Calcium wave pacemakers in eggs

Egg cells from every animal and plant species studied to date elicit single or repetitive Ca²⁺ transients in response to fertilizing sperm. As an important second messenger, Ca²⁺ triggers a broad range of cellular reactions, including contraction, secretion and gene expression(Berridge, 1997;Bootman et al., 2001). In eukaryotic cells, a calcium wave starts with an initial increase in Ca²⁺ concentration in a restricted region of the cell (a pacemaker site), which then propagates, leading to a global Ca²⁺ wave(Berridge, 1997;Bootman et al., 2001;McDougall et al., 2000;Marchant and Parker, 2001). Several factors are important in determining the transition from non-propagated elementary Ca²⁺ release events (Ca²⁺`puffs') to the initiation and propagation of a global Ca²⁺ wave. They include the magnitude and kinetics of Ca²⁺ release during each elementary event, the Ca²⁺ sensitivity of the Ca²⁺release channels, the spatial organization of release sites, Ca²⁺sequestration and Ca²⁺ diffusion, as well as Ca²⁺buffering within the cytosol (Berridge,1997; Marchant et al.,1999; Marchant and Parker,2001; Bootman et al.,2001). In somatic cells such as hepatocytes, acinar cells or HeLa cells, all or some of these factors contribute to create a subcellular region of higher sensitivity to the Ca²⁺-releasing second messengers,which becomes the pacemaker site (Rooney et al., 1990; Lee et al.,1997; Petersen et al.,1999; Thomas et al.,1999; Ito et al.,1999). In eggs, global Ca²⁺ waves always initiate in the cortex and then propagate through the cortex or the whole cytoplasm(reviewed in Sardet et al.,1998; Stricker,1999; McDougall et al.,2000; Kline et al.,1999; Deguchi et al.,2000). After initiation of the Ca²⁺ wave in the cortex,propagation is due to the sequential activation by Ca²⁺ of Ca²⁺-release channels at the front of the wave(Berridge, 1997;Bootman et al., 2001). The Ca²⁺ wave speed, which mainly depends on the rate of passive Ca²⁺ diffusion between Ca²⁺ release sites, can be influenced by the spatial organization of these Ca²⁺ release sites(Bugrim et al., 1997).

In eggs, Ca²⁺ waves triggered by sperm entry result mainly from the release of Ca²⁺ from intracellular stores by inositol 1,4,5 trisphosphate [Ins(1,4,5)P₃]-induced Ca²⁺release (IICR) (reviewed in Miyazaki et al., 1993; Stricker,1999; McDougall et al.,2000). The mechanism underlying Ins(1,4,5)P₃production in eggs at the time of fertilization is still intensely debated. Similarly, the nature of the sperm factor(s) inducing Ca²⁺ release at fertilization remains elusive, although several competing groups agree that it must be a protein, possibly a form of phospholipase C or an activator of it(see Stricker, 1999;Swann and Parrington, 1999;Parrington et al., 2000;McDougall et al., 2000;Nixon et al., 2000;Runft and Jaffe, 2000;Mehlmann et al., 2001;Jaffe et al., 2001;Carroll, 2001;Runft et al., 2002). Very recently, a mammalian sperm factor was characterized and found to be a new form of PLC [PLCξ (Saunders et al.,2002)]. In most species, it seems reasonable to assume that the entering sperm delivers a factor into the egg and that this factor generates Ins(1,4,5)P₃ either directly or indirectly.

The first wave, which we will refer to as the `fertilization Ca²⁺ wave', is generally the largest and longest-lasting wave, and,in some species, it is followed by repetitive Ca²⁺ waves of lower amplitude and shorter duration. Ca²⁺ wave pacemakers elicit waves for minutes (20-30 minutes in ascidians, 45-60 minutes in some molluscs, 90 minutes in nemerteans) or hours (4 hours in mammals), and they stop operating at the end of the meiotic cell cycles (except in mammals, in which they stop several hours after completion of meiosis, at the time of pronuclei formation). The pacemaker site can be fixed in the cortex or undergo dramatic movements as the cortex is reorganized in preparation for development(Sardet et al., 2002). Ca²⁺ wave pacemakers are either located in a region of enhanced sensitivity to Ins(1,4,5)P₃ or reside in the vicinity of a local source of Ins(1,4,5)P₃. Here, we briefly examine how the subcellular organization of the Ca²⁺ release machinery may create stable Ca²⁺ wave pacemakers in the egg. We also discuss how spatially and temporally regulated production of Ins(1,4,5)P₃ can give rise to multiple calcium wave pacemakers in a single egg cell.

Two main types of egg can be distinguished with regards to their patterns of sperm-triggered Ca²⁺ signals. Eggs of sea urchins, amphibians,cnidarians, nematode and fish display a single Ca²⁺ increase upon fertilization. Conversely, eggs of nemerteans, some molluscs, annelids,ascidians and mammals display repetitive Ca²⁺ waves. In most of these eggs, the oscillations following the fertilization Ca²⁺ wave all emanate from cortical sites distinct from the initial sperm entry site(Eckberg and Miller, 1995;Kline et al., 1999;Deguchi et al., 2000)(reviewed in Sardet, 1998; Stricker,1999). Perhaps the most elaborate of these examples is the egg of ascidians (urochordates at the base of the vertebrate line). The mature ascidian egg is arrested in metaphase I before fertilization, and sperm entry induces two series of Ca²⁺ waves, driving, successively, the completion of meiosis I and meiosis II(Fig. 1). The first series originates from the mobile meiosis-I-associated Ca²⁺ wave pacemaker[the MI pacemaker (McDougall and Sardet,1995)]. The second meiotic cycle is entrained by a second pacemaker (the MII pacemaker) stably located in the vegetal cortex(Speksnijder, 1990b;McDougall and Sardet, 1995;Dumollard and Sardet, 2001)(Figs 1 and2).

We do not completely understand why some eggs display repetitive Ca²⁺ waves whereas others exhibit only a single wave. Recent work on ascidian and mouse eggs reveals that arresting the egg in meiotic metaphase is both sufficient and necessary to sustain sperm-triggered Ca²⁺oscillations (for details, see Jones,1998; Nixon et al.,2000; Carroll,2001). Meiotic `M-phase' thus favors repetitive Ca²⁺waves (as in nemerteans, some molluscs, annelids, ascidians and mammals). By contrast, only a single large fertilization Ca²⁺ wave is observed when fertilization causes a rapid transition to an interphasic cytoplasm(<20 minutes, as in fish or amphibians) and when fertilization takes place during interphase (as in cnidarians or sea urchins). It remains to be seen whether eggs of fish or amphibians can be made to undergo Ca²⁺oscillations when blocked in meiotic M-phase after fertilization. The data relating an M-phase stage of the cell cycle to the ability to generate multiple Ca²⁺ transients is compelling in ascidian and mouse eggs(reviewed in Nixon et al.,2000; Carroll,2001). In these eggs, regulation of the Ca²⁺ release machinery by cell cycle factors probably participates in determining the temporal pattern of the fertilization Ca²⁺ signals. However,whether such regulation proves to be universal requires further research.

The major organelles contributing to the regulation of intracellular Ca²⁺ levels are the endoplasmic reticulum (ER), the plasma membrane and mitochondria. Eggs also possess large numbers of specific vesicular organelles (yolk platelets, pigmented vesicles and cortical granules) that contain Ca²⁺ (Gillot et al.,1991); however their role in Ca²⁺ homeostasis is unknown.

The egg cortex and cytoplasm are filled with an extensive and continuous ER network (Speksnijder et al.,1993; Jaffe and Terasaki,1993; Terasaki et al.,1996; Terasaki et al.,2001; Stricker et al.,1998; Kline et al.,1999). The ER network contains the intracellular Ca²⁺channels — the Ins(1,4,5)P₃ receptors (IP3Rs) and ryanodine receptors (RyRs) — as well as the sarco-endoplasmic reticulum Ca²⁺-ATPases (SERCAs) that pump calcium back into the ER.

Among the three known isoforms of IP3R found in somatic cells(Taylor et al., 1999), IP3R1 is the most prevalent and functionally important isoform in the egg[Xenopus (Runft et al.,1999; Brind et al.,2000); ascidian (Kyozuka et al., 1998)]. Low levels of IP3R2 and IP3R3 have been reported in mouse eggs (Fissore et al.,1999), but their physiological roles remain unclear(Brind et al., 2000;Jellerette et al., 2000). The other family of Ca²⁺ release channels (RyRs) is present on the cortical ER of sea urchin eggs (McPherson et al., 1992) and in ascidian(Albrieux et al., 2000) and mouse eggs (Ayabe et al.,1995). However, except for sea urchins, the involvement of RyRs in the initiation and propagation of sperm-triggered calcium waves appears to be minor, and their role remains unclear (reviewed inMcDougall et al., 2000).

Eggs from all animal phyla seem principally to use Ca²⁺ release from internal stores to generate single or repetitive Ca²⁺ waves at fertilization. In some species, external Ca²⁺ is also used for the fertilization wave [in molluscs (Deguchi et al., 1996) (reviewed inSardet et al., 1998)] or contributes to the maintenance of the repetitive Ca²⁺ waves [in mice (McGuiness et al., 1996) (reviewed inStricker, 1999)]. In many species, voltage-operated Ca²⁺ channels [VOCC(Arnoult and Villaz, 1994;Leclerc et al., 2000)] and Ca²⁺-release-activated Ca²⁺ (CRAC) channels that mediate so-called `capacitative Ca²⁺ entry'(Arnoult et al., 1996;Jaconi et al., 1997;Csutora et al., 1999; Machaca et al., 2000; Putney et al.,2001) are also present, but their role at fertilization is still ill defined.

In the mature mouse egg, the physiological Ca²⁺ load is primarily cleared via SERCAs and plasma membrane Ca²⁺ ATPases(PMCAs). A minor contribution may also be provided by the plasma membrane Na⁺/Ca²⁺ exchanger(Carroll, 2000). PMCAs are probably responsible for Ca²⁺ efflux from ascidian eggs after each Ca²⁺ wave (Kuthreiber et al.,1993) as well as for the loss of total Ca²⁺ content after fertilization in sea urchin eggs(Gillot et al., 1991).

In the past few years, mitochondria have been shown to be major regulators of Ca²⁺ signals (reviewed inRutter and Rizzuto, 2000;Rizzuto et al., 2000;Duchen, 2000). Sequestration of Ca²⁺ by mitochondria has two regulatory effects on IICR,suppressing positive and negative Ca²⁺ feedback on the opening of the IP3R. In addition, ATP production by mitochondria might provide a further means of modulating Ca²⁺ signals: ATP^4- sensitizes the IP3R (Mak et al., 1999;Mak et al., 2001), whereas Mg²⁺-complexed ATP is consumed to refill the ER Ca²⁺stores. Mitochondria can thus provide negative or positive feedback on Ins(1,4,5)P₃-mediated Ca²⁺ signals. Such negative feedback has been reported in a wide range of somatic cells. For example, initiation of global Ca²⁺ waves in myocytes preferentially occurs in mitochondrion-poor regions of the cell(Boitier et al., 1999). A positive feedback effect of mitochondria on Ins(1,4,5)P₃-mediated signals has been reported only in oligodendrocytes, in which Ca²⁺ wave initiation and amplification sites are found in mitochondrion-rich regions of the cell(Simpson et al., 1997).

Except in sea urchins, in which mitochondria are a sink for cytosolic Ca²⁺ (Eisen and Reynolds,1985; Girard et al.,1991), the role mitochondria play in Ca²⁺ signalling in eggs remains largely obscure. In mouse eggs, collapsing mitochondrial potential impairs Ca²⁺ clearance from the cytosol(Liu et al., 2001), but no picture of the regulation of Ca²⁺ oscillations by mitochondria can be drawn from only this study. In ascidian eggs, mitochondria contribute to the activity of the second Ca²⁺ wave pacemaker both by buffering cytosolic Ca²⁺ and by locally providing ATP (R.D., unpublished). Nevertheless, an understanding of the role of mitochondria in regulating Ca²⁺ wave pacemakers will require measurement of the local intracellular Ca²⁺ concentration and the local mitochondrial ATP production in the vicinity of the IP3Rs. The recent development and subcellular targeting of GFP-based Ca²⁺ and Ins(1,4,5)P₃ indicators as well as luciferase-based ATP indicators should allow the direct measurement of mitochondrial Ca²⁺ levels, intracellular ATP concentration and the Ins(1,4,5)P₃ concentration in the living zygote(Hirose et al., 1999;Rutter and Rizzuto, 2000).

Given the central role played by Ca²⁺ release from intracellular stores, the organization of the ER in eggs has received much attention. In eggs that display repetitive Ca²⁺ waves, the interconnected network of ER sheets and tubes is organized into ER-rich domains (also called ER clusters). The cytoplasm also has ER-poor domains, which contain high densities of mitochondria and/or other vesicular organelles(Speksnijder et al., 1993;Stricker et al., 1998;Kline et al., 1999;Dumollard and Sardet, 2001). The ER clusters are made of densely packed tubes and sheets of ER membrane(Speksnijder et al., 1993;Fissore et al., 1999;Terasaki et al., 2001;Dumollard and Sardet, 2001). In oscillating eggs [nemertean (Stricker et al., 1998); mouse (Kline et al., 1999); ascidian(Dumollard and Sardet, 2001;Sardet et al., 2002)] as well as in Xenopus eggs (Terasaki et al., 2001), ER-rich domains are concentrated in the 2-6 μm thick layer beneath the plasma membrane and are more dispersed in the deeper cytoplasm (deeper than 5 μm). In ascidian eggs, meiosis II Ca²⁺waves initiate in a large cortical disc of concentrated ER tubes and sheets(20 μm in diameter and 2-5 μm thick) located in the vegetal contraction pole (Fig. 2B-F)(Speksnijder, 1992;McDougall and Sardet, 1995;Dumollard and Sardet, 2001). In mouse and nemerteans eggs, ER clusters also line the vegetal cortex of the egg where meiotic Ca²⁺ waves are initiated(Stricker et al., 1998;Kline et al., 1999).

In Xenopus and mouse eggs, at least, these ER-rich domains are rich in IP3R1s (Terasaki et al.,2001; Mehlmann et al.,1996; Fissore et al.,1999). The appearance of cortical ER-rich domains during maturation correlates with an increase in sensitivity to Ins(1,4,5)P₃ and to sperm-induced Ca²⁺ release(Chiba et al., 1990;Shiraishi et al., 1995;Mehlmann and Kline, 1994;Terasaki et al., 2001)(reviewed in Sardet et al.,2002). Local injections of Ins(1,4,5)P₃ and sperm extracts in mouse eggs have revealed that the egg cortex is a region of higher sensitivity to Ins(1,4,5)P₃ and to sperm extracts(Oda et al., 1999). Indeed,although the abundance of ER in the egg cortex renders this region more sensitive to Ins(1,4,5)P₃, it is also exposed to the highest concentrations of Ins(1,4,5)P₃ as it is closest to the source of PtdIns(4,5)P₂ in the plasma membrane(Halet et al., 2002;Sardet et al., 2002).

In several somatic cells, the location of the Ca²⁺ wave pacemakers corresponds to the area of the cell that is most sensitive to Ins(1,4,5)P₃ (Ito et al., 1999; Thomas et al.,1999; Petersen et al.,1999), and Ins(1,4,5)P₃, which diffuses rapidly in the cytoplasm, is thought to act as a global messenger(Albritton et al., 1992;Kasai and Petersen, 1994).

The mouse MII pacemaker appears to be an example of this type of pacemaker. In the mature mouse egg, ER-rich domains are larger in the vegetal cortex(Kline et al., 1999), whereas mitochondria are more abundant in the animal hemisphere(Calarco, 1995;Van Blerkom et al., 2002). The MII pacemaker of the mouse egg resides in the ER-enriched vegetal cortex,which is probably a site of enhanced sensitivity to Ins(1,4,5)P₃. Therefore, similarly to the somatic cell Ca²⁺ wave pacemakers, the mouse MII pacemaker site appears to be determined by the organization of the Ca²⁺ stores of the egg, with Ins(1,4,5)P₃ acting as global messenger(Fig. 1,Fig. 2A).

Interestingly, an artificial pacemaker can be induced in the ascidian egg by global uncaging of caged Ins(1,4,5)P₃(cIns(1,4,5)P₃) or its poorly metabolised analogue cgPtdIns(4,5)P₂ [caged 1-(a-Glycerophosphoryl)-D-myo-inositol 4,5-bisphosphate, P4(5)]. This artificial pacemaker, localized in the animal pole of the ascidian egg,functions under globally elevated Ins(1,4,5)P₃ levels and thus resides in the region of highest sensitivity to Ins(1,4,5)P₃(Dumollard and Sardet, 2001)(Figs 1 and2). In common with the mouse MII pacemaker, the location of this artificial pacemaker can be explained by asymmetries in the distribution of the ER along the animal-vegetal axis of the ascidian egg. In these eggs, the ER-rich domains invade the whole egg except for the vegetal subcortex, where most mitochondria accumulate(Fig. 2)(Dumollard and Sardet, 2001). The corollary of this is that the sperm-triggered MII pacemaker in the ascidian egg, located in the vegetal pole (the site opposite the artificial pacemaker), is not at a site of enhanced Ins(1,4,5)P₃sensitivity. This indicates that the general organization of the ER stores in these eggs is not sufficient to determine the pacemaker site. The pacemakers in the ascidian egg may then rely on mechanisms other than a global increase in Ins(1,4,5)P₃ levels.

In ascidians, the MI pacemaker stimulated by sperm entry resides in a cortical ER-rich domain that forms rapidly around the sperm nucleus and centrosome and moves with them towards the vegetal pole(Fig. 1)(Dumollard and Sardet, 2001). This suggests that the MI pacemaker of the ascidian relies on a localised moving source of Ins(1,4,5)P₃. The ascidian MII Ca²⁺ wave pacemaker does not reside in a region of enhanced sensitivity to Ins(1,4,5)P₃(Fig. 2)(Dumollard and Sardet, 2001);it might thus require local production of Ins(1,4,5)P₃ in the vegetal contraction pole. The contraction pole possesses numerous microvilli and is thus rich in PtdIns(4,5)P₂ (Figs1 and2)(Sardet et al., 2002). Therefore, the ascidian MII pacemaker may be different from the characterized pacemakers of somatic cells and the mouse MII pacemaker, since it would rely on the apposition of cortical ER-rich clusters to a local source of Ins(1,4,5)P₃ (Fig. 1) (Dumollard and Sardet,2001).

Hypothesizing local production of Ins(1,4,5)P₃ even in the large egg cell raises the question of how such a gradient is maintained. Indeed, as Ins(1,4,5)P₃ diffuses rapidly through the cytosol, locally produced Ins(1,4,5)P₃ would quickly invade the whole cell, making Ins(1,4,5)P₃ gradients energy consuming to maintain without dynamic Ins(1,4,5)P₃production. Theoretically, repetitive Ca²⁺ waves can result from either a sustained increase in Ins(1,4,5)P₃ levels or an oscillating production of Ins(1,4,5)P₃(Jacob, 1990). In ascidians, a single, large and sustained Ins(1,4,5)P₃ increase(achieved by uncaging cgPtdIns(4,5)P₂ in the whole egg,Fig. 1) mimics the first series of Ca²⁺ oscillations, indicating that the ascidian MI pacemaker is driven by a continuous moving source of Ins(1,4,5)P₃induced by sperm entry (Dumollard and Sardet, 2001). Similarly in mouse eggs, a slow and continuous uncaging of cIns(1,4,5)P₃ can reproduce the sperm-triggered Ca²⁺ oscillations(Jones and Nixon, 2000). Therefore, the mouse MII Ca²⁺ wave pacemaker can also be regulated by a single and sustained increase in Ins(1,4,5)P₃ levels(Fig. 1). Furthermore, the Ca²⁺ transients triggered by the ascidian MI and artificial pacemakers, as well as those triggered by the mouse MII pacemaker, are all preceded by a characteristic slow rise in Ca²⁺ levels called a`pacemaker Ca<sup>2+</sup> rise' (Fig. 1) (Jones and Nixon,2000; Dumollard and Sardet,2001). This `pacemaker Ca²⁺ rise' is a hallmark of low-frequency (period >20 seconds) Ca²⁺ oscillations generated under constantly elevated Ins(1,4,5)P₃ levels(Jacob, 1990;Marchant and Parker, 2001),which further suggests that a single and sustained Ins(1,4,5)P₃ increase regulates the ascidian MI pacemaker and the mouse MII pacemaker.

By contrast, the ascidian MII pacemaker cannot be reproduced by a long-lasting increase in Ins(1,4,5)P₃ levels in the egg,and no `pacemaker Ca²⁺ rise' precedes these Ca²⁺transients (Fig. 1)(Dumollard and Sardet, 2001). An oscillating production of Ins(1,4,5)P₃ from the contraction pole might underlie the activity of the ascidian Ca²⁺wave MII pacemaker (Fig. 1). The recent finding that the mammalian sperm factor is possibly a Ca²⁺-activated phospholipase C (PLC)(Rice et al., 2000;Saunders et al., 2002) argues in favor of Ins(1,4,5)P₃ oscillations driving sperm-triggered Ca²⁺ oscillations in eggs. Indeed, the prolonged stimulation of a Ca²⁺-activated PLC can result in Ca²⁺oscillations regulated by an oscillating production of Ins(1,4,5)P₃ (for details, seeMeyer and Stryer, 1988). In addition, Ins(1,4,5)P₃ oscillations regulating repetitive Ca²⁺ waves during prolonged exposure to agonists have now been observed in several types of somatic cells(Hirose et al., 1999;Nash et al., 2001). The issue of Ins(1,4,5)P₃ oscillations is an intensively debated topic in cell physiology, and the Ca²⁺ wave pacemakers in eggs could provide an invaluable experimental system to resolve such questions in the future.

Even though the ascidian and the mouse Ca²⁺ wave pacemakers seem to rely on different mechanisms, they are both located in the vegetal cortex of the egg (as is the pacemaker in eggs of the primitive nemertean). This suggests that the location the Ca²⁺ wave pacemaker may have developmental significance.

The polarized nature of the calcium signals may in itself influence embryonic patterning by regulating early embryonic cleavages. In ascidians,nemerteans and mouse, the egg cortex is polarized along the animal-vegetal axis and, in ascidians, this polarity amplifies after fertilization through actomyosin-driven cortical contractions(Sardet et al., 2002). Is the generation of repetitive Ca²⁺ waves from the vegetal cortical pacemaker a mechanism used to prime the vegetal pole region for later developmental events such as cleavage or gastrulation, which, in nemerteans and ascidians, takes place in the vegetal/dorsal pole of the embryo? Mouse embryos were long thought to have no significant polarity until the late cleavage stage, but recent marking experiments show that in fact, as in ascidians and nemerteans, although regulation can override this polarity,there is a relationship between the animal-vegetal axis, the sperm entry point and the developmental axes of pre- and post-implantation embryos (reviewed inLu et al., 2001).

Finding out whether Ca²⁺ wave patterns play a role in later development will require studies that interfere with the normal spatio-temporal pattern of Ca²⁺ waves without perturbing mitosis and cleavage. The rather simple ascidian embryo, which displays two different meiotic Ca²⁺ wave pacemakers and develops into a swimming tadpole within a day, is particularly suited to studies of the relationship between meiotic Ca²⁺ waves and development(Fig. 2)(Dumollard and Sardet, 2001). It should be possible in the future to relate patterns of Ca²⁺waves and phenotypic differences in embryos.

We thank Mark Larman and Karl Swann for their help with imaging calcium waves in mouse eggs injected with sperm extracts. We are also grateful to Christian Rouviere and Mohammed Khamla for technical assistance.