SEARCH

SEARCH BY CITATION

Keywords:

  • CLIMP-63;
  • endoplasmic reticulum;
  • huntingtin;
  • kinectin;
  • membrane fusion;
  • microtubules;
  • p22;
  • p97;
  • VAP-B

Abstract

  1. Top of page
  2. Abstract
  3. Endoplasmic Reticulum – Microtubule Connection
  4. Stable Attachment to MTs is Essential for Maintaining the ER Network
  5. Maintenance of the ER Network Requires Homotypic Membrane Fusion
  6. Newcomers Proposed to Mediate ER–MT Interaction
  7. Concluding Remarks
  8. NOTE ADDED IN PROOF
  9. Acknowledgments
  10. References

The endoplasmic reticulum (ER) of higher eukaryotic cells is a dynamic network of interconnected membrane tubules that pervades almost the entire cytoplasm. On the basis of the morphological changes induced by the disruption of the cytoskeleton or molecular motor proteins, the commonly accepted model has emerged that microtubules and conventional kinesin (kinesin-1) are essential determinants in establishing and maintaining the structure of the ER by active membrane expansion. Surprisingly, very similar ER phenotypes have now been observed when the cytoskeleton-linking ER membrane protein of 63 kDa (CLIMP-63) is mutated, revealing stable attachment of ER membranes to the microtubular cytoskeleton as a novel requirement for ER maintenance. Additional recent findings suggest that ER maintenance also requires ongoing homotypic membrane fusion, possibly controlled by the p97/p47/VICP135 protein complex. Work on other proteins proposed to regulate ER structure, including huntingtin, the EF-hand Ca2+-binding protein p22, the vesicle-associated membrane protein-associated protein B and kinectin isoforms further contribute to the new emerging concept that ER shape is not only determined by motor driven processes but by a variety of different mechanisms.


Endoplasmic Reticulum – Microtubule Connection

  1. Top of page
  2. Abstract
  3. Endoplasmic Reticulum – Microtubule Connection
  4. Stable Attachment to MTs is Essential for Maintaining the ER Network
  5. Maintenance of the ER Network Requires Homotypic Membrane Fusion
  6. Newcomers Proposed to Mediate ER–MT Interaction
  7. Concluding Remarks
  8. NOTE ADDED IN PROOF
  9. Acknowledgments
  10. References

The endoplasmic reticulum (ER) is a membrane-bound organelle present in all eukaryotic cells. It synthesizes lipids as well as secretory and membrane proteins, and ensures subsequent post-translational processing and quality control of the proteins (1–3). In addition, the ER is a multifunctional signaling organelle that controls entry and release of calcium, sterol biosynthesis, apoptosis and the release of arachidonic acid (4). The ER of higher eukaryotes comprises three main subdomains: the nuclear envelope, the ribosome-bound rough ER and the ribosome-free smooth ER. Both the absolute and relative abundance of rough and smooth ER varies with the cell type. ER membranes form flattened sheets, cisternae and tubules. In well-spread areas of the cell periphery, these tubules can be seen using light microscopy to form an irregular polygonal network with characteristic three-way junctions (5–7). Despite its complex organization, the ER is a continuous membrane compartment (8). The function of the cytoskeleton in ER architecture has been investigated in numerous studies, which revealed that microtubules (MTs) and ER are highly interdependent structures. Accordingly, the ER network retracts to the cell center in a number of experimental conditions that affect MTs or MT-associated proteins. These conditions include the depolymerization of the MT cytoskeleton by drugs or low temperature (9), the suppression of the heavy chain of the MT plus-end-directed motor protein kinesin (10) and the overexpression of the MT-associated protein tau that probably impedes kinesin-mediated transport (11). In addition, the extension of ER membranes along MTs has been imaged using video microscopy both in vivo (12) and in cell-free systems, where this process can be inhibited by antibodies against kinesin heavy chain (13,14). In the past few months, various nonmotor proteins have been proposed to regulate the structure of the ER in higher eukaryotes. Although some of these findings will require additional analysis, they are interesting leads toward a more complete understanding of the determinants controlling ER shape. Here, we discuss these novel candidates and their putative role in ER morphogenesis.

Stable Attachment to MTs is Essential for Maintaining the ER Network

  1. Top of page
  2. Abstract
  3. Endoplasmic Reticulum – Microtubule Connection
  4. Stable Attachment to MTs is Essential for Maintaining the ER Network
  5. Maintenance of the ER Network Requires Homotypic Membrane Fusion
  6. Newcomers Proposed to Mediate ER–MT Interaction
  7. Concluding Remarks
  8. NOTE ADDED IN PROOF
  9. Acknowledgments
  10. References

Recent studies on the cytoskeleton linking membrane protein of 63 kDa (CLIMP-63) have revealed a novel type of interaction between ER and MTs. CLIMP-63 is a type II ER membrane protein that was identified in a search for organelle-specific markers of the early secretory pathway (15,16). Rather unique among ER proteins, CLIMP-63 is excluded from the outer nuclear membrane by forming large immobile oligomers in the reticular ER (15,17). Purified CLIMP-63 binds to MTs, and when overexpressed, CLIMP-63 rearranges the ER along bundled MTs, suggesting that the protein functions in anchoring ER membranes to MTs (18). The significance of such an interaction became apparent when the MT-binding function of CLIMP-63 was abrogated. Cells that overexpress CLIMP-63 mutants impaired in MT binding exhibit a poorly extended ER without changes of the MT cytoskeleton (Figure 1) (19). These CLIMP-63-dependent ER rearrangements can be explained by the MT-binding properties of CLIMP-63: overexpression of wild-type CLIMP-63 increases the number of interacting points between ER membranes and MTs and therefore induces their co-alignment (Figure 1B), whereas overexpression of mutated CLIMP-63 that can no longer bind to MTs disturbs the anchoring of the ER to MTs (Figure 1C). The results suggest that CLIMP-63 is required to stabilize the extended ER network. Such a role is in line with several additional observations. CLIMP-63 or equivalent proteins do not exist in yeast, in which the maintenance of the ER relies on actin filaments rather than MTs (20). In higher eukaryotes, CLIMP-63 is ubiquitously expressed and specifically localizes to the reticular domain of the ER (15). Moreover, the binding of CLIMP-63 to MTs is regulated in a cell cycle-dependent manner. Binding is inhibited during mitosis (19), that is, in a cell stage where the ER network is less developed and in part switches from a MT- to an actin-based organization (21). Interestingly, the collapse of ER membranes induced by MT-binding-impaired CLIMP-63 mutants is strikingly similar to that induced by depolymerizing MTs or inhibiting MT-based motor activity (9–11). Thus, when the ER forms by active extension of membranes along MTs, CLIMP-63 may act in synergy with kinesin motors to stabilize ER tubules that would otherwise retract to the cell center due to membrane tension generated by the extensive tubule formation (22,23). CLIMP-63 thus appears as a key structural protein of the ER. Its characterization revealed that stable anchoring of the ER to MTs is essential to maintain the typical structure of this organelle.

image

Figure 1. Role of CLIMP-63 in maintaining ER structure. Localization of endogenous or transfected CLIMP-63 (green) and MTs (red) in COS cells using immunofluorescence microscopy (left panels). Corresponding schematic interpretations of the phenotypes are illustrated in the right panels. (A) Non-transfected cells. The ER appears as extended network superimposed to microtubules. The extension of ER membranes is powered by MT-based motors (yellow triangles). The ER network is proposed to be stabilized by CLIMP-63-mediated anchoring of ER membranes to MTs (black circles). (B) Wild-type CLIMP-63-transfected cells. In cells overexpressing CLIMP-63, ER membranes and MTs are bundled and form rope-like structures. This rearrangement is the result of an increased number of contact sites between the ER and MTs due to CLIMP-63 overexpression. (C) Cells overexpressing a mutant form of CLIMP-63 impaired for MT binding. Mutated CLIMP-63 that is unable to bind to MTs (19) displaces the endogenous protein from MTs. As a result, newly formed ER tubules are not stable enough to elongate, and the ER concentrates in the center of the cell. Scale bar, 10 µm.

Maintenance of the ER Network Requires Homotypic Membrane Fusion

  1. Top of page
  2. Abstract
  3. Endoplasmic Reticulum – Microtubule Connection
  4. Stable Attachment to MTs is Essential for Maintaining the ER Network
  5. Maintenance of the ER Network Requires Homotypic Membrane Fusion
  6. Newcomers Proposed to Mediate ER–MT Interaction
  7. Concluding Remarks
  8. NOTE ADDED IN PROOF
  9. Acknowledgments
  10. References

ER dynamics have been studied extensively both in living cells and in cell-free systems (5,12–14,22,24). Irrespective of whether de novo forming or steady-state networks are investigated, numerous homotypic membrane fusion events can frequently be observed that continuously lead to new polygon formation within the ER. Thus, membrane fusion may be required not only for ER biogenesis but also for ER maintenance. Two recent studies support this notion and provide important new information on the molecular mechanism underlying the fusion of ER membranes. In cytosol exchange experiments with streptolysin O-permeabilized semi-intact CHO cells, N-ethylmaleimide-treated interphase cytosol decreases the number of three-way junctions within the ER network, suggesting that membrane fusion activity is required for maintaining the ER network (25,26). Likewise, the ER is disrupted by the addition of mitotic cytosol in conjunction with the MT-depolymerizing drug nocodazole but can be reformed by adding interphase cytosol. Reformation of the ER network is proposed to be controlled by two sequential fusion reactions. The first process, mediated by NSF/α and γ-SNAP, would lead to vesicle aggregates and the formation of thin membrane tubules connecting the disrupted ER elements. The subsequent fusion process that leads to complete network formation (Figure 2C) would be mediated by the p97/p47/VCIP135 protein complex. Both reactions also require the t-SNARE syntaxin 18. Noteworthy, a knockdown of the syntaxin 18-associated protein BNIP1 or its inactivation by antibodies leads to a disintegration of the reticular ER, further supporting the importance of homotypic membrane fusion in ER maintenance (27).

image

Figure 2. Mechanisms proposed to be involved in ER formation. ER membranes are actively pulled along MTs by microtubule plus end-directed kinesin-type motors (a), and the resulting membrane extensions are stabilized by the ER membrane protein CLIMP-63 (b) and possibly huntingtin (not shown). If ER membranes get sufficiently close to each other, they fuse due to the combined action of SNAREs and peripheral fusion machineries (c). Fusion may be further facilitated by the MT-associated protein p22 that binds to the ER after a calcium-induced conformational change. Extension of ER membranes can be inhibited by blocking kinesin with neutralizing antibodies (b′) or by knocking down kinesin receptors, such as kinectin (not shown). ER extensions can also be inhibited by interfering with the MT-anchoring function of CLIMP-63 (see Figure 1). In all these cases, the formation of three-way junctions is no longer possible. (c′) Microinjection of antibodies against p47 or VIC135 greatly decreases the number of three-way junctions, suggesting that the p97/47 complex, in conjunction with syntaxin 18 (not shown), is required for the fusion of ER membranes.

These results suggest that membrane fusion events permanently remodel the ER network during interphase and that this process is inhibited during mitosis. Inversely, fusion processes dedicated to extend the reticulum need to be reactivated after mitosis. In support of such a mechanism, antibodies against VCIP135 or p47 microinjected into prometaphase cells severely impede ER network restructuring when cells exit mitosis while cell division is unaffected (28). Similarly, reformation of the nuclear envelope after mitosis also involves function of the p97 protein complex in two discrete steps (29). Reversible down-regulation of ER fusion is most likely controlled by phosphorylation as indicated by the inhibition of mitotic ER disruption by non-phosphorylated p47 (25). The p97/p47/VCIP135 complex also controls Golgi reassembly (28). Selective recruitment of p97/p47/VCIP135 to the Golgi involves the SNARE protein syntaxin 5, as opposed to its recruitment to the ER that appears to involve syntaxin 18.

Collectively, these observations point to an important role of the p97 fusion machinery complex in maintaining the ER. Currently, the involvement of p97 in this process is still indirect. It primarily relies on antibody injections that are difficult to control and the use of recombinant mutant p47. Irrespective of the required fusion machinery, there is little doubt that homotypic membrane fusion is continually required for maintaining the typical structure of the ER.

Newcomers Proposed to Mediate ER–MT Interaction

  1. Top of page
  2. Abstract
  3. Endoplasmic Reticulum – Microtubule Connection
  4. Stable Attachment to MTs is Essential for Maintaining the ER Network
  5. Maintenance of the ER Network Requires Homotypic Membrane Fusion
  6. Newcomers Proposed to Mediate ER–MT Interaction
  7. Concluding Remarks
  8. NOTE ADDED IN PROOF
  9. Acknowledgments
  10. References

A number of interesting candidates for the control of ER morphogenesis have recently emerged from diverse fields of research. One candidate is the ubiquitous Huntington's disease protein huntingtin, of which a polyglutamine expansion leads to a dominant progressive neurodegenerative disease. The normal function of huntingtin is not entirely understood and might be multiple, as its exact subcellular localization is a matter of debate (see below). Numerous studies, however, suggest a role in intracellular membrane trafficking and axonal transport along MTs as well as signaling (30–32). Surprisingly, the silencing of the huntington's disease gene by siRNA has recently been found to induce an aberrant ER network in mouse neuroblastoma and human glioblastoma cells (33). A huntingtin knockdown in these cells results in a poorly extended ER network leaving MTs intact (33). Because other organelles are not affected and this alteration of ER shape is typical of the disruption of the interaction between the ER and MTs (9–11,19), a possible role of huntingtin in ER morphogenesis deserves consideration. Unlike CLIMP-63, huntingtin is a cytoplasmic protein mainly associated with MTs (34,35), but it is also proposed to bind to various membranes, including synaptic and clathrin-coated vesicles, recycling endosomes, ER, Golgi and plasma membrane. Therefore, a specific ER phenotype because of knocking down huntingtin came as a surprise. Huntingtin interacts with a multitude of proteins (30), although it remains to be shown if all the huntingtin-binding proteins identified by yeast two-hybrid and biochemical analyses also interact in vivo. Huntingtin binds β-tubulin (35) and is also found in a protein complex with dynactin, to which it binds via its neuronal partner huntingtin-associated protein (HAP1) (36), which provides dual opportunities for direct and motor–based interactions with MTs. At least three mutually nonexclusive mechanisms can be proposed for the interaction of huntingtin with ER membranes: (i) Huntingtin may directly bind to phospholipids of the ER bilayer through electrostatic interactions. Such a mechanism has been proposed for plasma membrane targeting of huntingtin that is mediated by amino acids 172–372 (37) and may also contribute to ER binding, although by itself would provide insufficient organelle specificity. (ii) Huntingtin may become anchored to the ER membrane by palmitoylation. The protein is indeed palmitoylated by and binds to the palmitoyltransferase HIP14 (38). Again, palmitoylation may contribute to ER localization but is not a modification for specifically targeting a protein to the ER. Moreover, because HIP14 is mainly localized to the Golgi and cytoplasmic vesicles (39), this protein may mediate the binding of huntingtin to these organelles rather than the ER. (iii) Huntingtin may bind to an integral membrane protein of the ER. An interesting and as yet the only candidate is the Ca2+ release channel inositol-trisphosphate receptor InsP3R1. Huntingtin binds to InsP3R1 in a ternary complex with HAP1 and influences neuronal calcium signaling mediated by InsP3R1 (40). It will be interesting to see whether a specific disturbance of the huntingtin/InsP3R1 interaction leads to a change of ER morphology. In a broader context, the possible involvement of huntingtin in ER maintenance may point to a role of altered ER structure in the development of Huntington's disease, in addition to the disruption of axonal transport and the formation of inclusions resulting from aggregation of the N-terminally polyglutamine-expanded huntingtin.

Another candidate morphogen of the ER is the EF-hand Ca2+-binding protein p22 (41). p22 is a cytosolic protein that binds to MTs in an N-myristoylation-dependent manner (42,43). In vitro, the protein mediates interaction between microsomal membranes and MTs. p22 binds poorly to microsomes in the absence of calcium, but binding is increased by a calcium-induced conformational change of the protein. In vivo, similar to huntingtin, p22 primarily localizes to MTs and only minimally to the ER. In agreement with the MT localization, microinjection of N-myristoylated p22 bundles MTs, whereas microinjection of p22 antibodies disrupts MTs. Interestingly, the injection of p22 antibodies also vesiculates ER membranes (41). This effect might be an indirect result of MT disruption. However, because the vesiculation differs significantly from the typical retraction of the ER observed after MT polymerization, it appears more likely that p22 has a dual function in MT stabilization and ER network organization. An attractive possibility is that p22 promotes homotypic membrane fusion. Such a mechanism is supported by two findings. First, p22 was initially characterized as a protein required for targeting and/or fusion of membranes (44). Second, microinjection of myristoylated wild-type p22, but not myristoylated mutant p22 (that is unable to undergo the calcium-induced conformational change), increases the number of three-way junctions within the ER. Such a role in reticulum formation would explain why the interaction between p22 and ER is too weak to be detected at steady state in vivo. ER membranes would interact only transiently with MT-associated p22 in places where Ca2+ signals allow for membrane binding (Figure 2c). A role of p22 in coupling the formation of the reticulum to the MT cytoskeleton would also be in accordance with the absence of this protein in yeast.

On the basis of drastic morphological changes induced by protein overexpression, additional candidate proteins have been suggested to control ER structure. The ER is an extraordinarily plastic organelle. Large scale proliferation and/or reorganization of ER membranes frequently occur when exogenous or ER resident proteins are expressed at high levels, due to as diverse mechanisms as membrane protein stockpiling, the accumulation of misfolded, nondegradable proteins leading to aggresomes and Russel bodies, or lumenal osmotic changes (1,45–49). Accordingly, one has to be extremely careful in proposing an ER morphogenetic function for a protein solely based on its overexpression. Recently, Amarilio and co-workers proposed an intriguing mechanism for controlling ER morphology involving the vesicle-associated membrane protein-associated protein (VAP)-B (50,51). VAPs are type II integral membrane proteins proposed to function in secretory pathway trafficking (52–54). In yeast, the VAP homologue Scs2p acts as an ER anchor for cytosolic proteins that carry a specific peptidic signal, the FFAT domain (55). Mammalian VAP-B binds to the FFAT-bearing Nir proteins, a family of soluble proteins involved in membrane traffic (56), and in cell culture, this interaction results in differential remodeling of ER structure (50). Co-expression of tagged VAP-B and Nir2 leads to the formation of stacked membrane arrays, and co-expression of VAP-B and Nir3 bundles ER membranes along MTs. While membrane stockpiling may be the result of tight apposition of membranes mediated by VAP-B/Nir2 complexes, Nir3 is proposed to either bridge VAP-B to MTs or increase the affinity of VAP-B for MTs (Figure 3D). These results require further studies to ascertain the proposed regulatory function of VAP and Nir proteins in controlling ER structure. For instance, it is unknown whether endogenous VAP-B and Nir proteins interact within the cell, and endogenous Nir proteins need to be localized to exclude the possibility that overexpression leads to a nonrelevant association with the ER. Despite these limitations, the VAP-B data lead to an interesting concept: the possibility of modulating ER morphology via a receptor that can bind different partners. Such a mechanism is also supported by recent data showing that overexpression of an FFAT-binding-defective mutant of the VAP isoform A disrupts the ER network (51).

image

Figure 3. ER morphogenesis is regulated by diverse molecules.(A) Kinesin-1 motor-mediated ER extension along MTs. Kinesin-1 binds to its ER receptor kinectin, depicted here as the 160-kDa membrane-anchored form. Kinectin is thought to form dimers via extensive coiled-coil oligomerization of the cytoplasmic domain. Isoforms of kinectin lacking the kinesin-binding domain are suggested to regulate kinesin attachment by forming heterodimers with the full length protein (see text). (B) Stable attachment of ER to MTs by linker molecules. Attachment can be mediated by the integral membrane protein CLIMP-63. Its MT binding is regulated by multiple phosphorylation of the cytosolic tail. CLIMP-63 forms high molecular weight multimers (n) via coiled–coil interactions of the lumenal domain. Huntingtin, an MT-associated protein may also link ER and MTs. (C) ER reticulation by transient interaction with MTs. The MT-associated calcium-binding protein p22 can link MTs to the ER upon a calcium-induced conformational change and thereby promotes ER reticulation. (D) Modulation of ER structure by ER receptors interacting with different linkers. The ER membrane protein VAP-B can bind to proteins bearing an FFAT domain that may modulate ER shape. One of these proteins is Nir 3, which promotes ER association with MTs.

Yet, another possible way of regulating ER morphogenesis may be drawn from recent data on the kinesin-binding protein kinectin. Kinectin is a membrane protein that was initially characterized as the main receptor for the motor protein kinesin. Kinectin constitutes the majority of kinesin-binding sites on membrane organelles and is required for kinesin-driven motility (57–60). Such a general function has been questioned, however, because kinectin was undetectable in axons until recently (see below), where organelles are predominantly transported by kinesin, and due to the finding that organelle transport is not impaired in kinectin knock-out mice (61,62). It is worth noting that a similar discrepancy is known for kinesin. Ablation of the ubiquitous conventional kinesin heavy chain gene in the mouse has no effect on ER morphology (63), while inhibition of kinesin in cultured cells blocks ER extension. Such discrepancies may be due to back-up systems operating in intact animals as opposed to cultured cells. An essential role for kinectin in kinesin-mediated extension of the ER became evident recently in cell culture experiments, where silencing of kinectin led to a collapse of the ER to the center of the cell (64). This phenotype is strikingly similar to that obtained by inhibiting kinesin. Such a phenotype can also be induced by expressing soluble fragments encompassing the domains in kinectin or kinesin that mediate their interaction. More unexpectedly, the authors report that kinectin has at least 16 splice isoforms in the mouse and that one isoform is indeed expressed in axons. Surprisingly, most isoforms lack an intact kinesin-binding domain. Although these isoforms have only been detected at the RNA level and data concerning their relative abundance are missing, these results raise the question of the function of kinectins lacking a kinesin-binding domain. Santama et al. suggest that such isoforms may act as ER receptors for partners other than kinesin. Another mutually non-exclusive possibility is that these kinectins regulate the attachment of kinesin to ER membranes. Kinectin is thought to form dimers, and expressing an artificial construct lacking a kinesin-binding domain also prevents ER extension, pointing to a dominant negative effect. In summary, kinectin appears essential for ER morphogenesis, most probably via anchoring the motor kinesin (Figure 3A), as presumed for a long time. The novel idea, to be tested in detail, is that different isoforms of kinectin may positively and negatively regulate the binding of kinesin to the ER and thereby contribute to the shape of the ER.

Concluding Remarks

  1. Top of page
  2. Abstract
  3. Endoplasmic Reticulum – Microtubule Connection
  4. Stable Attachment to MTs is Essential for Maintaining the ER Network
  5. Maintenance of the ER Network Requires Homotypic Membrane Fusion
  6. Newcomers Proposed to Mediate ER–MT Interaction
  7. Concluding Remarks
  8. NOTE ADDED IN PROOF
  9. Acknowledgments
  10. References

Studies on ER morphogenesis have traditionally focused on molecular motors and supporting cytoskeletal structures, among which MTs and associated motors play a major role. While the full complement of motors that contribute to this process is still being defined, recent studies suggest that the interaction between ER and MTs is mediated by a variety of mechanistically diverse molecules (Figure 3). In particular, it appears that stable linker proteins, such as CLIMP-63, are critical for ER maintenance (Figure 3B). In addition, because the ER retracts only slowly upon depolymerization of MTs (9) and can form in vitro in the absence of any cytoskeleton (24), additional morphogenic mechanisms must exist. As illustrated by p22 and VAP-B, it is likely that many players involved in the regulation of ER structure operate only transiently which renders their analysis difficult (Figure 3C,D). Clearly, novel methods need to be developed to study such interactions by reversibly immobilizing the corresponding partner proteins in vivo. In addition, because the structure of the ER is artificially modified by many proteins when overexpressed, complementary approaches are required to substantiate the presumed morphogenetic role of a candidate protein, such as its inhibition or knockdown. Novel proteins controlling ER morphogenesis will probably be identified in global knockdown approaches that screen for changes in ER morphology. An additional emerging determinant for ER maintenance is homotypic membrane fusion. Extrapolating from this fusion requirement, it can be anticipated that ER lipids are also important. Lipid acylation has been proposed to inhibit ER membrane fusion (65), leading to the notion that lipid-modifying enzymes may contribute to regulating ER structure.

NOTE ADDED IN PROOF

  1. Top of page
  2. Abstract
  3. Endoplasmic Reticulum – Microtubule Connection
  4. Stable Attachment to MTs is Essential for Maintaining the ER Network
  5. Maintenance of the ER Network Requires Homotypic Membrane Fusion
  6. Newcomers Proposed to Mediate ER–MT Interaction
  7. Concluding Remarks
  8. NOTE ADDED IN PROOF
  9. Acknowledgments
  10. References

An additional morphogenetic mechanism has recently been proposed by Voeltz, G.K. et al. (Cell 124, 573–586, 2006). The authors found that the two membrane proteins Rtn4a/NogoA and DP1/Yop1p are responsible for the generation of tubular morphology in the ER. They propose that these membrane proteins act as morphogens by partitioning into and stabilizing curved ER membrane tubules.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Endoplasmic Reticulum – Microtubule Connection
  4. Stable Attachment to MTs is Essential for Maintaining the ER Network
  5. Maintenance of the ER Network Requires Homotypic Membrane Fusion
  6. Newcomers Proposed to Mediate ER–MT Interaction
  7. Concluding Remarks
  8. NOTE ADDED IN PROOF
  9. Acknowledgments
  10. References

We thank Martin Spiess and members of the Hauri group for critical reading of the manuscript and helpful comments. Research in the author's laboratory is supported by the University of Basel and the Swiss National Science Foundation.

References

  1. Top of page
  2. Abstract
  3. Endoplasmic Reticulum – Microtubule Connection
  4. Stable Attachment to MTs is Essential for Maintaining the ER Network
  5. Maintenance of the ER Network Requires Homotypic Membrane Fusion
  6. Newcomers Proposed to Mediate ER–MT Interaction
  7. Concluding Remarks
  8. NOTE ADDED IN PROOF
  9. Acknowledgments
  10. References