Rab6 is required for the exocytosis of cortical granules and the recruitment of separase to the granules during the oocyte-to-embryo transition in Caenorhabditis elegans

After fertilization, oocytes undergo a series of dramatic changes collectively called oocyte-to-embryo transition, which are essential for the onset of embryogenesis in many species (Ferrell, 1999; Wessel et al., 2001; Runft et al., 2002; Pellettieri et al., 2003; Horner and Wolfner, 2008). These changes include an increase in the intracellular calcium concentration (Ca²⁺), cytoskeletal rearrangement, the resumption of the cell cycle, and the exocytosis of secretory vesicles called ‘cortical granules’ (Ferrell, 1999; Wessel et al., 2001; Runft et al., 2002; Horner and Wolfner, 2008). Cortical granules are derived from the Golgi complex and contain extracellular matrix components such as proteases and glycoproteins (Wessel et al., 2001). Because of cortical granule exocytosis, the embryo surface is reorganized to protect embryos from polyspermy and osmotic and mechanical stress and to allow molecules involved in embryo functions such as cytokinesis to localize on the embryo surface (Wessel et al., 2001; Mizuguchi et al., 2003; Olson et al., 2006).

Cortical granule exocytosis occurs within seconds after fertilization in sea urchins, starfish, and frogs (Wessel et al., 2001). The increased Ca²⁺ concentration upon fertilization affects the sub-processes of cortical granule exocytosis, such as the translocation of cortical granules to near the plasma membrane in mice (Kline and Kline, 1992; Matson et al., 2006) and soluble N-ethylmaleimide sensitive factor attachment protein receptor (SNARE)-dependent membrane fusion in sea urchins (Conner et al., 1997; Conner and Wessel, 2001). It has been proposed that in the latter sub-process, the Ca²⁺-binding protein synaptotagmin promotes SNARE-dependent membrane fusion in response to Ca²⁺ (Leguia et al., 2006). However, cortical granule exocytosis does not always occur just after fertilization and the corresponding increase in Ca²⁺ levels, suggesting that Ca²⁺ is not the sole and direct trigger of the exocytosis. For example, in mice and Caenorhabditis elegans, cortical granule exocytosis occurs more than 5 min after fertilization (Fukuda and Chang, 1978; Sato et al., 2008), whereas the first Ca²⁺ increase occurs within 3 min after fertilization (Lawrence et al., 1997; Samuel et al., 2001).

C. elegans is a well-studied model for the oocyte-to-embryo transition and the subsequent cortical granule exocytosis (Marcello and Singson, 2010). In C. elegans, the timing of cortical granule exocytosis seems to be under the control of cell cycle regulators. In proximal oocytes, the cell cycle is arrested in meiotic prophase I, and cortical granules are located near the plasma membrane, but they are not secreted. Oocytes mature and proceed to metaphase I in response to a diffusive signal from the spermatheca (McCarter et al., 1999; Kosinski et al., 2005). The cyclin-dependent kinase CDK-1 is required for the cell cycle progression (Boxem et al., 1999). After fertilization, the anaphase-promoting complex/cyclosome (APC/C) is activated to induce meiotic anaphase I, which is when cortical granule exocytosis occurs (Sato et al., 2006; Bembenek et al., 2007) (supplementary material Fig. S1). CDK-1 and the subunits of APC/C as well as the downstream effector of APC/C, separase, are required for cortical granule exocytosis in C. elegans (Sato et al., 2006; Bembenek et al., 2007). Generally, separase is activated at the metaphase-to-anaphase transition by the degradation of its binding partner securin by APC/C (Nasmyth, 2002). Interestingly, the C. elegans ortholog of separase, SEP-1 (Siomos et al., 2001), translocates to cortical granules just before exocytosis (Bembenek et al., 2007). In addition, SEP-1 is required for the fusion of secretory vesicles into the plasma membrane during cytokinesis (Bembenek et al., 2010). These results collectively indicate that the timing of cortical granule exocytosis is regulated by cell cycle regulators (i.e. CDK-1, APC/C, and SEP-1). Further, the localization of SEP-1 on cortical granules is likely critical for exocytosis, because sep-1 alleles defective for cortical granule exocytosis are defective for SEP-1 localization on cortical granules, and a suppressor for the exocytosis defect also rescues the SEP-1 localization (Richie et al., 2011). However, the molecule that recruits separase to cortical granules remains unidentified.

Rab GTPases are important regulators of exocytosis in general, as they regulate budding, motility, tethering, and fusion of intracellular vesicles (Stenmark, 2009). During cortical granule exocytosis in C. elegans, RAB-11.1 (a Rab11 GTPase homolog) contributes to the targeting of cortical granules to the plasma membrane in oocytes before fertilization (Sato et al., 2008). However, the Rab GTPase involved in the fusion process during oocyte-to-embryo transition has not been identified.

Rab6 is known to play a key role in the membrane traffic around the Golgi complex. In mammalian cells, Rab6 cooperates with various effector proteins including motor proteins to regulate the fission of transport vesicles from the Golgi and the translocation of Golgi-derived vesicles along microtubules (Echard et al., 1998; Matanis et al., 2002; Grigoriev et al., 2007; Miserey-Lenkei et al., 2010; Grigoriev et al., 2011). In Drosophila oocytes, Rab6 helps establish cell polarity (Januschke et al., 2007). However, its role in cortical granule exocytosis has not been analyzed yet. In this study, we investigated the role of Rab6 in cortical granule exocytosis in C. elegans. We discovered a novel function of Rab6: the recruitment of separase on cortical granules for exocytosis during the oocyte-to-embryo transition.

The C. elegans genome has 2 Rab6 homologs, namely, rab-6.1 and rab-6.2. We first examined whether RAB-6.1 and RAB-6.2 localize to the Golgi complex. We found that RAB-6.1 (tagged with the red fluorescence protein mCherry; mCherry::RAB-6.1) and RAB-6.2 [tagged with green fluorescence protein (GFP); GFP::RAB-6.2] co-localized to vesicular structures throughout the oocyte cytoplasm (Fig. 1A). The vesicles were adjacent to the Golgi marker ManS [tagged with yellow fluorescence protein (YFP); ManS::YFP] (Fig. 1B), suggesting that RAB-6.1/RAB-6.2 vesicles are derived from the Golgi complex. We also confirmed that the mCherry::RAB-6.1 construct can rescue the phenotype resulting from a rab-6.1-deletion mutation (supplementary material Fig. S2).

To investigate the role played by Rab6 during C. elegans embryogenesis, we knocked down rab-6.1 and rab-6.2 via RNA interference (RNAi) treatment. Previous studies reported that the co-depletion of rab-6.1 and rab-6.2 results in embryonic lethality (Audhya et al., 2007; Zhang et al., 2008). We first confirmed that our RNAi treatment selectively depleted the RAB-6.1 or RAB-6.2 protein from the embryo (supplementary material Fig. S3). We found that the osmotic sensitivity of rab-6.1 (RNAi) embryos was high, but that of rab-6.2 (RNAi) embryos was not (Fig. 1C). rab-6.1 (RNAi) embryos, but not rab-6.2 (RNAi) embryos, were permeable to the lipophilic dye FM4-64 (Fig. 1C). We found a similar phenotype in the rab-6.1-deletion mutant rab-6.1 (tm2124), which lacks 64% of the rab-6.1 coding region, but not in the rab-6.2-deletion mutant, rab-6.2 (tm1924), which lacks 18% of the rab-6.2 coding region (Fig. 1C). High osmotic sensitivity and incorporation of the lipophilic dye is a characteristic of defective permeability barrier formation due to the failure of cortical granule exocytosis (Rappleye et al., 1999; Sato et al., 2008). The number of embryos produced by rab-6.1 (tm2124) worms was ∼29% of those produced by wild-type worms, and about 97% of the embryos were unhatched (supplementary material Fig. S2). Further, we confirmed that the cleavage furrow was hardly invaginated in the rab-6.1 (RNAi) embryos (Fig. 1D). This phenotype is also observed in permeability barrier-defective embryos (Olson et al., 2006). The results suggest that RAB-6.1 is involved in the remodeling of the embryo surface.

To further evaluate the involvement of RAB-6.1 in cortical granule exocytosis, we examined the effect of rab-6.1 RNAi on the dynamics of a cortical granule marker, a C. elegans caveolin-1 homolog CAV-1 (tagged with GFP; CAV-1::GFP) (Sato et al., 2006; Sato et al., 2008). In a previous study, all the 30 Rab GTPase genes were examined for their roles in cortical granule exocytosis, and the role of rab-11.1 was reported (Sato et al., 2008) but that of rab-6 was not. After the oocytes pass through the spermatheca, CAV-1 vesicles are tethered to the plasma membrane and incorporated into it (Fig. 2A; supplementary material Movie 1). We evaluated this incorporation process by using a transgenic strain expressing both CAV-1::GFP and the pleckstrin homology domain-containing membrane marker PH tagged with mCherry (mCherry::PH) (Kachur et al., 2008). Significant co-localization of CAV-1::GFP and mCherry::PH was observed when the cortical CAV-1 vesicles disappeared, indicating the incorporation of the CAV-1::GFP bodies into the plasma membrane (supplementary material Movie 1; Fig. 2A, middle column, 16 min). Later, during meiosis II, CAV-1 is re-internalized (supplementary material Movie 1; Fig. 2A, 24 min). In rab-6.1 (RNAi) embryos, however, the CAV-1::GFP vesicles were not incorporated into the plasma membrane (supplementary material Movie 2; Fig. 2B, 16 and 24 min). An abnormal behavior of CAV-1::GFP vesicles was observed in rab-6.1 (tm2124) embryos, but not in rab-6.2 (RNAi) or rab-6.2 (tm1924) embryos (supplementary material Fig. S4). These results indicate that RAB-6.1 is required for cortical granule exocytosis.

Our observation suggested that RAB-6.2 plays a minor role in cortical granule exocytosis. We detected delayed and partial exocytosis of cortical granules (supplementary material Fig. S4) in the rab-6.1-deletion mutant embryos [rab-6.1 (tm2124)], whereas the exocytosis was completely defective in rab-6.1 (RNAi) (Fig. 2). We hypothesized that rab-6.1 RNAi may slightly affect the expression of RAB-6.2, although we did not observe a detectable decrease in RAB-6.2 expression upon rab-6.1 RNAi (supplementary material Fig. S3). We also hypothesized that simultaneous knockdown of RAB-6.1 and RAB-6.2 leads to a complete defect in cortical granule exocytosis. Consistent with this scenario, we observed a complete defect in cortical granule exocytosis after RAB-6.2 knockdown by RNAi in the rab-6.1 (tm2124) strain (supplementary material Fig. S4). Of note is the finding that rab-6.1 (tm2124); rab-6.2 (tm1924) double deletion worms did not grow to adulthood, and we could not obtain embryos of this strain. These results indicate that RAB-6.2 can partially substitute the role of RAB-6.1 in cortical granule exocytosis. This role of RAB-6.2, however, is unlikely to be sufficient for functional exocytosis because rab-6.1 (tm2124) alone showed a significant permeability barrier defect (Fig. 1C).

To investigate the role played by RAB-6.1 in cortical granule exocytosis, we analyzed its subcellular localization during cortical granule exocytosis. We generated a transgenic worm expressing both mCherry::RAB-6.1 and CAV-1::GFP and found that more than 90% of the mCherry::RAB-6.1 protein localized to the CAV-1::GFP vesicles in the oocytes (supplementary material Movie 3; Fig. 3A). Following fertilization, the RAB-6.1-positive CAV-1 vesicles are tethered to the plasma membrane. When the vesicles fused with the plasma membrane during meiotic anaphase I, the mCherry::RAB-6.1 signals from the CAV-1 bodies disappeared and were found dispersed throughout the cytoplasm (supplementary material Movie 4; Fig. 3B). This behavior is similar to that of Rab GTPase in vesicle trafficking after membrane fusion in other species (see Discussion). These results suggest that RAB-6.1 plays a role in membrane fusion between the cortical granules and plasma membrane.

To examine the role of GTP in RAB-6.1 localization, we expressed GDP-locked (T25N) and GTP-locked (Q70L) forms of RAB-6.1. While the GTP-locked form localized to the cortical granules, the GDP-locked form failed to do so (Fig. 3C). This result suggests that RAB-6.1 is in its GTP-bound form during the process. The expression of the GTP-locked form did not show an apparent dominant effect, suggesting that GTP-GDP exchange regulation of RAB-6.1 is relevant for the cortical granule localization of RAB-6.1 but not specifically for exocytosis.

To obtain insights into the mechanisms of action of RAB-6.1, we compared the effect of rab-6.1 inhibition on the intracellular distribution of cortical granules with the effects produced by the inhibition of other genes reported to be essential for cortical granule exocytosis (Bembenek et al., 2007; Sato et al., 2008). We performed RNAi for the membrane traffic regulator rab-11.1 and the cell cycle regulators emb-27 cdk-1, and sep-1 in CAV-1::GFP-expressing worms (supplementary material Movies 5–10; Fig. 4). emb-27 and cdk-1 encode the C. elegans orthologs of CDC16 (a subunit of APC/C) (Golden et al., 2000) and Cdk1 (Boxem et al., 1999), respectively. RAB-11.1 is involved in recruiting cortical granules to the plasma membrane (Sato et al., 2008). We confirmed that GFP::RAB-11.1 co-localizes with mCherry::RAB-6.1 vesicles near the plasma membrane until exocytosis (supplementary material Fig. S5). Consistent with the previous report (Sato et al., 2008), CAV-1 vesicles were strongly accumulated around the nucleus in the rab-11.1 (RNAi) oocytes (supplementary material Movie 7; Fig. 4A). After fertilization, the CAV-1 vesicles were not tethered to the plasma membrane but were distributed throughout the cytoplasm in rab-11.1 (RNAi) embryos (Fig. 4B). In contrast, rab-6.1 (RNAi) embryos showed a different phenotype. After fertilization, the vesicles did not accumulated around the nucleus of the embryo; they are located near the plasma membrane but were not secreted (Fig. 4B). Of note, this phenotype was similar to that of the cdk-1, emb-27 and sep-1 (RNAi) embryos (Fig. 4B). On the basis of these results, we suspected that the role played by RAB-6.1 in cortical granule exocytosis is closely related to that of the cell cycle regulators downstream of the fertilization signal (CDK-1, EMB-27 and SEP-1).

Since the effects of RAB-6.1 inhibition on intracellular distribution of cortical granules were similar to those of CDK-1, EMB-27 and SEP-1 inhibition, we suspected a functional interaction between RAB-6.1 and cell cycle regulation. A previous study has revealed that SEP-1 localizes to the cortical granules during meiotic anaphase I and induced cortical granule exocytosis (Bembenek et al., 2007). SEP-1 has been also shown to be required for the fusion of secretory vesicles into the plasma membrane during cytokinesis (Bembenek et al., 2010). Thus, SEP-1 functions in membrane fusion, although the detailed mechanism is not known. We speculated that RAB-6.1 recruits SEP-1 to cortical granules to function as an effector protein during membrane fusion. We confirmed that GFP::SEP-1 translocates to RAB-6.1-positive cortical granules for 2 to 3 min before secretion (supplementary material Movie 11; Fig. 5A). The co-localization of SEP-1 and RAB-6.1 was observed even in rab-11.1 (RNAi) embryos where RAB-6.1-positive cortical granules were not tethered to the plasma membrane (Fig. 5B; supplementary material Fig. S6). We examined the effect of rab-6.1 RNAi on the recruitment of SEP-1 to the cortical granules and found that SEP-1 was not recruited in rab-6.1 (RNAi) embryos (Fig. 5B). Chromosome separation did occur in rab-6.1 (RNAi) embryos (Fig. 5B, arrows), indicating that the recruitment of SEP-1 to the cortical granules is independent of its chromosome segregation function. In contrast, polar body exclusion was defective in rab-6.1 (RNAi) embryos, which is consistent with a previous report that proposed that extracellular space formation via cortical granule exocytosis is required for polar body exclusion (Bembenek et al., 2007).

SEP-1 recruitment to the cortical granules was not observed in cdk-1 (RNAi) and emb-27 (RNAi) embryos, in which chromosome separation and subsequent polar body extrusion did not occur (Fig. 5B). This result suggests that entry into anaphase is required for SEP-1 recruitment to the cortical granules. To induce the chromosome segregation function of separase, APC/C activates the protease activity of separase by degrading securin (Nasmyth, 2002). To recruit SEP-1 to the cortical granules, APC/C is required but the activation of SEP-1's protease activity may not be required. The protease-dead form of SEP-1 (Bembenek et al., 2010) was still able to localize to the cortical granules (supplementary material Fig. S7). The role played by securin in SEP-1 localization to the cortical granules is not clear. Securin knockdown (ify-1) by RNAi did not induce ectopic recruitment of SEP-1 to the cortical granules but inhibited SEP-1 localization to the cortical granules and inhibited exocytosis (supplementary material Fig. S8). The stability of the SEP-1 protein may be reduced in ify-1 (RNAi) C. elegans embryos, as reported in the case of human cells (Jallepalli et al., 2001).

The GTP- and GDP-bound forms of Rab GTPases are generally the active and inactive states, respectively, and the former can interact with effector proteins (Stenmark, 2009). Consistent with this general trend of Rab GTPases, we observed the co-localization of the GTP-locked (Q70L), but not GDP-locked (T25N), form of RAB-6.1 with SEP-1 (supplementary material Fig. S9A).

Consistent with the suggested minor role of RAB-6.2 in cortical granule exocytosis (supplementary material Fig. S4), RAB-6.2 likely plays a minor role also in SEP-1 localization to the cortical granules. We detected some SEP-1 positive cortical granules in the rab-6.1-deletion mutant embryos [rab-6.1 (tm2124)], which were not detected after RAB-6.2 knockdown by RNAi in the rab-6.1 (tm2124) strain (supplementary material Fig. S9B).

We observed that in normal embryos, RAB-6.1 temporarily co-localized with SEP-1 in the cortical granules for a few minutes and then was dispersed in the cytoplasm concomitantly with membrane fusion (Fig. 5A). Interestingly, in sep-1 (RNAi) or cdk-1 (RNAi) embryos, mCherry::RAB-6.1 did not dissociate from the cortical granules (supplementary material Movie 12; Fig. 5C). The result indicates that not only cortical granule exocytosis but also the dissociation of RAB-6.1 from the cortical granules depends on SEP-1. This further raises the possibility that the exocytosis triggers RAB-6.1 dissociation or vice versa and that the upstream event is triggered by SEP-1 (see Discussion).

We showed for the first time, to the best of our knowledge, that a C. elegans homolog of RAB6 GTPase, RAB-6.1, is required for cortical granules exocytosis during the oocyte-to-embryo transition (Fig. 6). Previous studies have shown that in addition to membrane traffic regulators, some cell cycle regulators are also involved in cortical granule exocytosis, so a link between cell cycle regulation and cortical granule exocytosis has been suggested (Sato et al., 2006; Bembenek et al., 2007). Because SEP-1 is required for the exocytosis of cortical granules (Bard et al., 2006; Bembenek et al., 2007; Bembenek et al., 2010), the timing of cortical granule secretion is considered to depend on SEP-1 under the control of cell cycle progression. It has been postulated that the recruitment of SEP-1 to cortical granules is required for cortical granule exocytosis, but how SEP-1 is recruited is not known. Moreover, previous studies found that the only way to prevent SEP-1 localization to cortical granules was to mutate SEP-1 itself. Therefore, the importance of the SEP-1 localization to the cortical granules was not clear. In this study, we found that the recruitment of SEP-1 to the cortical granules depends on RAB-6.1. Because both cortical granule exocytosis and SEP-1 recruitment to cortical granules were defective on rab-6.1 knockdown, the hypothesis purporting the importance of SEP-1 localization to the cortical granules was supported. Our data indicate that RAB-6.1 is an important molecule that recruits separase after fertilization in meiotic anaphase I to the cortical granules to ensure timely exocytosis.

We were able to demonstrate that RAB-6.1 is essential for recruiting SEP-1 to the cortical granules, but the underlying molecular mechanism remains unclear. Because RAB-6.1 and SEP-1 co-localize on cortical granules, these 2 proteins may interact directly or indirectly. Generally, Rab GTPases interact with various effector proteins to recruit them to vesicles (Stenmark, 2009). Separase directly interacts with its regulators, securin and cyclin B (Gorr et al., 2005). We were not able to detect an interaction between RAB-6.1 and SEP-1 in a biochemical assay. Because the co-localization of RAB-6.1 and SEP-1 is observed in a very short period of time after fertilization (Fig. 5) even when RAB-6.1 is locked to its GTP form [RAB-6.1 (Q70L)] (supplementary material Fig. S9A), it may be difficult to demonstrate the biochemical interaction between the RAB-6.1 and SEP-1 proteins.

Our experiments using the GTP- and GDP-locked forms of RAB-6.1 indicated that RAB-6.1 should be in its GTP-bound form to be localized to the cortical granules (Fig. 3C). In humans, Rab6 is phosphorylated via a protein kinase C-dependent mechanism in activated platelets, and phosphorylated Rab6 has a higher affinity for GTP than GDP in vitro (Fitzgerald and Reed, 1999). The consensus phosphorylation site is conserved in C. elegans (Fitzgerald and Reed, 1999). However, no experimental support is currently available for the hypothesis that RAB-6.1 must be phosphorylated for its localization to the cortical granules. Interestingly, however, the depletion of protein phosphatase 5 (PPH-5) suppresses the localization defect for two of the three sep-1 alleles that have a specific defect in SEP-1 localization to the cortical granules (Bembenek et al., 2007; Richie et al., 2011). It has been proposed that phosphorylation is involved in the regulation of SEP-1 localization, and that SEP-1 is a plausible target of the phosphorylation (Richie et al., 2011). Alternatively, RAB-6.1 may be a target for phosphoregulation, and PPH-5 depletion may lead to increased levels of the GTP-bound form of RAB-6.1 and of the RAB-6.1 protein on cortical granules in order to suppress some of the sep-1 alleles.

The behavior of RAB-6.1 during cortical granule exocytosis would be a useful model to study the spatiotemporal regulation of Rab GTPases. We showed the dynamic behavior of a Rab GTPase, in that RAB-6.1 temporarily co-localized with SEP-1 on cortical granules for a few minutes and then was dispersed in the cytoplasm concomitantly with membrane fusion (Fig. 5A). We initially suspected that this dynamic change in the subcellular localization of RAB-6.1 might reflect conversion between GDP- and GTP-bound forms (Pfeffer and Aivazian, 2004; Stenmark, 2009). In general, guanine nucleotide exchange factors (GEFs) and GTPase-activating proteins (GAPs) switch between these two states (Stenmark, 2009; Barr and Lambright, 2010). However, we observed the normal dynamic behavior of the GTP-locked form of RAB-6.1 (Fig. 3C; supplementary material Fig. S9A), and thus, the spatiotemporal regulation of RAB-6.1 during cortical granule exocytosis may not involve GTP-GDP exchange. We found that in sep-1 (RNAi) embryos, the cortical granules did not fuse with the plasma membrane, and RAB-6.1 remained on the unsecreted cortical granules (Fig. 5C). Thus, a possible regulatory mechanism could be that SEP-1 induces cortical granule exocytosis, and the exocytosis triggers the dissociation of RAB-6.1 from the granules. Alternatively, SEP-1 may trigger the dissociation of RAB-6.1, which may in turn trigger membrane fusion. Further characterization of the relationship between SEP-1 and RAB-6.1 should provide a clue for understanding the spatiotemporal regulation of Rab GTPases and its consequences.

C. elegans strains were maintained using standard techniques (Brenner, 1974). The strains used in this study are listed in supplementary material Table S1. For construction of the CAL0384 and CAL0396 strains, the full-length DNA of rab-6.1 or rab-6.2 was amplified using polymerase chain reaction (PCR) from C. elegans genomic DNA. For construction of the CAL0621 and CAL0641 strains, the full-length rab-6.1 DNA was engineered to be in the GTP-locked (Q70L) or GDP-locked (T25N) forms by site-directed mutagenesis. The full-length gene was cloned into the pie-1-based vector pID3.01B or mCherry_N_GW vector. The latter was obtained by modifying the pie-1-based vector TH312-Cherry(n) (R. Arai, unpublished). The promoter and 3′ UTR regulatory sequences used to drive rab-6.1 and rab-6.2 expression were derived from pie-1. The pID3.01B and TH312-Cherry(n) vectors were provided by G. Seydoux and A. Hyman, respectively. Transgenic worms were created using the microparticle bombardment technique (Praitis et al., 2001). The rab-6.1 (tm2124) and rab-6.2 (tm1924) mutants were obtained from National Bioresource Project for the Experimental Animal ‘Nematode C. elegans’, and backcrossed to wild-type N2 four and five times, respectively. For rab-6.1 (tm2124), the first 556 bp including the start codon and upstream 348 bp was deleted. For rab-6.2 (tm1924), a 774-bp internal region was deleted, so it would generate a truncated protein. This truncated RAB-6.2 lacks one of the conserved GTP-binding domains and thus is probably not functional. The deletion mutation in the mutants analyzed in this study was confirmed using the single worm PCR method (Williams et al., 1992) with slight modifications and the following primer sets: 5′-TGGCTGATTTCACAAATAACGCT-3′ and 5′-AACACGGACATTGACGGCCT-3′ (Rual et al., 2004) for rab-6.1 (tm2124) and 5′-AGCGCTTTTAAGGGGGATAA-3′ and 5′-TGTACTTGGAGGACCGAACC-3′ (Sönnichsen et al., 2005) for rab-6.2 (tm1924). The rab-6.1 (tm2124) mutation was maintained by creating a balancer strain by crossing with HS732 (wrm-1(tm514) III/hT2[bli-4(e937) let-?(q782) qIs48] (I;III)). CAL0011, CAL0421, CAL0431, CAL0501, CAL0511, CAL0531, CAL0671, CAL0681, CAL0731, CAL0741, CAL0652, CAL0691, CAL0721, CAL0701, CAL0751, and CAL0711 were obtained by crossing AZ244 with OD58, RT688 with OD70, RT688 with CAL0396, WH347 with CAL0396, WH416 with CAL0396, CAL0384 with CAL0396, WH416 with CAL0621, RT688 with CAL0621, WH416 with CAL0641, RT688 with CAL0641, rab-6.1 (tm2124) with HS732, CAL0652 with RT688, CAL0652 with WH416, CAL0652 with CAL0396, rab-6.2 (tm1924) with RT688, and rab-6.2 (tm1924) with WH416, respectively.

RNAi was performed by injecting double-stranded RNAs (dsRNAs) as described previously (Hara and Kimura, 2009). The dsRNAs were amplified using PCR from C. elegans genomic DNA. The primer sets used for amplification of dsRNAs in this study are the same as those in PhenoBank (http://www.worm.mpi-cbg.de/phenobank2/cgi-bin/MenuPage.py) except in the case of rab-6.1. For rab-6.1, the following primers were used: 5′-TAATACGACTCACTATAGGCAGACAAGTTTCGACGGAGGATG-3′ and 5′-AATTAACCCTCACTAAAGGTTGACGGCCTTGGCTCTCT-3′. Worms injected with dsRNA were incubated at 22°C for 22–30 h before analysis of oocytes and embryos.

For microscopic analysis, the worms were placed in a drop of M9 buffer containing 1 mM levamisole on an agar pad (2%) and covered with a coverslip. For FM4-64 staining, embryos were dissected from gravid hermaphrodites in 150 mM KCl containing 5 mM HEPES (pH 7.2) and 33 µM FM4-64 (Molecular Probes, Carlsbad, CA) and incubated for 10 minutes at room temperature in the dark. Fluorescence images were visualized using a spinning-disk confocal system (CSU-X1; Yokogawa, Tokyo, Japan) mounted on a microscope (BX71; Olympus, Tokyo, Japan) equipped with a UPlanSApo 60×1.30 NA or 100×1.40 NA objective (Olympus) at room temperature. Digital images were obtained using a charge-coupled device (CCD) camera (iXon; Andor, Belfast, UK) controlled by the Metamorph software (Molecular Devices, Downington, PA). For analysis of osmotic sensitivity, embryos were dissected from gravid hermaphrodites in hypotonic or hypertonic solutions (50 or 250 mM KCl, respectively). To prevent compression by a coverslip, the embryos were placed in a drop of the solution in the wells of a slide glass (Multitest slide 8-well; MP Biomedicals LLC, Solon, OH) and covered with a coverslip. To obtain differential interference contrast (DIC) images of the embryos under each condition, the embryos were viewed under a microscope (BX51; Olympus) equipped with a UPlanSApo 60×1.20 NA objective at room temperature. Digital images were obtained with a CCD camera (ORCA-ER; Hamamatsu, Japan) controlled by the IPLab software (BD Biosciences, Rockville, MD).

To quantify the intracellular distribution of cortical granules, the average fluorescence intensity of CAV-1::GFP-expressing embryos was measured just before granule secretion (15–16 min after fertilization) (Fig. 4B) by using the IPLab software. The area of each embryo in the confocal section was divided into 4 zones from the outside: we first drew an oval along the edges of the embryos by judging visually. Then, we drew ovals that were 3/4, 1/2 and 1/4 the size of the original oval by using the software (Fig. 4B). The average fluorescence intensity was normalized by subtracting the average intensity of a background region.

The statistical difference in the ratio of the average intensity of CAV-1::GFP between the outside area (‘1’) and the middle area (‘2’) in each condition (Fig. 4B) was determined using the two-tailed paired Student's t-test. The normality of the data was assessed by using the Shapiro-Wilk test. P values <0.05 were considered statistically significant.

We are grateful to Drs K. Sato (Gunma University, Japan) and S. Ihara (National Institute of Genetics, Mishima, Japan) for technical advice, J. Ahringer (University of Cambridge, Cambridge, UK), A. Hyman (Max Planck Institute of Molecular Cell Biology and Genetics, Germany), H. Sawa and Y. Kohara (National Institute of Genetics, Mishima, Japan), G. Seydoux (Johns Hopkins University, USA), J. N. Bembenek (University of Michigan, USA) and National Bioresource Project (S. Mitani Laboratory, Tokyo Women's Medical University, Japan) for providing strains. Strains were also provided by the Caenorhabditis Genetics Center, funded by the National Institutes of Health. We thank Dr R. Arai (National Institute of Genetics) for sharing unpublished material, and the members of our laboratory for critical reading of the manuscript.