In vivo incorporation of radioactive metabolites by Golgi apparatus and other cell fractions of onion stem

Incorporation in vivo of various ¹⁴C-labeled substrates into dictyosomes of onion (Allium cepa) stem was determined, and comparisons were made with other cell fractions on a nitrogen basis. Tissue explants were incubated for varying times in the presence of the radioactive metabolites supplied in the external medium. Fractions were then obtained from homogenates stabilized with glutaraldehyde. Purified fractions containing dictyosomes (individual stacks of cisternae) of the Golgi apparatus were obtained by centrifugation in a sucrose gradient also yielding a smooth membrane fraction free of dictyosomes. Dictyosomes were preferentially labeled with choline-1,2-¹⁴C and acetate-2-¹⁴C, suggesting that plant Golgi apparatus participate in the synthesis or modification of membrane lipids. Dictyosomes were also labeled with glucose-U-¹⁴C and leucine-U-¹⁴C, but on a molar basis incorporation was less than with choline or acetate.     

The Golgi apparatus: defining the identity of Golgi membranes

The Golgi apparatus is a stack of compartments that serves as a central junction for membrane traffic, with carriers moving through the stack as well as arriving from, and departing toward, many other destinations in the cell. This requires that the different compartments in the Golgi recruit from the cytosol a distinct set of proteins to mediate accurate membrane traffic. This recruitment appears to reflect recognition of small GTPases of the Rab and Arf family, or of lipid species such as PtdIns(4)P and diacylglycerol, which provide a unique "identity" for each compartment. Recent work is starting to reveal the mechanisms by which these labile landmarks are generated in a spatially restricted manner on specific parts of the Golgi.     

Putative Glycosyltransferases and Other Plant Golgi Apparatus Proteins Are Revealed by LOPIT Proteomics

The Golgi apparatus is the central organelle in the secretory pathway and plays key roles in glycosylation, protein sorting, and secretion in plants. Enzymes involved in the biosynthesis of complex polysaccharides, glycoproteins, and glycolipids are located in this organelle, but the majority of them remain uncharacterized. Here, we studied the Arabidopsis (Arabidopsis thaliana) membrane proteome with a focus on the Golgi apparatus using localization of organelle proteins by isotope tagging. By applying multivariate data analysis to a combined data set of two new and two previously published localization of organelle proteins by isotope tagging experiments, we identified the subcellular localization of 1,110 proteins with high confidence. These include 197 Golgi apparatus proteins, 79 of which have not been localized previously by a high-confidence method, as well as the localization of 304 endoplasmic reticulum and 208 plasma membrane proteins. Comparison of the hydrophobic domains of the localized proteins showed that the single-span transmembrane domains have unique properties in each organelle. Many of the novel Golgi-localized proteins belong to uncharacterized protein families. Structure-based homology analysis identified 12 putative Golgi glycosyltransferase (GT) families that have no functionally characterized members and, therefore, are not yet assigned to a Carbohydrate-Active Enzymes database GT family. The substantial numbers of these putative GTs lead us to estimate that the true number of plant Golgi GTs might be one-third above those currently annotated. Other newly identified proteins are likely to be involved in the transport and interconversion of nucleotide sugar substrates as well as polysaccharide and protein modification.The Golgi apparatus is the central organelle in the secretory pathway, and in higher plants it is involved in the biosynthesis and transport of cell wall matrix polysaccharides, glycoproteins, proteoglycans, and glycolipids as well as in protein trafficking to different subcellular compartments. The last decade has produced substantial findings on the function of the Golgi apparatus: insights into the protein trafficking at the endoplasmic reticulum (ER)/Golgi interface, Golgi structural maintenance, its involvement in endocytosis, and its behavior during cell division (for review, see Faso et al., 2009). However, despite its importance, only a small proportion of the Golgi proteome has been studied: relatively few Golgi proteins have been localized, and even fewer have been functionally characterized.The Golgi apparatus is thought to contain a large and diverse group of membrane-bound glycosyltransferases (GTs). The current view is that different GT activities are required for synthesis of the linkage between different donor and acceptor sugars. Having in mind the diversity of linkage types found in cell wall polysaccharides, the number of different GTs involved is likely to be very large. For instance, it has been estimated that for the biosynthesis of pectin alone, the action of 65 different enzymatic activities is needed (Caffall and Mohnen, 2009). By the end of the year 2011, 468 Arabidopsis (Arabidopsis thaliana) sequences had been annotated in the Carbohydrate-Active EnZymes (CAZy) GT database (Cantarel et al., 2009; http://www.cazy.org). We estimate that two-thirds of these CAZy-classified GTs may be targeted to the Golgi. The remaining one-third are cytosolic or plastidic enzymes involved in processes including, secondary metabolism or starch synthesis. The reported sequences are classified into 43 CAZy families based on amino acid sequence similarities within which at least one member has been biochemically characterized. Each family is likely to have a common structural fold, and three-dimensional (3-D) structures have been resolved for 20 of these 43 families. These are divided mostly into two structural classes, having either a GT-A fold or a GT-B fold (Unligil and Rini, 2000; Bourne and Henrissat, 2001). Moreover, most of the structurally uncharacterized GT families are predicted to adopt either the GT-A or GT-B fold based on 3-D structural homology modeling (Coutinho et al., 2003; Lairson et al., 2008). Despite this conserved 3-D structure, different GT families have very low or undetectable sequence similarities. Consequently, predicting novel GTs based solely on their amino acid sequence similarities is not always achievable, and structural homology searches have also proven useful (Hansen et al., 2009).The length and properties of the transmembrane domain (TMD) of endomembrane proteins appear to play a role in protein sorting and location within the secretory pathway and can be used to predict protein localization (Hanton et al., 2005; Sharpe et al., 2010). In order to perform such predictions, a high number of experimentally localized proteins is required, but only limited data sets have been available for plants to date.In order to identify the most abundant CAZy-classified GTs as well as novel putative GTs, in this work we rigorously extended our proteomic studies of the Golgi apparatus. We have previously developed a high-throughput mass spectrometry (MS)-based quantitative proteomics technique for localization of organelle proteins by isotope tagging (LOPIT; Dunkley et al., 2004, 2006). Here, we report new LOPIT data sets and apply a new method of combining them with published LOPIT data sets, localizing an unprecedented number of plant organelle proteins. We have analyzed the TMD properties of the proteins assigned to the ER, Golgi, and plasma membrane (PM) and determined the organelle-specific features. Structural prediction analysis of the Golgi-localized proteins with unknown functions assessed the protein sequences for the potential to fold similarly to known GT structures. We found that the Golgi contains a substantial number of candidate GT families that have no characterized functions. These results yield a broader understanding of the Golgi function and its biochemical properties.     

Golgi Phosphoprotein 3 Triggers Signal-mediated Incorporation of Glycosyltransferases into Coatomer-coated (COPI) Vesicles

Newly synthesized membrane and secreted proteins undergo a series of posttranslational modifications in the Golgi apparatus, including attachment of carbohydrate moieties. The final structure of so-formed glycans is determined by the order of execution of the different glycosylation steps, which seems intimately related to the spatial distribution of glycosyltransferases and glycosyl hydrolases within the Golgi apparatus. How cells achieve an accurate localization of these enzymes is not completely understood but might involve dynamic processes such as coatomer-coated (COPI) vesicle-mediated trafficking. In yeast, this transport is likely to be regulated by vacuolar protein sorting 74 (Vps74p), a peripheral Golgi protein able to interact with COPI coat as well as with a binding motif present in the cytosolic tails of some mannosyltransferases. Recently, Golgi phosphoprotein 3 (GOLPH3), the mammalian homolog of Vps74, has been shown to control the Golgi localization of core 2 N-acetylglucosamine-transferase 1. Here, we highlight a role of GOLPH3 in the spatial localization of α-2,6-sialyltransferase 1. We show, for the first time, that GOLPH3 supports incorporation of both core 2 N-acetylglucosamine-transferase 1 and α-2,6-sialyltransferase 1 into COPI vesicles. Depletion of GOLPH3 altered the subcellular localization of these enzymes. In contrast, galactosyltransferase, an enzyme that does not interact with GOLPH3, was neither incorporated into COPI vesicles nor was dependent on GOLPH3 for proper localization.     

Growth kinetics of the Golgi apparatus during the cell cycle in onion root meristems

In onion root meristems, the number of dictyosomes per cell shows a kinetics of growth strongly related to the cell cycle. During the interphase of steady-state proliferative cells, the volume density and numerical density of the Golgi apparatus decrease to reach minimum values in late-interphase cells, characterized by their greatest length. This pattern is also found in the total volume occupied by Golgi apparatus. Once in mitosis, the above-mentioned parameters begin to increase reaching maximum mean values in telophase. After the experimental uncoupling of chromosome and growth cycles by presynchronization with hydroxyurea, we found a similar behaviour pattern in the Golgi apparatus: decreasing values during interphase and a triggering of Golgi-apparatus growth in prophase independently of the bigger cell sizes reached in mitosis as an effect of pretreatment with hydroxyurea. These results indicate a cyclic kinetics of this subcellular component in higher-plant meristems, coupled with early mitotic events.     

Glycosyltransferases with endogenous acceptor activity in plasma membranes isolated from rat liver

Plasma membrane fractions from rat liver exhibited glycosyltransferase activity with endogenous membrane-associated acceptors and either UDP-galactose, UDPglucose, UDP-N-acetylglucosamine, or GDPmannose donors. Of these, incorporation into non-lipid acceptors was most active with UDP-galactose and only with UDPgalactose and UDPmannose was there incorporation into endogenous lipid acceptors. CMP-N-acetylneuraminic acid was inactive as a donor with the isolated plasma membranes. In order to demonstrate transferase activity, low concentrations of substrate sugar nucleotides and short incubation times were used as well as sulfhydryl protectants and a phosphatase inhibitor (NaF) in the reaction mixtures. The findings support the concept of surface localization of at least a galactosyl transferase in cells of rat liver.     

Microtubule independent vesiculation of Golgi membranes and the reassembly of vesicles into Golgi stacks

We have recently shown that ilimaquinone (IQ) causes the breakdown of Golgi membranes into small vesicles (VGMs for vesiculated Golgi membranes) and inhibits vesicular protein transport between successive Golgi cisternae (Takizawa et al., 1993). While other intracellular organelles, intermediate filaments, and actin filaments are not affected, we have found that cytoplasmic microtubules are depolymerized by IQ treatment of NRK cells. We provide evidence that IQ breaks down Golgi membranes regardless of the state of cytoplasmic microtubules. This is evident from our findings that Golgi membranes break down with IQ treatment in the presence of taxol stabilized microtubules. Moreover, in cells where the microtubules are first depolymerized by microtubule disrupting agents which cause the Golgi stacks to separate from one another and scatter throughout the cytoplasm, treatment with IQ causes further breakdown of these Golgi stacks into VGMs. Thus, IQ breaks down Golgi membranes independently of its effect on cytoplasmic microtubules. Upon removal of IQ from NRK cells, both microtubules and Golgi membranes reassemble. The reassembly of Golgi membranes, however, takes place in two sequential steps: the first is a microtubule independent process in which the VGMs fuse together to form stacks of Golgi cisternae. This step is followed by a microtubule-dependent process by which the Golgi stacks are carried to their perinuclear location in the cell. In addition, we have found that IQ has no effect on the structural organization of Golgi membranes at 16 degrees C. However, VGMs generated by IQ are capable of fusing and assembling into stacks of Golgi cisternae at 16 degrees C. This is in contrast to the cells recovering from BFA treatment where, after removal of BFA at 16 degrees C, resident Golgi enzymes fail to exit the ER, a process presumed to require the formation of vesicles. We propose that at 16 degrees C there may be general inhibition in the process of vesicle formation, whereas the process of vesicle fusion is not affected.     

The Golgi apparatus and cell polarity: Roles of the cytoskeleton,the Golgi matrix,and Golgi membranes

Membrane trafficking plays a crucial role in cell polarity by directing lipids and proteins to specific subcellular locations in the cell and sustaining a polarized state. The Golgi apparatus, the master organizer of membrane trafficking, can be subdivided into three layers that play different mechanical roles: a cytoskeletal layer, the so-called Golgi matrix, and the Golgi membranes. First, the outer regions of the Golgi apparatus interact with cytoskeletal elements, mainly actin and microtubules, which shape, position, and orient the organelle. Closer to the Golgi membranes, a matrix of long coiled–coiled proteins not only selectively captures transport intermediates but also participates in signaling events during polarization of membrane trafficking. Finally, the Golgi membranes themselves serve as active signaling platforms during cell polarity events. We review here the recent findings that link the Golgi apparatus to cell polarity, focusing on the roles of the cytoskeleton, the Golgi matrix, and the Golgi membranes.     

Puromycin does not inactivate the galactrosyltransferase of Golgi membranes.

The effects of puromycin on galactosyltransferase from four sources, rat liver and lactating sheep mammary Golgi membranes, bovine colostrum and human serum have been investigated. We do not find that the synthesis of N-acetyllactosamine is much inhibited, even in the presence of 3mM puromycin. This is in contrast to previously reported results for rat liver Golgi membranes. We interpret the low level of inhibition observed in terms of pH effects.     

A microtubule-binding protein associated with membranes of the Golgi apparatus

A monoclonal antibody (M3A5), raised against microtubule-associated protein 2 (MAP-2), recognized an antigen associated with the Golgi complex in a variety of non-neuronal tissue culture cells. In double immunofluorescence studies M3A5 staining was very similar to that of specific Golgi markers, even after disruption of the Golgi apparatus organization with monensin or nocodazole. M3A5 recognized one band of Mr approximately 110,000 in immunoblots of culture cell extracts; this protein, designated 110K, was enriched in Golgi stack fractions prepared from rat liver. The 110K protein has been shown to partition into the aqueous phase by Triton X-114 extraction of a Golgi-enriched fraction and was eluted after pH 11.0 carbonate washing. It is therefore likely to be a peripheral membrane protein. Proteinase K treatment of an isolated Golgi stack fraction resulted in complete digestion of the 110K protein, both in the presence and absence of Triton X-100. A the 110K protein is accessible to protease in intact vesicles in vitro, it is presumably located on the cytoplasmic face of the Golgi membrane in vivo. The 110K protein was able to interact specifically with taxol-polymerized microtubules in vitro. These results suggest that the 110K protein may serve to link the Golgi apparatus to the microtubule network and so may belong to a novel class of proteins: the microtubule-binding proteins.     

Recycling of Golgi-resident Glycosyltransferases through the ER Reveals a Novel Pathway and Provides an Explanation for Nocodazole-induced Golgi Scattering

During microtubule depolymerization, the central, juxtanuclear Golgi apparatus scatters to multiple peripheral sites. We have tested here whether such scattering is due to a fragmentation process and subsequent outward tracking of Golgi units or if peripheral Golgi elements reform through a novel recycling pathway. To mark the Golgi in HeLa cells, we stably expressed the Golgi stack enzyme N-acetylgalactosaminyltransferase-2 (GalNAc-T2) fused to the green fluorescent protein (GFP) or to an 11–amino acid epitope, VSV-G (VSV), and the trans/TGN enzyme β1,4-galactosyltransferase (GalT) fused to GFP. After nocodazole addition, time-lapse microscopy of GalNAc-T2–GFP and GalT–GFP revealed that scattered Golgi elements appeared abruptly and that no Golgi fragments tracked outward from the compact, juxtanuclear Golgi complex. Once formed, the scattered structures were relatively stable in fluorescence intensity for tens of minutes. During the entire process of dispersal, immunogold labeling for GalNAc-T2–VSV and GalT showed that these were continuously concentrated over stacked Golgi cisternae and tubulovesicular Golgi structures similar to untreated cells, suggesting that polarized Golgi stacks reform rapidly at scattered sites. In fluorescence recovery after photobleaching over a narrow (FRAP) or wide area (FRAP-W) experiments, peripheral Golgi stacks continuously exchanged resident proteins with each other through what appeared to be an ER intermediate. That Golgi enzymes cycle through the ER was confirmed by microinjecting the dominant-negative mutant of Sar1 (Sar1p^dn) blocking ER export. Sar1p^dn was either microinjected into untreated or nocodazole-treated cells in the presence of protein synthesis inhibitors. In both cases, this caused a gradual accumulation of GalNAc-T2–VSV in the ER. Few to no peripheral Golgi elements were seen in the nocodazole-treated cells microinjected with Sar1p^dn. In conclusion, we have shown that Golgi-resident glycosylation enzymes recycle through the ER and that this novel pathway is the likely explanation for the nocodazole-induced Golgi scattering observed in interphase cells.     

Interaction of membranes of the Golgi complex with microtubules in vitro.

We have developed an in vitro assay for characterizing the binding of elements of the Golgi complex to microtubules. The binding assay comprises three distinct components, Golgi elements purified from Vero cells by subcellular fractionation, taxol-polymerized tubulin from bovine brain coupled to magnetic beads and cytosol from HeLa cells. Binding of Golgi elements to microtubules is quantitated by measuring the activity of the Golgi marker enzyme, galactosyltransferase, associated with the microtubule-coated beads retrieved with a magnet. In the presence of cytosol, 35 to 45% of the total input of galactosyltransferase activity (Golgi elements) bind to microtubules; only 3% of the Golgi elements bind to microtubules, however, in the absence of cytosolic factors. This binding is saturable at a cytosol concentration of approximately 5 mg/ml or at a high input of Golgi elements. Cytosol-stimulated binding of Golgi elements to microtubules is decreased to less than 15% when cytosol is pretreated with 2 mM N-ethylmaleimide (NEM) and it is abolished when cytosolic proteins are inactivated by heat or when microtubules have been coated with heat-stable microtubule-associated proteins (MAPs). Trypsinization of the membranes of the Golgi elements abolishes their ability to bind to microtubules. Furthermore, inactivation of cytoplasmic dynein by UV/vanadate treatment does not affect the binding. This suggests that the interaction of Golgi elements with microtubules depends on NEM-sensitive cytosolic factors and membrane-associated receptors, but not on the microtubule-based motor protein cytoplasmic dynein.     

Tension in tubulovesicular networks of Golgi and endoplasmic reticulum membranes

The endoplasmic reticulum (ER) and Golgi have robust bidirectional traffic between them and yet form distinct membrane compartments. Membrane tubules are pulled from large aggregates of ER or Golgi by microtubule motors to form ER tubulovesicular networks or Golgi tubules both in vivo and in vitro. The physical properties of membranes are critical for membrane traffic and organelle morphology. For example, tension applied to membranes can create tethers, drive membrane flow, and set the diameter of the tubules. Here, we formed ER and Golgi membrane networks in vitro and used optical tweezers to measure directly, for the first time, the membrane tensions of these organelles to clarify the possible role of tension in membrane flow. We report that higher forces are needed to form tethers from ER (18.6 +/- 2.8 pN) than from Golgi (11.4 +/- 1.4 pN) membrane tubules in vitro. Since ER tubules are smaller in diameter than Golgi tubules, it follows that Golgi networks have a lower tension than ER. The lower tension of the ER could be an explanation of how Golgi tubules can be rapidly drawn into the ER by tension-driven flow after fusion, as is observed in vivo.     

Transbilayer movement of monohexosylsphingolipids in endoplasmic reticulum and Golgi membranes

The transbilayer movement of glycosphingolipids has been characterized in Golgi, ER, plasma, and model membranes using spin-labeled and fluorescent analogues of the monohexosylsphingolipids glucosylceramide and galactosylceramide and of the dihexosylsphingolipid lactosylceramide. In large unilamellar lipid vesicles, monohexosylsphingolipids underwent a slow transbilayer diffusion (half-time between 2 and 5 h at 20 degrees C). Similarly, the inward redistribution of these sphingolipids in the plasma membrane of the hepatocyte-like cell line HepG2 and of erythrocytes was slow. However, in rat liver ER and Golgi membranes, we found a rapid transbilayer movement of spin-labeled monohexosylsphingolipids (half-time of approximately 3 min at 20 degrees C), which suggests the existence of a monohexosylsphingolipid flippase. The transbilayer movement of glucosylceramide in the Golgi and the ER displayed a saturable behavior, was inhibited by proteolysis, did not require Mg-ATP, and occurs in both directions. Treatment with DIDS inhibited the flip-flop of glucosylceramide but not that of phosphatidylcholine. These data suggest that the transbilayer movement of monoglucosylceramide in the ER and in the Golgi involves a protein that could be distinct from that previously evidenced for glycerophospholipids in the ER. In vivo, transbilayer diffusion should promote a symmetric distribution of monohexosylsphingolipids which are synthesized in the cytosolic leaflet. This should allow glucosylceramide rapid access to the lumenal leaflet where it is converted to lactosylceramide. No significant transbilayer movement of lactosylceramide occurred in both artificial and natural membranes over 1 h. Thus, lactosylceramide, in turn, is unable to diffuse to the cytosolic leaflet and remains at the lumenal leaflet where it undergoes the subsequent glycosylations.     

Identification of lipoprotein-binding proteins in rat liver Golgi apparatus membranes

The low density lipoprotein (LDL) receptor has been shown to be a plasma membrane glycoprotein responsible for the cellular binding and endocytosis of plasma lipoproteins. Inasmuch as the Golgi apparatus has been shown to participate in glycoprotein processing and in the assembly of plasma lipoproteins by hepatic and intestinal epithelial cells, the present studies were designed to test the hypothesis that lipoprotein receptors are present within Golgi membranes. Utilizing ligand blotting with a variety of iodinated lipoproteins, several lipoprotein-binding proteins were identified in rat liver Golgi membranes at apparent molecular weights (Mr) 200,000, 160,000, 130,000, 120,000, 100,000, 80,000, and 70,000. The 130,000 protein was the most prominent and was identified as the mature LDL receptor by its binding characteristics and an Mr characteristic of the plasma membrane receptor. Enzymatic deglycosylation studies suggested that the 120,000 and 100,000 proteins were LDL receptor precursors lacking sialic acid. Antibody to the LDL receptor recognized all the bands on immunoblots except the 70,000 protein, with the 130,000 protein being the most prominent. Isolation of the Golgi fractions in the presence of protease inhibitors did not eliminate any of the proteins recognized by the antibody but did result in sharper bands on the blots. Additionally, we investigated the hypothesis that conditions that regulate plasma membrane LDL receptors also cause detectable changes in receptors in Golgi membranes. All the binding proteins were increased in Golgi membranes from rats treated with 17-alpha-ethynylestradiol. Colchicine caused an accumulation of 120,000 Mr protein, suggesting blockage of final sialylation in the trans Golgi. When protein synthesis was inhibited by cycloheximide, there was no reduction of mature LDL receptors in Golgi membranes, consistent with recycling of receptors through this organelle.     

Enzymatic O-glycosylation of kappa-caseinomacropeptide by ovine mammary Golgi membranes

The results of subcellular fractionation of sheep mammary gland membranes indicate that N-acetylgalactosaminyl polypeptide transferase and galactosyl-N-acetylgalactosaminyl transferase, which are involved in the assembly of disaccharide units of kappa-casein, are localized chiefly in Golgi membranes. The glycosyltransferase activities incorporating N-acetyl [1-14C] galactosamine and [U-14C] galactose from uridine diphosphate N-acetyl [1-14C] galactosamine and uridine diphosphate [U-14C] galactose, respectively, were measured after membrane solubilization with Triton X-100 either with unglycosylated caseinomacropeptide, or with this polypeptide containing the N-acetylgalactosamine side chain residues (desialylated and degalactosylated caseinomacropeptide). Radioactive N-acetylgalactosamine was incorporated in the unglycosylated acceptor peptide, and the glycosidic bonds in the product were alkali labile, suggesting that they were linked to the hydroxyamino acid residues. In addition radioactive N-acetylgalactosamine was released after alpha N-acetyl-D-galactosaminidase treatment of labelled caseinomacropeptide. [U-14C] galactose was incorporated in the desialylated and degalactosylated acceptor peptide. Reductive alkaline treatment of [U-14C] galactose peptide resulted in the release of a major product, the chromatographic properties of which in TLC were identical with authentic galactosyl (1 leads to 3) N-acetylgalactosaminitol. The structure of the labelled disacchariditol determined after periodate oxidation (two equivalents) by gas liquid chromatography-mass spectrometry revealed that the [U-14C] galactose was linked to position C-3 on the N-acetylgalactosaminyl-residue. The anomery of the galactose, as determined by a chemical method, indicates unambiguously a beta configuration.     

Biochemical dissection of AP-1 recruitment onto Golgi membranes

Recruitment of the Golgi-specific AP-1 adaptor complex onto Golgi membranes is thought to be a prerequisite for clathrin coat assembly on the TGN. We have used an in vitro assay to examine the translocation of cytosolic AP-1 onto purified Golgi membranes. Association of AP-1 with the membranes required GTP or GTP analogues and was inhibited by the fungal metabolite, brefeldin A. In the presence of GTP gamma S, binding of AP-1 to Golgi membranes was strictly dependent on the concentration of cytosol added to the assay. AP-1 recruitment was also found to be temperature dependent, and relatively rapid at 37 degrees C, following a lag period of 3 to 4 min. Using only an adaptor-enriched fraction from cytosol, purified myristoylated ARF1, and Golgi membranes, the GTP gamma S-dependent recruitment of AP-1 could be reconstituted. Our results show that the association of the AP-1 complex with Golgi membranes, like the coatomer complex, requires ARF, which accounts for the sensitivity of both to brefeldin A. In addition, they provide the basis for a model for the early biochemical events that lead to clathrin-coated vesicle formation on the TGN.