00l3-7227/90/1265-2635$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society
Vol. 126, No. 5 Printed in U.S.A.
Immunohistochemical Localization of the Glucocorticoid Receptor in Pancreatic /?- Cells of the Rat* BRIT FISCHER, ULRICH RAUSCH, PETRA WOLLNY, HANNES WESTPHAL, JURGEN SEITZ, AND GERHARD AUMULLER Departments of Anatomy (B.F., U.R., J.S., G.A.), Cell Biology (P. W.), and Molecular Biology (H. WJ, University of Marburg, Marburg, West Germany
T
ABSTRACT. We used a monoclonal antibody against an epitope located in the N-terminal moiety of the rat glucocorticoid receptor to identify the glucocorticoid receptor-containing cells in the rat pancreas. Monospecific polyclonal antisera against insulin, glucagon, somatostatin, and amylase were applied to serial sections in colocalization studies to identify the respective endocrine and exocrine cells. Glucocorticoid receptor immunoreactivity was exclusively present in nuclei and cytoplasm of the j8-cells of pancreatic islets. Western blots using the glucocorticoid receptor antibody resulted in identical 94K immunoreactive proteins in both liver and pancreas. After adrenalectomy, the glucocorticoid receptor immunoreactivity of /3-cells decreased significantly. A computer-assisted method of semiquantitative
evaluation of the glucocorticoid receptor immunoreactivity demonstrated a significant decrease in the staining intensity of the /3-cells by 23.5% and in that of insulin antibodies by 10.4%, while amylase immunoreactivity was only slightly decreased. Serum levels of corticosterone determined by RIA decreased from 225 Mg/ml in sham-operated animals to 55 Mg/ml in animals 14 days after adrenalectomy, while the tissue content of amylase decreased by 45%. The immunohistochemical findings give circumstantial evidence of the presence of glucocorticoid receptor in /3-cells. We interpret our data as indicating an indirect effect of glucocorticoids on amylase synthesis via a glucocorticoid-insulin-exocrine cell pathway. (Endocrinology 126: 2635-2641,1990)
HE EXOCRINE pancreas is capable of modulating
cortisone, or aldosterone administration (12).
the synthesis and secretion of exocrine digestive
In an immunohistochemical approach, we have used a
enzymes according to the chemical composition of the ingested diet (1). It is now well documented that this regulation is mediated by the peptide hormones insulin, cholecystokinin, and secretin for carbohydrates, proteins, and lipids, respectively (2-4). In addition to peptide hormones, steroid hormones seem to be involved in the regulation of the exocrine pancreas (5, 6). The physiological role of glucocorticoids in the regulation of gene expression of exocrine pancreatic enzymes is not well understood. There are some indications for an indirect effect of glucocorticoids on amylase secretion. Insulin has been shown to be involved in the regulation of pancreatic amylase secretion (7-9); it is unclear, however, if this is a direct or an indirect effect (10). Corticosterone is capable of an acute and effective inhibition of glucoseand arginine-induced secretion of insulin (11). Also, glucose-induced insulin secretion of isolated pancreatic islets is inhibited after dexamethasone, prednisone, hydro-
monospecific monoclonal antibody against the rat glucocorticoid receptor (13) to identify the receptor-containing cells. The /3-cells of Langerhans islets were the only cell type in the pancreas that contained glucocorticoid receptor. The receptor content is down-regulated after glucocorticoid deprivation, as shown by adrenalectomy experiments. There was an only slight decrease in insulin and amylase immunoreactivity in the pancreas of adrenalectomized rats, indicating a preferentially indirect action of glucocorticoids on insulin-dependent amylase secretion via a glucocorticoid-insulin-acinar cell pathway. Materials and Methods Animal care and experimental groups
Received August 28, 1989. Address all correspondence and requests for reprints to: U. Rausch, Department of Anatomy and Cell Biology, Robert-Koch-Street 6, D3550 Marburg, West Germany. * This work was supported by the Deutsche Forschungsgemeinschaft (SFB 215/B4/B5/C3).
Young male Wistar rats, weighing 180-200 g (8-11 weeks old) were purchased from Ivanovas (Kieslegg, West Germany) and kept at 24 C on a 12-h light, 12-h dark cycle in heat- and humidity-controlled animal rooms. Standard rat chow (Altromin, Lage, West Germany) and tap water were freely available. Rats were anesthetized with chloral hydrate (1 ml/100 g BW) and bilaterally adrenalectomized or sham-operated by the dorsal approach. Body weights were monitored before surgery and 6,10, and 14 days later.
2635
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
GLUCOCORTICOID RECEPTOR IN PANCREATIC /3-CELLS
2636
Endo • 1990 Vol 126 • No 5
FIG. 1. Immunocytochemical localization of glucocorticoid receptor, insulin, glucagon, somatostatin, and amylase in rat pancreas on paraffin sections from 6-day sham-operated animals, a, Characteristic immunocytochemical reaction of /3-cells with the glucocorticoid receptor antibody. Compared to the cytoplasm, the nuclei are intensively stained. The peripheral cells of the pancreatic islet and exocrine cells are unstained. Magnification, X100. b, Immunoreaction of /3-cells in the same islet as those in a, using an insulin antibody. The distribution patterns of immunoreactive cells in a and b are similar. The intracellular distribution and staining intensities are different. X200. c, A sequential section is stained with the glucagon antibody; acells located in the periphery of the islet show an intense immunoreaction. These cells are clearly unstained in a. X200. d, Immunoreaction of 5-cells using a somatostatin antibody. In this case the cells are restricted to the periphery of the islet and are devoid of glucocorticoid receptor immunoreactivity in a serial section (not shown). X200. e, Immunoreaction of the exocrine pancreatic tissue using the amylase antibody. The exocrine cells show a moderately reacting cytoplasm. The intraluminal secretion in a small pancreatic duct {righthand portion of figure) is strongly stained, while the pancreatic islet {center) is clearly unstained. X100.
Tissue collection Rats were killed in the animal room at 1400 h by decapitation 6,10, and 14 days after surgery. Blood was collected into chilled heparinized plastic centrifuge tubes, and plasma was frozen and stored at —70 C for corticosterone determination. Pancreata were removed from the animals, dissected free of extraneous tissue, and cut into two halves. One portion was immersion fixed in Bouin's fixative (24 h at room temperature). The remainder of the gland was frozen on dry ice and stored at —70 C for later determination of pancreatic enzyme activity and protein and DNA contents. Corticosterone determination in plasma Corticosterone was measured in unextracted plasma using a rabbit polyclonal antibody against corticosterone and [3H]corticosterone (ICN, Carson, CA). The antiserum cross-reacts 100% with corticosterone, 6% with desoxycorticosterone, and less than 0.3% with other steroids. A 3H RIA (ICN) was used to determine the corticosterone content in blood plasma (ID50, 146 ng/ml; intraassay coefficient of variance, 3%; n = 30). Biochemical analysis of pancreatic contents One quarter of each pancreas was homogenized in 25 mM Tris-HCl (pH 8.9), 0.1% Triton X-100, 0.1 mM phenylmethyl-
sulfonylfluoride, and 10 nM Foy 305 (Sanol-Schwarz, Monheim, West Germany). Samples were precipitated in 0.5 N perchloric acid for DNA determination, as previously described (14). Protein was determined, using BSA as standard (15). Further samples were assayed for amylase (16). Cytoplasmic proteins were isolated immediately after removing the tissue by a 1-h 100,000 X g centrifugation in a Beckman 50Ti rotor (Fullerton, CA) and boiled in sample buffer (17). Sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) (12% polyacrylamide) was performed (17) and visualized by Coomassie blue staining. Transfer of proteins on nitrocellulose membranes was performed in a carbon electrode transblot chamber (CTI, Idstein, West Germany) (18). Nonspecific binding sites were blocked with nonfat dry milk (19), and the sheets were incubated with the glucocorticoid receptor antibody. Antibodies Commercial polyclonal antisera against insulin (raised in guinea pig; IBL, Hamburg, West Germany), glucagon (raised in rabbit; IBL), and somatostatin (raised in rabbit; Sigma, Munich, West Germany) were used at dilutions (in PBS) ranging from 1:200-1:1000. Secondary antibodies [peroxidaselabeled antiguinea pig immunoglobulin (IgG) and peroxidaselabeled antirabbit IgG (Dakopatts, Hamburg, West Germany)]
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
GLUCOCORTICOID RECEPTOR IN PANCREATIC /3-CELLS
:
2637
^mm
FlG. 2. Immunocytochemical localization of glucocorticoid receptor in pancreatic islets from control and experimental animals: a-c, 6, 10, and 14-day sham-operated animals; d-f, 6, 10, and 14-day adrenalectomized animals. Magnification: 21 and b, X200; c-f, X100. a, b, and c, No essential differences in the staining intensity and distribution of immunoreaction with the glucocorticoid receptor antibody are seen in the j8-cells of pancreatic islets from sham-operated animals. The endocrine cells at the periphery of the islet and the exocrine cells remain unstained, d, Six days after adrenalectomy the glucocorticoid receptor immunoreactivity of the /3-cells appears in the control range; the number of strongly reacting cells is reduced, e, Ten days after adrenalectomy the decrease in glucocorticoid receptor immunoreactivity becomes obvious; most nuclei are only slightly immunoreactive. f, Only a few glucocorticoid receptor-immunoreactive cells are left in a pancreatic islet 14 days after adrenalectomy. In other cases, the decrease in immunoreactivity was less pronounced.
were used at dilutions of 1:30 and 1:200, respectively. A polyclonal antibody raised against highly purified rat pancreatic amylase in rabbits was prepared in our laboratory and proved to be monospecific by two-dimensional Western blotting. The antibody against the rat glucocorticoid receptor has been extensively characterized and described previously (13). It recognizes an epitope at the N-terminal portion of the receptor. The secondary antibody was a commercial peroxidase-labeled antimouse IgG (Amersham-Buchler, Braunschweig, West Germany), which was diluted 1:30.
V ^ H S3K* tkl.
?s$t- *
a chromogen and hydrogen peroxide as a substrate resulted in a brown stain. All incubations took place in a humid chamber at 22 C. Sections were then dehydrated and mounted in synthetic resin. Primary antibodies were applied to serial sections to yield a serial sequence of immunoreactions with glucocorticoid receptor, insulin, glucagon, somatostatin, and amylase. Control incubations included 1) substitutions of the respective primary antisera with nonimmune serum, 2) progressive decrease in antibody concentrations and eventually replacement by PBS, and 3) use of irrelevant antibodies (antiprostatic binding protein) (20) as antibody controls.
Immunohistochemistry Specimens from sham-operated as well as adrenalectomized animals (6, 10, and 14 days after surgery) were used. Bouinfixed paraffin-embedded pancreata were serially cut at 4- to 5juni thickness, deparaffinized in xylene, and rehydrated in graded alcohol baths. The indirect immunoperoxidase method was applied. Sections were first treated with a 3% hydrogen peroxide solution, followed by a 10% BSA solution in PBS to block endogenous peroxidase activity. Subsequently, sections were incubated sequentially with normal swine serum, the respective primary antisera, and the pertinent peroxidase-labeled secondary antibodies. Reaction with diaminobenzidine as
Semiquantitatiue evaluation of immunoreactions Photomicrographs of each of the immunoreactions were taken in a Zeiss Axiomat photomicroscope (New York, NY) at a final magnification of about X100. The individual cell types (A-, B-, and D-cells) of the islets of Langerhans were classified according to their staining intensity (light-medium-dark) and counted. Since this method allows no quantitation, an image analysis system (Zeiss IBAS, Kontron Software) was used. This system allows the discrimination of 25 different classes of grey values in photomicrographs. A grey value of 1-10 (class 1) represents black; a grey value of 240-250 (class 25) represents
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
2638
GLUCOCORTICOID RECEPTOR IN PANCREATIC /3-CELLS
spectively, and 2 sham-operated animals per group. About 100300 cells (cytoplasm or nuclei) per islet were marked with a cursor, and the grey values are measured with the IBAS system, using the exocrine pancreas as background. The number of grey values per group was calculated (mean ± SD), expressed as a percentage of the total number of all cells counted, and plotted (see Figs. 3 and 4). Immunoreactions encompassed the glucocorticoid receptor and insulin stains. In the case of the glucocorticoid receptor, only the nuclei were measured, while in case of insulin, only the cytoplasm (excluding the nuclei) was measured.
Glucocorticoid Receptor Immunreactivity Computer assisted quantitation (IBAS) % frequency
Results
10 IS grey values ADX-6d
ADX-10d
ADX-14d
Endo • 1990 Voll26«No5
SHAM
FIG. 3. Semiquantitative evaluation of glucocorticoid receptor immunoreactivity in /3-cells from control and experimental animals. Photomicrographs from pancreatic islets of three animals of each adrenalectomy group (6,10, and 14 days) and 2 animals of each sham operation group (6, 10, and 14 days) processed for glucocorticoid receptor immunoreactivity were used for image analysis (Zeiss IBAS). The nuclei of cells in five islets per group (equalling about 100-300 nuclei/islet) were labeled with a pointer and monitored for staining intensity (expressed as grey values relative to the nuclei of the exocrine tissue that served as the control). The distribution of grey values (representing glucocorticoid receptor immunoreactivity) was expressed as a percentage of the total number of nuclei counted per group. A shift in the grey values was detected only by day 10 in adrenalectomized animals. This shift in grey values, i.e. decrease in glucocorticoid receptor immunoreactivity, was still more pronounced in 14-day adrenalectomized animals. The effect of adrenalectomy was calculated from the maxima of the adrenalectomy plot relative to the sham plot and was about 23.5%. Insulin Immunoreactivity Computer assisted quantitation (IBAS) % frequency
Immunolocalization
of the glucocorticoid receptor
At low magnification, the islets of Langerhans were
the only tissue components of the pancreas that reacted specifically with the glucocorticoid receptor antibody. At higher magnifications, the remnants of adipose tissue also displayed a positive immunoreaction. Since there was an uneven distribution of glucocorticoid receptor immunoreactivity among the endocrine islet cells, sequential staining with antisera against insulin, glucagon, somatostatin, and a-amylase was performed (Fig. 1). /3Cells were immunoreactive with both the insulin antiserum and the glucocorticoid antiserum. In A- and Dcells immunoreactivity with the glucocorticoid receptor was in the background range. At higher antibody concentrations (1:200-1:400), and short incubation periods (1 h at room temperature) the cytoplasm of the /3-cells also reacted with the glucocorticoid receptor antibody. Cytoplasmic staining decreased considerably at higher dilutions (1:1000) of the antibody and prolonged incubation at low temperature (24-48 h at 4 C). All control incubations were negative. Effects of adrenalectomy on immunoreactivity
grey values SHAM-14d
ADX-14d
FIG. 4. Semiquantitative evaluation of insulin immunoreactivity of /?cells in pancreatic islets from control and experimental animals. The approach described in Fig. 3 was used by labeling the cytoplasm (nuclei excluded) of /3-cells by a pointer and determining the staining intensity. Animals 14 days after adrenalectomy were compared to sham-operated animals. The decrease in insulin immunoreactivity, comparing the maxima of the adrenalectomy plot to the sham plot, was 10.4%.
white. A total of 15 islets (5 islets/animals; 3 animals/group) per immunoreaction were evaluated. The groups comprised 3 animals after each adrenalectomy for 6, 10, and 14 days, re-
Comparison of the immunoreactions with the glucocorticoid receptor antibody in sham-operated and adrenalectomized rats clearly resulted in a decrease in nuclear staining intensity of /3-cells in animals 14 days after adrenalectomy. Using a quantitative computer-assisted approach, no significant changes in the glucocorticoid receptor immunoreactivity of /3-cell nuclei was observed in sham-operated animals 6, 10, or 14 days after surgery (Fig. 2, a-c). Contrary to that, in adrenalectomized animals a significant decrease in immunostaining with the glucocorticoid receptor antiserum was observed (Fig. 2, d-f). Comparing 14-day sham-operated with 14-day adrenalectomized animals, a significant 23.5% decrease in staining of /3-cells was calculated (Fig. 3). The quantitative evaluation of insulin immunoreactivity of /3-cells (Fig. 4) in sham-operated (14 days; Fig. 5a) and adrenalectomized (14 days; Fig. 5b) animals resulted in a de-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
GLUCOCORTICOID RECEPTOR IN PANCREATIC tf-CELLS
2639
FlG. 5. Insulin immunoreactivity in pancreatic islets from control and experimental animals, a, In a 14-day shamoperated animal no difference is observed from the control situation (Fig. lb). b, In an animal 14 days after adrenalectomy the staining intensity of the /5-cells is very low; this is obviously partly due to the low avidity of the insulin antibody used.
FIG. 6. Amylase immunoreactivity of the exocrine pancreas from control and experimental animals, a, The exocrine portion of the pancreas is homogeneously stained by a moderately strong immunoreaction in a 14-day sham-operated animal, b, No significant differences in amylase immunoreactivity from the sham-operated situation (a) is observed in the exocrine pancreas from an animal 14 days after adrenalectomy. TABLE 1. Effect of 14-day adrenalectomy (ADX) on body weight, pancreas weight, a-amylase activity, and corticosterone content kDa
BW(g)
Pancreas (g)
Amylase (U/ Mg DNA)
Corticosterone (ng/ml)
SHAM 307 ± 2 1.05 ± 0.1 8.05 ± 1.4 225 ±9.9 ADX 239 ± 3 0.92 ± 0.1 55.7 ± 7.7 4.2 ± 1.1 The relatively high values of corticosterone may be due to the time of death of the rats (1400 h).
- : _ _ • •
W
**
"
•*-
66
crease in staining of /3-cells by about 10.4% (Fig. 4). No obvious changes in the immunoreaction of amylase antiserum with exocrine pancreatic cells were observed (Fig. 6). Biochemical data Table 1 summarizes the experimental data obtained from animals after glucocorticoid deprivation. While the pancreas weight changed only slightly in adrenalectomized rats, body weights decreased by 22% compared to those in sham-operated animals. Amylase content decreased by 50% after 14 days, whereas other exocrine proteins (trypsinogen, chymotrypsinogen, and lipase) remained unchanged (data not shown). The corticosterone content in blood of adrenalectomized animals decreased by 75% compared to that in sham-operated animals. To get more information about the identity of the pancreatic glucocorticoid receptor, cytosolic samples of pancreas and liver homogenates were separated by SDSPAGE, transblotted on nitrocellulose, and incubated with the glucocorticoid receptor antibody (Western blotting). Figure 7 shows the identification of the 94-kD form of the glucocorticoid receptor in both liver and pancreas.
116
45
29
a b c M FIG. 7. Immunoidentification of the glucocorticoid receptor by Western blotting. Rat pancreas and liver cytosolic fractions were prepared as described in Materials and Methods. Cytosolic fractions were separated by SDS-PAGE, transferred to nitrocellulose sheets, and stained with the glucocorticoid receptor antibody, a, Pancreas cytosol (198 ng total protein), b, Liver cytosol (225 fig total protein), c, Purified rat liver cytosolic glucocorticoid receptor (200 ng). M, Marker proteins; GR, The 94-kD form of the glucocorticoid receptor. The 37-kD protein is an immunoreactive fragment of the receptor appearing after freezing and thawing of the samples. Due to proteolytic degradation a 37-kD immunoreactive fragment appeared.
Discussion The presence of glucocorticoid receptor exclusively in nuclei of pancreatic /3-cells has been shown immunohis-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
2640
GLUCOCORTICOID RECEPTOR IN PANCREATIC jS-CELLS
tochemically, while other islet cell types and the exocrine part of the pancreas remained unstained. The hormone dependence of the immunoreactivity has been crosschecked by hormone deprivation experiments. Hormonedependent down-regulation of the glucocorticoid receptor in /3-cells subsequent to corticosteroid deprivation was followed by a decrease in peripheral levels and intracellular concentrations of insulin. Steroid hormone receptors can be detected autoradiographically (5, 21), using hybridization techniques (22), or by immunocytochemistry at both the light (23-25) and electron microscopic levels (26). Our immunohistochemical results in the pancreas are comparable to those in the liver (22) in that the reaction was preferentially, but not exclusively, confined to nuclei (of /3-cells), and the antibody reacted well with Bouin-fixed paraffin-embedded tissue. As has been shown in the liver (22), the intensity of the immunostaining is hormone dependent, i.e. it decreases in adrenalectomized animals. These findings have been confirmed by evaluation of the immunoreaction. The semiquantitative computer-assisted measurement of grey values (representing the immunoreactive material) proved to be effective. It requires, however, a highly standardized immunohistochemical technique so that the staining reaction adequately reflects the true antigen content of the respective structure. While the staining intensity representing the content of /3-cells in immunoreactive insulin in adrenalectomized animals decreased by only 10%, the respective value for the glucocorticoid receptor immunoreaction decreased by about 25%, indicating a downregulation of the glucocorticoid receptor after adrenalectomy. Decreases of 90% in basal insulin serum levels and of 70% in tissue amylase activity after adrenalectomy have been described (27) in normal rats. It was uncertain, however, whether this means a direct or an indirect (via glucocorticoid) action of insulin on amylase synthesis or vice versa. Addition of glucose to the drinking water of adrenalectomized rats increased both plasma insulin and glucose, but had no effect on amylase content, indicating a direct regulatory effect of glucocorticoids (28). Pancreatic acinar cell amylase gene expression has been studied after adrenalectomy and corticosterone replacement (10). It was found that physiological levels of corticosterone have a selective effect on pancreatic amylase gene expression, which may account for the effect of adrenalectomy and replacement on amylase synthesis. Glucocorticoid-induced gene regulation in the azaserineinduced rat pancreatic acinar cell carcinoma resulted in a decrease in amylase (29) or most cellular proteins (30). These data were explained as an indirect effect via inducing regulatory proteins in the cell (29), but it seems more likely that these AR4-2J tumor cells serve as a useful model to study general differentiative effects of
Endo • 1990 Vol 126 • No 5
glucocorticoids (30). Although considerable evidence suggests that insulin is an effective modulator of pancreatic amylase synthesis (7, 10), it is not clear whether there is a direct effect of glucocorticoids on pancreatic amylase gene expression or an indirect effect mediated via an influence of insulin. Our findings of a glucocorticoid-regulated glucocorticoid receptor exclusively present in /3-cells of the islets of Langerhans are clearly in favor of an indirect effect of glucocorticoids on amylase synthesis via a glucocorticoidinsulin-exocrine pathway.
Acknowledgments The authors thank Drs. J. Wager and M. Kirsch, Department of Anatomy, University of Marburg, for advice and assistance in the morphometrical analysis. Mrs. I. Dammshauser for photographic work, M. Dreher for excellent immunocytochemical assistance, Drs. G. Sturm and U. Bieker for corticosterone determinations, and Dr. H. F. Kern, Department of Cell Biology, University of Marburg, for critical reading of the manuscript.
References 1. Grossman MI, Greengard H, Ivy IC 1943 On the mechanism of the adaptation of pancreatic enzymes to dietary composition. Am J Physiol 141:38 2. Soling HD, Unger KG 1972 The role of insulin in the regulation of a-amylase synthesis in the rat pancreas. Eur J Clin Invest 2:199 3. Schick J, Kern HF, Scheele GA 1984 Hormonal stimulation in the exocrine pancreas results in a coordinate and anticoordinate regulation of protein synthesis. J Cell Biol 99:1559 4. Rausch U, Riidiger K, Vasiloudes P, Kern HF, Scheele GA 1986 Lipase synthesis in the rat exocrine pancreas is regulated by secretin. Pancreas 1:522 5. Winborn WB, Sheridan PJ, McGill Jr HC 1987 Localization of progestin receptors in the islets of Langerhans. Pancreas 2:289 6. Logsdon CD, Mossner J, Williams JA, Goldfine I 1985 Glucocorticoids increase amylase mRNA levels, secretory organelles, and secretion in pancreatic acinar AR42J cells. J Cell Biol 100:1200 7. Korc M, Owerbach D, Quinto C, Rutter WJ 1981 Pancreatic isletacinar cell interaction: amylase messenger RNA levels are determined by insulin. Science 213:351 8. Abdeljlil AB, Palla JC, Desnuelle P 1965 Effect of insulin on pancreatic amylase and chymotrypsinogen. Biochem Biophys Res Commun 18:71 9. Mossner J, Logsdon CD, Williams JA, Goldfine ID 1985 Insulin, via its own receptor, regulates growth and amylase synthesis in pancreatic acinar AR 42J cells. Diabetes 34:891 10. Logsdon CD, Akada SF, Meyer C, Dallman MF, Williams JA 1987 Pancreatic acinar cell amylase gene expression: selective effects of adrenalectomy and corticosterone replacement. Endocrinology 121:1242 11. Barseghian G, Levine R, Epps P 1982 Direct effect of cortisol and cortisone on insulin and glucagon secretion. Endocrinology 111:1648 12. Pierluissi J, Navas FO, Ashcroft SJH 1986 Effect of adrenal steroids on insulin release from cultured rat islets of Langerhans. Diabetologia 29:119 13. Westphal HM, Moldenhauer G, Beato M 1982 Monoclonal antibodies to the rat liver glucocorticoid receptor. EMBO J 1:1467 14. Richards GM 1974 Modifications of the diphenylamine reaction giving increased sensitivity and simplicity in the estimation of DNA. Anal Biochem 57:369 15. Bradford MM 1976 A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the the proteindye binding. Anal Biochem 72:248
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
GLUCOCORTICOID RECEPTOR IN PANCREATIC /3-CELLS 16. Bernfeld P 1955 Amylase a and 0. Methods Enzymol 1:149 17. Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680 18. Towbin H, Staehelin T, Gordon J 1979 Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76:4350 19. Johnson DA, Gautsch JW, Sportsman JR, Elder JH 1984 Improved technique utilizing nonfat dry milk for analysis of proteins and nucleic acids transferred to nitrocellulose. Gene Anal Tech 1:3 20. Aumuller G, Seitz J, Heyns W, Flickinger CJ 1982 Intracellular localization of prostatic binding protein (PBP) in rat prostate by light and electron microscopic immunocytochemistry. Histochemistry 76:497 21. Stumpf WE 1971 Autoradiographic techniques for the localization of hormones and drugs at the cellular or subcellular level. Acta Endocrinol (Copenh) 153:205 22. Pfeifer A, Barden N 1988 Glucocorticoid receptor gene expression in rat pituitary gland intermediate lobe following ovariectomy. Mol Cell Endocrinol 55:115 23. Antakly T, Eisen HJ 1984 Immunocytochemical localisation of glucocorticoid receptor in target cells. Endocrinology 115:1984
2641
24. Teasdale J, Lewis FA, Barrett ID, Abbott AC, Wharton J, Bird CC 1986 Immunocytochemical application of monoclonal antibodies to rat liver glucocorticoid receptor. J Pathol 150:227 25. LaFond RE, Kennedy SW, Harrison RW, Villee CA 1988 Immunocytochemical localization of glucocorticoid receptors in cells, cytoplasts, and nucleoplasts. Exp Cell Res 175:52 26. Liposits ZS, Uht RM, Harrison RW, Gibbs FP, Paull WK, Bohn MC 1987 Ultrastructural localization of glucocorticoid receptor (GR) in hypothalamic paraventricular neurons synthesizing corticotropin releasing factor (CRF). Histochemistry 87:407 27. Mossner J, Bohm SK, Secknus R, Fischbach W, Dammrich J, Wunsch PH 1987 Einflu/3 von Glukokortikoiden auf das exokrine Pankreas der Ratte. Verh Dtsch Ges Pathol 71:108 28. Mossner J, Bohm S, Fischbach W 1989 Role of glucocorticosteroids in the regulation of pancreatic amylase synthesis. Pancreas 4:194 29. Logsdon CD, Perot KJ, McDonald AR 1987 Mechanism of glucocorticoid-induced increase in pancreatic amylase gene transcription. J Biol Chem 262:15765 30. Swarovsky B, Steinhilber W, Scheele GA, Kern HF 1988 Coupled induction of exocrine proteins and intracellular compartments involved in the secretory pathway in AR4-2J cells by glucocorticoids. Eur J Cell Biol 47:101
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
Vol. 126, No. 5 Printed in U.S.A.
Immunohistochemical Localization of the Glucocorticoid Receptor in Pancreatic /?- Cells of the Rat* BRIT FISCHER, ULRICH RAUSCH, PETRA WOLLNY, HANNES WESTPHAL, JURGEN SEITZ, AND GERHARD AUMULLER Departments of Anatomy (B.F., U.R., J.S., G.A.), Cell Biology (P. W.), and Molecular Biology (H. WJ, University of Marburg, Marburg, West Germany
T
ABSTRACT. We used a monoclonal antibody against an epitope located in the N-terminal moiety of the rat glucocorticoid receptor to identify the glucocorticoid receptor-containing cells in the rat pancreas. Monospecific polyclonal antisera against insulin, glucagon, somatostatin, and amylase were applied to serial sections in colocalization studies to identify the respective endocrine and exocrine cells. Glucocorticoid receptor immunoreactivity was exclusively present in nuclei and cytoplasm of the j8-cells of pancreatic islets. Western blots using the glucocorticoid receptor antibody resulted in identical 94K immunoreactive proteins in both liver and pancreas. After adrenalectomy, the glucocorticoid receptor immunoreactivity of /3-cells decreased significantly. A computer-assisted method of semiquantitative
evaluation of the glucocorticoid receptor immunoreactivity demonstrated a significant decrease in the staining intensity of the /3-cells by 23.5% and in that of insulin antibodies by 10.4%, while amylase immunoreactivity was only slightly decreased. Serum levels of corticosterone determined by RIA decreased from 225 Mg/ml in sham-operated animals to 55 Mg/ml in animals 14 days after adrenalectomy, while the tissue content of amylase decreased by 45%. The immunohistochemical findings give circumstantial evidence of the presence of glucocorticoid receptor in /3-cells. We interpret our data as indicating an indirect effect of glucocorticoids on amylase synthesis via a glucocorticoid-insulin-exocrine cell pathway. (Endocrinology 126: 2635-2641,1990)
HE EXOCRINE pancreas is capable of modulating
cortisone, or aldosterone administration (12).
the synthesis and secretion of exocrine digestive
In an immunohistochemical approach, we have used a
enzymes according to the chemical composition of the ingested diet (1). It is now well documented that this regulation is mediated by the peptide hormones insulin, cholecystokinin, and secretin for carbohydrates, proteins, and lipids, respectively (2-4). In addition to peptide hormones, steroid hormones seem to be involved in the regulation of the exocrine pancreas (5, 6). The physiological role of glucocorticoids in the regulation of gene expression of exocrine pancreatic enzymes is not well understood. There are some indications for an indirect effect of glucocorticoids on amylase secretion. Insulin has been shown to be involved in the regulation of pancreatic amylase secretion (7-9); it is unclear, however, if this is a direct or an indirect effect (10). Corticosterone is capable of an acute and effective inhibition of glucoseand arginine-induced secretion of insulin (11). Also, glucose-induced insulin secretion of isolated pancreatic islets is inhibited after dexamethasone, prednisone, hydro-
monospecific monoclonal antibody against the rat glucocorticoid receptor (13) to identify the receptor-containing cells. The /3-cells of Langerhans islets were the only cell type in the pancreas that contained glucocorticoid receptor. The receptor content is down-regulated after glucocorticoid deprivation, as shown by adrenalectomy experiments. There was an only slight decrease in insulin and amylase immunoreactivity in the pancreas of adrenalectomized rats, indicating a preferentially indirect action of glucocorticoids on insulin-dependent amylase secretion via a glucocorticoid-insulin-acinar cell pathway. Materials and Methods Animal care and experimental groups
Received August 28, 1989. Address all correspondence and requests for reprints to: U. Rausch, Department of Anatomy and Cell Biology, Robert-Koch-Street 6, D3550 Marburg, West Germany. * This work was supported by the Deutsche Forschungsgemeinschaft (SFB 215/B4/B5/C3).
Young male Wistar rats, weighing 180-200 g (8-11 weeks old) were purchased from Ivanovas (Kieslegg, West Germany) and kept at 24 C on a 12-h light, 12-h dark cycle in heat- and humidity-controlled animal rooms. Standard rat chow (Altromin, Lage, West Germany) and tap water were freely available. Rats were anesthetized with chloral hydrate (1 ml/100 g BW) and bilaterally adrenalectomized or sham-operated by the dorsal approach. Body weights were monitored before surgery and 6,10, and 14 days later.
2635
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
GLUCOCORTICOID RECEPTOR IN PANCREATIC /3-CELLS
2636
Endo • 1990 Vol 126 • No 5
FIG. 1. Immunocytochemical localization of glucocorticoid receptor, insulin, glucagon, somatostatin, and amylase in rat pancreas on paraffin sections from 6-day sham-operated animals, a, Characteristic immunocytochemical reaction of /3-cells with the glucocorticoid receptor antibody. Compared to the cytoplasm, the nuclei are intensively stained. The peripheral cells of the pancreatic islet and exocrine cells are unstained. Magnification, X100. b, Immunoreaction of /3-cells in the same islet as those in a, using an insulin antibody. The distribution patterns of immunoreactive cells in a and b are similar. The intracellular distribution and staining intensities are different. X200. c, A sequential section is stained with the glucagon antibody; acells located in the periphery of the islet show an intense immunoreaction. These cells are clearly unstained in a. X200. d, Immunoreaction of 5-cells using a somatostatin antibody. In this case the cells are restricted to the periphery of the islet and are devoid of glucocorticoid receptor immunoreactivity in a serial section (not shown). X200. e, Immunoreaction of the exocrine pancreatic tissue using the amylase antibody. The exocrine cells show a moderately reacting cytoplasm. The intraluminal secretion in a small pancreatic duct {righthand portion of figure) is strongly stained, while the pancreatic islet {center) is clearly unstained. X100.
Tissue collection Rats were killed in the animal room at 1400 h by decapitation 6,10, and 14 days after surgery. Blood was collected into chilled heparinized plastic centrifuge tubes, and plasma was frozen and stored at —70 C for corticosterone determination. Pancreata were removed from the animals, dissected free of extraneous tissue, and cut into two halves. One portion was immersion fixed in Bouin's fixative (24 h at room temperature). The remainder of the gland was frozen on dry ice and stored at —70 C for later determination of pancreatic enzyme activity and protein and DNA contents. Corticosterone determination in plasma Corticosterone was measured in unextracted plasma using a rabbit polyclonal antibody against corticosterone and [3H]corticosterone (ICN, Carson, CA). The antiserum cross-reacts 100% with corticosterone, 6% with desoxycorticosterone, and less than 0.3% with other steroids. A 3H RIA (ICN) was used to determine the corticosterone content in blood plasma (ID50, 146 ng/ml; intraassay coefficient of variance, 3%; n = 30). Biochemical analysis of pancreatic contents One quarter of each pancreas was homogenized in 25 mM Tris-HCl (pH 8.9), 0.1% Triton X-100, 0.1 mM phenylmethyl-
sulfonylfluoride, and 10 nM Foy 305 (Sanol-Schwarz, Monheim, West Germany). Samples were precipitated in 0.5 N perchloric acid for DNA determination, as previously described (14). Protein was determined, using BSA as standard (15). Further samples were assayed for amylase (16). Cytoplasmic proteins were isolated immediately after removing the tissue by a 1-h 100,000 X g centrifugation in a Beckman 50Ti rotor (Fullerton, CA) and boiled in sample buffer (17). Sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) (12% polyacrylamide) was performed (17) and visualized by Coomassie blue staining. Transfer of proteins on nitrocellulose membranes was performed in a carbon electrode transblot chamber (CTI, Idstein, West Germany) (18). Nonspecific binding sites were blocked with nonfat dry milk (19), and the sheets were incubated with the glucocorticoid receptor antibody. Antibodies Commercial polyclonal antisera against insulin (raised in guinea pig; IBL, Hamburg, West Germany), glucagon (raised in rabbit; IBL), and somatostatin (raised in rabbit; Sigma, Munich, West Germany) were used at dilutions (in PBS) ranging from 1:200-1:1000. Secondary antibodies [peroxidaselabeled antiguinea pig immunoglobulin (IgG) and peroxidaselabeled antirabbit IgG (Dakopatts, Hamburg, West Germany)]
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
GLUCOCORTICOID RECEPTOR IN PANCREATIC /3-CELLS
:
2637
^mm
FlG. 2. Immunocytochemical localization of glucocorticoid receptor in pancreatic islets from control and experimental animals: a-c, 6, 10, and 14-day sham-operated animals; d-f, 6, 10, and 14-day adrenalectomized animals. Magnification: 21 and b, X200; c-f, X100. a, b, and c, No essential differences in the staining intensity and distribution of immunoreaction with the glucocorticoid receptor antibody are seen in the j8-cells of pancreatic islets from sham-operated animals. The endocrine cells at the periphery of the islet and the exocrine cells remain unstained, d, Six days after adrenalectomy the glucocorticoid receptor immunoreactivity of the /3-cells appears in the control range; the number of strongly reacting cells is reduced, e, Ten days after adrenalectomy the decrease in glucocorticoid receptor immunoreactivity becomes obvious; most nuclei are only slightly immunoreactive. f, Only a few glucocorticoid receptor-immunoreactive cells are left in a pancreatic islet 14 days after adrenalectomy. In other cases, the decrease in immunoreactivity was less pronounced.
were used at dilutions of 1:30 and 1:200, respectively. A polyclonal antibody raised against highly purified rat pancreatic amylase in rabbits was prepared in our laboratory and proved to be monospecific by two-dimensional Western blotting. The antibody against the rat glucocorticoid receptor has been extensively characterized and described previously (13). It recognizes an epitope at the N-terminal portion of the receptor. The secondary antibody was a commercial peroxidase-labeled antimouse IgG (Amersham-Buchler, Braunschweig, West Germany), which was diluted 1:30.
V ^ H S3K* tkl.
?s$t- *
a chromogen and hydrogen peroxide as a substrate resulted in a brown stain. All incubations took place in a humid chamber at 22 C. Sections were then dehydrated and mounted in synthetic resin. Primary antibodies were applied to serial sections to yield a serial sequence of immunoreactions with glucocorticoid receptor, insulin, glucagon, somatostatin, and amylase. Control incubations included 1) substitutions of the respective primary antisera with nonimmune serum, 2) progressive decrease in antibody concentrations and eventually replacement by PBS, and 3) use of irrelevant antibodies (antiprostatic binding protein) (20) as antibody controls.
Immunohistochemistry Specimens from sham-operated as well as adrenalectomized animals (6, 10, and 14 days after surgery) were used. Bouinfixed paraffin-embedded pancreata were serially cut at 4- to 5juni thickness, deparaffinized in xylene, and rehydrated in graded alcohol baths. The indirect immunoperoxidase method was applied. Sections were first treated with a 3% hydrogen peroxide solution, followed by a 10% BSA solution in PBS to block endogenous peroxidase activity. Subsequently, sections were incubated sequentially with normal swine serum, the respective primary antisera, and the pertinent peroxidase-labeled secondary antibodies. Reaction with diaminobenzidine as
Semiquantitatiue evaluation of immunoreactions Photomicrographs of each of the immunoreactions were taken in a Zeiss Axiomat photomicroscope (New York, NY) at a final magnification of about X100. The individual cell types (A-, B-, and D-cells) of the islets of Langerhans were classified according to their staining intensity (light-medium-dark) and counted. Since this method allows no quantitation, an image analysis system (Zeiss IBAS, Kontron Software) was used. This system allows the discrimination of 25 different classes of grey values in photomicrographs. A grey value of 1-10 (class 1) represents black; a grey value of 240-250 (class 25) represents
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
2638
GLUCOCORTICOID RECEPTOR IN PANCREATIC /3-CELLS
spectively, and 2 sham-operated animals per group. About 100300 cells (cytoplasm or nuclei) per islet were marked with a cursor, and the grey values are measured with the IBAS system, using the exocrine pancreas as background. The number of grey values per group was calculated (mean ± SD), expressed as a percentage of the total number of all cells counted, and plotted (see Figs. 3 and 4). Immunoreactions encompassed the glucocorticoid receptor and insulin stains. In the case of the glucocorticoid receptor, only the nuclei were measured, while in case of insulin, only the cytoplasm (excluding the nuclei) was measured.
Glucocorticoid Receptor Immunreactivity Computer assisted quantitation (IBAS) % frequency
Results
10 IS grey values ADX-6d
ADX-10d
ADX-14d
Endo • 1990 Voll26«No5
SHAM
FIG. 3. Semiquantitative evaluation of glucocorticoid receptor immunoreactivity in /3-cells from control and experimental animals. Photomicrographs from pancreatic islets of three animals of each adrenalectomy group (6,10, and 14 days) and 2 animals of each sham operation group (6, 10, and 14 days) processed for glucocorticoid receptor immunoreactivity were used for image analysis (Zeiss IBAS). The nuclei of cells in five islets per group (equalling about 100-300 nuclei/islet) were labeled with a pointer and monitored for staining intensity (expressed as grey values relative to the nuclei of the exocrine tissue that served as the control). The distribution of grey values (representing glucocorticoid receptor immunoreactivity) was expressed as a percentage of the total number of nuclei counted per group. A shift in the grey values was detected only by day 10 in adrenalectomized animals. This shift in grey values, i.e. decrease in glucocorticoid receptor immunoreactivity, was still more pronounced in 14-day adrenalectomized animals. The effect of adrenalectomy was calculated from the maxima of the adrenalectomy plot relative to the sham plot and was about 23.5%. Insulin Immunoreactivity Computer assisted quantitation (IBAS) % frequency
Immunolocalization
of the glucocorticoid receptor
At low magnification, the islets of Langerhans were
the only tissue components of the pancreas that reacted specifically with the glucocorticoid receptor antibody. At higher magnifications, the remnants of adipose tissue also displayed a positive immunoreaction. Since there was an uneven distribution of glucocorticoid receptor immunoreactivity among the endocrine islet cells, sequential staining with antisera against insulin, glucagon, somatostatin, and a-amylase was performed (Fig. 1). /3Cells were immunoreactive with both the insulin antiserum and the glucocorticoid antiserum. In A- and Dcells immunoreactivity with the glucocorticoid receptor was in the background range. At higher antibody concentrations (1:200-1:400), and short incubation periods (1 h at room temperature) the cytoplasm of the /3-cells also reacted with the glucocorticoid receptor antibody. Cytoplasmic staining decreased considerably at higher dilutions (1:1000) of the antibody and prolonged incubation at low temperature (24-48 h at 4 C). All control incubations were negative. Effects of adrenalectomy on immunoreactivity
grey values SHAM-14d
ADX-14d
FIG. 4. Semiquantitative evaluation of insulin immunoreactivity of /?cells in pancreatic islets from control and experimental animals. The approach described in Fig. 3 was used by labeling the cytoplasm (nuclei excluded) of /3-cells by a pointer and determining the staining intensity. Animals 14 days after adrenalectomy were compared to sham-operated animals. The decrease in insulin immunoreactivity, comparing the maxima of the adrenalectomy plot to the sham plot, was 10.4%.
white. A total of 15 islets (5 islets/animals; 3 animals/group) per immunoreaction were evaluated. The groups comprised 3 animals after each adrenalectomy for 6, 10, and 14 days, re-
Comparison of the immunoreactions with the glucocorticoid receptor antibody in sham-operated and adrenalectomized rats clearly resulted in a decrease in nuclear staining intensity of /3-cells in animals 14 days after adrenalectomy. Using a quantitative computer-assisted approach, no significant changes in the glucocorticoid receptor immunoreactivity of /3-cell nuclei was observed in sham-operated animals 6, 10, or 14 days after surgery (Fig. 2, a-c). Contrary to that, in adrenalectomized animals a significant decrease in immunostaining with the glucocorticoid receptor antiserum was observed (Fig. 2, d-f). Comparing 14-day sham-operated with 14-day adrenalectomized animals, a significant 23.5% decrease in staining of /3-cells was calculated (Fig. 3). The quantitative evaluation of insulin immunoreactivity of /3-cells (Fig. 4) in sham-operated (14 days; Fig. 5a) and adrenalectomized (14 days; Fig. 5b) animals resulted in a de-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
GLUCOCORTICOID RECEPTOR IN PANCREATIC tf-CELLS
2639
FlG. 5. Insulin immunoreactivity in pancreatic islets from control and experimental animals, a, In a 14-day shamoperated animal no difference is observed from the control situation (Fig. lb). b, In an animal 14 days after adrenalectomy the staining intensity of the /5-cells is very low; this is obviously partly due to the low avidity of the insulin antibody used.
FIG. 6. Amylase immunoreactivity of the exocrine pancreas from control and experimental animals, a, The exocrine portion of the pancreas is homogeneously stained by a moderately strong immunoreaction in a 14-day sham-operated animal, b, No significant differences in amylase immunoreactivity from the sham-operated situation (a) is observed in the exocrine pancreas from an animal 14 days after adrenalectomy. TABLE 1. Effect of 14-day adrenalectomy (ADX) on body weight, pancreas weight, a-amylase activity, and corticosterone content kDa
BW(g)
Pancreas (g)
Amylase (U/ Mg DNA)
Corticosterone (ng/ml)
SHAM 307 ± 2 1.05 ± 0.1 8.05 ± 1.4 225 ±9.9 ADX 239 ± 3 0.92 ± 0.1 55.7 ± 7.7 4.2 ± 1.1 The relatively high values of corticosterone may be due to the time of death of the rats (1400 h).
- : _ _ • •
W
**
"
•*-
66
crease in staining of /3-cells by about 10.4% (Fig. 4). No obvious changes in the immunoreaction of amylase antiserum with exocrine pancreatic cells were observed (Fig. 6). Biochemical data Table 1 summarizes the experimental data obtained from animals after glucocorticoid deprivation. While the pancreas weight changed only slightly in adrenalectomized rats, body weights decreased by 22% compared to those in sham-operated animals. Amylase content decreased by 50% after 14 days, whereas other exocrine proteins (trypsinogen, chymotrypsinogen, and lipase) remained unchanged (data not shown). The corticosterone content in blood of adrenalectomized animals decreased by 75% compared to that in sham-operated animals. To get more information about the identity of the pancreatic glucocorticoid receptor, cytosolic samples of pancreas and liver homogenates were separated by SDSPAGE, transblotted on nitrocellulose, and incubated with the glucocorticoid receptor antibody (Western blotting). Figure 7 shows the identification of the 94-kD form of the glucocorticoid receptor in both liver and pancreas.
116
45
29
a b c M FIG. 7. Immunoidentification of the glucocorticoid receptor by Western blotting. Rat pancreas and liver cytosolic fractions were prepared as described in Materials and Methods. Cytosolic fractions were separated by SDS-PAGE, transferred to nitrocellulose sheets, and stained with the glucocorticoid receptor antibody, a, Pancreas cytosol (198 ng total protein), b, Liver cytosol (225 fig total protein), c, Purified rat liver cytosolic glucocorticoid receptor (200 ng). M, Marker proteins; GR, The 94-kD form of the glucocorticoid receptor. The 37-kD protein is an immunoreactive fragment of the receptor appearing after freezing and thawing of the samples. Due to proteolytic degradation a 37-kD immunoreactive fragment appeared.
Discussion The presence of glucocorticoid receptor exclusively in nuclei of pancreatic /3-cells has been shown immunohis-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
2640
GLUCOCORTICOID RECEPTOR IN PANCREATIC jS-CELLS
tochemically, while other islet cell types and the exocrine part of the pancreas remained unstained. The hormone dependence of the immunoreactivity has been crosschecked by hormone deprivation experiments. Hormonedependent down-regulation of the glucocorticoid receptor in /3-cells subsequent to corticosteroid deprivation was followed by a decrease in peripheral levels and intracellular concentrations of insulin. Steroid hormone receptors can be detected autoradiographically (5, 21), using hybridization techniques (22), or by immunocytochemistry at both the light (23-25) and electron microscopic levels (26). Our immunohistochemical results in the pancreas are comparable to those in the liver (22) in that the reaction was preferentially, but not exclusively, confined to nuclei (of /3-cells), and the antibody reacted well with Bouin-fixed paraffin-embedded tissue. As has been shown in the liver (22), the intensity of the immunostaining is hormone dependent, i.e. it decreases in adrenalectomized animals. These findings have been confirmed by evaluation of the immunoreaction. The semiquantitative computer-assisted measurement of grey values (representing the immunoreactive material) proved to be effective. It requires, however, a highly standardized immunohistochemical technique so that the staining reaction adequately reflects the true antigen content of the respective structure. While the staining intensity representing the content of /3-cells in immunoreactive insulin in adrenalectomized animals decreased by only 10%, the respective value for the glucocorticoid receptor immunoreaction decreased by about 25%, indicating a downregulation of the glucocorticoid receptor after adrenalectomy. Decreases of 90% in basal insulin serum levels and of 70% in tissue amylase activity after adrenalectomy have been described (27) in normal rats. It was uncertain, however, whether this means a direct or an indirect (via glucocorticoid) action of insulin on amylase synthesis or vice versa. Addition of glucose to the drinking water of adrenalectomized rats increased both plasma insulin and glucose, but had no effect on amylase content, indicating a direct regulatory effect of glucocorticoids (28). Pancreatic acinar cell amylase gene expression has been studied after adrenalectomy and corticosterone replacement (10). It was found that physiological levels of corticosterone have a selective effect on pancreatic amylase gene expression, which may account for the effect of adrenalectomy and replacement on amylase synthesis. Glucocorticoid-induced gene regulation in the azaserineinduced rat pancreatic acinar cell carcinoma resulted in a decrease in amylase (29) or most cellular proteins (30). These data were explained as an indirect effect via inducing regulatory proteins in the cell (29), but it seems more likely that these AR4-2J tumor cells serve as a useful model to study general differentiative effects of
Endo • 1990 Vol 126 • No 5
glucocorticoids (30). Although considerable evidence suggests that insulin is an effective modulator of pancreatic amylase synthesis (7, 10), it is not clear whether there is a direct effect of glucocorticoids on pancreatic amylase gene expression or an indirect effect mediated via an influence of insulin. Our findings of a glucocorticoid-regulated glucocorticoid receptor exclusively present in /3-cells of the islets of Langerhans are clearly in favor of an indirect effect of glucocorticoids on amylase synthesis via a glucocorticoidinsulin-exocrine pathway.
Acknowledgments The authors thank Drs. J. Wager and M. Kirsch, Department of Anatomy, University of Marburg, for advice and assistance in the morphometrical analysis. Mrs. I. Dammshauser for photographic work, M. Dreher for excellent immunocytochemical assistance, Drs. G. Sturm and U. Bieker for corticosterone determinations, and Dr. H. F. Kern, Department of Cell Biology, University of Marburg, for critical reading of the manuscript.
References 1. Grossman MI, Greengard H, Ivy IC 1943 On the mechanism of the adaptation of pancreatic enzymes to dietary composition. Am J Physiol 141:38 2. Soling HD, Unger KG 1972 The role of insulin in the regulation of a-amylase synthesis in the rat pancreas. Eur J Clin Invest 2:199 3. Schick J, Kern HF, Scheele GA 1984 Hormonal stimulation in the exocrine pancreas results in a coordinate and anticoordinate regulation of protein synthesis. J Cell Biol 99:1559 4. Rausch U, Riidiger K, Vasiloudes P, Kern HF, Scheele GA 1986 Lipase synthesis in the rat exocrine pancreas is regulated by secretin. Pancreas 1:522 5. Winborn WB, Sheridan PJ, McGill Jr HC 1987 Localization of progestin receptors in the islets of Langerhans. Pancreas 2:289 6. Logsdon CD, Mossner J, Williams JA, Goldfine I 1985 Glucocorticoids increase amylase mRNA levels, secretory organelles, and secretion in pancreatic acinar AR42J cells. J Cell Biol 100:1200 7. Korc M, Owerbach D, Quinto C, Rutter WJ 1981 Pancreatic isletacinar cell interaction: amylase messenger RNA levels are determined by insulin. Science 213:351 8. Abdeljlil AB, Palla JC, Desnuelle P 1965 Effect of insulin on pancreatic amylase and chymotrypsinogen. Biochem Biophys Res Commun 18:71 9. Mossner J, Logsdon CD, Williams JA, Goldfine ID 1985 Insulin, via its own receptor, regulates growth and amylase synthesis in pancreatic acinar AR 42J cells. Diabetes 34:891 10. Logsdon CD, Akada SF, Meyer C, Dallman MF, Williams JA 1987 Pancreatic acinar cell amylase gene expression: selective effects of adrenalectomy and corticosterone replacement. Endocrinology 121:1242 11. Barseghian G, Levine R, Epps P 1982 Direct effect of cortisol and cortisone on insulin and glucagon secretion. Endocrinology 111:1648 12. Pierluissi J, Navas FO, Ashcroft SJH 1986 Effect of adrenal steroids on insulin release from cultured rat islets of Langerhans. Diabetologia 29:119 13. Westphal HM, Moldenhauer G, Beato M 1982 Monoclonal antibodies to the rat liver glucocorticoid receptor. EMBO J 1:1467 14. Richards GM 1974 Modifications of the diphenylamine reaction giving increased sensitivity and simplicity in the estimation of DNA. Anal Biochem 57:369 15. Bradford MM 1976 A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the the proteindye binding. Anal Biochem 72:248
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
GLUCOCORTICOID RECEPTOR IN PANCREATIC /3-CELLS 16. Bernfeld P 1955 Amylase a and 0. Methods Enzymol 1:149 17. Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680 18. Towbin H, Staehelin T, Gordon J 1979 Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76:4350 19. Johnson DA, Gautsch JW, Sportsman JR, Elder JH 1984 Improved technique utilizing nonfat dry milk for analysis of proteins and nucleic acids transferred to nitrocellulose. Gene Anal Tech 1:3 20. Aumuller G, Seitz J, Heyns W, Flickinger CJ 1982 Intracellular localization of prostatic binding protein (PBP) in rat prostate by light and electron microscopic immunocytochemistry. Histochemistry 76:497 21. Stumpf WE 1971 Autoradiographic techniques for the localization of hormones and drugs at the cellular or subcellular level. Acta Endocrinol (Copenh) 153:205 22. Pfeifer A, Barden N 1988 Glucocorticoid receptor gene expression in rat pituitary gland intermediate lobe following ovariectomy. Mol Cell Endocrinol 55:115 23. Antakly T, Eisen HJ 1984 Immunocytochemical localisation of glucocorticoid receptor in target cells. Endocrinology 115:1984
2641
24. Teasdale J, Lewis FA, Barrett ID, Abbott AC, Wharton J, Bird CC 1986 Immunocytochemical application of monoclonal antibodies to rat liver glucocorticoid receptor. J Pathol 150:227 25. LaFond RE, Kennedy SW, Harrison RW, Villee CA 1988 Immunocytochemical localization of glucocorticoid receptors in cells, cytoplasts, and nucleoplasts. Exp Cell Res 175:52 26. Liposits ZS, Uht RM, Harrison RW, Gibbs FP, Paull WK, Bohn MC 1987 Ultrastructural localization of glucocorticoid receptor (GR) in hypothalamic paraventricular neurons synthesizing corticotropin releasing factor (CRF). Histochemistry 87:407 27. Mossner J, Bohm SK, Secknus R, Fischbach W, Dammrich J, Wunsch PH 1987 Einflu/3 von Glukokortikoiden auf das exokrine Pankreas der Ratte. Verh Dtsch Ges Pathol 71:108 28. Mossner J, Bohm S, Fischbach W 1989 Role of glucocorticosteroids in the regulation of pancreatic amylase synthesis. Pancreas 4:194 29. Logsdon CD, Perot KJ, McDonald AR 1987 Mechanism of glucocorticoid-induced increase in pancreatic amylase gene transcription. J Biol Chem 262:15765 30. Swarovsky B, Steinhilber W, Scheele GA, Kern HF 1988 Coupled induction of exocrine proteins and intracellular compartments involved in the secretory pathway in AR4-2J cells by glucocorticoids. Eur J Cell Biol 47:101
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 21:12 For personal use only. No other uses without permission. . All rights reserved.
