Portal de Periódicos da CAPES

Regulated Apical Secretion of Zymogens in Rat Pancreas

2001; Elsevier BV; Volume: 276; Issue: 17 Linguagem: Inglês

10.1074/jbc.m006221200

1083-351X

Katja Schmidt, Michael Schrader, Horst-Franz Kern, Ralf Kleene,

Lysosomal Storage Disorders Research

We examined the role of glycosphingolipid- and cholesterol-enriched microdomains, or rafts, in the sorting of digestive enzymes into zymogen granules destined for apical secretion and in granule formation. Isolated membranes of zymogen granules from pancreatic acinar cells showed an enrichment in cholesterol and sphingomyelin and formed detergent-insoluble glycolipid-enriched complexes. These complexes floated to the lighter fractions of sucrose density gradients and contained the glycosylphosphatidylinositol (GPI)-anchored glycoprotein GP-2, the lectin ZG16p, and sulfated matrix proteoglycans. Morphological and pulse-chase studies with isolated pancreatic lobules revealed that after inhibition of GPI-anchor biosynthesis by mannosamine or the fungal metabolite YW 3548, granule formation was impaired leading to an accumulation of newly synthesized proteins in the Golgi apparatus and the rough endoplasmic reticulum. Furthermore, the membrane attachment of matrix proteoglycans was diminished. After cholesterol depletion or inhibition of glycosphingolipid synthesis by fumonisin B1, the formation of zymogen granules as well as the formation of detergent-insoluble complexes was reduced. In addition, cholesterol depletion led to constitutive secretion of newly synthesized proteins,e.g. amylase, indicating that zymogens were missorted. Together, these data provide first evidence that in polarized acinar cells of the exocrine pancreas GPI-anchored proteins, e.g.GP-2, and cholesterol-sphingolipid-enriched microdomains are required for granule formation as well as for regulated secretion of zymogens and may function as sorting platforms for secretory proteins destined for apical delivery. We examined the role of glycosphingolipid- and cholesterol-enriched microdomains, or rafts, in the sorting of digestive enzymes into zymogen granules destined for apical secretion and in granule formation. Isolated membranes of zymogen granules from pancreatic acinar cells showed an enrichment in cholesterol and sphingomyelin and formed detergent-insoluble glycolipid-enriched complexes. These complexes floated to the lighter fractions of sucrose density gradients and contained the glycosylphosphatidylinositol (GPI)-anchored glycoprotein GP-2, the lectin ZG16p, and sulfated matrix proteoglycans. Morphological and pulse-chase studies with isolated pancreatic lobules revealed that after inhibition of GPI-anchor biosynthesis by mannosamine or the fungal metabolite YW 3548, granule formation was impaired leading to an accumulation of newly synthesized proteins in the Golgi apparatus and the rough endoplasmic reticulum. Furthermore, the membrane attachment of matrix proteoglycans was diminished. After cholesterol depletion or inhibition of glycosphingolipid synthesis by fumonisin B1, the formation of zymogen granules as well as the formation of detergent-insoluble complexes was reduced. In addition, cholesterol depletion led to constitutive secretion of newly synthesized proteins,e.g. amylase, indicating that zymogens were missorted. Together, these data provide first evidence that in polarized acinar cells of the exocrine pancreas GPI-anchored proteins, e.g.GP-2, and cholesterol-sphingolipid-enriched microdomains are required for granule formation as well as for regulated secretion of zymogens and may function as sorting platforms for secretory proteins destined for apical delivery. In endocrine and exocrine cells secretory proteins are stored at high concentrations in secretory granules and are released in response to external stimuli. During the course of granule formation proteins destined for storage and regulated secretion are segregated from those constitutively secreted (1Tooze S.A. Biochim. Biophys. Acta. 1998; 1404: 231-244Crossref PubMed Scopus (185) Google Scholar, 2Huttner W.B. Ohashi M. Kehlenbach R.H. Barr F.A. Bauerfeind R. Braunling O. Corbeil D. Hannah M. Pasolli H.A. Schmidt A. Cold Spring Harbor Symp. Quant. Biol. 1995; 60: 315-327Crossref PubMed Scopus (45) Google Scholar, 3Arvan P. Castle D. Biochem. J. 1998; 332: 593-610Crossref PubMed Scopus (444) Google Scholar). Granule formation and apical sorting in pancreatic acinar cells involves the selective aggregation of a mixture of regulated secretory proteins and the association of these aggregates with specific membrane domains of the trans-Golgi network (TGN),1 which then pinch off as condensing vacuoles. While the selective aggregation of the pancreatic secretory proteins has been well documented (4Leblond F.A. Viau G. Laine J. Lebel D. Biochem. J. 1993; 291: 289-296Crossref PubMed Scopus (63) Google Scholar, 5Freedman S.D. Scheele G.A. Biochem. Biophys. Res. Commun. 1993; 197: 992-999Crossref PubMed Scopus (41) Google Scholar, 6Dartsch H. Kleene R. Kern H.F. Eur. J. Cell Biol. 1998; 75: 211-222Crossref PubMed Scopus (36) Google Scholar), how these aggregates become bound to the luminal side of the TGN membrane and are ultimately sorted into condensing vacuoles is poorly understood. The zymogen granule (ZG) glycoprotein GP-2, which is linked to the luminal surface via a glycosylphosphatidylinositol (GPI) anchor and represents up to 30% of the total membrane proteins within the ZG membrane, was the first candidate postulated to mediate this sorting event (7Scheele G.A. Fukuoka S. Freedman S.D. Pancreas. 1994; 9: 139-149Crossref PubMed Scopus (55) Google Scholar). However, two studies have demonstrated that during embryonic development and in partially differentiated acinar carcinoma cell lines, granule formation can occur in the complete absence of GP-2 (8Hansen L.J. Reddy M.K. Reddy J.K. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 4379-4383Crossref PubMed Scopus (18) Google Scholar, 9Dittie A. Kern H.F. Eur. J. Cell Biol. 1992; 58: 243-258PubMed Google Scholar, 10Laine J. Pelletier G. Grondin G. Peng M. LeBel D. J. Histochem. Cytochem. 1996; 44: 481-499Crossref PubMed Scopus (9) Google Scholar). This argues against a role as specific sortase of GP-2, but does not exclude its involvement in granule formation and in apical sorting (7Scheele G.A. Fukuoka S. Freedman S.D. Pancreas. 1994; 9: 139-149Crossref PubMed Scopus (55) Google Scholar), at least in the adult exocrine pancreas. Scheele et al. (7Scheele G.A. Fukuoka S. Freedman S.D. Pancreas. 1994; 9: 139-149Crossref PubMed Scopus (55) Google Scholar) have put forward the hypothesis that GP-2, together with attached proteoglycans, forms a submembranous matrix at the luminal side of ZGM. We have recently identified and characterized proteoglycan and glycoprotein components of this matrix (11Schmidt K. Dartsch H. Linder D. Kern H. Kleene R. J. Cell Sci. 2000; 113: 2233-2242PubMed Google Scholar) and have evidence that it functions in the binding of aggregated enzyme proteins to membranes, a process that is considered as condensation-sorting (12Kleene R. Dartsch H. Kern H.F. Eur. J. Cell Biol. 1999; 78: 79-90Crossref PubMed Scopus (37) Google Scholar). The acinar cells of the pancreas are highly polarized cells with regulated secretion directed to the apical domain surrounding the acinus lumen. From studies of polarized protein trafficking performed mainly with MDCK cells, it is known that protein sorting to the apical domain of the plasma membrane is controlled by N- and/orO-glycosylation of protein ectodomains (13Fiedler K. Simons K. Cell. 1995; 81: 309-312Abstract Full Text PDF PubMed Scopus (275) Google Scholar, 14Rodriguez-Boulan E. Gonzalez A. Trends Cell Biol. 1999; 9: 291-294Abstract Full Text Full Text PDF PubMed Scopus (147) Google Scholar), by apical sorting determinants present in the cytoplasmic tail of seven transmembrane proteins (15Chuang J.Z. Sung C.H. J. Cell Biol. 1998; 142: 1245-1256Crossref PubMed Scopus (134) Google Scholar, 16Sun A.Q. Ananthanarayanan M. Soroka C.J. Thevananther S. Shneider B.L. Suchy F.J. Am. J. Physiol. 1998; 275: G1045-G1055PubMed Google Scholar), or by the incorporation of apically sorted proteins into lipid microdomains, called rafts, in the Golgi complex (17Harder T. Simons K. Curr. Opin. Cell Biol. 1997; 9: 534-542Crossref PubMed Scopus (716) Google Scholar, 18Simons K. Ikonen E. Nature. 1997; 387: 569-572Crossref PubMed Scopus (8050) Google Scholar). Inclusion into rafts and subsequent apical sorting has been shown for GPI-anchored proteins at the plasma membrane (17Harder T. Simons K. Curr. Opin. Cell Biol. 1997; 9: 534-542Crossref PubMed Scopus (716) Google Scholar) and for some transmembrane proteins (19Kundu A. Avalos R.T. Sanderson C.M. Nayak D.P. J. Virol. 1996; 70: 6508-6515Crossref PubMed Google Scholar, 20Lin S. Naim H.Y. Rodriguez A.C. Roth M.G. J. Cell Biol. 1998; 142: 51-57Crossref PubMed Scopus (164) Google Scholar). Rafts are membrane microdomains enriched in glycosphingolipids, sphingomyelin, and cholesterol that are thought to be assembled within the exoplasmic leaflet of the Golgi membrane and that have been proposed to act as a sorting platform for the apical delivery of plasma membrane proteins (18Simons K. Ikonen E. Nature. 1997; 387: 569-572Crossref PubMed Scopus (8050) Google Scholar, 21Simons K. van Meer G. Biochemistry. 1988; 27: 6197-6202Crossref PubMed Scopus (1083) Google Scholar). Rafts are defined biochemically by their insolubility in non-ionic detergents such as Triton X-100 at 4 °C (22Brown D.A. Rose J.K. Cell. 1992; 68: 533-544Abstract Full Text PDF PubMed Scopus (2604) Google Scholar). The detergent-insoluble glycosphingolipid complexes (DIGs) float to lighter fractions on sucrose density gradients and have been identified in TGN-derived apical transport vesicles (22Brown D.A. Rose J.K. Cell. 1992; 68: 533-544Abstract Full Text PDF PubMed Scopus (2604) Google Scholar). Several apical cargo proteins, e.g. the influenza virus proteins hemagglutinin and neuraminidase (19Kundu A. Avalos R.T. Sanderson C.M. Nayak D.P. J. Virol. 1996; 70: 6508-6515Crossref PubMed Google Scholar, 20Lin S. Naim H.Y. Rodriguez A.C. Roth M.G. J. Cell Biol. 1998; 142: 51-57Crossref PubMed Scopus (164) Google Scholar), or GPI-anchored proteins (17Harder T. Simons K. Curr. Opin. Cell Biol. 1997; 9: 534-542Crossref PubMed Scopus (716) Google Scholar) are found in DIGs during their transport to the apical domain, while basolaterally delivered cargo proteins are excluded from DIGs (23Keller P. Simons K. J. Cell Biol. 1998; 140: 1357-1367Crossref PubMed Scopus (470) Google Scholar). Raft association of proteins is mediated either directly by the GPI moiety or by the transmembrane domains of proteins, e.g. the lectin VIP36, or indirectly by binding to other raft-associated proteins. It is not clear which of the above mechanisms exist in polarized acinar cells of the pancreas to sort the 15–20 different zymogens to the apical surface and what role the GPI-anchored GP-2 and lipid rafts play in this phenomenon. As lipid microdomains have been described in secretory granules of neuroendocrine cells (24Thiele C. Huttner W.B. Semin. Cell Dev. Biol. 1998; 9: 511-516Crossref PubMed Scopus (43) Google Scholar, 25Dhanvantari S. Loh Y.P. J. Biol. Chem. 2000; 275: 29887-29893Abstract Full Text Full Text PDF PubMed Scopus (108) Google Scholar), we investigated whether they were also present in the membranes of zymogen granules of the exocrine pancreas. We obtained evidence that GP-2, the secretory lectin ZG16p, and sulfated proteoglycans were associated with cholesterol-glycosphingolipid-enriched microdomains isolated from membranes of zymogen granules. Furthermore, GP-2 and lipid microdomains were required for proper granule formation and regulated secretion of zymogens. Polyclonal antibodies to rat GP-2 and pig amylase were used as reported previously (9Dittie A. Kern H.F. Eur. J. Cell Biol. 1992; 58: 243-258PubMed Google Scholar). Polyclonal antibody ZG16 to recombinant rat ZG16p was described in Cronshagen et al. (26Cronshagen U. Voland P. Kern H.F. Eur. J. Cell Biol. 1994; 65: 366-377PubMed Google Scholar). Polyclonal antibodies to carboxyl ester lipase were raised in chicken. A polyclonal antibody to carboxypeptidase α was obtained from Rockland Immunochemicals (Gilbertsville, PA). Species-specific anti-IgG antibodies conjugated to horseradish peroxidase were obtained from Bio-Rad. ZG were isolated as described previously (6Dartsch H. Kleene R. Kern H.F. Eur. J. Cell Biol. 1998; 75: 211-222Crossref PubMed Scopus (36) Google Scholar). Briefly, the pancreas was removed from fasted male Wistar rats (200–230 g; Charles River, Sulzfeld, Germany) and homogenized on ice in the following buffer: 0.25 m sucrose, 5 mmMES at pH 6.25, 0.1 mm MgSO4, 1 mmdithiothreitol, 10 μm FOY-305 (Sanol Schwarz, Mannheim, Germany), 2.5 mm Trasylol (Bayer, Leverkusen, Germany), 0.1 mm phenylmethylsulfonyl fluoride. The homogenate was centrifuged at 500 × g for 10 min at 4 °C, and the resulting postnuclear supernatant was further centrifuged at 2000 × g for 10 min at 4 °C to pellet the ZG. After removal of the brownish layer of mitochondria on top of the pellet, the ZG were lysed in 50 mm Hepes, pH 8.0, and stored at −20 °C. After thawing, the soluble zymogen granule content proteins were separated from the ZGM by ultracentrifugation (100,000 × g for 30 min, 4 °C). ZGM were resuspended in 50 mm Hepes, pH 8.0, and stored at −20 °C. For bicarbonate treatment of ZGM, 500 μl of 300 mm NaHCO3, pH 11.5, were added to 500 μl of ZGM (corresponding to 200 μg of protein), and the mixture was incubated on ice for 2 h. For PI-PLC treatment of ZGM, 500 milliunits of PI-PLC (Roche Molecular Biochemicals, Mannheim, Germany) were added to ZGM (corresponding to 800 μg of protein), and the samples were incubated at 37 °C for 1 h in a final volume of 500 μl. Treated membranes were reisolated by high-speed centrifugation at 100,000 × g for 30 min. For the preparation of Golgi- and rER-enriched fractions, the postnuclear supernatant was adjusted to 0.5 m sucrose, layered onto a step gradient of 1.4 and 0.9 m sucrose, and was overlaid with 0.25 m sucrose solution containing 10 mmHepes, pH 6.8, 5 mm MgCl2, and 3 mmCaCl2. After centrifugation for 1 h at 110,000 ×g, a Golgi-enriched faction was obtained at the 0.5/0.9m interphase together with a rER-enriched fraction at the 0.9 m/1.4 m interphase. The collected fractions were adjusted to 0.2 m sucrose with 50 mmHepes, pH 8.0, and stored at −20 °C. After thawing and centrifugation at 100,000 × g for 30 min at 4 °C, the supernatants containing the soluble Golgi or rER content proteins and the pellets containing Golgi (GM) or rER membranes (rERM) were stored at −20 °C in aliquots for further use. Isolated membranes corresponding to 400 μg of protein were resuspended in 100 μl of 50 mm Hepes, pH 8.0, and mixed with 200 μl of methanol/chloroform (1:1). After Bligh-Dyer two-phase extraction (27New R.R.C. Rickwood D. Hames B.D. Liposomes-A Practical Approach. IRL Press, Oxford1990Google Scholar) lipids were separated by thin layer chromatography (TLC) using a mixture of chloroform:methanol:H2O (65:25:4). To extract polar lipids (28Baldoni E. Bolognani L. Vitaioli L. Eur. J. Histochem. 1995; 39: 253-257PubMed Google Scholar) membranes corresponding to 2 mg of protein were resuspended in 20 μl of 50 mm Hepes, pH 8.0, and incubated with 500 μl of methanol and 1 ml of n-hexane. After centrifugation at 3000 × g for 5 min, 1 ml of chloroform was added to the methanol phase, and centrifugation was repeated. The resulting supernatant was submitted to TLC using a mixture of chloroform:methanol:0.25% KCl (50:54:10). Lipids were visualized by 20% (v/v) H2SO4 at 150 °C and quantitated. Cholesterol, glucosyl ceramide, lactosyl ceramide, sulfatides, phosphatidylcholine, and sphingomyelin were used as standards. 500 μg of membranes were incubated on ice for 30 min in 50 mm Hepes, pH 8.0, containing 1% (v/v) Triton X-100 and were centrifuged at 120,000 × g for 30 min. The Triton X-100-insoluble pellet fraction was resuspended in 50 mm Hepes, pH 8.0, and adjusted to 1.2 m sucrose. 500 μl of the pellet fraction were overlayed with 1 ml of 1.1 m and 0.5 ml of 0.15m sucrose. After centrifugation for 1 h at 120,000 × g in a swing out rotor (TLA 120.1, Beckman Instruments, Munich, Germany) fractions of 200 μl were collected from the top of the gradient. For flotation experiments with ZGM and GM, membranes were bicarbonate-treated prior to Triton X-100 extraction to partially remove membrane-associated matrix components which were found to inhibit flotation. Proteins were precipitated by methanol (29Wessel D. Flugge U.I. Anal. Biochem. 1984; 138: 141-143Crossref PubMed Scopus (3141) Google Scholar) and analyzed by gel electrophoresis. Lipids were extracted by Bligh-Dyer two-phase extraction (27New R.R.C. Rickwood D. Hames B.D. Liposomes-A Practical Approach. IRL Press, Oxford1990Google Scholar) and analyzed by TLC. Using an orogastric tube rats were fed with a single dose of 50 mg/kg of FOY-305 (Sanol-Schwarz, Mohnheim, Germany) dissolved in 1 ml of tap water and were sacrificed 6–9 h after feeding. FOY treatment for up to 6 h results in partial degranulation (40–60%) of the pancreas (30Rausch U. Adler G. Weidenbach H. Weidenbach F. Rudolff D. Koop I. Kern H.F. Cell Tissue Res. 1987; 247: 187-193Crossref PubMed Scopus (40) Google Scholar). Pancreatic lobules were prepared as described previously (31Scheele G.A. Palade G.E. J. Biol. Chem. 1975; 250: 2660-2670Abstract Full Text PDF PubMed Google Scholar). The lobules were incubated at 37 °C under agitation and were supplied with 100% O2 every 15 min. For cholesterol depletion the lobules were incubated for 2.5 h in 5 ml MEM (Sigma, Munich, Germany) containing 25 mm mevalonate acid-lactone (Sigma) and 0.4 mm lovastatin (Merck Sharp & Dohme, Haar, Germany) and thereafter treated with 10 mm methyl-β-cyclodextrin (CD) for 30 min (Sigma). For inhibition of GPI-anchor biosynthesis lobules were incubated for 3 h either in 5 ml of glucose-deficient MEM (Life Technologies, Inc.) containing 50 mm d-mannosamine (Sigma) or were first incubated in 5 ml of MEM for 2.5 h and afterwards in MEM containing 5 μg/ml YW 3548 (Novartis, Basel, Switzerland) for an additional 30 min. Ceramide biosynthesis was inhibited by treatment with 28 μmfumonisin B1 for 3 h. Treated lobules and controls were either used for electron microscopy or for pulse-chase experiments. Drug-treated lobules and controls were washed twice with Met/Cys-deficient medium (Life Technologies, Inc.) and were pulse-labeled for 10 min at 37 °C in 5 ml of pulse medium containing 200 μCi/ml Tran35S-label (ICN, Eschwege, Germany) and the appropriate drugs (mevalonate/lovastatin, fumonisin B1, mannosamine, or YW 3548). After washing in MEM the lobules were incubated for different chase times in 5 ml of chase medium (MEM containing 10 × Met/Cys and the appropriate drugs). Exocytosis of zymogen granules was hormonally induced by the addition of 1 ng/ml caerulein, 25 μmcarbamoylcholine chloride, 130 nm phorbol 12-myristate 13-acetate (Sigma), and 1 mm cyclic 8-bromo-adenosine-3′:5′-monophosphate (Roche Molecular Biochemicals) to the chase medium. For determination of secreted amylase the proteins of 200–500 μl of chase medium were precipitated by trichloroacetic acid, resuspended in the same volume of Laemmli buffer, and separated by SDS-PAGE. Quantitation of radiolabeled amylase was performed on the gels using a Bio-imager FUJIX BAS 1500 and TINATM software, version 2.0 (Raytest, Straubenhardt, Germany). Total amylase was quantitated after Coomassie staining or immunoblotting of the corresponding gels. Alternatively, amylase activity in the medium was assayed according to Ref. 32Rick W. Stegbauer H.P. Bergmeyer H.U. Methods of Enzymatic Analysis. Verlag Chemie, Weinheim, Germany1974: 149-158Google Scholar.35S-Amylase secretion after different experimental conditions was determined in relation to total normalized amylase secretion. Similar results were obtained when total amylase was either determined by quantitation of gels or by enzyme activity. In Fig. 5data obtained after quantitation of gels are shown. Lactate dehydrogenase activity was assayed using commercially available test kits (Roche Molecular Biochemicals) and expressed as percent of total lactate dehydrogenase activity in the medium and cell homogenate. Trypsin activity was measured according to Ref. 33Hummel B.C. Can. J. Biochem. 1959; 37: 1393-1399Crossref PubMed Google Scholar. For determination of protein-bound radioactivity, samples were applied to 3MM filters (Whatman, Maidstone, United Kingdom). The filters were swirled in 10% trichloroacetic acid on ice for 30 min and washed with ice-cold 5% trichloroacetic acid and ethanol. After drying of the filters the radioactivity was determined by liquid scintillation counting (Raytest, Straubenhardt, Germany). Protein synthesis and incorporation of radioactivity was reduced after cholesterol depletion (70% of control levels) and after treatment with mannosamine (60% of control levels). However, the ratios of the total protein-bound radioactivity present in the ZG, rER, and Golgi subfractions to total radioactivity incorporated in the postnuclear supernatant were similar after all treatments, indicating that protein delivery to the secretory pathway was not disturbed. Therefore, the protein-bound radioactivity of the individual subfractions was expressed as percent of total protein-bound radioactivity of all subfractions (Σ cpm ZG, rER, Golgi) (Table I).Table IEffects of inhibition of GPI-anchor biosynthesis and cholesterol depletion on de novo granule formationInhibitionMock (control lobules)Cholesterol biosynthesis (extraction by CD) (lovastatin/mevalonate, CD)Ceramide biosynthesis (fumonisin B1)GPI-anchor biosynthesis (N-glycosylation) (mannosamine)GPI-anchor biosynthesis (YW 3548)cpm (% of total), mean ± S.D.ZG24.4 ± 7.15.4 ± 0.910 ± 2.81.2 ± 0.214.8 ± 0.8Golgi70.1 ± 8.595 ± 0.785.8 ± 3.285 ± 0.276.5 ± 2.7rER4.9 ± 1.34.2 ± 0.34.3 ± 0.414 ± 0.38.5 ± 2.6Pancreatic lobules were prepared from rats fed with a single dose of FOY-305 leading to hormonally stimulated release of zymogen granules. Degranulated lobules were incubated in the absence (control lobules) or in the presence of mannosamine or the terpenoid lactone YW 3548, two potent inhibitors of GPI-anchor biosynthesis. To disrupt lipid rafts, degranulated lobules were incubated for 2.5 h in the presence of mevalonate/lovastatin followed by incubation with CD for additional 30 min to deplete intracellular cholesterol levels or were treated with fumonisin B1, an inhibitor of glycosphingolipid synthesis. After pulse labeling with Tran35S label, Golgi- and rER-enriched fractions as well as ZG were isolated after 30-min chase. The protein-bound radioactivity of the subfractions was determined and expressed as percent of total (Σ cpm ZG, rER, Golgi). The data are from three to four independent experiments and are expressed as mean ± S.D. Open table in a new tab Pancreatic lobules were prepared from rats fed with a single dose of FOY-305 leading to hormonally stimulated release of zymogen granules. Degranulated lobules were incubated in the absence (control lobules) or in the presence of mannosamine or the terpenoid lactone YW 3548, two potent inhibitors of GPI-anchor biosynthesis. To disrupt lipid rafts, degranulated lobules were incubated for 2.5 h in the presence of mevalonate/lovastatin followed by incubation with CD for additional 30 min to deplete intracellular cholesterol levels or were treated with fumonisin B1, an inhibitor of glycosphingolipid synthesis. After pulse labeling with Tran35S label, Golgi- and rER-enriched fractions as well as ZG were isolated after 30-min chase. The protein-bound radioactivity of the subfractions was determined and expressed as percent of total (Σ cpm ZG, rER, Golgi). The data are from three to four independent experiments and are expressed as mean ± S.D. Protein samples were separated by SDS-PAGE (12.5–15% acrylamide gels) according to Laemmli (34Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (206615) Google Scholar), transferred to nitrocellulose (Schleicher and Schüll, Dassel, Germany) using a semidry apparatus (Bio-Rad), and analyzed by immunoblotting. Immunoblots were processed using either enhanced chemiluminescence reagents (Amersham Pharmacia Biotech) or diaminobenzidine reagent (0.05% diaminobenzidine, 0.04% CoCl3, 0.05% H2O2 in 10 mm Tris-HCl, pH 7.5). Silver staining of gels was performed according to Hempelmann (35Hempelmann E. Kaminsky R. Electrophoresis. 1986; 7: 481Crossref Scopus (27) Google Scholar). For quantitation immunoblots were scanned and processed using PC bas software. Sulfated glycosaminoglycans (GAGs) were detected using the Blyscan assay according to the manufacturer's protocol (Biocolor, Belfast, Ireland). Briefly, a precipitable complex was formed by binding of the Blyscan dye to free GAGs or those bound to proteoglycans. After dissociation of the precipitated complex the samples were photometrically quantified, or in the case of radiolabeled material, samples were analyzed by liquid scintillation counting. Pancreatic lobules were fixed in 0.1% cacodylate buffer, pH 7.3, containing 1% glutaraldehyde (Serva, Heidelberg, Germany), postfixed in 1% osmium tetroxide, and embedded in Epon according to standard procedures. Thin sections (70 nm) were stained with uranyl acetate/lead citrate and examined using a Zeiss EM 9S electron microscope. For immunolabeling thin sections of Epon-embedded samples were etched with 10% H2O2 for 5 min, blocked with 1% bovine serum albumin in phosphate-buffered saline, and afterward incubated with an antibody to carboxypeptidase α and visualized using 10 nm protein A-gold. This antibody showed specific labeling of ZG and Golgi cisternae, but was not found to label other intracellular organelles. Quantitative analysis of the intactness of junctional complexes at the apical pole, the integrity of the apical and intracellular membranes (ER, mitochondria, granules), and of morphological alterations of the Golgi complex after different experimental treatments was performed on photographs at ×3000–12,000 (Table II). Sections were systematically scanned at the electron microscope, and each acinus encountered was photographed. A cell (acinus) was defined as intact (polarized) if the apico-lateral junctions and the plasma membrane showed no sign of disruption and if the ZG were still concentrated apically. Furthermore, the integrity of other internal organelles (see above) was comprised.Table IIQuantitative analysis of ultrastructural dataControlChol. dep.Fumonisin B1MannosamineYW 3548% of totalJunctional complexes + apical membrane intact92.9 ± 5.583.1 ± 693.3 ± 1.387.5 ± 7.790.1 ± 6.8n = 510n = 372n = 298n = 325n = 325Intracellular membranes intact95 ± 2.287 ± 3.694 ± 2.289.8 ± 7.392.8 ± 4.3n = 582n = 396n = 336n = 387n = 348Golgi cisternae altered31.3 ± 5.472.5 ± 8.775.3 ± 7.772.3 ± 3.679.2 ± 9.4n = 294n = 257n = 204n = 144n = 260Pancreatic lobules were prepared and incubated with the different inhibitors as described in Table I. Lobules were processed for electron microscopy afterwards, embedded in Epon, and sectioned. The intactness of the junctional complexes at the apical pole, the integrity of the apical and intracellular membranes (ER, mitochondria, granules), and morphological alterations of the Golgi cisternae of photographed cells were quantitated as described under "Experimental Procedures" and expressed as percent of total (mean ± S.D.). Cholesterol depletion or fumonisin B1 treatment caused dilation of individual Golgi cisternae, whereas mannosamine or YW 3548 treatment resulted in the accumulation of vesicles (see Figs. 2 and 6). For the control all unusual alterations of the Golgi complex were counted. The data for each experimental condition are from three to four individual experiments. Chol. dep. = cholesterol depletion; n = number of cells or Golgi complexes. Open table in a new tab Pancreatic lobules were prepared and incubated with the different inhibitors as described in Table I. Lobules were processed for electron microscopy afterwards, embedded in Epon, and sectioned. The intactness of the junctional complexes at the apical pole, the integrity of the apical and intracellular membranes (ER, mitochondria, granules), and morphological alterations of the Golgi cisternae of photographed cells were quantitated as described under "Experimental Procedures" and expressed as percent of total (mean ± S.D.). Cholesterol depletion or fumonisin B1 treatment caused dilation of individual Golgi cisternae, whereas mannosamine or YW 3548 treatment resulted in the accumulation of vesicles (see Figs. 2 and 6). For the control all unusual alterations of the Golgi complex were counted. The data for each experimental condition are from three to four individual experiments. Chol. dep. = cholesterol depletion; n = number of cells or Golgi complexes. Significant differences between experimental groups were detected by analysis of variance for unpaired variables using SigmaStat for Windows (Statistical Products and Service Solutions Inc., Chicago, IL). Data are presented as mean ± S.D., with an unpaired t test used to determine statistical differences. p values <0.05 are considered as significant, and p values <0.01 are considered as highly significant. Glycosphingolipid-cholesterol rafts form Triton X-100-insoluble glycolipid-rich complexes (DIGs) at 4 °C, which float to lighter fractions on sucrose density gradients (17Harder T. Simons K. Curr. Opin. Cell Biol. 1997; 9: 534-542Crossref PubMed Scopus (716) Google Scholar). To identify DIGs and possible DIG-associated proteins in membranes of isolated ZG, we extracted bicarbonate-treated ZGM with Triton X-100 at 4 °C and analyzed the detergent-insoluble fraction by density gradient centrifugation (Fi