Regulation of vesicular and tubular membrane traffic of the Golgi complex by coat proteins.

Transport of cargo through and from the Golgi complex is mediated by vesicular carriers and transient tubular connections. Two classes of vesicle have been implicated in the biosynthetic or anterograde membrane traffic of this organelle. Both classes of vesicle are coated on the cytoplasmic surface with proteins, of which at least one component is related. Tubular connections also enable exchange of material between membrane-bounded compartments associated with the Golgi complex, most obviously in cells that have been treated with the drug, brefeldin A. Coat proteins appear to be involved in the regulation of these transport processes. Their putative functions include sorting of cargo, as well as regulation of budding, fusion or targeting of the membrane carriers.

Coat proteins in intracellular membrane transport.

Transport of newly synthesized material from the endoplasmic reticulum (ER) towards the Golgi complex, through the Golgi cisternae, and out of the trans-Golgi network (TGN) is thought to be mediated by vesicular carriers. Different types of vesicle are involved in this biosynthetic membrane traffic. All are coated with protein complexes on their cytoplasmic surface. COP-coated vesicles have recently been implicated in transport of cargo from the ER to the TGN, and clathrin-coated vesicles from the TGN to endosomes, but the carriers moving material to the cell surface are still unknown. Sequence homologies between subunits of the COP- and the clathrin-adaptor complexes suggest that coat proteins may belong to a family of proteins with related functions. The precise role of the coat proteins is not fully understood, although they have been implicated in clustering of cargo into buds and in budding of vesicles. In addition, coat proteins may play an essential role in targeting of transport intermediates and may serve to regulate membrane fusion.

Motors and membrane traffic.

The cytoskeleton is essential for the proper function of many components of secretory and endocytic pathways, and the importance both of microtubule motors (kinesins and dyneins) and of actin motors (myosins) in these pathways is becoming apparent. Recent experiments have improved our understanding of which members of these motor protein families are involved in membrane traffic. Multiple motors are probably involved in the control of the morphology and dynamics of many membranes, and intriguing hints about how these motors are coordinated are appearing.

Probing the mobility of membrane proteins inside the cell.

Studies using a variety of microscopy-based approaches have led to a consensus that most cell-surface proteins are highly mobile and diffuse rapidly within fenced microdomains. Little attention, however, has so far been given to the analysis of the mobility of intracellular membrane proteins because of their comparative inaccessibility. Recent advances in microinjection, confocal microscopy and the construction of epitope-tagged proteins or of hybrids with an intrinsically fluorescent protein have allowed intracellular membrane proteins to be studied using approaches previously applied to characterize the mobility of cell-surface proteins. Confocal fluorescence recovery after photobleaching (c-FRAP) experiments show that intracellular membrane proteins may also be highly mobile.

CLIPs for organelle-microtubule interactions.

Interactions of intracellular membranes with microtubules play a fundamental role in the dynamic organization of cytoplasmic organelles. The microtubule-based motors kinesin and cytoplasmic dynein are responsible for directed movement of vesicles and organelles, but in vitro assays indicate the existence of another class of proteins linking membranes to microtubules. CLIP-170, a cytoplasmic linker protein that mediates binding of endosomes to microtubules, provides a paradigm for understanding how these proteins may complement the role of motors in regulating microtubule-dependent membrane trafficking.

Translocation and clustering of endosomes and lysosomes depends on microtubules.

Indirect immunofluorescence labeling of normal rat kidney (NRK) cells with antibodies recognizing a lysosomal glycoprotein (LGP 120; Lewis, V., S.A. Green, M. Marsh, P. Vihko, A. Helenius, and I. Mellman, 1985, J. Cell Biol., 100:1839-1847) reveals that lysosomes accumulate in the region around the microtubule-organizing center (MTOC). This clustering of lysosomes depends on microtubules. When the interphase microtubules are depolymerized by treatment of the cells with nocodazole or during mitosis, the lysosomes disperse throughout the cytoplasm. Lysosomes recluster rapidly (within 30-60 min) in the region of the centrosomes either upon removal of the drug, or, in telophase, when repolymerization of interphase microtubules has occurred. During this translocation process the lysosomes can be found aligned along centrosomal microtubules. Endosomes and lysosomes can be visualized by incubating living cells with acridine orange. We have analyzed the movement of these labeled endocytic organelles in vivo by video-enhanced fluorescence microscopy. Translocation of endosomes and lysosomes occurs along linear tracks (up to 10 microns long) by discontinuous saltations (with velocities of up to 2.5 microns/s). Organelles move bidirectionally with respect to the MTOC. This movement ceases when microtubules are depolymerized by treatment of the cells with nocodazole. After nocodazole washout and microtubule repolymerization, the translocation and reclustering of fluorescent organelles predominantly occurs in a unidirectional manner towards the area of the MTOC. Organelle movement remains unaffected when cells are treated with cytochalasin D, or when the collapse of intermediate filaments is induced by microinjected monoclonal antivimentin antibodies. It can be concluded that translocation of endosomes and lysosomes occurs along microtubules and is independent of the intermediate filament and microfilament networks.

Binding of E-MAP-115 to microtubules is regulated by cell cycle-dependent phosphorylation.

Expression levels of E-MAP-115, a microtubule-associated protein that stabilizes microtubules, increase with epithelial cell polarization and differentiation (Masson and Kreis, 1993). Although polarizing cells contain significant amounts of this protein, they can still divide and thus all stabilized microtubules must disassemble at the onset of mitosis to allow formation of the dynamic mitotic spindle. We show here that binding of E-MAP-115 to microtubules is regulated by phosphorylation during the cell cycle. Immunolabeling of HeLa cells for E-MAP-115 indicates that the protein is absent from microtubules during early prophase and progressively reassociates with microtubules after late prophase. A fraction of E-MAP-115 from HeLa cells released from a block at the G1/S boundary runs with higher apparent molecular weight on SDS-PAGE, with a peak correlating with the maximal number of cells in early stages of mitosis. E-MAP-115 from nocodazole-arrested mitotic cells, which can be obtained in larger amounts, displays identical modifications and was used for further biochemical characterization. The level of incorporation of 32P into mitotic E-MAP-115 is about 15-fold higher than into the interphase protein. Specific threonine phosphorylation occurs in mitosis, and the amount of phosphate associated with serine also increases. Hyperphosphorylated E-MAP-115 from mitotic cells cannot bind stably to microtubules in vitro. These results suggest that phosphorylation of E-MAP-115 is a prerequisite for increasing the dynamic properties of the interphase microtubules which leads to the assembly of the mitotic spindle at the onset of mitosis. Microtubule-associated proteins are thus most likely key targets for kinases which control changes in microtubule dynamic properties at the G2- to M-phase transition.

Identification and molecular characterization of E-MAP-115, a novel microtubule-associated protein predominantly expressed in epithelial cells.

A novel microtubule-associated protein (MAP) of M(r) 115,000 has been identified by screening of a HeLa cell cDNA expression library with an anti-serum raised against microtubule-binding proteins from HeLa cells. Monoclonal and affinity-purified polyclonal antibodies were generated for the further characterization of this MAP. It is different from the microtubule-binding proteins of similar molecular weights, characterized so far, by its nucleotide-insensitive binding to microtubules and different sedimentation behavior. Since it is predominantly expressed in cells of epithelial origin (Caco-2, HeLa, MDCK), and rare (human skin, A72) or not detectable (Vero) in fibroblastic cells, we name it E-MAP-115 (epithelial MAP of 115 kD). In HeLa cells, E-MAP-115 is preferentially associated with subdomains or subsets of perinuclear microtubules. In Caco-2 cells, labeling for E-MAP-115 increases when they polarize and form blisters. The molecular characterization of E-MAP-115 reveals that it is a novel protein with no significant homologies to other known proteins. The secondary structure predicted from its sequence indicates two domains connected by a putative hinge region rich in proline and alanine (PAPA region). E-MAP-115 has two highly charged regions with predicted alpha-helical structure, one basic with a pI of 10.9 in the NH2-terminal domain and one neutral with a pI of 7.6 immediately following the PAPA region in the acidic COOH-terminal half of the molecule. A novel microtubule-binding site has been localized to the basic alpha-helical region in the NH2-terminal domain using in vitro microtubule-binding assays and expression of mutant polypeptides in vivo. Overexpression of this domain of E-MAP-115 by transfection of fibroblasts lacking significant levels of this protein with its cDNA renders microtubules stable to nocodazole. We conclude that E-MAP-115 is a microtubule-stabilizing protein that may play an important role during reorganization of microtubules during polarization and differentiation of epithelial cells.

Identification of a novel nucleotide-sensitive microtubule-binding protein in HeLa cells.

A protein of Mr 170,000 (170K protein) has been identified in HeLa cells, using an antiserum raised against HeLa nucleotide-sensitive microtubule-binding proteins. Affinity-purified antibodies specific for this 170K polypeptide were used for its characterization. In vitro sedimentation of the 170K protein with taxol microtubules polymerized from HeLa high-speed supernatant is enhanced in the presence of an ATP depleting system, but unaffected by the non-hydrolyzable ATP analogue AMP-PNP. In addition, it can be eluted from taxol microtubules by ATP or GTP, as well as NaCl. Thus it shows microtubule-binding characteristics distinct from those of the previously described classes of nucleotide-sensitive microtubule-binding proteins, the motor proteins kinesin and cytoplasmic dynein, homologues of which are also present in HeLa cells. The 170K protein sediments on sucrose gradients at approximately 6S, separate from kinesin (9.5S) and cytoplasmic dynein (20S), further indicating that it is not associated with these motor proteins. Immunofluorescence localization of the 170K protein shows a patchy distribution in interphase HeLa cells, often organized into linear arrays that correlate with microtubules. However, not all microtubules are labeled, and there is a significant accumulation of antigen at the peripheral ends of microtubules. In mitotic cells, 170K labeling is found in the spindle, but there is also dotty labeling in the cytoplasm. After depolymerization of microtubules by nocodazole, the staining pattern is also patchy but not organized in linear arrays, suggesting that the protein may be able to associate with other intracellular structures as well as microtubules. In vinblastine-treated cells, there is strong labeling of tubulin paracrystals, and random microtubules induced in vivo by taxol are also labeled by the antibodies. These immunofluorescence labeling patterns are stable to extraction of cells with Triton X-100 before fixation, further suggesting an association of the protein with cytoplasmic structures. In vivo, therefore, the 170K protein appears to be associated with a subset of microtubules at discrete sites. Its in vitro behavior suggests that it belongs to a novel class of nucleotide-sensitive microtubule-binding proteins.

Control of microtubule nucleation and stability in Madin-Darby canine kidney cells: the occurrence of noncentrosomal, stable detyrosinated microtubules.

The microtubule-nucleating activity of centrosomes was analyzed in fibroblastic (Vero) and in epithelial cells (PtK2, Madin-Darby canine kidney [MDCK]) by double-immunofluorescence labeling with anti-centrosome and antitubulin antibodies. Most of the microtubules emanated from the centrosomes in Vero cells, whereas the microtubule network of MDCK cells appeared to be noncentrosome nucleated and randomly organized. The pattern of microtubule organization in PtK2 cells was intermediate to the patterns observed in the typical fibroblastic and epithelial cells. The two centriole cylinders were tightly associated and located close to the nucleus in Vero and PtK2 cells. In MDCK cells, however, they were clearly separated and electron microscopy revealed that they nucleated only a few microtubules. The stability of centrosomal and noncentrosomal microtubules was examined by treatment of these different cell lines with various concentrations of nocodazole. 1.6 microM nocodazole induced an almost complete depolymerization of microtubules in Vero cells; some centrosome nucleated microtubules remained in PtK2 cells, while many noncentrosomal microtubules resisted that treatment in MDCK cells. Centrosomal and noncentrosomal microtubules regrew in MDCK cells with similar kinetics after release from complete disassembly by high concentrations of nocodazole (33 microM). During regrowth, centrosomal microtubules became resistant to 1.6 microM nocodazole before the noncentrosomal ones, although the latter eventually predominate. We suggest that in MDCK cells, microtubules grow and shrink as proposed by the dynamic instability model but the presence of factors prevents them from complete depolymerization. This creates seeds for reelongation that compete with nucleation off the centrosome. By using specific antibodies, we have shown that the abundant subset of nocodazole-resistant microtubules in MDCK cells contained detyrosinated alpha-tubulin (glu tubulin). On the other hand, the first microtubules to regrow after nocodazole removal contained only tyrosinated tubulin. Glu-tubulin became detectable only after 30 min of microtubule regrowth. This strongly supports the hypothesis that alpha-tubulin detyrosination occurs primarily on "long lived" microtubules and is not the cause of the stabilization process. This is also supported by the increased amount of glu-tubulin that we found in taxol-treated cells.

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.

Microtubules are stabilized in confluent epithelial cells but not in fibroblasts.

Rhodamine-tagged tubulin was microinjected into epithelial cells (MDCK) and fibroblasts (Vero) to characterize the dynamic properties of labeled microtubules in sparse and confluent cells. Fringe pattern fluorescence photobleaching revealed two components with distinct dynamic properties. About one-third of the injected tubulin diffused rapidly in the cytoplasm with a diffusion coefficient of 1.3-1.6 x 10(-8) cm2/s. This pool of soluble cytoplasmic tubulin was increased to greater than 80% when cells were treated with nocodazole, or reduced to approximately 20% upon treatment of cells with taxol. Fluorescence recovery of the remaining two-thirds of labeled tubulin occurred with an average half-time (t1/2) of 9-11 min. This pool corresponds to labeled tubulin associated with microtubules, since it was sensitive to treatment of cells with nocodazole and since taxol increased its average t1/2 to greater than 22 min. Movement of photobleached microtubules in the cytoplasm with rates of several micrometers per minute was shown using very small interfringe distances. A significant change in the dynamic properties of microtubules occurred when MDCK cells reached confluency. On a cell average, microtubule half-life was increased about twofold to approximately 16 min. In fact, two populations of cells were detected with respect to their microtubule turnover rates, one with a t1/2 of approximately 9 min and one with a t1/2 of greater than 25 min. Correspondingly, the rate of incorporation of microinjected tubulin into interphase microtubules was reduced about twofold in confluent MDCK cells. In contrast to the MDCK cells, no difference in microtubule dynamics was observed in sparse and confluent populations of Vero fibroblasts, where the average microtubule half-life was approximately 10 min. Thus, microtubules are significantly stabilized in epithelial but not fibroblastic cells grown to confluency.

Dissociation of coatomer from membranes is required for brefeldin A-induced transfer of Golgi enzymes to the endoplasmic reticulum.

Addition of brefeldin A (BFA) to mammalian cells rapidly results in the removal of coatomer from membranes and subsequent delivery of Golgi enzymes to the endoplasmic reticulum (ER). Microinjected anti-EAGE (intact IgG or Fab-fragments), antibodies against the "EAGE"-peptide of beta-COP, inhibit BFA-induced redistribution of beta-COP in vivo and block transfer of resident proteins of the Golgi complex to the ER; tubulo-vesicular clusters accumulate and Golgi membrane proteins concentrate in cytoplasmic patches containing beta-COP. These patches are devoid of marker proteins of the ER, the intermediate compartment (IC), and do not contain KDEL receptor. Interestingly, relocation of KDEL receptor to the IC, where it colocalizes with ERGIC53 and ts-O45-G, is not inhibited under these conditions. While no stacked Golgi cisternae remain in these injected cells, reassembly of stacks of Golgi cisternae following BFA wash-out is inhibited to only approximately 50%. Mono- or divalent anti-EAGE stabilize binding of coatomer to membranes in vitro, at least as efficiently as GTP(gamma)S. Taken together these results suggest that enhanced binding of coatomer to membranes completely inhibits the BFA-induced retrograde transport of Golgi resident proteins to the ER, probably by inhibiting fusion of Golgi with ER membranes, but does not interfere with the disassembly of the stacked Golgi cisternae and recycling of KDEL receptor to the IC. These results confirm our previous results suggesting that COPI is involved in anterograde membrane transport from the ER/IC to the Golgi complex (Pepperkok et al., 1993), and corroborate that COPI regulates retrograde membrane transport between the Golgi complex and ER in mammalian cells.

Beta-COP is essential for transport of protein from the endoplasmic reticulum to the Golgi in vitro.

Using a novel in vitro assay which allows us to distinguish vesicle budding from subsequent targeting and fusion steps, we provide the first biological evidence that beta-COP, a component of non-clathrin-coated vesicles believed to mediate intraGolgi transport, is essential for transport of protein from the ER to the cis-Golgi compartment. Incubation in the presence of beta-COP specific antibodies and F(ab) fragments prevents the exit of vesicular stomatitis virus glycoprotein (VSV-G) from the ER. These results demonstrate that beta-COP is required for the assembly of coat complexes mediating vesicle budding. Fractionation of rat liver cytosol revealed that a major biologically active form of beta-COP was found in a high molecular pool (> 1,000 kD) distinct from coatomer and which promoted efficient vesicle budding from the ER. Surprisingly, rab1B could be quantitatively coprecipitated with this beta-COP containing complex and was also essential for function. We suggest that beta-COP functions in an early step during vesicle formation and that rab1B may be recruited as a component of a precoat complex which participates in the export of protein from the ER via vesicular carriers.

Accumulation of a microtubule-binding protein, pp170, at desmosomal plaques.

The establishment of epithelial cell polarity correlates with the formation of specialized cell-cell junctions and striking changes in the organization of microtubules. A significant fraction of the microtubules in MDCK cells become stabilized, noncentrosomally organized, and arranged in longitudinal bundles in the apical-basal axis. This correlation suggests a functional link between cell-cell junction formation and control of microtubule organization. We have followed the distribution of pp170, a recently described microtubule-binding protein, during establishment of epithelial cell polarity. This protein shows the typical patchy distribution along microtubules in subconfluent fibroblasts and epithelial cells, often associated with the peripheral ends of a subpopulation of microtubules. In contrast to its localization in confluent fibroblasts (A72) and HeLa cells, however, pp170 accumulates in patches delineating the regions of cell-cell contacts in confluent polarizing epithelial cells (MDCK and Caco-2). Double immunolocalization with antibodies specific for cell-cell junction proteins, confocal microscopy, and immunoelectron microscopy on polarized MDCK cells suggest that pp170 accumulates at desmosomal plaques. Furthermore, microtubules and desmosomes are found in close contact. Maintenance of the desmosomal association of pp170 is dependent on intact microtubules in 3-d-old, but not in 1-d-old MDCK cell cultures. This suggests a regulated interaction between microtubules and desmosomes and a role for pp170 in the control of changes in the properties of microtubules induced by epithelial cell-cell junction formation.

Beta-COP localizes mainly to the cis-Golgi side in exocrine pancreas.

We examined the distribution of the non-clathrin-coated vesicle-associated coat protein beta-COP in rat exocrine pancreatic cells by immunogold cytochemistry. Labeling for beta-COP was found in the Golgi region (48%) where it was associated with vesicles and buds of approximately 50 nm, showing a characteristic approximately 10-nm-thick coat. The other half of the label was present in the cytoplasm, not associated with visible coats or membranes, with a minor fraction present on small clusters of tubules and vesicles. Clathrin-coated vesicles were typically located at the trans-side of the Golgi complex, and showed a thicker coat of approximately 18 nm. Of the total beta-COP labeling over the Golgi region, 68% occurred on the cis-side, 6% on the cisternae, 17% on the rims of the cisternae, and only 9% on the trans-side. For clathrin these figures were 16, 2, 4, and 78%, respectively. At the cis-Golgi side beta-COP was present in transitional areas (TA), on so-called peripheral elements (PE), consisting of tubules and vesicles located between the cup-shaped transitional elements (TE) of the RER and the cis-most Golgi cisternae. Label for Sec23p was also present in TA but was located closer to the TE, while beta-COP labeled PE were located near the cis-Golgi cisternae. Upon energy depletion, Golgi associated beta-COP was almost exclusively (86%) in spherical aggregates of 200-500 nm in diameter, whereas the cis-side (6%), the cisternae (1%), the rims (4%) and trans-side (3%) of the Golgi complex, were barely labeled; 50% of the total label remained in the cytoplasm. The aggregates were predominantly located at the cis-side of the Golgi stack, next to, but distinct from the Sec23p positive TA, that were devoid of beta-COP and had only a few recognizable vesicles left. Incubation with aluminum fluoride resulted in fragmentation of the Golgi complex into large clusters of beta-COP positive vesicles, while 50% of the label remained in the cytoplasm, as in control cells. After 10 min of Brefeldin A treatment 91% of beta-COP was cytoplasmic and only 7% associated with membranes of the Golgi complex. The total label for beta-COP over exocrine cells remained unchanged during the incubation with either of the drugs, indicating that the drugs induce reallocation of beta-COP. Our data suggest that beta-COP plays a role in membrane transport at the cis-side of the Golgi complex.

Dynamic interactions of fluorescently labeled microtubule-associated proteins in living cells.

Microtubule-associated proteins (MAPs) from calf brain were fluorescently labeled with 6-iodoacetamido fluorescein (I-AF). The modified MAPs (especially enriched for MAP2) were fully active in promoting tubulin polymerization in vitro and readily associated with cytoplasmic filaments when microinjected into living cultured cells. Double-labeling experiments indicated that the microinjected AF-MAPs were incorporated predominantly, if not exclusively, into cytoplasmic microtubules in untreated cells or paracrystals induced within vinblastine-treated cells. Similar results were obtained with different cell types (neuronal, epithelial, and fibroblastic) of diverse origin (man, mouse, chicken, and rat kangaroo). Mobility measurements of the microinjected AF-MAPs using the method of fluorescence-photobleaching recovery (FPR) revealed two populations of AF-MAPs with distinct dynamic properties: One fraction represents the soluble pool of MAPs and is mobile with a diffusion coefficient of D = 3 X 10(-9) cm2/s. The other fraction of MAPs is associated with the microtubules and is essentially immobile on the time scale of FPR experiments. However, it showed slow fluorescence recovery with an apparent half time of approximately 5 min. The slow recovery of fluorescence on defined photobleached microtubules occurred most probably by the incorporation of AF-MAPs from the soluble cytoplasmic pool into the bleached area. The bleached spot on defined microtubules remained essentially immobile during the slow recovery phase. These results suggest that MAPs can associate in vivo with microtubules of diverse cell types and that treadmilling of MAP2-containing microtubules in vivo, if it exists, is slower than 4 micron/h.

Dissociation of a 110-kD peripheral membrane protein from the Golgi apparatus is an early event in brefeldin A action.

Brefeldin A (BFA) has a profound effect on the structure of the Golgi apparatus, causing Golgi proteins to redistribute into the ER minutes after drug treatment. Here we describe the dissociation of a 110-kD cytoplasmically oriented peripheral membrane protein (Allan, V. J., and T. E. Kreis. 1986. J. Cell Biol. 103:2229-2239) from the Golgi apparatus as an early event in BFA action, preceding other morphologic changes. In contrast, other peripheral membrane proteins of the Golgi apparatus were not released but followed Golgi membrane into the ER during BFA treatment. The 110-kD protein remained widely dispersed throughout the cytoplasm during drug treatment, but upon removal of BFA it reassociated with membranes during reformation of the Golgi apparatus. Although a 30-s exposure to the drug was sufficient to cause the redistribution of the 110-kD protein, removal of the drug after this short exposure resulted in the reassociation of the 110-kD protein and no change in Golgi structure. If cells were exposed to BFA for 1 min or more, however, a portion of the Golgi membrane was committed to move into and out of the ER after removal of the drug. ATP depletion also caused the reversible release of the 110-kD protein, but without Golgi membrane redistribution into the ER. These findings suggest that the interaction between the 110-kD protein and the Golgi apparatus is dynamic and can be perturbed by metabolic changes or the drug BFA.

Dynamic localization of CLIP-170 to microtubule plus ends is coupled to microtubule assembly.

CLIP-170 is a cytoplasmic linker protein that localizes to plus ends of microtubules in vivo. In this study, we have characterized the microtubule-binding properties of CLIP-170, to understand the mechanism of its plus end targeting. We show that the NH2-terminal microtubule-interacting domain of CLIP-170 alone localizes to microtubule plus ends when transfected into cells. Association of CLIP-170 with newly-formed microtubules was observed in cells microinjected with biotinylated tubulin, used as a tracer for growing microtubules. Using in vitro assays, association of CLIP-170 with recently polymerized tubulin is also seen. Cross-linking and sedimentation velocity experiments suggest association of CLIP-170 with nonpolymerized tubulin. We conclude from these experiments that the microtubule end targeting of CLIP-170 is closely linked to tubulin polymerization.

Characterization of a cis-Golgi matrix protein, GM130.

Antisera raised to a detergent- and salt-resistant matrix fraction from rat liver Golgi stacks were used to screen an expression library from rat liver cDNA. A full-length clone was obtained encoding a protein of 130 kD (termed GM130), the COOH-terminal domain of which was highly homologous to a Golgi human auto-antigen, golgin-95 (Fritzler et al., 1993). Biochemical data showed that GM130 is a peripheral cytoplasmic protein that is tightly bound to Golgi membranes and part of a larger oligomeric complex. Predictions from the protein sequence suggest that GM130 is an extended rod-like protein with coiled-coil domains. Immunofluorescence microscopy showed partial overlap with medial- and trans-Golgi markers but almost complete overlap with the cis-Golgi network (CGN) marker, syntaxin5. Immunoelectron microscopy confirmed this location showing that most of the GM130 was located in the CGN and in one or two cisternae on the cis-side of the Golgi stack. GM130 was not re-distributed to the ER in the presence of brefeldin A but maintained its overlap with syntaxin5 and a partial overlap with the ER-Golgi intermediate compartment marker, p53. Together these results suggest that GM130 is part of a cis-Golgi matrix and has a role in maintaining cis-Golgi structure.