Membrane traffic: a glitch in the Golgi matrix

Golgins are coiled-coil proteins thought to form a matrix important for shaping and organising Golgi cisternae and directing long-range recognition events in vesicular transport. This model is brought into question by new evidence that two golgins, GM130 and golgin-84, contribute to but are not essential for protein transport and Golgi structure.     

Membrane traffic: Arl GTPases get a GRIP on the Golgi

A subset of the golgin family of large coiled-coil proteins have a GRIP domain that mediates their localization to the trans-Golgi. Two recent papers show that the Arl3p and Arl1p small GTPases act sequentially to recruit GRIP domain proteins to the Golgi.     

Rafting through traffic: Membrane domains in cellular logistics

The intricate and tightly regulated organization of eukaryotic cells into spatially and functionally distinct membrane-bound compartments is a defining feature of complex organisms. These compartments are defined by their lipid and protein compositions, with their limiting membrane as the functional interface to the rest of the cell. Thus, proper segregation of membrane proteins and lipids is necessary for the maintenance of organelle identity, and this segregation must be maintained despite extensive, rapid membrane exchange between compartments. Sorting processes of high efficiency and fidelity are required to avoid potentially deleterious mis-targeting and maintain cellular function. Although much molecular machinery associated with membrane traffic (i.e. membrane budding/fusion/fission) has been characterized both structurally and biochemically, the mechanistic details underlying the tightly regulated distribution of membranes between subcellular locations remain to be elucidated. This review presents evidence for the role of ordered lateral membrane domains known as lipid rafts in both biosynthetic sorting in the late secretory pathway, as well as endocytosis and recycling to/from the plasma membrane. Although such evidence is extensive and the involvement of membrane domains in sorting is definitive, specific mechanistic details for raft-dependent sorting processes remain elusive.     

Membrane glycoproteins involved in hepatocyte adhesion to collagen type I

Liver membrane glycoproteins with affinity for immobilized collagen type I were subjected to preparative sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) followed by electroelution of the separated proteins. Electroeluted glycoproteins with ability to neutralize the inhibitory effect of anti-CollCAM antibodies on hepatocyte adhesion to collagen were collected from several consecutive runs and used to raise a high titer antiserum, denoted anti-CollCAM II. IgG from this antiserum inhibited the attachment of hepatocytes to dishes coated with collagen type I, but not to fibronectin- or collagen type IV-coated dishes. When the antibodies were immobilized to Sepharose CL-4B they bound three sets of glycoproteins with apparent Mr's of 105,000, 115,000, and 130,000 as analyzed by SDS-PAGE under nonreducing (NR) conditions. Upon reduction (R) the glycoproteins migrated with apparent Mr's of 115,000, 130,000, and 160,000, respectively. The Mr 105,000-115,000 (NR) glycoproteins effectively neutralized the inhibitory effect exerted by both anti-CollCAM and anti-CollCAM II antibodies, on hepatocyte spreading and attachment to collagen type I substrates. Peptide mapping suggested the Mr 160,000 (R) species to be different from the Mr 115,000 (R).     

Biosynthesis and modification of Golgi mannosidase II in HeLa and 3T3 cells

The biosynthesis and post-translational modification of mannosidase II, an enzyme required in the maturation of asparagine-linked oligosaccharides in the Golgi complex, has been investigated. Antibody raised against this enzyme purified from rat liver Golgi membranes was used to immunoprecipitate mannosidase II from rat liver, 3T3 cells, or HeLa cells. Mannosidase II immunoprecipitated from rat liver Golgi membranes, when analyzed by polyacrylamide gel electrophoresis, migrated with an apparent molecular weight of approximately 124,000. In contrast, the enzyme purified from rat liver Golgi membranes was shown to contain both the 124,000-dalton component and a 110,000-dalton polypeptide believed to result from degradation of intact mannosidase II during purification. Mannosidase II from 3T3 and HeLa cells migrated on polyacrylamide gels with apparent molecular weights of approximately 124,000 and 134,000-136,000, respectively. When immunoprecipitated from radiolabeled cultures, mannosidase II from both cell types was similar in the following respects: (a) the initial synthesis product had an apparent molecular weight of approximately 124,000; (b) in cultures treated with tunicamycin the initial synthesis product had an apparent molecular weight of approximately 117,000; (c) endoglycosidase H digestion of the initial synthesis product gave an apparent molecular weight similar to the tunicamycin-induced polypeptide; (d) the mature enzyme was mostly (HeLa) or entirely (3T3) resistant to digestion by endoglycosidase H. Loss of [35S]methionine from intracellular mannosidase II occurred with a half-life of approximately 20 h; there was no appreciable accumulation of labeled immuno-reactive material in the medium. HeLa mannosidase II, but not the 3T3 enzyme, was additionally modified 1-3 h after synthesis, the initial synthesis product being converted to a doublet with an apparent molecular weight of approximately 134,000-136,000. Evidence is presented that this mobility shift may result from O-glycosylation. Mannosidase II from both cell types could be labeled with [32P]phosphate or [35S]sulfate. The latter is apparently attached to oligosaccharide as indicated by inhibition of labeling by tunicamycin; the former was shown with the HeLa enzyme to be present as serine phosphate moieties. In addition, [3H]palmitate could be incorporated into the enzyme in 3T3 cells.(ABSTRACT TRUNCATED AT 400 WORDS)     

Membrane targeting: getting Arl to the Golgi

Post-translational modification with myristoyl or prenyl groups is essential for membrane association of many small GTPases in the Ras superfamily. Two recent papers show that, rather than myristoylation, amino-terminal acetylation of the Arf-like protein Arl3p is required for Golgi targeting via an interaction with an integral membrane protein called Sys1.     

Membrane traffic in polarized epithelial cells

Epithelial cells contain apical and basolateral surfaces with distinct compositions. Sorting of certain proteins to the basolateral surface involves the epithelial-specific mu 1b clathrin adaptor subunit. Recent results have shown that targeting to the basolateral surface utilizes the exocyst, whereas traffic to the apical surface uses syntaxin 3. Endocytosis at the apical surface is regulated by ARF6. Transcytosis of IgA is regulated by the p62Yes tyrosine kinase.     

Membrane traffic and the cellular uptake of cholera toxin.

In nature, cholera toxin (CT) and the structurally related E. coli heat labile toxin type I (LTI) must breech the epithelial barrier of the intestine to cause the massive diarrhea seen in cholera. This requires endocytosis of toxin-receptor complexes into the apical endosome, retrograde transport into Golgi cisternae or endoplasmic reticulum (ER), and finally transport of toxin across the cell to its site of action on the basolateral membrane. Targeting into this pathway depends on toxin binding ganglioside GM1 and association with caveolae-like membrane domains. Thus to cause disease, both CT and LTI co-opt the molecular machinery used by the host cell to sort, move, and organize their cellular membranes and substituent components.     

Recycling glycoproteins do not return to the cis-Golgi

Recycling of a number of glycoproteins along the site of action of mannosidase I (the distal endoplasmic reticulum/cis-Golgi) was followed in several different cell lines. Treatment of cells with 1-deoxymannojirimycin (dMM) produced high mannose oligosaccharides at positions otherwise occupied by complex-type oligosaccharides in these glycoproteins. Conversion of high-mannose-type oligosaccharides to complex oligosaccharides of proteins initially synthesized in the presence of dMM was used as a marker for recycling of glycoproteins along the site of action of dMM. In contrast to findings reported by Snider and Rogers (Snider, M. D., and O. C. Rogers. 1986. J. Cell Biol. 103:265-275), removal of dMM did not result in reconversion of high-mannose oligosaccharides to complex-type sugars, even after prolonged periods of culture. We conclude that surface glycoproteins do not recycle through the cis-medial Golgi elements.     

Characterization of Clofibrate-induced Retrograde Golgi Membrane Movement to the Endoplasmic Reticulum: Clofibrate Distinguishes the Golgi from the Trans Golgi Network

Clofibrate-induced retrograde Golgi membrane movement was blocked or retarded when NRK cells were treated with sodium azide/2-deoxyglucose, nocodazole, taxol, and destruxin B, indicating that it depends on energy, and the dynamic state of microtubules, and being acidic or vacuolar-type ATPase function. PDMP and phospholipase A₂ inhibitors also blocked it. These characteristics are similar to those of brefeldin A (BFA) and nordihydroguaiaretic acid (NDGA), inducers of retrograde Golgi membrane movement. However, clofibrate was distinguished from BFA in that BFA action was insensitive to phospholipase A₂ inhibitors and from NDGA in that NDGA stabilized microtubules against nocodazole and its action was almost insensitive to taxol. The trans Golgi network (TGN) was resistant to clofibrate, while BFA and NDGA dispersed it. To our knowledge, clofibrate is the first drug to show such different effects on the Golgi and TGN and, therefore, is expected to be a useful tool to distinguish their architecture and/or membrane dynamics.     

Non-carrier-mediated uptake of the mannosidase I inhibitor 1-deoxymannojirimycin by K562 erythroleukemic cells

A 3H label was introduced at the C-1 position of the mannosidase I inhibitor 1-deoxymannojirimycin (dMM) by catalytic hydrogenolysis of benzyl-2,3-O-isopropylidene-5-N-benzyl-6-O-benzyl-alpha-D-mannofurano side with 3H2. 1-[3H]dMM as well as its precursor 1-[3H]2,3-O-isopropylidene-dMM had identical Rf as the nonradioactive compounds on TLC. Furthermore, alpha 1-antitrypsin secreted by HepG2 cells was modified indistinguishably by treatment of the cells with dMM and 1-[3H]dMM. Thus, 1-[3H]dMM had chemical and biological properties identical with authentic dMM. Uptake of [14C]mannose by K562 cells could be inhibited by glucose but not by the mannose analogue dMM. Thus, dMM does not enter the cell through hexose transporter(s). Uptake of 1-[3H]dMM by K562 cells could not be inhibited by increasing concentrations of nonradioactive dMM (from 1-32,000 microM), showing transport of dMM into cells through nonfacilitated diffusion. Furthermore, uptake of 1-[3H]dMM by K562 cells was observed at 0 degrees C.     

A novel fluorescent ceramide analogue for studying membrane traffic in animal cells: accumulation at the Golgi apparatus results in altered spectral properties of the sphingolipid precursor

A series of ceramide analogues bearing the fluorophore boron dipyrromethene difluoride (BODIPY) were synthesized and evaluated as vital stains for the Golgi apparatus, and as tools for studying lipid traffic between the Golgi apparatus and the plasma membrane of living cells. Studies of the spectral properties of several of the BODIPY-labeled ceramides in lipid vesicles demonstrated that the fluorescence emission maxima were strongly dependent upon the molar density of the probes in the membrane. This was especially evident using N-[5-(5,7-dimethyl BODIPY)-1-pentanoyl]-D-erythro-sphingosine (C5-DMB-Cer), which exhibited a shift in its emission maximum from green (integral of 515 nm) to red (integral of 620 nm) wavelengths with increasing concentrations. When C5-DMB-Cer was used to label living cells, this property allowed us to differentiate membranes containing high concentrations of the fluorescent lipid and its metabolites (the corresponding analogues of sphingomyelin and glucosylceramide) from other regions of the cell where smaller amounts of the probe were present. Using this approach, prominent red fluorescent labeling of the Golgi apparatus, Golgi apparatus-associated tubulovesicular processes, and putative Golgi apparatus transport vesicles was seen in living human skin fibroblasts, as well as in other cell types. Based on fluorescence ratio imaging microscopy, we estimate that C5-DMB-Cer and its metabolites were present in Golgi apparatus membranes at concentrations up to 5-10 mol %. In addition, the concentration-dependent spectral properties of C5-DMB-Cer were used to monitor the transport of C5-DMB-lipids to the cell surface at 37 degrees C.     

Intracellular movement of two mannose 6-phosphate receptors: return to the Golgi apparatus

We have used Chinese hamster ovary (CHO) cells and a murine lymphoma cell line to study the recycling of the 215-kD and the 46-kD mannose 6-phosphate receptors to various regions of the Golgi to determine the site where the receptors first encounter newly synthesized lysosomal enzymes. For assessing return to the trans-most Golgi compartments containing sialyltransferase (trans-cisternae and trans-Golgi network), the oligosaccharides of receptor molecules on the cell surface were labeled with [3H]galactose at 4 degrees C. Upon warming to 37 degrees C, the [3H]galactose residues on both receptors were substituted with sialic acid with a t1/2 approximately 3 hrs. Other glycoproteins acquired sialic acid at least 8-10 times slower. Return of the receptors to the trans-Golgi cisternae containing galactosyltransferase could not be detected. Return to the cis/middle Golgi cisternae containing alpha-mannosidase I was measured by adding deoxymannojirimycin, a mannosidase I inhibitor, during the initial posttranslational passage of [3H]mannose-labeled glycoproteins through the Golgi, thereby preserving oligosaccharides which would be substrates for alpha-mannosidase I. After removal of the inhibitor, return to the early Golgi with subsequent passage through the Golgi complex was measured by determining the conversion of the oligosaccharides from high mannose to complex-type units. This conversion was very slow for the receptors and other glycoproteins (t1/2 approximately 20 h). Exposure of the receptors and other glycoproteins to the dMM-sensitive alpha-mannosidase without movement through the Golgi apparatus was determined by measuring the loss of mannose residues from these proteins. This loss was also slow. These results indicate that both Man-6-P receptors routinely return to the Golgi compartment which contains sialyltransferase and recycle through other regions of the Golgi region less frequently. We infer that the trans-Golgi network is the major site for lysosomal enzyme sorting in CHO and murine lymphoma cells.     

Endoplasmic reticulum (ER)-associated degradation of misfolded N-linked glycoproteins is suppressed upon inhibition of ER mannosidase I

To examine the role of early carbohydrate recognition/trimming reactions in targeting endoplasmic reticulum (ER)-retained, misfolded glycoproteins for ER-associated degradation (ERAD), we have stably expressed the cog thyroglobulin (Tg) mutant cDNA in Chinese hamster ovary cells. We found that inhibitors of ER mannosidase I (but not other glycosidases) acutely suppressed Cog Tg degradation and also perturbed the ERAD process for Tg reduced with dithiothreitol as well as for gamma-carboxylation-deficient protein C expressed in warfarin-treated baby hamster kidney cells. Kifunensine inhibition of ER mannosidase I also suppressed ERAD in castanospermine-treated cells; thus, suppression of ERAD does not require lectin-like binding of ER chaperones calnexin and calreticulin to monoglucosylated oligosaccharides. Notably, the undegraded protein fraction remained completely microsome-associated. In pulse-chase studies, kifunensine-sensitive degradation was still inhibitable even 1 h after Tg synthesis. Intriguingly, chronic treatment with kifunensine caused a 3-fold accumulation of Cog Tg in Chinese hamster ovary cells and did not lead to significant induction of the ER unfolded protein response. We hypothesize that, in a manner not requiring lectin-like activity of calnexin/calreticulin, the recognition or processing of a specific branched N-linked mannose structure enhances the efficiency of glycoprotein retrotranslocation from the ER lumen.     

Membrane nanotubes: dynamic long-distance connections between animal cells

Membrane nanotubes are transient long-distance connections between cells that can facilitate intercellular communication (for example, by trafficking vesicles or transmitting calcium-mediated signals), but they can also contribute to pathologies (for example, by directing the spread of viruses). Recent data have revealed considerable heterogeneity in their structures, processes of formation and functional properties, in part dependent on the cell types involved. Despite recent progress in this young research field, further research is sorely needed.     

Membrane hydration correlates to cellular biophysics during freezing in mammalian cells

Cell survival during freezing applications in biomedicine is highly correlated to the temperature history and its dependent cellular biophysical events of dehydration and intracellular ice formation (IIF). Although cell membranes are known to play a significant role in cell injury, a clear correlation between the membrane state and the surrounding intracellular and extracellular water is still lacking. We previously showed that lipid hydration in LNCaP tumor cells is related to cellular dehydration. The goal of this study is to build upon this work by correlating both the phase state of the membrane and the surrounding water to cellular biophysical events in three different mammalian cell types: human prostate tumor cells (LNCaP), human dermal fibroblasts (HDF), and porcine smooth muscle cells (SMC) using Fourier Transform Infrared spectroscopy (FTIR). Variable cooling rates were achieved by controlling the degree of supercooling prior to ice nucleation (− 3 °C and − 10 °C) while the sample was cooled at a set rate of 2 °C/min. Membranes displayed a highly cooperative phase transition under dehydrating conditions (i.e. NT = − 3 °C), which was not observed under IIF conditions (NT = − 10 °C). Spectral analysis showed a consistently greater amount of ice formation during dehydrating vs. IIF conditions in all cell types. This is hypothesized to be due to the extreme loss of membrane hydration in dehydrating cells that is manifested as excess water available for phase change. Interestingly, changes in residual membrane conformational disorder correlate strongly with cellular volumetric decreases as assessed by cryomicroscopy. A strong correlation was also found between the activation energies for freezing induced lyotropic membrane phase change determined using FTIR and the water transport measured by cryomicroscopy. Reduced lipid hydration under dehydration freezing conditions is suggested as one of the likely causes of what has been termed as “solution effects” injury in cryobiology.     

Membrane traffic in Dictyostelium discoideum: plasma membrane glycoconjugates internalized and recycled during fluid phase pinocytosis enter the Golgi complex

Plasma membrane glycoconjugates, enzymatically labelled with [3H]galactose, were used as an autoradiographic membrane marker for a morphometric analysis of membrane traffic during fluid phase pinocytosis in the amoeba, Dictyostelium discoideum. The fraction of grains associated with the plasma membrane decreased exponentially from 99% for cells fixed directly after labelling on the cell surface, to a steady-state value of 45% after about 1 h of pinocytotic activity. The complementary fraction of grains was observed on vacuolar membranes. Only after a lag of about 20 min, a small but significant fraction (3%) of the total grains, was found in the region of the Golgi membranes. During two subsequent doublings of cell number, over a period of 24 h, label was lost into the medium at a constant rate of 1% per h. The cell bound label remained fully membrane bound over the entire period. The beta(1-4) linkage was not noticeably modified, as judged by its susceptibility to hydrolytic release by beta-galactosidase. An analysis by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) showed an identical labelling pattern for total membrane fractions when prepared directly after labelling or after 24 h of membrane flow.