Requirement of V-ATPase for Ovulation and Embryogenesis inCaenorhabditis elegans *

Immunofluorescence analysis indicated that VHA-11, the C subunit of Caenorhabditis elegansV-ATPase, was localized in dot-like structures around the nuclei of early embryonic cells and was also detected in embryonic intestinal cells after comma stage. Vital staining with acridine orange showed that the intestinal cells had acidic compartments generated by V-ATPase, consistent with the intracellular localization of VHA-11. RNA interference could efficiently silence vha-11 gene expression: introduction of vha-11 double strand RNA led to embryonic lethality. Worms injected with the vha-11 double strand RNA produced embryos that became lethal. The development of embryos was arrested at various stages. However, their numbers gradually decreased, and the worms eventually became sterile due to the failure of ovulation. Similar results were obtained for RNA interference of the V-ATPase proteolipid genes. These results suggest that V-ATPases, and thus inside-acidic organelles, are required for ovulation and embryogenesis.

Vacuolar-type proton translocating ATPase (V-ATPase) 1 is a ubiquitous enzyme responsible for the acidification of cellular compartments in eukaryotic cells (for reviews, see Refs. [1][2][3]. The enzyme is a multi-subunit complex formed from a peripheral V1 sector with catalytic sites and a membrane-bound Vo sector with a proton pathway. At least 13 subunits are required for V-ATPase activity in yeast (1). Acidification generated by V-ATPase is required for intracellular processes such as receptor-mediated endocytosis, protein sorting, and zymogen activation (2). The proton electrochemical gradient established by V-ATPase energizes transporters for neurotransmitter uptake into synaptic vesicles and neurosecretory granules (4). Furthermore, the same enzyme localized in the plasma membranes has critical functions in renal intercalated cells for regulation of transepithelial acid-base transport (5), in osteoclasts for bone remodeling (6,7), and in seminal ducts for spermatogenesis (8).
In Caenorhabditis elegans, five vha genes (vha-1, vha-2, and vha-3 for c subunit; vha-4 for cЉ; vha-11 for C) encoding V-ATPase subunits have been found (9,10). They are predomi-nantly expressed in H-shaped excretory cells contributing to the excretory system in adult worms. Although all vha genes are highly expressed in embryonic stages, their functions are unknown.
During C. elegans embryogenesis, embryos undergo a series of four unequal divisions to produce five somatic founders (AB, E, MS, C, and D) (11). At the beginning of gastrulation, two E daughter cells move to the interior of an embryo to eventually generate intestinal cells. We observed that VHA-11, the C subunit of V-ATPase, was predominantly localized in the intracellular organelles of intestinal cells during late embryogenesis and that the generation of acidic compartments in the intestine was dependent on vha-11 expression. In addition, double strand RNA interference was used to silence V-ATPase gene-specific expression. Worms injected with vha-11 double strand RNA became sterile, due to the failure of ovulation. The oocytes underwent multiple rounds of DNA replication without cytokinesis and became polyploid, possibly because ovulation was not coupled with oocyte maturation in the injected worms, suggesting that V-ATPase activity or inside-acidic organelles are essential for oocyte maturation.

EXPERIMENTAL PROCEDURES
Preparation of Antibodies against VHA-11-A cDNA fragment encoding VHA-11 was cloned into the pET-32 expression vector and introduced in Escherichia coli BL21(DE3) cells (Novagen). After induction, a recombinant protein (corresponding to 99 -384 residues of VHA-11) was purified using an Ni-NTA column (Qiagen) and then injected into albino rabbits. The polyclonal antibodies were affinity-purified using the same recombinant protein.
RNA Interference-A dsRNA solution (0.2 mg/ml) was injected into the gonads of wild type worms, and the progenies arising 8 -24 h after injection were collected. At 48 h after injection, the embryonic lethality of the progenies was scored as the ratio of unhatched embryos to total progenies. For immunoblot analysis, the injected worms and the progenies were collected at 24 h after injection and treated with an alkaline hypochlorite solution (0.25 N KOH and 6% NaClO) for 7 min to isolate eggs (14). The eggs were suspended in the lysis buffer after washing with M9 buffer and then disrupted by rapid mixing in a vortex mixer. Samples were heated at 95°C for 5 min, subjected to gel electrophoresis, and then transferred onto nitrocellulose membranes. Immunodetection was carried out using an ImmunoPure Ultra-Sensitive ABC peroxidase staining kit (Pierce) and 3,3Ј-diaminobenzidine.
DAPI Staining-To stain chromosomal DNA, nematodes were picked up at 48 h after injection of dsRNA, transferred to fixing buffer (1% glutaraldehyde, 40 mM sodium phosphate buffer (pH 7.5) and 4 mM MgCl 2 ), and then incubated for more than 10 h. Fixed worms were stained for 6 h at room temperature with 1 g/ml 4Ј,6-diamidino-2phenylindole dihydrochloride n-hydrate (DAPI) in 40 mM sodium phosphate buffer (pH 7.5) and 5% glycerol.
Immunofluorescence Microscopy-Embryos were treated with the alkaline hypochlorite solution, washed extensively, fixed with methanol and acetone at Ϫ20°C, and then stained using affinity-purified anti-VHA-11 antibodies as described by Zwaal et al. (15). Immunostaining of adult worms was performed as described by Finney and Ruvkun (16). Fluorescence images were acquired with an LSM 510 (Carl Zeiss) or a BX50 (Olympus) microscope.
Vital Fluorescence Staining-Embryos prepared by treatment with the alkaline hypochlorite solution were incubated for 30 min in staining buffer (40 M acridine orange and 1% dimethyl sulfoxide) in the presence or absence of 25 M bafilomycin A1 (Wako). After washing with the phosphate buffer three times, the embryos were mounted on slide glasses. Differential interference contrast images were obtained with an LSM 510 (Carl Zeiss) microscope. For fluorescence images, a 560-to 615-nm band path was used for excitation with the LSM510, and a band path longer than 515 nm was selected for the BX50 (Olympus).

Localization of V-ATPase in Embryonic Intestine and Adult
H-shaped Excretory Cells-We have identified five vha genes (vha-1, vha-2, vha-3, vha-4, and vha-11) encoding V-ATPase subunits in C. elegans (9,10). Their expression was determined using the corresponding promoter sequences connected upstream of GFP or lacZ reporter gene. These promoters were highly active in H-shaped excretory cells and the rectum of adult worms (9,10).
To identify the cells and organelles in which V-ATPase is localized, polyclonal antibodies against VHA-11 were raised. VHA-11 is a functional V-ATPase subunit and has no isoforms in nematodes (10). As shown in Fig. 1A, the affinity-purified antibodies recognized a single protein band (40 kDa) on an immunoblot of a wild type nematode lysate. A band corresponding to the same size was also detected for the lysate of yeast cells harboring the vha-11 expression plasmid but not for that of yeast cells with a vector. These results indicated that the antibodies specifically recognized a 40-kDa vha-11 gene product.
Using the anti-VHA-11 antibodies, immunohistochemical analysis was carried out. In the adult stage, VHA-11 was expressed mainly in an H-shaped excretory cell (Fig. 1B), and also in intestinal cells (data not shown), consistent with expression of the vha-11::GFP reporter gene (10). In embryonic stages, VHA-11 was detected as dot-like intracellular compartments around the nuclei (Fig. 1C, arrowheads). Diffuse staining in the cytoplasm could be seen at all embryonic stages. Beginning at the comma stage, dense dot-like staining became clearly visible in intestinal cells (Fig. 1D, arrowheads) and was detectable during embryonic development. These results indicate that V-ATPase is localized in intracellular compartments FIG. 1. Immunochemical analysis of VHA-11. A, detection of VHA-11 in C. elegans total lysate. A total lysate of yeast cells carrying either a vector (Vector) or the vha-11 expression plasmid (pVHA-11), and a C. elegans lysate (C. elegans) were subjected to polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate, blotted onto nitrocellulose membranes, and then incubated with affinity-purified anti-VHA-11 antibodies. The amounts of the yeast and C. elegans lysates used were 30  of embryos, especially those of intestinal cells.
Generation of Acidic Compartments in the Embryonic Intestine by V-ATPase-The predominant expression of V-ATPase in the embryonic intestine suggests that acidic organelles are present in intestinal cells. Consistent with this suggestion, acridine orange, a weak basic dye, was highly accumulated in embryonic intestinal cells (Fig. 2, A and B). The accumulation was detected predominantly in the cytoplasmic perinuclear regions of intestinal cells in confocal sections (Fig. 3), indicating that inside-acidic compartments are present in the embryonic intestine. The acridine-orange staining disappeared upon either treatment with bafilomycin A1, a specific inhibitor of V-ATPase (Fig. 2C), or the addition of 5 mM ammonium chloride (data not shown). Inactivation of vha-11 expression by RNA interference prevented the accumulation of acridine orange in the intestine ( Fig. 2D and also see below). These results indicate that V-ATPase mediates the acidification of intracellular compartments in the embryonic intestine.
Embryonic Lethality Caused by Silencing of vha-11 Expression-RNA interference is effective for abolishing expression of a specific gene in C. elegans (13). We injected vha-11 dsRNA into adult worms to disrupt the gene expression in progenies. At 24 h after injection of the vha-11 dsRNA, the VHA-11 protein in embryos became undetectable, whereas introduction of the control lacZ dsRNA was ineffective (Fig. 4A). Silencing of vha-11 expression was observed not only in embryos retained in the uterus but also in eggs harvested from plates (Fig. 4A,  lanes 2 and 4). These results suggest that introduction of the vha-11 dsRNA completely turned the gene expression off during embryogenesis.
All embryos lacking vha-11 expression were essentially lethal (Table I), and their development was arrested at various stages. 15-30% of the embryos arrested at the one-cell stage. More than 30% of the embryos became lethal after gastrulation. The development of the other embryos was arrested between two-cell stage and gastrulation. The disappearance of VHA-11 in the embryonic intestine of vha-11 dsRNA-injected worms was confirmed immunochemically: no dot-like staining was detected in the intestine of comma-stage embryos (Fig. 4D). On the other hand, significant staining was observed in embryos from lacZ dsRNA-injected worms as well as those  without injection (Fig. 4, B and C). It was noteworthy that diffuse staining in the cytoplasm was also decreased in embryos from vha-11 dsRNA-injected worms. Consistent with the loss of VHA-11, acridine-orange staining was not observed in intestinal cells of embryos from the vha-11 dsRNA-injected worms (Fig. 2D). These results indicate that vha-11 is an essential gene for embryogenesis, particularly the formation of acidic compartments in the intestine.
Vo Subunit Genes vha-1 and vha-4 Are Required for Embryogenesis but Not vha-2 and vha-3-Because worms became lethal upon silencing of gene expression of vha-11 for the C subunit in the peripheral V1 sector, it was of interest to examine the genes for Vo subunits. A series of dsRNAs for proteolipid subunits was used to examine the effects on embryogenesis. C. elegans V-ATPase has two proteolipids, i.e. proteins of 23 kDa (VHA-4) and 16 kDa (VHA-1, VHA-2, and VHA-3) (9, 10). Introduction of vha-4 dsRNA led to embryonic lethality, similar to the case of vha-11 dsRNA (Table I), indicating that 23-kDa proteolipid is required for embryonic development.
Embryonic lethality was also observed with the vha-1 or vha-2 dsRNA coding for a 16-kDa proteolipid. However, introduction of the vha-1 or vha-2 dsRNA may abolish the expression of other 16-kDa proteolipid genes, because they exhibit high identity: 60% amino acid identity between VHA-1 and VHA-2; 100% between VHA-2 and VHA-3 (9, 10). Expected from the high identity of VHA-1, VHA-2, and VHA-3, no specific antibodies could be produced (data not shown). Thus, non-identical 3Ј-untranslated region (UTR) of each vha gene (vha-1, vha-2, or vha-3) was examined for embryonic development together with the presence of the corresponding transcript. The dsRNA for vha-1 UTR caused embryonic lethality, but those for vha-2 UTR and vha-3 UTR did not (Table I). Consistent with the results, the vha-1 transcript was essentially undetectable after injection of the vha-1 UTR dsRNA, although the vha-2 and vha-3 transcripts were present (Fig. 5). In contrast, injection of the dsRNA for vha-2 UTR caused the vha-2 and vha-3 expressions to silence, but significant amounts of vha-1 were expressed. The dsRNA for vha-3 UTR lowered only the corresponding transcript. Upon injections of vha-2 UTR ϩ vha-3 UTR dsRNA, no embryonic lethality was observed (Table I), and the vha-1 and vha-3 transcripts were detectable in embryos from the injected worms. These results indicate that the vha-1 expression is essential for embryogenesis, whereas vha-2 and vha-3 are not required.
Defective Ovulation in Worms Injected with the dsRNAs for V-ATPase Subunits-Because the acidic organelles generated by V-ATPase are important for embryogenesis, we were also interested in their roles in oogenesis. Nematode gonads display distal-proximal polarity in hermaphrodites (17), germ cells proliferate in the distal gonads, and oogenesis occurs in the loop and proximal regions (Fig. 6A). Condensation of chromosomes in diakinesis is observed in the proximal gonad (Fig. 6B,  arrowheads).
We observed that vha-11 dsRNA-injected worms were able to produce eggs. However, the egg numbers gradually decreased, and the worms finally became sterile after 24 h. This sterility continued for about 4 days, and then the worms started pro- FIG. 4. Immunofluorescence analysis of embryos from vha-11 dsRNAinjected worms. A, VHA-11 became undetectable after injection of vha-11 dsRNA. Embryos (1-g proteins) from worms injected with either lacZ (lanes 1,  3) or vha-11 dsRNA (lanes 2, 4) were subjected to gel electrophoresis and immunoblotting. Proteins were extracted from embryos either prepared from gravid worms (lanes 1, 2) or harvested from agar plates (lanes 3, 4). The arrowhead indicates the position of VHA-11. B-D, comma stage embryos from control (B), lacZ dsRNA-(C), and vha-11 dsRNAinjected (D) worms were immunochemically stained with antibodies against VHA-11: arrowheads indicate dot-like structures associated with VHA-11 in the intestine. Scale bar indicates 10 m.

TABLE I Injection of dsRNAs for V-ATPase subunits leads to embryonic
lethality Each dsRNA (0.2 mg/ml) was injected into hermaphrodites. To exclude prefertilized eggs, the injected worms were transferred to new plates at 8 h after injection. They were incubated for further 16 h, and then eggs were collected. Embryonic lethality was scored upon hatching of the progenies at 48 h after injection. UTR, 3Ј-untranslated region. ducing viable eggs again. To visualize the germ cell nuclei in the sterile worms, they were fixed at 48 h after injection of the vha-11 dsRNA and then stained with DAPI. Oocytes with condensed chromosomes at the diakinesis stage of meiotic prophase ⌱ were also detected (Fig. 6C, arrowheads). They were found in the distal gonads near the loop region, but in the proximal gonads of non-injected worms (Fig. 6B). The location was different in the injected worms, because abnormal oocytes with strong DAPI staining filled the proximal gonad (Fig. 6C). The chromosomes became loose and dispersed in the cytoplasm, showing H-shaped staining (Fig. 6D). Spermathecae were observed near the vulva due to the accumulation of oocytes, and no fertilized embryos were found in the uterus (Fig. 6, E and F). These findings indicate that oocytes are trapped in the gonads and then undergo endomitotic DNA replication, because ovulation fails. Endomitotic oocytes contain polyploid nuclei possibly due to the multiple rounds of DNA replication without cytokinesis (18 -20). Essentially the same results were obtained on injection of the vha-1, vha-2, or vha-4 dsRNA. These results indicate that V-ATPase, and thus acidic organelles, are essential for oocyte ovulation. DISCUSSION V-ATPase was localized immunochemically in dot-like structures of embryonic cells using antibodies against VHA-11. It may be associated with intracellular organelles such as lysosomes and endosomes, similar to in other eukaryotic cells (2). In comma stage embryos, V-ATPase was highly expressed in intestinal cells being localized in their intracellular compart-ments. The distribution in embryos was similar to that of the acidic compartments stained with acridine orange. The C. elegans intestine is known to contain numerous storage granules (21). Furthermore, labeled probes are taken through endocytosis and accumulated within granules in which acridine orange is stored (22), suggesting that the acidic pH generated by V-ATPase is required for nutrient intake by intestinal cells.
We used RNA interference (13) as a powerful tool for silencing V-ATPase gene-specific expressions. The progenies exhibited defects dependent on the time after injection of dsRNA. C. elegans germ cells in the gonads exhibit distal-proximal polarity during proliferation and meiotic prophase progression (17). Thus, each germ cell may show a defect, depending on the progress of oogenesis.
vha-11 dsRNA-injected worms produced eggs up to 24 h, but the eggs did not hatch. The development of the eggs was randomly arrested at different stages. The amount of vha tran-

FIG. 5. Detection of transcripts for vha-1 and vha-2 in embryos.
Transcripts for vha-1, vha-2, and vha-3 were detected in embryos from worms injected with dsRNA for vha-1 UTR, vha-2 UTR, vha-3 UTR, or vha-2 UTR ϩ vha-3 UTR by RT-PCR using the corresponding primers. As a control, transcript for ribosomal protein (rp21) was amplified. The PCR products were separated by polyacrylamide gel electrophoresis, and the gel was stained with SYBR Green I (Molecular Probes).

FIG. 6. Worms injected with vha-11 dsRNA have a defect in ovulation.
A, the anatomy of a hermaphrodite gonad. Germ cells in gonads display distal-proximal polarity. The distal cells proliferate in the mitotic cell cycle. Moving proximally, germ cells enter the pachytene stage of the meiotic prophase. In the loop and proximal regions, they progress to the diakinesis stage and undergo nuclear envelope breakdown before fertilization. Ovulation transports mature oocytes to the uterus through the lumen of the spermathecae where fertilization occurs (17). B, a posterior gonad of a wild-type adult hermaphrodite. A control worm (non-injected) was harvested, fixed with glutaraldehyde, and stained with DAPI. Small and dense nuclei in the distal gonad were found in the mitotic cell cycle. Oocytes in the proximal gonad contained condensed nuclei (arrowheads) of the diakinesis stage. C, a posterior gonad of a vha-11 dsRNA-injected worm. Endomitotic oocytes brightly stained with DAPI were found in the proximal gonad. Oocytes at the diakinesis stage (arrowheads) were observed in the loop and distal regions, because of occupation of the proximal gonad by endomitotic oocytes. D, a high magnification view of endomitotic oocytes of a vha-11 dsRNA-injected worm. The chromosomes became loose and dispersed in the cytoplasm. An H-shaped structure of chromosomes was often observed when it was dispersed. E, proximal gonads near the vulva of a vha-11 dsRNA-injected worm. Endomitotic oocytes accumulated up to the proximal region before spermathecae (arrowheads), and no fertilized embryos were observed in the uterus. scripts is high at the embryo rather than the larval and adult stages (10). Thus, the difference in the stage at the time of arrest may depend on the amount of maternal mRNA present in each embryo. 15-30% of the embryos arrested as one-cell stage, suggesting that the cell division cycle may require acidic compartments generated by V-ATPase. Inactivation of proteolipid subunit genes also caused embryonic lethality. Furthermore, similar results were obtained on injection of the dsRNA for F55H2.2, which is homologous to genes for the D subunit in the V1 sector. 2 These results suggest that V-ATPase activity itself is essential for the progress of embryonic development.
Yolk granules are embryonic organelles and store essential nutrients that support embryonic development (30). Their contents, yolk proteins, are derived from vitellogenins and have been shown to be taken up into acidic compartments of Xenopus oocytes through endocytosis (31). YP170, one of the C. elegans vitellogenins, is also taken up by oocytes through an endocytotic pathway mediated by the RME-2 receptor protein (32). Thus, it is reasonable to assume that disruption of the genes required for endocytosis affected oogenesis. As expected, worms injected with the dsRNA for clathrin or adaptin gene produced non-viable embryos (32). Because V-ATPase acidifies the endocytotic compartments, one of the reasons for the embryonic lethality due to the dsRNAs of the vha genes may be the inhibition of yolk protein uptake by oocytes.
The accumulation of endomitotic oocytes was observed in the gonads of worms injected with the dsRNAs for V-ATPase subunits. So far, several mutants defective in ovulation have been reported (18 -20). One of them, the emo-1 mutant, has a zygotic defect in ovulation and eventually produces oocytes with polyploid nuclei (18). Because the emo-1 gene encodes a homologue of Sec61 gamma, which is a component of the protein translocation pore in the endoplasmic reticulum in eukaryotic cells, it is proposed that the emo-1 mutation disrupts the transport of oocyte proteins required for ovulation (18). In addition to emo-1, the introduction of the dsRNAs for small GTPase (ATP-ribosylation factor) and coatmer genes necessary for clathrin-mediated endocytosis also leads to the accumulation of endomitotic oocytes (32), supporting the hypothesis that the inhibition of protein transport causes ovulation failure and the production of polyploid nuclei. A defect in oocyte ovulation caused by the dsRNAs for V-ATPase subunits is consistent with reports that acidic compartments are essential for protein trafficking such as endocytosis (3,33).