Dissection of the Golgi complex. I. Monensin inhibits the transport of viral membrane proteins from medial to trans Golgi cisternae in baby hamster kidney cells infected with Semliki Forest virus

Baby hamster kidney (BHK) cells were infected with Semliki Forest virus (SFV) and, 2 h later, were treated for 4 h with 10 microM monensin. Each of the four to six flattened cisternae in the Golgi stack became swollen and separated from the others. Intracellular transport of the viral membrane proteins was almost completely inhibited, but their synthesis continued and they accumulated in the swollen Golgi cisternae before the monensin block. In consequence, these cisternae bound large numbers of viral nucleocapsids and were easily distinguished from other swollen cisternae such as those after the block. These intracellular capsid-binding membranes (ICBMs) were not stained by cytochemical markers for endoplasmic reticulum (ER) (glucose-6-phosphatase) or trans Golgi cisternae (thiamine pyrophosphatase, acid phosphatase) but were labeled by Ricinus communis agglutinin I (RCA) in thin, frozen sections. Since this lectin labels only Golgi cisternae in the middle and on the trans side of the stack (Griffiths, G., R. Brands, B. Burke, D. Louvard, and G. Warren, 1982, J. Cell Biol., 95:781-792), we conclude that ICBMs are derived from Golgi cisternae in the middle of the stack, which we term medial cisternae. The overall movement of viral membrane proteins appears to be from cis to trans Golgi cisternae (see reference above), so monensin would block movement from medial to the trans cisternae. It also blocked the trimming of the high-mannose oligosaccharides bound to the viral membrane proteins and their conversion to complex oligosaccharides. These functions presumably reside in trans Golgi cisternae. This is supported by data in the accompanying paper, in which we also show that fatty acids are covalently attached to the viral membrane proteins in the cis or medial cisternae. We suggest that the Golgi stack can be divided into three functionally distinct compartments, each comprising one or two cisternae. The viral membrane proteins, after leaving the ER, would all pass in sequence from the cis to the medial to the trans compartment.     

Viral membrane proteins acquire galactose in trans Golgi cisternae during intracellular transport

Frozen, thin sections of baby hamster kidney (BHK) cells were incubated with either concanavalin A (Con A) or Ricinus communis agglutinin I (RCA) to localize specific oligosaccharide moieties in endoplasmic reticulum (ER) and Golgi membranes. These lectins were then visualized using an anti-lectin antibody followed by protein A conjugated to colloidal gold. All Golgi cisternae and all ER membranes were uniformly labeled by Con A. In contrast, RCA gave a uniform labeling of only half to three-quarters of those cisternae on the trans side of the Golgi stack; one or two cis Golgi cisternae and all ER membranes were essentially unlabeled. This pattern of lectin labeling was not affected by infection of the cells with Semliki Forest virus (SFV). Infected cells transport only viral spike glycoproteins from their site of synthesis in the ER to the cell surface via the stacks of Golgi cisternae where many of the simple oligosaccharids on the spike proteins are converted to complex ones (Green, J., G. Griffiths, D. Louvard, P. Quinn, and G. Warren. 1981. J. Mol. Biol. 152:663-698). It is these complex oligosaccharides that were shown, by immunoblotting experiments, to be specifically recognized by RCA. Loss of spike proteins from Golgi cisternae after cycloheximide treatment (Green et al.) was accompanied by a 50% decrease in the level of RCA binding. Hence, about half of the RCA bound to Golgi membranes in thin sections was bound to spike proteins bearing complex oligosaccharides and these were restricted to the trans part of the Golgi stack. Our results strongly suggest that complex oligosaccharides are constructed in trans Golgi cisternae and that the overall movement of spike proteins is from the cis to the trans side of the Golgi stack.     

Attachment of terminal N-acetylglucosamine to asparagine-linked oligosaccharides occurs in central cisternae of the Golgi stack

Using monoclonal antibodies and electron microscopy, we have localized N-acetylglucosamine transferase I within the Golgi apparatus. This enzyme initiates the conversion of asparagine-linked oligosaccharides to the complex type. We have found that the enzyme is concentrated in the central (or medial) cisternae of the Golgi stack. Cisternae at the cis and trans ends of the Golgi complex appear to lack this protein. These experiments establish a function for the medial portion of the Golgi and imply that the Golgi is partitioned into at least three biochemically and morphologically distinct cisternal compartments.     

Tomographic evidence for continuous turnover of Golgi cisternae in Pichia pastoris

The budding yeast Pichia pastoris contains ordered Golgi stacks next to discrete transitional endoplasmic reticulum (tER) sites, making this organism ideal for structure-function studies of the secretory pathway. Here, we have used P. pastoris to test various models for Golgi trafficking. The experimental approach was to analyze P. pastoris tER-Golgi units by using cryofixed and freeze-substituted cells for electron microscope tomography, immunoelectron microscopy, and serial thin section analysis of entire cells. We find that tER sites and the adjacent Golgi stacks are enclosed in a ribosome-excluding "matrix." Each stack contains three to four cisternae, which can be classified as cis, medial, trans, or trans-Golgi network (TGN). No membrane continuities between compartments were detected. This work provides three major new insights. First, two types of transport vesicles accumulate at the tER-Golgi interface. Morphological analysis indicates that the center of the tER-Golgi interface contains COPII vesicles, whereas the periphery contains COPI vesicles. Second, fenestrae are absent from cis cisternae, but are present in medial through TGN cisternae. The number and distribution of the fenestrae suggest that they form at the edges of the medial cisternae and then migrate inward. Third, intact TGN cisternae apparently peel off from the Golgi stacks and persist for some time in the cytosol, and these "free-floating" TGN cisternae produce clathrin-coated vesicles. These observations are most readily explained by assuming that Golgi cisternae form at the cis face of the stack, progressively mature, and ultimately dissociate from the trans face of the stack.     

Transferrin receptors recycle to cis and middle as well as trans Golgi cisternae in Ig-secreting myeloma cells

The recycling itinerary of plasma membrane transferrin receptors (TFR) was charted in IgG-secreting mouse myeloma cells (RPC 5.4) by tagging surface receptors with either bound anti-transferrin receptor antibodies (anti-TFR) or Fab fragments thereof and determining the intracellular destinations of the tagged receptors by immunocytochemistry. By immunofluorescence, TFR tagged with either probe were seen to be rapidly internalized and translocated from the cell surface to the juxtanuclear (Golgi) region. When localized by immunoperoxidase procedures at the electron microscopic level, the anti-TFR-labeled receptors were detected in all cisternae (cis, middle, and trans) of the Golgi stacks as well as in endosomes and trans Golgi reticular elements. There was no difference in the routing of TFR tagged with monovalent Fab and those tagged with divalent IgG. Tagged receptors were detected in Golgi stacks of approximately 50% of the cells analyzed. The position of the labeled cisternae within a given stack was found to be quite variable with cis and middle cisternae more often labeled at 5 min and trans cisternae at 30 min of antibody uptake. The finding that recycling plasmalemmal TFR can visit all or most Golgi subcompartments raises the likely possibility that any Golgi-associated posttranslational modification can occur during recycling as well as during the initial biosynthesis of plasmalemma receptors and other membrane proteins.     

Vesicular stomatitis virus glycoprotein, albumin, and transferrin are transported to the cell surface via the same Golgi vesicles

Human hepatoma cells, infected by vesicular stomatitis virus, offer a good system to study simultaneously the intracellular localization of a well defined transmembrane glycoprotein (VSV-G), a secretory glycoprotein (transferrin), and a nonglycosylated secretory protein (albumin). We used monospecific antibodies in combination with 5- and 8- nm colloidal gold particles complexed with protein A to immunolabel these proteins simultaneously in thin frozen sections of hepatoma cells. VSV-G, transferrin, and albumin are present in the same rough endoplasmic reticulum cisternae, the same Golgi compartments, and the same secretory vesicles. In the presence of the ionophore monensin intracellular transport is blocked at the trans cisternae of the Golgi complex, and VSV-G, transferrin, and albumin accumulate in dilated cisternae, which are apparently derived from the trans-Golgi elements. Glycoproteins, synthesized and secreted in the presence of monensin, are less acidic than those in control cultures. This is probably caused by a less efficient contact between the soluble secretory proteins and the membrane-bound glycosyltransferases that are present in the most monensin-affected (trans) Golgi cisternae.     

Detection of glycosaminoglycans in the Golgi complex of chondrocytes

Elongation and sulfation of glycosaminoglycans are pivotal roles of the Golgi complex during the biosynthesis of proteoglycan monomers. In the present work the spatial relationship between these processes has been investigated by using a combination of immunocytochemical and cytochemical techniques. Chondroitin sulfate and keratan sulfate glycosaminoglycans were immunocytochemically localized in 1 to 2 transmost cisternae, also in a system of narrow tubules at the trans face of the Golgi complex of chick epiphyseal chondrocytes. At these same locations sulfate groups were revealed with the high iron diamine (HID) method, proteoglycan monomers being visualized with ruthenium red. Several treatments were assayed in order to reversibly block the secretory pathway. Chondrocytes incubated at a low temperature, 15 degrees C, before fixation, showed both glycosaminoglycans in the middle cisternae of the Golgi stack as well as the above mentioned locations. After low temperature treatment both HID and ruthenium red stained the middle, but not the cis cisternae. Incubation of the cells for 30 min with either diethylcarbamazine or monensin before fixation permitted detection of glycosaminoglycans and proteoglycan monomers in the middle cisternae, whereas HID staining of the Golgi complex, but not that of secretory vesicles, was abolished. The results show that elongation of both chondroitin sulfate and keratan sulfate glycosaminoglycans takes place in the same Golgi compartments. These include the middle cisternae and probably also the trans cisternae and tubules. Also suggested is that sulfation of one or both types of glycosaminoglycans begins in the middle cisternae.     

Monensin and brefeldin A differentially affect the phosphorylation and sulfation of secretory proteins.

Chromogranin B and secretogranin II, two members of the granin family, are known to be post-translationally modified by the addition of O-linked carbohydrates to serine and/or threonine, phosphate to serine and threonine, and sulfate to carbohydrate and tyrosine residues. In the present study, chromogranin B and secretogranin II were used as model proteins to investigate in which subcompartment of the Golgi complex secretory proteins become phosphorylated. Monensin, a drug known to block the transport from the medial to the trans cisternae of the Golgi stack, inhibited the phosphorylation of the granins, indicating that this modification occurred distal to the medial Golgi. Monensin also blocked the addition of galactose to O-linked carbohydrates and the sulfation of the granins, confirming previous data that these modifications take place in the trans Golgi. To distinguish, within the trans Golgi, between the trans cisternae of the Golgi stack and the trans Golgi network, we made use of the previous observation that brefeldin A results in the redistribution to the endoplasmic reticulum of membrane-bound enzymes of the trans cisternae of the Golgi stack, but not of the trans Golgi network. Brefeldin A treatment abolished granin sulfation but resulted in the accumulation of phosphorylated and galactosylated granins. Differential effects of brefeldin A on membranes of the Golgi stack versus the trans Golgi network were also observed by immunofluorescence analysis of marker proteins specific for either compartment. Our results suggest that the phosphorylation of secretory proteins, like their galactosylation, largely occurs in the trans cisternae of the Golgi stack, whereas the sulfation of secretory proteins on both carbohydrate and tyrosine residues takes place selectively in the trans Golgi network.     

Distribution of glycosyltransferases among Golgi apparatus subfractions from liver and hepatomas of the rat.

Glycosyltransferase activities of highly purified fractions of Golgi apparatus, plasma membrane and endoplasmic reticulum, all from the same homogenates, were analyzed and compared. Additionally, Golgi apparatus were unstacked and the individual cisternae separated into fractions enriched in cis, median and trans elements using the technique of preparative free-flow electrophoresis. Golgi apparatus from both liver and hepatomas were enriched in all glycosyltransferases compared to endoplasmic reticulum and plasma membranes. However, Golgi apparatus from hepatomas showed both elevated fucosyltransferase and galactosyltransferase activities but reduced sialyltransferase and dipeptidyl peptidase IV (DPP IV) activities compared to liver. Activity of N-acetylglucosaminyltransferase was approximately the same in both liver and hepatoma Golgi apparatus. With normal liver, sialyl- and galactosyltransferase activities and DPP IV showed a marked cis-to-trans gradient of activity. Fucosyltransferase was concentrated in two regions of the electrophoretic separations, one corresponding to cis cisternae and one corresponding to trans cisternae. N-Acetylglucosaminyltransferase activity was more widely distributed but the endogenous acceptor activity was predominantly cis. With hepatoma Golgi apparatus, the pattern for DPP IV was similar to that for liver but those of sialyl- and galactosyltransferases differed markedly from liver. Instead of activity increasing cis to trans, the activities for sialyl- and galactosyltransferases decreased. For fucosyltransferases, activity dependent on exogenous acceptor was medial whereas with endogenous acceptor, two activity peaks, cis and trans, still were observed. For N-acetylglucosaminyltransferase the pattern for hepatoma was similar to that for liver. The results indicate alterations in the distribution of glycosyltransferase activities within the Golgi apparatus in hepatotumorigenesis that may reflect altered cell surface glycosylation patterns.     

Transport of horseradish peroxidase from the cell surface to the Golgi in insulin-secreting cells: preferential labelling of cisternae located in an intermediate position in the stack.

We have used serial sectioning to study the topology of Golgi cisternae in insulin-secreting cells during secretion-stimulated endocytotic uptake of exogenous horseradish peroxidase (HRP). HRP-labelled cisternae were followed on several series of consecutive sections. This revealed that labelled cisternae could always be traced to a position in the Golgi stack intermediate between the cis and the trans poles. This occurred in spite of the apparent cis or trans locations of HRP-containing cisternae on some sections. The latter images could be explained by the lack of the true cis or trans (clathrin-coated) cisternae at certain levels of the stack.     

The Golgi apparatus ofPhytophthora cinnamomi makes three types of secretory or storage vesicles concurrently

Summary The formation of three types of vesicles in the oomycetePhytophthora cinnamomi was investigated using ultrastructural and immunocytochemical techniques. All three vesicles are synthesised at the same time; one type serves a storage role; the others undergo regulated secretion. A monoclonal antibody Lpv-1 that is specific for glycoproteins contained in the storage vesicles labelled the endoplasmic reticulum (ER), elements in the transition region between ER and Golgi stack, and cis, medial and trans Golgi cisternae. Cpa2, a monoclonal antibody specific for glycoproteins contained within secretory dorsal vesicles labelled the transition region, cis cisternae and a trans-Golgi network. Vesicles possessing a structure characteristic of mature secretory ventral vesicles were observed in close association with the trans face of Golgi stacks. The results suggest that all three vesicles are formed by the Golgi apparatus. Double immunogold labelling with Lpv-1 and Cpa-2 showed that these two sets of glycoproteins occurred within the same Golgi cisternae, indicating that both products pass through and are sorted concurrently within a single Golgi stack.     

Asymmetrical distribution of clathrin in the Golgi apparatus of polypeptide-secreting cells

Using an antibody revealed by the protein A-gold technique, we have studied the distribution of clathrin antigenic sites in the Golgi area of pancreatic B-cells. Golgi compartments showing an immunolabelling comprised extensive segments of cisternae, typical coated vesicles, dilated extremities of cisternae with condensing secretory material, and newly formed secretory granules. Most of the labelled membranes were observed at the trans Golgi pole while little immunoreactivity was found on the cis pole.     

Structure of Golgi apparatus

Summary Golgi apparatus (GA) of eukaryotic cells consist of one or more stacks of flattened saccules (cisternae) and an array of fenestrae and tubules continuous with the peripheral edges of the saccules. Golgi apparatus also are characterized by zones of exclusion that surround each stack and by an assortment of vesicles (or vesicle buds) associated with both the stacks and the peripheral tubules of the stack cisternae. Each stack (sometimes referred to as Golgi apparatus, Golgi complex, or dictyosome) is structurally and functionally polarized, reflecting its role as an intermediate between the endoplasmic reticulum, the cell surface, and the lysosomal system of the cell. There is probably only one GA per cell, and all stacks of the GA appear to function synchronously. All Golgi apparatus are involved in the generation and movement of product and membrane within the cell or to the cell exterior, and these functions are often reflected as structural changes across the stacks. For example, in plants, both product and membrane appear to maturate from the cis to the trans poles of the stacks in a sequential, or serial, manner. However, there is also strong ultrastructural evidence in plants for a parallel input to the stack saccules, probably through the peripheral tubules. The same modes of functioning probably also occur in animal GA; although here, the parallel mode of functioning almost surely predominates. In some cells at least, GA stacks give rise to tubular-vesicular structures that resemble the trans Golgi network. Rudimentary GA, consisting of tubular-vesicular networks, have been identified in fungi and may represent an early stage of GA evolution.     

The ion channel activity of the influenza virus M2 protein affects transport through the Golgi apparatus

High level expression of the M2 ion channel protein of influenza virus inhibits the rate of intracellular transport of the influenza virus hemagglutinin (HA) and that of other integral membrane glycoproteins. HA coexpressed with M2 is properly folded, is not associated with GRP78- BiP, and trimerizes with the same kinetics as when HA is expressed alone. Analysis of the rate of transport of HA from the ER to the cis and medial golgi compartments and the TGN indicated that transport through the Golgi apparatus is delayed. Uncleaved HA0 was not expressed at the cell surface, and accumulation HA at the plasma membrane was reduced to 75-80% of control cells. The delay in intracellular transport of HA on coexpression of M2 was not observed in the presence of the M2-specific ion channel blocker, amantadine, indicating that the Golgi transport delay is due to the M2 protein ion channel activity equilibrating pH between the Golgi lumen and the cytoplasm, and not due to saturation of the intracellular transport machinery. The Na+/H+ ionophore, monensin, which also equilibrates pH between the Golgi lumen and the cytoplasm, caused a similar inhibition of intracellular transport as M2 protein expression did for HA and other integral membrane glycoproteins. EM data showed a dilation of Golgi cisternae in cells expressing the M2 ion channel protein. Taken together, the data suggest a similarity of effects of M2 ion channel activity and monensin on intracellular transport through the Golgi apparatus.     

Golgi apparatus cisternae of monensin-treated cells accumulate in the cytoplasm of liver slices

Protein transport via the endoplasmic reticulum Golgi apparatus-cell surface export route was blocked when slices (6-15 cells thick) of livers of 10-day-old rats were incubated with 1 microM monensin. Production of secretory vesicles by Golgi apparatus was reduced or eliminated and, in their place, swollen cisternae accumulated in the cytoplasm at the trans Golgi apparatus face. The swelling response was restricted to the six external cell layers of the liver slices, and the number of cells showing the response was little increased by either a greater concentration of monensin or by longer times of incubation. When monensin was added post-chase to the slices, flux of radioactive proteins to the cell surface was inhibited by about 80% as determined from standard pulse-chase analyses with isolated cell fractions. Radioactive proteins accumulated in both endoplasmic reticulum and Golgi apparatus and in a fraction that may contain monensin-blocked Golgi apparatus cisternae released from the stack. The latter fraction was characterized by galactosyltransferase/thiamine pyrophosphatase ratios similar to those of Golgi apparatus from control slices. The use of monensin with the tissue slice system may provide an opportunity for the cells to accumulate monensin-blocked Golgi apparatus cisternae in sufficient quantities to permit their isolation and purification by conventional cell fractionation methods.     

Compartmentation of asparagine-linked oligosaccharide processing in the Golgi apparatus

Golgi-associated processing of complex-type oligosaccharides linked to asparagine involves the sequential action of at least six enzymes. By equilibrium sucrose density gradient centrifugation of membranes from Chinese hamster ovary cells, we have partially resolved the set of four initial enzymes in the pathway (Mannosidase I, N-acetylglucosamine (GlcNAc) Transferase I, Mannosidase II, and GlcNAc Transferase II) from two later-acting activities (galactosyltransferase and sialyltransferase). In view of the recent demonstration that galactosyltransferase is restricted to the trans face of the Golgi complex in HeLa cells (Roth, J., and E.G. Berger, 1982, J. Cell Biol., 93:223-229), our results suggest that removal of mannose and attachment of peripheral N-acetylglucosamine may occur in some or all of the remaining cisternae on the cis side of the Golgi stack.     

Tyrosine sulfation is not the last modification of entactin before its secretion from 3T3-L1 adipocytes

Tyrosine sulfation of entactin was studied by labeling of 3T3-L1 adipocytes with [35S]methionine or H2 35SO4 in the presence or absence of tunicamycin or monensin. Four precursors (EN1-4) at different steps of modification were detected in addition to mature entactin. Under normal conditions, EN2 and mature entactin were intracellular species, and the latter was sulfated and secreted. Inhibition of co-translational transfer of N-linked oligosaccharides by tunicamycin produced EN1 and EN3 as intracellular species, and EN3 was sulfated and secreted. Interruption of protein transport from medial to trans (distal) Golgi cisternae by monensin, and consequent blockage of terminal glycosylation caused intracellular accumulation of EN4. EN4 was sulfated and of different size compared to mature entactin. These facts suggested that tyrosine sulfation of entactin occurs in medial Golgi cisternae and is not the last modification before its secretion. Our results appeared inconsistent with recent observations by Baeuerle and Huttner [(1987) J. Cell Biol. 105, 2655-2664] that tyrosine sulfation of IgM occurred within the trans (distal) Golgi cisternae as the last modification before its exit from the Golgi complex.     

Rapid redistribution of Golgi proteins into the ER in cells treated with brefeldin A: evidence for membrane cycling from Golgi to ER

In cells treated with brefeldin A (BFA), movement of newly synthesized membrane proteins from the endoplasmic reticulum (ER) to the Golgi apparatus was blocked. Surprisingly, the glycoproteins retained in the ER were rapidly processed by cis/medial Golgi enzymes but not by trans Golgi enzymes. An explanation for these observations was provided from morphological studies at both the light and electron microscopic levels using markers for the cis/medial and trans Golgi. They revealed a rapid and dramatic redistribution to the ER of components of the cis/medial but not the trans Golgi in response to treatment with BFA. Upon removal of BFA, the morphology of the Golgi apparatus was rapidly reestablished and proteins normally transported out of the ER were efficiently and rapidly sorted to their final destinations. These results suggest that BFA disrupts a dynamic membrane-recycling pathway between the ER and cis/medial Golgi, effectively blocking membrane transport out of but not back to the ER.     

Cytochemical localization of terminal N-acetyl-D-galactosamine residues in cellular compartments of intestinal goblet cells: implications for the topology of O-glycosylation

The O-linked oligosaccharides of mucin-type glycoproteins contain N- acetyl-D-galactosamine (GalNAc) that is not found in N-linked glycoproteins. Because Helix pomatia lectin interacts with terminal GalNAc, we used this lectin, bound to particles of colloidal gold, to localize such sugar residues in subcellular compartments of intestinal goblet cells. When thin sections of low temperature Lowicryl K4M embedded duodenum or colon were incubated with Helix pomatia lectin- gold complexes, no labeling could be detected over the cisternal space of the nuclear envelope and the rough endoplasmic reticulum. A uniform labeling was observed over the first and several subsequent cis Golgi cisternae and over the last (duodenal goblet cells) or the two last (colonic goblet cells) trans Golgi cisternae as well as forming and mature mucin droplets. However, essentially no labeling was detected over several cisternae in the central (medial) region of the Golgi apparatus. The results strongly suggest that core O-glycosylation takes place in cis Golgi cisternae but not in the rough endoplasmic reticulum. The heterogenous labeling for GalNAc residues in the Golgi apparatus is taken as evidence that termination of certain O- oligosaccharide chains by GalNAc occurs in trans Golgi cisternae.     

Wheat germ agglutinin fracture-label of Golgi apparatus membranes in proliferating cells

The thin section fracture-label technique has been recently used to analyze the distribution and compartmentalization of fully glycosylated components on intracellular membranes. Labelling with the lectin wheat germ agglutinin over the freeze-fractured membranes of Golgi apparatus in various secretory and non-secretory cells as well as in human peripheral lymphocytes was always very weak or absent even over the trans-most cisternae. In order to investigate if the labelling density may reflect the cellular activity in membrane biogenesis, we used in this study wheat germ agglutinin fracture-label of rapidly proliferating cells and mitogen-activated lymphocytes. Labelling over the fractured cisternae of the medial and trans portions of the Golgi apparatus was intense. Treatment with cycloheximide of proliferating cells induced a drastic reduction of the labelling over the Golgi cisternae.