JXB Advance Access originally published online on September 21, 2006
Journal of Experimental Botany 2007 58(1):49-64; doi:10.1093/jxb/erl135
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
RESEARCH PAPER |
The plant ER–Golgi interface: a highly structured and dynamic membrane complex
1Laboratoire de Biogenèse membranaire, UMR 5200 CNRS-Université de Bordeaux II, case 92, 146 rue Léo-Saignat, F-33076 Bordeaux-Cedex, France
2Department of Biology, 112 Science Place, University of Saskatchewan, Saskatoon, SK, Canada S7N 5E2
3Research School of Biological and Molecular Sciences, Oxford Brookes University, Oxford OX3 0BP, UK
4Laboratoire de Dynamique de la Compartimentation Cellulaire, Institut des Sciences du Végétal, CNRS UPR 2355, Avenue de la Terrasse, F-91400 Gif sur Yvette-Cedex, France
* To whom correspondence should be addressed. E-mail: pmoreau{at}biomemb.u-bordeaux2.fr
Received 20 January 2006; Accepted 26 July 2006
 |
Abstract |
|---|
 Top Abstract IntroductionSpecific molecular machineries...The SNARE machineryLipids, lipid-modifying enzymes,...Challenges for the futureReferences |
|---|
As compared with other eukaryotic cells, plants have developed an endoplasmic reticulum (ER)–Golgi interface with very specific structural characteristics. ER to Golgi and Golgi to ER transport appear not to be dependent on the cytoskeleton, and ER export sites have been found closely associated with Golgi bodies to constitute entire mobile units. However, the molecular machinery involved in membrane trafficking seems to be relatively conserved among eukaryotes. Therefore, a challenge for plant scientists is to determine how these molecular machineries work in a different structural and dynamic organization. This review will focus on some aspects of membrane dynamics that involve coat proteins, SNAREs (soluble N-ethylmaleimide-sensitive factor attachment receptor proteins), lipids, and lipid-interacting proteins.
Key words: Coat proteins, endoplasmic reticulum (ER), Golgi apparatus (GA), lipids, membrane biogenesis, membrane trafficking, SNAREs
|
Introduction |
|---|
|
TopAbstractIntroductionSpecific molecular machineries...The SNARE machineryLipids, lipid-modifying enzymes,...Challenges for the futureReferences |
|---|
Eukaryotic cells have developed a complex organization of biological membranes defining their intracellular compartments. In plant cells, membrane biology has to be adapted to specific requirements. For instance, it has to organize the photosynthetic pathways in chloroplasts, the formation of specific organelles for protein, glycan, or lipid storage, permit rapid changes in vacuolar content and membrane composition in response to osmotic or ionic stresses whilst maintaining the correct turgor pressure, and even organize de novo compartmentation during symbiotic processes. Plant cell endomembranes also participate in the organization of the transport and delivery of specific secretory molecules (transport and delivery of cell wall components, polarized distribution of specific plasma membrane transporters, etc). As in any eukaryotic cells, the plant cell secretory pathway is made up of a group of discrete membrane-bound organelles [i.e. endoplasmic reticulum (ER), Golgi apparatus (GA), endosomes/pre-vacuoles, and vacuoles], working along a secretory gradient to organize the processing, sorting, and delivery of cargo molecules to their final destination (Neumann et al., 2003; Hawes and Satiat-Jeunemaitre, 2005; Boutté et al., 2006).
The functional and structural identity of each membrane-bound compartment is related to the architecture/composition of its constitutive membranes (Moreau et al., 1998). However, early views of membrane-bound compartments as fixed structures are now being replaced by the concept that they are highly dynamic (Brandizzi et al., 2004). It is now clear that membrane composition and membrane-bound structures evolved to fit cellular requirements, and that membrane synthesis (and the turnover of membrane components) is a highly regulated process. It is well known that there can be a continuous exchange of membrane between the compartments of the secretory pathway, alongside some recycling processes to maintain compartment identity. The complexity of the mechanisms involved in such membrane dynamics and processing is increased further by the recent findings that chloroplast proteins may also transit through the GA, thus suggesting a possible membrane continuity between what was once considered to be two entirely independent organelles (Villarejo et al., 2005).
To understand the specificities of plant membrane biology along the secretory pathway, several questions must be addressed. (i) How is membrane identity acquired for each compartment? (ii) How is membrane composition maintained despite the vectorial membrane flow from ER to plasma membrane? (iii) How is the membrane composition adapted for differential developmental and environmental programmes? (iv) What are the molecular machineries regulating the passage of membrane from one compartment to the other along the secretory pathway?
Here, an overview of the recent data gained from studies on the ER–GA complex is provided, illustrating the complexity of the processes involved in the exchange between two distinct membrane interfaces. Moreover, since components of the molecular machinery regulating these processes appear to be conserved between eukaryotes, unravelling the exact roles of these molecular structures in the plant ER–GA interface, which appears to be structurally different from that of animal and yeast cells, is a tremendous challenge (Hawes and Satiat-Jeunemaitre, 2005). The requirement for specific sets of proteins and lipids that ensure efficient transit of cargo from the ER to the Golgi whilst maintaining the identities of the two organelles will be discussed. In this context, the focus is on the role of coat proteins, specific SNAREs (soluble N-ethyl-maleimide sensitive factor attachment receptor proteins) involved in membrane fusion, and the machinery that maintains the lipid identity of the two compartments.
The plant ER–GA interface: a dynamic channel
From ER to GA:
All proteins, membrane-bound or soluble, that are pre-destined for exocytosis or storage in intracellular compartments of the endomembrane system, have to enter the secretory pathway co-translationally via the Sec61 channel in the ER (Nicchita, 2002). Some may then undergo a first set of glycosylation events, and the quality of their processing will be checked by ER molecular chaperones (Trombetta and Parodi, 2003). Most will then have to reach the GA compartments, and transit through the Golgi stack before reaching their final destination (Munro, 2005).
The ER and GA are very distinct organelles in plant cells, having their own morphological, physiological, and biochemical features (Vitale and Denecke, 1999; Hawes and Satiat-Jeunemaitre, 2005). The ER is generally organized into a tubular reticulate network with occasional cisternae, although the precise 3D organization may appear to vary with the visualization techniques used and the cell types observed (Satiat-Jeunemaitre et al., 1999). The GA is usually made up of hundreds of discrete micron size stacks of membrane-bound cisternae, usually termed Golgi bodies, Golgi stacks, or dictyosomes. Direct connections between cisternae are often observed (Hawes and Satiat-Jeunemaitre, 2005; Képès et al., 2005). Moreover, membrane-like continuity between ER and GA has been observed using electron microscopy (see references in Képès et al., 2005), and the ER has to participate in the biogenesis of the GA (membrane synthesis, processing, and sorting of Golgi-resident proteins). Conversely, the biology of the GA will condition the structure of the ER as shown by some drug-induced effects of GA disruption on ER structure and function (Satiat-Jeunemaitre et al., 1996). This tight functional and structural interaction between ER and GA facilitates a mandatory passage between the two compartments for secreted and membrane-bound proteins.
The next question to address is the mode of transfer of cargo molecules between the two compartments.
Crossing the ER–GA interface: different transportation systems?:
Biochemical studies suggest that several distinct mechanisms may effect the ER–GA passage of cargo molecules (Moreau et al., 1998; see also additional references in Mérigout et al., 2002), but their mechano-structural support is still a question for further study.
Two main models, not mutually exclusive, have been proposed to explain the passage of ER-processed molecules to the GA. As electron microscopy studies have often indicated membrane connections between the two compartments in plant cells, it was first proposed that material exchange could take place by tubular connections, either transient or permanent. A major stream would be the unspecific transport of cargo molecules (membrane-bound or soluble), in so-called bulk flow (Denecke et al., 1990). Alternatively, movement from one compartment to the other could be selective and carried out by specific structures such as transport vesicles (Contreras et al., 2004b; see also above).
Arguments in favour of or against each of the proposed models are outlined in many reviews, with comments on the methodological approaches, the biological material, and, in some cases, the specific school of thought favoured in this controversial debate (Neumann et al., 2003; Hanton et al., 2005; Hawes, 2005; Képès et al., 2005). Therefore, it will not be discussed further here. The recent development of bioimaging techniques mainly using tobacco leaf epidermal cells has contributed much to this debate, as they demonstrate beautifully the close relationship between the ER and GA (Boevink et al., 1998; Brandizzi et al., 2002b, Saint-Jore et al., 2002; Chatre et al., 2005). These bioimaging approaches on leaf epidermal cells or BY-2 cells have also revealed the amazing dynamics of the ER–GA complex (Boevink et al., 1998; Nebenführ et al., 1999). Different types of movements of the ER–GA complex have been observed: (i) rapid movement of protein on the ER surface (Runions et al., 2006); (ii) organelle movements as both ER and Golgi stacks can move with cytoplasmic streams; (iii) more organized movements, as Golgi stacks move over ER/actin tracks (Boevink et al., 1998); and (iv) intraorganelle movement as ER contents can also move within the lumen of the network (Hawes et al., 2001). Are such movements necessary for the operation of ER to Golgi transport? It has clearly been shown by photobleaching experiments that Golgi movement is not necessary for the passage of cargo molecules from the ER to the GA (Brandizzi et al., 1992c), although transport can take place towards moving Golgi (daSilva et al., 2004).
A second clear message from these live-cell imaging data is that the ER–GA interface can be crossed in both directions, and these transport processes are independent of a functional cytoskeleton (Saint-Jore et al., 2002). Therefore, the regulation of ER–GA exchange is a subtle balance between anterograde and retrograde transport, and the regulation of such bidirectional movement of proteins and lipids is mediated by distinct molecular machineries.
Quest for the molecular mechanisms associated with the ER–GA interface:
The ER–GA interface in plant cells differs from the one described in animal cells. In the latter case, an intermediate compartment is a cargo carrier between ER and GA, and the role of microtubules in ER to GA transport is well established (Murshid and Presley, 2004). Therefore, the molecular machineries associated with the plant ER–GA interface must be plant specific at least at the level of their fine tuning.
Whatever the nature of the transport intermediate (tubular or vesicular), several types of structural and regulatory molecules would most probably be required to ensure efficient crossing of the ER–GA interface. (i) Components that will facilitate the deformation of ER membrane into tubules or vesicles (incidentally, one may note that budding profiles on ER membranes are scarcely observed in higher plant cells; meanwhile tubular extensions are commonly seen). (ii) A machinery that would facilitate the recruitment of specific cargo molecules either membrane-bound or soluble, unless bulk flow is the only mechanism associated with cargo exit in plant cells. (iii) Components that will mediate the docking and fusion of ER-derived membranes to Golgi membranes.
Different families of molecules can be involved in such actions, and here the focus is on three important groups: the coat proteins COPI and COPII with their associated GTPases (ARF1 and Sar1), which may be involved in specific recruitment processes, the SNARE proteins, which may have a role in mediating specific membrane fusion events, and finally the protein–lipid interactions, where proteins may induce lipid remodelling and membrane deformation (phospholipases, acyltransferases). Upstream molecules such as the Rab GTPases (Batoko et al., 2000), involved in regulating membrane fusion events, and putative membrane-tethering factors (Latijnhouwers et al., 2005) will not be discussed here.
|
Specific molecular machineries recruited at the ER and Golgi membranes |
|---|
|
TopAbstractIntroductionSpecific molecular machineries...The SNARE machineryLipids, lipid-modifying enzymes,...Challenges for the futureReferences |
|---|
The COPI and COPII coats
It is generally assumed that in yeasts and mammals, proteins are exported from the ER in transport intermediates, which then fuse with the GA and release their contents into this organelle. A common hypothesis proposes that the formation of these carriers is mediated by a protein coat made up of several proteins (Sar1, Sec23/Sec24, and Sec13/Sec31), termed COPII (Barlowe et al., 1994). Although direct evidence for the existence of COPII transport intermediates in plants has not yet been presented, the identification of plant homologues of COPII coat proteins implies that this pathway is present (d'Enfert et al., 1992; Bar-Peled and Raikhel, 1997; Takeuchi et al., 1998; Movafeghi et al., 1999; Belles-Boix et al., 2000). Furthermore, COPII components have been shown to influence protein transport between ER and Golgi (Takeuchi et al., 1998, 2000; Andreeva et al., 2000; Phillipson et al., 2001; daSilva et al., 2004; Yang et al., 2005). It is also generally assumed that the anterograde route is paralleled by a retrograde pathway that transports selected cargo and membrane from the Golgi back toward the ER, and mediated by the COPI protein coat. In contrast to COPII carriers, plant COPI vesicles have been visualized in situ (Pimpl et al., 2000), although their function in retrograde transport remains to be proven and even their exact role in mammalian cells has yet to be elucidated (for a review see Duden et al., 2005). Disruption of one route affects the other, meaning that interruption of COPI-mediated transport in the retrograde direction can prevent anterograde transport mediated by COPII (Lee et al., 2002; Ritzenthaler et al., 2002; Takeuchi et al., 2002; Pimpl et al., 2003; Stefano et al., 2006). Whether this occurs in a direct or an indirect manner has yet to be established (Stefano et al., 2006).
In both directions of transport, coat assembly is initiated by the activation of a small GTPase, achieved by the interaction of the GTPase with a guanine-nucleotide exchange factor (GEF), and its subsequent recruitment to the membrane from which the vesicle eventually buds. In non-plant systems, it has been shown that COPII carriers are formed in response to the activation of Sar1, followed by the recruitment of the larger coat subunits Sec23/Sec24 and Sec13/Sec31, which form the structural components of the coat (Barlowe et al., 1994). The ER domains where Sar1 (and other COPII proteins) are recruited define ER export sites (ERES), with the Sar1 exchange factor, Sec 12, being distributed over the surface of the ER (daSilva et al., 2004). ARF1 plays a similar role in COPI vesicle formation, recruiting the coatomer complex, made up of seven different subunits within the cytosol and targeting the resulting complex to the cis-Golgi membrane (Waters et al., 1991; Palmer et al., 1993). The coat may have several roles in membrane exchange: impose a curvature, and/or specifically select some cargo molecules. The coat has to be removed to allow fusion (and possibly transport) of the carrier to its final destination. It was thought that GTP hydrolysis was necessary for dissociation of the coat from the vesicle, thus exposing the membrane in order for fusion with the target membrane to occur. However, a recent study in yeast has shown that GTP hydrolysis is required in order for COPII vesicle fission from the ER membrane to occur (Bielli et al., 2005). This hydrolysis is instigated by the interaction of a GTPase-activating protein (GAP) with the GTPase. In the case of COPII, the GAP function can be performed by Sec23, a structural part of the coat itself (Yoshihisa et al., 1993). However, in the case of COPI, a Golgi-localized protein known as GAP1 catalyses GTP hydrolysis (Randazzo, 1997). The GAP activity of GAP1 appears to be partially dependent on the presence of coatomer (Szafer et al., 2000, 2001), which would reduce the possibility for unproductive cycling of ARF1 between the GTP- and GDP-bound states.
Interestingly, ARF1 may play several distinct roles within the plant cell. It has also been shown to play a role in a vacuolar sorting pathway in plants (Pimpl et al., 2003). Fluorescent protein fusions of Arabidopsis thaliana ARF1 have been localized to compartments of unknown identity derived from the GA (Stefano et al., 2006) and on putative endocytic structures (Xu and Scheres, 2005). An ARF-GEF that localizes to the endosomes has also been identified (Geldner et al., 2003), indicating that different GEFs may mediate alternative functions for GTPases within the cell. It is not clear whether the function of the GEF is only to recruit the GTPase to the membrane, or whether different GEFs induce subtly different conformations of the GTPase, resulting in modified functionality. No data have yet been presented to suggest similar multiple functions for Sar1, but this possibility cannot be ruled out.
Export mechanisms—diffusion (non-selective flow) or selection?
Whatever the nature of the transport intermediate in plant cells, the mechanism by which proteins are recruited into this structure has long been a subject of intense debate and may vary depending on the nature of the proteins investigated. It has been shown for instance that soluble proteins can exit the ER via a bulk flow mechanism (Denecke et al., 1990; Phillipson et al., 2001), although the existence of receptors to concentrate and increase the rate of export of certain soluble cargo molecules cannot be ruled out. Saturation of specific transport steps along the secretory pathway has also been observed, where high levels of expression of a secretory cargo molecule can reduce the efficiency by which it is transported (Phillipson et al., 2001). Further investigations are required to discern at which stage in the secretory pathway such saturation occurs, as it may indicate the existence of a previously unidentified cargo receptor. Additional evidence for the bulk flow mechanism of ER export of soluble proteins is provided by the existence of a retrieval mechanism for rescuing soluble ER-resident proteins from the GA. A tetrapeptide H/KDEL motif is found at the extreme C-terminus of many soluble ER-residents protein (Denecke et al., 1992) and is thought to interact with a receptor molecule (most probably a yeast ERD2p homologue) in the cis-Golgi, as described in other systems (Lewis et al., 1990; Semenza et al., 1990). This interaction apparently triggers the formation of COPI vesicles (Lewis and Pelham, 1992; Aoe et al., 1997), leading to the transport of the cargo molecule back to the ER.
In a similar way, some ER-resident transmembrane-spanning proteins travel to the GA and are later retrieved back to the ER. In these cases, a di-lysine motif is thought to interact with the COPI coat at the cis-Golgi (Contreras et al., 2004a). This retrieval system initially indicated that proteins with transmembrane domains travel through the secretory pathway by means of diffusion. Support for this model was presented by Brandizzi et al. (2002a), who showed that the length of the transmembrane domain plays a role in the final destination of type I membrane-spanning proteins within the secretory pathway. Short transmembrane domains restrict such proteins to the ER, whereas longer ones permit export to distal locations. This indicated that a diffusion mechanism might exist, whereby a protein would travel through the secretory pathway until it reached a membrane of sufficient thickness to mask its hydrophobic transmembrane domain. However, two recent publications have shown that certain amino acid motifs in the cytosolic domains of transmembrane proteins can also influence ER export in vivo (Hanton et al., 2005; Yuasa et al., 2005). Contreras et al. (2004b) have shown that a di-hydrophobic motif interacts with COPII coat subunits in vitro, although the relevance of such an interaction has yet to be established in vivo. It has been reported that a di-basic motif is involved in the ER–Golgi transport of a prolyl hydroxylase–green fluorescent protein (GFP) fusion (Yuasa et al., 2005). A role for di-acidic motifs was addressed by Hanton et al. (2005) for different types of Golgi-localized proteins. The authors demonstrated that not only did the di-acidic motifs of cargoes influence their rate of transport from ER to Golgi in plant cells, but also that they were dominant over transmembrane domain length in defining ER export. These findings indicate that although diffusion is one way by which proteins can exit the ER, other mechanisms also exist to ensure efficient export. Cytosolic signals may also contribute to allow fast tracking of certain proteins out of the ER, although it is not yet clear how these signals facilitate export of proteins. Studies in non-plant systems have suggested that COPII coat components interact with certain motifs and induce the formation of transport carriers. These findings are supported by the data presented by Contreras et al. (2004b) regarding di-hydrophobic signals, although further investigations are required to elucidate the role of other signals in ER export. If cytosolic export motifs induce formation of COPII carriers, it seems likely that other transmembrane proteins might enter these carriers by means of diffusion and exit the ER by this process.
|
The SNARE machinery |
|---|
|
TopAbstractIntroductionSpecific molecular machineries...The SNARE machineryLipids, lipid-modifying enzymes,...Challenges for the futureReferences |
|---|
General characteristics of SNAREs
SNAREs in eukaryotic cells favour the fusion of two apposed lipid bilayers and, through their specificities, contribute to the targeting of membrane components to maintain membrane identity and functions (Hong, 2005). Compared with Homo sapiens (35 SNAREs), Drosophila melanogaster (20 SNAREs), and Saccharomyces cerevisiae (21 SNAREs), the Arabidopsis genome contains a greater number of SNAREs, since at least 54 genes have been identified (Sanderfoot et al., 2000; Pratelli et al., 2004; Uemura et al., 2004). This large number of SNARE-related proteins in plants may reflect a plant-specific diversity in their cellular functions (for instance cell growth and development, autophagy, gravitropism, stress responses, and resistance to pathogens) that can be related or not to their fusogenic properties (Pratelli et al., 2004; Surpin and Raikhel, 2004).
SNAREs are mostly anchored to the cytosolic side of the membrane by a C-terminal trans-membrane domain (TMD) or a prenyl group, and a few, such as SNAP25, are bi-palmitoylated (Hong, 2005). SNAREs are characterized by one or rarely two SNARE motifs (coiled-coil domains) which are responsible for SNARE pairing by forming a four-helical bundle (Hong, 2005), and such domains are sufficient for triggering membrane fusion (Hu et al., 2005a). SNAREs were first classified into v-SNAREs and t-SNAREs (v for vesicle and t for target; Weber et al., 1998). Depending on the presence of an arginine or a glutamine in a central position of the helical bundle (the ‘zero layer’) of the SNARE motifs, SNAREs have since been classified either as R-SNAREs or Q-SNAREs (Fasshauer et al., 1998). Five subfamilies have since emerged: Qa-SNAREs (syntaxin, t-SNAREs with Q in the zero layer and a single SNARE motif), Qb-SNAREs (v-SNAREs with Q in the zero layer and a single SNARE motif similar to the N-terminal SNARE motif of SNAP25), Qc-SNAREs (v-SNAREs or syntaxin-like with Q in the zero layer and a single SNARE motif similar to the C-terminal SNARE motif of SNAP25), R-SNAREs (VAMPs, v-SNAREs with R in the zero layer and a single SNARE motif), and finally the SNAP25 family with two SNARE motifs (Hong, 2005). Rossi et al. (2004a) have proposed dividing VAMPs into two additional subfamilies, RD-SNAREs and RG-SNAREs, according to the conserved flanking amino acid (D or G) in the C-terminus of their zero layer.
Finally, SNAREs can also be classified according to their N-terminal domains. The most complex organization is found in the syntaxins which have three helical structures (Ha, Hb, and Hc) preceded by an N-terminal low complexity domain (Hong, 2005). R-SNAREs/VAMPs are also subdivided into short VAMPs with a very short N-terminus (brevins) or long VAMPs with a longer N-terminus (longins). The longin domains are structurally related to profilin (Rossi et al., 2005b).
Plant ER–Golgi SNAREs
The subcellular localizations of putative or established SNAREs of the ER–Golgi interface, based on protein expression, effects of brefeldin A treatments, and in vivo light microscopy analyses of fluorescent protein fusions (Uemura et al., 2004; Chatre et al., 2005), are given in Table 1. The fusogenic potential of plant SNAREs has rarely been studied in vitro, and the published studies are restricted to SNAREs from the trans-Golgi and the Golgi/prevacuolar compartments (Sanderfoot et al., 2001; Chen et al., 2005). In this review, the focus is on SNAREs predicted to function at the ER–Golgi interface and within the Golgi cisternae.
|
Some of these SNAREs are remarkable in the composition of their zero layer. Indeed AtBet11 (BS14a, Qc-SNARE) contains a histidine in the zero layer instead of a glutamine (Tai and Banfield, 2001) and AtSec22 (R-SNARE) contains a valine instead of an arginine. The same observation can be made for the yeast SNARE Bet1 which harbours a serine at the zero layer (Tai and Banfield, 2001), and for the mammalian Vti1a (Qb-SNARE) and Slt1 (Qc-SNARE) which have an aspartate (Hong, 2005). Therefore, SNARE interactions are either not affected by such changes in the nature of this residue or these ER–GA SNAREs perform their function by interacting differently.
Compared with mammals where SNARE expression can be highly tissue specific (Rossi et al., 2004b), it appears that SNAREs are widely expressed in plant tissues, and this is particularly true for the SNAREs of the ER–Golgi interface (Uemura et al., 2004). However, since the number of SNAREs in plants is higher and considering that these proteins may have very different functions in various tissues (Pratelli et al., 2004; Surpin and Raikhel, 2004), it may be expected that post-transcriptional or post-translational regulations confer tissue-specific activities to these proteins.
In plants, the role of the SNAREs has only recently been determined in vacuolar transport, cell surface assembly, and cell-plate formation during cytokinesis (Pratelli et al., 2004; Surpin and Raikhel, 2004; Jürgens, 2005). However, the role of the different SNAREs in the early secretory pathway of plants has yet to be established. Due to their homology with mammalian and yeast proteins, and their location in A. thaliana suspension-cultured cells (Uemura et al., 2004), the SNAREs AtSec22, AtMemb11, and AtSYP31/AtSed5 (related to the yeast syntaxin Sed5) were expected to act at the early ER–Golgi step. Figure 1 shows the intracellular location of AtSec22–YFP and AtMemb11–YFP in tobacco leaf epidermal cells and the effects of brefeldin A on their distribution, suggesting that these SNAREs are at the ER–Golgi interface (Chatre et al., 2005). Overexpression of AtSec22-C–YFP or AtMemb11-C–YFP [SNAREs translationally fused to either the YFP or CFP] in this system affected the dynamics of AtERD2-Y–CFP (the putative Arabidopsis H/KDEL receptor, translationally fused to either the YFP or CFP) and another Golgi reporter fusion protein (ST–YFP, partial rat sialyltransferase fused to YFP) with subsequent redistribution of these markers into the ER (Chatre et al., 2005). In addition, the trafficking of a secreted marker (secYFP) was mainly blocked at the level of the ER by co-expression of the two v-SNAREs (Chatre et al., 2005). Thus, AtSec22 and AtMemb11 appear to be critical v-SNAREs of the ER–Golgi interface in tobacco leaf epidermal cells, and are at least involved in the anterograde pathway. Whether these SNAREs are also involved in the retrograde pathway as reported for Sec22p in yeast (Burri et al., 2003) has to be determined.
|
AtBet11 (translationally fused to CFP), another v-SNARE closely related to animal Bet1 (regulating ER–Golgi transport in other eukaryotic cells; Hong, 2005), did not affect the distribution of either AtERD2–YFP or ST–YFP, an observation which is in agreement with its estimated trans-Golgi location (Uemura et al., 2004; Chatre et al., 2005). On the contrary, the trafficking of secYFP was partly blocked at the level of the ER by the expression of AtBet11–CFP (Chatre et al., 2005). As a consequence, it is suggested that AtBet11 is also involved in the anterograde pathway but probably at a later stage in the Golgi than AtSec22 and AtMemb11. Due to a cis-Golgi distribution, the AtBet11 isoform AtBet12 could be required in earlier steps in the Golgi (Uemura et al., 2004).
The situation is perhaps more complex for AtSYP31 (named AtSed5 in Chatre et al., 2005). Brefeldin A treatment was found to change the labelling of AtSYP31 to a more typical ER pattern in A. thaliana (Uemura et al., 2004) than in tobacco leaf epidermal cells where aggregated membranes (cAggr. in Table 1) were also observed. Overexpression of AtSYP31–CFP in tobacco leaf epidermal cells did not affect the dynamics of AtERD2–YFP and ST–YFP to the same extent as did overexpression of AtSec22 and AtMemb11, but the trafficking of secYFP was predominantly blocked at the level of the ER (Chatre et al., 2005). Since AtSYP31 is believed to be the t-SNARE syntaxin in the cis-Golgi, it could be predicted that when AtSYP31 is overexpressed, Golgi reporter proteins are less retained into the ER but more at the ERES, which cannot be distinguished from the Golgi bodies (daSilva et al., 2004). It has recently been shown in yeast that AtSYP31 may, under certain conditions, be by-passed in the activation of SNARE pairing for the formation of fusogenic SNARE complexes (Peng and Gallwitz, 2004). Such discrepancies could be due to different levels or distributions of the expressed proteins. A 3–5 times higher level of expressed tagged SNAREs compared with the endogenous proteins can lead to a substantial shift in the distribution of the expressed proteins (Cosson et al., 2005), and the consequence of such a partial shift on the functionality of the tagged protein may vary according to the fusion protein considered. In addition, these authors have shown that the location of SNAREs is not due to retention mechanisms but proceeds from dynamic transport equilibrium. The disturbance of such an equilibrium and overconcentration at specific sites may perturb membrane dynamics and protein function. Therefore, further investigations by electron microscopic immunocytochemistry need to be conducted to determine the precise location of the endogenous SNAREs between ER, ERES, and the different Golgi cisternae. This discussion may question the validity of localization/functional analyses using such overexpressed tagged proteins. The use of endogenous promoters may solve this problem, but insofar as the level of expression is sufficient to visualize the fusion proteins. It has also to be considered that, besides approaches using loss of function, approaches using gain of function such as overexpression can report complementary information.
No studies are yet available on the other putative ER–Golgi SNAREs presented in Table 1. We can only speculate that the ER-localized AtSYP81, which is a homologue of the animal Syn18 and the yeast Ufe1 known to be active in Golgi–ER retrograde transport, could be involved in any retrograde pathway ending in the ER. Finally, the existence of several syntaxin-like SNAREs such as AtSYP71, 72, and 73 in the ER could suggest their requirement in multiple retrograde pathways and/or fission events of the ER in other membrane biogenesis processes. All the SNAREs so far reported at the ER and Golgi level seem to be expressed ubiquitously (Uemura et al., 2004), but their abundance and tissue specificity will have to be determined through quantitative polymerase chain reaction (QPCR).
SNARE targeting
Although the central role of SNAREs in membrane fusion has been elucidated, only a few studies have been devoted to investigating how theses proteins are targeted to their specific compartments.
It has been shown that the targeting of the mammalian YKT6, a C-terminal isoprenylated SNARE, is dependent on the N-terminus profiling-like longin domain and not driven by the isoprenylation per se (Hasegawa et al., 2003). In fact, it has been proposed that the longin domain interacts with membrane lipids and the SNARE motif in the cytosol so that only the longin domain is available for interaction with the target membrane (Hasegawa et al., 2004). It has also been reported that targeting of mammalian SNAP-25, a plasma membrane palmitoylated SNARE, is independent of both palmitoylation and its interaction with the syntaxin 1A, and that an interaction with another membrane factor must be responsible for its targeting (Loranger and Linder, 2002).
It is known that type I membrane proteins can be anchored in specific membranes according to the length of the TMD (Brandizzi et al., 2002a), but the situation for C-terminal anchored proteins such as SNAREs is unknown. In yeast, Bet1p, Gos1p, Sft1p, and Sed5p have a 15 amino acid TMD and are located in the Golgi, whereas Sec22p and Slt1p, located in the ER, have a 19–20 amino acid TMD (Burri and Lithgow, 2004). In plants, AtSec22 and AtSYP31 both have a 17 amino acid TMD with different subcellular localizations (Uemura et al., 2004; Chatre et al., 2005). Therefore, it is unlikely that the length of the TMD of the integral SNAREs explains their location. This may be due to the different orientation of the proteins in the ER membranes with respect to type I proteins. It may have to be asked instead whether the N-terminal cytosolic domain(s) and the SNARE motifs of these proteins are required in SNARE targeting. The SNARE motif of Bet1 has been shown to contribute its subcellular targeting in NRK cells, and this function is independent of the heteromeric SNARE interactions of the protein (Joglekar et al., 2003). This highlights a very important point: that the SNARE motif has the ability to interact with different partners for different tasks. In the case of the animal and yeast Sec22, it has been shown that both the longin and the SNARE domains are required for an efficient packaging of Sec22p into COPII vesicles (Liu et al., 2004). In addition, the longin domain is not a regulatory factor for SNARE complex assembly, and SNARE pairing is not required for COPII-dependent Sec22p export from the ER (Liu et al., 2004). It appears that a sequence of 10 amino acids in the SNARE motif of Sec22p is critical for recognition by COPII, and Liu et al. (2004) have proposed that the longin and the SNARE motifs interact with the Sec23/Sec24 complex at two distinct sites.
Although AtSec22 has a KXKXX sequence at the C-terminus which could be an ER targeting motif, since AtSec22 is a C-terminus tail-anchored protein, the C-terminus is in the luminal side of the membrane and thus unlikely to be an ER targeting determinant. Mutagenesis approaches are required to unravel the motifs or domains involved in its targeting. For example, it would be interesting to investigate if its longin domain is critical since the vacuolar targeting of VAMP7 SNAREs appears to be dependent on their entire N-terminus (Uemura et al., 2005).
It is clear that SNARE–coat protein interactions can be critical for SNARE targeting and dynamics, but also for COP protein recruitment. This has recently been highlighted for the COPII and COPI machineries, and the ER–Golgi SNAREs. Specific interactions were first observed between Sed5p and the COPII component Sec24p in yeast (Peng et al., 1999). More recently, several binding sites on the Sec23/Sec24 complex and especially on Sec24p have been identified which govern the selection of several yeast SNAREs (Bet1, Sec22, and Sed5) and other cargo proteins (Miller et al., 2003; Mossessova et al., 2003). In addition, the selective Sec23/Sec24-dependent packaging of the SNAREs was determined to be specific for the fusogenic monomeric or complexed forms of the SNAREs, suggesting a tight control of fusion starting at the budding level (Mossessova et al., 2003). Finally, additional sites must exist in the COPII proteins to explain the efficiency and the selectivity of the ER–Golgi step for different cargo proteins including the SNAREs (Miller et al., 2005). Interestingly, it has also been suggested that v-SNAREs may function as ARF receptors on membranes (Rein et al., 2002; Honda et al., 2005). Therefore, different sets of regulators (GTP-binding proteins such as ARF, Sar1, Rab, etc., SNAREs, COPI and COPII coat proteins, and tethering factors) may provide multiple specific interactions between several partners, and drive the specificities of the targeting and fusion events. In addition, it has recently been observed that different SNAREs can have various affinities for specific membrane domains (Salaün et al., 2005a), and that the degree of their association with these domains may regulate the efficiency of exocytosis (Salaün et al., 2005b). As a consequence, domain-forming lipids and lipid-modifying enzymes will also have to be considered in the driving and regulation of these events.
SNAREs in fusion and beyond?
The physical action of SNAREs in membrane fusion has been investigated in several animal and yeast models, and all studies conclude that fusion can proceed through a hemifusion intermediate (Xu et al., 2005; Reese et al., 2005; Ungermann and Langosh, 2005). In addition, hemifusion may not be just an intermediate step to complete fusion but may also participate in transient fusion reactions during reversible kiss-and-run events (Giraudo et al., 2005). In all fusion events, v-SNAREs may control the successive steps involved: the N-terminus regulates vesicle priming, the SNARE motif drives initiation of fusion, and the TMD may control fusion pore formation (Borisovska et al., 2005; Ungermann and Langosh, 2005). The SNARE complexes comprise four helices with a Qa/Qb/Qc/R stoichiometry, and replacing one Q by an R in the central zero layer can be lethal or induces a growth defect in yeast that can be rescued by replacing the R of the R-SNARE with a Q (Graf et al., 2005). However, all zero level positions do not seem to be critical to the same extent as the fact that the functionality of some of these proteins can be retained after residue substitution. In addition, it has been observed that only some SNARE combinations are functional. For example, Sec22p can be replaced by YKT6p in yeast (Liu and Barlowe, 2002) but the reverse is not true. The extent to which some redundancy exists in the SNAREs of Arabidopsis encoding 54 of these proteins is a relevant question waiting to be addressed. As a support observation, Niihama et al. (2005) have shown that a single mutation in the SNARE VTI12 can rescue the zig-1 mutant of Arabidopsis which lacks the SNARE VTI11, indicating some flexibility in SNARE functions through gene duplication.
In addition, it has recently been suggested that SNAREs themselves and SNARE-like proteins (amysin, tomosyn) may compete to participate in non-functional SNARE complexes that will inhibit and regulate fusion (Scales et al., 2002; Varlamov et al., 2004; Constable et al., 2005). The specificities of SNARE complexes in the different membrane trafficking pathways are, to some extent, recapitulated in cell-free fusion assays with isolated SNAREs. It has been suggested that non-functional complexes may also have a physiological relevance, and the so-called ‘i-SNAREs’ (inhibitory SNAREs) may contribute to control the specificity of membrane fusion (Varlamov et al., 2004). For example, according to SNARE concentration gradients in the Golgi, SNAREs may have a normal SNARE function or an i-SNARE function in different cisternae, and this would confer a sort of buffering effect on the fusion capacity in the cisternae (Varlamov et al., 2004). As proposed for Bet1 in animal cells (Varlamov et al., 2004), Bet11 in the trans-Golgi of plant cells could also have such an i-SNARE role. Finally, animal proteins containing longin domains homologous to Sec22b (Sec22a and Sec22c) are expressed without any genuine SNARE motif and may participate in the regulation of SNARE complex formation (Gonzales et al., 2001).
Lastly, another function which could be managed by SNAREs, especially by syntaxins, is a more structural action on membrane architecture. It has recently been found that mutations affecting the phosphorylation of Sed5p cannot only disturb ER–Golgi traffic in yeast but can also affect the morphology of the Golgi (Weinberger et al., 2005). AtSYP31 seems to have a wide distribution across plant Golgi (Uemura et al., 2004; Chatre et al., 2005). Its cis-Golgi location is compatible with a function as a syntaxin in ER–Golgi transport, but a distribution from the cis- to the trans-cisternae could indicate that this syntaxin is also participating in structural aspects of plant Golgi.
|
Lipids, lipid-modifying enzymes, and organelle dynamics |
|---|
|
TopAbstractIntroductionSpecific molecular machineries...The SNARE machineryLipids, lipid-modifying enzymes,...Challenges for the futureReferences |
|---|
Backgound
As stated by Engelman (2005), ‘membranes are more mosaic than fluid’. The random state of membrane organization of the Singer–Nicholson model is now replaced by one reflecting a more patchy organization with different domains (composition, thickness, turnover, and, therefore, homeostasis). In such a view, membrane properties are highly dependent on protein–lipid interactions where lipids organize proteins and vice versa. As a consequence, membrane structure and membrane trafficking are controlled by close relationships between protein-based and lipid-based machineries (De Matteis and Godi, 2004). Beside the roles of coat proteins and SNAREs in membrane organization and dynamics (structure, deformation, fusion), lipids, lipid-modifying enzymes (phospholipases, acyltransferases), and lipid-modified proteins (acylated or isoprenylated GTP-binding proteins for example) also have key roles to play in membrane dynamics, for instance through the control of membrane curvature (McMahon and Gallop, 2005). In this last section, lipids are considered not as membrane bricks but more as molecular architects (by themselves or through protein modifications) of membrane dynamics regulating trafficking.
Sterols, oxysterols, and related proteins
The structure of the sterol molecules can affect and regulate the curvature of the membranes which will help to prepare a membrane for budding or fusion (Bacia et al., 2005). In animal cells, cholesterol has been shown to be required for the formation of both constitutive and regulated post-Golgi secretory vesicles (Wang et al., 2000), and in the biogenesis of synaptic vesicles (Thiele et al., 2000). It has been shown that cholesterol may facilitate vesicle fusion in exocytosis by virtue of its intrinsic negative curvature, and contributes to fast Ca2+-triggered membrane fusion (Churchward et al., 2005). In yeast, ergosterol has also been found to be a key regulator of endocytosis, and several mutants of this pathway, the erg mutants, are affected in the ergosterol metabolic pathway (Heese-Peck et al., 2002). On the other hand, the cholesterol levels in mammalian Golgi membranes must be tightly regulated because an excess of cholesterol can lead to a vesiculation of the Golgi complex itself, and this process is dynamin and phospholipase A2 (PLA2)-dependent (Grimmer et al., 2005).
Oxysterols are naturally occurring hydroxylated sterol derivatives of interest in the secretory pathway of eukaryotic cells. Oxysterol-binding proteins (OSBP) are effectively lipid receptors involved in sterol homeostasis, being able to regulate vesicular transport in both animal and yeast cells (Xu et al., 2001; Li et al., 2002). In yeast, the related OSBP (Kes1p) has been shown to be targeted to the Golgi complex through binding to phosphoinositides formed by a phosphatidylinositol kinase (Pik1p), and may regulate Golgi secretory function through interactions with the ARF(s) and Sec14p pathways (Li et al., 2002). In animal cells, one OSBP and several OSBP-related proteins were found to interact with a syntaxin-like VAMP-associated protein-A, and have been shown to participate in the organization of the COPII-dependent ER–Golgi pathway (Wyles et al., 2002; Wyles and Ridgway, 2004).
In plant cells, the sterol biosynthetic pathways lead not only to mature membrane sterols such as sitosterol and stigmasterol, but also to cholesterol or the plant steroid hormones (brassinosteroids). The study of dwarf mutants disturbed in brassinosteroid synthesis has clearly demonstrated the role of these steroid hormones in plant development (Fujioka and Yokota, 2003; Schaller, 2003). Interestingly, it has then been found that membrane sterols are also critical for controlling cell growth, cell polarity, and embryonic development (Schaeffer et al., 2001; Schrick et al., 2002; Willemsen et al., 2003; Schaller, 2004). On the contrary, little is known about the role of sterols in the plant secretory pathway and Golgi dynamics. Preliminary approaches to answer these questions, based upon pharmacological experiments, are yielding some promising results. For instance, blocking cyclopropylsterol maturation at the level of the cycloeucalenol–obtusifoliol isomerase by fenpropimorph disturbs the secretory pathway and induces a fenestration of the Golgi bodies (Hartmann et al., 2002). Interestingly, a strong perturbation of the morphology of the Golgi by brefeldin A leads to a reduced synthesis of phytosterols (Mérigout et al., 2002). Therefore, some relationships may exist between sterol metabolism and Golgi morphology in plant cells. Recently, lipid rafts have been isolated from the plasma membrane of plant cells, and as expected, the phytosterols are enriched in these domains (Mongrand et al., 2004; Borner et al., 2005). We have recently found evidence of lipid rafts in the Golgi membranes of plant cells, and observed that a perturbation of sterol metabolism by fenpropimorph can block sterol precursors and lipid rafts in the Golgi (Laloi et al., unpublished results). As a consequence, mature sterols are also probably critical for membrane trafficking as in other eukaryotes.
Sphingolipid metabolism and organelle structure
Although the metabolism of cerebrosides and inositolsphingolipids (the two families of sphingolipids in plants) have been extensively studied (Dunn et al., 2004; Sperling et al., 2004), nothing is known about their roles in the secretory pathway of plants. These lipids are also enriched in plasma membrane lipid rafts (Mongrand et al., 2004; Borner et al., 2005), but their action in membrane traffic has yet to be established in plants.
Several studies have, however, investigated such a role in other eukaryotic cells. For example, natural long chain ceramides, either incorporated or formed upon sphingomyelinase treatment in animal cells, enhance the disassembly of the Golgi induced by brefeldin A, suggesting that the levels of ceramides may be critical for Golgi stability (Fukunaga et al., 2000). Pharmacological approaches using inhibitors of glucosylceramide synthase have confirmed a role for sphingolipids in the maintenance of Golgi architecture, and in anterograde membrane trafficking (Nakamura et al., 2001). It has been shown that short chain ceramides can decrease the binding of ARF to Golgi membranes and therefore affect the formation of COPI vesicles (Abousalham et al., 2002). Finally, recent investigations in animal cells have suggested that high concentrations of sphingosine formed from the hydrolysis of ceramides can induce Golgi fragmentation (Hu et al., 2005b). Overall, these studies highlight a likely role for sphingolipids in Golgi morphodynamics as found for cholesterol. However, the situation is less clear in yeast where the various post-Golgi pathways appear not to be systematically dependent on sphingolipid biosynthesis (Lisman et al., 2004).
Phospholipases, acyltransferases, and lipid-binding proteins
More and more studies reveal the critical impact of phospholipid modifications (acylations or de-acylations) on membrane dynamics and particularly in the case of the Golgi membranes in animal cells. Evidence has been obtained for the regulation of Golgi structure and membrane trafficking by PLA2 which cleaves the fatty acids of phospholipids at the sn-2 position of the glycerol (Choukroun et al., 2000), and the phospholipase activity induces membrane tubulation (De Figueiredo et al., 1998). Such properties of PLA2 to promote deformation of membranes have been confirmed on giant liposomes where PLA2-induced budding and fission events have been observed (Staneva et al., 2004), and these events are caused by the formation of positive membrane curvatures induced by the lysophospholipids (Brown et al., 2003). Interestingly, the reverse reaction of phospholipid synthesis by acyltransferases stabilizes membrane bilayers, and inhibition of such enzymes also leads to Golgi membrane tubulation (Drecktrah et al., 2003). PLA1-related enzymes (which cleave the fatty acids of phospholipids at the sn-1 position of the glycerol) have been found to induce dispersion of the Golgi complex and/or aggregation of the ER membranes (Nakajima et al., 2002), and they can interact with proteins of the COPII machinery to participate in the organization of the ER exit sites (Shimoi et al., 2005). No evidence for the involvement of such enzymes in Golgi dynamics and the early secretory pathway of plant cells has yet been shown.
Another critical phospholipase in Golgi function is phospholipase D (PLD) that cleaves the polar head to produce phosphatidic acid (PA). Inhibition of its activity (stimulated by ARF and/or Sar1p) decreases the level of PA but also of phosphoinositides that are required for structural integrity of the Golgi complex in animal cells (Siddhanta et al., 2000). Interestingly, PA regulates the synthesis of phosphoinositides but the latter can also regulate the PLD activity. It has been found that PLD activation participates in the formation of COPII complexes in ER export (Pathre et al., 2003). An isoform of PLD has been localized to the rims of the Golgi complex and may be involved in Golgi morphodynamics playing a role in vesicular transport from the Golgi (Freyberg et al., 2002). The critical role of PLD has also been determined in yeast (Routt et al., 2005 and references therein). Beside the role of PA in stimulating phosphoinositide synthesis, a role for PA and its derivatives lyso-PA and diacylglycerol in regulating membrane curvature has also been proposed (Kooijman et al., 2003).
In plant cells, it is clear from the literature that PLD and the formation of PA are required for normal plant growth and polarized cell expansion (Gardiner et al., 2003; Potocky et al., 2003). It has recently been shown that inhibition of PLD in the pollen tube considerably reduces the number of secretory vesicles, and that PA and phosphoinositides regulate pollen tube growth (Monteiro et al., 2005). Despite numerous studies on PLD and PA signalling in plant physiology, no studies have yet addressed the putative role of PLD in the early secretory pathway of plants.
Another family of proteins relevant to Golgi function are the phospholipid-binding proteins. It has been shown that Sec14p and Nir2p are both involved in the control of diacylglycerol levels in Golgi membranes in yeast and mammal cells, respectively (Kearns et al., 1997; Litvak et al., 2005). Maintenance of diacylglycerol concentrations in Golgi membranes is certainly related to the property of diacylglycerol in membrane fission. In plant cells, a Sec14 candidate has been cloned from Arabidopsis which can complement the yeast mutant (Jouannic et al., 1998), but no role at the Golgi level has been established. Recently, a Sec14p-nodulin domain PITP from Arabidopsis (a Sec14p-like protein) has been demonstrated to be key regulator of polarized membrane growth in root hairs (Vincent et al., 2005).
Lipid domains and selective transport at the ER–Golgi interface
Phosphatidylserine accumulates in the plasma membrane of leek cells and originates from both the ER after intracellular trafficking and a local synthesis by the serine exchange enzyme (Bessoule and Moreau, 2004). Both phosphatidylserine synthase and the serine exchange enzyme can synthesize phosphatidylserine in the ER (Bessoule and Moreau, 2004). As a first example of phospholipid sorting at the ER–Golgi interface, it has been determined that only the serine exchange enzyme synthesizes phosphatidylserine with very long chain fatty acids that are targeted to the secretory pathway (Sturbois-Balcerzak et al., 1999; Vincent et al., 2001). Phosphatidylserine was also found to be targeted to ER-derived vesicles in rat liver membranes (Moreau et al., 1992, 1993). This suggests the existence of specific ER domains where such lipids are concentrated before transport, and this poses the question as to whether these domains correspond to the ER export sites revealed by the COPII machinery (da Silva et al., 2004; Stefano et al., 2006). Interestingly, C-tail-anchored proteins with a moderate hydrophobic TMD can reside in the ER, whereas addition of non-polar amino acids in the TMD can translocate the protein to the plasma membrane. This protein sorting toward the secretory pathway may be linked to lipid-based sorting mechanisms, and specific interactions with saturated fatty acid-bearing acidic phospholipids such as phosphatidylserine may be relevant (Ceppi et al., 2005). Another interesting finding was the requirement for phosphatidylserine in the formation of ER-derived COPII vesicles in vitro (Matsuoka et al., 1998). Finally, the well-established association of GTP-binding proteins (such as Sar1, ARFs, and Rabs) with membranes, governed by lipid modifications and lipid interactions, indicates that lipids are crucial at several levels in membrane trafficking.
|
Challenges for the future |
|---|
|
TopAbstractIntroductionSpecific molecular machineries...The SNARE machineryLipids, lipid-modifying enzymes,...Challenges for the futureReferences |
|---|
We can wonder whether Golgi bodies mature from the ER export sites in a similar manner to that occurring in Pichia pastoris, S. cerevisiae, or Drosophila oocytes (Bevis et al., 2002; Glick, 2002; Herpers and Rabouille, 2004), and whether the cis-Golgi cisternae may progressively mature into trans-Golgi cisternae as in P. pastoris (Mogelsvang et al., 2003). Figure 2 proposes several questions about the organization of the ER–GA interface, and the interactions between the different families of proteins (GTP-binding proteins, coat proteins, SNAREs, etc.) that may drive the specificic exchange of material between the ER and the Golgi.
|
Another challenge for the future will be to unravel the physical role of lipids and lipid/membrane-interacting proteins in membrane domain formation and dynamics related to membrane trafficking in plants. As highlighted by McMahon and Gallop (2005), membrane deformation and regulation of membrane curvature are fundamental events that must be considered. Such physical deformations can be managed by changes in lipid composition and asymmetry, integral membrane proteins, cytoskeletal proteins, scaffolding by peripheral membrane proteins, and finally by active helix insertion of membrane-associated proteins (Lee et al., 2005; McMahon and Gallop, 2005). In addition, membrane curvature may regulate the strength of association of peripheral proteins, and membrane curvature-sensing proteins may contribute to the architecture and dynamics of membrane domain formation (De Matteis and Godi, 2004). In the case of lipid asymmetry, its loss in cis-Golgi membranes may affect Golgi retrograde transport to the ER by perturbing protein recruitment (Hua and Graham, 2003). Promising investigations in plants on interactions between lipids and proteins of the traffic machineries have begun to emerge from the literature (Jensen et al., 2000; Lam et al., 2002).
The complexity and diversity of protein and lipid machineries involved at the ER–GA level will undoubtedly keep researchers busy for a long while.
|
Abbreviations |
|---|
ARF, ADP-ribosylation factor; CFP, cyan fluorescent protein; ER, endoplasmic reticulum; ERES, endoplasmic reticulum export site; GA, Golgi apparatus; GAP, GTPase-activating protein; GEF, guanine-nucleotide exchange factor; GFP, green fluorescent protein; OSBP, oxysterol-binding protein; PA, phosphatidic acid; PLA, phospholipase A; PLD, phospholipase D; SNARE, soluble N-ethylmaleimide-sensitive factor attachment receptor protein; TMD, trans-membrane domain; YFP, yellow fluorescent protein.
|
References |
|---|
|
TopAbstractIntroductionSpecific molecular machineries...The SNARE machineryLipids, lipid-modifying enzymes,...Challenges for the futureReferences |
|---|
Abousalham A, Hobman TC, Dewald J, Garbutt M, Brindley DN. (2002) Cell-permeable ceramides preferentially inhibit coated vesicle formation and exocytosis in Chinese hamster ovary compared with Madin–Darby canine kidney cells by preventing the membrane association of ADP-ribosylation factor. Biochemical Journal 361 653–661.[CrossRef][ISI][Medline]
Andreeva AV, Zheng H, Saint-Jore CM, Kutuzov MA, Evans DE, Hawes CR. (2000) Organization of transport from endoplasmic reticulum to Golgi in higher plants. Biochemical Society Transactions 28 505–512.[ISI][Medline]
Aoe T, Cukierman E, Lee A, Cassel D, Peters PJ, Hsu VW. (1997) The KDEL receptor, ERD2, regulates intracellular traffic by recruiting a GTPase-activating protein for ARF1. The EMBO Journal 16 7305–7316.[CrossRef][ISI][Medline]
Bacia K, Schwille P, Kurzchalia T. (2005) Sterol structure determines the separation of phases and the curvature of the liquid-ordered phase in model membranes. Proceedings of the National Academy of Sciences, USA 102 3272–3277.
Barlowe C, Orci L, Yeung T, Hosobuchi M, Hamamoto S, Salama N, Rexach MF, Ravazzola M, Amherdt M, Schekman R. (1994) COPII: a membrane coat formed by Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell 77 895–907.[CrossRef][ISI][Medline]
Bar-Peled M and Raikhel NV. (1997) Characterization of AtSEC12 and AtSAR1. Proteins likely involved in endoplasmic reticulum and Golgi transport. Plant Physiology 114 315–324.[Abstract]
Batoko H, Zheng HQ, Hawes C, Moore I. (2000) A rab1 GTPase is required for transport between the endoplasmic reticulum and Golgi apparatus and for normal Golgi movement in plants. The Plant Cell 12 2201–2218.
Belles-Boix E, Babiychuk E, Montagu MV, Inzé D, Kushnir S. (2000) CEF, a sec24 homologue of Arabidopsis thaliana, enhances the survival of yeast under oxidative stress conditions. Journal of Experimental Botany 51 1761–1762.
Bessoule JJ and Moreau P. (2004) Phospholipid synthesis and dynamics in plant cells. Topics in Current Genetics 6 89–124.
Bevis BJ, Hammond AT, Reinke CA, Glick BS. (2002) De novo formation of transitional ER sites and Golgi structures in Pichia pastoris. Nature Cell Biology 4 750–756.[CrossRef][ISI][Medline]
Bielli A, Haney CJ, Gabreski G, Watkins SC, Bannykh SI, Aridor M. (2005) Regulation of Sar1 NH2 terminus by GTP binding and hydrolysis promotes membrane deformation to control COPII vesicle fission. Journal of Cell Biology 171 919–924.
Boevink P, Oparka K, Santa Cruz S, Martin B, Betteridge A, Hawes C. (1998) Stacks on tracks: the plant Golgi apparatus traffics on an actin/ER network. The Plant Journal 15 441–447.[CrossRef][ISI][Medline]
Borisovska M, Zhao Y, Tsytsyura Y, Glyvuk N, Takamori S, Matti U, Rettig J, Südhof T, Bruns D. (2005) v-SNAREs control exocytosis of vesicles from priming to fusion. The EMBO Journal 24 2114–2126.[CrossRef][ISI][Medline]
Borner GH, Sherrier DJ, Weimar T, Michaelson LV, Hawkins ND, Macaskill A, Napier JA, Beale MH, Lilley KS, Dupree P. (2005) Analysis of detergent-resistant membranes in Arabidopsis. Evidence for plasma membrane lipid rafts. Plant Physiology 137 104–116.
Boutté Y, Crosnier M-T, Carraro N, Traas J, Satiat-Jeunemaitre B. (2006) Immunocytochemistry of the plasma membrane recycling pathway and cell polarity in plants: studies on PIN proteins. Journal of Cell Science 119 1255–1265.
Brandizzi F, Frangne N, Marc-Martin S, Hawes C, Neuhaus JM, Paris N. (2002a) The destination for single-pass membrane proteins is influenced markedly by the length of the hydrophobic domain. The Plant Cell 14 1077–1092.
Brandizzi F, Fricker M, Hawes CR. (2002b) A greener world: the revolution in plant bioimaging. Nature Reviews Molecular and Cell Biology 3 520–530.
Brandizzi F, Irons SL, Johansen J, Kotzer A, Neumann U. (2004) GFP is the way to glow: bioimaging of the plant endomembrane system. Journal of Microscopy 214 138–158.[ISI][Medline]
Brandizzi F, Snapp EL, Roberts AG, Lippincott-Schwartz J, Hawes C. (2002c) Membrane protein transport between the endoplasmic reticulum and the Golgi in tobacco leaves is energy dependent but cytoskeleton independent: evidence from selective photobleaching. The Plant Cell 14 1293–1309.
Brown WJ, Chambers K, Doody A. (2003) Phospholipase A2 (PLA2) enzymes in membrane trafficking: mediators of membrane shape and function. Traffic 4 214–221.[ISI][Medline]
Burri L and Lithgow T. (2004) A complete set of SNAREs in yeast. Traffic 5 45–52.[CrossRef][ISI][Medline]
Burri L, Varlamov O, Doege CA, Hofmann K, Beilharz T, Rothman JE, Söllner T, Lithgow T. (2003) A SNARE required for retrograde transport to the endoplasmic reticulum. Proceedings of the National Academy of Sciences, USA 100 9873–9877.
Ceppi P, Colombo S, Francolini M, Raimondo F, Borgese N, Masserini M. (2005) Two tail-anchored protein variants, differing in transmembrane domain length and intracellular sorting, interact differently with lipids. Proceedings of the National Academy of Sciences, USA 102 16269–16274.
Chatre L, Brandizzi F, Hocquellet A, Hawes C, Moreau P. (2005) Sec22 and Memb11 are v-SNAREs of the anterograde endoplasmic reticulum–Golgi pathway in tobacco leaf epidermal cells. Plant Physiology 139 1244–1254.
Chen Y, Shin YK, Bassham DC. (2005) YKT6 is a core constituent of membrane fusion machineries at the Arabidopsis trans-Golgi network. Journal of Molecular Biology 350 92–101.[CrossRef][ISI][Medline]
Choukroun GJ, Marshansky V, Gustafson CE, McKee M, Hajjar RJ, Rosenzweig A, Brown D, Bonventre JV. (2000) Cytosolic phospholipase A2 regulates Golgi structure and modulates intracellular trafficking of membrane proteins. Journal of Clinical Investigation 106 983–993.