VHA-19 Is Essential in Caenorhabditis elegans Oocytes for Embryogenesis and Is Involved in Trafficking in Oocytes

There is an urgent need to develop new drugs against parasitic nematodes, which are a significant burden on human health and agriculture. Information about the function of essential nematode-specific genes provides insight to key nematode-specific processes that could be targeted with drugs. We have characterized the function of a novel, nematode-specific Caenorhabditis elegans protein, VHA-19, and show that VHA-19 is essential in the germline and, specifically, the oocytes, for the completion of embryogenesis. VHA-19 is also involved in trafficking the oocyte receptor RME-2 to the oocyte plasma membrane and is essential for osmoregulation in the embryo, probably because VHA-19 is required for proper eggshell formation via exocytosis of cortical granules or other essential components of the eggshell. VHA-19 may also have a role in cytokinesis, either directly or as an indirect effect of its role in osmoregulation. Critically, VHA-19 is expressed in the excretory cell in both larvae and adults, suggesting that it may have a role in osmoregulation in C. elegans more generally, probably in trafficking or secretion pathways. This is the first time a role for VHA-19 has been described.


Introduction
Parasitic nematodes are a significant burden on human health [1] and cause large losses to agriculture [2]. Nematode-specific genes are of interest because they may provide selective drug targets for treatment of parasitic nematodes, or highlight essential pathways in these important organisms. The free-living nematode Caenorhabditis elegans is an excellent model in which to study nematode-specific genes because its genome is fully sequenced and well annotated (unlike many of the parasitic nematodes of interest), and it is easy to manipulate experimentally. We identified VHA-19 in a screen for essential nematode-specific proteins in C. elegans. VHA-19 satisfied the criteria of the screen because it had less than 30% sequence identity over 150 amino acids to any mammalian proteins; it had 30% or greater sequence identity to predicted proteins from nematode parasites; and had a reported maternal sterile and lethal phenotype when gene expression was silenced by RNA interference (RNAi) [3,4,5].
This work shows that VHA-19 has a role in C. elegans reproduction. The reproductive system of the C. elegans hermaphrodite consists of two U-shaped gonad arms that each terminate with a spermatheca, connecting to a central uterus. Oocytes and spermatozoa are generated from the primordial germ cell, which resides at the distal end of each gonad arm and proliferates by mitosis. The proliferation of the primordial germ cell is controlled by the distal tip cell (Kimble & White, 1981). Hermaphrodite C. elegans produce sperm during the fourth larval (L4) stage, and switch to producing oocytes in early adulthood. Oocytes migrate from the distal gonad and line up in the proximal gonad where they are squeezed one by one through the spermatheca. Once fertilized in the spermatheca, the embryo is passed into the uterus and subsequently laid via the vulva. The egg hatches into the first larval stage, and the worm progresses through four larval stages in total before moulting into the adult hermaphrodite. Male C. elegans also occur; they exclusively produce sperm and are capable of mating with hermaphrodite C. elegans. When this occurs, male sperm are used preferentially [6,7].
In addition to reproduction, we also demonstrate that VHA-19 is involved in osmoregulation and osmotic support of the embryo. The embryo is considered protected from osmotic shock by the inner lipid layer of the eggshell [8], which is thought to be formed by the extrusion of cytoplasmic granules [9]. Osmotic support is important for early cytokinesis in the embryo, particularly in hypertonic or hypotonic media [10]. A lack of osmotic protection can lead to cytokinesis defects [10].
Here we show that VHA-19 is essential in C. elegans oocytes for the production of viable progeny and in early larval development. Within the oocyte, VHA-19 is involved in the trafficking of the oocyte receptor protein RME-2 to the plasma membrane. Since some vha-19(RNAi) embryos appear to be fertilized but are osmotically sensitive, we also hypothesize that VHA-19 is required for integrity of the eggshell and, possibly also cytokinesis, in the early embryo.

Key Features of VHA-19
VHA-19 is a 451 amino acid protein (www.wormbase.org) for which the bioinformatics applications TMPred [11] and HMMTOP [12] predict a transmembrane domain between amino acids 408-426. Also, the ProP 1.0 server [13] predicts that there is an N-terminal signal cleavage site between amino acids 16 and 17 in VHA-19 and that it is cleaved by furin between amino acids 159 and 160. Interestingly, the furin cleavage site and signal sequence are also present in translated expressed sequence tags of putative VHA-19 homologues from parasitic nematodes (not shown).
Based on a small region of similar sequence to the mammalian protein Ac45, VHA-19 is predicted to bind to the V0 subunit of the vacuolar ATPase (www.wormbase.org). This is because mammalian Ac45 has been shown to bind to several subunits of the V0 domain [14]. However, any association between VHA-19 and the C. elegans V-ATPase is yet to be shown experimentally.

vha-19 Expression is Essential for C. elegans Post Embryonic Larval Development and Embryogenesis
To investigate the role of VHA-19 in C. elegans, we fed worms on vha-19 dsRNA, commencing at the first larval stage (L1) and also from the fourth larval stage (L4). We used RNA interference (RNAi) because vha-19(tm2205) mutant C. elegans are inviable (not shown).
C. elegans fed vha-19 dsRNA from the L1 stage arrested and died as larvae by the L3/L4 stage (Table 1), indicating that VHA-19 is essential for early larval development. Feeding with vha-19 dsRNA, commencing at the L4 stage, resulted in adults that appeared grossly normal. However, only 26% (n = 233) of the progeny produced by these adults hatched into larvae (Table 2), and all of these F1 larvae arrested and died by the L3 stage.
To investigate why the eggs of vha-19(RNAi) adults failed to hatch we examined the reproductive organs of these worms and observed compressed eggs present in their uterus ( Figure 1A). These compressed eggs were first observed after 16 hours of feeding on vha-19 dsRNA and were typically located closest to the spermatheca ( Figure 1A-B). Eggs with a normal appearance were also observed in the uterus of vha-19(RNAi) adults, but they were usually located close to the vulva, indicating that they had emerged from the spermatheca earlier than the compressed eggs ( Figure 1A).
The compressed eggs observed in the uteri of vha-19(RNAi) adults probably corresponded to embryos that failed early embryogenesis. In agreement with this hypothesis, no vha-19(RNAi) embryo hatched into larvae if it was laid beyond 24 hours of exposure to dsRNA (Table 2), and beyond this time point nearly all vha-19(RNAi) adults had switched to producing compressed eggs ( Figure 1B). Thus, in addition to post-embryonic larval development, expression of vha-19 is also required for early embryogenesis and/or earlier events in C. elegans reproduction.

VHA-19 is Required in the Germline for Embryogenesis
To determine whether vha-19 expression was required exclusively in the germline for the production of normal embryos, we silenced vha-19 expression in rrf-1 mutant C. elegans [NL2098]. RRF-1 is essential for RNAi in somatic tissue, but is not required for RNAi in the germline [15]. Similar to wild type C. elegans adults fed vha-19 dsRNA ( Figure 1A), 94% (n = 83) of rrf-1; vha-19(RNAi) adults had compressed eggs present in their uterus after feeding on dsRNA for 24 hours from L4 ( Figure 1C) whereas the compressedegg phenotype was almost absent (1%; n = 76) in rrf-1; control(RNAi) adults ( Figure 1C). This implies that VHA-19 has an essential role in the germline, because silencing expression of vha-19 in the oocytes and/or sperm alone was sufficient to induce the compressed-egg phenotype.

A Sperm Defect in vha-19(RNAi) Adults is Unlikely to Explain the Presence of Compressed Eggs
Given that vha-19 expression is required in the germline, we examined the spermatozoa, oocytes and process of ovulation in more detail in vha-19(RNAi) adults to identify abnormalities. Spermatozoa observed in the spermatheca of vha-19(RNAi) adults appeared to be normal, were present at the same time as compressed eggs ( Figure 1D) and were still present after 72 hours of exposure to dsRNA. We therefore examined the ability of these spermatozoa to trigger ovulation, which occurs via the secretion by sperm of major sperm protein [16,17].
The rate of ovulation in vha-19(RNAi) adults was investigated by counting the number of eggs laid every 24 hours. As oocytes must pass though the spermatheca to be laid (whether fertilized or not), comparing the average number of eggs laid between individual control(RNAi) and vha-19(RNAi) worms allowed the rate of ovulation to be determined, and thus, whether there was a backup in the oviduct of vha-19(RNAi) worms. Interestingly, the average number of eggs laid by vha-19(RNAi) adults was not significantly different from control(RNAi) adults in the first 24 hours of exposure to dsRNA (p.0.05; Figure 2A), but was significantly lower than the number of eggs laid by control(RNAi) worms in the subsequent time periods (p,0.001; Figure 2A). This indicates that the rate of ovulation was not reduced in the time period where compressed eggs were first observed in the uterus of vha-19(RNAi) worms (0-24 hours; Figure 1B), but was reduced in subsequent time periods. There are three possible explanations for this: first, while the sperm that were observed in the spermathecae of vha-19(RNAi) worms at the 24 hour time-point ( Figure 1D) were probably able to trigger ovulation normally, vha-19(RNAi) oocytes may have been released at a slower rate beyond this time point. Second, there may have been fewer sperm in the spermatheca of vha-19(RNAi) adults overall, which would result in the rate of ovulation decreasing more quickly in vha-19(RNAi) adults. Third, vha-19(RNAi) oocytes may have become less responsive to sperm in the later time periods.

VHA-19 has a Role in the Oocytes
Notably, vha-19(RNAi) adults had significantly fewer oocytes in the oviduct than control(RNAi) worms at the 48-hour time point (p,0.05; Figure 2B), despite the presence of spermatozoa in the spermatheca. This was the case for both wild type C. elegans and H2B::GFP C. elegans [AZ212] fed vha-19 dsRNA ( Figure 2B) (H2B::GFP C. elegans were used to better visualize the nuclei of individual oocytes). This suggests that the rate of oocyte production declines prematurely in vha-19(RNAi) adults and explains why fewer eggs were laid by vha-19(RNAi) adults compared with control(RNAi) adults in the 48-72 hour time period ( Figure 2A).
Not only did the rate of oocyte production decline in vha-19(RNAi) adults compared with control(RNAi) adults; the oocytes also exhibited uncontrolled replication (endomitosis) ( Figure 2C). As early as the 24-hour time point, when the rate of oocyte production did not differ between the experimental groups ( Figure 2A), a significantly greater number of vha-19(RNAi) adults had endomitotic oocytes in the oviduct compared with control(R-NAi) adults (p,0.0001; Figure 2D). Endomitotic oocytes were visualized as oocytes with multiple nuclei, or oocytes in which the chromosomes had begun to decondense, indicating that nuclear envelope breakdown had occurred [18] ( Figure 2C). Endomitotic oocytes were also present in the oviducts of vha-19(RNAi) adults at the 48-hour time point, at a higher frequency ( Figure 2D). These observations suggest that some vha-19(RNAi) oocytes are defective before fertilization.
Collectively, the requirement of vha-19 expression in the germline for normal embryogenesis ( Figure 1C), the decline in oocyte production ( Figure 2A-B), and the presence of endomitotic oocytes in the oviducts of vha-19(RNAi) adults ( Figure 2C-D), suggest that VHA-19 has an essential role in the development of the oocytes. To confirm this, a mating experiment was designed. Virgin female fog-2 C. elegans [CB4108], which do not produce sperm and, thus, are unable to reproduce without mating [19], were fed on vha-19 or control dsRNA and mated with male fog-2 C. elegans fed control or vha-19 dsRNA. The resulting progeny from each mating were then tracked throughout their development and compared.
Only 13% (n = 415) of the progeny sired by fog-2; control(RNAi) males reached adulthood when they were produced by fog-2; vha-19(RNAi) females (Table 3), whereas all progeny sired by fog-2; control(RNAi) males and produced by control(RNAi) females reached adulthood (Table 3). Hence fog-2; control(RNAi) males were capable of siring normal progeny. Similarly, the number of eggs laid by fog-2; vha-19(RNAi) females mated with fog-2; control(RNAi) males was significantly lower than the average number of eggs laid by fog-2; control(RNAi) females mated with fog-2; control(RNAi) males (p,0.01; Table 3).These results suggest that both the proportion of progeny that reached adulthood and the number of eggs that were laid were independent of the dsRNA on which the male was fed, but rather depended on the dsRNA on which the female was fed. Thus, VHA-19 was required in the oocytes rather than the sperm, indicating that expression of vha-19 in the oocytes is essential for successful embryogenesis.

vha-19(RNAi) Oocytes do not Endocytose Vitellogenin
To explore further the cause of defective oocytes in vha-19(RNAi) adults we assayed the ability of vha-19(RNAi) oocytes to endocytose vitellogenin, using a transgenic strain in which one of the C. elegans vitellogenins, VIT-2, is fused to GFP [DH1033] [20]. Vitellogenins are lipid-binding proteins, synthesized in the intestine, that are secreted as yolk granules from the intestine into the pseudocoelom (body cavity), where they diffuse into the gonad via pores in the somatic gonadal sheath cells [21,22]. In the somatic gonadal sheath cells, yolk granules come into direct contact with developing oocytes, which endocytose the granules using the low density lipoprotein receptor, RME-2, by receptormediated endocytosis [20,21].

VHA-19 is Required for Trafficking of RME-2 to the Oocyte Plasma Membrane
To determine whether VHA-19 is involved in the traffic of RME-2, the localization of RME-2::GFP in the oocytes was compared between transgenic C. elegans adults [RT408] fed control or vha-19 dsRNA from the L4 stage. Examination of the oocytes of worms from each experimental group showed that the majority of RME-2::GFP aggregated around the nucleus of vha-19(RNAi) oocytes and did not reach the plasma membrane ( Figure 3B). The aggregates that formed at the periphery of the nucleus ( Figure 3B) resembled previously reported ''Golgi -like'' blocks in secretion [23], suggesting that a block in the trafficking of RME-2::GFP (which occurs via a secretory pathway) [20] may have occurred. In contrast, RME-2::GFP was localized to the plasma membrane of the oocytes in RME-2::GFP; control(RNAi) C. elegans ( Figure 3B). The observed block in the trafficking of RME-2 could explain why vha-19(RNAi) adults are unable to endocytose sufficient vitellogenin ( Figure 3A).

vha-19(RNAi) Embryos have a Faulty Eggshell
We questioned whether VHA-19 had a role in trafficking or secretion at other stages of reproduction and embryonic develop- ment. Because vha-19(RNAi) embryos were compressed ( Figure 1A), we investigated their eggshell. The C. elegans eggshell forms shortly after fertilization [7], provides mechanical and osmotic support for the embryo [24], and has been shown to be faulty when expression of known components of the secretory pathway are silenced by RNAi [20].
Regardless of the time point at which they were isolated, approximately 80% (n = 59) of isolated vha-19(RNAi) embryos had a visible eggshell ( Figure 4A). Some vha-19(RNAi) embryos had a clearly visible eggshell even when they were compressed ( Figure 4A), indicating that the structural fidelity, but not the presence, of the eggshell is dependent on vha-19 expression. In PLoS ONE | www.plosone.org agreement with this hypothesis, vha-19(RNAi) embryos were osmotically sensitive in hypertonic and hypotonic media ( Figure 4B), from which protection would normally be provided by the eggshell, and had a tendency to swell within their eggshell in all media types ( Figure 4B-C), suggesting that osmoregulation was also abnormal in these embryos.
vha-19 was recently identified as a candidate gene in a screen for permeable embryos [26]. Thus vha-19(RNAi) embryos might be osmotically sensitive because their eggshells were permeable. Rappeleye and colleagues [27] have shown that permeability of the embryo can reflect a defect in the lipid layer of the eggshell. While 11% (n = 136) and 34% (n = 71) of vha-19(RNAi) embryos were permeable to Trypan Blue after 24 hours and 48 hours of exposure to dsRNA, respectively, the majority of isolated vha-19(RNAi) embryos were impermeable. Hence the lipid layer may be intact. However, many vha-19(RNAi) embryos laid after 48 hours of exposure to dsRNA disintegrated in the medium when they were isolated, which may have represented a population of vha-19(RNAi) embryos that were permeable. We therefore determined whether vha-19(RNAi) embryos in the uterus were permeable to the dye Acridine Orange at this late time point. These in utero embryos were permeable ( Figure 4D), suggesting that VHA-19 may be involved in the secretion of the lipid layer of the eggshell in the embryo. However, as most vha-19(RNAi) embryos laid at earlier time points were impermeable, it is possible that the contents of the uterus of vha-19(RNAi) adults after longer periods of exposure to dsRNA represented unfertilized eggs, which would have lacked the permeability barrier. It is also possible that mechanical pressure in the uterus rendered vha-19(RNAi) embryos permeable.

vha-19(RNAi) Embryos Exhibit Signs of Defective Cytokinesis
An intact eggshell is important for early cytokinesis events in the embryo [27,28,29]. In particular, successful cytokinesis in wild type embryos requires an impermeable eggshell to provide protection from osmotic pressure: embryos with permeable eggshells exhibit cytokinesis defects in hypotonic and hypertonic media, but not in isotonic media [10]. Because vha-19(RNAi) embryos had a defect in osmoregulation in addition to being osmotically sensitive ( Figure 4C), we hypothesized that they may exhibit cytokinesis defects as well. We found that eggs cut out of the uteri of vha-19(RNAi) adults had multiple nuclei that varied in number from two upwards ( Figure 4A-B,E-F). Interestingly, some vha-19(RNAi) embryos resembled the dead embryos produced when expression of a known subunit of the C. elegans V-ATPase, vha-8, was silenced [30]. In 11% (n = 91) of isolated vha-19(RNAi) embryos, furrows were present that randomly separated clusters of nuclei ( Figure 4F), which were never observed in control(RNAi) embryos. These furrows resembled cytokinesis defects observed when wild type embryos with permeable eggshells were placed in hypotonic media [10], as well as cytokinesis defects described in rab-11(RNAi) embryos [31] and syn-4(RNAi) embryos [32], which are genes involved in the secretory pathway and membrane fusion, respectively. Interestingly, a polar body was clearly visible for a Figure 2. vha-19(RNAi) adults produce fewer oocytes beyond 24 hours of exposure to vha-19 dsRNA. A) The average number of eggs laid every 24 hours by control(RNAi) or vha-19(RNAi) adults feeding on dsRNA. Error bars represent the standard error of the mean (s.e.m). ***indicates p,0.001. B) The average number of oocytes in the proximal oviduct of either wild type or AZ212 (pie-1::H2B::GFP) C. elegans fed on control or vha-19 dsRNA for 24 or 48 hours. C) The proximal oviduct of AZ212 C. elegans fed control or vha-19 dsRNA for 24 hours. White arrows indicate normal oocyte nuclei. Endomitotic oocytes in the oviduct are indicated by a square bracket. Note that both oocytes with multiple nuclei (most proximal oocyte in square bracket), and oocytes in which the chromosomes have started to decondense prematurely (indicated by the green background of the two most distal oocytes in the square bracket), are endomitotic. The irregular, bright, particulate dots in the distal gonad (dg) are autofluorescent gut granules. Scale bars indicate 10 mm. D) The percentage of endomitotic oocytes observed in the oviducts of wild type or AZ212 C. elegans fed control or vha-19 dsRNA from the L4 stage for 24 or 48 hours. doi:10.1371/journal.pone.0040317.g002 Table 3. The progeny produced by mating fog-2 male C. elegans with fog-2 female C. elegans fed on either control or vha-19 dsRNA.    Figure 4E), indicating that they were fertilized and that cytokinesis of the oocyte pronuclei had successfully occurred.

Tissue Expression Pattern of VHA-19
Since vha-19(RNAi) embryos were osmotically fragile and exhibited osmoregulation defects, we determined whether vha-19 mRNA is expressed in tissues that control osmoregulation. We constructed a transgenic strain [WT029] in which a fragment of the presumed vha-19 promoter sequence was fused to a GFP reporter and detected GFP in the excretory cell, intestine and hypodermis of vha-19p::GFP larvae and adults ( Figure 5). These are tissues in which other known osmoregulation genes are expressed [33]. We could not detect vha-19 expression in the germline ( Figure 5), but transgenes are often silenced in the C. elegans germline [34,35]. Indeed, there is consistent evidence from genome-wide expression studies that vha-19 is both maternally and embryonically expressed [36,37,38] and in two recent studies vha-19 mRNA has been detected in the germline [39] and in germline precursor cells [40]. To confirm the GFP expression pattern we attempted to create an antibody against VHA-19, but binding of this antibody was non-specific (not shown).

Discussion
We have characterized the physiological and developmental roles of the novel C. elegans protein VHA-19 and have shown that it is essential in the oocytes for successful embryogenesis. This role somewhat resembles the described role of known V-ATPase subunits in C. elegans reproduction [30,41], which is consistent with the hypothesised association between VHA-19 and the C. elegans V-ATPase. As well as the oocytes, VHA-19 is likely to have an important role in the developing embryo. This is because most vha-19(RNAi) embryos were clearly fertilized, as indicated by the presence of an eggshell ( Figure 4A), which forms after fertilisation, and, in some embryos, a polar body ( Figure 4E). Despite being fertilized, these embryos failed to complete embryogenesis. We therefore hypothesize that in this case a post-fertilization defect occurred during formation of the eggshell and/or cytokinesis, because vha-19(RNAi) embryos were osmotically fragile and responded abnormally to osmotic media ( Figure 4B-C), and because some vha-19(RNAi) embryos resembled those observed by others when a cytokinesis defect occurs ( Figure 4E-F) [32].
There are likely to be two explanations for the compressed eggs observed in the uterus of vha-19(RNAi) adults ( Figure 1A). Eggs laid by vha-19(RNAi) adults after 24 hours of exposure to dsRNA are likely to be compressed because the structure of their eggshell is compromised, as indicated by their osmotic sensitivity ( Figure 4B-C) and their compressed nature ( Figure 4A). Eggs laid by vha-19(RNAi) adults after longer periods of exposure to dsRNA (and observed in the uterus) are likely to be compressed because the oocytes were defective, and so may not have been fertilized. In support of this explanation, eggs observed in the uterus of vha-19(RNAi) adults after 48 hours of exposure to dsRNA had often mixed together and could not be distinguished individually (e.g. Figures 1C, 4D).
An explanation for the osmosensitivity of, and general osmoregulation defect observed in, vha-19(RNAi) embryos could be that VHA-19 is involved in the trafficking of cortical granules. These form in developing oocytes and are exocytosed in anaphase I of oocyte meiosis, shortly after oocytes are activated by major sperm protein [42]. Disrupted trafficking of cortical granules results in embryos with defective eggshells and osmolarity defects [42], as observed in vha-19(RNAi) embryos ( Figure 4B-C). Further, as VHA-19 was required for proper trafficking of the receptor RME-2 in oocytes ( Figure 3B), and RME-2 is trafficked via the secretory pathway, a role for VHA-19 in the traffic or exocytosis of cortical granules is possible.
If VHA-19 were involved in a secretory pathway this could be relevant to cytokinesis as well, as regression of the cleavage furrow can occur when secretion from the Golgi to the plasma membrane is inhibited in embryos [31] and when hemicentin (a protein that is secreted into the pseudocoelom from the intestine) is absent from the germline [43]. Interestingly, when embryos fail to complete cytokinesis in the early stages of embryogenesis, nuclear division continues, resulting in embryos with multiple nuclei that are not cellularized [29,44], which was observed for vha-19(RNAi) embryos ( Figure 4A-B,E-F). Thus, is it possible that VHA-19 may be important for cytokinesis via a role in the secretory/trafficking pathway. Alternatively, since cytokinesis can be disrupted by osmotic pressure [10], and vha-19(RNAi) embryos are osmotically fragile, this may explain why some vha-19(RNAi) embryos appear to have a defect in cytokinesis.
We attempted to film vha-19(RNAi) embryos to ascertain precisely when in cytokinesis defects might occur, but could not detect actively dividing nuclei. In a few rare instances we observed pronuclei in vha-19(RNAi) embryos fusing and attempting to divide, but the cleavage furrow regressed and the nuclei collapsed together, which supports the hypothesis that VHA-19 may be involved in a critical secretion step in cytokinesis. Since most vha-19(RNAi) embryos had multiple nuclei (Figs 4A-B,E-F), this implies that nuclear division was not prevented. While it has been shown that providing osmotic support for embryos can rescue their cytokinesis defects [10], we could not rescue cell division using 150 mM KCl (a medium used for osmotic support of fragile embryos [28]) or Egg Buffer [45]. An intact eggshell is required for extrusion of the polar body and partitioning of P granules [10,28], but we observed polar bodies in at least a few vha-19(RNAi) embryos ( Figure 4E). Thus, while vha-19(RNAi) embryos are osmotically sensitive and clearly have defects in osmoregulation ( Figure 4B-C), it is not clear whether the failure of cytokinesis in these embryos is an indirect result of osmotic sensitivity or because VHA-19 has a separate role in cytokinesis.
It is also possible that VHA-19 may have a role in cytokinesis in the oocytes, because endomitotic oocytes were observed in the oviducts of vha-19(RNAi) adults ( Figure 2C-D), and vha-19(RNAi) adults produced fewer oocytes overall ( Figure 2B). Interestingly, these phenotypes were also observed when expression of known subunits of the V-ATPase were silenced [41], although the onset of the phenotypes was earlier and more severe when individual V-ATPase genes were silenced. Efficient germline proliferation requires functioning cytokinesis [46], and expression of vha-19 mRNA has been detected in the germline precursor cell [40], suggesting that VHA-19 may have a role in germline proliferation. Loss of function of cytokinesis genes also induces multinucleate oocytes [46], which were observed in the oviducts of 20% of vha-19(RNAi) adults (not shown), but this was not significantly different from control(RNAi) adults (10/46 vs. 2/34). If there was a defect in cytokinesis in vha-19(RNAi) oocytes, this could also be an indirect consequence of oocyte osmotic fragility.
Acridine Orange. Note that embryos in the uterus of control(RNAi) adult excluded Acridine Orange (arrows) whereas the embryos in the uterus of the vha-19(RNAi) adult were permeable to Acridine Orange (bracket). Scale bars represent 100 mm in D and 10 mm in A-B,E and F. doi:10.1371/journal.pone.0040317.g004 In summary, we propose that VHA-19 is involved in trafficking of molecules in the oocytes and the embryo that are essential for osmotic homeostasis, and possibly, cytokinesis. Such a role for VHA-19 would explain why multinucleate, but non-cellularized, embryos that were osmotically sensitive were observed earlier in the egg-laying cycle and also why endomitotic oocytes and a late decline in oocyte production were observed. This is the first time that a role for VHA-19 has been described in C. elegans.

C. elegans Strains and Culture
C. elegans were cultured using standard methods [47] at 20uC unless stated otherwise. The N2(Bristol) strain was used as the C. elegans wild type strain. Transgenic and mutant strains were obtained from the Caenorhabditis Genetics Center (University of Minnesota, Minneapolis, MN, USA) and were as follows: NL2098:

RNAi Experiments
C. elegans were fed on HT115(DE3) E. coli transformed with an RNAi feeding vector as previously described [3,48]. Synchronized L1 or L4 C. elegans were placed on RNAi plates expressing either negative control or vha-19 dsRNA and incubated at 20uC for 12-72 hours. The negative control was pCB19, which expresses dsRNA from a fragment of the Arabidopsis thaliana light-harvesting gene, lhcb4.3, that has no significant sequence identity to any gene in the C. elegans genome [49]. C. elegans were fed on vha-19 dsRNA using a construct (IV-1D10) purchased from Source BioScience (Nottingham, UK).

Microscopy of C. elegans
Whole C. elegans and eggs were visualized using a Leica DMLB microscope fitted with Nomarski optics (Leica Microsystems GmbH, Wetzlar, Germany) and a halogen lamp for fluorescent microscopy, or a Leica ultraviolet SP2 Confocal microscope (Leica Microsystems GmbH, Wetzlar, Germany). Images were captured using a SPOT RT-SE6 1.4 MP Slider digital camera and SPOT Advance software (Meyer Instruments, Houston, Texas, USA) for the Leica DMLB microscope and Leica confocal software (Leica Microsystems GmbH) for the SP2 UV confocal microscope. Images were analysed using either ImageJ software or the GIMP 2.2 software.

Time Course of the Number of Eggs Laid and Fate of their Progeny
Individual L4 C. elegans were fed on control or vha-19 dsRNA, incubated at 20uC and transferred to fresh RNAi plates every 24 hours for a total of 72 hours. The number of eggs and progeny produced by each adult in each time period were counted and monitored for 96 hours beyond hatching, with the developmental stage of progeny scored every 24 hours.
Embryonic arrest was defined as eggs that had not hatched after 96 hours. The average number of eggs laid at each time point was compared among experimental groups using a one-way analysis of variance (ANOVA) and a Tukey-Kramer multiple comparisons test using InStat (GraphPad Software, La Jolla, California, USA). Results were pooled from three separate experiments performed at least in duplicate.

Counting Oocytes in the Oviduct and Scoring for Endomitotic Oocytes
Wild type or AZ212 C. elegans were fed on control or vha-19 dsRNA from the L4 stage for 24 and 48 hours. The oviducts of the resulting individual adult worms were examined using either Nomarski microscopy (wild type) or confocal microscopy (AZ212) for endomitotic oocytes or multinucleate oocytes, and the number of oocytes in the proximal oviduct was also counted. Endomitotic oocytes were defined as oocytes with multiple nuclei, or oocytes in which the chromosomes had started to decondense, indicating that a breakdown of the nuclear envelope had occurred. Numbers of oocytes were compared using a Dunn's multiple comparisons test (InStat). The incidence of endomitotic oocytes was compared among experimental groups using a Fisher's exact test (nonparametric ANOVA: InStat). Results were pooled from three separate experiments.

Mating of Males with Females
Individual virgin female and male fog-2 C. elegans [CB4108] were fed on control or vha-19 dsRNA from the L4 stage for 24 hours at 20uC. The following crosses were then set up on individual plates seeded with OP50 E. coli: 1) fog-2 male and female, both fed on control dsRNA, 2) fog-2 male fed on control dsRNA, fog-2 female fed on vha-19 dsRNA and 3) fog-2 male and female both fed on vha-19 dsRNA. To ensure that mating occurred, three males were used in each cross with one female. Virgin fog-2 females were also picked to individual OP50 plates to ensure that these females could not produce eggs without mating; no eggs were ever observed (not shown). All mating plates were incubated at 20uC for 24 hours, adults were removed, and the number of eggs laid by each female (including any progeny that had already hatched into larvae) was determined. The progeny resulting from each mating were then followed for a further 8-9 days, with their stage of development (L1-adult) scored every 24 hours. The experiment was performed three times with at least five replicates of each cross per experiment and results were pooled for each experimental group. The percentage of worms that reached adulthood was compared among each group using a Fisher's exact test (Instat). The average number of eggs laid in each experimental group was compared using a Kruskal-Wallis Test (nonparametric ANOVA, Instat).

Eggshell Assays
Wild type or AZ212 C. elegans L4s were fed either control (pCB19) or vha-19 (IV-1D10) dsRNA and incubated for either 24 or 48 hours at 20uC. Approximately ten vha-19(RNAi) or control(RNAi) adults were picked into separate depression slides containing 10 ml of 150 mM KCl (Sigma-Aldrich, St. Louis, MO, USA) as recommended by Johnston et al. [28] and the sharp edge of a 26 G needle (BD, Franklin Lakes, New Jersey, USA) was used to cut adult worms in half, allowing the eggs to spill out into the medium. Eggs were scored for the presence of an eggshell (defined as a clear, translucent ring around the egg). The experiment was repeated twice for each time-point and results from experiments larva showing that vha-19 is expressed in the hypodermis (hyp) and intestine (int) but is apparently absent from the developing gonad (dg). vha-19 is also expressed in the excretory cell body (exc) and canals (thin arrowheads) of vha-19p::GFP larvae. The orientation of the larva is such that only one excretory canal is visible. Scale bar in A represents 50 mm and 100 mm in B and C. doi:10.1371/journal.pone.0040317.g005 were pooled for each time-point. The presence of an eggshell was also determined using 10 ml Egg Salts buffer [45].

Acridine Orange Staining of Eggs in the Uterus
C. elegans were fed on vha-19 or control dsRNA from the L4 stage for 48 hours, harvested by flooding plates with phosphate buffered saline (PBS), centrifuged and stained with 200 ml of 100 mg/ml Acridine Orange (Sigma-Aldrich) in 1% (v/v) dimethyl sulfoxide (Sigma-Aldrich) and incubated in the dark with agitation for 4-5 hours. Worms were washed with PBS four times to remove excess dye. Eggs in the uteri of adult C. elegans were visualised at excitation wavelength 490 nm and detected at 530 nm. The experiment was repeated and results were pooled.

Osmotic Sensitivity of Embryos
Wild type C. elegans were fed control or vha-19 dsRNA, commencing at the L4 stage, for 24 hours. At least six adults were cut at the head with a 26 G needle (BD) to release the embryos into each of 10 ml of deionised water, 150 mM KCl (Sigma-Aldrich) and 300 mM KCl (Sigma-Aldrich) on an agarose pad. Embryos were then visualised using a Leica DMLB microscope with Nomarski optics and scored for shrinking or swelling. The incidence of eggs that swelled or shrunk in the media was compared among experimental groups using a Fisher's exact test (InStat). Results were pooled from two separate experiments.

Generating Transgenic Strains
The transgenic vha-19prom::GFP construct was cloned using GatewayH cloning system [50]. Primers were designed to amplify a region encompassing 761 bp upstream of the vha-19 ATG start codon and the first 30 bp of sequence of the vha-19 exon 1. Primers were flanked with the attB4 and attB1R recombination sites (Forward primer: 59CATGGTTAAGAGCGTGCTGGCGT-CAC; Reverse primer: 59ATGAGCGAGAAAACGGCAAAAAG-TACCCTCAT). The 791 fragment was amplified from isolated C. elegans genomic DNA using KOD polymerase (Merck Chemicals, Darmstadt, Germany) in a standard reaction. The resulting polymerase chain reaction product was cloned into pDONR(P4-P1R) in a BP reaction (Invitrogen Australia; Mulgrave, Victoria, Australia) and then into pDESTDD04 using a LR reaction (Invitrogen Australia), using half the total reaction volumes recommended by the manufacturer. The final product was confirmed by sequencing and transformed into C. elegans using microparticle bombardment of unc-119(ed3) C. elegans. The resulting genotype of strain WT029 was unc-119(ed3), wtEx029[y55-H10a.1::gfp, unc-119(+)].