Journal of Histochemistry & Cytochemistry

Rabbits were primed intravenously with human serum albumin (HSA) and boosted with the same antigen 2 months later. Cells producing specific antibodies against HSA could be detected in vivo and it could be determined whether or not they belonged to the immunoglobulin (Ig) G class using a combined peroxidase (HRP) and alkaline phosphatase (AP) immunocytochemical technique. HRP-HSA conjugate was used for detection of anti-HSA-producing cells and AP-sheep anti-rabbit IgG (SRIgG) was used to determine the IgG class of the antibodies produced by these cells in the same spleen section. After performing both HRP and AP cytochemistry, cells with a red-stained cytoplasm represent anti-HSA-producing cells not stained for their antibody class and cells with a blue-stained cytoplasm represent cells producing IgG antibodies not directed against HSA. Cells with a double-stained cytoplasm represent cells producing anti-HSA antibodies belonging to the IgG class. We also attempted to determine whether or not part of the anti-HSA-producing cells belonged to the IgM class using AP-sheep anti-rabbit IgM (SRIgM). In this case no double-stained cells were detected, indicating that the affinity of intracellular IgM-anti-HSA antibodies is too low to allow detection using the present technique.

Labeled lectins with binding specificity to the hexose components of mucus glycoproteins (HPA, RCA I, PNA, Con A, WGA, and UEA I) were used to demonstrate structural differences in the glycoprotein composition of various cell types of the normal, benign and malignant gastrointestinal mucosa. While in the RCA I, UEA I, and WGA binding of normal mucus secreting cell types only quantitative differences were observed, the mucus in the surface epithelial cells of gastric mucosa and in the colonic goblet cells was characterized by the absence of PNA, Con A, and PNA, HPA binding sites, respectively. These lectins, however, showed a strong binding to the supranuclear, Golgi-region in the undifferentiated or activated forms of these cells. Even the staining intensity of the luminal membrane surfaces of the non mucinous parietal and chief cells was often stronger by PNA, HPA, and RCA I lectins than that of the mucus secretions in the highly differentiated mucus cells. These results indicate the existence of either heterogeneous glycoprotein components or mucus molecules with variations in the degree of glycosylation of their oligosaccharide chains in the different cells. The latter seems more likely since in benign and malignant alterations lectin binding sites appear in great density, which were found to be characteristic of the undifferentiated mucus cells or for the non mucinous cells of the normal gastric mucosa. Similarly in some gastric cancers which do not stain with the periodic acid-Schiff reaction at all, large amount of free or neuraminic acid substituted PNA binding sites can be detected.

Using affinity chromatography on HPA-, PNA-, Con A, and WGA-agarose columns only a part (10-30%) of the high molecular weight mucous glycoproteins could be isolated from the Triton X-100 solubilized components of normal as well as carcinomatous gastric mucosa. The main part of the mucus was not bound by the lectins, which corresponds to our earlier lectin histochemical observations on paraffin-embedded tissue sections. The lectin-bound mucous glycoproteins had a relatively lower molecular weight, ranging from about 250-1,000 kilodaltons, as indicated by polyacrylamide gradient gel electrophoresis and by gel filtration on Biogel A 1.5 m column. In gas chromatographic analysis the molar ratio of aminohexoses to galactose was found to be much higher (3:1) in the lectin-bound mucous substances than in the whole high molecular weight mucus (1:1). This finding indicates that lectins have a higher affinity to the hexosamine rich components of mucus, which may be special forms of mucous glycoprotein molecules or the incompletely glycosylated core and backbone regions of the oligosaccharide chains of mucus. Extremely high hexosamine values (10:1) were found in the PNA isolated mucus of gastric adenocarcinoma. Since it is known that PNA binds to the terminal disaccharide, beta-galactose-(1-3)-N-acetylgalactosamine, which is localized at the reducing end of the oligosaccharide chains of mucus, it is highly probable that the elongation of the oligosaccharide side chains is disturbed in gastric cancer cells.

The protein A-gold immunocytochemical technique was applied to the localization of vitellogenin in the hepatocyte of the bullfrog, Rana catesbeiana, eight days after treatment with estradiol-17 beta. Specific labeling was present in cellular compartments involved in protein secretion and was shown to progress in sequence through RER, Golgi apparatus, immature secretory granules, and mature secretory granules. Labeling intensities were quantitated and the values ranged from 34.6 to 172 gold particles/micron 2. In contrast, low background labeling was observed over mitochondria, nuclei, lipid droplets, and bile canaliculi. These observations support the hypothesis that vitellogenin synthesis and secretion in the frog hepatocyte lies exclusively along the RER-Golgi-granule secretory pathway. In addition to the cellular compartments involved in protein secretion, labeling was found over the majority of the lysosomes. The intensity of lysosomal labeling was intermediate between that of RER and Golgi apparatus. This labeling of lysosomes may be a consequence of the high blood plasma concentrations of vitellogenin that occur in the frog model, or to the well-known crinophagy phenomenon present in secretory cells.

The value of the freeze-substitution (FS) method for preparing tissues for electron microscopic immunohistochemistry was studied by comparing anterior pituitary cells prepared by this method and by a conventional method. Ultrathin sections of tissues embedded in Epon were subjected to immunostaining. The antigens adrenocorticotropin (ACTH) and prolactin (PRL) in a single ultrathin section were demonstrated by a simple double-labeling technique using a protein A-colloidal gold-antibody (pAG-Ab) complex. The preservation of cellular ultrastructure was superior in preparations obtained by FS. Gold-labeling was seen over secretory granules, and in ACTH cells also over the cytoplasmic matrix. The labeling was more intense in preparations obtained by FS, judging from the numbers of gold particles. In the double-labeling procedure, in which the pA-small colloidal gold-anti-PRL complex and pA-large colloidal gold-anti-ACTH complex were applied sequentially to sections, no cross-labeling with small and large gold particles was observed. It is concluded that if the antisera are sufficiently specific, the use of FS and the pAG-Ab complex is very effective in peptide immunohistochemistry. However, in the double-labeling procedure it is essential that the Fc-binding sites of pAG are saturated by the use of excess amounts of antibodies.

A visual test for detection of granulocyte surface markers using the avidin-biotin complex (ABC) has been developed. That this assay is highly specific, reproducible, and sensitive was determined by studying the expression of HLA antigens on granulocytes with monoclonal antibodies. Further, using granulocyte specific alloantisera, the results of the ABC test compared well to data from leukoagglutination assays and indirect immunofluorescence tests. The assay is particularly advantageous in that granulocytes can be stored, only small amounts of cells and sera are needed, and heterogeneous cell populations can easily be studied.

The tissue distribution of rat alpha 1-microglobulin (alpha 1-m) was studied by indirect immunofluorescence in various rat tissues using a polyvalent rabbit antiserum to the purified antigen and a monoclonal antibody (H23) to the human homologue, in parallel with a polyclonal anti-rat IgA antiserum. It was found that all tissues stained by anti-IgA were also alpha 1-m positive; these tissues included tissues of the stomach, duodenum, ileum, colon, pancreas, trachea, esophagus and jejunum. However, the observation that IgA plasma cells as well as secretory cells, while positively stained by anti-IgA, are alpha 1-m negative suggests that the association between IgA and alpha 1-m occurs at a postsecretory stage, after the IgA molecules have been transported across the epithelial cells. Additionally, hepatocytes were intensely stained by anti-alpha 1-m antibodies, indicating that the liver, as already suggested by metabolic studies on isolated guinea-pig liver explants, may be responsible for the synthesis of this protein. Among lymphoid tissues, an intense and homogeneous staining was observed in the thymus and the white pulp of the spleen. Sections of lymph nodes, however, showed differential staining; apart from a few isolated dendritic cells in the mantle region of the lymphoid follicles, the germinal centers and medullary cords showed no staining with anti-alpha 1-m antibodies. The paracortical cells, macrophages in the subcapsular sinus, and interfollicular lymphocytes showed intense cytoplasmic staining with anti-alpha 1-m antibodies. In other tissues, macrophages, monocytes, tissue histiocytes, and dendritic cells were alpha 1-m positive. Although they confirm the presence of alpha 1-m in the lymphoid tissues, as already reported in man, these results show that the protein is also present in hepatocytes and in exocrine fluids containing IgA. Since alpha 1-m, like secretory component, can bind to IgA to form stable complexes, these two heavily glycosylated proteins may have similar biologic properties.

Specific iron-binding sites in the gut were visualized by the sequential incubation of glutaraldehyde-fixed specimens in iron nitrilotriacetate (FeNTA) and acid ferro-cyanide (AF). In rat duodenum, jejunum, and ileum, approximately half of the enterocytes from the crypt to the distal villus contained FeNTA-AF-reactive material. The staining intensity of individual enterocytes appeared to decrease progressively in more distal locations in the gut. Maximal FeNTA-AF staining was observed in cells in the upper half of the villus, and was localized primarily in microvilli, apical cytoplasm, and lateral membranes. In duodenal crypt cells, stain deposits were present primarily in the microvilli. FeNTA-AF stained sites in the cytoplasmic and microvillus matrix were approximately 100-fold greater in number than were sites of intrinsic iron stained with AF alone. FeNTA-AF and AF staining in human duodenal enterocytes was similar to that observed in rat duodena, demonstrating the applicability of this methodology to human samples. In rat duodena, the distribution of AF staining alone in specimens taken 10 min after the in vivo intraduodenal administration of FeCl2 was similar to that of FeNTA-AF staining in tissue from fasted animals not given iron. The distribution and frequency of iron-binding sites stainable with the FeNTA-AF method which occur in a subpopulation of enterocytes can be correlated closely with physiologic, histologic, and ultrastructural parameters of inorganic iron absorption previously reported.

Using quantitative methods, citrate synthase (CS), fumarase, beta-hydroxyacyl-coenzyme A (CoA) dehydrogenase (beta OAC), 3-keto-acid CoA transferase (KCT), malic dehydrogenase (MDH), and malic enzyme were measured in seven defined parts of the nephron and in thin limb and papilla areas dissected from freeze-dried microtome sections of rat kidney. The results not only show a wide range of activity along the nephron for each of the enzymes, but that the proportions between the enzymes vary markedly among the different parts of the nephron. This suggests the existence of major regional differences in the capacity to oxidize specific metabolites. The ratio between two citrate cycle enzymes, fumarase and CS, was 4- or 5-fold higher in proximal segments than in the glomerulus or thin limb areas. The ratio between beta OAC (an enzyme of fatty acid oxidation) and CS was 3- to 5-fold higher in the middle proximal segments than in glomeruli or thin limb and papilla areas. The key enzyme for ketone body metabolism, KCT, was essentially confined to the thick tubule segments. Malic enzyme, in contrast to the other five enzymes, was highest in the proximal straight segments. New methods, sufficiently sensitive for this histochemical study, are described for malic enzyme and 3-keto-acid CoA transferase.

The application of a morphometric method to the quantification of peptide-containing nerves in the gut is described. It allows a simple estimation of the nerve fiber supply per unit volume of tissue (length density) and the calculation of the total nerve fiber supply per unit of intestinal mucosa (villus/crypt unit).

The authors describe a system for the quantitation of fluorescence light output by individual cells using the signal obtained from a silicon intensifier target video camera. The video image is digitized to 4 bits (16 levels), and a 512 X 512 matrix is constructed and stored in 128K of video memory. Areas to be measured are user-specified by means of a light pen entry system. Recorded intensities are integrated under microprocessor control to provide a measure of total fluorescence in the selected areas. The distribution of light intensities per pixel over the 16-level gray scale as well as morphometric data are also obtainable. Linear response to transmission and fluorescence standards was verified. Reproducibility of the system was evaluated using fluorescent beads and glutaraldehyde-fixed chick red blood cells, which gave coefficients of variation comparable to those obtainable from other systems. Measurements of DNA per nucleus of human diploid fibroblasts using Hoechst dye 33258 yielded the two sharp peaks corresponding to the 2C and 4C values of DNA expected from such cells. We conclude that digital video measurement of fluorescence provides meaningful data and has considerable promise as a sensitive tool for the quantitation of fluorescence at the cellular level.

In conjunction with the immunoperoxidase and the immunoferritin methods, antielastin antibody was used to study the localization of elastin in untreated and elastase-treated elastic fibers of the nuchal ligament and the aorta of fetal and young adult sheep. In tissues not treated with elastase, the staining reaction for antielastin antibody was localized in the outer zones of the amorphous components and along the surfaces of the microfibrils ; the central zones of the amorphous components were unreactive. After mild elastase treatment, incompletely digested amorphous components showed staining both in their central and outer zones, and some of the microfibrils became unreactive. After extensive elastase treatment, small scattered amorphous components were still found in association with bundles of microfibrils. These components were stained diffusely by the antielastin antibody method but were not detectable by staining with uranyl acetate and lead citrate or with Kajikawa 's method for elastin; elastin was not detected on the surfaces of the microfibrils by any of the methods used. These findings were interpreted as indicating that the surfaces of the microfibrils are associated with small amounts of elastin, and that evenly stained amorphous components are composed of elastin, which is loosely arranged and allows the penetration of antielastin antibody. These observations support the concept that microfibrils serve an important role as a scaffold for elastin deposition in elastogenesis. Because of their high sensitivity, immunohistochemical methods for detecting elastin are useful to study partially degraded elastic fibers.

A cytochemical technique for the ultrastructural localization of unique nuclear antigens is reported. Using a post-embedding indirect immunogold labeling procedure, nuclear antigens in electron-dense regions of the nucleus are localized with a minimum of nonspecific staining. Using this technique and indirect immunofluorescence, a panel of antinuclear monoclonal antibodies is shown to recognize preferentially cell cycle-dependent nuclear substructures. The antigenic domains recognized include specific regions in condensed chromatin, interchromatin granules, euchromatin, and chromosomes. The specificity of antigen recognition is demonstrated with qualitative and quantitative immunogold electron microscopy and immunoblot analysis. These results reveal the existence of previously undefined supramolecular organization within the nucleus and demonstrate the utility of the immunogold procedure when monoclonal antibodies are used.

Immunocytochemical localization of tyrosine hydroxylase (TH) was performed on paraffin sections pretreated with various proteolytic enzymes. It was found that pretreatment with trypsin (1.2 mg/ml) for 5 min resulted in a dramatic increase in the number of TH-positive terminals throughout the brain, especially in the cerebellum, which contains fine preterminal and terminal axons that are difficult to stain. This pretreatment also led to a significant reduction in background staining and allowed for the use of the TH antiserum at high working dilutions. Several other proteolytic enzymes were tested and only chymotrypsin was nearly as effective as trypsin with respect to TH staining.

Use of unfixed fresh frozen tissue sections for immunocytochemical studies reduces the possibility of denaturation of antigenic determinants compared to formalin fixation and paraffin embedding procedures. However, tissue and cellular morphology can be extensively altered in the numerous application and washing steps with frozen tissue sections. We tested a number of buffer solutions and showed that the use of dextran-containing buffers and fixation by glutaraldehyde after primary antibody application preserves tissue morphology. The procedures described here are also applicable to ascertaining the presence of Fc receptors of leukocytes in sections of carcinoma tissues. The buffered dextran washes and post-primary antibody fixation method was used to demonstrate the presence of immunoglobulin associated with squamous carcinoma cells. The immunoglobulin was not removed by washing of tissue sections at 37 degrees C but could be removed by low or high pH buffer washes, suggesting that the immunoglobulin is bound in a specific manner.

Affinity-purified goat anti-rabbit immunoglobulin G (GAR) was conjugated with (3H)-propionyl succinimidate and used to localize substance P (SP), enkephalin (ENK), and serotonin immunoreactive sites in the spinal dorsal horn and medulla of the rat and cat. Autoradiographic localization was demonstrated on paraffin, frozen, Vibratome, and 2 micron plastic sections. The latter were obtained from radiolabeled Vibratome sections that were embedded in epoxy resin. The distribution of SP, ENK, and serotonin demonstrated by radioimmunocytochemistry was comparable to that observed on semiadjacent sections using peroxidase-antiperoxidase (PAP) immunocytochemistry. The autoradiograms, however, were generated using primary antibody concentrations up to five times more dilute than concentrations used for the PAP procedure. Indirect radioimmunocytochemistry using a (3H) anti-immunoglobulin G second antibody can be used to localize a variety of monoclonal and polyclonal antisera. It is quantifiable at the light microscopic level and can be potentially used with peroxidase histochemistry to double label immunoreactive structures at the ultrastructural level.

A method is described for immunohistochemical demonstration of purine nucleoside phosphorylase (PNP: EC 2.4.2.1) in paraffin sections from routine surgical histology specimens. A peroxidase-antiperoxidase (PAP) method was employed, using specific rabbit antiserum against human PNP, which was purified from postmature human erythrocytes. In human lymph nodes, intensive staining for PNP was observed in the vast majority of small lymphocytes in paracortical areas, in many small lymphocytes in medullary cords, and in a few small-to medium-sized lymphocytes in germinal centers. Small lymphocytes in the primary follicles and those in the mantle zones of secondary follicles were negative for PNP staining. Tingible body macrophages, lymphatic sinus cells, and most of the large cells in germinal centers did not stain with anti-human PNP (hPNP) antibody. Endothelial cells of small vessels in the cortex and plasma cells did not show any constant pattern of PNP staining intensity. Histochemistry revealed that the distribution pattern of PNP activity was quite similar to that demonstrated on paraffin sections by the PAP method.

The specificity of the combined oxalate-pyroantimonate (OPA) technique for subcellular localization of calcium was examined by laser microprobe mass analysis (LAMMA) of the same tissue sections. The recorded spectra strongly suggest that the cytochemically detected precipitates contained calcium. This could be confirmed by LAMMA analysis of ethylene glycol tetraacetic acid-treated sections. The detected calcium was localized mainly in the rod outer segments, more particularly in the middle part. The validity of the OPA and LAMMA methods is discussed. A combination of both techniques is found to be a valuable tool to elucidate the role of calcium in physiological mechanisms.