Linkage of the ubiquitin-conjugating system and the endocytic pathway in ligand-induced internalization of the growth hormone receptor

The major function of the ubiquitin-conjugating system is the targeting of cytosolic and nuclear proteins for degradation by the proteasome. Recently, ubiquitin conjugation has been implicated in the downregulation of signalling receptors such as the mammalian growth hormone receptor (GHR) and the alpha-factor receptor in yeast. By examining truncated receptors, the internalization-deficient receptor mutant F327A and conditions under which clathrin-mediated GHR endocytosis is inhibited, we show here that GHR ubiquitination and ligand-induced GHR internalization are coupled events. Previously, we had shown that GHR endocytosis is dependent on an intact ubiquitination system. Here we present evidence that GHR ubiquitination depends on an intact endocytic pathway. Our data indicate that the ubiquitin-conjugating system and the endocytic pathway interact at the cytoplasmic tail of the GHR at the plasma membrane, where they cooperate to regulate internalization of the GHR.

Epidermal growth factor induces ubiquitination of Eps15

Epidermal growth factor (EGF) receptor pathway substrate clone 15 (Eps15) has been described as a 142-kDa EGF receptor substrate. It has been shown to bind to the EGF receptor, adaptor protein-2, and clathrin and is present at clathrin-coated pits and vesicles. Upon stimulation of cells with EGF or transforming growth factor alpha, Eps15 becomes rapidly and transiently phosphorylated on tyrosine residues. This phosphorylation coincides with an increase of 8 kDa in molecular mass. Here we show that this increase in molecular mass is not due to tyrosine phosphorylation. Instead, we found both by Western blotting and protein sequencing that this EGF-induced increase in molecular mass is the result of monoubiquitination. Eps15 ubiquitination but not tyrosine phosphorylation was inhibited under conditions that blocked EGF-induced internalization of the EGF receptor. Our results establish ubiquitination as a second form of EGF-stimulated covalent modification of Eps15.

Growth hormone-induced signal tranduction depends on an intact ubiquitin system

The growth hormone receptor (GHR) is a ubiquitinated cell surface protein. Ligand binding and receptor dimerization activate the cytosolic kinase Jak2. This event initiates signal transduction via STAT proteins. Expression of GHR in a Chinese hamster ovary (CHO) cell line, which exhibits a temperature-sensitive defect in ubiquitin conjugation (CHO-ts20), as well as in wild type cells (CHO-E36) has shown that endocytosis of the receptor requires an intact ubiquitin conjugation system (Strous G. J., van Kerkhof, P., Govers, R., Ciechanover A., and Schwartz, A. L. (1996) EMBO J. 15, 3806-3812). We have now examined the requirement for ubiquitin conjugation in growth factor-mediated signal transduction. In CHO-E36 and in CHO-ts20 cells at the permissive temperature, STAT proteins were activated in a growth factor-dependent fashion. However, no activation of STAT proteins was observed at the nonpermissive temperature in CHO-ts20 cells. Neither tyrosine phosphorylation of GHR nor of Jak2 was inhibited at the nonpermissive temperature. When tyrosine phosphorylation was inhibited following treatment with staurosporin, ubiquitination of the receptor proceeded normally. Furthermore, mutation of GHR phenylalanine-327, which prevents GHR endocytosis, inhibited receptor ubiquitination but allowed normal Jak/STAT-mediated signal transduction. Thus, these data provide evidence that the ubiquitin conjugation system is involved in the Jak/STAT signaling pathway, be it not at the initial stage(s) of Jak2 activity.

Mouse hepatitis virus strain A59 is released from opposite sides of different epithelial cell types

Coronaviruses infect humans and animals through epithelial cells of the gastrointestinal and respiratory tracts that serve as their primary target. When studying infections in cultured polarized epithelial cells, we found previously that coronaviruses are released from specific plasma-membrane domains; thus, mouse hepatitis virus (strain A59; MHV-A59) leaves murine epithelial kidney cells from the basolateral surface, whereas release of transmissible gastroenteritis virus from porcine epithelial kidney cells is confined to the apical membrane. This observation begged the question whether a particular coronavirus is consistently shed through the same membrane, irrespective of the nature of the epithelial cell. We therefore extended our studies with MHV-A59 to Madin-Darby canine kidney (MDCK) strain I and human colon carcinoma (Caco-2) cells, both of which are naturally refractory to MHV-A59 but were made susceptible to infection by transfection with recombinant MHV receptor cDNA. The release of MHV-A59 from Caco(MHVR) cells occurred preferentially from the basolateral side, consistent with our previous observations. In contrast, release from MDCK(MHVR) cells occurred almost exclusively from the apical surface. Because of this difference, we studied MHV-A59 infection of MDCK(MHVR) cells in more detail. The virus entered the cells preferentially from the apical side, a situation similar to that in murine epithelial cells, where the highest density of MHV receptor glycoprotein was found. The results from this and previous studies show that targeting of vesicles containing MHV-A59 to a specific side of epithelial cells may vary in different epithelial cell types.

A murine and a porcine coronavirus are released from opposite surfaces of the same epithelial cells

Epithelial cells are important target cells for coronavirus infection. Earlier we have shown that transmissible gastroenteritis coronavirus (TGEV) and mouse hepatitis coronavirus (MHV) are released from different sides of porcine and murine epithelial cells, respectively. To study the release of these viruses from the same cells, we constructed a porcine LLC-PK1 cell line stably expressing the recombinant MHV receptor cDNA (LMR cells). The MHV and TGEV receptor glycoproteins were shown by immunofluorescence to appear at the surface of the cells and to be functional so that the cells were susceptible to both MHV and TGEV infection. Both coronaviruses entered polarized LMR cells only through the apical surface. Remarkably, while the cells remained susceptible to TGEV for long periods, infectability by MHV decreased with time after plating of the cells onto filters. This was not due to a lack of expression of the MHV receptor, since this glycoprotein was still abundant on the apical surface of these cells. TGEV and MHV appeared to exit LMR cells from opposite sides. Whereas TGEV was released preferentially at the apical membrane, MHV was released preferentially at the basolateral surface. These results show that vesicles containing the two coronaviruses are targeted differently in LMR cells. We propose that the viruses are sorted at the Golgi complex into different transport vesicles that carry information directing them to one of the two surface domains. The apical release of TGEV and the basolateral release of MHV might be factors contributing to the difference in virus spread found between TGEV and MHV in their respective natural hosts, the former causing mainly a localized enteric infection, the latter spreading through the body to other organs.

The human intestinal cell lines Caco-2 and LS174T as models to study cell-type specific mucin expression

Mucin expression was studied during proliferation and differentiation of the enterocyte-like Caco-2 and goblet cell-like LS174T cell lines. Caco-2 cells express mRNAs of MUC1, MUC3, MUC4 and MUC5A/C whereas MUC2 and MUC6 mRNAs are virtually absent. Furthermore, MUC3 mRNA is expressed in a differentiation dependent manner, as is the case for enterocytes. Concomitantly MUC3 protein precursor (approximately 550 kDa) was detected in Caco-2 cells. In LS174T cells mucin mRNAs of MUC1, MUC2 and MUC6 are constitutively expressed at high levels, whereas MUC3, MUC4 and MUC5A/C mRNAs are present at low levels. At the protein level LS174T cells express the goblet cell specific mucin protein precursors MUC2, MUC5A/C and MUC6 with apparent molecular masses of about 600 kDa, 470/500 kDa and 400 kDa respectively. MUC3 protein is not detectable. Furthermore, human gallbladder mucin protein (approximately 470 kDa precursor), of which the gene has not yet been identified, is expressed in LS174T cells. In addition, synthesis and secretion of the goblet cell specific mature MUC2, MUC5A/C and human gallbladder mucin was demonstrated in LS174T cells. It is concluded that Caco-2 and LS174T cell lines provide excellent in vitro models to elucidate the cell-type specific mechanisms responsible for mucin expression.

The ubiquitin conjugation system is required for ligand-induced endocytosis and degradation of the growth hormone receptor

The ubiquitin-dependent protein degradation system has recently been implicated in downregulation of signal transducing receptors. Growth hormone receptor (GHR) cDNA was transfected into Chinese hamster ovary cells, which exhibit a temperature-sensitive defect in ubiquitin conjugation (CHO-ts20), as well as into wild-type cells (CHO-E36). Upon binding of growth hormone (GH), two GHR polypeptides dimerize and initiate signal transduction. In CHO-E36 and in CHO-ts20 at the permissive temperature the GHR was ubiquitinated and degraded in a GH-dependent fashion. However, at the non-permissive temperature in CHO-ts20 cells, neither GH-dependent uptake nor degradation of the GHR was observed, while in CHO-E36 cells both GHR uptake and degradation were accelerated. Incubation of CHO-E36 cells with inhibitors of endosomal/lysosomal function (NH4Cl, bafilomycin A1) markedly reduced ligand-induced GHR degradation. Our results indicate that a functional ubiquitin conjugating system is required for GH-induced endocytosis and that degradation of both the exoplasmic and cytoplasmic portions of the GHR occurs within the endosomal/lysosomal compartment.

Synthesis of non-hydroxy-galactosylceramides and galactosyldiglycerides by hydroxy-ceramide galactosyltransferase

Galactosylceramide (GalCer) is the major glycolipid in brain. In order to characterize the activity of brain UDPgalactose: ceramide galactosyltransferase (CGalT), it has been stably expressed in CGalT-negative Chinese hamster ovary (CHO) cells. After fractionation of transfected cells, CHO-CGT, on sucrose gradients, the activity resides at the density of endoplasmic reticulum and not of Golgi. A lipid chromatogram from CHO-CGT cells revealed two new iodine-staining spots identified as GalCer, since they comigrate with GalCer standards, can be metabolically labelled with [3H]galactose, are recognized by anti-GalCer antibodies, and are resistant to alkaline hydrolysis. A third [3H]galactose lipid was identified as galactosyldiglyceride. In the homogenate CGalT displays a 25-fold preference for hydroxy fatty acid-containing ceramides. Remarkably, endogenous GalCer of transfected cells contains exclusively non-hydroxy fatty acids: fast atom bombardment and collision-induced dissociation mass spectrometric analysis revealed mainly C16:0 in the lower GalCer band on TLC and mainly C22:0 and C24:0 in the upper band. Our results suggest that CGalT galactosylates both hydroxy- and non-hydroxy fatty acid-containing ceramides and diglycerides, depending on their local availability. Thus, CGalT alone may be responsible for the synthesis of hydroxy- and non-hydroxy-GalCer, and galactosyldiglyceride in myelin.

Endocytosis of GPI-linked membrane folate receptor-alpha

GPI-linked membrane folate receptors (MFRs) have been implicated in the receptor-mediated uptake of reduced folate cofactors and folate-based chemotherapeutic drugs. We have studied the biosynthetic transport to and internalization of MFR isoform alpha in KB-cells. MFR-alpha was synthesized as a 32-kD protein and converted in a maturely glycosylated 36-38-kD protein 1 h after synthesis. 32-kD MFR-alpha was completely soluble in Triton X-100 at 0 degree C. In contrast, only 33% of the 36-38-kD species could be solubilized at these conditions whereas complete solubilization was obtained in Triton X-100 at 37 degrees C or in the presence of saponin at 0 degree C. Similar solubilization characteristics were found when MFR-alpha at the plasma membrane was labeled with a crosslinkable 125I-labeled photoaffinity-analog of folic acid as a ligand. Triton X-100-insoluble membrane domains containing MFR-alpha could be separated from soluble MFR-alpha on sucrose flotation gradients. Only Triton X-100 soluble MFR-alpha was internalized from the plasma membrane. The reduced-folate-carrier, an integral membrane protein capable of translocating (anti-)folates across membranes, was completely excluded from the Triton X-100-resistant membrane domains. Internalized MFR-alpha recycled slowly to the cell surface during which it remained soluble in Triton X-100 at 0 degree C. Using immunoelectron microscopy, we found MFR-alpha along the entire endocytic pathway: in clathrin-coated buds and vesicles, and in small and large endosomal vacuoles. In conclusion, our data indicate that a large fraction, if not all, of internalizing MFR-alpha bypasses caveolae.

MHV-A59 enters polarized murine epithelial cells through the apical surface but is released basolaterally

Coronaviruses have a marked tropism for epithelial cells. Entry and release of the porcine transmissible gastroenteritis virus (TGEV) is restricted to apical surfaces of polarized epithelial cells, as we have recently shown (J. W. A. Rossen, C. P. J. Bekker, W. F. Voorhout, G. J. A. M. Strous, A. van der Ende, and P. J. M. Rottier, 1994, J. Virol. 68, 7966-7973). In this paper we analyze the interactions of mouse hepatitis coronavirus A59 (MHV-A59) with polarized murine kidney cells (mTAL) grown on permeable supports. After inoculation from the apical or basolateral side, virus entry was found to take place only through the apical membrane. The virus utilized a protein of the carcinoembryonic antigen family as its receptor. In contrast to TGEV, MHV-A59 was released preferentially from the basolateral plasma membrane domain, as evidenced by the accumulation of viral proteins and infectivity in the basolateral culture fluid as well as by electron microscopical observations. In the mouse, MHV initially replicates in the nasal epithelium before being disseminated throughout the body; the basolateral release of MHV from epithelial cells into the animal's circulation may be the first step in the establishment of a systemic infection.

Cloning and analysis of human gastric mucin cDNA reveals two types of conserved cysteine-rich domains

Human gastric mucin was isolated by successive CsCl-gradient ultracentrifugation in the presence of guanidinium hydrochloride to prevent degradation of the polypeptide moieties of the molecules. The amino acid sequence of a tryptic fragment of this molecule was identical to that of a tryptic fragment of tracheobronchial mucin. An oligonucleotide based on this sequence hybridized specifically to human stomach mRNA and was subsequently used to screen a human stomach lambda ZAPII cDNA library. The largest of 10 positive clones encoded 850 amino acid residues, including the tryptic fragment, with high amounts of threonine, serine and proline residues. Interestingly, cysteine accounted for almost 8% of the amino acid residues. The 3' part of the sequence was very similar but not identical to the 3' region of human tracheobronchial cDNA. No tandem repeated sequences were present and the deduced polypeptide sequence contained two potential N-linked glycosylation sites. Four cysteine-rich clusters were detected, one of which was apparently homologous to the D-domains present in other mucins and in von Willebrand factor. The arrangement of the cysteines in three other cysteine-rich clusters was conserved in the human gastric mucin cDNA in a similar fashion as in two domains in the MUC2 gene product. The cysteine-rich domains were separated by short stretches of non-repetitive amino acid residues with a very high content of threonine and serine residues. These data suggest that the encoded polypeptide of this clone may be involved in disulphide-bond-mediated oligomerization of the mucin, and provide new insights into the molecular organization of mammalian apomucins.

39 kDa receptor-associated protein is an ER resident protein and molecular chaperone for LDL receptor-related protein

The low density lipoprotein receptor-related protein (LRP) is a multifunctional endocytic receptor with the ability to bind and endocytose several structurally and functionally distinct ligands. A 39 kDa receptor-associated protein (RAP) inhibits all ligand interactions with LRP in vitro. In the present study, we demonstrate that RAP is an endoplasmic reticulum (ER) resident protein. The tetrapepetide sequence HNEL at the C-terminus of RAP is both necessary and sufficient for RAP retention within the ER. Metabolic labeling combined with cross-linking studies show that RAP interacts with LRP in vivo. Pulse-chase analysis reveals that this association is transient early in the secretory pathway and coincides with LRP aggregation and reduced ligand binding activity. Both internal triplicated LRP binding domains on RAP and multiple RAP binding domains on LRP appear to contribute to the aggregation of LRP and RAP. Dissociation of RAP from LRP results from the lower pH encountered later in the secretory pathway and correlates with an increase in LRP ligand binding activity. Taken together, our results thus suggest that RAP functions intracellularly as a molecular chaperone for LRP and regulates its ligand binding activity along the secretory pathway.

Preparation of anti-mucin polypeptide antisera to study mucin biosynthesis

Mucins are very heavily O-glycosylated glycoproteins. For in depth studies on the cell biological aspects of mucins, anti-polypeptide antibodies are essential. We therefore developed a method for the preparation and screening of polyclonal antisera against mucin peptide epitopes. Mucins from five different tissues were isolated using CsCl/guanidinium.HCl density gradient centrifugation, and polyclonal antisera were prepared. Specificity for mucin peptide epitopes was determined by Western blotting, immunohistochemistry, and immunoprecipitation. The versatility of each anti-mucin antiserum for the study of mucin biosynthesis was tested in metabolic labeling experiments on tissue explants. All polyclonal antisera were directed primarily against peptide epitopes of mucin precursors as well as of fully glycosylated mucins. Each of the polyclonal antisera enabled us to study the mucin biosynthesis in the organ where the mucin was isolated from originally. Our mucin isolation method yields very pure mucins with sufficiently intact polypeptides to reproducibly elicit polyclonal anti-polypeptide antisera. As the sera recognized the polypeptides, primarily independent of the state of O-glycosylation, the intermediate steps in the biosynthesis of the mucins could be identified.

Coronaviruses in polarized epithelial cells

Coronaviruses have a marked tropism for epithelial cells. In this paper the interactions of the porcine transmissible gastroenteritis virus (TGEV) and mouse hepatitis virus (MHV-A59) with epithelial cells are compared. Porcine (LLC-PK1) and murine (mTAL) epithelial cells were grown on permeable supports. By inoculation from the apical or basolateral side both TGEV and MHV-A59 were found to enter the polarized cells only through the apical membrane. The release of newly synthesized TGEV from LLC-PK1 cells occurred preferentially from the apical plasma membrane domain, as evidenced by the accumulation of viral proteins and infectivity in the apical culture fluid. In contrast, MHV was released preferentially from the basolateral membrane of mTAL cells. The apical release of TGEV and the basolateral release of MHV may explain the in vivo establishment of a local and systemic infection, respectively.

Biosynthesis of a human gall-bladder mucin

Mucin glycoproteins play an important role in the initial stages of gall-stone formation by a currently largely unknown mechanism. Understanding the structure of gall-bladder mucin is necessary to comprehend the mechanism by which cholesterol monohydrate crystals aggregate. Three successive CsCl-gradient-ultracentrifugation steps were used to purify human gall-bladder mucin from gall-bladder tissue. The isolated macromolecules had a typical mucin-like monosaccharide composition and appeared as heterogeneous high-M(r) glycoproteins on SDS/PAGE. A polyclonal antiserum was raised against these molecules and the specificity of the antiserum was ascertained by immunoblotting. The antiserum specifically stained mucous granules at the apical side of all gall-bladder epithelial cells in neck, fundus and body. The antibody was subsequently used to immunoprecipitate the mucin and biosynthetic intermediates from gall-bladder-tissue homogenates. An early biosynthetic precursor of the isolated mucin was identified by SDS/PAGE as a single polypeptide with an apparent M(r) of approx. 470,000. This precursor protein was converted after 1 h into a heterogeneous high-M(r) glycoconjugate with an electrophoretic mobility similar to that of the purified mucin. The mature mucin, but not the precursor, was secreted into the culture medium, starting at 1 h. As shown by SDS/PAGE under non-reducing conditions, the precursors form disulphide-linked oligomers. Using the N-glycosylation inhibitor tunicamycin, the apparent M(r) of the precursor was decreased to approx. 410,000, indicating that N-linked glycan chains are attached to the precursor polypeptide.

Identification of a human gastric mucin precursor: N-linked glycosylation and oligomerization

Gastric mucin plays an important role in the protection of the stomach wall from chemical, microbiological and mechanical damage. We have previously isolated human gastric mucus glycoproteins and raised a polyclonal antiserum against these macromolecules. This antiserum specifically reacted with gastric mucins in immunoblotting experiments and stained mucous granules at the apical side of gastric surface epithelial cells. A similar staining pattern was obtained after incubation with an antiserum against rat gastric mucin. Next we used the antiserum in pulse-chase experiments of human stomach tissue explants. After short labelling periods with [35S]methionine and [35S]cysteine, the antiserum reacted with a polypeptide with an apparent molecular mass of approx. 500 kDa as determined by SDS/PAGE, which was converted after 90 min into a heterogeneous high-molecular-mass glycoprotein. This high-molecular-mass form, but not the 500 kDa polypeptide, was detectable in the culture medium after 2 h. This strongly suggests that the 500 kDa polypeptide is the precursor of the purified gastric mucin. Analysis of pulse-chase experiments by non-reducing SDS/PAGE revealed that the precursors form disulphide-linked oligomers early in biosynthesis, before the addition of O-linked sugars. After preincubation with the N-glycosylation inhibitor, tunicamycin, the apparent molecular mass of the precursor decreased marginally but consistently, indicating that N-linked glycan chains are present on the mucin precursor.

Entry and release of transmissible gastroenteritis coronavirus are restricted to apical surfaces of polarized epithelial cells

The transmissible gastroenteritis coronavirus (TGEV) infects the epithelial cells of the intestinal tract of pigs, resulting in a high mortality rate in piglets. This study shows the interaction of TGEV with a porcine epithelial cell line. To determine the site of viral entry, LLC-PK1 cells were grown on permeable filter supports and infected with TGEV from the apical or basolateral side. Initially after plating, the virus was found to enter the cells from both sides. During further development of cell polarity, however, the entry became restricted to the apical membrane. Viral entry could be blocked by a monoclonal antibody to the viral receptor aminopeptidase N. Confocal laser scanning microscopy showed that this receptor protein was present at both the apical and basolateral plasma membrane domains just after plating of the cells but that it became restricted to the apical plasma membrane during culture. To establish the site of viral release, the viral content of the apical and basolateral media of apically infected LLC-PK1 cells was measured by determining the amount of radioactively labelled viral proteins and infectious viral particles. We found that TGEV was preferentially released from the apical plasma membrane. This conclusion was confirmed by electron microscopy, which demonstrated that newly synthesized viral particles attached to the apical membrane. The results support the idea that the rapid lateral spread of TGEV infection over the intestinal epithelia occurs by the preferential release of virus from infected epithelial cells into the gut lumen followed by efficient infection of nearby cells through the apical domain.

Expression of functional growth hormone receptor in a mouse L cell line infected with recombinant vaccinia virus

The growth hormone receptor is a member of a large family of receptors including the receptors for prolactin and interleukins. Upon binding to one molecule of growth hormone two growth hormone receptor polypeptides dimerize. We have expressed the rabbit growth hormone receptor DNA in transfected mouse L cells infected with polymerase T7-producing vaccinia virus. The growth hormone receptor was synthesized as a 85-kDa protein and transported to the cell surface. Western blotting and metabolic labeling combined with immunoprecipitation using a rabbit antibody probably directed against the cytosolic domain of the receptor showed that its expression was dependent on both transfection of the growth hormone receptor DNA and vaccinia infection. Binding studies with 125I-labeled growth hormone demonstrated specific binding sites at the cell surface 20 h after transfection. Permeabilization with saponin showed that the growth hormone receptor binding sites were almost exclusively present at the cell surface with little intracellularly. Chemical crosslinking of the 125I-growth hormone complex resulted in a 180-kDa complex which could specifically be immunoprecipitated with the antiserum. Electron microscopic immunocytochemistry confirmed the presence of growth hormone receptor at the cell surface. Furthermore, specific growth hormone receptor antigen was also associated with intracellular membranes. These results thus show that this transient transfection system will be useful for cell biological studies of growth hormone receptor regulation.

Bidirectional transcytosis determines the steady state distribution of the transferrin receptor at opposite plasma membrane domains of BeWo cells

Trophoblast-like BeWo cells form well-polarized epithelial monolayers, when cultured on permeable supports. Contrary to other polarized cell systems, in which the transferrin receptor is found predominantly on the basolateral cell surface, BeWo cells express the transferrin receptor at both apical and basolateral cell surfaces (Cerneus, D.P., and A. van der Ende. 1991. J. Cell Biol. 114: 1149-1158). In the present study we have addressed the question whether BeWo cells use a different sorting mechanism to target transferrin receptors to the cell surface, by examining the biosynthetic and transcytotic pathways of the transferrin receptor in BeWo cells. Using trypsin and antibodies to detect transferrin receptors at the cell surface of filter-grown BeWo cells, we show that at least 80% of newly synthesized transferrin receptor follows a direct pathway to the basolateral surface, demonstrating that the transferrin receptor is efficiently intracellularly sorted. After surface arrival, pulse-labeled transferrin receptor equilibrates between apical and basolateral cell surfaces, due to ongoing transcytotic transport in both directions. The subsequent redistribution takes over 120 min and results in a steady state distribution with 1.5-2.0 times more transferrin receptors at the basolateral surface than at the apical surface. By monitoring the fate of surface-bound 125I-transferrin, internalized either from the apical or basolateral surface transcytosis of the transferrin receptor was studied. About 15% of 125I-transferrin is transcytosed in the basolateral to apical direction, whereas 25% is transcytosed in the opposite direction, indicated that the fraction of receptors involved in transcytosis is roughly twofold higher for the apical receptor pool, as compared to the basolateral pool. Upon internalization, both apical and basolateral receptor pools become redistributed on both surfaces, resulting in a twofold higher number of transferrin receptors at the basolateral surface. Our results indicate that in BeWo cells bidirectional transcytosis is the main factor in surface distribution of transferrin receptors on apical and basolateral surfaces, which may represent a cell type-specific, post-endocytic, sorting mechanism.

Detergent insolubility of alkaline phosphatase during biosynthetic transport and endocytosis. Role of cholesterol

Alkaline phosphatase is anchored to the outer leaflet of the plasma membrane by a covalently attached glycosyl-phosphatidylinositol anchor. We have studied the biosynthetic transport and endocytosis of alkaline phosphatase in the choriocarcinoma cell line BeWo, which endogenously expresses this protein. It was demonstrated that the protein was synthesized as a Triton X-100-soluble precursor. During transport to the cell surface the enzyme was converted in a mature form, which was insoluble in Triton X-100 at 0 degrees C. Once at the cell surface 85% of alkaline phosphatase remained in the detergent-insoluble form. Under steady state conditions 15% of alkaline phosphatase was endocytosed. Most interestingly, this fraction of internalized alkaline phosphatase was completely soluble in Triton X-100 at 0 degrees C. After depletion of membrane cholesterol by saponin, alkaline phosphatase became completely soluble in Triton X-100 at 0 degrees C, suggesting that cholesterol plays a critical role in the formation and maintenance of Triton X-100-resistant membrane domains.

Differential effects of brefeldin A on transport of secretory and lysosomal proteins

Brefeldin A (BFA) rapidly blocks anterograde exocytotic transport through the Golgi complex. Sustained retrograde traffic induced by brefeldin A causes redistribution of constituents of the Golgi, but not the trans-Golgi network (TGN), to the endoplasmic reticulum (ER). In the present study on HepG2 cells, we have observed a differential effect of BFA on transport from the TGN of two soluble proteins: alpha 1-antitrypsin as a representative of secretory proteins and cathepsin D as a prototype of lysosomal enzymes. The Golgi complex of HepG2 cells is sensitive to BFA, as within minutes after its addition nearly all activity of three resident Golgi enzymes was recovered in the ER as monitored by cell fractionation on sucrose density gradients. In accordance with this, "high mannose"-glycosylated alpha 1-antitrypsin was retained in or transported back to the ER. "Complex"-glycosylated alpha 1-antitrypsin was neither secreted into the medium nor transported back to the ER. Most of it was retained in vesicles with the buoyant density of Golgi. These vesicles contained the fluid phase endocytotic marker horseradish peroxidase when this was added to the culture medium prior to the BFA, suggesting that the vesicles derived from the TGN. After BFA addition, the compartment became inaccessible to endocytosed horseradish peroxidase. In contrast to blocking transport of complex alpha 1-antitrypsin, BFA did not affect processing of newly synthesized complex-glycosylated procathepsin D (53 kDa) to the mature 31-kDa form. Neither did it interfere with processing of endocytosed procathepsin D. That the mature cathepsin D had indeed reached the lysosomes was verified by Percoll density gradient fractionation. In conclusion, in HepG2 cells, BFA induces two blocks in the secretory pathway: one at the level of the ER-Golgi juncture and the other in the TGN. In contrast, transport from the Golgi complex to the lysosomes and from the plasma membrane to the lysosomes continued.

The differential degradation of two cytosolic proteins as a tool to monitor autophagy in hepatocytes by immunocytochemistry

The major pathway for cytosolic constituents to enter lysosomes is by autophagy. We used two cytosolic proteins, CuZn superoxide dismutase (SOD) and carbonic anhydrase III (CAIII), as autophagic markers in male rat hepatocytes. We took advantage of the differential presence of the two proteins in autophagic vacuoles because of the high resistance of SOD to lysosomal degradation as compared with CAIII. This allows us to determine the sequence of autophagic vacuole formation. We have double immunogold-labeled SOD and CAIII in cryosections of fasted rat liver and calculated the ratios of SOD over CAIII labeling densities (SOD/CAIII) in autophagic vacuoles (AV), as compared with the cytoplasm. Different classes of AV were defined according to their SOD/CAIII, their morphology, and their additional immunolabeling for the lysosomal markers lgp120 and cathepsin D. Of all AV, 15% exhibited a cytosol-like SOD/CAIII, indicating that degradation had not yet begun. Most of these initial AV (AVi) showed two enveloping membranes. The formation of AVi was prevented by 3-methyladenine, a potent inhibitor of autophagy. Of all AV, 85% showed a SOD/CAIII that exceeded the cytosolic ratio. These single membrane-bound vacuoles were called degradative AV (AVd). Labeling for lysosomal markers allowed the characterization of AV that shared features with both AVi and AVd. These AVi/d had a cytosol-like SOD/CAIII and a double membrane, but showed some labeling for lysosomal markers. Probably these AVi/d represent the recipient compartment for lysosomal components. AVd were positive for cathepsin D and lgp120. We discerned two AVd subclasses. Early AVd with cytosol-like SOD labeling density while CAIII labeling density was consistently lower than in the cytosol. Their size was similar to AVi and AVi/d. Late AVd contained higher SOD concentrations and were mostly larger. Our findings suggest that AV acquire lysosomal constituents by fusion with small nonautophagic structures and that after subsequent elimination of the inner membrane of AVi, degradation starts resulting in the formation of early AVd and late AVd.

Distribution of CuZn superoxide dismutase in rat liver

Morphological and cell fractionation approaches were used to establish unambiguously the distribution of CuZn superoxide dismutase (CuZn SOD) in rat hepatocytes. Immunocytochemical observations revealed a primarily cytoplasmic localization of the enzyme. While only trace amounts were found in cell organelles like mitochondria and peroxisomes, lysosomes were labelled stronger than the cytoplasm. The presence of CuZn SOD in lysosomes was also identified in cell fractions of normal and Triton WR-1339-treated rats. Microscopic studies showed that the distribution of CuZn SOD was not affected by Triton, but the Triton treatment induced an apparent increase in the number and size of lysosomes with electron lucent contents which corresponded with a shift of lysosomes to low buoyant density fractions. The majority of CuZn SOD originally present in the mixed mitochondrial-peroxisomal-lysosomal fractions of the normal liver comigrated with the main peak of the lysosomal marker, beta-hexosaminidase, after Triton treatment. The peroxisomal marker, catalase, did not migrate with the CuZn SOD/beta-hexosaminidase-rich fractions in livers from Triton-treated animals. These results confirm earlier observations in rat liver cells, showing that CuZn SOD, a primarily cytosolic enzyme, accumulates in lysosomes. It is not present in significant amounts in rat hepatocyte peroxisomes.

Identification of subcellular compartments involved in biosynthetic processing of cathepsin D

We have assigned the biosynthetic processing steps of cathepsin D to intracellular compartments which are involved in its transport to lysosomes in HepG2 cells. Cathepsin D was synthesized as a 51-kDa proenzyme. After formation of 51-55-kDa intermediates due to processing of N-linked oligosaccharides, procathepsin D was proteolytically processed to an intermediate 44-kDa and the mature 31-kDa enzyme. The intersection of the biosynthetic pathway of cathepsin D with the endocytic pathway was labeled with horseradish peroxidase and monitored biochemically by 3,3'-diaminobenzidine cytochemistry. Horseradish peroxidase was used either as a fluid-phase marker to label the entire endocytic pathway or conjugated to transferrin (Tf) to label endosomes only. Directly after biosynthesis cathepsin D was accessible neither to horseradish peroxidase nor Tf-horseradish peroxidase. Newly synthesized 51-55-kDa species of cathepsin D present in the trans-Golgi reticulum were accessible to both horseradish peroxidase and Tf-horseradish peroxidase. The accessibility of trans-Golgi reticulum to both endocytosed horseradish peroxidase and Tf-horseradish peroxidase was monitored by colocalization with a secretory protein, alpha 1anti-trypsin. The proteolytic processing of 51-55-kDa to 44-kDa cathepsin D occurred in compartments which were fully accessible to fluid-phase horseradish peroxidase. Tf-horseradish peroxidase had access to only 20% of 44-kDa cathepsin D while it had no access to 31-kDa cathepsin D. In contrast, the 31-kDa species was completely accessible to fluid-phase horseradish peroxidase. We conclude that proteolytic processing of 51-55-kDa to 44-kDa cathepsin D occurs in endosomes, whereas the processing of 44-31-kDa cathepsin D takes place in lysosomes.

Mucin-type glycoproteins

Considerable advances have been made in recent years in our understanding of the biochemistry of mucin-type glycoproteins. This class of compounds is characterized mainly by a high level of O-linked oligosaccharides. Initially, the glycoproteins were solely known as the major constituents of mucus. Recent studies have shown that mucins from the gastrointestinal tract, lungs, salivary glands, sweat glands, breast, and tumor cells are structurally related to high-molecular-weight glycoproteins, which are produced by epithelial cells as membrane proteins. During mucin synthesis, an orchestrated sequence of events results in giant molecules of Mr 4 to 6 x 10(6), which are stored in mucous granules until secretion. Once secreted, mucin forms a barrier, not only to protect the delicate epithelial cells against the extracellular environment, but also to select substances for binding and uptake by these epithelia. This review is designed to critically examine relations between structure and function of the different compounds categorized as mucin glycoproteins.

Mannose 6-phosphate-independent targeting of cathepsin D to lysosomes in HepG2 cells

We have studied the role of N-linked oligosaccharides and proteolytic processing on the targeting of cathepsin D to the lysosomes in the human hepatoma cell line HepG2. In the presence of tunicamycin cathepsin D was synthesized as an unglycosylated 43-kDa proenzyme which was proteolytically processed via a 39-kDa intermediate to a 28-kDa mature form. Only a small portion was secreted into the culture medium. During intracellular transport the 43-kDa procathepsin D transiently became membrane-associated independently of binding to the mannose 6-phosphate receptor. Subcellular fractionation showed that unglycosylated cathepsin D was efficiently targeted to the lysosomes via intermediate compartments similar to the enzyme in control cells. The results show that in HepG2 cells processing and transport of cathepsin D to the lysosomes is independent of mannose 6-phosphate residues. Inhibition of the proteolytic processing of 53-kDa procathepsin D by protease inhibitors caused this form to accumulate intracellularly. Subcellular fractionation revealed that the procathepsin D was transported to lysosomes, thereby losing its membrane association. Procathepsin D taken up by the mannose 6-phosphate receptor also transiently became membrane-associated, probably in the same compartment. We conclude that the mannose 6-phosphate-independent membrane-association is a transient and compartment-specific event in the transport of procathepsin D.

Rat gastric mucin is synthesized and secreted exclusively as filamentous oligomers

Oligomeric gastric mucin was isolated from the fundic part of the rat stomach. Previously we have shown by biochemical analysis that this oligomeric mucin consists of disulphide-linked homo-oligomers, which contain no other covalently attached proteins [Dekker, Aelmans & Strous (1991) Biochem. J. 277, 423-427]. Electron-microscopic images of the oligomeric mucin revealed a heterogenous population of long filamentous molecules of 300-3000 nm length. After reduction and carboxymethylation the monomeric mucins displayed a length distribution with a single peak at about 279 nm. Length-distribution analysis of oligomeric molecules with length up to 1000 nm revealed three subpopulations with one, two or three times the length of the monomeric mucin. The oligomers displayed small globular domains of about 15 nm, which were equally spaced along the molecule's length. As the distance between these globular domains was similar to the monomer length, these domains most likely indicate attachment sites of the monomers. These results show that the mucin monomers attached end-to-end in the oligomer. Biosynthesis of the mucin oligomers was studied by labelling of stomach explants in vitro with [35S]methionine, [3H]galactose or [35S]sulphate and subsequent immunoprecipitation of the mucin with a specific antiserum. Analysis by electrophoresis and gel filtration revealed that the oligomerization takes place by formation of disulphide bonds between the 300 kDa mucin precursors. The mucin was exclusively synthesized and secreted as fully glycosylated oligomers, as neither precursor proteins nor monomeric mucin were detected in the culture medium. A model for the biosynthesis of rat gastric mucin is proposed in which the filamentous mucin monomers are linked end-to-end by disulphide bonds.

The oligomeric structure of rat and human gastric mucins

Intact oligomeric gastric mucins were isolated from the fundic part of rat and human stomach. Physicochemical properties of the oligomeric mucins from both species, such as buoyant density, molecular mass, proteinase-resistance, amino acid composition and monosaccharide composition were similar. Biochemical analysis showed that the oligomeric mucins from both species consist of disulphide-linked mucin monomers exclusively: no other covalently attached proteins were detected in purified monomeric mucin. Four major differences were found between the monomeric mucins of these species: (1) the human monomer is larger, (2) the proteolytic-digest peptides derived from proteinase-sensitive portions of the polypeptide backbone displayed no sequence similarity, (3) the human mucin was less sulphated compared with rat mucin, and (4) the proteinase-sensitive part of the human mucin was relatively larger. However, analyses of [3H]galactose-labelled mucin from both species on gel filtration revealed that both gastric mucins were exclusively synthesized as oligomers. The results indicate that the oligomeric structures of human and rat gastric mucin are similar and their biosyntheses are not affected by the differences in the subunits.

Late endosomes derive from early endosomes by maturation

Endocytosed proteins destined for degradation in lysosomes are targeted mainly to early endosomes following uptake. Late endosomes are the major site for entry of newly synthesized lysosomal hydrolases via the cation-independent mannose 6-phosphate receptor into the degradative pathway. No consensus exists as to the mechanism of transport from early to late endosomes. We used asialoorosomucoid and transferrin to label selected parts of the degradative and receptor-recycling pathways, respectively, in the human hepatoma cell line HepG2. Intracellular mixing of sequentially endocytosed 125I- and HRP-labeled ligands was monitored by using 3,3'-diaminobenzidine-mediated density perturbation. The entire endocytic pathway of asialoorosomucoid, except for the lysosomes, remained fully accessible to subsequently endocytosed transferrin conjugated to HRP with unchanged kinetics. These results together with immunoelectron microscopic data support a model in which early endosomes gradually mature into late endosomes.

A pool of intracellular phosphorylated asialoglycoprotein receptors which is not involved in endocytosis

One proposed function of the asialoglycoprotein receptor in hepatocytes is to mediate the endocytosis of galactose and N-acetylgalactosamine-exposing glycoproteins. Recently we defined a pool of intracellular H1 subunits of the asialoglycoprotein receptor (ASGPR) in the human hepatoma cell line HepG2 which appeared not to be involved in endocytosis (Stoorvogel, W., Geuze, H. J., Griffith, J. M., Schwartz, A. L., and Strous, G. J. (1989) J. Cell Biol. 108, 2137-2148). In addition, a pool of stably phosphorylated intracellular ASGPR has been detected (Fallon, R. J., and Schwartz, A. L. (1988) J. Biol. Chem. 263, 13159-13166). In the current study we integrate these findings and provide evidence for the existence of two types of intracellular nonexchangeable compartments containing ASGPR. A transiently phosphorylated pool of ASGPR shuttles between the plasma membrane and endosomes, via a pathway identical to that of the transferrin receptor. The second pool comprises 20% of the total intracellular ASGPR, is stably phosphorylated at a serine residue, and is located in intracellular compartments devoid of recycling transferrin receptor. We refer to this ASGPR pool as the "silent pool." We furthermore show that the two receptor pools are confined to compartments exhibiting different buoyant densities on sucrose density gradients. ASGPR in the "silent pool" is fully glycosylated, suggesting a post-Golgi sorting mechanism for trafficking to this compartment. Possible functions of the "silent" ASGPR pool are discussed.

Mannose 6-phosphate-independent membrane association of cathepsin D, glucocerebrosidase, and sphingolipid-activating protein in HepG2 cells

The membrane association of the lysosomal enzymes cathepsin D and glucocerebrosidase and its naturally occurring sphingolipid activating protein was studied in HepG2 cells. We differentially permeabilized cells with low concentrations of saponin, at which secretory proteins rinsed out completely, whereas integral membrane proteins were not released. All relevant intracellular compartments were shown to be permeabilized by saponin. Metabolic labeling showed that early precursors of cathepsin D, sphingolipid activating protein, and glucocerebrosidase were completely released from the cells, whereas more than 80% of the high molecular mass intermediates were retained by the cells. Treatment of permeabilized cells with 10 mM mannose 6-phosphate released only 50% of the cell-associated cathepsin D. Glucocerebrosidase remained membrane-associated, but cathepsin D and sphingolipid activating protein were released from the cells after proteolytic processing. Sphingolipid activating proteins and cathepsin D behaved similarly during biosynthesis and showed similar sensitivity to mannose 6-phosphate. The membrane association of the intermediate form of cathepsin D was independent of the presence of N-linked oligosaccharides. Subcellular fractionation on sucrose gradients showed that the lysosomal proteins became membrane-associated probably in the Golgi complex, and that both mannose 6-phosphate-dependent and mannose 6-phosphate-independent membrane association occur in the same compartments. We conclude that, in HepG2 cells, cathepsin D, sphingolipid activating protein, and glucocerebrosidase exhibit MPR-independent membrane association which is acquired in the same compartments beyond the rough endoplasmic reticulum.

Brefeldin A induces a microtubule-dependent fusion of galactosyltransferase-containing vesicles with the rough endoplasmic reticulum

The fungal drug brefeldin A (BFA) has recently been found to induce a redistribution of medial- and cis-Golgi components to the endoplasmic reticulum (ER), raising the possibility of the existence of a retrograde pathway from the Golgi complex to the ER. Here, we demonstrate a BFA-induced reversible rearrangement of the trans-Golgi membrane protein galactosyltransferase (Gal-T) to the ER in HeLa cells. With immunofluorescence microscopy we have shown that BFA first caused a rapid change of Gal-T immunolabelling from a normal Golgi complex pattern to long and slender structures emanating from the cell centre and co-localizing with tubulin. Then immunofluorescence became ER-like. This effect was not dependent on ongoing protein synthesis and was reversed to normal within 120 min after removal of the drug. Restoration of the Golgi complex after removal of brefeldin A was energy-dependent but not mediated by microtubules nor dependent on protein synthesis. BFA-induced backflow of Gal-T was inhibited by nocodazole, a microtubule-disrupting agent. Immunoelectron microscopy showed that BFA treatment resulted in the fusion of Gal-T-containing vesicles with the ER. Furthermore, sucrose gradient centrifugation showed a significant shift in density of mature Gal-T polypeptides upon BFA treatment: about 40% of the enzyme migrated from its original density (1.13 g/ml) to the density of rough ER (1.19 g/ml). Thus, BFA caused microtubule-dependent vesicular backflow from a trans-Golgi component to the ER followed by fusion of the Golgi-derived vesicles with the ER.

Covalent oligomerization of rat gastric mucin occurs in the rough endoplasmic reticulum, is N-glycosylation-dependent, and precedes initial O-glycosylation

Rat gastric mucin undergoes extensive modifications during biosynthesis, including oligomerization, N- and O-glycosylation, and sulfation. We characterized the events in the rough endoplasmic reticulum (RER) and Golgi complex and studied how these steps are interrelated, using specific inhibitors of cellular processes. The mucin precursors oligomerize in the RER by forming intermolecular disulfide bonds. The oligomers comprise a mixture of predominantly di- and trimers of molar ratio 3:2. The oligomerized precursors are transported to the Golgi complex to form mature, oligomeric mucin by extensive O-glycosylation, and sulfation. N-Glycosylation of the precursor is required for efficient oligomerization. Brefeldin A, which inhibits protein transport between RER and Golgi complex, allows oligomerization and concomitantly induces initial O-glycosylation. Oligomerization and egrees from the RER precedes initial O-glycosylation and are therefore independent of the latter process.

Biosynthesis of gastric mucus glycoprotein of the rat

We have studied the biosynthesis of rat gastric mucin in stomach segments using an antiserum against rat gastric mucin specific for peptide epitopes. Pulse-chase experiments were performed with [35S]methionine, [3H]galactose, and [35S]sulfate to label mucin precursors in different stages of biosynthesis, which were analyzed after immunoprecipitation. The earliest mucin precursor that could be detected by sodium dodecyl sulfate-polyacrylamide gel electrophoresis was a 300-kDa protein. The occurrence of N-linked "high-mannose" oligosaccharides on this protein was shown by susceptibility to degradation by endo-beta-N-acetylglucosaminidase H. This precursor could be labeled with [35S]methionine and not with [3H]galactose or [35S]sulfate. The 300-kDa precursor was converted into mature mucin after extensive glycosylation and sulfation. The mature mucin but not the 300-kDa precursor was in part secreted into the medium. Specific inhibition of sulfation with sodium chlorate had no effect on rate and amount of mucin secretion. In addition, we show that two core proteins are expressed in rats, slightly varying in Mr among individual animals.

Isolation and structural analysis of rat gastric mucus glycoprotein suggests a homogeneous protein backbone

We isolated monomeric gastric mucus glycoprotein from the rat stomach by applying three successive CsCl-density-gradient steps in the continuous presence of guanidinium chloride. The rat gastric mucin was pure as compared with mucin isolated without the chaotropic reagent. In addition, the presence of guanidinium chloride resulted in a better preservation of the protein moiety. The purified mucin was fractionated according to buoyant density and chemically radiolabelled on tyrosine or cysteine residues and digested with specific proteinases. Analysis of mucin fractions of various densities gave identical peptide patterns, suggesting that the fractions contain a common protein backbone. Electron-microscopic images of the individual mucin molecules were recorded using rotary shadowing. They showed large filamentous molecules with a mean length of 208 nm that, after proteolytic digestion, yielded glycopeptides with a mean length of 149 nm. Heterogeneity in buoyant density and electrophoretic mobility is located in this large glycopeptide which remains after proteolytic digestion. Metabolic labelling of the mucin with [35 S]sulphate and [3H]galactose, followed by purification and proteolytic digestion, revealed that this glycopeptide accounts for most of the mass and contains relatively little protein, but probably all the oligosaccharides and sulphate. As this protein part is masked by the oligosaccharides, detailed study by the methods described was not possible. The results indicate that rat gastric mucin is homogeneous in a major part of the protein backbone and that the heterogeneity of the molecule originates most likely from differences in sulphate and/or sugar composition.

Relations between the intracellular pathways of the receptors for transferrin, asialoglycoprotein, and mannose 6-phosphate in human hepatoma cells

We compared the intracellular pathways of the transferrin receptor (TfR) with those of the asialoglycoprotein receptor (ASGPR) and the cation-independent mannose 6-phosphate receptor (MPR)/insulin-like growth factor II receptor during endocytosis in Hep G2 cells. Cells were allowed to endocytose a conjugate of horseradish peroxidase and transferrin (Tf/HRP) via the TfR system. Postnuclear supernatants of homogenized cells were incubated with 3,3'-diaminobenzidine (DAB) and H2O2. Peroxidase-catalyzed oxidation of DAB within Tf/HRP-containing endosomes cross-linked their contents to DAB polymer. The cross-linking efficiency was dependent on the intravesicular Tf/HRP concentration. The loss of detectable receptors from samples of cell homogenates treated with DAB/H2O2 was used as a measure of colocalization with Tf/HRP. To compare the distribution of internalized plasma membrane receptors with Tf/HRP, cells were first surface-labeled with 125I at 0 degrees C. After uptake of surface 125I-labeled receptors at 37 degrees C in the presence of Tf/HRP, proteinase K was used at 0 degrees C to remove receptors remaining at the plasma membrane. Endocytosed receptors were isolated by means of immunoprecipitation. 125I-TfR and 125I-ASGPR were not sorted from endocytosed Tf/HRP. 125I-MPR initially also resided in Tf/HRP-containing compartments, however 70% was sorted from the Tf/HRP pathway between 20 and 45 min after uptake. To study the accessibility of total intracellular receptor pools to endocytosed Tf/HRP, nonlabeled cells were used, and the receptors were detected by means of Western blotting. The entire intracellular TfR population, but only 70 and 50% of ASGPR and MPR, respectively, were accessible to endocytosed Tf/HRP. These steady-state levels were reached by 10 min of continuous Tf/HRP uptake at 37 degrees C. We conclude that 30% of the intracellular ASGPR pool is not involved in endocytosis (i.e., is silent). Double-labeling immunoelectron microscopy on DAB-labeled cells showed a considerable pool of ASGPR in secretory albumin-positive, Tf/HRP-negative, trans-Golgi reticulum. We suggest that this pool represents the silent ASGPR that has been biochemically determined. A model of receptor transport routes is presented and discussed.

Effect of ATP depletion and temperature on the transferrin-mediated uptake and release of iron by BeWo choriocarcinoma cells

We have recently described the transferrin-mediated uptake and release of iron by BeWo cells [van der Ende, du Maine, Simmons, Schwartz & Strous (1987) J. Biol. Chem. 262, 8910-8916]. We now extend our studies of the mechanisms responsible for uptake and release of iron by these cells. Following preloading, 59Fe release was maximal (about 12%) after about 4 h. Replacement of the extracellular medium with an equal volume of fresh medium either prior to or following the time at which equilibrium was reached further stimulated 59Fe release. Both the rate and maximum amount of iron release decreased if longer loading times were used. Preincubation of BeWo cells for 15 min with 10 mM-sodium cyanide and 50 mM-2-deoxyglucose prior to the determination of 59Fe release did not alter the amount released into medium (which did not contain a high-affinity iron chelator). However, under these conditions, the uptake of 59Fe was dramatically inhibited as a result of prolongation of the transferrin-transferrin-receptor complex recycling time. These results demonstrate that the release of iron from BeWo cells is independent of cellular ATP levels, whereas iron uptake is ATP-dependent. Rates of both 59Fe release and 59Fe uptake were temperature-dependent. Analysis of these data via an Arrhenius plot suggests a single rate-limiting step for the release and uptake processes between 0 and 37 degrees C. The apparent energies of activation of these processes are very similar (approx. 59.0 kJ/mol for iron release and 50.6 kJ/mol for iron uptake), which raises the possibility that the release and uptake of iron share a common thermodynamically rate-limiting step. Possible mechanisms involved in iron release out of the cell and out of the endosome are discussed.

Sulphation causes heterogeneity of gastric mucins

The synthesis of mucus glycoprotein in rat stomach was studied in stomach segments, which were pulse-labelled with both [3H]galactose and [35S]sulphate and chased for various times. The radioactive glycoproteins were analyzed by CsCl centrifugation and by agarose gel electrophoresis. After a pulse-labelling for 15 min with [3H]galactose, a possible intermediate with an Mr of 200,000 and a buoyant density of 1.60 g/ml could be demonstrated. Following chase periods of 1 and 4 h, [3H]galactose and [35S]sulphate were present in glycoproteins with a mean buoyant density of 1.50 g/ml. This is clearly different from the main density of glycoproteins isolated from mucosal scrapings (1.46 g/ml). Another difference is the high electrophoretic mobility on gel electrophoretic analysis of newly synthesized glycoproteins compared to that of the major portion of the glycoproteins from mucosal scrapings. When sulphation of glycoproteins was inhibited by sodium chlorate, electrophoretic mobility and buoyant density both decreased. Sodium chlorate had no effect on glycoprotein synthesis nor on glycoprotein secretion. We conclude from our data that the heterogeneity in electrophoretic mobility and buoyant density can be attributed to a different degree of sulphation of the same glycoprotein.

Release of soluble resident as well as secretory proteins from HepG2 cells by partial permeabilization of rough-endoplasmic-reticulum membranes

Secretory proteins migrate from the rough endoplasmic reticulum (ER) to the Golgi complex at different rates. Selective retention of specific proteins to rough-ER membrane constituents could explain this phenomenon. We have permeabilized HepG2 cells with low concentrations of saponin. Release of newly synthesized proteins was studied after brief labelling in the presence of [35S]methionine. The efflux of several secretory proteins was studied at various saponin concentrations; a 2-fold higher saponin concentration was required to release transferrin compared with that required to release albumin and orosomucoid. Glucosidase II, a soluble resident protein of the ER, is released at the same saponin concentration as albumin. Saponin did not destroy the membrane skeleton structure; at the concentrations used, the integral membrane protein G of vesicular-stomatitis virus remained fully associated with the cells.

Metalloendoprotease inhibitors block protein synthesis, intracellular transport, and endocytosis in hepatoma cells

Fusion of membrane vesicles has been implicated in many intracellular processes including the transport of proteins destined for secretion or storage. Vesicular transport coupled with membrane fusion has been demonstrated for rough endoplasmic reticulum to Golgi and Golgi to plasma membrane transport as well as receptor mediated endocytosis and receptor recycling. Recent studies with inhibitors suggest that metalloendoproteases may mediate a wide variety of intracellular fusion events. Thus, in order to examine the potential role of metalloendoproteases in both transport/secretion and endocytosis/recycling we have used selected dipeptide substrates to probe these processes in human HepG2 cells. Using pulse-chase labeling, immunoprecipitation, and polyacrylamide gel electrophoresis we show that transport and secretion of newly synthesized proteins along the exocytotic route were completely inhibited by substrate dipeptides (e.g. Cbz-Gly-Phe-amide, where Cbz is benzyloxycarbonyl) but not by irrelevant dipeptides (e.g. Cbz-Gly-Gly-amide). The effect was rapid, reversible, and specific. The secretory pathway was blocked between the rough endoplasmic reticulum and Golgi as well as Golgi and plasma membrane as judged by the status of N-glycosylation intermediates. In addition, these inhibitors specifically inhibited protein synthesis without alterations in cellular ATP concentrations. However, cell-free amino acid incorporation was not inhibited. Receptor-mediated uptake of asialoglycoproteins was specifically and reversibly inhibited by dipeptide substrates. This effect appears to be secondary to inhibition of recycling as neither ligand binding nor internalization were affected. Thus the present observations suggest that metalloendoprotease activity may be involved in the regulation of multiple intracellular pathways perhaps at the level of vesicular fusion events.

Sorting of mannose 6-phosphate receptors and lysosomal membrane proteins in endocytic vesicles

The intracellular distributions of the cation-independent mannose 6-phosphate receptor (MPR) and a 120-kD lysosomal membrane glycoprotein (lgp120) were studied in rat hepatoma cells. Using quantitative immunogold cytochemistry we found 10% of the cell's MPR located at the cell surface. In contrast, lgp120 was not detectable at the plasma membrane. Intracellularly, MPR mainly occurred in the trans-Golgi reticulum (TGR) and endosomes. lgp120, on the other hand, was confined to endosomes and lysosomes. MPR was present in both endosomal tubules and vacuoles, whereas lgp120 was confined to the endosomal vacuoles. In cells incubated for 5-60 min with the endocytic tracer cationized ferritin, four categories of endocytic vacuoles could be discerned, i.e., vacuoles designated MPR+/lgp120-, MPR+/lgp120+, MPR-/lgp120+, and vacuoles nonimmunolabeled for MPR and lgp120. Tracer first reached MPR+/lgp120-, then MPR+/lgp120+, and finally MPR-/lgp120+ vacuoles, which are assumed to represent lysosomes. To study the kinetics of appearance of endocytic tracers in MPR-and/or lgp120-containing pools in greater detail, cells were allowed to endocytose horse-radish peroxidase (HRP) for 5-90 min. The reduction in detectability of MPR and lgp120 antigenicity on Western blots, due to treatment of cell homogenates with 3'3-diaminobenzidine, was followed in time. We found that HRP reached the entire accessible pool of MPR almost immediately after internalization of the tracer, while prolonged periods of time were required for HRP to maximally access lgp120. The combined data suggest that MPR+/lgp120+ vacuoles are endocytic vacuoles, intermediate between MPR+/lgp120-endosomes and MPR-/lgp120+ lysosomes, and represent the site where MPR is sorted from lgp120 destined for lysosomes. We propose that MPR is sorted from lgp120 by selective lateral distribution of the receptor into the tubules of this compartment, resulting in the retention of lgp120 in the vacuoles and the net transport of lgp120 to lysosomes.

Recycling of 5′-nucleotidase in a rat hepatoma cell line

Intracellular movement of cell surface 5'-nucleotidase was studied in H4S cells, a rat hepatoma cell line. Surface labelled cells were incubated for various periods at 37 degrees C and treated with neuraminidase at 0 degrees C. Removal of sialic acid residues from glycoproteins results in a change of their isoelectric points. Analysis with isoelectric focusing was then used to distinguish between cell surface and intracellular 5'-nucleotidase. Incubation of 125I-surface-labelled cells at 37 degrees C resulted in a gradual decrease of labelled 5'-nucleotidase at the plasma membrane until, at 60 to 90 min, a steady state was reached with 52% of the label on the cell surface and 48% intracellular. Pretreatment of the cells with the weak base primaquine had no influence on this distribution while at the same time uptake of iron via the transferrin receptor was inhibited. Using immunoelectron microscopy 5'-nucleotidase was found on the cell surface, in multivesicular endosomes and the Golgi complex. Preincubation of the cells in the presence of cycloheximide caused a reduction of labelling in the Golgi complex, whereas the label in the other compartments was retained. These results lead to the conclusion that 5'-nucleotidase does not recycle through the Golgi complex and that in contrast to the transferrin receptor the recycling of 5'-nucleotidase is not inhibited by primaquine.

The influence of pre-incubation on the rate of iron uptake and of transferrin synthesis by intact isolated rat seminiferous tubules

It has been reported recently that major kinetic differences exist in the rate of iron uptake and of transferrin synthesis between intact seminiferous tubules and cultured Sertoli cells. To investigate possible causes of these differences, intact isolated rat seminiferous tubules were isolated and pre-incubated overnight. Then the rates of iron uptake and of transferrin synthesis were compared with those of freshly isolated tubules. We found that overnight pre-incubation increased the rate of both processes. Iron depletion was not the cause of these changes, since pre-incubation under iron excess gave comparable results. Possible explanations of these phenomena are discussed.

The pathways of endocytosed transferrin and secretory protein are connected in the trans-Golgi reticulum

We used a conjugate of transferrin and horseradish peroxidase (Tf/HRP) to label the intracellular transferrin receptor route in the human hepatoma cell line HepG2. The recycling kinetics of [125I]Tf/HRP were similar to those of unmodified [125I]Tf, implying identical routes for both ligands. 3,3'Diaminobenzidine (DAB)-cytochemistry was performed on post-nuclear supernatants of homogenates of cells which were incubated with both Tf/HRP and [125I]Tf, and caused two different effects: (a) the equilibrium density of [125I]Tf containing microsomes in a Percoll density gradient was increased, and (b) the amount of immunoprecipitable [125I]Tf from density-shifted lysed microsomes was only 20% of that of nonDAB treated microsomes. The whole biosynthetic route of alpha 1-antitrypsin (AT), a typical secretory glycoprotein in HepG2 cells, was labeled during a 60-min incubation with [35S]methionine. DAB cytochemistry was performed on post-nuclear supernatants of homogenates of cells which were also incubated with Tf/HRP. DAB cytochemistry caused approximately 40% of microsome-associated "complex" glycosylated [35S]alpha 1-antitrypsin ([35S]c-AT) to shift in a Percoll density gradient. Only part of the density shifted [35S]c-AT could be recovered by immunoprecipitation. A maximum effect was measured already after 10 min of Tf/HRP uptake. The density distribution of the "high mannose" glycosylated form of 35S-alpha 1-anti-trypsin [( 35S]hm-AT) was not affected by Tf/HRP. If in addition to Tf/HRP also an excess of non-conjugated transferrin was present in the medium, [35S]c-AT was not accessible for Tf/HRP, showing the involvement of the transferrin receptor (TfR) in the process. Furthermore, we show that if Tf/HRP and [35S]c-AT were located in different vesicles, the density distribution of [35S]c-AT was not affected by DAB-cytochemistry. Pulse-labeling with [35S]methionine was used to show that [35S]c-AT became accessible to endocytosed Tf/HRP minutes after acquirement of the complex configuration. A common intracellular localization of endocytosed Tf/HRP and secretory protein could be confirmed by immuno-electron microscopy: cryosections labeled with anti-albumin (protein A-colloidal gold) as well as DAB reaction product showed double-labeling in the trans-Golgi reticulum.

Kinetics of transferrin endocytosis and iron uptake by intact isolated rat seminiferous tubules and Sertoli cells in culture

The receptor-mediated endocytotic cycle of rat and human transferrin has been studied in intact, isolated rat seminiferous tubules and Sertoli cells in culture. Double-labeled [( 59Fe125I]) transferrin has been used to study the fate of transferrin and iron. Diferric transferrin binds to the tubules and the cultured Sertoli cells and is internalized. The iron remains inside, while the transferrin recycles and is released into the medium. Although, as reported before (Wauben-Penris et al., 1986), "extra" binding sites for human transferrin exist as compared to rat transferrin, this does not result in extra uptake of transferrin or iron. Both rat and human transferrin transport iron into the cells and recycle back to the surface, and do so with identical kinetics. A striking difference has been found between the mean efficient recycling times of the transferrin receptors in intact tubules (90 min) and in Sertoli cells in culture (21 min). Possible explanations of this difference are discussed. Light-microscopic autoradiography of [125 I]-labeled transferrin has revealed that the transferrin protein is excluded from the adluminal compartment, even after 21 h of incubation. This indicates that externally added transferrin itself does not deliver iron to the postmeiotic germ cells in intact, isolated rat seminiferous tubules.

In vitro biosynthesis of two human galactosyltransferase polypeptides

HeLa cell galactosyltransferase is synthesized as two precursor polypeptides of Mr = 45,000 and Mr = 47,000. The enzyme is present in the Golgi complex as a (mature) Mr = 54,000 glycoprotein. If cells are treated with tunicamycin, two precursor polypeptides are synthesized without N-linked oligosaccharides with molecular weights of 42,000 and 44,000, respectively. To investigate whether the two precursor polypeptides are synthesized on different mRNAs total RNA from HeLa cells was translated in a wheat germ cell-free system. Galactosyltransferase polypeptides were isolated by immunoprecipitation and compared to the polypeptides synthesized in vivo in the presence of tunicamycin. The two in vitro translated polypeptides co-migrate exactly with the polypeptides made in the cells in the presence of tunicamycin, indicating two different mRNAs for galactosyltransferase. The results also indicate that translocation of galactosyltransferase through the membrane of the rough endoplasmic reticulum is not followed by signal peptide cleavage.