Introduction

Protein secretion is a fundamental process for communication between eukaryotic cells and therefore, for organismal homeostasis, reproduction and survival. Two types of secretory processes can be distinguished based on the rate and mode of regulation of secretory vesicle release: Constitutive secretion and regulated secretion (Morgan, 1995). In constitutive secretion, secretory vesicles are exocytosed as they are produced. This process takes place in all eukaryotic cells, mainly to maintain homeostasis of the plasma membrane and the extracellular matrix. In contrast, regulated secretion takes place in specialized cell types, such as endocrine and exocrine cells, as well as in cells of the immune system such as platelets and neutrophils (Aggarwal, Jennings, Manning, & Cameron, 2023; Ley et al., 2018). These cell types produce specialized secretory vesicles known as secretory granules (SGs), which are released in response to specific stimuli (Kogel & Gerdes, 2010). SGs sprout from the trans-Golgi network (TGN) as immature vesicles, incompetent for fusion with the plasma membrane. Maturation of SGs involves homotypic fusion and acquisition of membrane proteins, which are required for SG delivery and fusion with the plasma membrane. Finally, stimulus-driven fusion of SGs with the plasma membrane results in cargo release to the extracellular milieu (Omar-Hmeadi & Idevall-Hagren, 2021; Sugita, 2008).

The salivary gland of Drosophila melanogaster larvae provides several advantages for the identification of regulators of SG biogenesis, maturation and exocytosis (Burgess et al., 2012; de la Riva-Carrasco et al., 2021; Neuman & Bashirullah, 2018; Rousso, Schejter, & Shilo, 2016). At late third larval instar, salivary glands synthesize a series of mucins, collectively known as Glue proteins, which become packed in SGs. Approximately four hours before the onset of pupariation, concerted exocytosis of mucin-containing SGs, followed by extrusion of the mucins out of the prepupal body are essential for gluing the puparium to the substratum (Borne, Kovalev, Gorb, & Courtier-Orgogozo, 2020). Interestingly, these three events of SG development (biogenesis, maturation and exocytosis) occur sequentially and only once in each cell of the salivary gland, that will later on be degraded during metamorphosis (Duan et al., 2020; Tracy, Velentzas, & Baehrecke, 2016). The tagging of one of these mucins with fluorophores (Sgs3-GFP or Sgs3-dsRed), (Biyasheva, Do, Lu, Vaskova, & Andres, 2001) combined with the large size of salivary gland cells and their SGs, have allowed high resolution imaging and real time traceability of SG biogenesis, maturation and secretion, leading to the identification and characterization of dozens of factors required along the exocytic pathway (Burgess et al., 2011; Neuman et al., 2022; Reynolds, Zhang, Tran, & Ten Hagen, 2019; Torres, Rosa-Ferreira, & Munro, 2014; Tran, Masedunskas, Weigert, & Ten Hagen, 2015).

The exocyst is a hetero-octameric protein complex identified and initially characterized in the budding yeast, and later found to be conserved across all eukaryotic organisms (Hsu et al., 1996; Novick, Field, & Schekman, 1980; Novick et al., 1995; TerBush, Maurice, Roth, & Novick, 1996). Yeast cells bearing mutations in exocyst subunits display intracellular accumulation of secretory vesicles and defects in exocytosis (Govindan, Bowser, & Novick, 1995; TerBush et al., 1996). Molecularly, the exocyst complex participates in vesicle tethering to the plasma membrane prior to SNARE-mediated fusion (An et al., 2021). Exocyst complex malfunction has been therefore associated with tumor growth and invasion, as well as with development of ciliopathies, among other pathological conditions (Luo, Zhang, Luca, & Guo, 2013; Mavor et al., 2016; Thapa et al., 2012; Whyte & Munro, 2002; B. Wu & Guo, 2015).

In this work, we have utilized the Drosophila salivary gland to carry out a methodic analysis of the requirement of each of the eight subunits of the exocyst along the regulated exocytic pathway. By inducing temperature-dependent gradual downregulation of the expression of each of the subunits, we discovered novel functions of the exocyst complex in regulated exocytosis, namely in SG biogenesis, SG maturation and homotypic fusion, as well as in SG fusion with the plasma membrane. We propose that the exocyst complex participates in all these processes as a general tethering factor.

Results

Characterization of secretory granule progression during salivary gland development

Salivary glands of Drosophila melanogaster larvae fulfill different functions during larval development. Mostly, they act as exocrine glands producing non-digestive enzymes during the larval feeding period, as well as mucins when the larva is about to pupariate (Costantino et al., 2008; Farkas et al., 2014). More recently, salivary glands have been proposed to behave as endocrine organs as well, secreting a yet unidentified factor that regulates larval growth (Li et al., 2022). Biosynthesis of mucins produced by salivary glands, named Salivary gland secreted proteins (Sgs), begins at the second half of the third larval instar in response to an ecdysone peak (Biyasheva et al., 2001). After being glycosylated at the Endoplasmic Reticulum (ER) and the Golgi Complex (GC), mucins are packed in SGs. Following subsequent developmentally-controlled hormonal stimuli, SGs are massively exocytosed, releasing the mucins to the gland lumen, and finally expelling them outside of the puparium, and gluing it to the substratum (Biyasheva et al., 2001). We used Sgs3-GFP or Sgs3-dsRed transgenic lines to follow this process in vivo. In wandering larvae, Sgs3-dsRed can be detected in salivary glands (Figure 1A’), while in prepupae, Sgs3-dsRed has been secreted out from the puparium, and is no longer detectable in salivary glands (Figure 1B’). To have a better temporal and spatial resolution of this process, we dissected and analyzed by confocal microscopy salivary glands expressing Sgs3-GFP at 4 h intervals starting at 96 h after egg laying (AEL). Expression of Sgs3-GFP begins at 96-100 h AEL, and can be detected at the distal region of the gland (Figure 1C). Later on, Sgs3-GFP expression expands to more proximal cells, and at 116 h AEL, the mucin becomes detectable in the whole gland, with the exception of ductal cells which do not express Sgs3 (Figure 1C) (Biyasheva et al., 2001). Thereafter, at 116-120 h AEL, in response to an ecdysone peak, SGs fuse with the apical plasma membrane, and release their content to the gland lumen (Figure 1C).

Figure 1.

Detailed observation of distal cells of salivary glands revealed that SGs enlarge from 96 h AEL onwards (Figure 1D), (Ma & Brill, 2021b; Neuman, Lee, Selegue, Cavanagh, & Bashirullah, 2021), a phenomenon that we quantified by measuring SG diameter (Figure 1D and Supplementary Table 1). At 96-100 h AEL nascent SGs are smaller than 1 μm in diameter. Later, and up to 112 h AEL, most SGs are smaller than 3 μm in diameter, and classified them as “Immature SGs”. From 112 h AEL onwards most SGs are bigger than 3 μm in diameter and classified as “Mature SGs” (Figure 1D and Supplementary Table 1). The maximal SG diameter that we detected was 7.13 μm at 116-120 AEL, just before exocytosis (Supplementary Table 1).

The eight subunits of the exocyst complex are required for regulated exocytosis of secretory granules

Fluorophore-tagged Sgs3 can also be used to screen for exocytosis mutants in which Sgs3-GFP is retained inside salivary glands at the prepupal stage (de la Riva-Carrasco et al., 2021; Ma et al., 2020). Using this approach, we found that the exocyst complex is apparently required for exocytosis of SGs, since silencing of any of the eight subunits of the complex results in retention of Sgs3-GFP in salivary glands of prepupae (Figure 2 and Supplementary Figure 1B). We used the salivary gland specific driver forkhead-Gal4 (fkh-Gal4) to induce the expression of RNAis against each of the eight subunits of the exocyst complex in larvae that also expressed Sgs3-GFP. We analyzed Sgs3-GFP distribution in wandering larvae and prepupae (Figure 2A-B), and observed that whereas control prepupae were able to expel Sgs3 outside the puparium (Figure 2A’ and C), this process was blocked in most individuals expressing RNAis against any of the exocyst subunits, with Sgs3-GFP being detected inside salivary glands of prepupae (Figure 2B’-C and Supplementary Figure 1B). These observations suggest that all exocyst subunits are required for exocytosis of SGs.

Figure 2.

Having established that knock-down of each of the exocyst subunits can block secretion of Sgs3, leading to retention of the mucin in the salivary glands (Figure 2), we investigated if exocytosis is actually impaired in these larvae. We dissected salivary glands of control and exocyst knock-down individuals at 116 h AEL, and analyzed them under the confocal microscope. Whereas control salivary gland cells were filled with mature SGs (Figure 3A), a more heterogeneous situation was found in salivary glands cells expressing RNAi against exocyst subunits, with some cells displaying mature SGs, and other, immature SGs, while even some displayed Sgs3-GFP in a mesh or network-like structure (Figure 3B). Therefore, we investigated if this mosaic phenotypic manifestation was due to variations in cell-to-cell Gal4-UAS activation and therefore RNAi expression, being an indication of potentially different functions of the exocyst complex in the secretory pathway of salivary gland cells. We performed RNAi-mediated knock-down of each of the eight exocyst subunits at different temperatures (29, 25, 21 and 19°C), that were accurately controlled in water baths, to obtain different levels of silencing and likely different phenotypic manifestations of exocyst complex downregulation. Whereas expression of a control RNAi resulted in mature SGs irrespectively of the temperature, expression of exocyst subunits RNAis at high temperatures (29°C) resulted, as a general rule, in phenotypes consistent with early arrest of the secretory pathway, since in most cells Sgs3-GFP was retained in a reticular structure or packed in immature SGs (Figure 3C, Supplementary Figure 1A and Supplementary Table 3). By lowering the temperature of RNAi expression, and therefore moderating silencing, the most severe phenotypes (early arrest of the secretory pathway) gradually became less prominent and simultaneously, the proportion of cells displaying immature SGs, and even mature SGs, became more noticeable (Figure 3C and Supplementary Figure 1A and Supplementary Table 3), suggesting that lower expression of RNAis allowed the progression of the secretory pathway. Notably, for each of the RNAis tested there was a temperature at which each of the three phenotypic outcomes could be clearly identified, although this particular temperature could be different for each RNAi, likely due to expression levels of the transgenes (Figure 3C, Supplementary Figure 1A and Supplementary Table 3). RT-qPCR confirmed that different culturing temperatures resulted in different degrees of silencing of exocyst subunit mRNA as shown in Supplementary Figure 2 for exo70 and sec3 (Supplementary Figure 2A-B). Also, we found that different phenotypic outcomes resulting from expression of different RNAi transgenic lines that target the same subunit (exo70) are due to differences in the levels of mRNA downregulation generated by each particular (exo70 ^RNAiV and exo70^RNAiBL), and that there is a correlation between the level of mRNA downregulation and strength of the phenotype observed (Supplementary Figure 2C).

Figure 3.

To define more precisely the role of the exocyst, and rule out potential pleiotropic effects due to developmental defects derived from chronic exocyst downregulation, we made use of the Gal80 thermosensitive tool (Gal80^ts) ((Lee & Luo, 1999)). Larva were grown at a restrictive temperature (18°C) until they reached the 3^er instar (120 hours). In this manner, the exocyst complex could be functional up until that developmental stage. Then, larvae were transfered to the permissive temperature (29°C) and salivary glands were dissected 36 hours later. We found that temporally-restricted expression of sec3^RNAi or sec15^RNAi phenocopied unrestricted expression of the same RNAis at 29°C (Supplementary Figure 3), indicating that the phenotypes obtained are not due to pleiotropic effects caused by developmentally unrestricted downregulation of the exocyst. Finally, the MLS phenotype generated by expression of sec15^RNAi could be rescued by simultaneous expression of GFP-Sec15, supporting the notion that defective biogenesis of SGs was specifically provoked by sec15 loss of function (Supplementary Figure 4).

A comprehensive analysis of cell polarity, as well as a number of general markers of cellular homeostasis (Supplementary Figures 5 and 6), ruled out that defects in SG biogenesis or maturation observed after knock-down of exocyst subunits could stem from potential secondary effects derived from poor cellular health, but rather reflect genuine functions of the exocyst complex in the secretory pathway. Along these lines, a recent report showed that apical polarity defects generated by loss of polarity protein Crumbs, do not affect or interfere with SG exocytosis (Lattner, Leng, Knust, Brankatschk, & Flores-Benitez, 2019), further supporting the notion that there are parallel pathways controlling cell polarity and SG biogenesis, maturation and exocytosis.

The fact that the three phenotypic outcomes (1-Sgs3-GFP retained in a mesh; 2-immature SG and 3-mature-not exocytosed SG) could be retrieved with appropriate silencing any of the eight subunits lead us to propose that the holocomplex, and not subcomplexes or individual subunits, could function several times along the secretory pathway, and that each of these activities could require different amounts of exocyst complex. Importantly, irrespectively of the temperature of expression of RNAis, retention of SGs in salivary glands cells was always significantly higher in exocyst knock-down individuals than in controls (Figure 3D and Supplementary Figure 1B), indicating that ultimately, the exocyst is required for SG exocytosis.

The exocyst complex is required for secretory granule biogenesis

We decided to characterize each of the three phenotypic manifestations of exocyst loss-of-function in more detail. The early-most manifestation of the requirement of the exocyst in the secretory pathway was the reticular or mesh-like phenotype obtained by strong silencing (29°C) any of the subunits (Figure 4A). This phenotype was reminiscent of mutants in which Sgs3 is retained at endoplasmic reticulum exit sites (ERES) or at ER-Golgi Complex (GC) contact sites (Burgess et al., 2011; Reynolds et al., 2019), suggesting that knock-down of the exocyst may provoke Sgs3 retention at the ER or GC, blocking SG biogenesis. Indeed, we found that RNAi-mediated silencing of sec15,sec3 or sec10 provoked Sgs3 retention in the ER since it colocalized with ER markers, as indicated by Pearsońs coefficient, whereas in control larvae Sgs3-GFP was inside SGs, at a comparable developmental stage (Figure 4B-G and Supplementary Figures 7 and 8).

Figure 4.

We analyzed in detail GFP-Sec15 subcellular localization, and found that associated closely with the trans-Golgi marker RFP-Golgi, but not to the ER marker KDEL-RFP, in salivary glands just prior to SG biogenesis (Figure 5 and Supplementary Movie 1). This association was lost after the onset SG biogenesis, suggesting that the exocyst associates with the GC at the specific developmental stage when Sgs3 transits through that organelle (Figure 5B-E). Three-dimensional reconstruction of Sec15-Golgi foci confirmed the spatial association with the exocyst (Figure 5E and Supplementary Movie 2-3). In line with this, we found that exocyst silencing under conditions that block Sgs3-GFP in the ER (29°C), both, cis and trans-CG structures were severely affected (Figure 6 and Supplementary Figure 9). This suggests that at this stage of salivary gland development the exocyst complex localizes to the GC, where it is required to maintain cis- and trans-GC morphology and allow the correct transport of Sgs3 from the ER through the GC (Figure 6E-F).

Figure 5.

Figure 6.

Role of the exocyst in secretory granule maturation: Homotypic fusion

Given that appropriate silencing conditions of any of the exocyst subunits can result in accumulation of immature SGs (Figure 7A), we set out to investigate a potential role of the exocyst in SG maturation. SG maturation is a multifaceted process that involves, among other events, homotypic fusion between immature SGs (Du et al., 2016; Neuman & Bashirullah, 2018). We found that GFP-Sec15 often localized in discrete foci in between immature SGs (75% of total GFP-Sec15), but this localization drops dramatically when SGs have undergone maturation (15%) (Figure 7B, C and Supplementary Movie 4-5). This transient localization of the GFP-Sec15 supports the notion of a role of the exocyst in SG homofusion, which was confirmed by live imaging of salivary glands ex-vivo (Figure 7D and Supplementary Movie 6). In support of this notion, when GFP-Sec15 was overexpressed at 25 °C, unusually large SGs of up to 20 μm in diameter could be detected, a size never observed in control salivary glands overexpressing GFP alone (Figure 7E-H), indicating that overexpression of the Sec15 subunit alone is sufficient to induce homotypic fusion between SGs, which is in agreement with reports that indicate that Sec15 functions as a seed for exocyst complex assembly (Escrevente, Bento-Lopes, Ramalho, & Barral, 2021; Guo, Roth, Walch-Solimena, & Novick, 1999). Sec15 overexpression is expected to induce the formation of of exocyst holocomplexes, provoking excessive homofusion among SGs. In contrast, overexpression of Sec8, which is not expected to induce the formation of the whole complex, did not have an effect on homotypic fusion of SGs (Figure 7F, H).

Figure 7.

The characteristic localization of GFP-Sec15 foci in between adjacent immature SGs, the fact that Sec15 overexpression results in oversized SGs, and the observation that downregulation of any subunit of the exocyst complex at appropriate levels results in accumulation of immature SGs, weigh in favor of the notion that the exocyst plays a critical role in SG homotypic fusion.

Role of the exocyst in secretory granule maturation: acquisition of membrane proteins

Besides homotypic fusion, maturation of SGs involves the incorporation of specific proteins that are required for homotypic fusion, apical navigation or fusion with the apical plasma membrane. The mechanisms by which these maturation factors associate with SG are not well understood. One of such proteins is the calcium sensor Synaptotagmin-1 (Syt-1), which localized at the basolateral membrane of salivary gland cells before SG biogenesis (96 h AEL) (Supplementary Figure 10A) and later, became detectable on the membrane of nascent SGs (diameter<1μm), prior to homotypic fusion (Figure 8A and Supplementary Figure 10B). As SGs mature, the presence of Syt-1 on SGs became more prominent, with a sharp increase after SG homofusion (Supplementary Figure 10C-D), suggesting that recruitment of Syt-1 to SGs continued after homotypic fusion had occurred (Figure 8M). Downregulation of the exocyst subunits Sec5 or Sec3 to levels that generate immature SGs, significantly reduced the presence of Syt-1 on their membranes, as compared to immature SGs of control salivary glands (Figure 8A-C and Supplementary Figure 11A-C). Interestingly, weaker knock-down, which allowed the formation of mature SGs, improved but not completely restored Syt-1 recruitment, compared to control salivary cells (Supplementary Figure 11D-F), indicating that Syt-1 is recruited to mature SGs in an exocyst-dependent manner (Figure 8M).

Figure 8.

CD63, the Drosophila homolog of mammalian Tsp29Fa, is another protein required for SG maturation. CD63 localizes at the apical plasma membrane before SG formation and reaches the membrane of SGs through endosomal retrograde trafficking (Supplementary Figure 10E), (Ma et al., 2020). We investigated if the exocyst participates in recruitment of CD63 to SGs, and found that CD63 could be readily detected at the membrane of 1μm; 3μm or 5μm SGs (Figure 8D and Supplementary Figure 10F-H), while upon Sec5 or Sec3 downregulation CD63 was significantly reduced on immature SGs (Figure 8D-F and Supplementary Figure 11G-I). Interestingly, milder reduction of Sec3 expression, under conditions that allow the formation of mature SGs, did not affect CD63 recruitment to SGs (Supplementary Figure 11J-L), indicating that CD63 is recruited to SGs before homotypic fusion in an exocyst-dependent manner and that, unlike to Syt-1, recruitment ceases after homotypic fusion has occurred (Figure 8M).

The small GTPases Rab11 and Rab1 were reported to be required for SG maturation. Whereas Rab11 associates to SGs independently of their size, Rab1 was found transiently on immature SGs only (Ma & Brill, 2021a; Neuman et al., 2021), and moreover, Rab1 recruitment to SGs depends on Rab11 (Neuman et al., 2021). Given that Rab11 is a stable component of SGs and Rab1 is not, we reasoned that other players, perhaps the exocyst, might be involved in Rab1 association or dissociation from the SG membrane. In fact, silencing of Sec5 under conditions that allowed the formation of immature SGs, provoked significant reduction of Rab1-YFP on the SG membrane (Figure 8G-I), suggesting that the exocyst participated in Rab1 recruitment (Figure 8M).

Remarkably, immature SGs that resulted from Sec5 or Sec3 knock-down displayed higher-than-normal levels of YFP-Rab11 around them (Figure 8J-L and Supplementary Figure 11M-O), suggesting that the exocyst is a negative regulator of Rab11 recruitment. This observation is in apparent contradiction with previous reports by others and us that indicate that Rab11 is required for SG maturation (de la Riva-Carrasco et al., 2021; Ma & Brill, 2021a; Neuman et al., 2021). Noteworthy, overexpression of a constitutively active form of Rab11 (UAS-YFP-Rab11^CA), which was readily recruited to SGs, also provoked an arrest of SG maturation, thereby phenocopying exocyst complex knock-down (Supplementary Figure 12A-C). Thus, during maturation of SGs, the levels of active Rab11 need to be precisely regulated and this is achieved, at least in part, by the exocyst (Figure 8M and Supplementary Figure 12F).

Given that the exocyst is an effector of different Rab-GTPases during vesicle exocytosis (Guo et al., 1999; Novick et al., 1995; S. Wu, Mehta, Pichaud, Bellen, & Quiocho, 2005), we next investigated if Rab11 is required for recruiting the exocyst to SGs. Indeed, we found that Sec15 failed to localize on SGs following knock-down of Rab11 (Supplementary Figure 12D-E), indicating a crucial role of Rab11 in recruitment of Sec15, and probably of the whole exocyst to SGs.

Overall, the results described in this section suggest that Rab11 recruits Sec15, and perhaps the whole exocyst, to immature SGs to allow homotypic fusion and maturation (Supplementary Figure 12D-E) while in turn, the exocyts limits the levels and/or activity of Rab11 on SGs (Figure 8J-L and Supplementary Figure 11M-R), as too much Rab11 is apparently detrimental for SG maturation (Supplementary Figure 12A-C). We propose that a single-negative feedback loop precisely regulates Rab11 and exocyst complex activity/levels, thus controlling recruitment of maturation factors such as Syt-1, CD63 and Rab1, and therefore, the outcome of SG maturation (Supplementary Figure 12F).

The exocyst is required for secretory granule fusion with the apical plasma membrane

Under low silencing conditions of exocyst subunits the prevalent phenotype was mature SGs retained in salivary gland cells (Figure 3D, Figure 9A, and Supplementary Figure 1B), suggesting a function of the complex in SG fusion with the apical plasma membrane. The process of SG-plasma membrane fusion can be assessed by visualizing the incorporation of plasma membrane-specific components to the membrane of SGs (de la Riva-Carrasco et al., 2021; Rousso et al., 2016; Tran et al., 2015). Phosphatidylinositol 4,5-bisphosphate (PI(4,5)P₂) is a lipid of the inner leaflet of the apical plasma membrane, which is absent in endomembranes (Phan et al., 2019). Using a fluorophore-based reporter of PI(4,5)P₂ the SGs that have fused with the plasma membrane can be distinguished from those that have not. Using this approach we found that silencing of Exo70 impaired fusion with the plasma membrane (Figure 9B-D). Consistent with a role of the exocyst in fusion of SGs with the plasma membrane, at this stage of development GFP-Sec15 was no longer detected GFP-Sec15 was detected on SGs at the sites of contact with the plasma membrane (Figure 9E-G and Supplementary Movie 7). These data indicate that during regulated exocytosis, the exocyst complex is required for fusion of SGs to the apical plasma membrane, possibly acting as a tethering complex (Figure 9H-I).

Figure 9.

Overall, by utilizing the Drosophila larval salivary gland, we have made a comprehensive analysis of the role of the exocyst complex in the pathway of regulated exocytosis. We found that the exocyst is critically required for biogenesis of SGs, for their maturation and homotypic fusion and for mediating fusion between SGs and the plasma membrane.

Discussion

The exocyst was initially identified as crucial for secretion in yeast (Novick et al., 1980; TerBush et al., 1996) and later on, for basolateral trafficking in animal epithelial cells (Grindstaff et al., 1998; Lipschutz et al., 2000); (Langevin et al., 2005), with implications in various cellular processes, including cell migration, cytokinesis, ciliogenesis and autophagy (Bodemann et al., 2011; Park et al., 2010; Rogers et al., 2004; Thapa et al., 2012). Recently, an in-depth study of the mammalian exocyst was performed, showing that each of the eight subunits of the complex are essential for constitutive secretion of soluble proteins (Pereira et al., 2023). However, fewer is known about the function of the exocyst in regulated exocytosis. A recent work indicates that, in cultured pancreatic beta cells, the exocyst mediates tethering of insulin-containing granules to the plasma membrane and to the cortical F-actin network, being required for exocytosis of a subset of these granules (Zhao et al., 2023). Previous studies had implicated specific subunits of the complex in insulin-stimulated exocytosis of the glucose transporter Glut4 in cultured adipocytes (Inoue, Chang, Hwang, Chiang, & Saltiel, 2003), as well as in skeletal muscle cells (Fujimoto et al., 2019). In Drosophila, some subunits of the complex (Sec3, Sec5, Sec6, Sec8 and Sec15) were recently suggested to be required for SG development (Ma et al., 2020), although earlier, a role for the Sec10 in salivary gland exocytosis had been disregarded (Andrews, Zhang, Trotta, & Broadie, 2002). Therefore, a role of the exocyst in regulated exocytosis is less clear. In this study, we not only demonstrate that the eight subunits of the complex are required for regulated exocytosis of SGs of Drosophila larval salivary gland, but also that the holocomplex participates in multiple steps along the secretory pathway.

Most studies of the secretory pathway rely on a bimodal readout: Intracellular retention of SGs versus exocytosis of SG content. These approaches set a limit to our understanding of the specific functions that regulators and effectors exert on the exocytic pathway. By using fluorescently labeled versions of a SG cargo protein in Drosophila salivary glands, Sgs3, we have shown that knock-down of any of the exocyst subunits can bring about three distinct phenotypic outcomes: 1-Impairment of SG biogenesis, 2-impairment of SG homotypic fusion and maturation, or 3-impairment of SG fusion with the apical plasma membrane. Noteworthy, the frequency at which these defects occurs depends on the extent of exocyst downregulation, implying that each of the three functions depends on different levels of the complex: SG-plasma membrane fusion is highly sensitive to even slight reductions of exocyst subunits levels; SG maturation requires intermediate levels of the complex, and SG biogenesis seems to be the most robust of the three processes, and strong reduction of exocyst levels is required to bring about this defect.

Although exocyst subcomplexes or individual subunits have been suggested to play specific roles in cell biology (Mehta et al., 2005), it is generally believed that the functions of the exocyst are carried out by the holocomplex (Ahmed et al., 2018). Studies analyzing the requirement of all eight subunits in a single biological process are not abundant. In Drosophila, a study was performed in which the eight subunits were found to be required for developmentally regulated autophagy, during salivary cell death, but not for starvation-induced autophagy in the fat body (Tracy et al., 2016) and also for proper synaptic development at the neuromuscular junction (Kang et al., 2024). However, in other biological settings some subunits of the exocyst (Sec3, Sec5, Sec6, Sec8, and Sec10) but not others (Sec15, Exo70 and Exo84) were shown to be specifically required for general as well as specific autophagy in yeast (Singh et al., 2019). In mammalian cells, Exo84- and Sec5-induced exocyst complex activation have opposing roles on autophagy machinery activation (Bodemann et al., 2011). Our systematic analysis of the requirement of each of the exocyst subunits in biogenesis, maturation and exocytosis of SGs led us to conclude that all three biological processes are carried out by the holocomplex, and not by individual subunits or subcomplexes.

Consistent with the three biological functions that the exocyst exerts in the secretory pathway, we found that its subcellular localization during salivary gland development is dynamic. Immediately after the onset of Sgs3 synthesis, the exocyst localized at the Golgi complex; later, when SG biogenesis has begun the exocyst was mostly present at the fusion point between immature SGs, and finally it localized on mature SGs in close proximity to the apical plasma membrane. These different subcellular localizations are coherent with the three phenotypes that we have observed after precise modulation of exocyst levels. Exocyst localization at the trans-Golgi network and at the plasma membrane has been reported before in mammalian cells (Yeaman, Grindstaff, Wright, & Nelson, 2001), and although it was suggested in that study that the exocyst might be required at several steps of the secretory pathway from the TGN to the plasma membrane, the idea was not further explored.

By analyzing SG-plasma membrane fusion markers, we have shown that, paralleling the requirement of the exocyst in constitutive exocytosis, the complex participates in SG fusion with the plasma membrane in regulated exocytosis as well. Moreover, we have found that the Sec15 subunit localizes exactly at contact sites between SGs and the plasma membrane, further supporting the notion that the exocyst plays a role in this fusion process. It is accepted that the function exerted by the exocyst in the fusion between SGs and the plasma membrane depends on Ral GTPases (Brymora, Valova, Larsen, Roufogalis, & Robinson, 2001; Wang, Li, & Sugita, 2004). RalA is the only Ral GTPase described so far in Drosophila, and we have recently reported that RalA is involved in SG-plasma membrane fusion in the salivary gland (de la Riva-Carrasco et al., 2021). Therefore, it seems likely that, paralleling constitutive exocytosis, the interaction between RalA and the exocyst is critical for tethering SGs to the plasma membrane during regulated exocytosis as well. The molecular components regulating tethering and fusion of SGs to the plasma membrane remain largely unknown. Perhaps, this is because the proteins involved in these processes participate in earlier steps of the secretory pathway as well. Our temperature-dependent manipulation of the expression of exocyst subunits, in combination with differential requirements of the complex in terms of quantity/activity in each of the processes was instrumental for uncovering three different discrete steps in which the exocyst plays a role along the regulated secretory pathway.

Proteins that will be secreted are co-translationally translocated to the ER, and then transported into the cis-region of the GC in COPII-coated vesicles that are 60–90 nm in diameter and therefore, sufficient to accommodate most membrane and secreted molecules (Raote & Malhotra, 2021). Larger cargos, such as collagen and mucins might use alternative ER-cis-GC communication mechanism that are independent of vesicular carriers. Specifically, direct connections between the ER and cis-GC are formed in a Tango1 and COPII dependent manner (Reynolds et al., 2019)(Yang, Feng, & Pastor-Pareja, 2024). Disruption of these connections not only affects SG formation, but also has profound impact on GC structure (Bard & Malhotra, 2006; Rios-Barrera, Sigurbjornsdottir, Baer, & Leptin, 2017). We found that strong silencing of any of the exocyst subunits results in retention of Sgs3 in the ER, abrogation of SG biogenesis, and alteration of normal morphology of the GC. In Drosophila, as well as in mammalian cells, Golgi complex is polarized in cis and trans-cisternae that are held together by tethering complexed of the CATCHR family. Disruption of these connections result in altered Golgi complex morphology, reflected in fragmentation and swelling of cisternae and impairment of the secretory pathway (D’Souza, Taher, & Lupashin, 2020; Khakurel & Lupashin, 2023; Liu et al., 2019). The fact that exocyst silencing not only affected SG formation, but also had profound impact on GC structure, together with the observation that Sec15 specifically associated with the Golgi complex but not with the ER before the onset of SG biogenesis, argues in favor of a role of this complex as a tethering complex between cisternae. We propose that the exocyst might be redundant with other CATCHR complexes in this function since only severe downregulation of exocyst subunits expression can manifest this phenotype.

Maturation of SGs is a multidimensional process that involves homotypic fusion, acidification, cargo condensation, and acquisition of membrane proteins that will steer SGs to the apical plasma membrane and contribute to the recognition, tethering and fusion (Boda et al., 2023; Ji et al., 2018; Nagy et al., 2022; Syed, Zhang, Tran, Bleck, & Ten Hagen, 2022). Previously, the exocyst was shown to be required for maturation of Weibel-Palade bodies (Sharda et al., 2020). In the current work, we have shown that the exocyst participates in several aspects of SG maturation. One of such processes is homotypic fusion of immature SGs. We found that an adequate extent of downregulation of the expression of any of the exocyst subunits leads to accumulation of immature SGs. Furthermore, consistent with a function of the exocyst in SG homofusion, we found that Sec15 localizes preferentially at the fusion point between adjacent immature SGs. In support of this notion, we found that overexpression of Sec15 results in unusually large granules, likely derived from uncontrolled homotypic fusion. Therefore, our data suggest that the exocyst complex is the tethering factor responsible to bring immature SGs in close proximity to enable fusion.

Besides SG growth by homotypic fusion, several factors are recruited to the maturing SG, including the transmembrane proteins CD63 and Syt-1. Our temperature-dependent genetic manipulations revealed that the addition of Syt-1 to SGs occurs immediately after they have emerged from the TGN, and addition of this protein continues after they have attained a mature size. Syt-1 localizes at the basolateral membrane of salivary gland cells before SG biogenesis, and it is not known how Syt-1 reaches the maturing SGs. Given that the exocyst is required for loading SGs with Syt-1, one possibility is that a vesicular carrier transports Syt-1 from the basolateral membrane to nascent and maturing SGs, and that the contact between SGs and these vesicles is mediated by the exocyst. CD63, which is also required for SG maturation, localizes instead at the apical plasma membrane before SG biogenesis, and reaches the SGs from the endosomal retrograde pathway through endosomal tubes (Ma et al., 2020). We found that cells in which the exocyst has been knocked-down display reduced levels of CD63 around SGs, suggesting that the exocyst might contribute to the contact between endosomal tubes and SGs as well. Based on our analysis, this maturation event should take place mostly before, and not after, homotypic fusion. Thus, our results support the notion that the exocyst complex might contribute to tethering maturing SGs to different types of membrane-bound carriers that transport maturation factors like Syt-1 and CD63. These observations expand the notion that a crosstalk between the secretory and endosomal pathways occurs (Ma & Brill, 2021a; Ma et al., 2020; Papandreou & Tavernarakis, 2020), and that the exocyst is a critical factor linking the two pathways.

Vesicle maturation, either in the endocytic or the exocytic pathways, usually involves changes in vesicle-bound Rab proteins (Ailion et al., 2014; Ma & Brill, 2021a; Thomas, Highland, & Fromme, 2021). The mechanisms that mediate these Rab-switches are not completely understood. Specifically, in Drosophila larval salivary glands, Rab1 and Rab11 association to SGs appear to be crucial for SG maturation and timely exocytosis (Neuman et al., 2021). We found that the levels of Rab1 and Rab11 on SGs are respectively positively and negatively regulated by the exocyst, and that additionally, Rab11 is itself required for recruitment of Sec15 to immature SGs. These complex relationships between Rab11, Rab1 and the exocyst can be explained through a single-negative feedback loop that would guarantee adequate levels of Rab11 and the exocyst on maturing SGs. This possibility is supported by previous evidence from Drosophila and other systems that indicate that Rab11 and Sec15 interact physically and genetically (Escrevente et al., 2021; Guo et al., 1999; Novick, 2016). In fact, the exocyst is an effector of Rab11 (Sec4p in yeasts), as GTP-bound Rab11 recruits Sec15 to secretory vesicles, which in turn triggers formation of the holocomplex on the vesicle membrane as previously shown in yeast, Drosophila and mammalian cells (S. Wu et al., 2005) (Takahashi et al., 2012) (Zhang, Ellis, Sriratana, Mitchell, & Rowe, 2004).

An increasing number of publications reveal the complexity and variety of vesicle trafficking routes that feed into the biogenesis, maturation and exocytosis of SGs. Our study has revealed that the exocyst holocomplex acts at several steps of the pathway of regulated exocytosis: namely, maintaining Golgi complex morphology to allow protein exit from the ER and transit through the early secretory pathway; promoting homotypic fusion of SGs and acquisition of maturation factors; and finally, allowing SG-plasma membrane fusion. We propose that all these events depend on the activity of the exocyst complex as a tethering factor.

Materials and methods

Fly stocks and genetics

All fly stocks were kept on standard corn meal/agar medium at 25°C. Crosses were set up in vials containing two males and five females of the required genotypes at 25°C. Crosses were flipped every 24 hours to avoid larval overcrowding, and then moved to the desired temperature: 29°, 25°, 21°, 19° or 18°C. The temperature used for each experiment is specified in figure legends. In the experiments of Figures 2, 3, 8 and Supplementary Figures 1 and 11, all crosses were maintained in a bath water to reduce temperature fluctuations. D. melanogaster lines used in this work are listed in Supplementary Table 2, and were obtained from the Bloomington Drosophila Stock Center (http://flystocks.bio.indiana.edu) or from the Vienna Drosophila Stock Center (https://stockcenter.vdrc.at); Sgs3-dsRed was generated by A.J. Andres’ Lab (University of Nevada, United States).

Sgs3-GFP retention phenotype

Larvae or prepupae of the desired genotype were visualized and photographed inside glass vials under a fluorescence dissection microscope Olympus MVX10. In prepupae, localization of Sgs3-GFP or Sgs3-dsRed inside salivary glands or outside the puparium was determined. Each experiment was repeated at least three times.

Developmental staging and SG size

When larvae were cultured at 25°C, SG maturation progressed according to the timeline shown in Figure 1. Roughly, at 29°C larval development was shortened by 24 hours, and extended in 1, 3 and 4 days when larvae were cultured at 21, 19°C and 18 °C respectively. Precise physiological staging of each salivary gland was carried out according to Neuman and co-workers (Neuman et al., 2021) upon dissection and observation under the confocal microscope. In all experiments, only SGs from the distal-most cells of the salivary glands were imaged to avoid potential variations in SG size due to desynchronization in the synthesis of Sgs3. Data of temperatures and developmental staging from each experiment are synthesized in Supplementary Table 4.

RNA extraction and cDNA synthesis

Total RNA was isolated from dissected salivary glands of third-instar larvae using 500 μl of Quick-Zol reagent (Kalium Technologies, RA00201) following the manufacturer’s instructions. The concentration and integrity of the RNA were determined using NanoDrop (Thermo Fisher Scientific) spectrophotometry. RNA (1 μg) was reverse-transcribed using M-MLV Reverse Transcriptase (Invitrogen, 10338842) using oligo-dT as a primer (https://doi.org/10.1080/15548627.2021.1991191). Control reactions omitting reverse transcriptase were used to assess the absence of contaminating genomic DNA in the RNA samples. An additional control without RNA was included.

Real-time PCR

Gene expression was analyzed by quantitative PCR in a CFX96 Touch (Bio-Rad) cycler. The reactions were performed using HOT FIREPol EvaGreen qPCR Mix Plus (without ROX; Solis BioDyne, 08-25-00001), 0.40 μM primers, and 12–25 ng of cDNA, in a final volume of 10.4 μl (https://doi.org/10.1080/15548627.2021.1991191). Cycle conditions were initial denaturation at 95°C for 15 min, and 40 cycles of denaturation at 95°C for 20 s, annealing at 60°C for 1 min, and extension and optical reading stage at 72°C for 30 s, followed by a dissociation curve consisting of ramping the temperature from 65 to 95°C while continuously collecting fluorescence data (https://doi.org/10.1080/15548627.2021.1991191). Product purity was confirmed by agarose gel electrophoresis. Relative gene expression levels were calculated according to the comparative cycle threshold (CT) method (https://doi.org/10.1080/15548627.2021.1991191). Normalized target gene expression relative to rpl29 was obtained by calculating the difference in CT values, the relative change in target transcripts being computed as 2−ΔCT. The efficiencies of each target and housekeeping gene amplification were measured and shown to be approximately equal. Oligonucleotides were obtained from Macrogen (Seoul, Korea), and their sequences were the following: exo70: Fw 5′-GAAGTGGTTCTCCGATCGCT-3′, Rv 5′-ACGAGCGGAGGTTGTCTTTT-3′; sec3: Fw 5′-GAAGACGCAACACATGGACG-3′, Rv 5′-CTTTGCATATTGGCCCCATCC-3′; sec5: Fw 5′-GTCAATGAGACTGCCAAGAACT-3′, Rv 5′-CCTGCAGTGGAATGTGCCTA-3′; rpl29: Fw 5′-GAACAAGAAGGCCCATCGTA-3′, Rv 5′-AGTAAACAGGCTTTGGCTTGC-3′. Rpl29 was used as housekeeping gene. Specificity and quality of oligonucleotide sequences for exo70, sec3, sec5, and rpl29 were checked using Primer Blast Resource of the NCBI (http://www.ncbi.nlm.nih.gov/tools/primer-blast/).

Quantification of the penetrance of phenotypes upon knockdown of exocyst subunits

Sgs3-GFP intracellular distribution was analyzed in salivary gland cells, and one of three phenotypic categories was defined for each cell: 1) “mesh-like structure” when Sgs3-GFP was distributed in a network-like compartment; 2) “Immature SGs” when Sgs3-GFP was in SGs with a median diameter smaller than 3 μm; and 3) “Mature SGs” when Sgs3-GFP was in SGs with a median diameter equal or larger than 3 μm. The penetrance of each of the three phenotypes was calculated for each genotype of interest at the four different temperatures analyzed.

Salivary gland imaging

Salivary glands of the desired stage were dissected in cold PBS (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na₂HPO₄, 1.47 mM KH₂PO₄, [pH 8]), and then imaged directly under the confocal microscope without fixation for no more than five minutes. In the experiment of Figure 9G and Supplementary Figure 6E, salivary glands were fixed for 2 hours in 4% PFA (Paraformaldehyde, Sigma) at room temperature, and then washed three times for 15 minutes with PBS-0.1% Triton X-100. For filamentous actin staining, salivary glands were incubated for 1 hour with Alexa Fluor 647 Phalloidin (ThermoFisher Scientific 1:400) in PBS-0.1% Triton X-100; stained tissues were mounted in gelvatol mounting medium (Sigma) and imaged at a confocal microscope Carl Zeiss LSM 710 with a Plan-Apochromat 63X/1.4NA oil objective, or Carl Zeiss LSM 880 with a Plan-Apochromat 20X/0.8 NA air objective or a Plan-Apochromat 63X/1.4NA oil objective.

For live imaging, salivary glands were dissected in PBS, and then mounted in a 15 mm diameter plastic chamber with a glass bottom made of a cover slip, containing 40 µl of HL3.1 medium (70 mM NaCl, 5 mM KCl, 1.5 mM CaCl2, 2 mM MgCl2, 5 mM HEPES, 115 mM sucrose, 5 mM trehalose, and pH 7.2 with NaHCO3). The medium was removed allowing the tissue to adhere to the bottom of the chamber, and then a Biopore membrane hydrophilic PTFE with a pore size of 0.4 μm (Millipore, Sigma) was placed over the sample, and HL3.1 medium was added on top covering the membrane. Images were captured under an inverted Carl Zeiss LSM 880 confocal microscope with a Plan-Apochromat 20X/0.8 NA air objective or a Plan-Apochromat 63X/1.4NA oil objective. For Supplementary Movie 1, frames were obtained every 2.25 seconds, while for Supplementary Movie 6 and 7, frames were captured every 0.78 and 0.67 seconds respectively.

Image processing and analysis

Image deconvolution was performed using the Parallel Spectral Deconvolution plugin of the ImageJ software (NIH, Bethesda, MD) with standard pre-sets (Schneider, Rasband, & Eliceiri, 2012). Image analyses were made with ImageJ (Schneider et al., 2012), and graphs were generated with the R Studio software (Team, 2020). For SG quantification, a region of interest (ROI) from each cell was used. In each ROI, the area of SGs was assessed, and SG diameter was calculated assuming that SGs are circular, using the formula ((Area/π)^1/2)*2 = Diameter. In experiments of Figure 4B, D, F, H and Supplementary Figure 7A, C, E, G, I two-dimensional lines scans were generated with the ImageJ plot profile. Fluorescence intensity was determined related to maximal intensity of each marker, always within a linear range. For quantification of fluorescence intensity of Syt1, CD63, Rab1 and Rab11 (Figure 8 and Supplementary Figure 11), the mean intensity of three different ROIs of 5 μm² from each cell was measured. 17-35 cells from 4-11 salivary glands were used in the analysis. Association analyses in Figure 5B, E were performed considering structures as associated when the fluorescence spikes, of each fluorophore, had a distance minor or equal to ∼0.6 μm. In the experiment of Figure 7C, GFP-Sec15 foci between granules were measured over total foci in four different ROIs of 225 μm² from each cell. For nucleus quantification (Supplementary Figure 5E) the area from each one, was measured using Image J freehand section. In polarity experiments, the mean fluorescence intensity of apical markers (Supplementary Figure 6A, B) was obtained by drawing a plotted line, of ∼2 μm wide, over the apical membrane. In the case of PI(4,5)P₂ and F-actine (Supplementary Figure 6 D, E), localized in both compartments particularly enriched in apical membrane, a perpendicular line, of ∼9 μm wide, was draw from the basolateral region to the apical and the coefficient between the maximum fluorescence intensity of both regions was calculated. Pearson’s coefficient was obtained with JACoP plugging of Image J (Bolte & Cordelieres, 2006). The Golgi defect, upon exocyst knock down (Supplementary Figure 9), was measured as the number of cells with aberrant cis-Golgi or with swollen trans-Golgi vesicles over total cells within each salivary gland.

Statistical analyses

Statistical significance was calculated using one-way analysis of variance (ANOVA), a Likelihood Ratio test or a Wald Test, and followed by a Tuckey’s test with a 95% confidence interval (p < 0.05) when comparing multiple factors.

Acknowledgements

We are grateful to Dr. Andrew Andres, Dr. Gabor Juhasz, the Bloomington Stock Centre and the Vienna Drosophila Resource Centre for fly strains. Dr. Andrés Rossi and Dr Esteban Miglietta, from FIL microscopy facility, for technical support with confocal microscopy; Andrés Liceri for fly food preparation; the FIL personnel for assistance and members of the Wappner lab for fruitful discussions.

Additional information

Competing interests

The authors declare no competing or financial interests.

Funding

This work was supported by grants from Agencia Nacional de Promoción de Científica y Tecnológica: PICT-2018-1501 and PICT-2021-I-A-00240 to PW and PICT-2021-GRF-TII00418 to MM. SSF and SPP were supported with fellowships of the Consejo Nacional de Investigaciones Científicas y Técnicas.

Author contributions

SSF, Conceptualization, Data curation, Formal analysis, Validation, Investigation, Methodology, Writing – review and editing; SPP, Investigation, Methodology; SF, Investigation, Writing – review and editing; PW, Conceptualization, Funding acquisition, Project administration, Writing – review and editing; MM, Conceptualization, Supervision, Funding acquisition, Methodology, Writing – original draft, Project administration, Writing – review and editing

Supplementary tables and figures

Supplementary Table 1.

Supplementary Table 2.

Supplementary Table 3.

Supplementary Table 4.

Supplementary Figure 1.

Supplementary Figure 2.

Supplementary Figure 3.

Supplementary Figure 4.

Supplementary Figure 5.

Supplementary Figure 6.

Supplementary Figure 7.

Supplementary Figure 8.

Supplementary Figure 9.

Supplementary Figure 10.

Supplementary Figure 11.

Supplementary Figure 12.