SEARCH

SEARCH BY CITATION

Keywords:

  • evolution;
  • exocyst;
  • expression profiling;
  • plants;
  • protein complex (Sec3, Sec5, Sec6, Sec8, Sec10, Sec15, Exo70 and Exo84)

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • The exocyst is a complex of eight proteins (Sec3p, Sec5p, Sec6p, Sec8p, Sec10p, Sec15p, Exo70p and Exo84p) involved in tethering vesicles to the plasma membrane during regulated or polarized secretion. Here, the plant exocyst complex was explored in phylogenetic, expression, and subcellular localization studies.
  • Evolutionary relationships of predicted exocyst subunits were examined in the complete genomes of Arabidopsis thaliana, Oryza sativa, Populus trichocarpa and Physcomitrella patens. Furthermore, detailed expression profiling of the A. thaliana microarray databases was performed and subcellular localization patterns were studied.
  • Several plant exocyst subunit genes appear to have undergone gene expansion in a common ancestor and subsequent duplication events in independent plant lineages. Expression profiling revealed that the A. thaliana Exo70 gene family exhibits dynamic expression patterns, while the remaining exocyst subunit genes displayed more static profiles. Subcellular localization patterns for A. thaliana exocyst subunits ranged from cytosolic to endosomal compartments (with enrichment in the early endosomes and the trans-Golgi network). Interestingly, two endosomal-localized AtExo70 proteins also recruited other exocyst subunits to these compartments.
  • Overall subcellular localization patterns were observed that were also found in yeast and animal cells, and this, coupled with the evolutionary relationships, suggests that the exocyst may perform similar conserved functions in plants.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Vesicle trafficking is a conserved process that is essential to all eukaryotes and is involved the specific transport of macromolecules from one membrane-bound compartment to another within the cell. Plants, mammals and yeast, for instance, contain all of the necessary machinery for vesicle formation and fusion, and trafficking occurs through the formation of vesicles from the donor membrane, followed by their transport to the target site, and eventually membrane fusion. This process of directing vesicles to the correct compartment is tightly regulated, and the key players in tethering vesicles to the target membrane include the Rab GTPases and their effector proteins. For instance, Rab GTPases located on the surface of vesicles specifically interact with tethering complexes present at the target membrane, allowing for vesicle docking and subsequent membrane fusion via soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complex formation. Overall, these Rab GTPases and tethering complex proteins have been well characterized in yeast and animal systems, but much less is known about their roles in plants (reviewed in Carter et al., 2004a; Jürgens, 2004; Sutter et al., 2006).

One conserved complex involved in the tethering of Golgi-derived secretory vesicles at the plasma membrane is the exocyst complex (reviewed in Hsu et al., 2004; Munson & Novick, 2006). The exocyst complex was first identified in Saccharomyces cerevisiae, where it plays a role in secretion and polarized growth during budding (TerBush & Novick, 1995; TerBush et al., 1996). A homologous complex was subsequently identified in mammalian cells (Hsu et al., 1996; Kee et al., 1997). In both yeast and mammals, the exocyst complex is comprised of eight subunits: Secretory (Sec)3p, Sec5p, Sec6p, Sec8p, Sec10p, Sec15p, Exocyst (Exo)70p and Exo84p; that localize to specific regions of the plasma membrane where active vesicle fusion events, such as those occurring during polarized secretion, take place (Finger et al., 1998; Guo et al., 1999; Boyd et al., 2004). For example, in mammalian PC12 cells, Exo70 localizes to the growth cone during neurite outgrowth in response to nerve growth factor. In the absence of the signal for exocyst-mediated vesicle tethering, Exo70 is found in the perinuclear region of the undifferentiated PC12 cells (Vega & Hsu, 2001). The mammalian exocyst complex also plays a role in the selective tethering of vesicles to either the apical or the basolateral membranes (reviewed in Fölsch, 2005; Oztan et al., 2007), and the regulated vesicle trafficking of Glucose Transporter-4 (GLUT4) to the plasma membrane surface in response to insulin (Inoue et al., 2003).

Based on sequence homology, plant genomes are predicted to contain all of the eight exocyst complex subunits. Interestingly, in contrast to yeast and animal genomes which contain single copies of each exocyst component, the putative plant exocyst genes are encoded by expanded gene families (Elias et al., 2003; Synek et al., 2006). The role of these predicted exocyst complex proteins in plant growth and development, however, has only recently begun to be uncovered. For instance, a few plant mutants with defects in exocyst subunit genes have been identified and display abnormalities in tissues that undergo polarized secretion. For example, several Arabidopsis thaliana sec mutants (sec5a/5b, sec6, sec8 and sec15a) displayed reduced pollen tube growth rates in comparison to wild-type plants, resulting in a partial transmission defect (Cole et al., 2005; Hála et al., 2008). In maize (Zea mays), a transposon insertion disrupting a Sec3 gene results in a roothairless1 (rth1) mutation with defects in root hair elongation (Wen et al., 2005). An A. thaliana exo70A1 mutant displays a similar reduced root hair growth phenotype, in addition to shortened hypocotyls, alterations in inflorescence architecture, and impaired flower development (Synek et al., 2006). Also, Exo70A1 is essential in the stigmatic papillae of Brassica and A. thalianato promote compatible pollen hydration and pollen tube penetration through the stigmatic surface (Samuel et al., 2009). Synek et al. (2006) reported that insertional mutations in the AtExo70B2, AtExo70D3, AtExo70F1, AtExo70G1 and AtExo70H7 genes did not yield any obvious morphological phenotypes, despite all being highly expressed. Interestingly, mutations in A. thaliana AtExo70A1 with either Sec5A or Sec8 displayed a synergistic effect on hypocotyl elongation, indicating that a functional relationship exists between these two subunits in vivo (Hála et al., 2008). Indeed, biochemical data presented in the same study indicated that these and other components of the exocyst localize as a complex to the tips of growing pollen (Hála et al., 2008).

In this study, we have conducted a survey of the different plant exocyst complex genes by building phylogenies of the predicted exocyst complex subunit families from four whole sequenced plant genomes, including the genomes of A. thaliana, Oryza sativa and Populus trichocarpa, and the recently released Physcomitrella patens (moss) genome (Rensing et al., 2008). Detailed expression profiling was also conducted for the predicted A. thaliana exocyst complex genes, and finally, the subcellular distributions of the different A. thaliana exocyst complex subunit proteins were investigated in transient expression studies in suspension-cultured cells. Interestingly, two A. thaliana Exo70 members were found to co-localize with trans-Golgi network/endosomal marker proteins and recruited other exocyst complex subunits to these compartments. These localization patterns are consistent with the proposed role of the plant exocyst complex in vesicular trafficking with the recycling endosomal compartment (Zárskýet al., 2009).

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Identification of predicted plant exocyst complex genes and microarray expression profiling

Published yeast and mammalian exocyst complex subunit sequences were obtained from GenBank (National Centre for Biotechnology Information (NCBI), Supporting Information Table S1) to use in BLASTp searches for related predicted proteins in the Arabidopsis thaliana genome. Using the predicted A. thaliana exocyst complex subunit sequences, BLASTp analyses were then conducted against the predicted proteomes of poplar (Populus trichocarpa) (http://genome.jgi-psf.org/Poptr1_1/Poptr1_1.home.html), Physcomitrella patens(moss) (http://genome.jgi-psf.org/Phypa1_1/Phypa1_1.home.html), and rice (Oryza sativa) (http://rice.tigr.org/). All final data sets were downloaded in March 2007. All the identified plant, animal and yeast exocyst complex subunit sequences were then used in PSI-BLAST analyses against all the predicted plant proteomes to ensure that all the predicted exocyst complex subunit sequences were identified. Using the Blast Like Alignment Tool (BLAT) (Kent, 2002), genome locations for the predicted plant exocyst complex genes were identified, and this information was used to remove duplicates or incorrectly predicted gene models. Multiple sequence alignments were generated using muscle 3.6 (Edgar, 2004), and unreliably aligned regions and gaps were removed. Programs from the phylip suite were used for tree building (Felsenstein, 1993). seqboot was used to create the 10 000 bootstrapping alignments, which were run with proml to generate 10 000 trees. consense was used to combine the 10 000 trees into one tree with bootstrap support values for each of the branches. This tree was then inputed again into proml to determine branch lengths, and the resulting tree was displayed using TreeDyn (http://www.treedyn.org/) and Adobe Illustrator 11.0 (Adobe Systems, San Jose, CA, USA).

Publicly available A. thaliana microarray expression data sets were explored for the predicted A. thaliana exocyst complex genes using the Genevestigator suite of tools (Zimmermann et al., 2004) and the tools hosted at the BioArray Resource (BAR) (Toufighi et al., 2005). The various data sets were formatted into heat maps using the Meta Analyzer tool as hosted at the BioArray Resource (Toufighi et al., 2005). Expression patterns under different conditions and treatments were expressed as log-transformed values normalized to the controls. Additional expression data for A. thaliana genes and rice genes were obtained using the massively parallel signature sequencing (MPSS) rice and A. thaliana databases (Meyers et al., 2004). Searches against the dbEST database (http://www.ncbi.nlm.nih.gov/dbEST/) hosted at NCBI were conducted using the plant Exo70 sequences. Expressed sequence tags (ESTs) that were identified were then subjected to a BLAST search against the identified Exo70 sequences to determine genuine EST hits.

Cloning of the predicted A. thaliana exocyst complex cDNAs

The cDNAs encoding full-length open reading frames (ORFs) for A. thaliana Exo70A1 (At5g03540), Sec3a (At1g47550), Sec3b (At1g47560), Sec5a (At1g76850), Sec6 (At1g71820), Sec8 (At3g10380), Sec15a (At3g56640), Sec15b (At4g02350) and Exo84b (At5g49830) were amplified using the polymerase chain reaction (PCR) from Columbia-0 leaf tissue cDNA. Before subcloning into their destination vectors, the PCR-amplified cDNA products were first cloned into the pCR2.1 TOPO vector (Invitrogen) as per the manufacturer’s instructions. The ORFs were then transferred to the pRTL2ΔNS/mGFP:MCS vector (Shockey et al., 2006), a plant expression vector containing a modified version of the green fluorescent protein (GFP) in which the protein’s dimerization domain was disrupted by mutating the leucine at position 221 to lycine (Zacharias et al., 2002). All ORFs were cloned as N-terminal fusions with GFP, with AtExo84b, AtSec8, AtSec15a and AtSec15b ORFs being cloned into pRTL2ΔNS/mGFP:MCS as BamH1 fragments, while AtSec5a and AtSec6 ORFs were digested with XmaI and cloned into the XmaI (SmaI) site of pRTL2ΔNS/mGFP:MCS. Full-length and error-free AtSec3 and AtSec10 cDNAs could not be isolated, and thus were not included in this study.

cDNAs coding for full-length A. thaliana Exo70B2 (At1g07000), Exo70C2 (At5g13990), Exo70D2 (At1g54090), Exo70E2 (At5g61010), Exo70G1 (At4g31540), Exo70H1 (At3g55150) and Exo70H7 (At5g59730) were obtained from the Arabidopsis Biological Resource Center (ABRC; Ohio State University). The ORFs of AtExo70A1, AtExo70B2, AtExo70C2, AtExo70E2, AtExo70G1, AtExo70H1 and AtExo70H7 were amplified using PCR along with the appropriate forward and reverse primers that introduced in-frame NheI sites at the 5′ and 3′ ends. Next the NheI-digested PCR products were cloned into NheI-digested pRTL2ΔN/S myc-MCS, a plant expression vector that includes the 35S cauliflower mosaic virus promoter and sequences encoding an initiator methionine, glycine linkers, and the myc epitope tag (underlined in MGEQKLISEEDLG), followed by a multiple cloning site. pRTL2ΔN/S myc-MCS was constructed by inserting sequences encoding NheI, KpnI and XmaI restriction sites immediately downstream of the XbaI site in pRTL2/myc-X (Shockey et al., 2006). Similarly, AtExo70A1, AtExo70D2 PCR-amplified ORFs were cloned into the KpnI and XmaI sites in pRTL2ΔN/S myc-MCS. The myc-tagged version of AtSec15b was constructed by ligating the AtSec15b BamH1 fragment from pRTL2ΔNS/mGFP:Sec15b into BamH1-digested pRTL2/myc-X. The red fluorescent protein (RFP) versions of AtExo70G1 and AtExo70E2 were constructed using pRTL2/RFP-MCS, a plant expression vector containing the RFP ORF (but lacking a stop codon) followed by a multiple cloning site (Shockey et al., 2006). Specifically, for AtExo70G1, PCR amplification was used to add SacI sites flanking the ORF, and resulting PCR product was then cloned into the SacI site of pRTL2/RFP-MCS. For AtExo70E2, the ORF was excised with EcoRI, and pRTL2/RFP-MCS was digested with XbaI. The two DNA samples were treated with the Klenow fragment and then blunt end ligated together. Complete details of the template DNAs and oligonucleotide primers used in these and other cloning procedures are available upon request. Plasmids encoding GFP fused to the N-terminus of the A. thaliana SNARE protein Syntaxin of Plants (Syp)-21 (GFP:Syp21), Syp42 (GFP:Syp42) or Syp52 (GFP:Syp52) were kindly provided by Dr Masa H. Sato (Uemura et al., 2004).

Transient expression of A. thaliana exocyst complex subunits in Tobacco Bright Yellow-2 (BY-2) cells

Tobacco (Nicotiana tabacum cv. BY-2) suspension-cultured cells were maintained and prepared for biolistic bombardment as described previously (Banjoko & Trelease, 1995). Briefly, transient transformations were performed using 2 μg of plasmid DNA (or 2 μg of each plasmid for co-transformations) with the biolistic particle delivery system 1000/HE (Bio-Rad Laboratories, Mississauga, Canada). Bombarded cells were incubated for 6, 8 or 20 h to allow gene expression and sorting of the protein product(s), and then either viewed immediately (refer to Fig. S4) or fixed in formaldehyde, incubated with 0.01% (w/v) pectolyase Y-23 (Kyowa Chemical Products, Osaka, Japan), and permeabilized with 0.3% (v/v) triton X-100 (Sigma-Aldrich Ltd) (Lee et al., 1997).

Immunostaining of fixed and permeabilized cells was performed as previously described (Trelease et al., 1996). Antibodies used were as follows: mouse anti-myc antibodies in hybridoma medium (clone 9E10; Princeton University, Monoclonal Antibody Facility, Princeton, NJ, USA); goat anti-mouse Alexa Fluor 488 immunoglobulin G (IgG) (Jackson Immunoresearch Laboratories, West Grove, PA, USA), and goat anti-mouse rhodamine red-X IgG (Jackson ImmunoResearch Laboratories). Concanavalin A conjugated to Alexa 594 (Molecular Probes, Eugene, OR, USA) was added to BY-2 cells at a final concentration of 5 μg ml−1 during the final 20 min of incubation with secondary antibodies.

Epifluorescent images of BY-2 cells were acquired using a Zeiss Axioskop 2 MOT epifluorescence microscope (Carl Zeiss, Toronto, Canada) with a Zeiss ×63 Plan Apochromat oil-immersion objective. Image captures were performed using a Retiga 1300 charge-coupled device camera (Qimaging, Burnaby, Canada) and Northern Eclipse 7.0 software (Empix Imaging, Mississauga, Canada). Confocal laser-scanning microscopy (CLSM) images of BY-2 cells were acquired using either a LSM510 (Carl Zeiss) with a Zeiss ×63 Plan-Apochromat oil-immersion objective or a Leica DM RBE microscope with a Leica ×63 Plan Apochromat oil-immersion objective, a Leica TCS SP2 scanning head, and the Leica tcs nt software package (Version 2.61) (Leica, Heidelberg, Germany). All fluorescence images of cells shown in individual figures are representative of > 50 independent (transient) transformations from at least two independent transformation experiments. Figure compositions were generated using Adobe Photoshop CS (Adobe Systems).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Identification of predicted plant exocyst complex subunits

To examine the evolutionary relationships among predicted exocyst complex subunits in plant, yeast and animal genomes, BLASTp searches were conducted on the predicted proteomes of the four complete plant genomes: those of A. thaliana (Arabidopsis Genome Initiative, 2000), rice (O. sativa; Yuan et al., 2003), poplar (P. trichocarpa; Tuskan et al., 2006), and moss (P. patens; Rensing et al., 2008). The amino acid sequences from the yeast and mammalian exocyst complex subunits (Exo70, Sec3, Sec5, Sec6, Sec8, Sec10, Sec15 and Exo84) were first used to BLASTp search the predicted A. thaliana proteome. The predicted exocyst complex subunits emerging from this search were then used in BLASTp searches of the rice, poplar and moss genomes. While previous phylogenies have been published on predicted plant Exo70 proteins (Elias et al., 2003; Synek et al., 2006), updated gene models used in this study resulted in some notable differences in the Exo70 families. Similar to the previous studies, the plant Exo70 genes were found to be expanded relative to their yeast, Drosophila, and mammalian counterparts (Fig. 1, Table 1). With the exception of Sec6, gene duplications were also found in the remaining predicted plant exocyst subunits (Table 1, Fig. 2). Duplications of plant genes are not uncommon as polyploidization events have occurred in the evolution of almost every angiosperm (Paterson, 2005). However, the level of expansion was unique for Exo70 in comparison to the other exocyst complex subunit genes, and, interestingly, this expansion was also detected in moss, a nonvascular land plant (Table 1). However, while the moss genome is predicted to have approximately the same number of genes as the A. thaliana genome, it only appears to encode 13 Exo70 proteins, whereas the A. thaliana proteome has 23 predicted Exo70 proteins (Table 1). Many of the plant Exo70 genes are expressed, with ESTs identified for 32 rice Exo70 genes, 10 P. patens Exo70 genes, and six poplar Exo70 genes (Fig. 1). The majority of the A. thaliana Exo70 genes are also expressed, with more detailed expression described in Fig. 3.

image

Figure 1.  Phylogenetic tree of predicted Exocyst (Exo)70 proteins. A maximum likelihood phylogenetic tree was generated with 100 plant Exo70 protein sequences and rooted with three animal and one fungal Exo70 protein sequences. Exo70 protein sequences were identified from four predicted proteomes of different plant species with completely sequenced genomes. Arabidopsis thaliana and Populus trichocarpa are both predicted to encode 23 Exo70 genes, Oryza sativa is predicted to have 41 Exo70 genes, and Physcomitrella patens is predicted to have 13 Exo70 genes. For A. thaliana and O. sativa, the assigned gene identifiers for the genes are displayed on the tree. The P. trichocarpa and P. patens genomes were recently released, and the names currently assigned to the predicted gene models are shown. The underlined A. thaliana Exo70 proteins were further investigated in this study. The number of predicted exons for each Exo70 gene is summarized in the first column on the right-hand side. To assess if the genes of these predicted proteins were transcribed, searches against expressed sequence tag (EST) databases (National Centre for Biotechnology Information) were conducted and the results are summarized on the second column on the right-hand side. In the far right-hand side column are the names that were assigned to exocyst complex genes from A. thaliana and O. sativa in Cole et al. (2005) and Synek et al. (2006). At the branch nodes are 10 000 bootstrap replicates represented as percentage values.

Download figure to PowerPoint

Table 1.   Predicted copy number of exocyst subunit genes in plant genomes
 MossRiceArabidopsis thalianaPoplar
  1. Exo, Exocyst; Moss, Physcomitrella patens; rice, Oryza sativa; poplar, Populus trichocarpa; Sec, Secretory.

Exo7013412323
Sec33221
Sec54122
Sec61111
Sec83112
Sec103112
Sec152323
Exo847436
image

Figure 2.  Phylogenetic trees of predicted Secretory (Sec)3, Sec5, Sec6, Sec8, Sec10, Sec15 and Exo84 proteins. Maximum likelihood phylogenetic trees were constructed for (a) Sec3, (b) Sec5, (c) Sec6, (d) Sec8, (e) Sec10, (f) Sec15 and (g) Exocyst (Exo)84 amino acid sequences from Arabidopsis thaliana, Populus trichocarpa, Oryza sativa and Physcomitrella patens. For A. thaliana and O. sativa, the assigned gene identifiers for the genes are displayed in parentheses on the tree. The P. trichocarpa and P. patens genomes were recently released, and the names currently assigned to the predicted gene models are shown. The Sec3 gene from P. trichocarpa was incorrectly predicted as two gene models, and the names of both gene models are included. The underlined A. thaliana exocyst complex proteins were further investigated in this study. At the branch nodes are 10 000 bootstrap replicates.

Download figure to PowerPoint

image

Figure 3.  Expression analysis of the Arabidopsis thaliana exocyst. (a) Expression profiles of the A. thaliana exocyst complex genes in different tissues. To visualize the expression of the A. thaliana exocyst components in various tissue types, e-northern analyses were conducted using the Genevestigator Meta Analyser-Plant Organs data set (Zimmermann et al., 2004) and formatted using the Meta Analyzer at the BioArray Resource (BAR) site (Toufighi et al., 2005). (b) Expression profiles of selected A. thaliana Exo70 genes under different conditions: hormone treatments, pathogen infections and elicitors, and abiotic stress treatments. Data from the AtGenExpress series (Schmid et al., 2005; Kilian et al., 2007) were analysed using BAR (Toufighi et al., 2005). Expression values (compared with the controls) were retrieved and formatted using the Meta Analyzer hosted at BAR (Toufighi et al., 2005). See Supporting Information Figs S1–S3 for more detailed expression profiles.

Download figure to PowerPoint

Evolutionary relationships among predicted plant exocyst complex subunits

To explore the evolutionary relationships among the 100 predicted plant Exo70 proteins, maximum likelihood analyses were performed and rooted with one yeast and three animal sequences (Fig. 1). The resulting tree is an updated version of the previously published Exo70 trees (Elias et al., 2003; Synek et al., 2006), and the Exo70.1, Exo70.2 and Exo70.3 clades described in these previous trees have been well conserved here. However, the tree shown in Fig. 1 includes two additional predicted rice Exo70 genes and two fewer predicted poplar Exo70 genes. In addition, slight differences among the positions of ancestral lineages were observed. The tree topology generated from our analyses demonstrates strong bootstrap support for the arrangement of these clades (Fig. 1).

Phylogenetic analysis of the putative Exo70 amino acid sequences indicated that plants appeared to have acquired numerous Exo70 genes via ancient duplication events in a common ancestor (for the four plant lineages used in this study) as well subsequent duplication events in the independent plant lineages. This has been observed across all three major Exo70 clades, but a specific example of this can be seen in clade Exo70.1, where a duplication event of the ancestral Exo70.1 gene has resulted into a Physcomitrella subclade, distinct from the angiosperm Exo70 genes, and subsequently further duplication events have occurred within the nonvascular (P. patens), monocot (rice) and eudicot (A. thaliana and poplar) lineages. Within each of the three major clades of Exo70 genes, P. patens duplicates are monophyletic, while duplicates from the angiosperm exocyst components are paraphyletic. It is also interesting to note that, within the Exo70.2 clade, there is a major subclade of 16 rice Exo70 genes that appear to be undergoing rapid divergence, resulting in this subclade being distinct from the others. Individual maximum likelihood analyses were also conducted for the Sec3, Sec5, Sec6, Sec8, Sec10, Sec15 and Exo84 loci (Fig. 2). Similar trends were also seen in the Sec15 and Exo84 phylogenies, with the P. patens duplicates being monophyletic, while the angiosperm Sec15 and Exo84 duplicates were paraphyletic (Fig. 2).

Polyploidization/whole-genome duplications are common and have been key events in the evolution of almost all plants (Blanc & Wolfe, 2004). While whole-genome duplications can explain the presence of some duplicates in the plant exocyst component genes, other events have also led to gene duplications, such as unequal crossing over, retrotransposition and segmental chromosomal duplication (Zhang, 2003). Some key features that help identify gene duplications acquired through retrotransposition include the absence of introns, poly A tracts and loss of cis-regulatory sequences. Often genes acquired in this fashion cannot be expressed because of a lack of appropriate cis-elements and therefore become pseudogenes (Zhang, 2003). The plant Exo70 genes that reside in clades Exo70.2 and Exo70.3 have significantly fewer introns than those in clade Exo70.1 (Fig. 1). Of the A. thaliana Exo70 genes that reside in clades Exo70.2 and Exo70.3, three genes contain one intron while seventeen do not have any introns. By contrast, AtExo70A1, AtExo70A2 and AtExo70A3 reside in clade Exo70.1 and contain 12, 13 and 8 exons, respectively. A similar trend is observed in Exo70 genes found in rice, poplar and P. patens, with Exo70 genes that reside in clades Exo70.2 and Exo70.3 generally having no more than two introns. The exceptions are three rice Exo70 genes that have up to five introns and four P. patens genes that have up to eight introns, and these sequences cluster together in distinct subclades within Exo70.2 and Exo70.3, respectively (Fig. 1). Searches in the untranslated regions of the genes in clades Exo70.2 and Exo70.3 did not uncover long obvious tracts of poly A regions. It does, however, seem possible that ancient retrotransposition events could have been responsible for producing the common ancestor(s) of clades Exo70.2 and Exo70.3.

Unequal crossing over events allow duplicate genes to be in tandem, such as AtSec3a and AtSec3b, which share 91% sequence identity, whereas the presence of multiple copies of AtSec5, AtSec15 and AtExo84 appears to be the result of chromosomal/genome duplication events. Gene silencing through mutational decay is to be expected for a high percentage of duplicated genes and pseudogenization usually occurs within the first few million years subsequent to a duplication event (Lynch & Conery, 2000). Arabidopsis thaliana is reported to have undergone one genome duplication event 25–100 Ma (Vision et al., 2000; Blanc et al., 2003) and most of the genetic redundancy acquired appears to have been selectively lost (Blanc & Wolfe, 2004). The resulting single copies of AtSec6, AtSec8 and AtSec10 are probably a consequence of loss of duplicated copies of these genes through mutational decay. The presence of fewer AtSec and AtExo84 genes implies that duplicates (in addition to those currently observed) of these genes acquired through doubling events have also been selectively lost and only excess copies of AtExo70 genes have been selectively maintained.

Duplicated genes can have a number of fates: acquisition of novel function, silencing of one of the copies or retention of original or similar function, also known as subfunctionalization (Wendel, 2000). Duplicated genes are unlikely to maintain identical functions unless it is advantageous to have excess amounts of that gene product (Nowak et al., 1997). Maintenance of duplicated genes through subfunctionalization is thought to happen quite quickly after duplication events (Force et al., 1999) and can be achieved simply by differential expression, whether spatial or temporal (Gu et al., 2002). Subfunctionalization can also be achieved through alteration of function where the products of the duplicated genes assume part of the parental gene’s function (Zhang, 2003). Changes in the expression of genes can arise from multiple events such as sequence deletion, inter-chromosomal exchange or chromosomal rearrangement events leading to alterations of cis-regulatory elements, which ultimately can lead to different trans-regulatory factors (Chen & Ni, 2006). As seen in the expression analysis (in the next section) of the putative A. thaliana exocyst, differential expression in terms of timing and tissue type has probably played a significant role in the subfunctionalization of the putative A. thaliana Exo70 genes.

Plant exocyst expression analysis

Publicly available microarray data sets were searched to uncover the expression profiles for the A. thaliana exocyst complex genes across different tissues and treatments (Zimmermann et al., 2004; Schmid et al., 2005; Toufighi et al., 2005; Kilian et al., 2007). Electronic northern analyses of the A. thaliana transcriptome revealed that the Sec3, Sec5, Sec6, Sec8, Sec10, Sec15, and Exo84 genes were largely ubiquitously expressed in all plant organs as well as at different stages of plant development (Fig. 3a). When comparing the expression profiles for the AtSec5 and AtSec15 gene pairs, AtSec5b and AtSec15a expression levels were considerably lower across all tissues in comparison to their paralogues, AtSec5a and AtSec15b, respectively. For the AtSec3 gene pair, the oligonucleotide probe on the ATH1 whole-genome chip detects both AtSec3a and AtSec3b, and so the two genes could not be differentiated. However, AtSec3b is expressed, as the cDNA was isolated by RT-PCR, and searches of the A. thaliana MPSS database yielded unique signatures, particularly in A. thaliana tissues treated with salicylic acid (Meyers et al., 2004). AtExo84a was not present on the ATH1 whole-genome chip, and we were unable to amplify the cDNA through RT-PCR. Also, searches of the A. thaliana MPSS database did not yield any signatures for AtExo84a.

Electronic northern analyses of the A. thaliana Exo70 genes revealed that several Exo70 members are ubiquitously expressed while a few show more tissue-specific patterns of expression (Fig. 3a). For example, AtExo70A1, AtExo70B1 and AtExo70H7 are broadly expressed across a number of different tissues, while AtExo70A2, AtExo70C1, AtExo70C2, AtExo70H3 and AtExo70H5 are expressed in fewer tissues with highest levels detected in the stamen and pollen. AtExo70H6 was the only A. thaliana Exo70 gene that is not represented on the ATH1 whole-genome chip, but previous reports indicated that ESTs were available for this gene (Synek et al., 2006). AtExo70A3 expression was not detected in the different tissues surveyed in Fig. 3a and was previously reported as a potential pseudogene (Synek et al., 2006). However, searches conducted using the AtGenExpress tissue data set (Schmid et al., 2005) found AtExo70A3 to be expressed during embryogenesis, with the highest levels detected in the torpedo-stage root. Also, AtExo70A3 was found to be up-regulated in response to treatment with hormones such as ABA, 1-aminocyclopropane-1-carboxylate (ACC), zeatin and methyl jasmonate (Fig. 3b and Fig. S1a). A more extensive survey of the microarray expression databases for hormone and inhibitor treatments, pathogen infections, and abiotic stress treatments revealed that several of the A. thaliana Exo70 genes had increased expression levels for a number of these treatments (Fig. 3b and Figs S1–S3). Members of the Exo70 gene family appeared to be more responsive to these treatments while the remaining exocyst complex genes showed very little variation in mRNA levels (Figs S1–S3). The AtExo70 genes that reside in the A and H subclades appeared to be the most responsive, showing increases in mRNA levels when treated with hormones, abiotic stresses and pathogen infections. In particular, several members of the AtExo70H subclade were highly up-regulated during pathogen infections and elicitor treatments (Fig. 3b). The Sec3, Sec5, Sec6, Sec8, Sec10, Sec15 and Exo84 genes did not appear to show any significant expression changes in response to various treatments, exhibiting at most a twofold change in expression (Figs S1,S2).

In a comparison of the plant Exo70 phylogenetic tree with the expression profiles, a clear correlation was observed for the AtExo70C1 and AtExo70C2 subclades. These two Exo70 genes cluster together based on sequence and expression, with both being highly expressed in the stamen and pollen. Analysis of rice MPSS databases also revealed unique 20-bp signatures for OsExo70C1 and OsExo70C2 in mature pollen grains, suggesting conservation of expression patterns across monocots and dicots. It is probable that a gene duplication event before the monocot–dicot divergence produced this subclade which is found in A. thaliana, rice and poplar. Consistent with the idea of differential expression playing a significant role in subfunctionalization, it is interesting to note that, within some of the AtExo70 subclades, there are very divergent expression patterns, with the most notable examples being the AtExo70A and AtExo70H subclades. This is also observed for the AtSec5 and AtSec15 gene pairs.

Subcellular localization patterns of the putative A. thaliana exocyst complex proteins in transiently transformed tobacco BY-2 cells

In yeast and animal systems, specific subcellular localization patterns have been observed for exocyst complex proteins during processes of polarized or regulated secretion (Finger et al., 1998; Guo et al., 1999; Boyd et al., 2004). Also, some of the exocyst complex proteins (Sec6, Sec8 and Exo70A1) were shown to localize to the tips of growing pollen tubes in A. thaliana (Hála et al., 2008). To characterize the A. thaliana exocyst complex proteins in more detail, we conducted a detailed analysis of their subcellular localization patterns in tobacco BY-2 suspension-cultured cells, a well-characterized in vivo protein trafficking system (Banjoko & Trelease, 1995; Brandizzi et al., 2003; Miao & Jiang, 2007). Specifically, N-terminal GFP-tagged, full-length versions of AtSec5a, AtSec6, AtSec8, AtSec15a, AtSec15b and AtExo84b were expressed individually (via biolistic bombardment) in BY-2 cells and their resulting subcellular localization patterns were observed by standard epifluorescence microscopy. In order to avoid any potential mislocalizations as a result of protein overexpression, only cells displaying weak GFP fluorescence were assessed at 6 h after bombardment. While both AtSec6 and AtSec8 displayed a diffuse fluorescence pattern throughout the cytosol (Fig. 4a–d), AtSec5a, AtSec15a, AtSec15b and AtExo84b were localized to globular structures in the cytosol (Fig. 4f,j,n,r). To visualize this region of the cell more clearly, fluor-conjugated concanavalin A (ConA) was used to stain the endoplasmic reticulum (ER), including the nuclear envelope, in cells expressing GFP-tagged AtSec5a, AtSec15a, AtSec15b and AtExo84b (Fig. 4g,k,o,s). An examination of the resulting merged images revealed that the globular structures localized to the perinuclear region, but did not show good co-localization with the ConA-stained ER (Fig. 4h,l,p,t).

image

Figure 4.  Subcellular localization of A. thaliana Sec and Exo84 proteins. Arabidopsis thaliana GFP:Sec/Exo84 fusion proteins were transiently expressed in tobacco BY-2 cells for 6 h, fixed in formaldehyde and then processed for epifluorescence microscopy. ConA is an endogenous endoplasmic reticulum (ER) stain conjugated to Alexa 594. Bars, 10 μm. (a, b) Differential interference contrast (DIC) and GFP:Sec6, respectively; (c, d) DIC and GFP:Sec8, respectively; (e–h) DIC, GFP:Sec5a, ConA and merged image of (f) and (g), respectively; (i–l) DIC, GFP:Sec15a, ConA and merged image of (j) and (k), respectively; (m–p) DIC, GFP:Sec15b, ConA and merged image of (n) and (o), respectively; (q–t) DIC, GFP:Exo84b, ConA and merged image of (r) and (s), respectively. Exo, exocyst; Sec, secretory.

Download figure to PowerPoint

For the A. thaliana Exo70 family, eight Exo70 proteins representing different branches of the Exo70 phylogenetic tree were selected for analysis (Fig. 1, underlined). Briefly, a c-myc epitope tag was added to the N-termini of full-length AtExo70A1, AtExo70B2, AtExo70C2, AtExo70D2, AtExo70E2, AtExo70G1, AtExo70H1 and AtExo70H7 proteins, and each construct was transiently expressed individually in tobacco BY-2 cells, followed by immunostaining and confocal laser-scanning microscopic (CLSM) analysis. Overall, subcellularly, myc-tagged AtExo70A1, AtExo70B2, AtExo70C2, AtExo70D2, AtExo70H1 and AtExo70H7 were mostly distributed throughout the cytosol and within the nucleus (Fig. 5b,d,f,h,n,p). These proteins also localized to numerous small punctate structures (Fig. 5b,d,f,h,n,p) that did not co-localize with known mitochondrial or peroxisomal marker proteins. A similar punctate pattern of localization was previously observed for a GFP-tagged AtExo70 transiently expressed in A. thaliana protoplasts (Elias et al., 2003). By contrast, myc-tagged AtExo70E2 and AtExo70G1 appeared to localize to well-defined larger punctate structures throughout the cytosol (Fig. 5j,l).

image

Figure 5.  Subcellular localization of Arabidopsis thaliana exocyst (Exo)70 proteins. BY-2 cells were transformed transiently with myc:Exo70 fusion proteins, fixed in formaldehyde at 6 h post biolistic bombardment, and then processed for immunofluorescence confocal laser scanning microscopy (CLSM). Bars, 10 μm. (a, b) Differential interference contrast (DIC) and myc:Exo70A1, respectively; (c, d) DIC and myc:ExoB2, respectively; (e, f) DIC and myc:Exo70C2, respectively; (g, h) DIC and myc:Exo70D2, respectively; (i, j) DIC and myc:Exo70E2, respectively; (k, l) DIC and myc:Exo70G1, respectively; (m, n) DIC and myc:Exo70H1, respectively; (o, p) DIC and myc:Exo70H7, respectively.

Download figure to PowerPoint

To identify the subcellular structures to which AtSec15b (globular structures in the perinuclear region (Fig. 4j)) and AtExo70E2 (punctate structures throughout the cytosol (Fig. 5j)) localized, we compared their localization to the localization of several different endocytic pathway marker proteins, namely the A. thaliana SNAREs Syp21, Syp42 and Syp52 (Sanderfoot et al., 2001; Samaj et al., 2005). Syp21 is localized to the late endosome/prevacuolar compartment (PVC) and is believed to be involved in Golgi-to-PVC vesicular trafficking (Sanderfoot et al., 2001; Foresti et al., 2006). In contrast, Syp42 has been shown to reside in the trans-Golgi network (TGN) (Bassham et al., 2000; Sanderfoot et al., 2001; Uemura et al., 2004) which is also considered to function as an early endosomal compartment (Dettmer et al., 2006). Finally, Syp52 localizes to the TGN and the PVC (Sanderfoot et al., 2001) as well as the vacuole where it is proposed to be involved in vacuolar membrane fusion (Carter et al., 2004b; Uemura et al., 2004). As shown in Fig. 6, GFP-tagged versions of Syp21, Syp42 and Syp52 that were co-expressed with myc:Sec15b or myc:Exo70E2 in BY-2 cells displayed significant overlapping patterns of localization, as assessed by immunofluorescence CLSM. Specifically, when Syp21was co-expressed with either AtSec15b (Fig. 6b,c) or AtExo70E2 (Fig. 6n,o), partial co-localizations (visualized as yellow/orange colour) were observed in the merged images (Fig. 6d,p). A high degree of partial co-localization was also observed when Syp42 was co-expressed with either AtSec15b (Fig. 6f–h) or AtExo70E2 (Fig. 6r–t). Similarly, a partial overlap in localization patterns was observed when Syp52 was co-expressed with AtSec15b (Fig. 6j–l) or AtExo70E2 (Fig. 6v–x). AtExo70G1 was found to have behaved similarly to AtExo70E2; for example, an RFP-tagged version of Exo70G1 (RFP:Exo70G1) showed complete co-localization with myc:Exo70E2 (Fig. 6z–bb) and a large degree of overlapping localization with co-expressed GFP:Syp42 (Fig. 6dd–ff). Taken together, the co-localization of AtSec15b and AtExoE2/G1 with the various SNARE proteins suggests that these exocyst subunits localize primarily to TGN/early endosome compartments, but are also present in other compartments of the endocytic pathway (e.g. late endosomes/prevacuolar compartments).

image

Figure 6.  Co-expression of AtSec15b, AtExo70E2, and AtExo70G1with the endosomal marker soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins Syp21, Syp42, and Syp52. BY-2 cells were co-transformed transiently with selected exocyst subunits and GFP:Syp fusion proteins, fixed in formaldehyde at 6 h post biolistic bombardment, and then processed for immunofluorescence confocal laser scanning microscopy (CLSM). Syp21 is localized to the late endosome/prevacuolar compartment (PVC), Syp42 is found primarily in the early endosome/TGN, and Syp52 localizes to both the PVC and trans-Golgi network (TGN) (Sanderfoot et al., 2001; Samaj et al., 2005). As a control, Supporting Information Fig. S4 shows that the localization patterns of co-expressed RFP:Exo70E2 and GFP:Syp42 are similar in tobacco BY-2 and Arabidopsis thaliana suspension cells that were either formaldehyde-fixed or living during CLSM imaging. (a–d) myc:Sec15b with GFP:Syp21; (e–h) myc:Sec15b with GFP:Syp42; (i–l) myc:Sec15b with GFP:Syp52; (m–p) myc:Exo70E2 with GFP:Syp21; (q–t) myc:Exo70E2 with GFP:Syp42; (u–x) myc:Exo70E2 with GFP:Syp52; (y–bb) RFP:Exo70G1 with myc:Exo70E2; (cc–ff) RFP:Exo70G1 with GFP:Syp42. Exo, exocyst; Sec, secretory; Syp, Syntaxin of Plants.

Download figure to PowerPoint

Interestingly, similar co-localization patterns were observed between AtSec15b and AtExoE2/G1 and the co-expressed A. thaliana SNAREs despite the different structural patterns observed for AtSec15b and AtExoE2/G1 when they were expressed alone (cf. AtSec15b- and AtExoE2/G1-transformed cells in Figs 4,5 and 6), a difference that may be attributable to the effects of ecotopic expression of individual exocyst subunits on the structural configuration and/or distribution of the TGN/endosomal compartments. For example, treatment of BY-2 cells with the vesicle-trafficking inhibitor brefeldin A (BFA) induces the formation of a so-called ‘BFA compartment’ (Tse et al., 2006; Müller et al., 2007; Robinson et al., 2008) which consists of endosomal and TGN membranes and which displays a perinuclear globular-like fluorescence pattern similar to that observed for AtSec15b.

Co-localization of putative A. thaliana exocyst complex subunits

Given the ability of AtExo70E2 to localize to endocytic compartments, particularly the TGN/early endosome, we also tested whether AtExo70E2 could influence the subcellular localization patterns of other exocyst complex proteins. To test this, the GFP-tagged versions of AtSec5a, AtSec6, AtSec8, AtSec15a, AtSec15b and AtExo84b were each co-expressed with myc:Exo70E2 in BY-2 cells, and localization patterns were examined using immuno-epifluorescence microscopy. As shown in Fig. 7, AtExo70E2 had a distinct effect on the cellular distribution of most of the co-expressed exocyst complex proteins tested. When individually expressed, AtSec5a, AtSec15a, AtSec15b and AtExo84b were localized to perinuclear globular structures (Fig. 4). The addition of AtExo70E2 altered their distribution patterns to localization in punctate structures distributed throughout the cell (Fig. 7f,n,r,v,z); AtExo70E2 was also localized to these structures (Fig. 7h,p,t,v,z; co-localizations are visualised in yellow/orange). AtExo70E2 also recruited Sec8 to the punctate structures (Fig. 7n,p) from its diffuse cytosolic distribution when this protein was expressed on its own (Fig. 4d). AtSec6 was the only exocyst complex protein that remained unchanged in the presence of AtExo70E2, retaining the diffuse cytosolic distribution pattern (Fig. 7j) that was also seen when it was expressed alone (Fig. 4b). As expected, AtExo70E2 had no effect on the distribution of GFP alone (Fig. 7b). The GFP-tagged versions of AtSec5a, AtSec6, AtSec8, AtSec15a, AtSec15b and AtExo84b were also co-expressed with the other myc-tagged Exo70 proteins tested (AtExo70A1, AtExo70B2, AtExo70C2, AtExo70D2, AtExo70G1, AtExo70H1 and AtExo70H7). Interestingly, AtExo70G1, which had a similar localization pattern to AtExo70E2 (Fig. 5j,l) and co-localized with AtExo70E2 (Fig. 6y–bb), showed a similar pattern of recruitment of AtSec5a, AtSec8, AtSec15a, AtSec15b and AtExo84b to AtExo70G1-localized punctate structures. The remaining Exo70 constructs did not alter the cellular distribution patterns of the other exocyst complex proteins.

image

Figure 7.  Co-expression of Arabidopsis thaliana Secs and Exo84 with AtExo70E2. BY-2 cells were co-transformed transiently with myc:Exo70E2 and GFP:Sec/Exo84 fusion proteins, fixed in formaldehyde at 6 h post biolistic bombardment, and then processed for epifluorescence microscopy. Bars, 10 μm. (a–d) Differential interference contrast (DIC), GFP, myc:Exo70E2 and merged image of (b) and (c), respectively; (e–h) DIC, GFP:Sec5a, myc:Exo70E2 and merged image of (f) and (g), respectively; (i–l) DIC, GFP:Sec6, myc:Exo70E2 and merged image of (j) and (k), respectively; (m–p) DIC, GFP:Sec8, myc:Exo70E2 and merged image of (n) and (o), respectively; (q–t) DIC, GFP:Sec15a, myc:Exo70E2 and merged image of (r) and (s), respectively; (u–x) DIC, GFP:Sec15b, myc:Exo70E2 and merged image of (v) and (w), respectively; (y–bb) DIC, GFP:Exo84, myc:Exo70E2 and merged image of (z) and (aa), respectively. Exo, exocyst; Sec, secretory.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

In this study, we have conducted detailed phylogenetic analyses of the predicted exocyst complex subunits (Exo70, Sec3, Sec5, Sec6, Sec8, Sec10, Sec15, and Exo84) in four complete plant genomes: those of A. thaliana (Arabidopsis Genome Initiative, 2000), rice (O. sativa; Yuan et al., 2003), poplar (P. trichocarpa; Tuskan et al., 2006), and moss (P. patens; Rensing et al., 2008). As previously reported by Synek et al. (2006), the plant Exo70 genes have evolved into large expanded families, in contrast to their single-copy yeast, Drosophila, and mammalian counterparts. The remaining predicted plant exocyst subunit genes were also duplicated, with the exception of Sec6. Polyploidization is an ancient and ongoing process in plant genome evolution, and it has long been recognized that gene and genome duplications have been important factors in the evolution of eukaryotes as they have relaxed the selection pressure on duplicated genes and allowed diversification of function (Ohno, 1970; Wendel, 2000). Currently, there is only one report of duplicated exocyst components in a metazoan (Brymora et al., 2001), suggesting that duplicated exocyst complex subunit genes acquired during genome duplication events have been selectively lost in animals. Similarly, only some of the duplicated genes for the plant Sec3, Sec5, Sec6, Sec8, Sec10, Sec15 and Exo84 genes are retained in the genomes of modern-day plants. By contrast, plant Exo70 genes which were acquired through multiple rounds of genome/chromosomal duplications and/or retrotransposition events appear to have been retained. What is interesting from the phylogenetic analysis is that the presence of three major clades in the moss, rice, A. thaliana, and poplar genome suggests that plants have acquired a number of Exo70 genes via ancient duplication events in a common ancestor of the plants. In addition, subsequent duplication events have occurred before and after the monocot–dicot divergence, leading to further expansion of the plant Exo70 gene families.

The fact that the Exo70 genes have been greatly expanded in plant genomes suggests that the exocyst complex has acquired new regulatory roles in plants, and that Exo70 plays a key role in defining these functions as well as in recruiting other exocyst complex subunits to these processes. This is supported by our data showing that AtExo70E2 and AtExo70G1 were capable of recruiting other exocyst components to the TGN in tobacco BY-2 cells. Perhaps under other growth conditions or in different cell types, the other AtExo70 proteins would play similar roles. Further evidence that the AtExo70 proteins are key regulators comes from the analyses of the various A. thaliana microarray databases, where the expression profiles of the AtExo70 genes appear to be very dynamic under different treatment conditions. By contrast, the remaining exocyst complex genes showed very little variation in the same treatment regimes. It is reasonable to suggest that the plant exocyst may be controlled, in part, through differential expression of the AtExo70 genes.

Within the A. thaliana Exo70 gene family, subfunctionalization of the duplicated genes appears to have been, to some extent, achieved through the divergent expression patterns seen in the microarray expression databases. A second process, whereby duplicated genes each adopt a part of the parental gene’s function, can also result in duplicated genes being subfunctionalized and preserved (Wendel, 2000). This could explain why several of the AtExo70 genes are expressed in the same tissues; for example, AtExo70A2, AtExo70C1, AtExo70C2, AtExo70H3 and AtExo70H5 are all expressed in male reproductive organs. Another potential fate of retained duplicated genes is neofunctionalization, where the gene product adopts a novel function (Zhang, 2003). A clear example of this can be seen when comparing tethering complexes functioning at the Golgi. Work conducted by Whyte & Munro (2001) indicated that components from the exocyst, Vacuolar Protein Sorting (Vps)52/53/54, and Sec34/35 complexes appear to be related to each other, suggesting a common ancestry, yet these complexes are involved in different membrane trafficking steps and interact with different partners (Whyte & Munro, 2001). Similarly, the expanded plant Exo70 family may have evolved to interact with different partners to regulate distinct membrane trafficking steps in plants. This may also explain why we observed different subcellular distribution patterns for the AtExo70 proteins in the BY-2 cells.

In yeast and animal systems, the exocyst acts as a tethering complex to deliver exocytic vesicles to specific sites on the plasma membrane for polarized growth (e.g. yeast budding and neurite outgrowth) or membrane localization (e.g. polarized epithelial cells), and regulated secretion following a stimulus (e.g. insulin signalling and the delivery of GLUT4 transporters). Similar types of roles for the putative exocyst are also starting to emerge in plants. Analyses of plant exocyst subunit mutants predict functions in the polarized delivery of exocytic vesicles, such as the maize rth1 (sec3) mutant with defects in root hair elongation (Wen & Schnable, 1994; Wen et al., 2005); the A. thaliana sec5a/5b, sec6, sec8, and sec15a mutants with reduced pollen tube growth and partial transmission defects (Cole et al., 2005; Hála et al., 2008); and the A. thaliana exo70A1 mutant with reduced root hair growth and shortened hypocotyls (Synek et al., 2006; Hála et al., 2008). In addition, the Sec6, Sec8 and Exo70A1 proteins have been shown to co-localize to the apex of growing tobacco pollen tubes (Hála et al., 2008). With our microarray expression analyses revealing general trends of increased AtExo70 gene expression levels following pathogen infections, hormone treatments, and abiotic stress treatments, there may also be other potential roles for the exocyst in these processes. One could envision the exocyst regulating some type of secretory event in response to the treatment or stress. Notably, we have also uncovered a role for Exo70A1 in pollen–pistil interactions as it is required in the stigmatic papillae of Brassica and A. thalianato promote compatible pollen hydration and pollen tube penetration of the stigmatic surface (Samuel et al., 2009).

To better understand the plant exocyst structure and the precise mechanism underlying its assembly in vivo, we investigated the subcellular localization patterns of several of the exocyst subunits in tobacco BY-2 cells. Although BY-2 cells are undifferentiated and probably do not undergo polarized secretion (Nagata et al., 1992), their large size and transformation efficiency make them amenable to studies aimed at elucidating the intracellular localization of the exocyst components. Indeed, we observed a number of interesting patterns of localization of the exocyst proteins in the BY-2 cells. We observed the localization of AtSec5a, AtSec15a, AtSec15b and AtExo84b to large globular-like structures in the perinuclear region of the cell which probably included TGN/early endosomes membranes as well as late endosome membranes (based on the overlapping localization of AtSec15a with Syp21, 42, and 52). Interestingly, these perinuclear structures bore a resemblance to the perinuclear localized ‘BFA compartment’ pattern which consists of plant endosomal and TGN membranes (Tse et al., 2006; Müller et al., 2007;Robinson et al., 2008). Perinuclear localization has also been observed for the rat Exo70 protein in undifferentiated rat adrenal PC12 cells (Vega & Hsu, 2001), and for the human Sec10 and Sec15 proteins in Madin-Darby Canine Kidney (MDCK) cells (Lipschutz et al., 2003;Wang & Hsu, 2003; Zhang et al., 2004). For human Sec10, this distribution pattern was attributed to its interaction with a component of the ER translocon and proposed to be monitoring/regulating the cell’s secretory capacity (Lipschutz et al., 2003; Wang & Hsu, 2006). Yeast Sec8p and Sec15p were also found to interact with components of the ER translocon (Toikkanen et al., 2003). Interestingly, in rat adrenal PC12 cells, Exo70 was found to relocalize to the extending neurite during NGF-induced neuronal differentiation (Vega & Hsu, 2001). Finally, in polarized animal cells, exocyst subunits have been detected in perinuclear-localized apical recycling endosomes and are involved in the delivery of transport vesicles to specific plasma membrane regions (Blankenship et al., 2007; Oztan et al., 2007; Hubert et al., 2009). Given the undifferentiated state of the tobacco BY-2 cells, the perinuclear localization may be a default pattern for some exocyst complex components while waiting for a ‘polarize secretion’ cue, although one cannot rule out the possibility that this compartment was formed as a result of increasing the levels of (ectopic) expression of select exocyst subunits. Another default pattern in undifferentiated plant cells may be the diffuse cytosolic distribution observed for AtSec6, AtSec8 and AtExo70A1 in BY2 cells. In addition to this pattern, Hála et al. (2008) have demonstrated the enrichment of AtSec6, AtSec8 and AtExo70A1 at the tip of growing pollen tubes. It is very likely that BY-2 cells lack the necessary cue for polarized secretion that would be present in the growing pollen tube.

We also observed that AtExo70E2 and AtExo70G1 were localized to punctuate structures throughout the cell which corresponded with markers to the TGN/early endosome as well as the late endsome/prevacuolar compartment, and were able to recruit several other exocyst complex subunits (i.e. AtSec5a, AtSec8, AtSec15a, AtSec15b and AtExo84b) to the same locations. The ability of the Exo70E2 and Exo70G1 to recruit Sec5a, Sec8, Sec15a, Sec15b and Sec84b to these endomembrane-containing punctuate structures suggests that these subunits may be interacting and forming high-order complexes at these locations. This is consistent with biochemical data suggesting that A. thaliana Sec5, Sec6, Sec8, Sec15a and Exo70A1 can form a complex in vivo in plants (Hála et al., 2008). Yeaman et al. (2001) conducted a similar localization study in nonmotile fibroblastic NRK-49F cells, which do not undergo polarized secretion and hence are in a similar physiological state to tobacco BY-2 cells. In these NRK-49F cells, 60–70% of exocyst complexes were located at the TGN, whereas the remaining 20–30% of the exocyst complexes were located on the plasma membrane in cell–cell contact zones. Based on these data, the authors of this study speculated that the TGN membranes marked by the exocyst complexes represent sorting domains for a specific class of secretory vesicles (Yeaman et al., 2001). In addition to the TGN, the partial localization of AtExo70E2 and AtExo70G1 to the late endosome/prevacuolar compartments in BY-2 cells suggests that, in addition to the proposed role of the exocyst in polarized cells, some Exo70 proteins may be required for the general transport of vesicles through the secretory pathway in cells not actively undergoing polarized secretion. Evidence in support of this theory comes from mammalian systems, where the exocyst has been reported to facilitate docking of endosomal-derived vesicles at the plasma membrane (Prigent et al., 2003; Zhang et al., 2004). Thus, the A. thaliana exocyst complex localization patterns that we observed in BY-2 cells were similar to those found in yeast and animal cells, and this further supports the theory that the exocyst performs similar conserved functions in plants. Zárskýet al. (2009) have also proposed that the plant exocyst complex function in vesicular trafficking from the recycling endosomal compartment, which is thought to reside at the TGN/early endosome, and the patterns that we have observed are consistent with this function.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank May Grace Aldea-Brydges for technical assistance. We are very grateful to Dr Ralph Quatrano for providing access to the Physcomitrella genome sequence before publication and Masa H. Sato for providing plasmids expressing GFP:Syp21, GFP:Syp42 and GFP:Syp52. We are also indebted to Dr Steven Rothstein for his financial support of SKG through an Ontario Centre for Agriculture Genomics Grant. This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC) to DRG and RTM, and a Canada Research Chair to DRG.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information