Calcium signaling in mammalian egg activation and embryo development: Influence of subcellular localization

Abstract

Calcium (Ca²⁺) signals drive the fundamental events surrounding fertilization and the activation of development in all species examined to date. Initial studies of Ca²⁺ signaling at fertilization in marine animals were tightly linked to new discoveries of bioluminescent proteins and their use as fluorescent Ca²⁺ sensors. Since that time, there has been rapid progress in our understanding of the key functions for Ca²⁺ in many cell types and the impact of cellular localization on Ca²⁺ signaling pathways. In this review, which focuses on mammalian egg activation, we consider how Ca²⁺ is regulated and stored at different stages of oocyte development and examine the functions of molecules that serve as both regulators of Ca²⁺ release and effectors of Ca²⁺ signals. We then summarize studies exploring how Ca²⁺ directs downstream effectors mediating both egg activation and later signaling events required for successful preimplantation embryo development. Throughout this review, we focus attention on how localization of Ca²⁺ signals influences downstream signaling events, and attempt to highlight gaps in our knowledge that are ripe areas for future research.

Introduction

Calcium (Ca²⁺) is a ubiquitous second messenger that regulates secretion, contraction, metabolic enzymes, gene expression, and many other cell functions. Ca²⁺ has critical roles in fertilization in all animal species studied to date, from marine invertebrates to mammals. In all of these organisms, a rise in intracellular Ca²⁺ is responsible for releasing unfertilized eggs from arrest in meiosis and triggering the embryonic developmental program; this process is known as “egg activation”.

Our current knowledge regarding Ca²⁺ signaling during fertilization and embryo development in mammals was made possible by key observations and methodological developments largely generated by early studies of marine animals. Experiments in sea urchins in the 1930’s led to the first suggestion that Ca²⁺ was important at fertilization (Heilbrunn and Young 1930; Mazia 1937). In 1962, Shimomura et al. purified the bioluminescent, blue light-emitting protein aequorin from the jellyfish Aequorea victoria and reported that it could be utilized as an indicator of Ca²⁺ concentrations in biological fluids (Shimomura et al. 1962; Shimomura et al. 1963). Microinjection of purified aequorin into medaka eggs resulted in the first documentation of a large increase in intracellular Ca²⁺ levels in response to a fertilizing sperm (Ridgway et al. 1977). Aequorin was eventually used to demonstrate that in mouse and hamster eggs, unlike the single large increase in Ca²⁺ observed in most marine animals, multiple oscillations in intracellular Ca²⁺ occur and persist for several hours after fertilization (Cuthbertson and Cobbold 1985; Cuthbertson et al. 1981; Miyazaki et al. 1986). This finding was later shown to be true of all types of mammalian eggs.

In the early 1980’s, a series of optical indicators for Ca²⁺ were developed based on the Ca²⁺ chelator, 1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid, better known as BAPTA (Tsien 1983). The addition of acetoxymethyl ester groups to these tetracarboxylate indicators allowed cellular uptake without microinjection; cleavage of the ester group by intracellular esterases resulted in trapping of the indicator in the cytoplasm. The indicators had strongly enhanced ultraviolet light-induced fluorescence in response to Ca²⁺ binding, and allowed detection of sub-micromolar changes in cytoplasmic Ca²⁺ levels. The later development of fluorescent Ca²⁺ indicators with excitation and emission spectra in the visible range expanded their use dramatically (Minta et al. 1989). Finally, ratiometric Ca²⁺ indicators such as fura-2 were developed whose peak excitation wavelength altered in response to Ca²⁺ binding, allowing calculations of increases in Ca²⁺ over baseline levels (Grynkiewicz et al. 1985). Ratiometric Ca²⁺ indicators are currently used widely to study Ca²⁺ signaling at fertilization in mammals.

In the 1970’s, a second bioluminescent protein, green fluorescent protein, was purified from several Aequora species, and the protein was eventually cloned (Chalfie 1995; Prasher et al. 1992; Prendergast and Mann 1978). Expression of green fluorescent protein in prokaryotic cells was reported in 1994, beginning the modern era of using exogenously expressed fluorescent proteins as markers of gene expression and as genetically encoded sensors of Ca²⁺ and other small molecules. Together, these studies set the stage for rapid progress in our understanding of the key functions for Ca²⁺ during mammalian egg activation and embryo development, and the impact of cellular localization on these signaling pathways.

In this review, which focuses on mammals, we first consider in detail how Ca²⁺ itself is regulated and stored at different stages of oocyte development and how Ca²⁺ is released during egg activation. We then summarize the available literature regarding how Ca²⁺ directs the numerous downstream events of egg activation required for successful preimplantation embryo development. We focus attention on how the localization of various Ca²⁺ pools influences downstream signaling events. Throughout this review, we will attempt to highlight gaps in our knowledge that are ripe areas for future research.

Regulation of calcium stores

Both the endoplasmic reticulum (ER) and mitochondria regulate intracellular signaling through the storage and release of Ca²⁺. The ER is the major Ca²⁺ storage organelle in eggs and it plays a critical role in the initiation of Ca²⁺ release at fertilization. The ER contains Ca²⁺ pumps, intraluminal Ca²⁺ storage proteins, and specific Ca²⁺-releasing channels, which together regulate Ca²⁺ signaling while maintaining cellular Ca²⁺ homeostasis. Mitochondria mainly serve as energy-supplying organelles that in oocytes provide ATP for fertilization and preimplantation embryo development (Torner et al. 2004). However, mitochondria also regulate intracellular Ca²⁺ homeostasis by serving as a “buffer” for high cytoplasmic Ca²⁺ levels. Ca²⁺ enters mitochondria when the Ca²⁺ concentration in the cytoplasm is over a certain threshold, and once the cytoplasmic Ca²⁺ has returned to its resting level, Ca²⁺ is pumped back into cytoplasm by the Na⁺/Ca²⁺ exchanger (Bootman et al. 2001). In addition to the ER and mitochondria, there is evidence that Ca²⁺ can be stored and released by actin filaments in response to both extracellular and intracellular signals (Lange and Gartzke 2006).

Intracellular Ca²⁺ stores are regulated in part by “store operated Ca²⁺ entry” (SOCE), a mechanism to replenish intracellular Ca²⁺ from the extracellular environment (Putney 1986). Several molecules involved in mediating SOCE are now identified. Depletion of ER Ca²⁺ stores is recognized by one of two STIM proteins, STIM1 and STIM2, which are single transmembrane ER proteins that bind Ca²⁺ through an EF-hand motif located in the ER lumen (Cai 2007; Hewavitharana et al. 2007; Roos et al. 2005; Williams et al. 2001). STIM2 is activated by small decreases in ER luminal Ca²⁺, suggesting it functions to stabilize basal cytosolic and ER Ca²⁺ levels under non-stimulated conditions (Brandman et al. 2007). In contrast, STIM1 is only activated by larger decreases in ER luminal Ca²⁺ that occur during triggered Ca²⁺ signaling events, at least in part because of its higher Ca²⁺affinity. In somatic cells, significant lowering of ER Ca²⁺ levels causes oligomerization and redistribution of STIM1 into punctate ER structures near the plasma membrane (Liou et al. 2005; Smyth et al. 2008). These activated STIM1 oligomers signal to ORAI proteins, which are highly selective plasma membrane Ca²⁺ channel subunits, stimulating transport of extracellular Ca²⁺ into the cytosol (Feske et al. 2006; Vig et al. 2006; Zhang et al. 2005). Sarcoplasmic/endoplasmic reticulum Ca²⁺ ATPase (SERCA) pumps can then transport Ca²⁺ back into the ER, replenishing cellular stores. Mammalian oocytes express both STIM1 and ORAI1 (Gomez-Fernandez et al. 2009; Koh et al. 2009). Mouse oocytes also express STIM2, which co-localizes with the ER marker calreticulin (Fig. 1). STIM2 distribution changes during oocyte maturation, when it becomes strongly associated with the region surrounding the meiotic spindles and is lower in the vegetal poles. These findings indicate that similar to somatic cells, both STIM1 and STIM2 likely function to regulate ER Ca²⁺ stores in mouse and other mammalian oocytes.

Figure 1. Localization of STIM2 in maturing mouse oocytes.

Mouse oocytes at the germinal vesicle (GV), germinal vesicle breakdown (GVBD), metaphase I (MI), and metaphase II (MII) stages were fixed and stained for both STIM2 (red) and calreticulin, an endoplasmic reticulum (ER; green) marker. Antibodies used were anti-STIM2 (Cat# 4125; ProSci Inc., Poway, CA) and anti-calreticulin (cat# sc-7431; Santa Cruz Biotechnology, Santa Cruz, CA). Spindle region indicated on merged images with arrows.

It is not clear whether SOCE plays an important role in replenishing Ca²⁺ stores in mature mammalian oocytes during fertilization. In somatic cells, SOCE is suppressed during mitotic stages of the cell cycle by phosphorylation of STIM1 (Smyth et al. 2009). In maturing Xenopus oocytes, SOCE becomes uncoupled from Ca²⁺ depletion around the time of germinal vesicle (GV) breakdown (Machaca and Haun 2000; Machaca and Haun 2002). In addition, ORAI1 becomes internalized and STIM1 oligomerization does not occur, indicating that SOCE is essentially inactive during meiosis in Xenopus (Yu et al. 2009a). It is well documented that SOCE can be induced experimentally in mature mammalian oocytes (Machaty et al. 2002; Martin-Romero et al. 2008; McGuinness et al. 1996; Miao et al. 2012), but there are conflicting reports regarding whether SOCE is required for egg activation. During mouse fertilization, inhibiting SOCE using either gadolinium or the specific SOCE inhibitor Synta66 (Ng et al. 2008) does not prevent replenishment of intracellular Ca²⁺ stores after fertilization, indicating that Ca²⁺ influx occurs through alternative channels (Miao et al. 2012). In pig oocytes in which siRNA was used to knock down STIM1 protein levels during oocyte maturation, fertilization mainly induced only a single Ca²⁺ transient and no subsequent oscillations occurred (Lee et al. 2012). Of note, the amplitude of the first Ca²⁺ transient in the control eggs shown in this study was much higher than that in the STIM1-depleted eggs, indicating that the Ca²⁺ stores had been depleted before fertilization occurred. Although it is certainly possible that there are species differences in a requirement for SOCE at fertilization, our interpretation of these apparently conflicting results is that porcine oocytes require STIM1 to maintain ER Ca²⁺ stores during maturation; consequently, STIM1-depleted oocytes have diminished Ca²⁺ stores leading to the single low amplitude sperm-induced Ca²⁺ transient and failure of the sperm-induced egg activation cascade. In support of this interpretation, the same authors demonstrated that the amount of Ca²⁺ mobilized in response to ionomycin treatment is far less in STIM1-depleted pig oocytes than in controls (Lee et al. 2012). Further experiments in these and other mammalian systems may shed light on whether these results represent species differences or whether alternative mechanisms are required to promote Ca²⁺ influx during fertilization of mammalian oocytes.

Alterations in calcium release mechanisms during oocyte maturation

In most mammals, GV oocytes arrested in the diplotene stage of meiosis I are stimulated to resume meiosis in response to the LH surge. The oocyte subsequently undergoes GV breakdown, progresses through the remainder of meiosis I, extrudes the first polar body, and finally becomes arrested again at metaphase of meiosis II (MII). This process is referred to as “oocyte maturation”, and it includes a complex sequence of nuclear and cytoplasmic events that prepare the egg for fertilization and initiation of embryo development.

During maturation to MII, the ER undergoes a dramatic microtubule- and microfilament-dependent structural reorganization (FitzHarris et al. 2007; Mann et al. 2010; Mehlmann et al. 1995). In GV stage oocytes, the ER is continuous with the nuclear envelope and has a fine reticular structure throughout the cytoplasm, with few small accumulations in the cortex (Mehlmann et al. 1995). At GVBD, the ER begins to restructure into a dense reticular network intimately connected with the MI spindle as it migrates toward the oocyte cortex (FitzHarris et al. 2007). At MII, the ER is no longer tightly associated with the spindle. Instead, it is organized in a reticular network of large diameter (~2–3 Om) sub-plasma membrane accumulations mainly in regions away from the MII spindle (Mehlmann et al. 1995).

Concomitant with the changes in organization of Ca²⁺ storage organelles, the oocyte’s ability to release Ca²⁺ in response to various triggers changes with maturation. The Ca²⁺ transient induced by the Ca²⁺ ionophore, ionomycin, in MII eggs is approximately 3-fold larger than that induced in GV oocytes, and the greatest change occurs between MI and MII (Jones et al. 1995; Mehlmann and Kline 1994). Similarly, MII eggs release more Ca²⁺ than GV oocytes in response to either microinjected inositol 1,4,5-trisphosphate (IP₃) or fertilization by sperm (Fujiwara et al. 1993; Mehlmann and Kline 1994). One explanation for these findings would be that there are larger releasable Ca²⁺ stores in MII eggs than in GV oocytes. However, mouse GV oocytes treated with thimerosal, a chemical that sensitizes the IP₃-receptor so that more Ca²⁺ is released in response to the same amount of IP₃, release just as much Ca²⁺ as MII eggs, and more than that induced by treating GV oocytes with ionomycin (Mehlmann and Kline 1994). In oocytes of other species, such as the hamster, IP₃ sensitivity changes less dramatically over the course of maturation, with GV oocytes containing at least 80% of the IP₃-sensitive Ca²⁺ stores present in MII eggs. (Fujiwara et al. 1993). These findings indicate that at least in some species the Ca²⁺ stores in GV oocytes and MII eggs are similar, and that GV oocytes are much less sensitive to Ca²⁺ release mechanisms. The increase in sensitivity of MII eggs to IP₃-mediated Ca²⁺ release is likely caused in part by recruitment of IP₃-receptor type 1 (IP₃R1) mRNA for translation during maturation, resulting in increased protein levels in MII eggs (He et al. 1997; Mehlmann et al. 1996; Xu et al. 1994). In addition, changes in phosphorylation of IP₃R1 likely increase the efficiency of Ca²⁺ release in MII eggs (Wakai et al. 2012).

Egg activation overview

The transition of an unfertilized egg, which is arrested in metaphase of meiosis II, to a developing embryo, requires completion of a sequence of events known as egg activation. These events include cortical granule (CG) exocytosis, modifications of the zona pellucida and plasma membrane that prevent polyspermy, completion of meiosis, recruitment of maternal mRNAs into polysomes for translation, and formation of male and female pronuclei (Schultz and Kopf 1995). In all animal species studied to date, alterations in intracellular Ca²⁺ initiate egg activation.

In mammals, Ca²⁺ release occurs after sperm-egg plasma membrane fusion in response to release of sperm head-associated phospholipase C zeta (PLCζ) into the egg’s cytoplasm (Kashir et al. 2012; Knott et al. 2005; Saunders et al. 2002). PLCζ catalyzes breakdown of phosphatidylinositol 4,5-bisphosphate (PIP₂) to inositol 1,4,5-trisphosphate (IP₃) and diacylglycerol (DAG); endogenous egg PLCβ isoforms likely also contribute to IP₃ generation at fertilization (Igarashi et al. 2007). IP₃ subsequently binds to IP₃R1 in endoplasmic reticulum (ER) membranes, triggering Ca²⁺ release from IP₃-sensitive ER stores (Miyazaki et al. 1992; Xu et al. 1994). This initial Ca²⁺ release leads to the generation of low frequency oscillations in intracellular Ca²⁺ that result from the sequential release of Ca²⁺ from ER stores followed by reuptake of Ca²⁺ into the ER via SERCA pumps and extrusion of Ca²⁺ from the egg via plasma membrane Ca²⁺ ATPase pumps. These Ca²⁺ oscillations continue for several hours, and are dependent on replenishment of Ca²⁺ stores by influx of extracellular Ca²⁺ (Igusa and Miyazaki 1983; Kline and Kline 1992). The characteristic shapes of the first two Ca²⁺ transients are shown in Figure 2, and a schematic diagram showing how Ca²⁺ fluxes are initiated at fertilization is shown in Figure 3. In the following sections, we will review in detail the contributions of different signaling molecules to generating the Ca²⁺ oscillations.

Figure 2. Characteristic appearance of first and second sperm-induced Ca²⁺ transients.

Arrow indicates “shoulder” in first Ca²⁺ transient.

Figure 3. Ca²⁺ release and reuptake pathways during initial egg activation.

Schematic diagram of Ca²⁺ release from ER stores in response to PLCζ-mediated generation of IP₃ and the channels responsible for Ca²⁺ extrusion and reuptake. 1) PLCζ hydrolyzes PIP₂ in cortical pool to generate IP₃ and DAG. 2) IP₃ mediates Ca²⁺ release from ER stores via IP₃R1. 3) Increased cytosolic Ca²⁺ and possibly other signaling molecules activate endogenous egg PLC(s), causing generation of additional IP₃. 4) Ca²⁺ re-enters ER stores via SERCA pumps. 5) Cytoplasmic Ca²⁺ is extruded from egg by PMCA pumps. 6) DAG and Ca²⁺ promote plasma membrane translocation of cPKC. 7) Phosphorylation of cPKC-sensitive Ca²⁺ channels promotes Ca²⁺ influx. PB I, first polar body; ER, endoplasmic reticulum; PLCζ, phospholipase C zeta; PIP₂, phosphatidylinositol 4,5-bisphosphate; IP₃, inositol 1,4,5-trisphosphate; DAG, diacylglycerol; ePLC, egg phospholipase C; cPKC, conventional protein kinase C; IP₃R1, type 1 IP₃ receptor; Ca²⁺, calcium; SERCA, sarcoplasmic/endoplasmic reticulum Ca²⁺ ATPase pump; PMCA, plasma membrane Ca²⁺ ATPase pump; ?, cPKC-sensitive plasma membrane Ca²⁺ channel(s), identity unknown.

Phosphoinositide-specific phospholipase C

The 13 mammalian phosphoinositide PLCs are grouped into six sub-families, β, γ, δ, ε, η, and ζ, on the basis of sequence homology and the location of structural domains (Fig. 4)(Rhee and Bae 1997; Suh et al. 2008). Although these PLC isoforms all catalyze the hydrolysis of PIP₂ to IP₃ and DAG, the signals that activate them differ, and they are expressed in a tissue and cell-specific fashion that controls their contributions to modulating cellular responses. Several different PLC isoforms are found in the fertilized egg and likely contribute to IP₃-receptor-mediated Ca²⁺ fluxes. As mentioned above, the sperm delivers PLCζ to the egg at the time of sperm-egg fusion. In addition, mouse oocytes contain mRNAs encoding at least 8 of the 13 PLC isoforms, including members of all sub-families except for PLCζ (Fig. 4). Based on information deposited in GEO Profiles (http://www.ncbi.nlm.nih.gov/geoprofiles), human and cow oocytes appear to contain the same PLC isoforms as mouse oocytes, but there is much less information available regarding the relative quantities of the various isoforms.

Figure 4. Phospholipase C isoforms.

Schematic diagram of functional domains of the six mammalian PLC sub-families. The size ranges of the proteins and their splice forms within each PLC sub-family and known regulatory mechanisms are indicated. Based on 3 distinct microarray datasets deposited in GEO Profiles, mammalian oocytes contain the listed PLC isoforms. Isoforms listed in larger, bold font have the highest signal intensities; isoforms in smaller, plain font have lower signal intensities, suggesting relative abundance of the mRNAs. PLC isoforms with an average signal intensity in fully grown oocytes of < 50 across the three datasets are not listed. RasGEF, guanine nucleotide exchange factor for Ras-like small GTPases domain; PH, pleckstrin homology domain; EF, EF-hand-like domain; X and Y, catalytic domains; C2, C2-domain Ca²⁺-binding pocket; RA, Ras-association domain; SH2, Src-homology 2 domain; SH3, Src-homology 3 domain; PIP₃, phosphatidylinositol 3,4,5-trisphosphate; PA, phosphatidic acid; AA, arachidonic acid; Rac, Rac GTPases.

PLCζ contains the conserved Ca²⁺-binding EF hand domain, catalytic domains, and C2 domains found in all PLCs (Fig. 4). However, it lacks the phosphatidylinositol-binding pleckstrin homology domain, which can mediate plasma membrane localization of other PLCs via interactions with phospholipids [reviewed in (Ito et al. 2011)]. Instead, PLCζ localizes diffusely throughout the cytoplasm when exogenously expressed in mouse eggs, suggesting that the sperm PLCζ that enters the egg at fertilization is also freely diffusible (Yu et al. 2012). Indeed, there is evidence that PLCζ targets a cortical but non-plasma membrane pool of PIP₂ for hydrolysis at fertilization (Yu et al. 2012). A second unusual feature of PLCζ is its extremely high Ca²⁺ sensitivity, with 70% maximal activity at Ca²⁺ concentrations of 100 nM (Kouchi et al. 2004). This high Ca²⁺ sensitivity suggests that simple exposure of sperm-derived PLCζ to basal cytoplasmic egg Ca²⁺ levels could stimulate PLCζ to produce an amount of IP₃ sufficient to trigger the initial release of Ca²⁺ from ER stores.

PLCβ isoforms are typically activated by G-protein-dependent pathways linked to transmembrane receptors (Rhee and Bae 1997). A unique feature of PLCβ isoforms is their elongated C-terminal region that mediates interactions with Gq-family α subunits, whereas the PH domain appears to mediate Gβγ subunit interactions (Suh et al. 2008). Gαq does not appear to activate egg PLCβ isoforms at fertilization because a Gαq-specific function-blocking antibody does not prevent sperm-induced egg activation (Williams et al. 1998). However, blocking activity of Gβγ subunits by microinjecting eggs with phosducin partially inhibits pronucleus formation after fertilization (Moore et al. 1994), suggesting that Gβγ-mediated PLCβ activation could facilitate IP₃ generation. In addition, PLCβ isoform activity is enhanced by rising Ca²⁺ levels (Taylor et al. 1991) such as would be expected in the fertilized egg downstream of PLCζ activity. This property of PLCβ suggests that it could promote additional IP₃ production and subsequent IP₃R-mediated Ca²⁺ release, contributing to the large increase in cytoplasmic Ca²⁺ observed in the fertilized egg. Indeed, there is evidence in mouse oocytes that reduction of PLCβ1 protein results in alterations of Ca²⁺ oscillatory activity observed at fertilization, suggesting that PLCβ1 normally functions in collaboration with PLCζ to generate the IP₃ required for the observed Ca²⁺ oscillatory pattern (Igarashi et al. 2007).

The PLCγ isoforms participate in mitogenic responses via interactions with receptor- and non-receptor (cytosolic) tyrosine kinases, and can also be activated by lipid-derived second messengers such as those generated by phospholipase D or phospholipase A2 (PLA2) (Rhee and Bae 1997). The PLCγ SH2 domains mediate association with receptor tyrosine kinases, which phosphorylate and activate PLCγ. Plasma membrane targeting of PLCγ may also be mediated by interactions between the PH domain and PIP₃ generated by phosphoinositide 3-kinase activity. Both PLCγ1 and PLCγ2 are present in mouse eggs, and microinjecting recombinant PLCγ1 into mouse eggs can induce low frequency Ca²⁺ oscillations similar to those induced by sperm (Mehlmann et al. 2001). However, inhibition of phosphoinositide 3-kinases, Src-family kinases, or SH2 domain-mediated PLCγ activation does not affect sperm-induced Ca²⁺ release, indicating that PLCγ activation by these mechanisms is not required at fertilization (Kurokawa et al. 2004; Mehlmann et al. 1998; Mehlmann et al. 2001; Mehlmann and Jaffe 2005). The possibility remains; however, that egg PLCγ could contribute to IP₃ generation downstream of either cytosolic tyrosine kinase activity or lipid second messengers generated at fertilization. Indeed, there is evidence in the mouse that fertilization causes activation of Src-family tyrosine kinases in localized regions of the egg (McGinnis et al. 2007; McGinnis et al. 2011). In addition, a unique cytosolic PLA2 isoform, cPLA2γ, is highly abundant in the mouse egg (Vitale et al. 2005), and could generate lipid-signaling molecules such as arachidonic acid that activate PLCγ.

Like PLCζ, both PLCη isoforms are very sensitive to activation by Ca²⁺ (Stewart et al. 2007), so endogenous egg PLCη, if present, would likely contribute to IP₃ generation downstream of the initial PLCζ-mediated Ca²⁺ increase. Although PLCδ isoforms are about 10-fold less sensitive to Ca²⁺ than PLCη isoforms (Stewart et al. 2007), a contribution of these PLCs to fertilization-induced Ca²⁺ release is also possible. PLCε is activated by small GTPases, Gα12/13, and Gβγ subunits. PLCε could in theory promote IP₃ generation at fertilization, but expression in mouse eggs of a constitutively active form of RAS, which should activate PLCε, does not induce Ca²⁺ release or other egg activation events (Moore et al. 1994), making this an unlikely scenario.

It is now well established that PLCζ activity initiates Ca²⁺ release at fertilization, even though only a very small quantity of the protein is carried into the egg by the single fertilizing sperm (Saunders et al. 2002). However, the variety of endogenous PLC isoforms expressed in mammalian eggs, combined with the initial generation of several potential activators of these PLC isoforms, indicates that endogenous egg PLCs likely contribute significant amounts of enzymatic activity toward PIP₂ hydrolysis at fertilization (Fig. 3). The identity of endogenous egg PLCs and their relative contributions to controlling Ca²⁺ oscillations at fertilization are ripe areas for future research.

Protein kinase C (PKC)-mediated regulation of calcium

Numerous PKC isoforms are expressed in mammalian eggs, including members of the conventional, novel, and atypical PKC subfamilies (Kalive et al. 2010). Conventional PKCs (cPKCs), which include PKCα, β, and γ isoforms, undergo Ca²⁺-dependent activation by DAG and anionic phospholipids, resulting in plasma membrane translocation and phosphorylation of substrate proteins on serine or threonine residues. In somatic cells, cPKC movement between the plasma membrane and cytoplasm occurs in synchrony with alterations in intracellular Ca²⁺ (Oancea and Meyer 1998). In mouse eggs, cPKCs translocate to the plasma membrane in response to elevations in intracellular Ca²⁺ via interactions of Ca²⁺ with the regulatory C2 domain (Halet et al. 2004). DAG generated during PIP₂ hydrolysis by PLC also binds to the regulatory C1 domains, and likely increases the time of cPKC membrane association. As in somatic cells, when egg Ca²⁺ levels return to baseline, the cPKCs dissociate and return to the cytosol until the subsequent Ca²⁺ rise.

Modulation of PKC activity in the egg at fertilization by either over-expression of PKC isoforms or use of cell-permeant PKC activators significantly increases the duration and frequency of Ca²⁺ oscillations, and also promotes Ca²⁺ influx (Halet et al. 2004; Yu et al. 2008). Furthermore, inhibition of PKC activity reduces the persistence of Ca²⁺ oscillations; this effect is partially mitigated by increasing the extracellular Ca²⁺ concentration. These findings suggest that PKC activity regulates Ca²⁺ influx at fertilization by phosphorylating substrates at or near the plasma membrane (Fig. 3).

The nature of these cPKC substrates is not known. One possibility is that cPKC phosphorylates one or more cellular components involved in SOCE, e.g., STIM1 or ORAI1, promoting additional Ca²⁺ uptake via this route. This seems unlikely because in somatic cells, both STIM1 phosphorylation on serine residues and PKC-mediated phosphorylation of ORAI1 are associated with suppression of SOCE (Kawasaki et al. 2010; Smyth et al. 2009). In Xenopus oocytes, SOCE is inactive during meiosis as a result of internalization of ORAI1 into an intracellular vesicular compartment (Yu et al. 2009a). Furthermore, in mouse eggs, inhibiting SOCE has no effect on sperm-induced Ca²⁺ oscillations, suggesting that the majority of Ca²⁺ influx at fertilization occurs through non-SOCE-associated Ca²⁺ channels (Miao et al. 2012). There is precedent for this idea because somatic cell ion channels including transient receptor potential-family Ca²⁺-permeable channels can be activated by PKC or DAG (Dai et al. 2009; Large et al. 2009), but the molecular basis for Ca²⁺ influx across the egg plasma membrane at fertilization remains to be determined.

Impact of calcium oscillation parameters on egg activation

Based on studies of mouse eggs fertilized in vitro, the first Ca²⁺ transient occurs within a few minutes of sperm–egg fusion after a prolonged (~1 min) small rise in cytoplasmic Ca²⁺ (Deguchi et al. 2000). The characteristic shape of the first Ca²⁺ transient is an extremely rapid rise for 2–3 seconds, followed by an inflection point representing a very short decrease in the rate of Ca²⁺ rise, referred to as a “shoulder”, and then a resumption of a the very rapid rate of Ca²⁺ rise to a peak of 1–3 OM (Fig. 2). The Ca²⁺ level drops but remains significantly elevated above baseline for up to several minutes; there are often additional small Ca²⁺ spikes superimposed on the early period of the first transient. The first transient is completed with a rapid drop in Ca²⁺ to baseline. This multi-component first Ca²⁺ transient pattern is consistent with the idea that several distinct mechanisms contribute to Ca²⁺ release, including IP₃ generation by sperm PLCζ and multiple egg PLC isoforms. In addition, Ca²⁺-induced Ca²⁺ release likely contributes. Succeeding Ca²⁺ transients are typically of lower amplitude, shorter duration, either lack or have a less distinct shoulder and superimposed spike pattern, and slow in frequency over time. The Ca²⁺ oscillation frequency is highly variable in different species and even between different eggs from the same individual, ranging from as high as every ~3 min in the hamster to ~50 min in the cow (Jones 1998). The oscillations generally persist for several hours, ending around the time of pronucleus formation.

Spatiotemporal analysis of the first Ca²⁺ transient in the mouse indicates that it is initiated near the site of sperm-egg fusion and subsequently a wave of increased Ca²⁺ is propagated toward the opposite side of the egg (Deguchi et al. 2000). Subsequent Ca²⁺ transients are initiated from regions in the egg cortex not restricted to the sperm entry site, likely from cortical regions enriched in ER accumulations (Kline et al. 1999). Cytoplasmic movements reminiscent of those that occur after fertilization in numerous lower species appear to coincide with these Ca²⁺ waves (Deguchi et al. 2000). These cytoplasmic movements depend on actomyosin-based cytoskeletal alterations, and the extent of these movements has been proposed to correlate with developmental potential (Ajduk et al. 2011). This correlation would not be surprising because there is a clear correlation between the total Ca²⁺ signal experienced by the egg after fertilization and the success of both pre- and post-implantation embryo development (Ozil et al. 2006; Ozil et al. 2005). The importance of the Ca²⁺ oscillatory pattern for subsequent development has raised concerns regarding the use of intracytoplasmic sperm injection (ICSI) for clinical assisted reproduction. ICSI results in Ca²⁺ oscillations similar to those induced by standard in vitro fertilization, but spatiotemporal aspects of the signaling pattern differ somewhat (Sato et al. 1999).

Not surprisingly, qualities of the egg also strongly influence the Ca²⁺ oscillatory pattern. Eggs fertilized later than normal after ovulation, referred to as “post-ovulatory aged eggs” or “aged eggs”, have altered Ca²⁺ oscillatory patterns including higher frequency but lower amplitude Ca²⁺ transients that have slower rates of rise and fall (Hao et al. 2009; Igarashi et al. 1997; Jones and Whittingham 1996). These alterations appear to be due in part to decreases in egg ER Ca²⁺ stores with time after ovulation (Jones and Whittingham 1996; Takahashi et al. 2009), which is a finding consistent with a non-existent or poorly functioning SOCE system during meiotic stages of the cell cycle. Aged eggs have lower developmental potential than recently ovulated eggs, likely as a result of these perturbations in Ca²⁺ signaling.

Early responses to calcium signals at fertilization

Mitochondrial energy production

One of the earliest effects of sperm-induced Ca²⁺ release is stimulation of mitochondrial energy production (Campbell and Swann 2006; Dumollard et al. 2004). In somatic cells, mitochondrial function is regulated by Ca²⁺ uptake into the mitochondrial matrix through the high capacity Ca²⁺ uniporter. The increase in Ca²⁺ levels causes activation of several Ca²⁺-sensitive mitochondrial enzymes and subsequently increased NADH/NAD ratios and ATP production (Ashby and Tepikin 2001). In addition, Ca²⁺-dependent mitochondrial substrate carriers can be activated directly by increases in cytosolic Ca²⁺(Satrustegui et al. 2007). Based on work in sea urchin and ascidian eggs, Ca²⁺ uptake into mitochondria most likely also occurs at fertilization in mammalian eggs (Dumollard et al. 2006). In mouse eggs, coincident with each cytosolic Ca²⁺ transient there is a measurable increase in NADH and decrease in FAD²⁺, both of which occur during activation of the citric acid cycle (Dumollard et al. 2004). Furthermore, the amount of ATP in the egg begins to increase at the time of the first Ca²⁺ transient, and continues to increase during the entire period encompassed by continuing Ca²⁺ oscillations before returning to baseline with their cessation (Campbell and Swann 2006).

In addition to providing energy for egg activation and embryo development, mitochondria participate in intracellular Ca²⁺ homeostasis. Because of their active Ca²⁺ uptake and efflux mechanisms, mitochondria serve to buffer cytosolic Ca²⁺ levels while at the same time prolonging the time of cytosolic Ca²⁺ elevation. Mitochondria located in close proximity to the ER can also facilitate refilling of ER stores after IP₃-stimulated Ca²⁺ release (Malli et al. 2005). In mouse eggs, chemical inhibitors of mitochondrial function disrupt the sperm-induced Ca²⁺ oscillatory pattern, and cause a sustained increase in cytoplasmic Ca²⁺ (Liu et al. 2001). These findings support a model in which mitochondrial Ca²⁺ sequestration normally contributes to returning Ca²⁺ levels to baseline between the Ca²⁺ transients. Contributions of mitochondria to regulating spatiotemporal aspects of Ca²⁺ oscillatory behavior and the reciprocal, i.e., contributions of Ca²⁺ fluxes to regulating mitochondrial function, could explain why eggs that do not undergo appropriate maturation-associated mitochondrial redistribution or have abnormally functioning mitochondria have low developmental competence (Au et al. 2005; Fernandes et al. 2012; Hao et al. 2009).

Cortical granule exocytosis/zona pellucida block to polyspermy

Cortical granules (CGs) are membrane-bound vesicles located in the cortex of the mature egg, but restricted from the actin cap region overlying the metaphase II spindle (Deng et al. 2003). CG exocytosis can be observed within minutes of the first Ca²⁺ transient after fertilization, and the Ca²⁺ increase is both necessary and sufficient for this event (Abbott and Ducibella 2001). Additional Ca²⁺ oscillations continue to drive waves of CG exocytosis, so that the number of CGs remaining in the egg continues to decrease with time after fertilization, and most of the CGs are gone in pronuclear stage embryos (Ducibella et al. 2002). The Ca²⁺-dependent downstream mediators that transduce the Ca²⁺ signal to cause CG exocytosis in mammalian eggs have not yet been identified. By analogy to somatic cells, these could include calmodulin, calcium/calmodulin-dependent protein kinase II (CaMKII), PKC isoforms, and actin depolymerizing enzymes, among others.

Exposure of the zona pellucida to CG contents extruded into the perivitelline space results in biochemical modifications that comprise the zona pellucida block to polyspermy. In mammalian eggs only a single protein, the metalloendoprotease ovastacin, has been demonstrated to be a CG component (Burkart et al. 2012). Ovastacin cleaves the zona pellucida glycoprotein ZP2, which prevents additional sperm binding (Burkart et al. 2012). Thus, Ca²⁺ signaling upstream of CG exocytosis is a critical regulator of the zona pellucida block to polyspermy.

Calcium/calmodulin-dependent protein kinase II (CaMKII)/cell cycle resumption

CaMKII is a ubiquitous serine-threonine protein kinase that is activated by Ca²⁺ and calmodulin to phosphorylate diverse substrates. CaMKII oligomers undergo autophosphorylation after Ca²⁺/calmodulin binding to the autoinhibitory region relieves inhibition of the catalytic sites (Tombes et al. 2003). The phosphorylated form of CaMKII has “autonomous” activity independent of Ca²⁺. This complex autoregulatory behavior of CaMKII allows it to decode the frequency and amplitude of Ca²⁺ spikes (De Koninck and Schulman 1998). In mouse eggs, CaMKII activity rises with the first Ca²⁺ transient and then parallels the subsequent Ca²⁺ oscillations, indicating that this enzyme is a proximal decoder of the Ca²⁺ signals at fertilization (Markoulaki et al. 2004).

In mammals, four genes encode multiple isoforms of CaMKII (α, β, γ, and δ) that have distinct expression and localization patterns allowing for differential regulation of their downstream effectors (Tombes et al. 2003). CaMKIIγ is the predominant isoform in mouse eggs, although CaMKIIδ is also expressed to some degree (Backs et al. 2010). Several lines of evidence indicate that CaMKII is an important proximal effector of Ca²⁺ signaling during egg activation. First, CaMKII inhibitors prevent cell cycle progression in both parthenogenetically activated and fertilized mouse eggs (Johnson et al. 1998). Furthermore, expressing a constitutively active form of CaMKII (lacking the autoinhibitory domain) induces complete egg activation and subsequent formation of developmentally competent blastocysts in the absence of any measurable alterations in egg cytoplasmic Ca²⁺ (Knott et al. 2006). A definitive role for the CaMKII isoform in promoting cell cycle resumption at fertilization was demonstrated using CaMKIIγ knockout mice. Eggs from these mice exhibit a normal pattern of sperm-induced Ca²⁺ oscillations and undergo CG exocytosis; however, they completely fail to resume meiosis and as a result the females are completely infertile (Backs et al. 2010). CaMKII substrates whose phosphorylation is required for meiosis II resumption include Wee1b, an inhibitor of the cell cycle regulator cyclin dependent kinase 1, and endogenous meiotic inhibitor 2, which regulates function of the anaphase promoting complex (Oh et al. 2011; Perry and Verlhac 2008). These studies clearly indicate that CaMKIIγ is a requisite mediator of cell cycle resumption at fertilization.

Later calcium-dependent responses during egg activation

Polar body emission

Emission of the second polar body (PB2) is driven largely by the early Ca²⁺/CaMKII-driven cell cycle resumption events described above. However, there appears to be a continuing role for Ca²⁺ in mediating successful cytokinesis. Activation of the Ca²⁺/calmodulin-dependent enzyme myosin light chain kinase (MYLK2) and its downstream target, myosin II, are required to regulate spindle position and completion of PB2 emission (Matson et al. 2006). The Ca²⁺-dependent PKC isoform, PKCα, and the phosphorylated form of myristoylated alanine-rich protein kinase C substrate (MARCKS) both associate with the contractile ring of the forming PB2 (Kalive et al. 2010; Michaut et al. 2005). Because phosphorylation of MARCKS modulates actin crosslinking activity, these findings suggest a functional role for PKCα in modulating cytoskeletal activity required for PB2 emission.

Subcellular localization of the Ca²⁺ signal involved in PB2 emission appears to be critical for its success. Blocking Ca²⁺ influx in fertilized mouse eggs prevents PB2 emission, despite the occurrence of multiple intracellular Ca²⁺ oscillations and cell cycle resumption; the end result of this treatment is formation of embryos containing three pronuclei (Miao et al. 2012). There is precedent in the plant literature for the idea that Ca²⁺ influx is required for fertilization. In maize, Ca²⁺ influx is initiated locally at the site of sperm-egg cell fusion, is propagated to the opposite pole of the egg cell in association with actomyosin-based cortical contraction, and is followed by an overall increase in cytoplasmic Ca²⁺ (Antoine et al. 2000; Antoine et al. 2001). The Ca²⁺ influx is required for sperm-egg cell fusion (Antoine et al. 2001).

In the mouse, even when Ca²⁺ influx is prevented, PB2 emission occurs when eggs are parthenogenetically activated with a constitutively active form of CaMKII. This finding suggests that Ca²⁺ influx-mediated signaling in the sub-plasma membrane region may act via CaMKII to mediate PB2 emission. The set of molecular players involved in such a signaling pathway likely include PKCα, MARCKS, MYLK2, myosin II and other Ca²⁺-sensitive cytoskeletal elements, but how these molecules might be regulated by Ca²⁺influx and CaMKII remains an open question.

Membrane block to polyspermy

In additional to the relatively rapid establishment of the zona pellucida block to polyspermy, properties of the egg plasma membrane change slowly after fertilization in a way that impairs binding of supernumerary sperm. These changes comprise the “membrane block to polyspermy”, which is assessed experimentally by measuring the ability of fertilized zona pellucida-free eggs to bind and fuse with additional sperm. Like other egg activation events detailed above, the membrane block is a graded response to Ca²⁺ signals, with more effective/more quickly established membrane blocks occurring in eggs that have higher numbers of Ca²⁺ oscillations (Gardner et al. 2007). However, Ca²⁺ signals alone are not sufficient, because eggs that undergo ICSI or that artificially express PLCζ have normal patterns of Ca²⁺ oscillations but do not form a membrane block (Maleszewski et al. 1996; Wortzman-Show et al. 2007). These findings suggest that sperm-egg fusion provides an essential signal to the plasma membrane and/or cortical cytoskeleton that, along with Ca²⁺, is required for establishment of the membrane block.

Maternal mRNA recruitment

During oocyte maturation, maternal mRNAs are recruited to polysomes for translation. This process continues after fertilization so that translation of new proteins required for embryo development can be initiated before transcription begins from the newly formed embryonic genome. Like CG exocytosis and the membrane block to polyspermy, recruitment of maternal mRNAs appears to be a graded response to the number of Ca²⁺ oscillations experienced by the egg (Ducibella et al. 2002). However, artificial methods of inducing cell cycle resumption, in the absence of Ca²⁺ elevation, can result in maternal mRNA recruitment (Knott et al. 2006). These findings indicate that the effects of Ca²⁺ are not direct, but instead occur indirectly by a process downstream of cell cycle resumption.

Calcium signaling during preimplantation embryo development

Mitotic cleavage divisions

Ca²⁺ transients are observed during mitotic cleavage events in embryos of a wide variety of invertebrate and vertebrate species (Whitaker 2006). These Ca²⁺ fluctuations may occur in restricted microdomains, so can be difficult to measure, but appear to be essential for mitotic divisions in fruit fly, sea urchin, starfish, and frog embryos. Although Ca²⁺ transients have been observed in association with pronuclear envelope breakdown in the mouse, these transients are not observed in all embryos (Day et al. 2000), and are not observed in parthenotes that nevertheless undergo mitotic cleavages (Kono et al. 1996). However, artificially induced Ca²⁺ increases provided to pronuclear stage embryos close to the time of nuclear envelope breakdown can accelerate this process, suggesting that although Ca²⁺ transients are not essential in the mouse they can participate in signaling during mitosis (FitzHarris et al. 2005).

Compaction

Eight-cell stage embryos undergo a dramatic morphogenetic process known as ‘compaction’, in which embryos change into a tightly packed cluster of cells from a loosely connected grouping of blastomeres that remain associated with each other largely by virtue of their common location within the zona pellucida. Compaction occurs as a result of increases in cell membrane apposition and increased intercellular adhesive forces, and is dependent on the presence of extracellular Ca²⁺ (Ducibella and Anderson 1975). Indeed, a brief incubation in Ca²⁺-free medium allows for blastomere biopsy and experimental embryo dissociation procedures because this treatment significantly weakens intercellular adhesions. Compaction depends on alterations in both actin- and microtubule-based cytoskeletal structures, and involves cadherin 1, calmodulin and PKC (Kalive et al. 2010; Pey et al. 1998). Cadherin 1-mediated intercellular adhesive interactions depend on Ca²⁺ binding to its extracellular domains (Gumbiner 2005), explaining the reduction in blastomere adhesiveness in the absence of extracellular Ca²⁺. Alterations in intracellular Ca²⁺ modulate cytoskeletal architecture and the function of both calmodulin and PKC, indicating that intracellular Ca²⁺ fluxes are also likely to regulate compaction.

Implantation

Several lines of evidence support the involvement of Ca²⁺ fluxes in regulating implantation. Prior to implantation, or prior to hatching in vitro, blastocyst stage embryos develop “trophectoderm projections”, which are cytoplasmic projections that intermittently protrude through the zona pellucida to interact with the external environment (Gonzales et al. 1996). This process is reminiscent of the behavior of lamellopodia that form on the mobile edge of migrating cells. Cell migration is highly dependent on Ca²⁺ via the actions of numerous Ca²⁺-dependent effectors including calcineurin, calmodulin, actin, myosin, and actomyosin complex-associated proteins such as integrins (Prevarskaya et al. 2011). Both Ca²⁺ influx and Ca²⁺ release from intracellular stores modulate somatic cell migration (Prevarskaya et al. 2011). Based on these findings in somatic cells, it seems likely that Ca²⁺ fluxes modulate extension and retraction of trophectoderm projections. This idea is supported by recent evidence that a small conductance Ca²⁺-activated potassium channel, KCNN3, plays a role in mouse blastocyst hatching and F-actin assembly in vitro (Lu et al. 2012).

Delayed implantation in the mouse can be induced by performing an ovariectomy on the morning of pregnancy day 4 followed by progesterone treatment (Yoshinaga and Adams 1966). The resulting dormant blastocysts can then be “activated” by a single dose of estrogen, and implantation occurs. How estrogen activates the blastocyst is not clear, but there is evidence that Ca²⁺ plays a role. In GnRH neurons, estrogen exposure causes a very rapid increase in the frequency of intracellular Ca²⁺ spikes via a non-genomic signaling pathway (Terasawa et al. 2009). Similarly, intracellular Ca²⁺ is mobilized by treatment of dormant mouse blastocysts with either cell-permeant or cell-impermeant forms of estradiol (Yu et al. 2009b). This effect is not dependent on extracellular Ca²⁺ and is blocked by a PLC inhibitor, suggesting that it occurs as a result of Ca²⁺ release from IP₃-sensitive stores.

Implantation is promoted by interactions between embryonic trophectoderm cells and the endometrium that are in part dependent on Ca²⁺ signals. For example, interaction of extracellular matrix ligands with mouse trophectoderm integrins induces a transient elevation of intracellular Ca²⁺ that is required to strengthen this interaction (Wang et al. 2002). This Ca²⁺ release is apparently mediated in large part by activation of trophectoderm PLCγ2, indicating that it occurs via Ca²⁺ release from intracellular stores rather than downstream of Ca²⁺ influx (Wang et al. 2007). In human cytotrophoblast cells, histamine promotes invasive behavior by inducing PLC-mediated Ca²⁺ release, implicating Ca²⁺ signaling in human implantation (Liu et al. 2004). A role for Ca²⁺ influx-mediated signaling in implantation was suggested by a report describing inhibition of mouse implantation by uterine horn injection with the L-type Ca²⁺ channel blocker diltiazem (Banerjee et al. 2009). This idea was confirmed recently by a series of experiments demonstrating a role for a serine protease-activated, amiloride-sensitive sodium channel (ENaC) in implantation in the mouse (Ruan et al. 2012). Endometrial epithelial cells express ENaC, which is apparently activated by blastocyst-derived trypsin, leading to sodium influx and membrane hyperpolarization. Membrane hyperpolarization triggers activation of L-type Ca²⁺ channels and Ca²⁺ influx, which is the upstream signal for production of molecules essential for implantation including prostaglandin E2 and cyclooxygenase 2 (Ruan et al. 2012). Taken together, these experiments firmly establish roles for both intracellular Ca²⁺ release and Ca²⁺ influx in regulating the signaling pathways required for embryo implantation.

Concluding Remarks

Knowledge regarding the critical nature of the Ca²⁺ signal in driving the events surrounding fertilization and the activation of development has now existed for decades. Much has more recently been discovered regarding the molecular players generating, regulating, and serving as effectors of this signal, though a number of questions remain in this area. However, we are only now beginning to appreciate the importance of the spatiotemporal subtleties that influence how Ca²⁺ signals are generated and transduced. The development of genetically encoded sensors of Ca²⁺ and other signaling components that can now be targeted to various subcellular compartments, combined with the development of microscopes and analysis techniques that have an ever-increasing capacity for detection of molecular interactions, will likely support rapid expansion of this area of research. We anticipate that research using these new methods will give us a much better appreciation of the fundamental concepts underlying the signals that transform the sperm and egg into a totipotent embryo.

Abbreviations

Ca²⁺

calcium

CAMKII

calcium/calmodulin-dependent protein kinase II

CG

cortical granule

cPKC

conventional protein kinase C

DAG

diacylglycerol

ENaC

serine protease-activated, amiloride-sensitive sodium channel

ER

endoplasmic reticulum

ICSI

intracytoplasmic sperm injection

IP₃

inositol 1,4,5-trisphosphate

IP₃R1

inositol 1,4,5-trisphosphate receptor, type 1

MARCKS

myristoylated alanine-rich protein kinase C substrate

MII

metaphase of meiosis II

MYLK2

myosin light chain kinase

PB2

second polar body

PIP₂

phosphatidylinositol 4,5-bisphosphate

PKC

protein kinase C

PLA2

phospholipase A2

PLC

phospholipase C

PLCζ

phospholipase C zeta

SERCA

sarcoplasmic/endoplasmic reticulum calcium ATPase

SOCE

store operated calcium entry