The C. elegans eggshell^*

In the sections below, we will suggest some nomenclature changes based on conversations with those in the field and reviewers of this chapter. The first being that we number each layer according to the order in which they are deposited into the eggshell. The second suggestion is that we rename the spaces as layers as well.

For many years, the C. elegans eggshell was known to be composed of three strata (Hall and Altun, 2008; Rappleye et al., 1999; Wharton, 1980; Bird, 1971) and the perivitelline space [now referred to as the extra-embryonic matrix (EEM), as suggested by a recent review (Johnston and Dennis, 2012)]. Two additional layers have recently been observed and defined (Olson et al., 2012; Benenati et al., 2009) (Figure 1). These layers are defined visually through electron micrographs and also by diagnostic biochemical treatments that can remove individual layers from the embryo (Olson et al., 2012; Benenati et al., 2009; Rappleye et al., 1999; Schierenberg and Junkersdorf, 1992). The outermost layer is the vitelline layer (VL) and can be selectively removed by treatment with sodium hypochlorite, leaving the remainder of the eggshell intact. This layer is present on the oocyte surface prior to fertilization. The next layer is the chitin layer (CL), digestible by chitinase. When this layer is removed, the embryo loses its ovoid shape and rounds up into a sphere. The third layer was previously thought to be a lipid-rich layer, based on chemical composition experiments done in Ascaris (Foor, 1967). However, Olson et al. (2012) have recently shown that this third layer is composed of proteoglycans. It is now referred to as the chondroitin proteoglycan (CPG) layer (Olson et al., 2012). The three outer layers are now known as the trilaminar outer eggshell. The fourth layer was previously known as the perivitelline space because it originally sat adjacent to the vitelline membrane, a name no longer used for the fifth layer (see Section 6 and Section 7). Johnston and Dennis (2012) suggested calling this layer the extra-embryonic matrix (EEM), a name we will use throughout this chapter. The EEM is an asymmetric layer that sits below the CPG layer (of the trilaminar outer eggshell). This layer is larger at the anterior end where the first extruded polar body resides. It is thought that this layer is fluid-filled and contains a number of proteins (Olson et al., 2012; Johnston et al., 2006). The fifth layer, which is separated from the trilaminar outer eggshell by the EEM, is now thought to be the lipid-rich layer, which makes up the osmotic/permeability barrier for the developing embryo. Olson et al. (2012) refer to this layer as the permeability barrier and its function is to keep out most small molecules and keep water in so the embryo stays hydrated. This layer has also been called the embryonic layer, due to its close apposition to the embryo (Benenati et al., 2009). Residing between the permeability barrier layer and the embryo proper is an amorphous space now called the peri-embryonic layer (layer six; Figure 1) (Olson et al., 2012). While Benenati et al. refer to the entire region as the embryonic layer, Olson et al. distinguish the permeability barrier as a thin membrane-like structure that encloses the peri-embryonic layer beneath it. Please refer to Section 7 for an expanded definition of these two layers. All of these layers are thought to be synthesized by the embryo in a hierarchical assembly pattern (Olson et al., 2012). Another layer has been observed in numerous nematode eggshells and has been called the uterine layer, the uterine coat, or the post-ovulatory envelope (Wharton, 1980; Foor, 1967). This layer is secreted by uterine cells as the embryos pass through the uterus. All of these layers will be discussed in greater detail in Sections 3 through 7, and 9.

The impermeable eggshell is established by the end of anaphase II of meiosis, with final modifications being made as the embryo undergoes its first few mitotic divisions (Bembenek et al., 2007; Zhang et al., 2005). From studies with Ascaris, the vitelline layer is present on the oocytes in the oviduct (Foor, 1967). Upon entry into the spermatheca, the oocyte, which is in prophase of meiosis I, is fertilized by a single sperm. A single calcium wave is the earliest event observed in the spermatheca (Samuel et al., 2001); this calcium spike lasts 2-4 seconds. Recently, Jun Takayama and Shuichi Onami probed the fertilization-induced Ca²⁺ transient in C. elegans oocytes and found that fertilization triggers biphasic Ca²⁺ waves in oocytes. In addition to a fast local wave that appeared at the site of sperm entry, similar to that observed by Samuel et al. (2001), a slow global wave was also observed to spread from this site to the opposite pole (Takayama and Onami, 2016).

Each ovulated oocyte spends about 3-5 minutes in the spermatheca. At some time during this period it is fertilized before passing through the spermathecal/uterine valve into the uterus. In the uterus, the maternal chromosomes undergo both meiotic divisions in the span of 20 minutes; the first mitotic division takes place 40 minutes after fertilization (at 20^oC). Evidence of the chitin layer of the eggshell has been observed in the spermatheca (Maruyama et al., 2007). Fertilization is the most likely trigger for the activation of the enzyme chitin synthase (CHS-1), responsible for the synthesis of the chitin layer. Synthesis of chitin is not dependent on cell cycle progression: zygotes arrested in metaphase of the first meiotic division still show normal chitin distribution on the +1 zygote (+1 refers to the position of the newly fertilized zygote relative to the spermatheca; the -1 is the oocyte adjacent to the spermatheca that is the next one to be fertilized) (Johnston et al., 2010). The formation of the third layer, the chondroitin proteoglycan layer, does depend on cell cycle progression. Cortical granule exocytosis (CGE), which takes place during anaphase I, is responsible for establishing this layer (Olson et al., 2012; Bembenek et al., 2007). The next layer, which is separated from the outer three layers by the EEM, is the lipid-rich permeability barrier. It is synthesized during meiosis II (Olson et al., 2012). The zygote thus becomes impermeable to dyes by anaphase II (see Figure 1 for a schematic of these layers).

Very little is known about the origin or the composition of the vitelline layer (VL). Much of what we know about this layer comes from studies with Ascaris lumbricoides. There is some agreement that this layer is present on the oocyte and that it is maintained post-fertilization (Wharton, 1980; Foor, 1967). It appears to be coincident with the plasma membrane in the oocyte, but separates from it upon fertilization (Hall and Altun, 2008; Foor, 1967). This layer might mediate sperm-oocyte binding, similar to what has been observed in other species. How the sperm successfully penetrates this layer in C. elegans is unknown. There are no published observations of sperm-egg fusion or the subsequent fusion site to shed light on how this penetration might occur. Although there is a report of some lectin staining enriched in the VL (Bembenek et al., 2007), there is not, to our knowledge, a way to distinguish it from the plasma membrane of the oocyte. This layer can be removed from mitotic embryos using sodium hypochlorite (Schierenberg and Junkersdorf, 1992) without harming their development.

The cbd-1 (chitin-binding domain protein; H02I12.1) gene encodes the only known VL marker to date. This gene is essential for embryonic development (Johnston et al., 2010). CBD-1::GFP is localized to the surface of developing oocytes and embryos (Johnston et al., 2010). Observations from our lab with a transgenic CBD-1::mCherry line support the conclusions that CBD-1 is a vitelline layer marker. CBD-1::mCherry can be stripped off the embryo with hypochlorite treatment, further supporting the conclusion that it is a component of the vitelline layer (D. N. Levine, K. K. Stein, and A. Golden, unpublished observations).

CBD-1-depleted oocytes exhibited precocious phospho-histoneH3 antibody staining and the ovulated oocytes are small, often pinched in two as they enter the uterus, indicative of possible eggshell defects (Johnston et al., 2010). CBD-1 also appears to have a role in the proper formation of the eggshell as embryos from cbd-1 RNAi-treated hermaphrodites are fertilized by more than one sperm, and produce an eggshell that is misshapen with chitin localized only at one end of the embryo. Furthermore, CBD-1 is required for membrane localization of EGG-1 and EGG-2 on the oocyte (Johnston et al., 2010), which are, in turn, essential for the membrane localization of CHS-1, EGG-3 and MBK-2 (see Section 4).

Chitin is a polymer of β-(1,4)-linked N-acetyl-glucosamine residues (Figure 2) that is prevalent in the support structures of fungi (cell wall), nematode embryos (eggshell and pharynx), and arthropods (exoskeleton) (Merzendorfer, 2006). In nematodes, chitin has long been implicated as a component of the eggshell because treatment of the embryo with chitinase is necessary for the removal of the eggshell (Edgar, 1995; Wharton, 1980). The chitin layer (CL) provides the mechanical support for the oval embryo; embryos without chitin are round and misshapen. The chitin eggshell can be thought of as a meshwork or sponge-like structure. It is a porous, but rigid, structure that gives the embryo its ovoid shape.

The presence of chitin in the nematode eggshell has been directly demonstrated using either a fluorescently-conjugated or immunogold-labeled chitin-binding domain (CBD) probe (New England Biolabs) in C. elegans (Olson et al., 2012; Parry et al., 2009; Maruyama et al., 2007; Johnston et al., 2006) (Figure 1 and Figure 2). Chitin is evident in the eggshell of a newly fertilized zygote at metaphase I (Maruyama et al., 2007). Chitin is not detectable on oocytes, and is dynamically modified in the early embryo: the fluorescent CBD probe can easily detect chitin in the eggshell of the +1 zygote, but this binding diminishes with subsequent embryonic development, beginning at the 2-cell stage (Zhang et al., 2005). Treatment of embryos with sodium hypochlorite restores CBD probe binding throughout embryogenesis, indicating that the chitin layer is modified by the outer vitelline layer or a sodium hypochlorite-sensitive substance that is present on the zygote after its exit from the spermatheca (Zhang et al., 2005). Interestingly, WGA labeling of the eggshell brightens as the embryo ages, and thus serves as another marker for eggshell modification (Bembenek et al., 2007). The most likely source of modifications are proteins that are being secreted by the uterine cells and binding to the eggshell. To ask whether the chitin layer modifications of the eggshell are affected by vesicle trafficking, Sato et al. (2008) stained control, metaphase I-arrested zygotes, and RAB-11.1-depleted embryos with the CBD probe. Rab GTPases play key roles in vesicle transport in many organisms: of the 30 rab genes in C. elegans, RAB-11.1 depletion strongly inhibited cortical granule exocytosis (but not cortical granule formation). In control and metaphase I-arrested zygotes, the CBD probe is present on the +1 zygote and then disappears by the +2 embryo stage. This observation suggests that in the absence of cortical granule fusion (in the metaphase I-arrested zygotes), the modification of the chitin layer still occurs normally. Thus, whatever is responsible for this chitin modification is still present when cortical granule exocytosis is blocked. However, RAB-11.1-depleted embryos had heavy and persistent labeling of the eggshell with the CBD probe in all embryos. These findings suggest that the remodeling of the CL also relies on a RAB-11.1-dependent vesicle transport pathway that does not include the cortical granules.

The genes required for the synthesis of the CL all share a common phenotype when depleted. Embryos lacking the CL of the eggshell are misshapen and extremely fragile. The fertilized zygotes are often crushed as they exit the spermatheca; this is in stark contrast to an unfertilized oocyte, which can pass into the uterus with little damage. Chitin-deficient zygotes fail to extrude polar bodies even though they undergo the meiotic divisions. Most embryos accumulate in the uterus as an indistinct mass; those that make it out of the spermatheca and uterus without being crushed are multinucleate, polyploid one-cell embryos that fail to undergo cytokinesis. It should be no surprise that such embryos also do not establish embryonic polarity (Benenati et al., 2009; Johnston et al., 2006). The polarity phenotypes described for many eggshell synthesis genes could be due to the observation that polarity cues depend on microtubules. Embryos that fail polar body extrusion are left with an excess of chromosomes and microtubules. Such embryos would be expected to display defects in polarity. Even one-cell arrested embryos have polarity defects because of the persistence of the meiotic spindle (Wallenfang and Seydoux, 2000). Thus, it is likely that the meiotic phenotypes and polarity defects in many eggshell mutants are indirect, caused by the lack of a chitinous eggshell, rather than a result of a specific role of chitin synthesis genes in these processes.

Below is a description of a number of proteins involved in the synthesis of the CL of the eggshell.

The third layer of the trilaminar outer eggshell had been thought to be a lipid-rich layer for the past 100 years or more. The reason this layer was thought to be lipid-rich was based on chemical extraction and analysis (Chitwood, 1938; for review, see Christenson, 1950). Early observations by light microscopists could discern the VL and CL of the outer eggshell, as well as an inner layer, often referred to as the vitelline membrane (see Section 7). We now know this layer to be the lipid-rich permeability barrier. Later, when electron microscopists began looking at the nematode eggshell, their fixation techniques likely destroyed this permeability barrier. The trilaminar outer eggshell survived, and the innermost layer was assumed to be the lipid-rich permeability barrier (S. Olson, personal communication). The recent findings by Olson et al. (2012) have dramatically altered our view of the C. elegans eggshell; these new findings strongly suggest that this third layer of the trilaminar outer eggshell is composed of proteoglycans. The two best-characterized proteoglycans in this layer are CPG-1/CEJ-1 and CPG-2 and is thus termed the CPG layer (Figure 1 and Figure 3). This layer is the third layer to be synthesized. Due to the sequential synthesis of the eggshell, perturbations of the CPG layer also result in defects in the synthesis of the fifth layer, the lipid-rich layer (see Section 7). Since it is the lipid-rich layer that makes up the permeability barrier, RNAi conditions or mutants that disrupt the formation of the CPG layer will result in embryos that are permeable due to defects in the subsequent formation of the lipid-rich layer.

There is a fluid-filled layer that separates the trilaminar outer eggshell from the permeability barrier (see Section 7) and embryo. This layer has previously been called the perivitelline space, perimembrane space, or the extra-embryonic space. See Figure 7 for a nice example of this layer (Benenati et al., 2009). Historically, because the permeability barrier is visible by light microscopy, it was called the vitelline membrane, and the space surrounding this membrane was called the perivitelline space. Since the terminology of vitelline membrane is no longer in use, this fluid-filled layer is now called the extra-embryonic matrix (EEM) (Johnston and Dennis, 2012). This layer is most obvious when observed at the anterior end of young embryos where there is a lot of space between the eggshell and the permeability barrier of the embryo. This layer is much less obvious along the dorsal or ventral surfaces of the embryo. It is thought that this layer is fluid-filled and contains a number of proteins (Olson et al., 2012; Johnston et al., 2006). The CPG-2::mCherry protein, for example, is an excellent marker for the extra-embryonic matrix (Figure 6) (Olson et al., 2012). Defects in the synthesis of chondroitin account for abnormalities in the synthesis of this layer, as evident from the analysis of the sqv mutants (Hwang and Horvitz, 2002). Without this layer, there is no separation between the plasma membrane of the embryo and the trilaminar outer eggshell; this defect likely accounts for the cytokinesis failures in all sqv (and cpg) mutants. It is thought that the plasma membrane and eggshell adhere to each other in the absence of this layer (as observed when CPG-1 and CPG-2 are co-depleted), thus contributing to these cytokinesis failures (Olson et al., 2012).

The permeability barrier is the layer of the eggshell that is responsible for the osmotic integrity of the embryo. It serves as a permeability barrier to prevent large molecules or toxins in the external environment from penetrating into the embryo. It has been shown that embryos can develop normally (in osmotic support media) if both the VL and CL (and presumably the CPG layer) are removed, leaving just the permeability barrier to protect the embryo (Schierenberg and Junkersdorf, 1992).

Dyes such as OG-phalloidin, fluorescein, and fluorescein dextran have been used to help visualize the permeability barrier (Figure 8) (Olson et al., 2012). A useful assay to test the integrity of this barrier is to soak embryos in the lipophilic dye FM4-64. Wild-type embryos that have completed the meiotic divisions are impermeable to this dye, whereas zygotes arrested in metaphase I are permeable (Sato et al., 2008; Johnston et al., 2006; Rappleye et al., 1999). This dye will be incorporated into the plasma membrane of the embryo if no permeability barrier exists (Figure 9). Embryos depleted of RAB-11.1 (see Section 5) were permeable to this dye, further suggesting that the cargo of the CGs is responsible for indirectly establishing the permeability barrier (Sato et al., 2008).

This fifth layer has only recently been identified as the permeability barrier; it formerly was known as the vitelline membrane (Chitwood, 1938; Bird, 1971), a structure visible by DIC microscopy that surrounds the embryo proper (whose boundary is defined by its plasma membrane). It is separated from the trilaminar outer eggshell by the extra-embryonic matrix. The permeability barrier encloses the second polar body (Figure 1, Figure 10, and Figure 11). The definition of the permeability barrier differs in two recent publications (Olson et al., 2012; Benenati et al., 2009). The definition by Benenati et al. (2009) described a layer that completely surrounds the embryo and is approximately 100-200 nm thick (Benenati et al., 2009). While Benenati et al. refer to this as a single layer, Olson et al. (2012) consider this as two distinct layers. They describe the lipid-rich permeability barrier as a thin membrane-like structure (Olson et al., 2012), distinct from the fluid-filled layer that lies beneath it [which they call the peri-embryonic space (Figure 1, Figure 10, and Figure 11) and Benenati et al. refer to as a filamentous embryonic layer (Benenati et al., 2009)]. This layer also provides protection for the embryo and directly apposes the embryo proper. This layer appears to be quite fluid and amorphous (Figure 12) and likely accounts for the high mobility of the second polar body during early embryonic development. It is likely that novel processing and fixation conditions are responsible for the visualization of this previously unnoticed layer. We propose calling this sixth layer the peri-embryonic layer, rather than peri-embryonic space (Olson et al., 2012) or embryonic layer (Benenati et al., 2009). We also agree that the permeability barrier should be considered a separate layer of the eggshell.

Recent studies have shown that formation of the permeability barrier requires passage through anaphase of meiosis II (Olson et al., 2012). This timing is distinct from the timing of the formation of the trilaminar eggshell, which is established by anaphase of meiosis I. Historically, the permeability barrier was thought to be the innermost layer of the trilaminar eggshell and was thought to be lipid-rich. However, as stated earlier, studies by Olson et al. (2012) revealed that the permeability barrier was a layer separated from the trilaminar eggshell by the extra-embryonic matrix. Since most studies prior to 2009 did not observe this fifth inner layer in fixed samples, and since embryos stripped of the VL and CL were still impermeable (Schierenberg and Junkersdorf, 1992), it had been assumed that the third layer was the permeability barrier. Given the most recent studies (Olson et al., 2012), it now appears that the lipid-rich layer responsible for protecting the embryo apposes the embryo and is not part of the trilaminar outer eggshell.

Phenotypes formerly attributed to the third layer may need to be re-evaluated to determine whether it was really the CPG layer or the lipid-rich permeability barrier that was defective. Either way, it appears that the lipid-rich layer has consistently been assigned the function of permeability barrier. It is the positioning of this layer that has recently been more clearly demonstrated (Olson et al., 2012). Since the fifth layer had not been observed before in fixed embryos, it is difficult to know how many previous studies affected this layer. For example, how do defects in CG exocytosis affect the formation and structure of both the CPG layer and the lipid-rich permeability barrier? The observation that defects in CG exocytosis are correlated with osmotic integrity defects suggests that both the CPG layer and the permeability barrier must be compromised. Lacking in most of the earlier studies are EM images to determine the integrity of all the layers of the eggshell.

Evidence that the permeability barrier is a lipid-rich layer came from earlier studies in which the depletion of lipid biosynthetic enzymes caused the OID phenotype (Benenati et al., 2009; Rappleye et al., 2003; Tagawa et al., 2001) (Figure 7). pod-2 was recovered in a genetic screen for mutants with polarity and osmotic defects (Section 8) and encodes an acetyl-CoA carboxylase ortholog (Rappleye et al., 2003). Two other lipid regulators were discovered in the same class: emb-8, which encodes the NADPH cytochrome P450 reductase, an enzyme that catalyzes the modification of fatty acids, and fasn-1, a fatty acid synthase (Rappleye et al., 2003). It was demonstrated that the addition of specific fatty acids to the diet of pod-2 mutant animals could rescue the polarity phenotypes, but not the osmotic phenotype (Rappleye et al., 2003). However, embryos depleted of EMB-8 and POD-7/8 (cytochrome P450 family members involved in the hydroxylation of lipids) were completely rescued by a diet of C. elegans lipid extracts (Benenati et al., 2009). These experiments together suggest that different lipids may be responsible for the polarity defects versus the osmotic defects. Two other proteins recently identified, PERM-1 and DGTR-1, suggest that ascaroside glycolipids may be an essential component of the permeability barrier (Olson et al., 2012). Determining the specific lipid and protein components of this fifth layer will certainly shed light on how this layer protects the embryo.

It has been observed that an actin cap forms on the posterior cortex of the oocyte several minutes after its entry into the spermatheca (Maruyama et al., 2007). This F-actin cap is thought to mark the site of sperm entry. In embryos depleted of EGG-3 or EGG-4/5, this cap is not restricted to the posterior cortex and disperses aberrantly (Parry et al., 2009; Maruyama et al., 2007). Most other eggshell studies have not assayed for this actin cap and thus its role in the initiation of eggshell formation remains unclear. Currently this cap remains a useful marker of the sperm entry site. At the anterior pole of the embryo, actin is required for polar body extrusion (Velarde et al., 2007). Severe RNAi depletion of F-actin not only disrupts polar body extrusion, but also results in cytokinesis defects during the mitotic divisions (Velarde et al., 2007).

The Aroian laboratory conducted a screen to identify a Polarity and Osmotic Defective (pod) class of mutants whose embryos develop with osmotic sensitivity and polarity defects (Rappleye et al., 1999). POD-1 is a coronin-like protein, an actin-binding protein that localizes to the anterior cortex of one-cell embryos. Embryos lacking POD-1 fail to extrude polar bodies, swell in hypotonic buffer and shrink in hypertonic buffer, and are permeable to fluorescent dyes (Rappleye et al., 1999). These embryos also display polarity defects, however, the polarity defects are separable from the osmotic defects (Rappleye et al., 1999). EM analysis of the eggshells of wild-type and mutant embryos revealed that pod-1 mutants have a normal trilaminar outer eggshell but also possess an aberrant extra layer that surrounds the embryo proper (Figure 13). Olson et al. (2012) proposed that this extra layer was a second CPG layer that resulted from delayed CG exocytosis. How an actin-binding protein functions in the osmotic integrity of the embryo awaits further investigation.

The POD-1 studies discussed above were the first to suggest a connection between osmoregulation and the actin cytoskeleton. Defects in osmoregulation can certainly perturb the actin cytoskeleton and cytokinesis. However, the reverse has not been observed: studies of actin RNAi or inhibition have not described defects in the osmotic sensitivity or permeability of the early embryo.

Another link to the actin cytoskeleton and osmoregulation came from studies of the cyk-3 mutant. cyk-3 encodes a ubiquitin C-terminal hydrolase. Depletion of this gene results in oval, osmotically sensitive, dye-permeable embryos. In addition, polar bodies are not extruded, no membrane contractions are observed, and pseudocleavage and cytokinesis fail (Kaitna et al., 2002). Actin distribution in the embryonic cortex is also disrupted. The authors observed a partial rescue of the cytokinesis defects using osmotic support media. Vesicle trafficking in these mutants has not yet been investigated. Presumably there are defects as the dye permeability of the embryos suggests the improper formation of the permeability barrier. The authors hypothesize that the target of this hydrolase is a protein that controls osmolarity, such as an ion channel.

The block to polyspermy is a mechanism present in all sexually reproducing organisms to ensure that progeny have a euploid genome. There are typically two levels of block: a fast block at the membrane level that is based on membrane potential, and a slow block that generally involves the modification or generation of an extracellular matrix that acts as a physical barrier to sperm entry (Wong and Wessel, 2006). There is currently no evidence in C. elegans that a disruption of membrane potential in the fertilized embryo leads to a polyspermic phenotype. Furthermore, no mutants or RNAi depletion studies with ion channels have resulted in polyspermy. The only observations of polyspermy have come from RNAi and mutant studies on genes that contribute to the synthesis of the chitin layer of the eggshell. The synthesis of a chitinous eggshell is likely the slow block mechanism in C. elegans. Johnston et al. (2010) recently reported that CHS-1, GNA-2, and CBD-1 depletions lead to polyspermy. egg-4/5 double mutants are also polyspermic (Parry et al., 2009). Johnston et al. (2010) also assert that embryos depleted of SPE-11, EGG-1, EGG-2, and EGG-3 are fertilized by more than a single sperm. In contrast, Kadandale et al. (2005) had previously shown that depletion of EGG-1/2 by RNAi led to infertility (no gamete fusion). Currently there is no good explanation for the discrepancy between these reports. Interestingly, most polyspermic embryos have just two sperm pronuclei and the authors suggest that there must be a second, non-chitin-dependent block (Johnston et al., 2010).

In contrast to the polyspermy exhibited in the Johnston study (2010) and the egg-4/5 study (Parry et al., 2009), there is at least one RNAi condition in which oocytes appear to be refractive to fertilization (Allen et al., 2014; Burrows et al., 2006). The Wee1 kinases are responsible for the phosphorylation and inhibition of Cdk1 in eukaryotes. Of the two Wee1 kinase orthologs in C. elegans, WEE-1.3 is expressed in the germline (but not exclusively) and is required to delay maturation of the oocyte until it is in the -1 position in the oviduct (Allen et al., 2014; Burrows et al., 2006). When WEE-1.3 is depleted, oocytes precociously demonstrate hallmarks of maturation such as phospho-histone H3 antibody staining and nucleolar breakdown. Interestingly, these oocytes pass through a sperm-filled spermatheca yet do not get fertilized. Is this due to a premature block to polyspermy or a premature CG movement to the cortex (Bembenek et al., 2007)? And, if so, does this indicate that the polyspermy block is coupled to the cell cycle? Depletion of WEE-1.3 could precociously activate the cell cycle, resulting in the premature internalization of the EGG proteins and MBK-2, and thus prevent fertilization. Further studies of this rare phenotype will hopefully uncover some additional players in the block to polyspermy.

One suggestion of a mechanism for the block to polyspermy originates with what has been called a post-ovulatory envelope (David Greenstein, personal communication) or a fertilization membrane (Hall and Altun, 2008). In some TEM images, it appears that unfertilized oocytes and newly fertilized zygotes possess a thin crenellated covering on top of the vitelline layer (David Greenstein, personal communication) (Hall and Altun, 2008) (Figure 14). It is tempting to speculate that this envelope is briefly assembled on all oocytes that pass through the spermatheca (whether they are fertilized or not). This envelope must be added in the spermatheca or uterus. This so-called uterine layer has been observed for numerous nematode species (Foor, 1967). We thus propose, in most situations where sperm-oocyte fusion has occurred, the post-ovulatory envelope or fertilization membrane is disassembled following the synthesis of the eggshell. However, perhaps in situations where the oocytes are not fertilized (such as spe-9 mutants and feminizing mutants), the disassembly of this post-ovulatory envelope does not occur and remains to stably protect the unfertilized oocyte. This hypothesis is actually borne out by our conversations with David Greenstein, his observations of spe-8 oocytes, and the images presented by David Hall in the Worm Atlas (Hall and Altun, 2008). If a transient post-fertilization envelope does exist in C. elegans, it would help to explain a perplexing observation about the differential stability of unfertilized oocytes and fertilized 1-cell zygotes that have not progressed through their meiotic divisions. Unfertilized oocytes, whether from a spe mutant or from a hermaphrodite that has exhausted its sperm, appear round and stable on growth plates. Although they do not possess a chitinous eggshell, they remain intact during the process of going through two spermathecal valves and then the vulva to be deposited externally. Yet chs-1, gna-2, and spe-11 embryos are extremely fragile and often crushed by the time they are laid on the growth plate. Perhaps fertilization activates a process that removes this protective post-ovulatory envelope. But because chs-1, gna-2, and spe-11 embryos are defective in eggshell synthesis, these mutant embryos are subsequently crushed by the forces of egg laying. Thus it has been proposed that this post-ovulatory envelope is responsible for the structural integrity of unfertilized oocytes, and, in fertilized embryos, serves as a block to polyspermy while the chitin layer of the eggshell is being synthesized (Hall and Altun, 2008).

It is likely that there are several more important players in the process of eggshell formation in C. elegans. How will these genes be discovered? Many of the genes suspected of being involved in eggshell synthesis (directly or indirectly) were originally identified in genome-wide RNAi screens (Sonnichsen et al., 2005). These lists of genes with putative eggshell phenotypes should be examined and studied in greater detail. The key phenotypic classes include Osmotic Integrity Defective, or OID, and passage through meiosis defective (Sonnichsen et al., 2005). The OID class consisted of 109 genes and has since been expanded by additional screens. Depletion by RNAi of other genes that result in passage through meiosis defects also cause osmotic defects. It is now thought that these eggshell defects are an indirect result of post-fertilization meiotic defects (such as the APC mutants). All of these genes are listed in Table 1 (Section 12). In a recent screen to find genes that, when depleted by RNAi, cause early embryos to be more permeable for drug and small molecule uptake, Carvalho et al. (2011) identified a set of 310 candidate genes that, by bioinformatics criteria, might play a role in eggshell synthesis. Of these, 20 resulted in >75% eggshell permeability. These genes had already been identified as members of the OID class of genes, except for inx-9, which is included in Table 1.

The future for the study of the C. elegans eggshell remains bright. There are numerous questions that demand investigation. For example, what are the cortical granules components? Knowing the makeup of these granules will be very informative with regard to the fusion of the vesicles to the plasma membrane and the proteins that are delivered to the extracellular environment of the embryo. Likewise, identifying the components of the permeability barrier and the biosynthetic pathways responsible for this structure is key to our understanding of this barrier.

To more adequately follow the synthesis of the various eggshell layers, the field needs more markers for each of the layers. Besides our ability to detect the chitin layer and CPG-1/2, we have few available reagents to label the various eggshell layers. Missing in most studies of the eggshell are EM and immunoEM images of various mutants. Such studies would benefit from images at different stages of early embryonic development to observe the sequential synthesis of the eggshell and the modifications thereof. The observation that most OID embryos also show defects in polar body extrusion and cytokinesis warrants further investigations as to the connection between these phenotypes.

Another important question to address is whether there is a fast block to polyspermy and what role, if any, does the calcium wave play in this block? Few existing mutants or RNAi depletion experiments have resulted in polyspermy, and in those cases, none of the genes yet suggest a clear fast block mechanism. Perhaps a forward genetic screen aimed specifically at identifying polyspermic embryos would help answer this important question.

Is there a sperm component required for CHS-1 activation? Intriguingly, there is evidence from other organisms that chitin synthase needs to be proteolytically processed in order to become functional. In the context of C. elegans fertilization, it is tempting to speculate that this regulatory mechanism works via a protease carried by the sperm, which is introduced into the oocyte at fertilization. This mode of regulation would serve two functions: (1) the activating factor would not be introduced into the embryo until the time at which the eggshell was required, and (2) it would prevent CHS-1 from ectopically activating. A precociously assembled eggshell would be developmentally catastrophic because it would prevent fertilization.

In Ascaris, chitin synthase was found in two forms, an inactive form in unfertilized oocytes and an activated form converted by protease activity in fertilized embryos (Dubinsky et al., 1986). Identifying the protease that processes CHS-1 may help to identify other genes required for the process of eggshell formation and egg activation. The sperm may contain a protease necessary for chitin synthase activation, either by processing CHS-1 directly, by processing an inhibitor (which would lead to degradation), or by processing an activator (Merzendorfer, 2006). Evidence points to a serine or cysteine protease (Martinez-Rucobo et al., 2009). An analogous regulation is also found in Manduca sexta (insect) via a serine protease called CTLP1 (Broehan et al., 2007). The closest worm paralogs are try-1, svh-1, and try-3. Deletions exist for some of these genes but no reports of eggshell phenotypes have been made.

Are there additional paternal factors involved in eggshell formation? In order to identify such factors (i.e., the putative sperm protease), RNAi screens are not likely to be the most fruitful strategy since many sperm genes are refractory to RNAi-mediated knockdown. The analysis of the above-mentioned protease genes might prove valuable, as well as the genetic and biochemical pursuit of proteins that act in the SPE-11 pathway. With the recent successes of CRISPR/Cas9-mediated genome editing in C. elegans (e.g., Arribere et al., 2014; Paix et al., 2014), one can directly create mutant alleles of candidates that might have paternal roles in eggshell formation.

Of the >100 genes already identified in RNAi screens as yielding the OID phenotype (Table 1, Section 12) (http://www.wormbase.org/; Sonnichsen et al., 2005), what hidden details exist in this collection of genes that would further clarify how the eggshell is synthesized and modified during early embryogenesis? Surely some of these genes will provide insights into the biosynthetic pathway for the permeability barrier. Another source of eggshell synthesis candidates may come from screens where genes involved in eggshell synthesis might be scored as maternal sterile or 'ugly brown Mels' as well as OID (Green et al., 2011; A. Singson, personal communication). The depletion of CHS-1, for example, whether by RNAi or mutant analysis, results in such fragile embryos that the mothers may appear to be sterile. Thus, the examination of sterile mutants may uncover a few previously overlooked eggshell defective mutants.

Great strides have been made in the last 10 years to uncover the mechanism and composition of eggshell formation in C. elegans. The new ultrastructural information presented by Olson et al. (2012) should spur a reexamination of past data and a more focused investigation in the field going forward.

We would like to thank Sara Olson, Josh Bembenek, Andy Singson, Anna Allen, Abigail Fuchsman, Amy Fabritius, Aimee Jaramillo-Lambert, and two anonymous reviewers for their very helpful feedback on this chapter. Special thanks go to Sara Olson for sharing her list of OID genes and many of her images. Table 1 represents a compilation of genes from database and literature searches from her laboratory and ours. Special thanks also go to Miyuki Sato for remaking Figure 9 for this chapter. Thanks also to those who shared unpublished data with us: Chris Merritt and Geraldine Seydoux (Johns Hopkins University School of Medicine), Jun Takayama and Shuichi Onami (RIKEN Quantitative Biology Center, Kobe, Japan) and David Greenstein (University of Minnesota). This research was supported by the Intramural Research Program of the NIH, National Institute of Diabetes, Digestive and Kidney Diseases.