Silver stains demonstrating neuroendocrine cells

During the preimmunohistochemical era, silver stains were an important part of the staining arsenal for identifying certain tissue structures and cell types in tissue sections. Some of them were useful for demonstrating endocrine cells, especially in the gastrointestinal tract. Until the late 1950s, silver stains, particularly those identifying endocrine cells, were accompanied by a number of technical difficulties resulting from uncontrolled staining factors. In the 1960s, new silver stains were developed for endocrine cell types and these stains gave reproducible results. One of the “older” silver stains and two of the “newer” ones are emphasized in this presentation, namely the Masson, the Grimelius and the Sevier-Munger techniques. The Masson stain demonstrates the enterochromaffin (EC, serotonin) cells, the Grimelius stain is a broad endocrine cell marker, and the Sevier-Munger technique demonstrates EC and EC-like cells and the C-cells of the thyroid. Especially in the preimmunohistochemical era, these staining methods often were used for histopathological diagnosis, particularly the Grimelius technique. The silver stains were developed empirically, and with few exceptions the chemical background is not known. Staining protocols are included.     

Application of silver stains to cytologic specimens of neuroendocrine tumors metastatic to the liver

Cytologic specimens of neuroendocrine tumors metastatic to the liver were examined with regard to their silver staining properties after the application of argentaffin and argyrophil staining techniques (Masson, Grimelius and Sevier-Munger). In tumors with a content of serotonin (small intestine carcinoids), the presence of this substance was demonstrated cytologically as an argentaffin reaction in individual tumor cells; however, formalin fixation was a prerequisite for positive staining. Melanin in malignant melanoma cells displayed a positive argentaffin reaction, irrespective of the fixation used (air drying, formalin, Bouin's fluid or acetone-alcohol). Thus, serotonin and melanin can be distinguished in cytologic samples of neuroendocrine tumors by the use of the Masson argentaffin reaction with different fixatives. The nonargentaffin-positive neuroendocrine tumor cells were weakly stained or unreactive with the Grimelius argyrophil technique. The Sevier-Munger argyrophil technique was negative or gave a disturbing nonspecific background staining reaction that was difficult to interpret in the cytologic samples. Thus, the Grimelius method appears to be the most useful silver stain for identifying neuroendocrine tumor cells in cytologic material, irrespective of their hormone content, since both argentaffin-positive and argentaffin-negative cell samples were stained at least to some degree.     

Silver stains in the study of endocrine cells of the gut and pancreas

The usefulness of argentaffin (Masson) and argyrophil (Davenport, Sevier-Munger and Grimelius) staining methods for identification of endocrine cell types in the gastrointestinal tract and pancreas is discussed and comments are made on the techniques themselves. The applicability of silver impregnation methods in the histopathological investigations of endocrine tumours in the above-mentioned organs is outlined, and the chemical background of the silver reactions is briefly reviewed.     

The endocrine cells of the rectum of adult ox.

In order to identify the endocrine cell types in various parts of the Ruminant gut, we have applied ultrastructural, both morphological and cytochemical, techniques, in parallel to the histochemical ones, to study the rectal mucosa of the adult Ox. In these studies we show that: "EC" cells, of the intestinal type, contain predominantly pleiomorphic granules, which are very electron dense and heavily reactive to "Masson" and "Grimelius" methods; "L" cells are recognizable by their numerous granules, which are fairly homogeneous in shape and osmiophilia. They do not react with "Masson" and are weak or negative to Grimelius s reaction. These granules occur near to others that are less dense, unreactive to "Masson", and that contain an argyrophilic matrix, with an eccentric electron dense core, which does not react with silver; "F-like" cells contain granules which are variable in shape, size and osmiophilia. They are unreactive to "Masson" and weak or unreactive to Grimelius silver; "H" cells contain few, small and uniformly osmiophilic granules. These are unreactive to "Masson" and uniformly reactive to "Grimelius". Our data suggest that the morphology, frequency and distribution of the cell types we have identified in the mucosa of the bovine rectum correspond with those reported in large intestine and rectum of Monogastrics, as by other authors described.     

On argyrophil reactions of endocrine cells in the human fetal pancreas

Summary The endocrine cells in the pancreas of five human fetuses with gestational ages of 18–20 weeks were examined by light and electron microscopy with special regard to argyrophil reactions. B-cells and typical A and D-cells were easily identified electron microscopically on the basis of their typical secretory granules. In the Grimelius argyrophil silver stain, a concentration of silver grains over the less electron dense peripheral mantle of the A-cell secretory granules was observed by electron microscopy. In the Hellerström and Hellman modification of the argyrophil Davenport alcoholic silver stain, silver grains were concentrated over the internal structures of the D-cell secretory granules. With this stain an accumulation of silver grains was also seen at the surface of the A-cell secretory granules. The argyrophil reaction of the A-granules was less pronounced than in the D-cells. In addition to B-cells and A- and D-cells, two other types of endocrine cell were observed by electron microscopy. These cells were argyrophil with the silver impregnation method of Grimelius. The electron microscopic findings at least partly explain the frequent overlapping between the two staining methods observed at the light microscope level.This study was supported by the Swedish Medical Research Council (Project No. 102)     

Immunocytochemical study of argyrophil (Sevier Munger) endocrine cells of adult human pancreas

Summary Bouin-fixed tissues from non-diabetic adult human pancreata display an argyrophil reaction mainly in the periphery of the islets with the silver technique of Sevier-Munger. The nature of these argyrophil cells was examined after restaining by an indirect immunocytochemical method using antibodies against insulin, glucagon, somatostatin and pancreatic polypeptide. After this procedure the argyrophil cells were identified as glucagon (A-) cells and pancreatic polypeptide (PP-) cells, although the latter exhibited a weaker reaction. The insulin (B-) cells and somatostatin (D-) cells were unreactive. The results show that the Sevier-Munger stain is of equal value to the Grimelius silver nitrate stain in adult human pancreatic islets after fixation in Bouin's fluid.Supported by grants from the Swedish Medical Research Council (Project No. 102)     

Etude critique des distinctions tinctoriales effectuées parmi les cellules endocrines de la muqueuse fundique du Lapin

Résumé Si chez des animaux intacts ou ayant reçu de la L-Dopa on compare les images d'une même coupe traitée d'abord selon Falck, ensuite colorée ou imprégnée selon divers procédés, on peut constater les faits suivants: la formaldéhyde confère une fluorescence jaune terne (Jt) à des cellules GIC dépourvues d'argyrophilie selon Sevier et Munger mais imprégnées en technique de Grimélius, colorées par le picroponceau et l'hématoxyline au plomb. Il y a approximativement trois fois plus de cellules imprégnées selon Grimélius que de cellules fluorescentes; parmi elles 1/3 environ réagit fortement en technique de Sevier et Munger (cellules à sérotonine), 1.3 réagit faiblement (cellules SM), 1/3 ne réagit pas du tout (cellules D et Jt). Les cellules SM correspondent aux «ECL cells» bien que dépourvues d'amine et d'argentaffinité (Fontana-Masson); elles se divisent en GIC et non GIC, les premières sont ou non colorables par l'hématoxyline au plomb, les secondes ne le sont pas.

---

Critical study of tinctorial distribution proposed among endocrine cells of the fundic mucosa in the rabbit

Summary Whether the same endocrine like cells are reactive towards different staining reactions has been hardly examined in normal and L-Dopa treated rabbits for distinction of cell types. Samples were first freeze-dried, exposed to formaldehyde vapours (standard Falck technic) and sections photographied. Thereafter they became further hydrated and were stained with silver according to Fontana-Masson or Grimelius or Sevier-Munger, or with ferric ferricyanide or lead hematoxylin or picroponceau.Formaldehyde condensation induced a bright yellow fluorescence in serotonin cells and revealed a dim yellow (jaune terne: Jt) unexpected fluorescence in others scarce cells. The laters were non argentaffin, unreactive with Sevier-Munger method but argyrophilic in Grimelius technic, stained blue black by lead hematoxylin and red by picroponceau; there are a part of the GIC or APUD cells group.Grimelius reaction stained three times as many cells as fluorescence did. With Sevier-Munger method, heavily darkened (serotonin cells), slightly reactive (Sevier-Munger or SM cells) and unreactive cells (D and Jt cells) roughly represented 1/3 each of all Grimelius positive cells. ECL cells according to Capella et al. (1969) apparently corresponded to SM cells; since all fluorescent cells are strongly argentaffin and reactive with Sevier-Munger method, since they are too much scanty to include EC and ECL cells it is conclude that ECL cells failed to exhibit fluorescence and to store serotonin: furthermore taking hypothetically ECL cells into fluorescent cells it would imply to blend D cells and SM cells which is at variance with earlier observations of unreactivity of D cells in Sevier-Munger technic at the ultrastructural level. SM cells population seemed heterogenous: only some of them were GIC and only some of the laters were stained by lead hematoxylin; however it is not clear how much the L-Dopa administration effect the second staining possibly causing erroneous reactions.

---

---

Avec la collaboration technique de Mme A. de Gouyon.     

Immunocytochemical study of argyrophil (Sevier Munger) endocrine cells of adult human pancreas

Bouin-fixed tissues from non-diabetic adult human pancreata display an argyrophil reaction mainly in the periphery of the islets with the silver technique of Sevier-Munger. The nature of these argyrophil cells was examined after restaining by an indirect immunocytochemical method using antibodies against insulin, glucagon, somatostatin and pancreatic polypeptide. After this procedure the argyrophil cells were identified as glucagon (A-) cells and pancreatic polypeptide (PP-) cells, although the latter exhibited a weaker reaction. The insulin (B-) cells and somatostatin (D-) cells were unreactive. The results show that the Seiver-Munger stain is of equal value to the Grimelius silver nitrate stain in adult human pancreatic islets after fixation in Bouin's fluid.     

Effect of Temperature on Argyrophil Impregnation: Development of A High Temperature Rapid Argyrophil Procedure

Out studies on the effects of temperature on the demonstration of neurosecretory granules using argyrophil stains indicate an inverse relationship between the time needed for staining and temperature of the silver and reducing solutions. Careful monitoring of the temperature of silver solutions during the Grimelius procedure and its modifications show long incubation times serve in large part only to bring the solutions to reaction temperature. Tissue sections added when this temperature has been reached will stain with the same intensity as sections impregnated for the entire incubation period. We have modified the argyrophil procedure so that double-impregnation with solutions preheated to 60-70 C and development in Bodian's reducer prepared with preheated water rapidly demonstrates secretory granules. Our method does not require a microwave oven and much shorter incubation periods are required than with usual procedures. It is not necessary to incubate sections in hot solutions for extended periods of time, which can lead to detachment of sections, nonspecific staining and decomposition of the silver solution. Rinsing after impregnation and before development greatly increases contrast of argyrophil cells by reducing background staining. Our procedure results in more reliable staining of argyrophil and argentaffin cells and takes only ten minutes.     

Endocrine cells of the gastric mucosa of Rana temporaria L

The endocrine cells of the gastric mucosa of Rana temporaria have been studied according to the ultrastructure, the staining properties of the granules with Masson Fontana's and Grimelius' silver methods, silver impregnation of Davenport on deplasticised semithin sections and immunocytochemical techniques. Seven different types of endocrine cells have been described. Six were regarded as belonging to known types: G, A, EC, ECL, D and P cells. One type was considered as unclassifiable.     

Effect of temperature on argyrophil impregnation: development of a high temperature rapid argyrophil procedure

Our studies on the effects of temperature on the demonstration of neurosecretory granules using argyrophil stains indicate an inverse relationship between the time needed for staining and temperature of the silver and reducing solutions. Careful monitoring of the temperature of silver solutions during the Grimelius procedure and its modifications show long incubation times serve in large part only to bring the solutions to reaction temperature. Tissue sections added when this temperature has been reached will stain with the same intensity as sections impregnated for the entire incubation period. We have modified the argyrophil procedure so that double-impregnation with solutions preheated to 60-70 C and development in Bodian's reducer prepared with preheated water rapidly demonstrates secretory granules. Our method does not require a microwave oven and much shorter incubation periods are required than with usual procedures. It is not necessary to incubate sections in hot solutions for extended periods of time, which can lead to detachment of sections, nonspecific staining and decomposition of the silver solution. Rinsing after impregnation and before development greatly increases contrast of argyrophil cells by reducing background staining. Our procedure results in more reliable staining of argyrophil and argentaffin cells and takes only ten minutes.     

Stains for A, B, and D cells in fetal rat islets.

Ten techniques often used for identification of A, B, and D cells in adult islets of Langerhans were applied to fetal rat pancreas. Modifications were tried with many of these techniques. Two indole methods (xanthydrol and postocoupled benxylidene reactions) and a cryostat technique using o-phthaladehyde failed to stain fetal islets. Phosphotungstic acid hematoxylin and lead hematoxylin lightly stained fetal A cell granules in Helly's fixed tissue. The Grimelius silver nitrate technique stains adult rat A cells but failed to stain fetal cells. A modification of this technique stained fetal A cells and a possible 4th cell type. The specificity of this method was confirmed by restaining stained cells with a fluorescent antibody technique and with pseudoisocyanin. B cells, as previously reported, were readily stained by the aldehyde fuchsin technique. Fetal D cells were not stained by the Hellerstrom-Hellman alcoholic silver nitrate method, nor did they display pseudoisocyanin metachromasia after acid hydrolysis; they did fluoresce brightly with this technique when viewed with UV light. It was thus possible to distinguish the three usual cell types, plus a possible fourth type, in the fetal rat pancreas.     

Silver stains for identification of neuroendocrine cells. A study of the chemical background

Summary The chemical background of silver stains used for visualization and characterization of peripheral neuroendocrine cells in the gastrointestinal tract and pancreas, and of their corresponding tumours, was studied in tissue sections and by a dot-blot technique. Sequential staining of pancreatic islets with an immunohistochemical procedure and silver staining of the same tissue section revealed that chromogranin A immunostained cells also displayed an argyrophil reaction with the Grimelius method, but no argentaffin reaction with the Masson technique. Accordingly, purified chromogranin A (15 g or less) treated in formalin and applied to nitrocellulose did not show any argentaffin reaction but displayed a dose-related argyrophil reaction. Equal quantities of other polypeptide components did not give rise to any silver reaction. Further dot-blot studies showed that the tryptophan and tyrosine metabolites, dopamine, norepinephrine, 5-hydroxytryptamine and 5-hydroxindole caused strongly argentaffin and argyrophil reactions while epinephrine, 5-hydroxyindole-3-acetic acid and 5-hydroxytryptophan gave only the former reaction. Among other chemical components studied, only guanine displayed weak silver staining. The results indicate that the reaction products between aldehydes and the granular content of biogenic amines synthesized from tryptophan and tryosine display an argentaffin reaction and that the granular chromogranin A caused an argyrophil but no argentaffin reaction.     

Localization of chromogranin A-immunoreactivity in bovine gastrointestinal endocrine cells with special reference to Grimelius silver stain.

Immunohistochemical studies of the gastrointestinal tract were carried out to characterize the cells exhibiting immunoreactivity for chromogranin A (CGA), a glycosylated protein primarily found in secretory granules of the adrenal medulla. Double immunostaining for gastrointestinal hormones and CGA revealed that in the bovine gastrointestinal tract CGA immunoreactivity occurs in mucosal epithelial cells containing gastrin, glucagon, substance P or motilin, but not in those containing somatostatin. Combined staining with anti-CGA serum and Grimelius' silver demonstrated frequent association of the two stains in a variety of endocrine cells. However, intracellular distribution of the two stains was different: CGA-immunoreactivity was detected in both supra- and infranuclear cytoplasm, whereas Grimelius' silver was mostly localized in the infranuclear region. These results suggest that CGA is the target of Grimelius' silver, as postulated recently (Rindi et al., 1986), but that some subcellular structure-related modification of molecules such as sialation is necessary for the positive Grimelius reaction.     

Endogenous prostaglandins are physiological regulators of endocrine cells in the gastroduodenal mucosa of the rat.

BACKGROUND/AIM: To investigate whether endogenous prostaglandins participate in the regulation of the gastrointestinal endocrine cell system. METHODS: Sprague-Dawley rats were treated with 1 mg/kg indomethacin subcutaneously or indomethacin subcutaneously and 500 microg/kg oral prostaglandin E2 or solvents for 2 months. Endocrine cells were visualized by using immunohistochemistry and by the Sevier-Munger silver stain on specimens from the gastroduodenal mucosa, and their total volume was estimated, using standard stereological methods. Plasma and gastrointestinal tissue concentrations of regulatory peptides were analyzed by radioimmunoassay. RESULTS: Fundic mucosa. The total volume of cells stained with the Sevier-Munger silver stain (enterochromaffin-like) was increased by indomethacin, but reduced by the administration of prostaglandin E2 (P < 0.05 vs. indomethacin). Indomethacin increased the total volume of somatostatin-immunoreactive. Similarly, rats given indomethacin and prostaglandin E2 had higher values than controls. Indomethacin increased the tissue concentration of somatostatin in the gastric fundus whereas prostaglandin E2 prevented such changes (P < 0.05 vs. indomethacin). Antral mucosa. The total volume of serotonin-immunoreactive cells was reduced by indomethacin, but increased by prostaglandin E2 (P < 0.05 vs. controls and indomethacin, respectively). Duodenal mucosa. The total volume of somatostatin-immunoreactive cells was reduced in the rats given indomethacin and prostaglandin E2 (P < 0.05 vs. controls and indomethacin). Indomethacin reduced and simultaneous administration of prostaglandin E2 increased the total volume of CCK-immunoreactive cells (P < 0.05 vs. controls and indomethacin). Indomethacin reduced the total volume of serotonin-immunoreactive cells whereas the simultaneous administration of PGE2 comparatively increased their total volumes (P < 0.05 vs. indomethacin), although they were still lower than the control values. The total volume of GIP-immunoreactive cells was slightly increased in the rats given both indomethacin and indomethacin + prostaglandin E2. The tissue concentration of somatostatin in the duodenum was reduced in rats given indometacin and prostaglandin E2 (P < 0.05 vs. controls and indomethacin). CONCLUSION: Endogenous prostaglandins, particularly prostaglandin E2, regulate CCK-, enterochromaffin-like-, somatostatin-, GIP- and enterochromaffin cells in the gastroduodenal mucosa of the rat.     

Ultrastructural identification and immunocytochemical study of somatostatin cells in antral mucosa of the rabbit and the mouse (author's transl)]

In normal and L-Dopa treated rabbits and mice, combined immunochemical methods, photonic histological methods for endocrine cells and ultrastructural methods were used to elucidate ultrastructure and properties of somatostatin cells of the antral mucosa. In normal rabbits, immunoreactive cells giving no fluorescence with Falck's technic, they corresponded neither to serotonin cells nor gastrin cells; they were unreactive with Fontana, Hellerstr?m-Hellmann, Sevier-Munger and Mac Conaill methods but very slightly stained with Grimelius methods. In L-Dopa treated animals somatostatin cells gave formaldehyde induced fluorescence (they were included in GIC cells, thus in Apud group), exhibited a good reaction with Grimelius and Sevier-Munger methods. In order to carry out the alternate semi-thin/thin section procedure (semi-thin sections for immunofluorescence or immunoenzymatic detection and serial thin sections counter-stained for conventional ultrastructure studies), immunological treatment were performed on M.F.F.--glutaraldehyde fixed small fragments of mucosa before inclusion in Epon 812 or, after inclusion, on semi-thin sections. We succeeded in identifying ultrastructurally somatostatin cells. They displayed round or ovo?d shaped secretory granules, and three constant typical structures: numerous microfilaments--light and homogenous granules, often seeming like lipids---granules made up by coarsely filamentous cores surrounded by a large empty halo. Somatostatin cells seemed different of X cells because of their predominant localisation in the antral mucosa (in the rabbit X cells were predominantly in the fundus) and because of the lack of nuclear microfilaments; they also seemed ultrastructuraly different of D1 cells.     

Some Introductory Comments on Silver Staining

Silver staining has become a versatile method for the visualization of specific cell structures and products. The similarity of the impregnation “nuclei” of reduced silver staining to the silver “specks” or “nuclei” of the latent image in photography is noted. “Physical” development (reduction of ionic silver in solution) in silver staining as compared to “chemical” development (reduction of ionic silver remaining in a silver halide crystal) in photographic procedures is briefly discussed.     

The Application of Dyes in the Cancer Problem

Three fundamental requirements for the problem of developing a differential stain for cancer are discussed: I. the choice of a technic for the microscopic preparation of tissues; II. an analysis of the biological properties peculiar to cancer; and III. various groups of dyes adaptable to such peculiar properties of cancer tissue. Under I the disadvantages of intravitam staining are pointed out and the use of cell suspensions, frozen sections, and fixed material favored. Under II three characteristics of cancer tissue offering possibilities for differential staining are discussed, the cytological structure known as the “plastin reaction”, the histogenic cycle of cancer tissue, and the viability of cancer tissue under anaerobic conditions. Under III modifications of the Giemsa stain are suggested for application to the plastin reaction, specific tissue stains advocated for the use of indicating end points in histogenic cycles, and the vital dyes, congo red and trypan blue, suggested as indicators for the survival of malignant tissues because of the failure of these dyes to permeate living cancer cells.

The angle of approach thruout has been an attempt to avoid unconscious pitfalls inherent in certain microscopic technics, and to substitute analytical methods for the blind trial and error method of routinely applying dye after dye in endless succession.     

The endocrine cells of the pyloric glands of adult ox

As part of a project to identify the endocrine cells ("EC" and "APUD" series) of the gastroenteric apparatus of ruminants, the ultrastructure of the mucosa of the pyloric glands of adult ox was studied morphologically and cytochemically, in parallel with a light microscope histochemical analysis. The results show that: the "EC" cells (producing 5-HT) are recognizable by their secretory granules which are heavily osmiophilic, argentaffin ("Masson") and argyrophilic ("Grimelius"). A further distinction is possible on the basis of their morphological features: the "EC" cells of the gastric type (which belong to the "ECn" group) contain granules fairly homogeneous in shape and size, while the "EC" cells of the intestinal type (or "EC1") show granules which are more pleiomorphic and variable in size. Of particular interest is the presence in some cells of granules typical of the "EC" cells of the intestinal type, in the vicinity of a few others, which appear quite similar to those of the adjoining exocrine cells; the "G" cells (gastrin producing) contain medium sized granules, which are unreactive to "Masson" and poorly argyrophilic. Their morphology is rather diverse; some of them (these are the "typical" cells) have a granular and weakly electron dense content, others (which we consider "atypical") show a homogeneous and heavily osmiophilic core, with an eccentrical empty area. Also present are granules whose appearance is intermediate and empty vesicles; the "D cells" (somatostatin producting) show round, medium sized granules which have a granular, moderately osmiophilic core, tightly encircled by the membrane. These granules are unreactive to "Masson" and to "Grimelius"; the "D1" cells (whose function is yet unclear) contain small, round granules whose core is variously but discretely electron dense and not always homogeneous; they are unreactive to "Masson" and fairly argyrophilic. These granules may be numerous and packed, or scarce; in this latter instance the few granules are intermingled with variously running tufts of parallel filaments, thus resembling the "P" cells, whose function is still undefined. These data show therefore that the types of endocrine cells we have identified in the pyloric glands of adult ox correspond to those described in other mammals; "X" and "F" or "PP" cell appear to be lacking.     

Double simultaneous staining of B and argyrophil cells of the pancreatic islets: a sequential thiosulphation-aldehyde fuchsin and Grimelius silver procedure

The thiosulphation-aldehyde fuchsin (TAF) method for the insulin-producing B cells can be followed by the Grimelius silver impregnation for the argyrophil cells. This double staining is useful to study, in normal and pathological tissues, the spatial distribution of the two main endocrine cell populations of the pancreatic islets. A treatment with potassium ferrocyanide has been found to enhance the argyrophilia of A cells.