The problem of specificity of antisera is one of the most important points in immunohistochemistry as well as in radioimmunoassay (RIA). However, the evaluatiion about specificity required in these two methods are quite different. In immunohistochemical staining, antisera which can never be used in RIA have been successefully employed. For example, anti-hCG was used for staining rat pituitary LH cells, and anti-mammalian LHRH was used for staining avian hypothalamus. On the other hand, very useful antisera for RIA are often eliminated in immunohistochemistry because of poor specificity. Such different evaluation is understandable if we compare the principles of these methods. The principle of RIA is the competitive binding, while immunohistochemistry based on the non-competitive binding where substances with very low affinities to the antibody can be bound if the concentration of antibody is satisfactorily high. In immunohistochemistry, the coexistence of antibodies to substances other than target antigen is very significant, while in RIA only the crossreactivity is important, and the competitive binding test gives no information about that. So, the evaluation of the antiserum in immunohistochemistry should be carried out using the non-competitive binding test which is based on the same principle to that of immunohistochemistry, and the results should be compared with those obtained with the competitive binding test which gives important information about the difference of the affinity constants of crossreacting substances.
Backscattered electron image (BEI) of colloidal gold-labeled antibody was compared with that of osmium which has chelated with the reaction product of peroxidase (HRP)-labeled antibody. Male rat hypophyses were fixed with 4% formaldehyde in 0.1M phosphate buffer, pH 7.4 and embedded in JB-4. Tissue sections at 2μm in thickness were placed on carbon coated glass slides and were reacted with rabbit anti-ACTH followed with either colloidal gold-labeled or HRP-labeled goat anti-rabbit IgG. Those reacted with HRP-labeled antibody were incubated with a solution containing diaminobenzidine (DAB) and H2O2. The DAB reaction product was chelated then with OsO4→thiocar-bohydrazide→OsO4. After washing and drying, the sections were observed with BEI mode of a JEOL T200 scanning electron microscope at 25kV accelerating voltage. BEI from the colloidal gold-labeled antibody was visualized as very fine granules in the cytoplasm of ACTH cells at high magnification (>3, 000×), but was difficult to recognize at low magnification (<300×). On the other hand, BEI from the osmium chelated with the reaction product of HRP could be recognized as amorphous masses in the cells both at the low and high magnifications. From these observations, it was concluded that the concomitant use of colloidal goldlabeled antibody and HRP-labeled antibody should be useful for the immunohistochemical localization of more than one antigen on the same section mounted on an ordinary glass slide at the electron microscopic level.
The location of human T-cell leukemia virus type I (HTLV-I) antigens in HTLV-I-producing cells, HUT 102 and MT-2, was described on the basis of immunoelectron microscopic studies and immunoblotting analysis. The direct and indirect peroxidase-labeled antibody methods with monoclonal and polyclonal antibodies to various HTLV-I antigens as well as sera from patient with adult T-cell leukemia were used for the immunoelectron microscopic observation. In HTLV-I-producing cells fixed with paraformaldehyde and incubated in cell suspension with sera from ATL patients, the viral envelope and plasma membrane are positively stained. In the specimens frozen-sectioned and incubated with the patient sera after fixation, positive staining is observed mainly on the viral envelope, plasma membrane, nuclear envelope, and endoplasmic reticulum of these cells. gag gene product p24 was localized in the viral core, and p19 was in the viral envelope and core and some parts of the plasma membrane. env gene product gp46x was deduced to be localized on the viral envelope and the plasma membrane, and env precursor protein gp62x was deduced to be in the endoplasmic reticulum and the nuclear envelope. pX gene product p40x was localized mainly in the euchromatin regions of the nuclei of HTLV-I-producing cells, but p68x of MT-2 cells was localized mainly in the nuclear envelope and the endoplasmic reticulum. The p68x detected in MT-2 cells with anti-p40x serum was deduced to be a fused protein consisting of a part of pX and env gene products and to share epitopes with p40x.
The aim of this study is to examine secretory mechanisms of the human chorionic gonadotropin (HCG) with immunocytochemical staining techniques. In 10 weeks and term placentae, HCG and its subuniuts were localized in the perinuclear space (PNS) and rough endoplasmic reticula (RER) of syncytial trophoblsts, but microvilli on the maternal cell surface were negative. In addition, 10 week old placenta demonstrated the presence of small vesicles and large lysosome-like granules in the cytoplasm facing the maternal circulations. They were stained by using anti-HCG β subunit antibody. The large lysosome-like granules were predominantly demonstrated to contain free β subunits. Term placentae did not show these large lysosome-like granules. These data supported the theory that HCG secretion mechanism is mediated by transportation via small vesicles in the first trimester. The lysosome-like granules may be the site for digestion of excess free subunit. In term placenta, HCG and its subunits were localized in PNS and RER indicating their actual production, in spite of low HCG β level in the maternal blood. These PNS and RER were more prominent in the cytoplasm facing fetal circulation and suggested the possibility of their secretion toward fetal circulation.
As part of a search for a new staining agent with higher selectivity for particular metals, we examined the staining properties of thiazolylazo-benzoic acid derivatives for histochemical staining of metals in rat liver tissue. Thiazolylazodimethylaminobenzoic acid (TAMB) stained only two metals, palladium(Pd) and copper(Cu), among a variety of metals. The results obtained revealed that the TAMB method was superior to the conventional p-dimethylaminobenzylidenerhodanine method with regard to selectivity for various metals when staining for Pd in histochemical practice. The differentiation of Pd from Cu was achieved by using masking agents, pyrophosphate or tripolyphosphate.
Immunocytochemical localization of glutathione-peroxidase (GSH-PO) in the ventral prostate was studied in rats under the following experimental conditions; 1) normal untreated 2) 4 weeks after castration 3) testosterone-administration to castrated, in order to clarify the role of GSH-PO. In normal untreated group, GSH-PO was predominantly observed in glandular epithelial cells of the ventral prostate. Intracellular localization of GSH-PO in the glandular epithelial cells was mainly observed in cytoplasmic matrix near the rough endoplasmic reticulum, and it was occasionally noted as a small granular structure at the supranuclear region. In the castrated animal, the intensity of GSH-PO staining in the glandular epithelial cells of atrophic prostate was markedly decreased. By testosterone-administration to the castrated animal, GSH-PO was clearly detected in the glandular epithelial cells of stimulated prostate. Intracellular localization was mainly observed in cytoplasmic matrix and the number of GSH-PO-positive granules increased remarkably as compared with those of other experimental groups. These findings strongly suggest that GSH-PO in the glandular epithelial cells of the rat ventral prostate is testosterone-dependent, and that its staining pattern is a useful marker for testosterone action.
Immunohistochemically detectable human epidermal growth factor (hEGF)/γ-urogastrone (γ-UG) was demonstrated with the use of polyclonal anti-hEGF/γ-UG antiserum in normal human submandibular glands and in their obstructive lesions. hEGF/γ-UG was isolated and purified from human urine and tested for its biological specificity by radioreceptor assay, and prepared anti-hEGF Fab' fraction. Histochemical expression of hEGF/γ-UG was confined to the intercalated and striated duct cells, and staining intensities differed depending on the fixations used. PLP fixed frozen sections showed the highest staining for hEGF/γ-UG in ductal segments, and 10% formalin and/or Bouin's fixed paraffin sections also revealed moderate-to-strong deposition. No immunoreactive hEGF/γ-UG was detected in either serous or mucous acinar compartments. In specimens of the obstructive sialoadenitis, hEGF/γ-UG staining of degenerated ductal epithelia was reduced, and was completely negative in fully degenerated cells.
The purpose of this study was to clarify histochemically the relation of ultrastructure and function in osteoclasts treated with calcitonin during experimental tooth movement. The upper first molars of Wistar strain rats were moved with a wire spring to obtain a number of active osteoclasts. These rats were divided into a control group (not treated with calcitonin) and seven experimental groups (treated with calcitonin). For demonstration of cytochrome c oxidase activity in mitochondria of the osteoclasts, the DAB method of Seligman et al. was used. The following results were obtained. 1) There was no remarkable difference in the localization of the enzyme activity of osteoclasts between undecalcified tissues and tissues decalcified by 5.0% EDTA solution. 2) Active osteoclasts in the control group contained many mitochondria with reaction products of the enzyme in the inner membrane and outer compartment (type 1 mitochondria, 82.6%/100μm2 in an osteoclast except nuclei) and a few mitochondria with reaction products only in the inner membrane of the cristae including intracristal spaces or lacking them at all sites (type 2 mitochondria, 17.4%). 3) At 3hr after calcitonin administration, maximal morphological changes of the osteoclasts with disappearance of a ruffled border and striking decrease of type 1 mitochondria (12.2%) were observed. 4) Osteoclasts of the experimental group at 72hr after calcitonin administration were almost typical active osteoclasts as seen in the control group and contained 80.2% type 1 mitochondria. These results suggest that cytochrome c oxidase activity was closely associated with changes in form and function of the osteoclasts.