Cell, Vol. 61, 5-7.
April 6, 1990, Copyright
0 1990 by Cell Press
Minireview
Microtubules, Membrane Traffic, and Cell Organization Regis B. Kelly Department of Biochemistry and Biophysics University of California San Francisco, California 94143
Eukaryotic cells frequently disperse their endoplasmic reticulum throughout their cytoplasm and centralize their Golgi membranes in the vicinity of the nucleus, an organization that seems to be generated and maintained by microtubules. Because of the juxtanuclear position of the Golgi membranes, secretory proteins newly synthesized in the endoplasmic reticulum are first centralized and then dispersed to the cell surface. Recent experiments have shown that microtubules seem to be involved in some but not all of the membrane trafficking steps within the cell (Figure 1). The data suggest that only centralizing and dispersal steps involve microtubules and indicate a relationship between microtubule-based membrane traffic and cellular organization. In the biosynthetic pathway, membrane vesicles can leave the Golgi region carrying newly synthesized proteins to the cell surface along microtubule tracks. If microtubules are depolymerized, proteins are still secreted, immovement is not essential plying that microtubule-based for exocytosis- nor indeed is cellular organization since that is also disrupted by microtubule depolymerization. What is changed is that membrane traffic becomes much less directed. Newly synthesized protein is no longer targeted to the growing edge of fibroblasts, for example. When microtubules are disrupted in epithelial cells, delivery to the apical surface, but not the basolateral, is perturbed in most cases studied. Neurons use microtubules for axonal transport. When endocrine cells develop long processes, movement of secretory granules to the process tips clearly involves microtubules (Kreis et al., 1989). In the secretory pathway, then, movement to the cell surface along microtubules is not essential, but is used to deliver newly synthesized material to defined surface locations. Selective targeting of post-Golgi vesicles to a region of the cell surface-may require the selective laying down or stabilizing of microtubule pathways from the Pans-Golgi network to specific regions of the cell surface. Microtubules are also involved in a late step in endocytosis. Newly endocytosed material is delivered to the early endosome (Figure 1) where material to be returned to the cell surface is sorted from material to be centralized. Early endosomes accumulate endocytotic material for about 10 min and then generate transport vesicles which will be taken to the cell center. Electron microscopy (Gruenberg et al., 1989) identified the endocytotic transport vesicle and showed it to be large (0.5 frrn diameter) and to be targeted to the prelysosomal compartment (Figure 1). Gruenberg et al. (1989) provide strong evidence that their large endocytotic transport vesicle is indeed the vesicle involved in microtubule-based movement to the cell center. Depolymerizing microtubules with nocodazole prevents
the transport vesicle from fusing with prelysosomal compartments and protects the contents of the transport vesicle from lysosomal degradation. What remains a puzzle, however, is why the cell takes endocytosed material to the nuclear region first, before degrading it. Why doesn’t degradation take place at the periphery? An attractive speculation is that a major function of the transport vesicle (Figure 1) is to recycle secretory vesicle membrane to the tran.sQolgi network in addition to its role in delivering contents to the lysosome. The most recently identified pathway of microtubulebased vesicle movement involves the salvage (Pelham, 1989) or intermediate (Lippincott-Schwartz et al., 1990) compartment between endoplasmic reticulum and Golgi complex. The salvage compartment (Figure 1) is identified in the electron microscope as the tubules and vesicles that accumulate in cells maintained at 15%, and in light microscopy by specific antibodies (Saraste et al., 1987; Schweizer et al., 1988). The compartment is believed (Pelham, 1989) to trap soluble endoplasmic reticulum proteins that contain a KDEL carboxy-terminus. In this model, KDEL receptors residing in the salvage compartment bind proteins containing a terminal KDEL sequence, such as BP, and recycle them back to the endoplasmic reticulum, where they dissociate. A second function may be to concentrate newly synthesized proteins, since both membrane and soluble proteins are at a higher concentration in the Golgi cisternae than in the endoplasmic reticulum. An unexpected finding is that flow between the salvage compartment and the endoplasmic reticulum is along
Figure
1. A Map of Major Membrane
Trafficking
Routes
within the Cell
Newly synthesized proteins leave the endoplasmic reticulum at smooth surface specializations and are first targeted to a salvage or intermediate compartment when they sorted. After carbohydrate modification in the Golgi ciaternae, they are sorted again either to secretory vesicles, or to a prelysosomal compartment. Protein taken up from the cell surface undergoes a peripheral sorting event at the early endosome. In some cases, receptors are recycled and ligands are carried in a large transport vesicle back to the prelysosomal compartment. Lysosomal enzymes from the Golgi mix with endocytosed material to form the lysosome. The three steps that have been demonstrated to use microtubule pathways are indicated by hatched arrows.
Cell 6
microtubules. This is most simply shown by experiments (Lippincott-Schwartz et al., 1990) in which a temperature block results in accumulation of a unique 53 kd antigen in the salvage compartment. At 16% flow back to the endoplasmic reticulum is inhibited, while flow from the endoplasmic reticulum to the salvage compartment continues. At 16% the antigen is clearly seen to be leaving the salvage compartment along “necklace” structures, identical to those seen in more complicated experiments in which brefeldin A is used to fuse the salvage compartment to the Golgi cisterna. Electron microscopy shows the ‘hecklaces” to consist of vesicles and tubules that are being transported along microtubules. If the temperature block is released, a synchronized wave of the 53 kd protein leaves the salvage compartment and can be transiently detected in the endoplasmic reticulum. Treatment of ceils with microtubule depolymerizing drugs inhibits the flow of the 53 kd protein from salvage compartment to endoplasmic reticulum, but not from endoplasmic reticulum to Golgi. The pathways that may not involve microtubules (Figure 1) are short in length. Microtubules do not appear to be needed for transport of newly synthesized protein between salvage compartment, Golgi cisternae, or the transGolgi network. Transport between cell surface and the early endosome, a peripheral sorting compartment, is also shortrange and does not require microtubules (Gruenberg et al., 1989). In addition to transporting vesicles, microtubules help maintain the architecture of the cell. Suitably stained endoplasmic reticulum can be visualized in living cells (Lee and Chen, 1988) to extend and retract processes, forming and dissolving a tubular network. A kinesin-like motor moving to the plus ends of microtubules could generate tension on the endoplasmic reticulum, pulling it to the cell periphery (Figure 2). Depolymerization of the microtubules collapses the endoplasmic reticulum membranes into the nuclear region. In contrast, the same treatment causes fragmentation and dispersal of the Golgi apparatus from the nuclear region. When depolymerizing agents are removed, the Golgi fragments reassemble in a juxtanuclear position. Such movement along microtubules can be recorded in living cells using fluorescent ceramide analogs as Golgi membrane markers (Ho et al., 1989). Since in many cells microtubules radiate out from a juxtanuclear microtubule organizing center, an obvious mechanism for reassembling the Golgi is to transport the fragments to the minus ends of microtubules, using a dynein-like motor. A recent set of experiments (Cooper et al., 1990) has suggested that the Golgi membranes, like endoplasmic reticulum membranes, are dynamic in living cells (Lee and Chen, 1988), and that the movements are microtubule associated. Once again a fluorescent ceramide analog was used to preferentially label Golgi membranes. Processes were seen that extended from Golgi regions to approach and perhaps fuse with other Golgi regions. In addition, small vesicles were seen leaving the Golgi region and whizzing to the cell periphery. Both movements appear to be microtubule associated.
Figure 2. Movement along Microtubules Is Crucial for Cellular Architecture as well as Membrane Traffic Microtubules radiate from a microtubule organizing center with plus ends outward. The endoplasmic reticulum is stretched between the nuclear (N) membrane and the plus ends of microtubules. This pathway also removes recycling membrane from the salvage compartment (S). The Golgi stacks and the salvage compartment are held in position by movement toward the minus ends of microtubules; this keeps elements of the central processing compartment near the nucleus. Vesicles also move to and from the cell surface along microtubules. Plusend motors are fill-in circles, and minus-end motors, squares.
How can we relate these impressive images of fluorescent membranes with the structures seen in the electron microscope? Three levels of organization contribute to Golgi structure. The primary or electron microscopic level is, of course, the stack of Golgi cisternae with associated trans-Golgi network. Since a cisterna is about 1 pm across and a stack of cisternae less than 0.5 pm, it is impossible to distinguish stacks from each other or from the transGolgi network by light microscopy. The secondary level of organization, seen in the light microscope, is due to the assembly of many Golgi stacks, often into branched ribbon-like structures (Cooper et al., 1990). The conventional Golgi pictures are sections through the ribbon. The third and final level is the localization of the ribbons or assemblies of stacks to a juxtanuclear location. When microtubules are depolymerized, the primary level of organization is undisturbed but the secondary association of stacks is dissolved, and the stacks themselves can diffuse away from a juxtanuclear region (Ho et al., 1969). What generates these ribbon-like structures, what holds them near the nucleus and how microtubules are involved is a mystery. Thus, in the observations of Cooper et al. (1990), we cannot be sure if we are watching membrane leave the Golgi stacks or dynamic changes in the poorly understood ribbon structures. The data reviewed here suggest how the roles of microtubules in organizing the cytoplasm and organizing membrane traffic might be combined in a unified model. The endoplasmic reticulum is an extension of the nuclear
yinireview
membrane and may owe its uniform distribution throughout the cytoplasm to interaction with microtubules, SpeCifitally to kinesin-driven movement along microtubules anchored at their minus ends in a juxtanuclear microtubule organizing center (Figure 2). The necklaces observed by Lippincott-Schwartz et al. (1990) are thought to be generated by membrane recycling from the salvage compartment back into the endoplasmic reticulum. One interpretation of their data is that membrane recycling uses the same microtubule pathways that spread the endoplasmic reticulum through the cytoplasm (Figure 2). Note that microtubules are not needed to move from endoplasmic reticulum to the Golgi compartment since this is a shortrange transition. The newly synthesized proteins leave the rough endoplasmic reticulum in smooth-membrane transitional zones that are in close proximity to the Golgi complex. Unlike the endoplasmic reticulum, which is diffusely distributed in the cytoplasm, the elements of what might be called a central sorting compartment, comprising the salvage compartment, the Golgi cisternae, the trsns-Golgi network, and the prelysosomal compartment, are localized to a juxtanuclear region. Dynein-like motors moving to the minus ends of microtubules (Figure 2) probably play a role in clustering Golgi membranes (Ho et al., 1989) and salvage compartments (Lippincott-Schwartz et al., 1990). When the separate Golgi elements are brought into proximity, they fuse to form an extended reticulum-a level of organization that is presumably useful but has no obvious function at present. Although we know how microtubules might organize the cytoplasm and how they might guide membrane traffic over the long distances resulting from that organization, it is not obvious why cells centralize their sorting compartments. Since the endoplasmic reticulum is extended throughout the cytoplasm, one could well expect the other organelles to be similarly diffuse. One advantage of first centralizing newly synthesized proteins in the Golgi region and the delivering them to the periphery along microtubules (Figure 2) is to allow selective delivery of newly made material to specialized regions of the cell surface. Such a capacity is crucial in migrating cells or polarized cells. It is not yet obvious why endocytotic transport vesicles move along microtubules. Degradation in the lysosome% a relatively slow event, is unlikely to be the reason. Perhaps some membrane components in the exocytotic vesicles need to be recycled quickly to the central sorting compartment. Although several features of cell structure can be accounted for with microtubules radiating from a juxtanuclear microtubule organizing center (Figure 2), the story cannot be considered complete, especially for differentiated cells. In myotubes, for example, the material of the microtubule organizing center is not next to the nucleus but instead surrounds it, and disruption of the microtubules does not disrupt Golgi organization. In epithelial cells, the microtubule organizing center changes location when cell-cell contact is made. After contact is made, it is found associated with the apical surface, and
the major microtubule organization is not radial but consists of parallel microtubules running from apical to basolateral surface (Bacallao et al., 1989). In such cells apical delivery to the cell surface might involve transport to the minus ends of microtubules (Achier et al., 1989). These differences may best be seen, however, as adaptations of the basic oellular ground plan shown in Figure 2. References Achier, C., Filmer, D., Merte, C., and Drenckhahn, 109, 179-189.
D. (1989). J. Cell Biol.
Bacallao, Ft., Antony, C., Dotti, C., Karsenti, E., Stelzer, Simon& K. (1989). J. Cell Biol. 109, 2817-2832. Cooper, M. S., Cornell-Bell, A. l-i., Chernjavsky, Smith, S. J. (1990). Cell 81, this issue. Gruenberg, 1301-1318.
J., Griffith%
G., and Howell,
Lee, C., and Chen,
L. B. (1988).
A., Dani, J. W., and
K. E. (1989). J. Cell Biol. 708,
Ho, W. C., Allan, V. J., van Meer, G., Berger, (1989). Eur. J. Cell Biol. 48, 250-263. Kreis, T. E., Matteoni, R., Hollinshead, J. Cell Biol. 49, 128-139.
E. Ii. K., and
E. G., and Kreis, T. E.
M., and Tooze, J. (1989).
Eur.
Cell 54, 37-46.
Lippincott-Schwartz, J., Donaldson, J. G., Schweizer, A., Berger, E. G., Hauri, H.-P., Yuan, L. C., and Klausner, R. D. (1990). Cell 60, 821-838. Pelham,
H. R. 8. (1989). Annu.
Saraste, J.,
Rev. Cell Biol. 5, l-23.
Palade, G. E., and Farquhar,
M. G. (1987). J. Cell Biol. 105,
2021-2030. Schweizer, A., Fransen, J. A. M., BBchi, T., Ginsel, (1988). J. Cell Biol. 107, 1643-1653.
L., and Hauri, H.-P
April 6, 1990, Copyright
0 1990 by Cell Press
Minireview
Microtubules, Membrane Traffic, and Cell Organization Regis B. Kelly Department of Biochemistry and Biophysics University of California San Francisco, California 94143
Eukaryotic cells frequently disperse their endoplasmic reticulum throughout their cytoplasm and centralize their Golgi membranes in the vicinity of the nucleus, an organization that seems to be generated and maintained by microtubules. Because of the juxtanuclear position of the Golgi membranes, secretory proteins newly synthesized in the endoplasmic reticulum are first centralized and then dispersed to the cell surface. Recent experiments have shown that microtubules seem to be involved in some but not all of the membrane trafficking steps within the cell (Figure 1). The data suggest that only centralizing and dispersal steps involve microtubules and indicate a relationship between microtubule-based membrane traffic and cellular organization. In the biosynthetic pathway, membrane vesicles can leave the Golgi region carrying newly synthesized proteins to the cell surface along microtubule tracks. If microtubules are depolymerized, proteins are still secreted, immovement is not essential plying that microtubule-based for exocytosis- nor indeed is cellular organization since that is also disrupted by microtubule depolymerization. What is changed is that membrane traffic becomes much less directed. Newly synthesized protein is no longer targeted to the growing edge of fibroblasts, for example. When microtubules are disrupted in epithelial cells, delivery to the apical surface, but not the basolateral, is perturbed in most cases studied. Neurons use microtubules for axonal transport. When endocrine cells develop long processes, movement of secretory granules to the process tips clearly involves microtubules (Kreis et al., 1989). In the secretory pathway, then, movement to the cell surface along microtubules is not essential, but is used to deliver newly synthesized material to defined surface locations. Selective targeting of post-Golgi vesicles to a region of the cell surface-may require the selective laying down or stabilizing of microtubule pathways from the Pans-Golgi network to specific regions of the cell surface. Microtubules are also involved in a late step in endocytosis. Newly endocytosed material is delivered to the early endosome (Figure 1) where material to be returned to the cell surface is sorted from material to be centralized. Early endosomes accumulate endocytotic material for about 10 min and then generate transport vesicles which will be taken to the cell center. Electron microscopy (Gruenberg et al., 1989) identified the endocytotic transport vesicle and showed it to be large (0.5 frrn diameter) and to be targeted to the prelysosomal compartment (Figure 1). Gruenberg et al. (1989) provide strong evidence that their large endocytotic transport vesicle is indeed the vesicle involved in microtubule-based movement to the cell center. Depolymerizing microtubules with nocodazole prevents
the transport vesicle from fusing with prelysosomal compartments and protects the contents of the transport vesicle from lysosomal degradation. What remains a puzzle, however, is why the cell takes endocytosed material to the nuclear region first, before degrading it. Why doesn’t degradation take place at the periphery? An attractive speculation is that a major function of the transport vesicle (Figure 1) is to recycle secretory vesicle membrane to the tran.sQolgi network in addition to its role in delivering contents to the lysosome. The most recently identified pathway of microtubulebased vesicle movement involves the salvage (Pelham, 1989) or intermediate (Lippincott-Schwartz et al., 1990) compartment between endoplasmic reticulum and Golgi complex. The salvage compartment (Figure 1) is identified in the electron microscope as the tubules and vesicles that accumulate in cells maintained at 15%, and in light microscopy by specific antibodies (Saraste et al., 1987; Schweizer et al., 1988). The compartment is believed (Pelham, 1989) to trap soluble endoplasmic reticulum proteins that contain a KDEL carboxy-terminus. In this model, KDEL receptors residing in the salvage compartment bind proteins containing a terminal KDEL sequence, such as BP, and recycle them back to the endoplasmic reticulum, where they dissociate. A second function may be to concentrate newly synthesized proteins, since both membrane and soluble proteins are at a higher concentration in the Golgi cisternae than in the endoplasmic reticulum. An unexpected finding is that flow between the salvage compartment and the endoplasmic reticulum is along
Figure
1. A Map of Major Membrane
Trafficking
Routes
within the Cell
Newly synthesized proteins leave the endoplasmic reticulum at smooth surface specializations and are first targeted to a salvage or intermediate compartment when they sorted. After carbohydrate modification in the Golgi ciaternae, they are sorted again either to secretory vesicles, or to a prelysosomal compartment. Protein taken up from the cell surface undergoes a peripheral sorting event at the early endosome. In some cases, receptors are recycled and ligands are carried in a large transport vesicle back to the prelysosomal compartment. Lysosomal enzymes from the Golgi mix with endocytosed material to form the lysosome. The three steps that have been demonstrated to use microtubule pathways are indicated by hatched arrows.
Cell 6
microtubules. This is most simply shown by experiments (Lippincott-Schwartz et al., 1990) in which a temperature block results in accumulation of a unique 53 kd antigen in the salvage compartment. At 16% flow back to the endoplasmic reticulum is inhibited, while flow from the endoplasmic reticulum to the salvage compartment continues. At 16% the antigen is clearly seen to be leaving the salvage compartment along “necklace” structures, identical to those seen in more complicated experiments in which brefeldin A is used to fuse the salvage compartment to the Golgi cisterna. Electron microscopy shows the ‘hecklaces” to consist of vesicles and tubules that are being transported along microtubules. If the temperature block is released, a synchronized wave of the 53 kd protein leaves the salvage compartment and can be transiently detected in the endoplasmic reticulum. Treatment of ceils with microtubule depolymerizing drugs inhibits the flow of the 53 kd protein from salvage compartment to endoplasmic reticulum, but not from endoplasmic reticulum to Golgi. The pathways that may not involve microtubules (Figure 1) are short in length. Microtubules do not appear to be needed for transport of newly synthesized protein between salvage compartment, Golgi cisternae, or the transGolgi network. Transport between cell surface and the early endosome, a peripheral sorting compartment, is also shortrange and does not require microtubules (Gruenberg et al., 1989). In addition to transporting vesicles, microtubules help maintain the architecture of the cell. Suitably stained endoplasmic reticulum can be visualized in living cells (Lee and Chen, 1988) to extend and retract processes, forming and dissolving a tubular network. A kinesin-like motor moving to the plus ends of microtubules could generate tension on the endoplasmic reticulum, pulling it to the cell periphery (Figure 2). Depolymerization of the microtubules collapses the endoplasmic reticulum membranes into the nuclear region. In contrast, the same treatment causes fragmentation and dispersal of the Golgi apparatus from the nuclear region. When depolymerizing agents are removed, the Golgi fragments reassemble in a juxtanuclear position. Such movement along microtubules can be recorded in living cells using fluorescent ceramide analogs as Golgi membrane markers (Ho et al., 1989). Since in many cells microtubules radiate out from a juxtanuclear microtubule organizing center, an obvious mechanism for reassembling the Golgi is to transport the fragments to the minus ends of microtubules, using a dynein-like motor. A recent set of experiments (Cooper et al., 1990) has suggested that the Golgi membranes, like endoplasmic reticulum membranes, are dynamic in living cells (Lee and Chen, 1988), and that the movements are microtubule associated. Once again a fluorescent ceramide analog was used to preferentially label Golgi membranes. Processes were seen that extended from Golgi regions to approach and perhaps fuse with other Golgi regions. In addition, small vesicles were seen leaving the Golgi region and whizzing to the cell periphery. Both movements appear to be microtubule associated.
Figure 2. Movement along Microtubules Is Crucial for Cellular Architecture as well as Membrane Traffic Microtubules radiate from a microtubule organizing center with plus ends outward. The endoplasmic reticulum is stretched between the nuclear (N) membrane and the plus ends of microtubules. This pathway also removes recycling membrane from the salvage compartment (S). The Golgi stacks and the salvage compartment are held in position by movement toward the minus ends of microtubules; this keeps elements of the central processing compartment near the nucleus. Vesicles also move to and from the cell surface along microtubules. Plusend motors are fill-in circles, and minus-end motors, squares.
How can we relate these impressive images of fluorescent membranes with the structures seen in the electron microscope? Three levels of organization contribute to Golgi structure. The primary or electron microscopic level is, of course, the stack of Golgi cisternae with associated trans-Golgi network. Since a cisterna is about 1 pm across and a stack of cisternae less than 0.5 pm, it is impossible to distinguish stacks from each other or from the transGolgi network by light microscopy. The secondary level of organization, seen in the light microscope, is due to the assembly of many Golgi stacks, often into branched ribbon-like structures (Cooper et al., 1990). The conventional Golgi pictures are sections through the ribbon. The third and final level is the localization of the ribbons or assemblies of stacks to a juxtanuclear location. When microtubules are depolymerized, the primary level of organization is undisturbed but the secondary association of stacks is dissolved, and the stacks themselves can diffuse away from a juxtanuclear region (Ho et al., 1969). What generates these ribbon-like structures, what holds them near the nucleus and how microtubules are involved is a mystery. Thus, in the observations of Cooper et al. (1990), we cannot be sure if we are watching membrane leave the Golgi stacks or dynamic changes in the poorly understood ribbon structures. The data reviewed here suggest how the roles of microtubules in organizing the cytoplasm and organizing membrane traffic might be combined in a unified model. The endoplasmic reticulum is an extension of the nuclear
yinireview
membrane and may owe its uniform distribution throughout the cytoplasm to interaction with microtubules, SpeCifitally to kinesin-driven movement along microtubules anchored at their minus ends in a juxtanuclear microtubule organizing center (Figure 2). The necklaces observed by Lippincott-Schwartz et al. (1990) are thought to be generated by membrane recycling from the salvage compartment back into the endoplasmic reticulum. One interpretation of their data is that membrane recycling uses the same microtubule pathways that spread the endoplasmic reticulum through the cytoplasm (Figure 2). Note that microtubules are not needed to move from endoplasmic reticulum to the Golgi compartment since this is a shortrange transition. The newly synthesized proteins leave the rough endoplasmic reticulum in smooth-membrane transitional zones that are in close proximity to the Golgi complex. Unlike the endoplasmic reticulum, which is diffusely distributed in the cytoplasm, the elements of what might be called a central sorting compartment, comprising the salvage compartment, the Golgi cisternae, the trsns-Golgi network, and the prelysosomal compartment, are localized to a juxtanuclear region. Dynein-like motors moving to the minus ends of microtubules (Figure 2) probably play a role in clustering Golgi membranes (Ho et al., 1989) and salvage compartments (Lippincott-Schwartz et al., 1990). When the separate Golgi elements are brought into proximity, they fuse to form an extended reticulum-a level of organization that is presumably useful but has no obvious function at present. Although we know how microtubules might organize the cytoplasm and how they might guide membrane traffic over the long distances resulting from that organization, it is not obvious why cells centralize their sorting compartments. Since the endoplasmic reticulum is extended throughout the cytoplasm, one could well expect the other organelles to be similarly diffuse. One advantage of first centralizing newly synthesized proteins in the Golgi region and the delivering them to the periphery along microtubules (Figure 2) is to allow selective delivery of newly made material to specialized regions of the cell surface. Such a capacity is crucial in migrating cells or polarized cells. It is not yet obvious why endocytotic transport vesicles move along microtubules. Degradation in the lysosome% a relatively slow event, is unlikely to be the reason. Perhaps some membrane components in the exocytotic vesicles need to be recycled quickly to the central sorting compartment. Although several features of cell structure can be accounted for with microtubules radiating from a juxtanuclear microtubule organizing center (Figure 2), the story cannot be considered complete, especially for differentiated cells. In myotubes, for example, the material of the microtubule organizing center is not next to the nucleus but instead surrounds it, and disruption of the microtubules does not disrupt Golgi organization. In epithelial cells, the microtubule organizing center changes location when cell-cell contact is made. After contact is made, it is found associated with the apical surface, and
the major microtubule organization is not radial but consists of parallel microtubules running from apical to basolateral surface (Bacallao et al., 1989). In such cells apical delivery to the cell surface might involve transport to the minus ends of microtubules (Achier et al., 1989). These differences may best be seen, however, as adaptations of the basic oellular ground plan shown in Figure 2. References Achier, C., Filmer, D., Merte, C., and Drenckhahn, 109, 179-189.
D. (1989). J. Cell Biol.
Bacallao, Ft., Antony, C., Dotti, C., Karsenti, E., Stelzer, Simon& K. (1989). J. Cell Biol. 109, 2817-2832. Cooper, M. S., Cornell-Bell, A. l-i., Chernjavsky, Smith, S. J. (1990). Cell 81, this issue. Gruenberg, 1301-1318.
J., Griffith%
G., and Howell,
Lee, C., and Chen,
L. B. (1988).
A., Dani, J. W., and
K. E. (1989). J. Cell Biol. 708,
Ho, W. C., Allan, V. J., van Meer, G., Berger, (1989). Eur. J. Cell Biol. 48, 250-263. Kreis, T. E., Matteoni, R., Hollinshead, J. Cell Biol. 49, 128-139.
E. Ii. K., and
E. G., and Kreis, T. E.
M., and Tooze, J. (1989).
Eur.
Cell 54, 37-46.
Lippincott-Schwartz, J., Donaldson, J. G., Schweizer, A., Berger, E. G., Hauri, H.-P., Yuan, L. C., and Klausner, R. D. (1990). Cell 60, 821-838. Pelham,
H. R. 8. (1989). Annu.
Saraste, J.,
Rev. Cell Biol. 5, l-23.
Palade, G. E., and Farquhar,
M. G. (1987). J. Cell Biol. 105,
2021-2030. Schweizer, A., Fransen, J. A. M., BBchi, T., Ginsel, (1988). J. Cell Biol. 107, 1643-1653.
L., and Hauri, H.-P
