Actin microfilaments facilitate the retrograde transport from the Golgi complex to the endoplasmic reticulum in mammalian cells.

The morphology and subcellular positioning of the Golgi complex depend on both microtubule and actin cytoskeletons. In contrast to microtubules, the role of actin cytoskeleton in the secretory pathway in mammalian cells has not been clearly established. Using cytochalasin D, we have previously shown that microfilaments are not involved in the endoplasmic reticulum-Golgi membrane dynamics. However, it has been reported that, unlike botulinum C2 toxin and latrunculins, cytochalasin D does not produce net depolymerization of actin filaments. Therefore, we have reassessed the functional role of actin microfilaments in the early steps of the biosynthetic pathway using C2 toxin and latrunculin B. The anterograde endoplasmic reticulum-to-Golgi transport monitored with the vesicular stomatitis virus-G protein remained unaltered in cells treated with cytochalasin D, latrunculin B or C2 toxin. Conversely, the brefeldin A-induced Golgi membrane fusion into the endoplasmic reticulum, the Golgi-to-endoplasmic reticulum transport of a Shiga toxin mutant form, and the subcellular distribution of the KDEL receptor were all impaired when actin microfilaments were depolymerized by latrunculin B or C2 toxin. These findings, together with the fact that COPI-coated and uncoated vesicles contain beta/gamma-actin isoforms, indicate that actin microfilaments are involved in the endoplasmic reticulum/Golgi interface, facilitating the retrograde Golgi-to-endoplasmic reticulum membrane transport, which could be mediated by the orchestrated movement of transport intermediates along microtubule and microfilament tracks.

Endocytosis of NBD-sphingolipids in neurons: exclusion from degradative compartments and transport to the Golgi complex.

Sphingolipids are abundant constituents of neuronal membranes that have been implicated in intracellular signaling, neurite outgrowth and differentiation. Differential localization and trafficking of lipids to membrane domains contribute to the specialized functions. In non-neuronal cultured cell lines, plasma membrane short-chain sphingomyelin and glucosylceramide are recycled via endosomes or sorted to degradative compartments. However, depending on cell type and lipid membrane composition, short-chain glucosylceramide can also be diverted to the Golgi complex. Here, we show that NBD-labeled glucosylceramide and sphingomyelin are transported from the plasma membrane to the Golgi complex in cultured rat hippocampal neurons irrespective of the stage of neuronal differentiation. Golgi complex localization was confirmed by colocalization and Golgi disruption studies, and importantly did not result from conversion of NBD-glucosylceramide or NBD-sphingomyelin to NBD-ceramide. Double-labeling experiments with transferrin or wheat-germ agglutinin showed that NBD-sphingolipids are first internalized to early/recycling endosomes, and subsequently transported to the Golgi complex. The internalization of these two sphingolipid analogs was energy and temperature dependent, and their intracellular transport was insensitive to the NBD fluorescence quencher sodium dithionite. These results indicate that vesicles mediate the transport of internalized NBD-glucosylceramide and NBD-sphingomyelin to the Golgi complex.

The golgi-associated COPI-coated buds and vesicles contain beta/gamma -actin.

It has been shown previously that the morphology and subcellular positioning of the Golgi complex is controlled by actin microfilaments. To further characterize the association between actin microfilaments and the Golgi complex, we have used the Clostridium botulinum toxins C2 and C3, which specifically inhibit actin polymerization and cause depolymerization of F-actin in intact cells by the ADP ribosylation of G-actin monomers and the Rho small GTP-binding protein, respectively. Normal rat kidney cells treated with C2 showed that disruption of the actin and the collapse of the Golgi complex occurred concomitantly. However, when cells were treated with C3, the actin disassembly was observed without any change in the organization of the Golgi complex. The absence of the involvement of Rho was further confirmed by the treatment with lysophosphatidic acid or microinjection with the constitutively activated form of RhoA, both of which induced the stress fiber formation without affecting the Golgi complex. Immunogold electron microscopy in normal rat kidney cells revealed that beta- and gamma-actin isoforms were found in Golgi-associated COPI-coated buds and vesicles. Taken together, the results suggest that the Rho signaling pathway does not directly regulate Golgi-associated actin microfilaments, and that beta- and gamma-actins might be involved in the formation and/or transport of Golgi-derived vesicular or tubular intermediates.

Morphological changes in the Golgi complex correlate with actin cytoskeleton rearrangements.

In this report we have studied the morphological changes of the Golgi complex (GC) that specifically accompany F-actin reorganizations. In starved rat RBL-2H3 tumor mast cells, the GC, that was visualized at immunofluorescence level with antibodies raised against the Golgi-resident proteins giantin, mannosidase II, or TGN-38, showed a compacted morphology with a supranuclear positioning. Concomitant to membrane ruffle formation induced by epidermal growth factor (EGF) or phorbol 12-myristate 13-acetate (PMA), and stress fiber formation induced by lysophosphatidic acid (LPA), specific GC morphological changes were observed. When cells were stimulated with EGF or PMA, the compacted GC morphology was transformed into a reticular network that was extended towards the cell periphery. When cells were incubated with LPA, the GC acquired a characteristic ring-shaped morphology. Brefeldin A (BFA) did not affect the PMA- or LPA-induced membrane ruffling and stress fiber formation, respectively, indicating that actin rearrangements occurred independent of the presence of the GC. Upon BFA removal, the presence of PMA or LPA during the recovery process induced the GC to acquire the morphological appearance described above for each agent. Moreover, the PMA- but not the LPA-induced GC rearrangements were sensitive to the actin perturbing agents cytochalasin D and jasplakinolide. When cells were preincubated with the phosphatidylinositide 3-kinase (PI3K) inhibitors wortmannin or LY294002, the PMA-induced GC morphological changes were inhibited but not membrane ruffles. Finally, the PMA-induced increase in the post-Golgi transport of glycosaminoglycans to the cell surface was not altered by cytochalasin D or jasplakinolide. Altogether, these data suggest that: (1) the shape of the GC is influenced by the 3D arrangement of actin microfilaments; (2) PI3K regulates the association of the GC with actin microfilaments; and (3) actin microfilaments are not essential for the post-Golgi transport to the plasma membrane.

Morphological and biochemical analysis of the secretory pathway in melanoma cells with distinct metastatic potential.

In this report, we have investigated whether alterations of the morphological and functional aspects of the biosecretory membrane system are associated with the metastatic potential of tumor cells. To this end, we have analyzed the morphology of the Golgi complex, the cytoskeleton organization and membrane trafficking steps of the secretory pathway in two human melanoma A375 cell line variants with low (A375-P) and high metastatic (A375-MM) potential. Immunofluorescence analysis showed that in A375-P cells, the Golgi complex showed a collapsed morphology. Conversely, in A375-MM cells, the Golgi complex presented a reticular and extended morphology. At the ultrastructural level, the Golgi complex of A375-P cells was fragmented and cisternae were swollen. When the cytoskeleton was analyzed, the microtubular network appeared normal in both cell variants, whereas actin stress fibers were largely absent in A375-P, but not in A375-MM cells. In addition, the F-actin content in A375-P cells was significantly lower than in A375-MM cells. These morphological differences in A375-P cells were accompanied by acceleration and an increase in the endoplasmic reticulum to Golgi and the trans-Golgi network to cell surface membrane transport, respectively. Our results indicate that in human A375 melanoma cells, metastatic potential correlates with a well-structured morphofunctional organization of the Golgi complex and actin cytoskeleton.

N-Ras induces alterations in Golgi complex architecture and in constitutive protein transport.

Aberrant glycosylation of proteins and lipids is a common feature of many tumor cell types, and is often accompanied by alterations in membrane traffic and an anomalous localization of Golgi-resident proteins and glycans. These observations suggest that the Golgi complex is a key organelle for at least some of the functional changes associated with malignant transformation. To gain insight into this possibility, we have analyzed changes in the structure and function of the Golgi complex induced by the conditional expression of the transforming N-Ras(K61) mutant in the NRK cell line. A remarkable and specific effect associated with this N-Ras-induced transformation was a conspicuous rearrangement of the Golgi complex into a collapsed morphology. Ultrastructural and stereological analyses demonstrated that the Golgi complex was extensively fragmented. The collapse of the Golgi complex was also accompanied by a disruption of the actin cytoskeleton. Functionally, N-Ras-transformed KT8 cells showed an increase in the constitutive protein transport from the trans-Golgi network to the cell surface, and did not induce the appearance of aberrant cell surface glycans. The Golgi complex collapse, the actin disassembly, and the increased constitutive secretion were all partially inhibited by the phospholipase A2 inhibitor 4-bromophenylacyl bromide. The results thus suggest the involvement of the actin cytoskeleton in the shape of the Golgi complex, and intracellular phospholipase A2 in its architecture and secretory function.

Modulation of carcinoembryonic antigen release by glucosylceramide--implications for HT29 cell differentiation.

Previous work suggested that glucosylceramide (GlcCer) plays a role in the regulation of cell differentiation of HT29 human colon tumor cells. In the present study, we investigated the role of GlcCer in the cellular release of carcinoembryonic antigen (CEA), a marker for cell differentiation. This was done by modulating the intracellular level of the glycolipid, according to two different approaches. The cells were treated with D,L-threo-1-phenyl-2-decanoylamino-3-morpholino-1-propanol (PDMP), which resulted in a specific lowering of the cellular GlcCer pool. Alternatively, by exogenous addition of a short-chain analog of the lipid, hexanoyl(C6)-GlcCer, the cellular pool was enhanced. The results demonstrate that PDMP causes an increase in the release of CEA, while exogenous C6-GlcCer suppresses its release. Furthermore, the enhanced release of CEA in the presence of PDMP, could be completely reversed upon exogenous addition of C6-GlcCer. Control experiments reveal that a potential interference of the well-known modulator of cell physiology, ceramide (Cer), can be excluded. Long-term depletion of GlcCer resulted in a change in a morphological feature of differentiation of the cells, i.e. an increase in apical membrane surface with microvilli brush borders, accompanied by an enhanced expression of the cytoskeletal protein villin. These results, together with the observations on modulation of the differentiation marker CEA by GlcCer, provide support for the conclusion that GlcCer interferes with the differentiation of HT29 cells.

Actin microfilaments are essential for the cytological positioning and morphology of the Golgi complex.

The organization and function of the Golgi complex was studied in normal rat kidney cells following disruption of the actin cytoskeleton induced by cytochalasin D. In cells treated with these reagents, the reticular and perinuclear Golgi morphology acquired a cluster shape restricted to the centrosome region. Golgi complex alteration affected all Golgi subcompartments as revealed by double fluorescence staining with antibodies to the cis/middle Mannosidase II and the trans-Golgi network TGN38 proteins or vital staining with the lipid derivate C6-NBD-ceramide. The ultrastructural and stereological analysis showed that the Golgi cisternae remained attached in a stacked conformation, but they were swollen and contained electron-dense intra-cisternal bodies. The Golgi complex cluster remained linked to microtubules since it was fragmented and dispersed after treatment with nocodazole. Moreover, the reassembly of Golgi fragments after the disruption of the microtubuli with nocodazole does not utilize the actin microfilaments. The actin microfilament requirement for the disassembly and reassembly of the Golgi complex and for the ER-Golgi vesicular transport were also studied. The results show that actin microfilaments are not needed for either the retrograde fusion of the Golgi complex with the endoplasmic reticulum promoted by brefeldin A or the anterograde reassembly after the removal of the drug, or the ER-Golgi transport of VSV-G glycoprotein. However, actin microfilaments are directly involved in the subcellular localization and the morphology of the Golgi complex.

PDMP blocks brefeldin A-induced retrograde membrane transport from golgi to ER: evidence for involvement of calcium homeostasis and dissociation from sphingolipid metabolism.

In this study, we show that an inhibitor of sphingolipid biosynthesis, D,L-threo-1-phenyl-2- decanoylamino-3-morpholino-1-propanol (PDMP), inhibits brefeldin A (BFA)-induced retrograde membrane transport from Golgi to endoplasmic reticulum (ER). If BFA treatment was combined with or preceded by PDMP administration to cells, disappearance of discrete Golgi structures did not occur. However, when BFA was allowed to exert its effect before PDMP addition, PDMP could not "rescue" the Golgi compartment. Evidence is presented showing that this action of PDMP is indirect, which means that the direct target is not sphingolipid metabolism at the Golgi apparatus. A fluorescent analogue of PDMP, 6-(N-[7-nitro-2,1, 3-benzoxadiazol-4-yl]amino)hexanoyl-PDMP (C6-NBD-PDMP), did not localize in the Golgi apparatus. Moreover, the effect of PDMP on membrane flow did not correlate with impaired C6-NBD-sphingomyelin biosynthesis and was not mimicked by exogenous C6-ceramide addition or counteracted by exogenous C6-glucosylceramide addition. On the other hand, the PDMP effect was mimicked by the multidrug resistance protein inhibitor MK571. The effect of PDMP on membrane transport correlated with modulation of calcium homeostasis, which occurred in a similar concentration range. PDMP released calcium from at least two independent calcium stores and blocked calcium influx induced by either extracellular ATP or thapsigargin. Thus, the biological effects of PDMP revealed a relation between three important physiological processes of multidrug resistance, calcium homeostasis, and membrane flow in the ER/ Golgi system.

Ceramide transport from endoplasmic reticulum to Golgi apparatus is not vesicle-mediated.

Ceramide (Cer) transfer from the endoplasmic reticulum (ER) to the Golgi apparatus was measured under conditions that block vesicle-mediated protein transfer. This was done either in intact cells by reducing the incubation temperature to 15 degreesC, or in streptolysin O-permeabilized cells by manipulating the intracellular environment. In both cases, Cer transfer was not inhibited, as demonstrated by the biosynthesis of ceramide monohexosides and sphingomyelin (SM) de novo from metabolically (with [14C]serine) labelled Cer. This assay is based on the knowledge that Cer is synthesized, starting from serine and palmitoyl-CoA, at the ER, whereas glycosphingolipids and SM are synthesized in the (early) Golgi apparatus. Formation of [14C]glycosphingolipids and [14C]SM was observed under conditions that block vesicle-mediated vesicular stomatitis virus glycoprotein transport. These results indicate that [14C]Cer is transferred from ER to Golgi by a non-vesicular mechanism.

Fluorescent, short-chain C6-NBD-sphingomyelin, but not C6-NBD-glucosylceramide, is subject to extensive degradation in the plasma membrane: implications for signal transduction related to cell differentiation.

The involvement of the plasma membrane in the metabolism of the sphingolipids sphingomyelin (SM) and glucosylceramide (GlcCer) was studied, employing fluorescent short-chain analogues of these lipids, 6-[N-(7-nitro-2,1,3-benzoxadiazol-4-yl) amino]hexanoylsphingosylphosphorylcholine (C6-NBD-SM), C6-NBD-GlcCer and their common biosynthetic precursor C6-NBD-ceramide (C6-NBD-Cer). Although these fluorescent short-chain analogues are metabolically active, some caution is to be taken in view of potential changes in biophysical/biochemical properties of the lipid compared with its natural counterpart. However, these short-chain analogues offer the advantage of studying the lipid metabolic enzymes in their natural environment, since detergent solubilization is not necessary for measuring their activity. These studies were carried out with several cell types, including two phenotypes (differing in state of differentiation) of HT29 cells. Degradation and biosynthesis of C6-NBD-SM and C6-NBD-GlcCer were determined in intact cells, in their isolated plasma membranes, and in plasma membranes isolated from rat liver tissue. C6-NBD-SM was found to be subject to extensive degradation in the plasma membrane, due to neutral sphingomyelinase (N-SMase) activity. The extent of C6-NBD-SM hydrolysis showed a general cell-type dependence and turned out to be dependent on the state of cell differentiation, as revealed for HT29 cells. In undifferentiated HT29 cells N-SMase activity was at least threefold higher than in its differentiated counterpart. In contrast, in all cell types studied, very little if any biosynthesis of C6-NBD-SM from the precursor C6-NBD-Cer occurred. Moreover, in the case of C6-NBD-GlcCer, neither hydrolytic nor synthetic activity was found to be associated with the plasma membrane. These results are discussed in the context of the involvement of the sphingolipids SM and GlcCer in signal transduction pathways in the plasma membrane.

Transport of biosynthetic sphingolipids from Golgi to plasma membrane in HT29 cells: involvement of different carrier vesicle populations.

Intracellular transport of the sphingolipids glucosylceramide (GlcCer) and sphingomyelin (SM), was examined in HT29 human colon adenocarcinoma cells. After synthesis from a fluorescent precursor, 6-[N-(7-nitro-2,1,3-benzoxadiazol-4-yl)amino]hexanoylceramide++ + (C6-NBD-Cer), transfer of SM from the Golgi complex to the plasma membrane can occur independently of that of GlcCer, as revealed by temperature-dependent experiments. Thus, at 20 degrees C, SM trafficking to the cell surface is essentially unaffected, whereas GlcCer transport to the plasma membrane is inhibited by approximately 75%, when compared to the transfer of both lipids at 37 degrees C. The mechanism by which SM and GlcCer are transported to the cell surface involves at least in part a vesicular mechanism. Transport vesicles, containing both lipids at their luminal surface, as revealed by the inaccessibility of the NBD fluorescence to the quencher sodium dithionite, have been isolated from cells, permeabilized by filter stripping. As evidenced by electron microscopic and biochemical criteria, no vesicles or lipids were released when cell permeabilization had been carried out with streptolysin. Density gradient analysis indicates the potential existence of several vesicle populations, distinctly enriched in either lipid, involved in transport of sphingolipids to the plasma membrane in HT29 cells.

Differential metabolism and trafficking of sphingolipids in differentiated versus undifferentiated HT29 cells.

Trafficking and metabolism of sphingolipids were examined in undifferentiated (G+) and differentiated (G+ reversed) HT29 human colon adenocarcinoma cell lines. Metabolic experiments employing a fluorescently labeled sphingolipid precursor, 6-[N-(7-nitro-2,1,3-benzoxadiazol-4-yl)amino]hexanoylceramide++ + (C6-NBD-ceramide) revealed that both qualitative and quantitative differences exist in sphingolipid synthesis between the 2 cell lines. One of the C6-NBD-sphingolipids synthesized in G+ cells is not found in the G+ reversed cells. Furthermore, the ratio of the 2 main products, C6-NBD-glucosylceramide and C6-NBD-sphingomyelin, differs: in G+ cells glucosylceramide is by far the main product, whereas G+ reversed cells synthesize C6-NBD-sphingomyelin in slight excess. Once established, these ratios of sphingolipids are quickly restored metabolically when distortion of the ratio is caused by experimental manipulation. This indicates that they represent a true metabolic equilibrium situation of the 2 sphingolipids in these cells, while the distinct ratios are mainly determined by the NBD-lipid pool in the plasma membrane. Preferential synthesis and transfer of glucosylceramide from its site of synthesis to the cell surface do not occur when the plasma membrane pool of glucosylceramide is selectively removed. This suggests that instantaneous replenishment via specific signalling is probably not involved as a mechanism in re-establishing perturbed lipid pools. In conjunction with observations on distinct lipid trafficking pathways of glucosylceramide in G+ and G+ reversed cells, the present metabolic studies emphasize a relation between the expression of this glycolipid and the state of differentiation of HT29 cells.

Sorting of sphingolipids in the endocytic pathway of HT29 cells.

The intracellular flow and fate of two fluorescently labeled sphingolipids, 6-[N-(7-nitro-2,1,3-benzoxadiazol-4-yl) amino]hexanoyl glucosyl sphingosine (C6-NBD-glucosylceramide) and C6-NBD-sphingomyelin, was examined in the human colon adenocarcinoma cell line HT29. After their insertion into the plasma membrane at low temperature and subsequent warming of the cells to 37 degrees C, both sphingolipid analogues were internalized by endocytosis, but their intracellular site of destination differed. After 30 min of internalization, C6-NBD-glucosylceramide was localized in the Golgi apparatus, as demonstrated by colocalization with fluorescently labeled ceramide, a Golgi complex marker, and by showing that monensin-induced disruption of the Golgi structure was paralleled by a similar perturbation of the fluorescence distribution. By contrast, C6-NBD-sphingomyelin does not colocalize with the tagged ceramide. Rather, a colocalization with ricin, which is internalized by endocytosis and predominantly reaches the lysosomes, was observed, indicating that the site of delivery of this lipid is restricted to endosomal/lysosomal compartments. Also, in monensin-treated cells no change in the distribution of fluorescence was observed. Thus, these results demonstrate that (sphingo)lipid sorting can occur in the endocytic pathway. Interestingly, the observed sorting phenomenon was specific for glucosylceramide, when compared to other glycolipids, while only undifferentiated HT29 cells displayed the different routing of the two lipids. In differentiated HT29 cells the internalization pathway of sphingomyelin and glucosylceramide was indistinguishable from that of transferrin.

Insulin controls key steps of carbohydrate metabolism in cultured HT29 colon cancer cells.

Effects of insulin on key steps of carbohydrate metabolism were investigated in cultured HT29 colon cancer cells by two different approaches, i.e. incubation of the cells either in the absence or in the presence of glucose in the medium. In glucose-deprived cells, insulin decreased glycogen breakdown, but did not affect polysaccharide levels when glucose was present. Glycogen synthase became activated after insulin treatment in both conditions, even though the activation was more evident when glucose was omitted. No effect on glycogen phosphorylase activity was evident under our experimental conditions. In cells incubated with glucose, the hormone stimulated in a dose-dependent manner the rates of glucose uptake and lactate release. Concomitantly with the increase in glycolytic rate, insulin caused a strong increase in fructose 2,6-bisphosphate. This effect was not observed in the absence of glucose. It is concluded that the carbohydrate metabolism of cultured HT29 cells responds to insulin, making this biological model suitable for investigations in vitro on the mechanism of insulin action.