The Exocyst at a Glance

The exocyst complex comprises eight subunits: Sec3, Sec5, Sec6, Sec8, Sec10, Sec15, Exo70 and Exo84. It was first identified in the budding yeast Saccharomyces cerevisiae by genetic and biochemical approaches (Novick et al., 1980; TerBush and Novick, 1995; TerBush et al., 1996). The mammalian exocyst complex was first purified from rat brain, and was found in all the tissues examined (Hsu et al., 1996, 1998). In cells, the exocyst is recruited to sites of active exocytosis and membrane expansion, where it mediates the tethering of secretory vesicles to the plasma membrane in preparation for soluble N-ethylmaleimide-sensitive factor (NSF) attachment protein receptor (SNARE)-mediated membrane fusion.

Most of the exocyst subunits interact with multiple other subunits within the complex (Guo et al., 1999a; Matern et al., 2001; Dong et al., 2005). Deep-etch electron microscopy (EM) has shown the glutaraldehyde-fixed exocyst complex to be a uniform T- or Y-shaped structure, whereas the unfixed complex displays variable shapes, which suggests conformational changes (Hsu et al., 1998). In Arabidopsis thaliana, electron tomography studies of vesicle fusion during cell plate formation showed ‘Y’- or ‘L’-shaped structures linking vesicles (Segui-Simarro et al., 2004). These structures are likely to represent the exocyst complex undergoing vesicle tethering, although further verification of their identity is needed.

The crystal structure of parts of several exocyst subunits have been resolved (Baek et al., 2010; Dong et al., 2005; Hamburger et al., 2006; Liu and Guo, 2012; Moore et al., 2007; Sivaram et al., 2006; Wu et al., 2005; Yamashita et al., 2010; Munson and Novick, 2006; Mott et al., 2003; Fukai et al., 2003; Jin et al., 2005). Many exocyst subunits contain rod-like structures that are composed of α-helical bundles, which are also found in other complex associated with tethering containing helical rods (CATCHR) family proteins (Yu and Hughson, 2010). These rod-shaped subunits are likely to pack together side by side into the ‘Y’-shaped complex, as observed by EM (Hsu et al., 1998). Structural studies have also revealed that the N-terminus of yeast Sec3 has a PH-domain-like structure that is evolutionarily conserved from yeast to mammals and plants (Baek et al., 2010; Yamashita et al., 2010), consistent with the proposed interactions of this domain with phosphatidylinositol 4,5-bisphosphate PI(4,5)P₂ and the Rho family of small GTPases (Guo et al., 2001; Zhang et al., 2001, 2008). Moreover, regions of Sec5 and Exo84 have been crystalized in complex with the GTPase Ral (Mott et al., 2003; Fukai et al., 2003; Jin et al., 2005).

In addition to the holo-complex, free pools of certain subunits and sub-complexes have also been identified by biochemical fractionation experiments (Guo et al., 1999a; Jin et al., 2005; Moskalenko et al., 2003; Morgera et al., 2012; Zhao et al., 2013). The physiological significance of the different sub-complexes still needs to be investigated. In vivo studies have shown unique expression patterns for different subunits in fly and zebrafish (Mehta et al., 2005; Murthy et al., 2005; Thisse et al., 2004). Therefore, the exocyst complex is likely to be assembled and activated with spatiotemporal specificity. Understanding how the assembly of the complex is regulated is crucial to the understanding of the function of the exocyst in diverse physiological processes.

It has been speculated that the exocyst tethers secretory vesicles to the plasma membrane (Pfeffer, 1999; Guo et al., 2000; Whyte and Munro, 2002). However, direct evidence for the role the exocyst in tethering is missing. Live-cell imaging in HeLa cells showed that Sec8 is transported to the plasma membrane on vesicles, and remains there for seconds until fusion (Rivera-Molina and Toomre, 2013). This observation may reflect the transport and tethering of vesicles mediated by the exocyst before SNARE-mediated fusion. In yeast, it has recently been shown that ectopic targeting of Sec3 to mitochondria or peroxisomes resulted in the recruitment of secretory vesicles to these surrogate organelles, supporting the role of the exocyst in vesicle targeting and tethering (Luo et al., 2014). Future imaging studies and in vitro reconstitution experiments are needed to ultimately establish the role of the exocyst in tethering.

At the molecular level, how is the exocyst associated with the secretory vesicles and the plasma membrane? It has been shown that two subunits, Sec3 and Exo70, directly bind to PI(4,5)P₂, which is located at the inner leaflet of the plasma membrane (Liu et al., 2007; He et al., 2007b; Zhang et al., 2008). Sec3 interacts with PI(4,5)P₂ through an evolutionarily conserved PH-domain-like region at its N-terminus (Baek et al., 2010; Yamashita et al., 2010), whereas Exo70 interacts with PI(4,5)P₂ through a patch of basic residues at its C-terminus (He et al., 2007,b; Liu et al., 2007; Dong et al., 2005; Hamburger et al., 2006; Moore et al., 2007). In yeast, simultaneous disruption of Sec3 and Exo70 binding to PI(4,5)P₂ dissociates the exocyst from the plasma membrane, and results in a block in secretion (He et al., 2007,b; Zhang et al., 2008). On the secretory vesicle, Sec15 directly interacts with the Rab protein Sec4 in its GTP-bound form (Guo et al., 1999a). Sec6 binds to the v-SNARE protein Snc2 (Shen et al., 2013). Moreover, Exo84 binds to Ral (see below) and might interact with phospholipids (Moskalenko et al., 2003; Jin et al., 2005). Overall, these combinatorial interactions might mediate exocyst–vesicle associations. Results from these molecular interaction studies are consistent with the localization analyses of individual exocyst subunits in yeast. Immuno-EM studies, coupled with fluorescence imaging experiments, have demonstrated that Sec3 and Exo70 associate with the plasma membrane, whereas the other subunits – including some Exo70 proteins – associate with secretory vesicles (Boyd et al., 2004). Furthermore, Sec3 and some of the Exo70 proteins are localized to the bud tip even in the absence of actin cables, which mediate the polarized delivery of secretory vesicles (Finger et al., 1998; Boyd et al., 2004; Zhang et al., 2008). In fission yeast and mammals, however, Sec3 and Exo70 were found to associate with vesicles as well as the plasma membrane (Yeaman et al., 2004; Liu et al., 2007; Andersen and Yeaman, 2010; Bendezu et al., 2012; Zhao et al., 2013). It is possible that, in these cells, all of the exocyst subunits can be transported on vesicles; once they arrive at the destination, Sec3 and Exo70 might then ‘anchor’ the exocyst and vesicles to the plasma membrane through their physical association with PI(4,5)P₂.

Membrane fusion immediately follows vesicle tethering. Genetic and biochemical studies in yeast indicate that the exocyst functions upstream of SNAREs (Brennwald et al., 1994; Wiederkehr et al., 2004, Grote et al., 2000). The yeast exocyst subunit Sec6 was reported to interact – as a dimer – with the t-SNARE protein Sec9 (the SNAP-25 homolog) (Sivaram et al., 2005). However, it is perplexing that Sec6 has an inhibitory effect on SNARE assembly in an in vitro experiment (Sivaram et al., 2005). Sec6 also binds to the v-SNARE protein Snc2 through part of its SNARE assembly motif; a mutation of SNC2 that disrupts this interaction led to mis-localization of the exocyst and a block in exocytosis (Shen et al., 2013). Finally, Sec6 also interacts with the Sec1/Munc18 family protein Sec1 (Morgera et al., 2012). In addition to Sec6, Exo84 interacts with Sro7 and Sro77, the homologs of mammalian Lethal giant larvae (Lgl), which bind to the SNARE proteins and regulate exocytosis (Lehman et al., 1999; Zhang et al., 2005). Despite the interactions of exocyst subunits with multiple SNARE components or SNARE regulators, no data have so far provided clear support for the role of the exocyst in promoting SNARE assembly and membrane fusion. Future reconstitution experiments are called for, in order to establish the role of the exocyst in SNARE-mediated fusion.

As vesicle tethering precedes fusion, spatial and temporal control of exocytosis in cells can be executed through the regulation of the exocyst. Indeed, subunits of the exocyst have been found to be direct targets of a number of small GTPases and kinases.

The first reported interaction between the exocyst and small GTPases was that of yeast Sec15 and the exocytic Rab protein Sec4 (Guo et al., 1999a). This interaction is specific for Sec4 because other Rab proteins, such as those functioning at the endoplasmic reticulum (Ypt1) or endosomes (Ypt5p), do not interact with Sec15. Activated Sec4 might mediate the recruitment of the exocyst to secretory vesicles as well as the assembly of the complex (Guo et al., 1999a; Luo et al., 2014). In Drosophila melanogaster, Sec15 was shown to interact with a subset of Rab proteins (Rab3, Rab8, Rab11 and Rab27) that are involved in exocytic trafficking (Wu et al., 2005). In mammalian cells, Sec15 was shown to interact with Rab11, which is involved in the generation of vesicles from the trans-Golgi network (TGN) or recycling endosomes for subsequent delivery to the plasma membrane (Zhang et al., 2004). The interaction of Sec15 with Rab proteins has been implicated in E-cadherin trafficking, Notch signaling and primary ciliogenesis (Classen et al., 2005; Jafar-Nejad et al., 2005; Langevin et al., 2005; Feng et al., 2012). In yeast, Sec15 was also reported to interact with the type-V myosin Myo2p, which binds to both Sec4 and Ypt32 (the yeast homolog of Rab11), suggesting that Rab proteins orchestrate the exocyst function for vesicle transport along the cytoskeleton (Lipatova et al., 2008; Jin et al., 2011). In addition to the interaction between Rab and Sec15, another exocyst subunit, Sec10, was shown to bind to GTP–Arf6, and this interaction is important for the recycling of proteins to the plasma membrane (Prigent et al., 2003). The study by Prigent et al. further highlights the potential role of the exocyst in the trafficking from recycling endosomes to the plasma membrane.

The Rho family of small GTPases control many cellular processes, including morphogenesis and proliferation. In budding yeast, the Rho proteins are master regulators of cell polarity through their interactions with actin regulators (for a review, see Park and Bi, 2007). Rho proteins have also been implicated in exocytosis, as certain mutant alleles of Cdc42 and Rho3 display defects in secretion (Adamo et al., 1999, 2001). Cdc42, in its GTP-bound form, directly interacts with Sec3 (Zhang et al., 2001). The dual interaction of the Sec3 N-terminal PH-domain-like region with Cdc42 and PI(4,5)P₂ controls the polarized localization of Sec3, and might also kinetically regulate exocytosis (Zhang et al., 2001, 2008; Adamo et al., 2001; Baek et al., 2010). In mammalian cells, Cdc42 functions together with the exocyst to regulate phagocytosis (Mohammadi and Isberg, 2013). In addition to Cdc42, another Rho family protein, Rho1, interacts with Sec3 through the same N-terminal region (Guo et al., 2001; Yamashita et al., 2010). Cdc42 and Rho1 compete with each other for their binding to Sec3, suggesting that they control Sec3 function at different stages of cell growth (Zhang et al., 2001). In addition to Sec3, Exo70 interacts with Cdc42 and Rho3 in yeast (Robinson et al., 1999; Wu et al., 2010). The interaction with Rho3 was mapped to the C-terminus of Exo70 (Dong et al., 2005), however, exactly which part of Exo70 binds to Cdc42 in yeast is unclear. It has been reported that prenylation of Cdc42 is needed for this interaction (Wu et al., 2010). In mammalian cells, Exo70 interacts, through its N-terminus, with Rho-related GTP-binding protein RhoQ (also known as TC10), a homolog of Cdc42, and this interaction is implicated in the incorporation of Glut4 to the plasma membrane in adipocytes in response to insulin stimulation (Inoue et al., 2003). Although the data for the above-mentioned interactions are quite clear, how these Rho GTPases coordinate with each other in their interaction with different exocyst subunits remains unclear.

In metazoans, Ral GTPases (RalA and RalB) interact with both Sec5 and Exo84, and these interactions have been implicated in many processes, including cell migration, autophagy, neurogenesis and cancer (Brymora et al., 2001; Sugihara et al., 2002; Moskalenko et al., 2002, 2003; Polzin et al., 2002; Moskalenko et al. ; Balakireva et al., 2006; Bodemann et al., 2011; Chen et al., 2006, 2007; Chien et al., 2006; Hazelett et al., 2011; Lalli, 2009; Rosse et al., 2006). Cell fractionation studies have shown that Sec5 and Exo84 are in separate sub-complexes (Moskalenko et al., 2003), and structure analysis has demonstrated that Sec5 and Exo84 competitively bind to GTP–Ral (Jin et al., 2005). Recent studies have shown different subcellular localizations of these two sub-complexes, and suggested their involvement in different cellular functions, such as autophagy and innate immunity (Bodemann et al., 2011). Furthermore, ubiquitylation of RalB promotes the switch from an interaction between RalB, Exo84 and beclin1 to that beteen RalB, Sec5 and TBK1, which further differentiates the roles of Ral and the exocyst in different processes (Simicek et al., 2013).

Components of the exocyst are targets for a number of kinases. In mammalian cells, Exo70 is phosphorylated by ERK1/2 in response to epidermal growth factor (EGF) signaling (Ren and Guo, 2012). Phosphorylation of Exo70 promotes its association with the rest of the exocyst complex and stimulates exocytosis. Because ERK1 and ERK2 are pivotal kinases in the classic Ras–MEK–ERK signaling cascade, ERK1/2 phosphorylation might coordinate exocytosis with other cellular processes – such as cell migration – in response to growth factor signaling. In budding yeast, Exo84 is phosphorylated by the cyclin-dependent kinase Cdk1 during mitosis (Luo et al., 2013). Phosphorylation of Exo84 disrupts its association with the rest of the exocyst subunits, resulting in a block of exocytosis before the transition from metaphase to anaphase. The study by Luo et al. might provide a molecular mechanism for the mitotic cell-growth arrest observed in most eukaryotic cells. In the human fungal pathogen Candida albicans, Exo84 is phosphorylated by Cdk1, which is bound to the hypha-specific cyclin Hgc1, at residues that are specific to this pathogen, which is important for filamentous growth (Caballero-Lima and Sudbery, 2014). Phosphorylation of Exo84 by Cdk1–Hgc1 alters its affinity for phosphatidylserine, although how this change affects filamentous growth remains unclear. Another subunit, Sec5, can be phosphorylated by PKC at its Ral-binding domain, which allosterically leads to dissociation of Sec5 from RalA in mammalian cells (Chen et al., 2011). This cycle of engagement and disengagement with RalA is likely to allow continuous rounds of exocytic trafficking to the plasma membrane. Overall, these studies demonstrate that subunits of the exocyst can be phosphorylated in different modes that serve a diverse array of cellular processes.

The basic function of the exocyst complex is exocytosis. However, recent studies implicate the exocyst in many other functions. A rationale is that many cellular processes involve, or need to coordinate with, exocytosis through the exocyst complex. In other cases, studies link the exocyst to cellular functions through binding partners that have no obvious connection to exocytosis (see supplementary material Table S1 for a list of proteins and molecules that interact with the exocyst complex). Next, we summarize several major cellular processes involving the exocyst.

In budding yeast, mutations in the exocyst subunits lead to accumulation of post-Golgi secretory vesicles in the cytoplasm (Novick, et al., 1980; Finger and Novick, 1997; Guo et al., 1999b; He et al., 2007a,b). The exocyst is localized to the emerging bud tip, where it mediates exocytosis for the asymmetric expansion of daughter cell surfaces during polarized cell growth. During cytokinesis, the exocyst is localized to the mother–daughter cell junction to mediate abscission (TerBush and Novick, 1995; Finger et al., 1998; Guo et al., 1999b; Zhang et al., 2008). The polarized localization of the exocyst stands in contrast to the localization of t-SNARE proteins, which are distributed along the entire plasma membrane (Brennwald et al., 1994). In Arabidopsis thaliana, the exocyst is localized to growing pollen tubes and root tips, which are regions of active membrane expansion (Hala et al., 2008; Fendrych et al., 2013). In epithelial cells, the exocyst was shown to mediate vesicle trafficking to the basolateral domain, and is important for the establishment of epithelial polarity (Grindstaff et al., 1998; Lipschutz et al., 2000; Langevin et al., 2005; Yeaman et al., 2004; Andersen and Yeaman, 2010; Xiong et al., 2012). The recruitment of the exocyst complex to the vesicle is facilitated by the clathrin adaptor complex AP-1B (Folsch et al., 2003). Genetic analyses in worms suggest that the exocyst functions together with Rab10 in basolateral recycling; ablation of the exocyst subunits disrupts endosomal tubules (Chen et al., 2014). Studies have also shown that the exocyst functions together with Rab8 and the Par3/6 complex in membrane trafficking to the apical domain (Blankenship et al., 2007; Oztan et al., 2007; Bryant et al., 2010). In neurons, the exocyst is localized to the growth cones, tips of neurites or points of branching, where exocytosis is needed for membrane addition and surface expansion (Hazuka et al., 1999; Vega and Hsu, 2001; Lalli and Hall, 2005; Das et al., 2014). Inhibition of the exocyst does not affect neural transmitter release at mature synapses (Andrews et al., 2002; EauClaire and Guo, 2003; Murthy et al., 2003). However, it has been shown that delivery of the NMDA and AMPA receptors to the post-synaptic membrane is regulated by the exocyst complex (Sans et al., 2003; Riefler et al., 2003; Gerges et al., 2006).

Directional cell migration requires the coordination of actin cytoskeleton remodeling, plasma membrane reorganization and polarized exocytosis. The exocyst, together with Rab and Ral, mediates the trafficking of signaling proteins and adhesion molecules to the leading edge (Rosse et al., 2006; Rosse et al., 2009; Spiczka and Yeaman, 2008; Assaker et al., 2010; Thapa et al., 2012). In addition, the secretion of matrix metalloproteinases (MMPs) is important for the formation of invadopodia during tumor cell invasion (Sakurai-Yageta et al., 2008; Liu et al., 2009; Yamamoto et al., 2013). ERK1/2 phosphorylation of Exo70 promotes MMP secretion by promoting the assembly of the exocyst complex (Ren and Guo, 2012). The exocyst was also reported to interact with the endosomal Wiskott–Aldrich syndrome protein and Scar homolog (WASH) complex, thus coupling generation of membrane Type-1 MMP-containing late endosomes to their exocytosis at the plasma membrane (Monteiro et al., 2013). In neurons, the exocyst interacts with the Par3/6 complex under the control of Ral to regulate polarized neuronal cell migration (Lalli, 2009; Das et al., 2014).

Actin remodeling is required for membrane protrusion at the leading edge. The Arp2/3 complex is the core machinery for the generation of a branched actin network that pushes the plasma membrane (Goley and Welch, 2006; Pollard and Borisy, 2003; Suraneni et al., 2012). Exo70 directly interacts with the ARPC1 subunit of the Arp2/3 complex, and promotes actin branching (Liu et al., 2012). Interaction between Exo70 and Arp2/3 is stimulated by EGF signaling, which is known to promote cell migration and tumor invasion (Liu et al., 2009; Zuo et al., 2006). In addition to modulating actin dynamics, Exo70 interacts with PI(4,5)P₂ and induces a high curvature on the plasma membrane (Zhao et al., 2013). The coordinated regulation of actin polymerization and membrane deformation through Exo70 leads to effective membrane protrusion during cell migration. The ability of Exo70 to induce high membrane curvature might also contribute to the formation of tunneling nanotubes, which mediate cell–cell communication (Hase et al., 2009; Schiller et al., 2013).

The exocyst functions in multiple stages of cytokinesis. As discussed above, CDK1 regulates the assembly of the exocyst complex through phosphorylation of Exo84, thus arresting exocytosis before the transition from metaphase to anaphase (Luo et al., 2013). At telophase, the exocyst directs the trafficking of vesicles to the cell–cell junction for abscission (Martin-Cuadrado et al., 2005; Wang et al., 2002 and 2003). In Arabidopsis, the exocyst subunits are localized to the separation sites during initiation of the cell plate and maturation, when vesicle fusion is in high demand (Fendrych et al., 2010). In mammalian cells, the exocyst is involved in the trafficking of secretory vesicles to the cleavage furrow and midbody under the control of Rab11 and RalA (Chen et al., 2006; Fielding et al., 2005; Neto et al., 2013). Because cleavage often takes place at one side of the midbody (also known as asymmetric abscission), the vesicles appear to be recruited to the cleavage site of one of the daughter cells (Gromley et al., 2005; Schiel et al., 2013).

Ral GTPases are crucial for tumorigenesis (Camonis and White, 2005; Hamad et al., 2002; Lim et al., 2005; Rangarajan et al., 2004). As a direct downstream effector of Ral, the exocyst has been shown to participate in tumorigenic signaling processes. Knockdown of Sec5 and Exo84, which interact with GTP–Ral, reduces oncogenic transformation and tumor cell growth (Issaq et al., 2010). The exocyst is also proposed to function as a scaffold for signaling molecules. RalB–Sec5 recruits and activates TBK1, which in turn activates Akt to overcome apoptosis in cancer cells (Chien et al., 2006; Chien and White, 2008; Ou et al., 2011). The interaction between RalB, Sec5 and TBK1 also initiates innate immune responses following viral infection (Chien et al., 2006). In addition, the RalB–Exo84 complex was shown to interact with beclin-1 to regulate autophagosome assembly in response to starvation (Bodemann et al., 2011). The switch of RalB binding from Exo84 to Sec5 is mediated by ubiquitylation of RalB (Simicek et al., 2013). It is interesting to note that the exocyst mostly plays a role as a signaling scaffold in the above mentioned process. It is not clear whether and how this scaffold role is linked to exocytosis.

The primary cilium is a microtubule-based membrane protrusion on the cell surface that functions as a signaling organelle. Cilia malfunction has been implicated in many diseases, such as Bardet-Biedl syndrome, Joubert syndrome and polycystic kidney diseases. Exocyst subunits have been detected at the base or inside of primary cilia by immunofluorescence microscopy or EM; inhibition of exocyst function affected ciliogenesis (Rogers et al., 2004; Zuo et al., 2009; Feng et al., 2012). In addition, an interaction network between Rab11, Rabin8 (the guanine nucleotide exchange factor for Rab8), Rab8 and the exocyst complex regulates primary ciliogenesis – probably by mediating the transport of transmembrane proteins or basal bodies to the cilia membrane, although further experimental evidence is needed to support this hypothesis (Knodler, et al., 2010; Das and Guo, 2011; Feng et al., 2012). Rab10 also interacts with the exocyst and mediate ciliogenesis in renal epithelial cells, probably through a similar mechanism (Babbey et al., 2010). The interaction of exocyst with the Par3/6 complex might also contribute to primary ciliogenesis, although direct evidence for this notion is lacking (Zuo et al., 2009; Lalli, 2009; Das et al., 2014).

As discussed above, the exocyst is involved in many biological processes from cell migration to primary ciliogenesis. Although its basic role in exocytosis is linked to many of these processes, its functional diversification likely involves many different regulatory mechanisms. In addition to small GTPases and kinases, alternative splicing of the exocyst at the mRNA level could serve as a means to diversify exocyst function. In humans, Exo70 has five splicing isoforms. During epithelialization, epithelial splicing regulatory proteins 1 and 2 (ESRP1 and ESRP2) mediate the generation of a particular isoform of Exo70 (also known as Exo70-E; Lu et al., 2013) that differs from other Exo70 isoforms by a 23-amino-acid insertion. This Exo70 isoform cannot interact with the Arp2/3 complex and fails to stimulate cell motility (Warzecha et al., 2010; Lu et al., 2013). Isoform switching of Exo70 takes place during epithelial-to-mesenchymal transition, and is implicated in breast cancer metastasis (Lu et al., 2013). Exo70 was also found to be the most diversified subunit of the exocyst complex in plants – many of them have dozens of Exo70 paralogues per genome. This might correlate with the many unique structures of endomembranes and surface domains observed in terrestrial plants and their adaptation to different environments (Cvrckova et al., 2012; Zarsky et al., 2009). For example, the Exo70 isoform Exo70E2 marks a recently discovered organelle, named exocyst-positive organelle (EXPO), that is implicated in an unconventional secretion pathway in plants (Wang et al., 2010). In addition to Exo70, many of the other exocyst subunits also have multiple isoforms or paralogues. The existence of isoforms or paralogues of house-keeping proteins, such as Exo70, might account for the many different functions observed for this evolutionarily conserved complex.

Since its initial purification (TerBush and Novick, 1995), the last two decades have seen exciting progress in the understanding of the exocyst complex. Research in the field has not only led to a better understanding of the exocyst as a tether complex in exocytosis, but also revealed the role of the exocyst in a wide range of cell biological processes. The expansion of our knowledge about the exocyst also raises new questions. To elucidate the fundamental mechanism of exocyst function in exocytosis, we must understand how the exocyst complex interacts with SNARE proteins and potentially activates them for vesicle fusion with the plasma membrane. Despite the efforts from a number of laboratories, a clear model is still lacking. What is the structure of the complex and how is it assembled? With the recent revolution in cryo-EM technology, it is possible that a near-atomic-level-resolution structure of the holo-complex can be obtained. In fact, cryo-EM is probably the only approach for a high-resolution structure of this macromolecular complex. Does assembly of the exocyst complex (or sub-complexes) coordinate vesicle tethering with upstream vesicle budding from the TGN and recycling endosomes? What are the molecular consequences of the exocyst interactions with small GTPases? At the organismal level, how is the exocyst complex involved in tissue and organ development? With the advent of new technologies in cell biology, the next decade will see more exciting progress in the understanding of the exocyst, exocytosis and beyond.