Search

An in vitro system to study the transport of plasma membrane proteins from the TGN to the basolateral plasma membrane of polarized MDCK cells has been developed in which purified cell fractions are combined and transport between them is studied under controlled conditions. In this system, a donor Golgi fraction derived from VSV or influenza virus-infected MDCK cells, in which 35S-labeled viral glycoproteins were allowed to accumulate in the TGN during a low temperature block, is incubated with purified immobilized basolateral plasma membranes that have their cytoplasmic face exposed and are obtained by shearing-lysis of MDCK monolayers grown on cytodex beads. Approximately 15-30% of the labeled glycoprotein molecules are transferred from the Golgi fraction to the acceptor plasma membranes and are recovered with the sedimentable (1 g) beads. Transport is temperature, energy and cytosol dependent, and is abolished by alkylation of SH groups and inhibited by the presence of GTP-gamma-S, which implicates GTP-binding proteins and the requirement for GTP hydrolysis in one or more stages of the transport process. Endo H-resistant glycoprotein molecules that had traversed the medial region of the Golgi apparatus are preferentially transported and their luminal domains become accessible to proteases, indicating that membrane fusion with the plasma membrane takes place in the in vitro system. Mild proteolysis of the donor or acceptor membranes abolishes transport, suggesting that protein molecules exposed on the surface of these membranes are involved in the formation and consumption of transport intermediates, possibly as addressing and docking proteins, respectively. Surprisingly, both VSV-G and influenza HA were transported with equal efficiencies to the basolateral acceptor membranes. However, low concentrations of a microtubular protein fraction preferentially inhibited the transport of HA, although this effect was not abolished by microtubule depolymerizing agents. This system shows great promise for elucidating the mechanisms that effect the proper sorting of plasma membrane proteins in the TGN and their subsequent targeting to the appropriate acceptor membrane.

The synchronized directed transfer of the envelope glycoproteins of the influenza and vesicular stomatitis viruses from the Golgi apparatus to the apical and basolateral surfaces, respectively, of polarized Madin-Darby canine kidney (MDCK) cells can be achieved using temperature-sensitive mutant viruses and appropriate temperature shift protocols (Rindler, M. J., I. E. Ivanov, H. Plesken, and D. D. Sabatini, 1985, J. Cell Biol., 100:136-151). The microtubule-depolymerizing agents colchicine and nocodazole, as well as the microtubule assembly-promoting drug taxol, were found to interfere with the normal polarized delivery and exclusive segregation of hemagglutinin (HA) to the apical surface but not with the delivery and initial accumulation of G on the basolateral surface. Immunofluorescence analysis of permeabilized monolayers of influenza-infected MDCK cells treated with the microtubule-acting drugs demonstrated the presence of substantial amounts of HA protein on both the apical and basolateral surfaces. Moreover, in cells infected with the wild-type influenza virus, particles budded from both surfaces. Viral counts in electron micrographs showed that approximately 40% of the released viral particles accumulated in the intercellular spaces or were trapped between the cell and monolayer and the collagen support as compared to less than 1% on the basolateral surface of untreated infected cells. The effect of the microtubule inhibitors was not a result of a rapid redistribution of glycoprotein molecules initially delivered to the apical surface since a redistribution was not observed when the inhibitors were added to the cells after the HA was permitted to reach the apical surface at the permissive temperature and the synthesis of new HA was inhibited with cycloheximide. The altered segregation of the HA protein that occurs may result from the dispersal of the Golgi apparatus induced by the inhibitors or from the disruption of putative microtubules containing tracks that could direct vesicles from the trans Golgi apparatus to the cell surface. Since the vesicular stomatitis virus G protein is basolaterally segregated even when the Golgi elements are dispersed and hypothetical tracks disrupted, it appears that the two viral envelope glycoproteins are segregated by fundamentally different mechanisms and that the apical surface may be incapable of accepting vesicles carrying the G protein.

The demonstration that the envelope glycoproteins G of vesicular stomatitus virus and hemagglutinin of influenza virus synthesized in polarized epithelial cells transfected with the corresponding genes are effectively segregated to the basolateral or apical plasma membrane domains, respectively, implies that the information determining this segregation resides within the structures of the proteins themselves. To localize the sorting information within these proteins, the polarity of secretion of truncated hemagglutinin and G glycoproteins secreted from confluent monolayers of MDCK cells transformed with vectors containing the corresponding truncated cDNAs was examined. It was found that, even though the transformed cells continued to secrete a major endogenous glycoprotein exclusively from the apical surface, the modified viral glycoproteins were secreted in a nonpolarized fashion from both sides of the monolayers. These observations suggest that important information for the sorting of the viral glycoprotein is contained within their membrane anchoring or cytoplasmic segments or that, if sorting signals are luminally located, these signals must be present in a conformation that is not attainable when the polypeptides are not attached to the membrane.

The intracellular route followed by viral envelope glycoproteins in polarized Madin-Darby canine kidney cells was studied by using temperature-sensitive mutants of vesicular stomatitis virus (VSV) and influenza, in which, at the nonpermissive temperature (39.5 degrees C), the newly synthesized glycoproteins (G proteins) and hemagglutinin (HA), respectively, are not transported out of the endoplasmic reticulum. After infection with VSV and incubation at 39.5 degrees C for 4-5 h, synchronous transfer of G protein to the plasma membrane was initiated by shifting to the permissive temperature (32.5 degrees C). Immunoelectron microscopy showed that under these conditions the protein moved to the Golgi apparatus and from there directly to a region of the lateral plasma membrane near this organelle. G protein then seemed to diffuse progressively to basal regions of the cell surface and, only after it had accumulated in the basolateral domain, it began to appear on the apical surface near the intercellular junctions. The results of these experiments indicate that the VSV G protein must be sorted before its arrival at the cell surface, and suggest that passage to the apical domain occurs only late in infection when tight junctions are no longer an effective barrier. In complementary experiments, using the temperature-sensitive mutant of influenza, cultures were first shifted from the nonpermissive temperature (39.5 degrees C) to 18.5 degrees C, to allow entrance of the glycoprotein into the Golgi apparatus (see Matlin, K.S., and K. Simons, 1983, Cell, 34:233-243). Under these conditions HA accumulated in Golgi stacks and vesicles but did not reach the plasma membrane. When the temperature was subsequently shifted to 32.5 degrees C, HA rapidly appeared in discrete regions of the apical surface near, and often directly above, the Golgi elements, and later diffused throughout this surface. To ensure that the anti-HA antibodies had access to lateral domains, monolayers were treated with a hypertonic medium to dilate the intercellular spaces. Some labeling was then observed in the lateral plasma membranes soon after the shift, but this never increased beyond 1.0 gold particle/micron, whereas characteristic densities of labeling in apical surfaces soon became much higher (approximately 10 particles/micron). Our results suggest that the bulk of HA follows a direct pathway leading from the Golgi to regions of the apical surface close to trans-Golgi cisternae.

Ribophorins are two transmembrane glycoproteins characteristic of the rough endoplasmic reticulum, which are thought to be involved in the binding of ribosomes. Their biosynthesis was studied in vivo using lines of cultured rat hepatocytes (clone 9) and pituitary cells (GH 3.1) and in cell-free synthesis experiments. In vitro translation of mRNA extracted from free and bound polysomes of clone 9 cells demonstrated that ribophorins are made exclusively on bound polysomes. The primary translation products of ribophorin messengers obtained from cultured hepatocytes or from regenerating livers co-migrated with the respective mature proteins, but had slightly higher apparent molecular weights (2,000) than the unglycosylated forms immunoprecipitated from cells treated with tunicamycin. This indicates that ribophorins, in contrast to all other endoplasmic reticulum membrane proteins previously studied, contain transient amino-terminal insertion signals which are removed co-translationally. Kinetic and pulse-chase experiments with [35S]methionine and [3H]mannose demonstrated that ribophorins are not subjected to electrophoretically detectable posttranslational modifications, such as proteolytic cleavage or trimming and terminal glycosylation of oligosaccharide side chain(s). Direct analysis of the oligosaccharides of ribophorin l showed that they do not contain the terminal sugars characteristic of complex oligosaccharides and that they range in composition from Man8GlcNAc to Man5GlcNAc. These findings, as well as the observation that the mature proteins are sensitive to endoglycosidase H and insensitive to endoglycosidase D, are consistent with the notion that the biosynthetic pathway of the ribophorins does not require a stage of passage through the Golgi apparatus.

Infection of two different lines of polarized epithelial cells grown as monolayers with several types of enveloped viruses results, for each virus type, in a characteristic asymmetric budding of virions. Influenza virus (WSN strain), simian virus 5, and Sendai virus bud exclusively from the free (apical) surface of the cells, while vesicular stomatitis virus acquires its envelope only from the basolateral plasma membrane. Because different viruses select specific domains of plasma membrane in the same cell type, virus-infected epithelial monolayers can provide an excellent model system for studies of the mechanisms that generate regional differences in the distribution of plasma membrane components of epithelial cells.

Treatment of rat liver rough microsomes (3.5 mg of protein/ml) with sublytical concentrations (0.08%) of the neutral detergent Triton X-100 caused a lateral displacement of bound ribosomes and the formation of ribosomal aggregates on the microsomal surface. At slightly higher detergent concentrations (0.12-0.16%) membrane areas bearing ribosomal aggregates invaginated into the microsomal lumen and separated from the rest of the membrane. Two distinct classes of vesicles could be isolated by density gradient centrifugation from microsomes treated with 0.16% Triton X-100: one with ribosomes bound to the inner membrane surfaces ("inverted rough" vesicles) and another with no ribosomes attached to the membranes. Analysis of the fractions showed that approximately 30% of the phospholipids and 20-30% of the total membrane protein were released from the membranes by this treatment. Labeling with avidin-ferritin conjugates demonstrated that concanavalin A binding sites, which in native rough microsomes are found in the luminal face of the membranes, were present on the outer surface of the inverted rough vesicles. Freeze-fracture electron microscopy showed that both fracture faces had similar concentrations of intramembrane particles. SDS PAGE analysis of the two vesicle subfractions demonstrated that, of all the integral microsomal membrane proteins, only ribophorins I and II were found exclusively in the inverted rough vesicles bearing ribosomes. These observations are consistent with the proposal that ribophorins are associated with the ribosomal binding sites characteristic of rough microsomal membranes.

The distribution of newly synthesized proteolipid protein (PLP, 23 kdaltons) and myelin basic proteins (MBPs, 14-21.5 kdaltons) was determined in microsomal and myelin fractions prepared from the brainstems o1 10-30 d-old rats sacrificed at different times after an intracranial injection of 35S-methionine. Labeled MBPs were found in the myelin fraction 2 min after the injection, whereas PLP appeared first in the rough microsomal fraction and only after a lag of 30 min in the myelin fraction. Cell-free translation experiments using purified mRNAs demonstrated that PLP and MBPs are synthesized in bound and free polysomes, respectively. A mechanism involving the cotranslational insertion into the ER membrane and subsequent passage of the polypeptides through the Golgi apparatus is consistent with the lag observed in the appearance of the in vivo-labeled PLP in the myelin membrane. Newly synthesized PLP and MBPs are not proteolytically processed, because the primary translation products synthesized in vitro had the same electrophoretic mobility and N-terminal amino acid sequence as the mature PLP and MBP polypeptides. It was found that crude myelin fractions are highly enriched in mRNAs coding for the MBPs but not in mRNA coding for PLP. This suggests that whereas the bound polysomes synthesizing PLP are largely confined to the cell body, free polysomes synthesizing MBPs are concentrated in oligodendrocyte processes involved in myelination, which explains the immediate incorporation of MBPs into the developing myelin sheath.

Dog pancreas rough microsomes were solubilized in 1% octyl beta-glucoside, and membrane vesicles were reconstituted by slow 30-fold dilution with a buffer of low ionic strength. Asymmetric assembly of the membranes occurred during reconstitution since the vesicles formed contained ribosomes bound only to the vesicular outer surfaces. The reconstituted vesicles were similar in protein composition to native rough microsomes, although these vesicles were largely devoid of luminal-content proteins. These reconstituted vesicles could translocate and process nascent secretory (human placental lactogen) and membrane proteins (influenza hemagglutinin and rat liver ribophorin I) synthesized in cell-free translation systems programmed with the corresponding mRNAs. Signal cleavage and N-glycosylation only occurred when the reconstituted membranes were present during translation, providing evidence that the translocation apparatus was asymmetrically assembled into the reconstituted membranes. When a supernatant lacking ribosomes and particles greater than 50S from centrifuging the detergent-solubilized microsomes at high speed was used for reconstitution, smooth-surfaced membrane vesicles were obtained that, except for the absence of ribosomal proteins, were similar in protein composition to that of the reconstituted vesicles from total solubilized rough microsomes. The reconstituted smooth-surfaced vesicles, however, were totally inactive in cotranslational processing and translocation of nascent polypeptides. These findings suggest that ribosomes and/or large macromolecular complexes, not dissociated under our solubilization conditions, are essential for in vitro assembly of a functional translocation apparatus.

Administration of the thyroid hormone 3,3,5'-triiodo-L-thyronine (T3) to rats leads to a marked increase in hepatic levels of mRNA for cytochrome c. Messenger RNA prepared from the free polysomes of T3-treated rats directed the in vitro synthesis of a polypeptide which only differed in amino acid sequence from mature cytochrome c in that it contained an NH2-terminal methionine. The in vitro product was incorporated specifically into purified rat liver mitochondria and became inaccessible to added trypsin when the mitochondria were added after translation was completed. Horse heart apocytochrome c, but not the holocytochrome, could compete with the in vitro synthesized polypeptide for its uptake into mitochondria. This suggests that the primary structural features of apocytochrome c, which serve as an addressing signal for mitochondria, are masked after the acquisition of heme and that this process occurs in the mitochondria. The addressing signal seems to be contained in a specific segment of the cytochrome polypeptide because only one fragment generated by CNBr cleavage of horse apocytochrome c, extending from residue 66 to the carboxy end of the molecule, could compete with the in vitro product for its transfer into mitochondria.

We have selected the rat preputial gland beta-glucuronidase as a model protein to study the sorting of newly synthesized lysosomal hydrolases to the lysosome. The complete coding sequence of beta-glucuronidase messenger RNA was determined from the sequences of a group of overlapping cDNA clones isolated from preputial gland cDNA libraries. The beta-glucuronidase mRNA primary translation product contains 648 amino acids, including an amino-terminal signal sequence of 22 residues. The polypeptide has four potential sites for the addition of asparagine-linked core oligosaccharides. A 376-residue segment of beta-glucuronidase shows extensive homology (23% sequence identity) to a portion of Escherichia coli beta-galactosidase. This homology most likely reflects an evolutionary relationship between the bacterial and eukaryotic enzymes and the conservation of structural features necessary for the glycosidase activity of both proteins. Translation of mRNA synthesized in vitro by transcription of a cDNA containing the entire beta-glucuronidase coding region yielded a polypeptide that was immunoprecipitated with anti-beta-glucuronidase antiserum and had the same electrophoretic mobility as the primary translation product of natural beta-glucuronidase mRNA. In the presence of microsomal membranes, the in vitro-synthesized beta-glucuronidase underwent cotranslational incorporation into the microsomes, as indicated by removal of the signal sequence and the addition of several oligosaccharide chains. The beta-glucuronidase cDNA will provide a useful tool to study the mechanism of mannose phosphorylation and other aspects of the sorting of lysosomal enzymes to lysosomes.

When synthesized in polarized epithelial cells, the envelope glycoproteins hemagglutinin of influenza and G of vesicular stomatitis virus are targeted to the apical and basolateral plasma membranes, respectively. To determine which portions of these transmembrane proteins contain information necessary for their sorting, the behavior of two different G-hemagglutinin chimeric polypeptides, consisting of all or nearly all the luminal portion of the vesicular stomatitis virus G protein linked to C-terminal segments of influenza hemagglutinin that included its transmembrane and cytoplasmic domains, was studied in MDCK cells transformed with the corresponding cDNAs. Both chimeras were transported from the endoplasmic reticulum to the Golgi apparatus and from there to the cell surface with the same rapid kinetics as the intact G protein. By using a cell surface immunoprecipitation assay with monolayers cultured on permeable filters that allows the recovery of labeled protein molecules present in each cell surface domain, it was found that both chimeric proteins as well as the intact G protein were delivered almost exclusively to the basolateral surface. This polarized distribution of the polypeptides did not change during a subsequent 90-min chase period, although during this time a large fraction of the glycoprotein molecules underwent degradation. In addition, a small fraction of the cell surface-associated glycoprotein molecules shed their ectoplasmic segments into the basolateral compartment, apparently as a result of a proteolytic cleavage. Immunofluorescence on transverse frozen sections and immunoelectron microscopy revealed a prominent accumulation of the chimeric polypeptides in the lateral cell membranes, with lesser amounts on the basal and apical surfaces. These results indicate that information specifying the basolateral transport of the G glycoprotein is located within the first 426 N-terminal amino acids of its ectoplasmic portion.

In confluent monolayers of the dog kidney epithelial cell line Madin-Darby canine kidney (MDCK) assembly of RNA enveloped viruses reflects the functional polarization of the cells. Thus, influenza, Sendai, and Simian virus 5 bud from the apical (free) surface, while vesicular stomatitis virions (VSV) are assembled at basolateral plasma membrane domains (Rodriguez-Boulan, E., and D.D. Sabatini, 1978, Proc. Natl. Acad. Sci. U.S.A., 75:5071-5075). MDCK cells derived from confluent monolayers by dissociation with trypsin-EDTA and maintained as single cells in spinner medium for 12-20 h before infection, lose their characteristic structural polarity. Furthermore, when these cells were infected with influenza or VSV, virions assembled in a nonpolarized fashion over most of the cell surface. However, when dissociated MDCK cells infected in suspension were sparsely plated on collagen gels to prevent intercellular contact and the formation of junctions, the characteristic polarity of viral budding observed in confluent monolayers was again manifested; i.e., VSV budded preferentially from adherent surfaces and influenza almost exclusively from free surface regions. Similar polarization was observed in cells which became aggregated during incubation in spinner medium: influenza budded from the free surface, while VSV was produced at regions of cell-cell contact. It therefore appears that in isolated epithelial cells attachment to a substrate or to another cell is sufficient to trigger the expression of plasma membrane polarity which is manifested in the asymmetric budding of viruses.

Madin-Darby canine kidney (MDCK) cells can sustain double infection with pairs of viruses of opposite budding polarity (simian virus 5 [SV5] and vesicular stomatitis virus [VSV] or influenza and VSV), and we observed that in such cells the envelope glycoproteins of the two viruses are synthesized simultaneously and assembled into virions at their characteristic sites. Influenza and SV5 budded exclusively from the apical plasma membrane of the cells, while VSV emerged only from the basolateral surfaces. Immunoelectron microscopic examination of doubly infected MDCK cells showed that the influenza hemagglutinin (HA) and the VSV G glycoproteins traverse the same Golgi apparatus and even the same Golgi cisternae. This indicates that the pathways of the two proteins towards the plasma membrane do not diverge before passage through the Golgi apparatus and therefore that critical sorting steps must take place during or after passage of the glycoproteins through this organelle. After its passage through the Golgi, the HA accumulated primarily at the apical membrane, where influenza virion assembly occurred. A small fraction of HA did, however, appear on the lateral surface and was incorporated into the envelope of budding VSV virions. Although predominantly found on the basolateral surface, significant amounts of G protein were observed on the apical plasma membrane well before disruption of the tight junctions was detectable. Nevertheless, assembly of VSV virions was restricted to the basolateral domain and in doubly infected cells the G protein was only infrequently incorporated into the envelope of budding influenza virions. These observations indicate that the site of VSV budding is not determined exclusively by the presence of G polypeptides. Therefore, it is likely that, at least for VSV, other cellular or viral components are responsible for the selection of the appropriate budding domain.

Previous studies (Rindler, M. J., I. E., Ivanov, H. Plesken, and D. D. Sabatini, 1985, J. Cell Biol., 100: 136-151; Rindler, M. J., I. E. Ivanov, H. Plesken, E. J. Rodriguez-Boulan, and D. D. Sabatini, 1984, J. Cell Biol., 98: 1304-1319) have demonstrated that in polarized Madin-Darby canine kidney cells infected with vesicular stomatitis virus (VSV) or influenza virus the viral envelope glycoproteins G and HA are segregated to the basolateral and apical plasma membrane domains, respectively, where budding of the corresponding viruses takes place. Furthermore, it has been shown that this segregation of the glycoproteins reflects the polarized delivery of the newly synthesized polypeptides to each surface domain. In transfection experiments using eukaryotic expression plasmids that contain cDNAs encoding the viral glycoproteins, it is now shown that even in the absence of other viral components, both proteins are effectively segregated to the appropriate cell surface domain. In transfected cells, the HA glycoprotein was almost exclusively localized in the apical cell surface, whereas the G protein, although preferentially localized in the basolateral domains, was also present in lower amounts, in the apical surfaces of many cells. Using transfected and infected cells, it was demonstrated that, after reaching the cell surface, the G protein, but not the HA protein, undergoes interiorization by endocytosis. Thus, in the presence of chloroquine, a drug that blocks return of interiorized plasma membrane proteins to the cell surface, the G protein was quantitatively trapped in endosome- or lysosome-like vesicles. The sequestration of G was a rapid process that was completed in many cells by 1-2 h after chloroquine treatment. The fact that in transfected cells the surface content of G protein was not noticeably reduced during a 5-h incubation with cycloheximide, a protein synthesis inhibitor that did not prevent the effect of chloroquine, implies that normally, G protein molecules are not only interiorized but are also recycled to the cell surface.

Various aspects of the biogenetic mechanisms that are involved in the insertion of nascent plasma membrane proteins into the endoplasmic reticulum (ER) membrane and their subsequent distribution through the cell have been investigated. For these studies chimeric genes that encode hybrid proteins containing carboxy-terminal portions of the influenza virus hemagglutinin (154 amino acids) or the vesicular stomatitis virus envelope glycoprotein (G) (60 amino acids) linked to the carboxy terminus of a nearly complete secretory polypeptide, growth hormone (GH), were used. In in vitro transcription-translation experiments, it was found that the insertion signal in the GH portion of the chimeras led to incorporation of the membrane protein segments into the ER membrane. Effectively, GH became part of the luminal segment of membrane proteins of which only very small segments, corresponding to the cytoplasmic portions of the G or HA proteins, remained exposed on the surface of the microsomes. When the chimeric genes were expressed in transfected cells, the products, as expected, failed to be secreted and remained cell-associated. These results support the assignment of a halt transfer role to segments of the membrane polypeptides that include their transmembrane portions. The hybrid polypeptide containing the carboxy-terminal portion of HA linked to GH accumulated in a juxtanuclear region of the cytoplasm within modified ER cisternae, closely apposed to the Golgi apparatus. The location and appearance of these cisternae suggested that they represent overdeveloped transitional ER elements and thus may correspond to a natural way station between the ER and the Golgi apparatus, in which further transfer of the artificial molecules is halted. The GH-G hybrid could only be detected in transfected cells treated with chloroquine, a drug that led to its accumulation in the membranes of endosome or lysosome-like cytoplasmic vesicles. Although the possibility that the chimeric protein entered such vesicles directly from the Golgi apparatus cannot be ruled out, it appears more likely that it was first transferred to the cell surface and was then internalized by endocytosis.

Ribophorin I is a type I transmembrane glycoprotein specific to the rough endoplasmic reticulum. We have previously shown that, when expressed in transfected HeLa cells, a carboxyl-terminally truncated form of ribophorin I that contains most of the luminal domain (RI332) is, like the native protein, retained in the endoplasmic reticulum (ER). Brefeldin A (BFA) treatment of these HeLa cells leads to O-glycosylation of RI332 by glycosyltransferases that are redistributed from the Golgi apparatus to the ER (Ivessa, N. E., De Lemos-Chiarandini, C., Tsao, Y.-S., Takatsuki, A., Adesnik, M., Sabatini, D. D., and Kreibich, G. (1992) J. Cell Biol. 117, 949-958). Using the state of glycosylation of RI332 as a measure for the BFA-induced backflow of enzymes of the Golgi apparatus to the ER, we now demonstrate that the retrograde transport is inhibited when cells are treated with various agents that affect intracellular Ca2+ concentrations, such as the dipeptide benzyloxycarbonyl (Cbz)-Gly-Phe-amide, the Ca2+ ionophore A23187, and thapsigargin, an inhibitor of the Ca(2+)-transporting ATPase of the ER. These treatments prevent the BFA-induced O-glycosylation of RI332. Immunofluorescence localization of the Golgi markers, MG-160 and galactosyltransferase, shows that when BFA is applied in the presence of Ca2+ modulating agents, the markers remain confined to the Golgi apparatus and are not redistributed to the ER, as is the case when BFA alone is used. Cbz-Gly-Phe-amide does not, however, interfere with the BFA-induced release of beta-COP from the Golgi apparatus. We conclude that the maintenance of a Ca2+ gradient between the cytoplasm and the lumen of the ER and the Golgi apparatus is required for the BFA-induced retrograde transport from the Golgi apparatus to the ER to occur.

Messenger RNA (mRNA) of membrane-bound polysomes in a membrane fraction of WI-38 cells remains associated with the microsomal membranes even after ribosomes and their nascent polypeptide chains are removed by using puromycin in a high salt buffer or by disassembling the ribosomes in a medium of high ionic strength lacking magnesium. mRNA either was specifically labeled in the presence of actinomycin D, or it was recognized by virtue of its affinity for oligo-dT. Poly A segments in bound mRNAs have an electrophoretic mobility in acrylamide gels which is characteristic of cytoplasmic mRNAs and corresponds to 150-200 adenyl residues. Extensive RNase treatment did not lead to release of the poly A segments of membrane-associated mRNA molecules either from an intact membrane fraction or from a membrane fraction previously stripped of ribosomes. On the other hand, RNase treatment led to the release and digestion of the nonpoly A segments of the mRNA molecules, indicating that the site of attachment of mRNA to the ER membranes is located near or at the 3' end of the molecule which contains the poly A. A direct association of mRNAs and endoplasmic reticulum membranes is considered in a modelto explain the assembly of bound polysomes and protein synthesis in a membrane-associated apparatus.

Membrane-bound ribosomes and messenger RNA remained associated with the microsomal membranes of human fibroblasts after cultures were treated with Verrucarin A, an inhibitor of initiation which led to extensive run-off of ribosomes from polysomal structures. When a membrane fraction from Verrucarin-treated cells containing such inactive ribosomes and mRNA was suspended in a medium of high salt concentration, extensive release of ribosomal subunits occurred without the need for puromycin. The mRNA nevertheless remained associated with the membranes. These results add support to the conclusion that, in human fibroblasts, mRNA is bound directly to ER membranes, independently of the ribosomes and nascent polypeptide chains.

The lateral mobility of ribosomes bound to rough endoplasmic reticulum (RER) membranes was demonstrated under experimental conditions. High-salt-washed rough microsomes were treated with pancreatic ribonuclease (RNase) to cleave the mRNA of bound polyribosomes and allow the movement of individual bound ribosomesmfreeze-etch and thin-section electron microscopy demonstrated that, when rough microsomes were treated with RNase at 4 degrees C and then maintained at this temperature until fixation, the bound ribosomes retained their homogeneous distribution on the microsomal surface. However, when RNase-treated rough microsomes were brought to 24 degrees C, a temperature above the thermotropic phase transition of the microsomal phospholipids, bound ribosomes were no longer distributed homogeneously but, instead, formed large, tightly packed aggregates on the microsomal surface. Bound polyribosomes could also be aggregated by treating rough microsomes with antibodies raised against large ribosomal subunit proteins. In these experiments, extensive cross-linking of ribosomes from adjacent microsomes also occurred, and large ribosome-free membrane areas were produced. Sedimentation analysis in sucrose density gradients demonstrated that the RNase treatment did not release bound ribosomes from the membranes; however, the aggregated ribosomes remain capable of peptide bond synthesis and were released by puromycin. It is proposed that the formation of ribosomal aggregates on the microsomal surface results from the lateral displacement of ribosomes along with their attached binding sites, nascent polypeptide chains, and other associated membrane proteins; The inhibition of ribosome mobility after maintaining rough microsomes at 4 degrees C after RNase, or antibody, treatment suggests that the ribosome binding sites are integral membrane proteins and that their mobility is controlled by the fluidity of the RER membrane. Examination of the hydrophobic interior of microsomal membranes by the freeze-fracture technique revealed the presence of homogeneously distributed 105-A intramembrane particles in control rough microsomes. However, aggregation of ribosomes by RNase, or their removal by treatment with puromycin, led to a redistribution of the particles into large aggregates on the cytoplasmic fracture face, leaving large particle-free regions.

The co-translational insertion of polypeptides into endoplasmic reticulum membranes may be initiated by cleavable amino-terminal insertion signals, as well as by permanent insertion signals located at the amino-terminus or in the interior of a polypeptide. To determine whether the location of an insertion signal within a polypeptide affects its function, possibly by affecting its capacity to achieve a loop disposition during its insertion into the membrane, we have investigated the functional properties of relocated insertion signals within chimeric polypeptides. An artificial gene encoding a polypeptide (THA-HA), consisting of the luminal domain of the influenza hemagglutinin preceded by its amino-terminal signal sequence and linked at its carboxy-terminus to an intact prehemagglutinin polypeptide, was constructed and expressed in in vitro translation systems containing microsomal membranes. As expected, the amino-terminal signal initiated co-translational insertion of the hybrid polypeptide into the membranes. The second, identical, interiorized signal, however, was not recognized by the signal peptidase and was translocated across the membrane. The failure of the interiorized signal to be cleaved may be attributed to the fact that it enters the membrane as part of a translocating polypeptide and therefore cannot achieve the loop configuration that is thought to be adopted by signals that initiate insertion. The finding that the interiorized signal did not halt translocation of downstream sequences, even though it contains a hydrophobic region and must enter the membrane in the same configuration as natural stop-transfer signals, indicates that the HA insertion signal lacks essential elements of halt transfer signals that makes the latter effective membrane-anchoring domains. When the amino-terminal insertion signal of the THA-HA chimera was deleted, the interior signal was incapable of mediating insertion, probably because of steric hindrance by the folded preceding portions of the chimera. Several chimeras were constructed in which the interiorized signal was preceded by polypeptide segments of various lengths. A signal preceded by a segment of 111 amino acids was also incapable of initiating insertion, but insertion took place normally when the segment preceding the signal was only 11-amino acids long.(ABSTRACT TRUNCATED AT 400 WORDS)

A monoclonal antibody (2C5) raised against rat liver lysosomal membranes was used to identify a 78-kD glycoprotein that is present in the membranes of both endosomes and lysosomes and, therefore, is designated endolyn-78. In cultures of rat hepatoma (Fu5C8) and kidney cells (NRK), this glycoprotein could not be labeled with [35S]methionine or with [32P]inorganic phosphate but was easily labeled with [35S]cysteine and [3H]mannose. Pulse-chase experiments and determinations of endoglycosidase H (endo H) sensitivity showed that endolyn-78 is derived from a precursor of Mr 58-62 kD that is processed to the mature form with a t1/2 of 15-30 min. The protein has a 22-kD polypeptide backbone that is detected after a brief pulse in tunicamycin-treated cells. During a chase in the presence of the drug, this is converted into an O-glycosylated product of 46 kD that despite the absence of N-linked oligosaccharides is effectively transferred to lysosomes. This demonstrates that the delivery of endolyn-78 to this organelle is not mediated by the mannose-6-phosphate receptor (MPR). Immunocytochemical experiments showed that endolyn-78 is present in the limiting membranes and the interior membranous structures of morphologically identifiable secondary lysosomes that contain the lysosomal hydrolase beta-glucuronidase, lack the MPR, and could not be labeled with alpha-2-macroglobulin at 18.5 degrees C, a temperature which prevents appearance of endocytosed markers in lysosomes. Endolyn-78 was present at low levels in the plasma membrane and in peripheral tubular endosomes, but was prominent in morphologically diverse components of the endosomal compartment (vacuolar endosomes and various types of multivesicular bodies) which acquired alpha-2-macroglobulin at 18.5 degrees C, and frequently contained substantial levels of the MPR and variable levels of beta-glucuronidase. On the other hand, the MPR was very rarely found in endolyn-containing structures that were not labeled with alpha-2-macroglobulin at the low temperature. Thus, the process of lysosomal maturation appears to involve the progressive delivery of lysosomal enzymes to various types of endosomes that may have already received some of the lysosomal membrane proteins. Although endolyn-78 would be one of the proteins added early to endosomes, other lysosomal membrane proteins may be added only to multivesicular endosomes that represent very advanced stages of maturation.

Fifteen peptides, ranging in length from 6 to 31 amino acids and corresponding in sequence to portions of the major phenobarbital-inducible form of rat liver cytochrome P-450 (P-450 PB-4), were previously synthesized chemically and used to prepare site-specific rabbit antibodies (Frey, A. B., D.J. Waxman, and G. Kreibich, 1985, J. Biol. Chem., 260:15253-15265). The antipeptide antibodies were affinity purified using Sepharose resins derivatized with the respective peptides and 14 preparations were obtained that in an ELISA assay showed affinities to immobilized P-450 judged to be adequate for binding studies on intact rat liver microsomes. The binding of these antibodies to rough microsomes from the livers of phenobarbital treated rats was assessed using 125I-labeled IgG and by immunoelectron microscopy employing protein A-gold as a marker. It was found that many of the antibodies bound to the cytoplasmic surface of the membrane but none bound to the luminal face of ruptured or inverted microsomal vesicles or to contaminating membranes of other organelles present in the preparations. These observations eliminate previously proposed models for the transmembrane disposition of P-450 that postulate the existence of multiple transmembrane domains and the exposure of several polar segments of the polypeptide on the luminal side of the membrane. The fact that an antibody raised to the first 31 residues of P-450 bound well to the purified P-450 but very poorly to rough microsomes, whereas an antibody to a peptide comprising residues 24-38 showed relatively strong binding to intact microsomes, is consistent with the proposal that the amino terminal segment of P-450 extending approximately to residue 20 is embedded in the phospholipid bilayer and the immediately following segment is exposed on the cytoplasmic surface of the membrane. All these results favor a model in which the cytochrome P-450 molecule is largely exposed on the cytoplasmic surface of the endoplasmic reticulum membrane to which it is anchored by its short amino terminal hydrophobic segment.