffected, but when followed by digestion with phosphatase, staining was diminished with three out of five antibodies. In contrast, digestion with trypsin abolished all staining with each antibody from group iii. If followed by digestion with phosphatase, staining reappeared, but the group iii pattern was replaced by a group ii pattern. Staining of this pattern was again abolished upon a second treatment with trypsin. The antibody from group iv lost most of its groups ii and iii staining patterns when sections were digested with trypsin. The group ii pattern reappeared and, indeed, was enhanced upon a subsequent phosphatase treatment and was reduced again upon a second trypsin treatment. Staining by four out of five antibodies from group ii was inhibited by inorganic phosphate. The data indicate that certain nerve cell bodies, their dendrites, and at least proximal axons possess nonphosphorylated neurofilaments and that long fibers, including terminal axons, possess phosphorylated neurofilaments. We propose that phosphorylation may be a factor in stabilizing compacted forms of neurofilaments and that heterogeneity of the compacted structures may play a role in a possible multiplicity of function within individual nerve cells.

Why monoclonals?

Until recently, antibodies used in a variety of immunochemical techniques have been raised in laboratory animals by repeated injections of purified antigen and adjuvant. These antisera are a mixture of antibodies that may recognize unique or common antigenic sites and have different affinities. Such polyclonal antisera can vary considerably from animal to animal or even with different bleeds from one animal. The hybridoma technique of Kohler and Milstein (1975) permits the selection of a monoclonal antibody with a specific affinity that recognizes a specific antigenic determinant. Monoclonal antibodies to a variety of neural antigens have been produced by injection of "impure" antigen. Immunocytochemical findings with these monoclonal antibodies demonstrate some of the advantages and problems of hybridoma technology. Studies with these antibodies in normal and pathological (B,B'-iminodipropionitrile intoxication, Alzheimer's disease) conditions illustrate some of the potential applications of monoclonal antibodies.