glycogen in pancreatic islets of steroid diabetic rats

Histochemistry (1981) 73:225 232 Histochemistry �9 Springer-Verlag 1981

Glycogen in Pancreatic Islets of Steroid Diabetic Rats

Carbohydrate Histochemical Detection and Localization Using an Immunocytochemical Technique *

R. Graf** and C. Klessen Institute of Anatomy of the University of Tiibingen, Osterbergstrasse 3, D-7400 Tiibingen, Federal Republic of Germany

Summary. In the islets of the rat pancreas, steroid diabetes induced by triamcinolon-acetonid leads to degranulation of the B cells and glycogen infiltration. The glycogen cannot be satisfactorily detected using methods like the chromic acid technique according to Bauer, staining with Best's carmine, or the usually applied periodic acid-Schiff (PAS) reaction. Glycogen detection is improved, however, when lead tetraacetate is used in place of periodic acid as oxidizing agent. When combining the carbohydrate detection method with the peroxidase - antiperoxidase (PAP) method used for immunocytochemical detection of the various pancreatic islet hormones, paraffin sections reveal that glycogen is primarily localized in granulated B cells; the degranulated B cells also contain glycogen, though in smaller amounts. In contrast, the islet cells containing somatostatin, glucagon and pancreatic polypeptide are nearly free of glycogen.

Introduction

In the course of spontaneous or experimentally induced diabetes mellitus, glycogen is stored in the islands of Langerhans in man and other mammals (Toreson 1951; Lazarus and Volk 1958; Fagundes and Cohen 1966). Glycogen was detected in the pancreatic islets of the rat after administration of a synthetic corticosteroid (triamcinolon-acetonid) during an investigation of the genesis of experimentally-induced steroid diabetes (Klessen and Leitz, unpublished results).

The aim of the present study was to identify the islet cells (A, B, D, or PP cells) in which this glycogen is localized. This was accomplished using a method designed to detect glycogen and pancreatic islet hormones in the same tissue section.

* This study was supported by the Deutsche Forschungsgemeinschaf~ K1 426/2 ** Two whom requests for offprints should be directed

0301- 5564/81/0073/0225/$01.60

226 R. Graf and C. Klessen

Material and Methods

Experiments were performed on 150-200 g adult male Sprague-Dawley rats (SPF-rats, S/iddeutsche Versuchstierfarm, Tuttlingen, FRG). The animals received a standard diet (Altromin | and water ad libitum. A single dosage of 5 mg/200 g B.W. triamcinolon-acetonid in crystalline suspension (Volon A | Squibb, von Heyden, Munich, FRG) was administered subcutaneously to 30 animals; 10 control animals received an equal dosage of isotonic saline. The gIucocorticoid was administered, and the animals sacrificed at the same time of day (between 9 a.m. and 11 a.m.). 24 h after glucocorti- coid application, the pancreas was removed under deep ether anesthesia from 5 experimental animals and 2 control animals. This procedure was repeated on 4 consecutive days.

The splenic portion of the pancreas was fixed for 24 h in Bouin's fluid (Romeis 1968, entry 304). After ethanol dehydration, the specimens were embedded in paraptast via methylbenzoate and benzol. 5 gm thick sections were mounted on microscope slides coated with gelatin containing alumchrome (method of Pappas 1971 modified according to Klessen and Rehbein 1974).

Histochemical and Immunocytochemical Reactions

a) Glycogen Detection. Chromic acid technique according to Bauer (Romeis, entry 1117), Best's carmine staining (Romeis, entry 1115), periodic acid-Schiff (PA S) reaction (McManus 1946, modified according to Graumann 1954), lead tetraacetate-Schiff (LTS) reaction (Graumann 1953 a; Malinin 1976). The Schiff's reagent was prepared according to Graumann (1953b) using basic fuchsin (Serva, Heidelberg, FRG) or pararosaniline (Merck, Darmstadt, FRG). Glycogen detection with silver proteinate (Thi6ry I967, modified according to Maxwell 1978): oxidation with 1% aqueous periodic acid solution, subsequent treatment of the sections with thiocarbohydrazide and silver proteinate. Alternatively 0.5% lead tetraacetate was used in place of 1% periodic acid as oxidizing agent. The specificity of the glycogen detection was checked with the diastase test (Graumann and Clauss 1959) using Diastase 2500 E/g (Merck). Control sections were treated with enzyme-free solution under otherwise identical conditions.

b) Immunocytochemical Method. For immunocytochemical detection of pancreatic islet hormones, the peroxidase-antiperoxidase (PAP) method was used (Sternberger 1979). The PAS or LTS reaction was performed prior to or following the PAP reaction. The following antibodies were used: normal sheep serum (Byk Mallinkrodt, Dietzenbach, FRG), normal rabbit serum (Byk Mallinkrodt), rabbit anti-guinea pig IgG 1:30 (Chemie Brunschwig, Basel, Switzerland), sheep anti-rabbit IgG 1:30 (Chemic Brunschwig), guinea pig anti-bovine insulin 1 : 10000 (Novo, Bagsvaerd, Denmark), rabbit anti-bovine glucagon 1:3000 (Novo), rabbit anti-bovine pancreatic polypepfide 1:8000 (Novo), rabbit anti-somatostatin 1:7500 (Ferring, Kiel, FRG), guinea pig peroxidase-antiperoxidase 1:5 (Fresenius, Oberursel, FRG), rabbit peroxidase-antiperoxidase 1:60 (Medac, Hamburg, FRG). Peroxidase activity is revealed after immune reaction by incubating the sections in diaminobenzidine (DAB) medium (Graham and Karnovsky 1966). Controls: a) Incubation with normal rabbit serum instead of anti-insulin - or normal sheep serum instead of anti-glucagon, anti-somatostatin or anti-pancreatic polypeptide; b)Incubation with the first antibody adsorbed with insulin (Novo), somatostatin (Curamed, Freiburg, FRG), glucagon (Novo) or pancreatic polypeptide (Nov@ Ad- sorption with 0.1% hormone solution, c) Incubation in DAB medium without addition of H~O2.

Blood Glucose Determination.

The presence of hyperglycemia was determined using the Reflotest | (Boehringer, Mannheim, FRG), which is reliable up to 350 rag% blood glucose. Blood samples were taken from anesthetized animals by cardiac puncture prior to removal of the pancreas.

Results

H y p e r g l y c e m i a c a n b e d e t e c t e d in e x p e r i m e n t a l a n i m a l s 24 h a f t e r a d m i n i s t r a t i o n

o f t r i a m c i n o l o n - a c e t o n i d a n d i n c r e a s e s d u r i n g the c o u r s e o f t he e x p e r i m e n t s

Glycogen in Pancreatic Islets 227

Figs. 1-4. Rat pancreas. Fixation: Bouin's fluid. 4 days after administration oftriamcinolon-acetonid

Figs. 1 a, b. Glycogen detection, a lead tetraacetate-Schiff reaction (LTS). b periodic acid-Schiff reaction (PAS)

(controls: 110 150 mg% blood glucose; experimental animals: 190-300 mg% on 1st day, subsequent further increase, > 350 mg% on 4th and 5th day). De- granulation of a portion of the B cells also occurs within 24 h, which does not affect all islets equally. In addition to islets with substantially degranulated cells, other islets contain numerous granulated B cells. Degranulation does not increase during the course of the experiments. Pancreatic islet glycogen can be detected by the 4th day after triamcinolon application. It occurs in fine and rough granula in cells lying in the center of the islets (Fig. 1 a). Granula may accumulate markedly in individual cells. Islet glycogen content varies notice- ably: in addition to islets very rich in glycogen some regularly contain little or no detectable glycogen. Pancreatic islet glycogen was not detected in any of the control animals.

The PAS reaction usually used to detect glycogen only reveals slight amounts in Bouin-fixed paraffin sections of pancreatic islets (Fig. 1 b). Techniques such as the chromic acid method according to Bauer or staining with the carmine method of Best yield even smaller amounts.

Substantially more islet glycogen can be detected, however, when using the LTS reaction (Fig. 1 a) with dimethyl sulfoxide (DMSO)/acetic acid instead of aqueous acetic acid solution as a lead tetraacetate solvent. In glycogen detection according to Maxwell, the reaction is also greatly enhanced when lead tetraace-


Figs. 2-4. Combined carbohydrate-immunocytochemical technique. Fig, 2a, combination glycogen detection (LTS) with insulin detection; glycogen is localized primarily in the granulated B cells. Fig, 2b, enlargement of a granulated B cell marked in Fig, 2a with an arrow. Fig. 3, Combination of glycogen detection with glncagon detection. Fig, 4. Combination of glycogen detection with pancreatic polypeptide detection; A and PP cells are free of glycogen

ta te is used ins tead of per iod ic acid as oxidiz ing agent. G lycogen de tec t ion is also more p r o n o u n c e d when pa ra rosan i l i ne is used ins tead o f basic fuchsin to p roduce the Schiff 's reagent .

A subs tan t ia l po r t i on of the pancrea t ic islets reveals c lear signs o f glycogen fl ight and d i sp lacement : cy top lasmic d i s t r ibu t ion o f g lycogen granu la is not un i fo rm; it is shifted in a cer ta in d i rec t ion within cells and the islets as a

whole.


In all of the glycogen detection methods employed in the present study, the reaction was negative following diastase pretreatment of sections.

In order to clarify the localization of glycogen in pancreatic islet cells, insulin detection was initially performed before glycogen detection. In all such cases glycogen detection was negative. When the sequence was reversed however - i.e., first glycogen, then insulin detection - both tests were positive. In order to determine the point at which the multistep PAP reaction inhibits glycogen detection, the PAP reaction was systematically interrupted step-by-step and immediately followed by the glycogen reaction. Detection of both insulin and glycogen was positive up to the step involving incubation with normal serum. Glycogen detection was negative subsequent to incubation with the first antibody.

The findings with the present combined detection method indicate that glycogen is primarily localized in the granulated B cells; degranulated B cells contain lesser amounts of glycogen (Figs. 2a and 2b). In contrast to the round to oval shape of the B cells in control animals, the B cells in steroid diabetic rats are polygonal and possess processes, in which also glycogen granula occur. Somatostatin, glucagon, and pancreatic polypeptide cells, however, either lack, or reveal only a few glycogen granula (Figs. 3 and 4).

D i s c u s s i o n

Microchemical measurements have revealed that glycogen is a regular constituent of mammalian pancreatic islets (Idahl and Hellman 1968; Matschinsky and Ellerman 1968; Hellman and Idahl 1969, 1970). However, glycogen is generally only detectable with histological-histochemical techniques in the presence of diabetic metabolic disturbances (Duff and Toreson 1951 ; Lazarus and Benscome 1955; Volk and Lazarus 1958a; Lazarow 1963; Fagundes and Cohen 1966). Nevertheless pancreatic islet glycogen cannot always be detected in the presence of even severe diabetic metabolic disturbances such as persistant hyperglycemia and glucosuria (Brolin and Berne 1970).

Despite the rapid induction and persistence of hyperglycemia, the histochemical methods used in the present study only detected glycogen in the pancreatic islets of the rat beginning on the 4th day after application of triamcinolon. This finding is consistent with that of Petkov et al. (1965), who also observed a delay between the start of hyperglycemia and glycogen infiltration in the islets of rats subjected to alloxan-induced diabetes. Glycogen infiltration in the rabbit and dog likewise does not begin until a few days after cortisone application; the glycogen content of the pancreatic islets continues to increase with increased length and intensity of the diabetes (Lazarus and Benscome 1955; Volk and Lazarus 1958a; Lazarus and Volk 1959; Volk and Lazarus 1964). The cause of the delay between the onset of hyperglycemia and the appearance of detectable amounts of islet glycogen could be the relatively low glycogen synthetase activity in the islets, which would only permit slow synthesis of glycogen (Brolin and Berne 1970).


Glycogen was localized in the B cells in earlier investigations using compara- bly unspecific and frequently also poorly reproducable staining techniques which, however, do permit selective B cell detection (e.g., aldehyde fuchsin staining, Volk and Lazarus 1958 b; Lazarow 1963).

The combined carbohydrate-immunocytochemical technique permits more precise localization of the glycogen. It is now possible to determine whether or not glycogen occurs in the granulated and/or degranulated B cells, or also in the A, D, or PP cells. A particular advantage is that the combination of glycogen detection with the immunocytochemical technique makes restaining or preparation of serial sections superfluous.

A significant finding of the present study is that the glycogen reaction must be conducted prior to the immune reaction. It is interesting that glycogen detection, which is performed subsequent to the multistep PAP reaction, is only negative after incubation with the first antibody. Glycogen, and thereby oxidiz- able glycol groups, may be simultaneously "masked" by the first antibody in a manner which is not clear at present.

The present results indicate that glycogen predominantly infiltrates granulated B cells. This finding contradicts the reports of Reynold et al. (1970) and Volk and Lazarus (1964) on the mouse and dog, respectively, whose electron optical results showed the occurrence of glycogen in degranulated cells. Whether or not glycogen was also detected in granulated B cells in the latter studies cannot be concluded from the published reports.

In contrast to the high glycogen content in B cells, merely a few granula are evident in the other endocrine cells. The fact that such granula are scarce and that the islet cells reveal glycogen flight and displacement, suggests that these few granules in the A, D, and PP cells may have been displaced from B cells. I.e., their localization may be an artifact. It cannot be ruled out, however, that glycogen could be infiltrated in these cells in the course of prolonged diabetes. An indication of such infiltration was provided by the findings of Volk and Lazarus (1963, 1964) on the dog. They found that glycogen was also infiltrated in A cells during prolonged diabetes induced by partial pancre- atectomy, and chronic administration of growth hormone.

One should keep in mind that all histochemical techniques for glycogen detection depend substantially on the specific detection method (Toreson 1951). Compared to other fixatives (Gendre, Rossman), Bouin's fluid is not optimal (Graumann 1958). It was used in the present study, however, because it permitted the best islet hormone detection results with the present immunoeytochemical technique. With regard to the PAS reaction, the present findings clearly demon- strate that Bouin-fixed paraffin sections treated with lead tetraacetate as oxidant reveal substantially more glycogen in the pancreatic islets than the usual glycogen detection techniques.

The dye used to produce the Schiff's reagent is also of significance: pararosaniline (Merck) yields much more glycogen than basic fuchsin. The suitability of other pararosanilines for producing the Schiff's reagent for glycogen detection in the pancreatic islets is currently being checked.

The present results illustrate the need for cautious evaluation of reports


on the histochemically detectable content of glycogen in pancreatic islets in the presence of diabetic metabolic disturbances.

Acknowledgements. We wish to thank Mrs. M. T61ken and Mrs. E. Fuchs for their skillful technical assistance.

References

Brolin SE, Berne C (1970) The activity pattern of enzymes associated with glycogen metabolism in the islets of Langerhans. In: Falkmer S, Hellman B, Tfiljedal I-B (eds) The structure and metabolism of the pancreatic islets. Pergamon Press, Oxford New York, pp 497 513

Duff GL, Toreson WE (1951) Prevention and reversal despite hyperglycemia of glycogen infiltration ("hydropic degeneration") in the pancreas in aIloxan diabetes in the rabbit. Endocrinology 48:298-312

Fagundes LA, Cohen RB (1966) Observations on amylophosphorylase activity and native glycogen in islet cells. Lab Invest 15 : 312

Graham RC, Jr, Karnovsky MJ (1966) The early stages of absorption of injected horseradish peroxidase in the proximal tubules of mouse kidney. Ultrastructural cytochemistry by a new technique. J Histochem Cytochem 14:291

Graumann W (1953a) Erfahrungen mit Bleitetraacetat als Oxidans ftir 1,2-Glykole. Microsc Acta 61:361 364

Graumann W (1953 b) Zur Standardisierung des Schiffschen Reagens. Microsc Acta 61:225 226 Graumann W (1954) Die histochemische Perjodatreaktion der Reticulin- und Kollagenfasern. Acta

Histochem (Jena) 1:116-126 Graumann W (1958) Untersuchungen zum cytochemischen Glykogennachweis. II. Mitteilung.

Chemische Fixation auf Pikrins~urebasis. Histochemie 1:97-108 Graumann W, Clauss W (1959) Untersuchungen zum cytochemischen Glykogennachweis. III. Mitt.

Versuche zum Diastasetest. Histochemie 1:241 246 Hellman B, Idahl LA (1969) Presence and mobilisation of glycogen in mammalian pancreatic

/?-cells. Endocrinology 84:1 8 Hellman B, Idahl LA (1970) On the functional significance of the pancreatic /Lcell glycogen.

In: Falkmer S, Hellman B, Tfiljedal I-B (eds) The structure and metabolism of the pancreatic islets. Pergamon Press, Oxford New York, pp 253-262

Idahl LA, Hellman B (1968) Microchemical assays of glucose and glucose-6-phosphate in mammalian pancreatic/?-cells. Acta Endocrinol (Kbh.) 59:479-486

Klessen Ch, Rehbein A (1974) Zur Schnittmontierung mit Chrom-Gelatine bei Behandlung yon Gewebsschnitten mit Perameisensfiure. Histochemistry 42:175-179

Lazarow A (1963) Functionai characterization and metabolic pathways of the pancreatic islet tissue. Recent Prog Horm Res 19:489 546

Lazarus SS, Benscome SA (1955) Alterations of pancreas during cortisone diabetes in rabbits. Proc Soc Exp Biol 89:114 I18

Lazarus SS, Volk BW (1958) Glycogen infiltration (hydropic degeneration) in the pancreas: A Review. Arch Pathol Lab Med 66:59-71

Lazarus SS, Volk BW (1959) Pancreatic adaptation to diabetogenic hormones. Arch Pathol Lab Med 67 : 456-467

Malinin GI (1976) Dimethyl sulfoxide-lead tetraacetate method for histochemical oxidation of polysaccharides. J Histochem Cytochem 24:443-447

Matschinsky FM, Ellerman JE (1968) Metabolism of glucose in the islets of Langerhans. J Biol Chem 243:2730 2736

Maxwell H (1978) An on-grid method for the specific demonstration of glycogen in electron micros- copy. Med Lab Sci 35:201

Petkov PE, Verne J, Wegmann R (1965) Recherches histoenzymatiques sur le pancr6as endocrine au cours du diab6te alloxanique chez le rat blanc. Ann Histochim 10:257 274

Reynold AE, Orci L, Stauffacher W, Junod A, Rouiller Ch (1970) Remarks on pancreatic /?-cells in spontaneous and experimental diabetes in small laboratory rodents. In: Falkmer S, Hellman


B, Tfiljedal I-B (eds) The structure and metabolism of the pancreatic islets. Pergamon Press, Oxford New York, pp 497-513

Romeis B (1968) Mikroskopische Technik. 16th ed. R. Oldenbom-g Verlag, Mfinchen Sternberger LA (1979) Immunocytochemistry, 2nd ed. John Wiley and Sons, New York Toreson WE (1951) Glycogen infiltration (so-called hydropic degeneration) in the pancreas in

human and experimental diabetes mellitus. Am J Pathol 27:327-347 Volk BW, Lazarus SS (1958 a) Zur Frage der hydropischen Regeneration des Pankreas beim experi-

mentellen Diabetes Mellitus. Ges Exp Med 130: 319-327 Volk BW, Lazarus SS (1958b) The effect of various diabetogenic hormones on the structure of

the rabbit pancreas. Am J Pathol 34:121-135 Volk BW, Lazarus SS (1963) Glycogenization of canine pancreatic A-cells during growth hormone

diabetes. Endocrinology 72:496 498 Volk BW, Lazarus SS (1964) Ultramicroscopic evolution of B-cell destruction in diabetic dogs.

In: Brolin SE, Hellman B, Knutson H (eds) The structure and metabolism of the pancreatic islets. Pergamon Press, Oxford London, pp 143-155

Received August 7, 1981

glycogen in pancreatic islets of steroid diabetic rats

Documents