Ca²⁺ oscillations induced by a cytosolic sperm protein factor are mediated by a maternal machinery that functions only once in mammalian eggs

In all mammalian species so far studied, the sperm activates the egg by triggering a series of Ca²⁺ oscillations, which are essential for egg activation and entry into the first mitotic cell cycle (Kline and Kline, 1992; Sun et al., 1992, 1994; Fissore et al., 1992; Fissore and Robl, 1994; Miyazaki et al., 1993; Whitaker and Swann, 1993; Schultz and Kopf, 1995; Swann and Lai, 1997). These oscillations are spatially organized as propogating waves (Miyazaki et al., 1992) and persist for several hours until pronuclear formation (Jones et al., 1995a; Kono et al., 1996).

The precise mechanism as to how the fertilizing sperm causes Ca²⁺ oscillations in the egg remains unclear. One hypothesis proposes that the sperm introduces a soluble cytosolic factor, which triggers Ca²⁺ release in the egg following gamete membrane fusion (Swann, 1990; Swann and Lai, 1997). This is directly supported by the finding that cytosolic sperm extracts could induce fertilization-like Ca²⁺ oscillations when microinjected into eggs (Swann, 1990; Homa and Swann, 1994; Sun and Moor, 1995; Parrington et al., 1996; Wu et al., 1997; Stricker, 1997; Kyozuka et al., 1998). It is evident that the sperm factor is protein-based and its Ca²⁺ oscillation-inducing activity is not species specific in mammals (reviewed by Swann and Lai., 1997). This factor appears to be sperm specific and is functional only when microinjected (Swann, 1990, 1994). So far several sperm factor candidates have been proposed, but no recombinant protein has yet been found to induce Ca²⁺ oscillations similar to those seen at fertilization (Parrington et al., 1996; Sette et al., 1997; Wolosker et al., 1998; Jones et al., 1998).

The mechanism of action of the sperm factor in causing Ca²⁺ oscillations is also a mystery. The sperm factor mobilizes Ca²⁺ release from the intracellular stores and the maintenance of Ca²⁺ oscillations is dependent upon Ca²⁺ influx (Swann, 1994). The sperm factor may cause Ca²⁺ release through InsP₃ receptor-mediated mechanisms, since blocking the function of the InsP₃ receptor completely inhibited sperm-induced Ca²⁺ release (Miyazaki et al., 1992, 1993). However, it is not known whether the sperm factor mobilizes intracellular release by producing InsP₃ or by acting directly with the InsP₃ receptor (Swann and Lai, 1997). The specificity of action of the sperm factor in triggering the long-lasting Ca²⁺ oscillations in mammalian eggs is also unknown.

The present study was undertaken to study the specificity of action of the sperm factor in triggering Ca²⁺ oscillations in mammalian eggs. We show that the Ca²⁺ oscillatory response of mouse eggs to the sperm factor stimulation depends on their previous history: once fertilized or stimulated by the sperm extracts, their ability to generate Ca²⁺ oscillations upon re-stimulation with the sperm extracts is diminished. Our results suggest that sperm-factor-induced Ca²⁺ oscillations in mammalian eggs require a maternal machinery. This machinery functions only once in mammalian oocytes and eggs, and is inactivated by the sperm or sperm extracts injection but not by parthenogenetic activation. In addition, we found that the loss of the oscillatory ability after fertilization is not due to changes in either InsP₃ receptor sensitivity to InsP₃ or Ca²⁺ pool size.

Cytosolic sperm extracts were prepared from either bovine or mouse sperm. Fresh bovine spermatozoa were obtained from a local Dairy Farm. Mouse sperm was taken from the cauda epididymis of sexually mature males.

Preparation of sperm cytosolic extracts was conducted following the procedures described by Swann (1990) with some modifications. Briefly, the spermatozoa were suspended in PBS and then centrifuged at 5,000 g for 10 minutes. After three repeated washing in PBS, the spermatozoa were then washed into an extracting buffer (120 mM KCl, 20 mM Hepes, 200 μM EDTA, 0.05% Brij35, 300 mM PMSF, 10 mM Leupeptin, pH 7.3), and lysed by sonication using an ultrasonic homogenizer (Cole Parmer, USA).

It should be stressed here that for defining the injection dosage of the sperm extracts, we always determined the sperm density of each preparation before sonication. This is done by hemacytometer readings of the final sperm suspension. After sonication, homogenates were then spun at 40,000 g for 30 minutes at 4°C. The clear supernatant was concentrated on centricon-30 ultrafiltration membranes (Amicon, UK). The concentrated fraction was diluted into an intracellular-like medium (ICM, the extracting medium without Brij35 and PMSF) and re-concentrated on the centricon-30 ultrafiltration membranes.

The retentate was then adjusted to a final concentration equivalent to 1-1.5×109 sperm/ml, namely, 1-1.5 sperm in every picoliter. This is calculated based on the equation: sperm equivalents per picoliter = (total sperm number in the suspension symbol ÷ final volume (ml) of the retentate) symbol ÷ 109. The protein concentration of the sperm extracts was estimated 0.2∼0.5 mg/ml. This collection of retentate, which contains 1-1.5 sperm equivalents per picoliter, is called ‘the sperm extracts’ throughout our paper. In our studies, the physiological dosage of the sperm extracts refers to injection of 2 pl of the above sperm extracts into a single mouse egg or oocyte. Inactivated sperm extracts were obtained by heating the bovine sperm extracts at 60°C for 30 minutes (Swann, 1990).

Female ICR mice (4-6 weeks old, bred in the Institute of Developmental Biology) were superovulated by serial injection of pregnant mares’ serum gonadotropin (PMSG) and human chorionic gonadotropin (hCG) 48 hours apart. MII eggs were collected from the oviducts at 14-16 hours post-hCG into prewarmed H6 medium (Nasr- Esfahani et al., 1990) supplemented with 4 mg/ml BSA (Fraction V, Sigma). The cumulus cells were removed by a brief incubation at 37°C in hyaluronidase (0.3 mg/ml, Sigma) in H6+BSA. The MII eggs were then washed three times in H6+BSA and transferred to microdrops of H6+BSA under paraffin oil pending further treatments. To obtain fertilized eggs (zygotes) at different cell cycle stages, female mice were paired with males immediately following hCG injection. G1-, S- and M-phase zygotes were subsequently collected at 18-19, 23-24 and 30 hours post-hCG (Howlett and Bolton, 1985). It should be noticed here that, in our study, the M-phase zygotes refer to those fertilized eggs that had already undergone complete nuclear membrane breakdown.

Parthenogenetic eggs were produced by exposure of MII eggs at 17-18 hours post-hCG either to 7% ethanol in M2 medium (Fulton and Whittingham, 1978) for 7 minutes at room temperature (Cuthbertson, 1983), or to Ca2+-free H6 medium containing 20 mM Sr2+ for 3 hours (Kline and Kline, 1992). After activation, the MII eggs were cultured in microdrops of CZB medium (Chatot et al., 1989) at 37°C, 5% CO2 in air. G1-, S- and M-phase of parthenogenetic eggs were obtained at 5-6, 10-11 and 16-17 hours after ethanol activation. Those eggs that had already undergone complete nuclear membrane breakdown were referred as M-phase parthenogenetic eggs.

Sperm-extract-activated eggs were produced by injecting mouse MII eggs at 14-16 hours post-hCG with 2 pl (equivalent to 2-3 sperm) of the sperm extracts in a protein concentration of 0.2∼0.5 mg/ml. This injection dosage is sufficient to induce fertilization-like Ca2+ oscillations in mouse MII eggs. Interphase and metaphase eggs activated by the sperm extracts were obtained at 10-12 hours and 18- 19 hours after the sperm extract injection.

MI oocytes were obtained by isolating germinal vesicle (GV)-stage oocytes from fully grown follicles. GV oocytes were cultured in T6 medium supplemented with 4 mg/ml BSA (Howlett et al., 1987). After 6-7 hours of culture, the majority of GV oocytes reached MI phase. The MI oocytes were thereafter either treated with 20 mM Sr2+ to induce artificial Ca2+ increases, or injected with active or inactivated bovine sperm extracts. They were then cultured for another 8-9 hours for maturation to MII stage. Those mature oocytes showing the first polar body were used for further sperm extract injection.

Micromanipulation of eggs was conducted following the procedures described by Sun et al. (1991, 1994). All injections were carried out using a pair of manually operated pressure microinjectors (IM-6,Narishige, Japan) filled with paraffin oil. Microinjection pipettes (Clark Electromedical Instruments, UK) were pulled by a micropipette puller (Sutter Instrument company, Model P-87, USA) to give an open tip (inner diameter 1-2 μm) and a long taper of about 10 mm in length. The microinjection pipette was connected to the microinjector and then filled with paraffin oil.

For quantitative injections, it is essential to measure the injection volume of samples accurately. In our experiments, the injection volume is measured as follows: under a 100× objective using an eyepiece fitted with a graticule, (1) the pipette was arranged horizontally to the injection chamber, which contained a microdrop (5 μl) of Ca2+-free H6 medium covered by light paraffin oil, (2) a small amount of H6 medium was sucked into the injection pipette, followed by oil to a defined length (from the tip to the interface between oil and H6 medium in the pipette), (3) the pipette was moved back to the H6 medium and the oil gently expelled into the medium to generate a ball-shaped microdrop of oil at the tip of the pipette, and (4) the diameter of the microdrop of oil was measured with the precalibrated graticule and its volume calculated based on the equation: V=4/3 × πR3. In this way, each defined volume in the pipette corresponds to a defined length in the pipette determined by the eye graticule. Once a defined volume and its length in the pipette was determined, a range of injection volumes (from 1 pl to 10 pl) and their corresponding lengths in the pipette against the graticule can be defined following the step 1 to 4. The calibrated data were then checked with different batches of pipettes pulled by the same puller using the same pulling parameters. We found that there is no need to check each pipette before injection, because the shape of micropipettes pulled with the same puller using the same parameters are almost identical.

All micromanipulations were conduced in a chamber containing microdrops of Ca²⁺-free H6 medium supplemented with 4 mg/ml BSA under paraffin oil unless otherwise stated. Before injection, we loaded a small drop of the sperm extracts or InsP₃ into the chamber, then moved the injection pipette to the microdrop containing the injection sample and very gently sucked a defined volume of sample into the pipette. After this, (1) the injection pipette was moved back to the microdrop of H6 containing the cells to be injected, (2) the egg or zygote or oocyte was secured by a holding pipette, (3) the injection pipette was inserted into the egg until its tip almost reached the opposite side of the egg’s cortex, and (4) immediately after the oolemma was broken by gentle negative-pressure sucking, the quantified sample was expelled into cytoplasm of the cell and the micropipette was immediately withdrawn. Intracellular concentrations were calculated from the concentration and volume of solution injected, assuming uniform distribution and an egg or zygote volume of 200 pl. Immediately after injection, the cell was moved to a chamber of our imaging system to detect intracellular Ca²⁺changes. In our hands, approximately 60∼80% injected eggs could survive after the injection. It should be pointed out here, when responses of zygotes and eggs to the stimulation of InsP₃ injection were analyzed, the injection was directly performed on the stage of a Nikon Diaphot 200 inverted epifluorescence microscope of the Ca²⁺imaging system. In this study, intracellular concentrations of 5 nM or 10 nM InsP3 were achieved by injecting into eggs or zygotes 5 pl of 200 nM or 400 nM InsP3 stock solution, respectively.

Eggs or zygotes were loaded with 2 μM fura-2 acetoxymethlyester (fura-2/AM, Molecular Probes) for 30 minutes in H6 containing 4 mg/ml BSA at 37°C immediately before measurement (Deng et al., 1998). After loading, the cells were washed three times in H6 and then transferred to a chamber containing H6 medium covered by light paraffin oil. The chamber was placed in a well on the stage of a Nikon Diaphot 200 inverted epifluorescence microscope (Nikon Instruments, Garden City, NY) for imaging, and maintained at 37°C by a thermostatic controller (Life Sciences Resources, Cambridge, UK). The system used in our laboratory for [Ca²⁺]i measurements was a MiraCal imaging system equipped with MiraCal Version 2.3 Software (Life Sciences Resources, Cambridge, UK). The emitted fluorescence intensities at 510 nm were recorded at 340 nm and 380 nm excitation wavelengths by Mira-1000TE low-light-level CCD camera. The fluorescence signal is displayed as the ratio of fluorescence intensities for the 340 nm/380 nm excitation wavelengths. [Ca²⁺]i was estimated from the ratio equation described by Grynkiewicz et al. (1985), which is calculated by computer simultaneously. The intensity of the excitation ultraviolet light was reduced by neutral density filters. Parameters required for the ratio equation were obtained according to the method of Poenie et al. (1985). The [Ca²⁺]i image was recorded every 2-10 seconds for up to 2-4 hours.

Injection of cytosolic extracts derived from mouse or bovine sperm will cause Ca²⁺ oscillations in mouse MII eggs (Fig. 1), confirming that the sperm extracts contain a Ca²⁺ oscillation- inducing factor. In this study, we chose to use bovine sperm as producers of the sperm protein factor, because (1) the Ca²⁺ oscillation-inducing ability of the sperm factor is not species specific in mammals (Swann, 1990; Homa and Swann, 1994; Wu et al., 1997), and (2) they can be easily obtained in sufficient amounts compared to the mouse.

To determine whether the ability of the sperm factor in triggering Ca²⁺ oscillations is seen specifically in unfertilized eggs or universally in zygotes, parthenogenetic eggs and maturing oocytes, we subsequently injected the cells with the physiological dosage of the sperm extracts and compared their patterns of calcium signaling after stimulation. Responses of the cells to the sperm extract injection are summarized in Table1. We found that injecting the sperm extracts into MII eggs induced Ca²⁺ oscillations in all eggs examined (n=8). The oscillatory process in the injected eggs persisted for 2-3 hours with a mean amplitude of 444±107 nM and a spiking interval of 2.6±2.0 minutes. Similarly, all of the injected MI oocytes (n=20) showed Ca²⁺ oscillations, which lasted for 1-3 hours with a mean amplitude of 344±70 nM and a spiking interval of 4.3±2.4 minutes. However, when the same amount of sperm extracts was injected into zygotes, none of them showed Ca²⁺ oscillations (n=84). These observations suggest that, once the egg is fertilized, its ability to generate Ca²⁺ oscillations upon stimulation with the sperm factor is diminished.

To find out whether parthenogenetically activated eggs and zygotes share a similar property of generating sperm-factor- induced Ca²⁺ oscillations, we injected parthenogenetic eggs and zygotes at various cell cycle stages with the physiological dosage of the sperm extracts and analyzed their patterns of Ca²⁺ signaling. It is important to note here that the M-phase zygotes and parthenogenetic eggs used throughout this investigation refer to cells that had already undergone complete nuclear envelope breakdown (NEBD). We chose to use those zygotes because after completion of NEBD no further endogenous Ca²⁺ release was detected in them during the first mitosis (our unpublished data). From the results in Table 1, we found that when the sperm extracts were injected into interphase (G1- and S-phase) cells, neither fertilized nor parthenogenetically activated eggs showed Ca²⁺ oscillations, suggesting that interphase marks a period that is insensitive to the sperm factor stimulation in triggering Ca²⁺ oscillations. Furthermore, we found that, although interphase parthenogenetic eggs failed to generate Ca²⁺ oscillations, the majority of them (17 out of 18 injected eggs) could generate a single Ca²⁺ transient. In contrast, only 7% of the injected zygotes (3 out of 45 cells examined) could generate one Ca²⁺ rise, the others did not show any detectable Ca²⁺ increase, indicating that the interphase fertilized eggs are less sensitive to sperm factor’s stimulation than the parthenogenetic eggs. More interestingly, we have found that, at metaphase of the first mitosis, the response of parthenogenetic eggs to the sperm extract injection is strikingly different from that of the injected zygotes. Majority of injected parthenogenetic eggs (16 out of 18 eggs examined) underwent Ca²⁺ oscillations while none of the injected zygotes (n=39) did. It is observed that the first Ca²⁺ transient in the M-phase parthenogenetic eggs has a higher amplitude and a longer duration than the following ones (see Table 1). These injected parthenogenetic eggs could oscillate 1-2 hours with a mean amplitude of 282±70 nM and a spiking interval of 10.2±6.9 minutes. Our results indicate that the ability to generate sperm-factor- induced Ca²⁺ oscillations is preserved in the M-phase parthenogenetically activated eggs, but not in the M-phase fertilized eggs.

In conclusion, our results suggest that fertilization results in a functional alteration of the egg’s Ca²⁺ releasing property so that Ca²⁺ oscillations cannot occur when the sperm factor is injected into already fertilized eggs.

We postulated that the inability of zygotes to generate sperm- factor-induced Ca²⁺ oscillations is probably attributed to inactivation of a maternal machinery which functions only once in mammalian oocytes and eggs. To test this possibility, we have performed three sets of experiments.

In the first set, we examined whether activated eggs that had already undergone Ca²⁺ oscillations were still able to generate Ca²⁺ oscillations when microinjected with the sperm extracts at M-phase of the first mitosis. In doing so, we injected the experimental eggs with the physiological dosage of the sperm extracts to induce fertilization-like Ca²⁺ oscillations (Fig. 2A), while the control eggs were activated by Sr²⁺, an artificial stimulus that can also cause Ca²⁺ oscillations (Kline and Kline, 1992). The general pattern and major characteristics of Sr²⁺- induced Ca²⁺ oscillations in MII eggs are shown in Fig. 2C and Table 2. We found that, after exposure in Ca²⁺-free H6 medium containing 20 mM Sr^2+, all treated MII eggs underwent repetitive Ca²⁺ transients (n=30). The activated eggs (judged by pronuclear formation) in both control and experimental group were cultured to the next metaphase and then injected with the physiological dosage of the sperm extracts. Table 3 summaries the responses of the two groups of eggs to the sperm extract injection at M-phase of the first mitosis. It was found that after sperm extract injection majority of the eggs in the control group (12 out of 19 cells examined) underwent Ca²⁺ oscillations (Fig. 2D). On the contrary, none of the injected eggs in the experimental group was able to generate repetitive Ca²⁺ transients (Fig. 2B; n=8). This suggests that the ability to generate sperm-factor-induced Ca²⁺ oscillations is abolished by the sperm extracts but not by parthenogenetic activation.

The second set of experiments was designed to determine whether the abolition of Ca²⁺ oscillatory ability by the sperm extracts is cell cycle dependent. To test it, we used Sr²⁺- activated parthenogenetic eggs as the model and the ability to generate Ca²⁺ oscillations at metaphase of the first mitotic division as the parameter. It is shown in Fig. 3 and Table 4 that injecting the parthenogenetic eggs at interphase with the sperm extracts diminished the egg’s ability to generate Ca²⁺ oscillations, while the control eggs injected with heat- inactivated sperm extracts was still sensitive to the sperm extract injection. In combination with the data in Fig. 2, our results indicate that the ability of the parthenogenetic eggs to generate sperm-factor-induced Ca²⁺ oscillations can be abolished by the sperm extracts independently of the cell cycle stages.

In the third set of experiments, we tested whether the MII egg would lose its ability of generating sperm-factor-induced Ca²⁺ oscillations if the cells had been stimulated by the sperm extracts during oocyte maturation. To do this, we injected MI oocytes in the experimental group with the functional sperm extracts while the control eggs were either injected with heat-inactivated sperm extracts or stimulated by Sr²⁺. The general pattern and major characteristics of sperm extract-induced or Sr²⁺-induced Ca²⁺ oscillations in MI oocytes are shown in Fig. 4Aa and Table 1, and Fig. 4Ca and Table 2, respectively. After the treatments, both the experimental and control eggs were cultured to MII stage and then microinjected with the physiological dosage of the sperm extracts. The responses of the experimental and control groups of eggs to the sperm extract injection are shown in Table 5 and Fig. 4. We found that the injected MII eggs in the experimental group could undergo a single Ca²⁺ transient but no oscillations (n=9). On the contrary, the eggs injected with heat-inactivated sperm extracts (Fig. 4B; 11 out of 11 cells examined) or prestimulated with Sr²⁺ (Fig. 4C; 9 out of 10 eggs) showed long-lasting Ca²⁺ oscillations. These observations therefore clearly demonstrate that the egg’s ability to generate sperm-factor-induced Ca²⁺ oscillations can be abolished if the cells were stimulated by the sperm extracts during oocyte maturation.

In conclusion, our results suggest that the long-lasting Ca²⁺ oscillations at fertilization are critically dependent upon a functional machinery that functions only once in mouse oocytes and eggs, and is abolished by sperm-derived factors but not by parthenogenetic activation.

The striking difference between metaphase parthenogenetic eggs and zygotes in their ability to generate Ca²⁺ oscillations upon stimulation with the sperm extracts may provide clues as to how the sperm factor triggers Ca²⁺ oscillations in mammalian eggs at fertilization. Since the sperm factor may induce Ca²⁺ release via the InsP₃ pathway (Miyazaki et al., 1992, 1993; Jones et al., 1998), we consider that any differences between the zygotes and the parthenogenetic eggs in their InsP₃ receptor sensitivity to InsP₃ may influence Ca²⁺ signaling induced by the sperm extracts. To determine this, we compared the responses of zygotes and parthenogenetic eggs to physiological dosages of InsP₃ injection. We chose to test two dosages (10 nM and 5 nM, final intracellular concentrations), because we had found in the mouse eggs that 10 nM InsP₃ was just sufficient to caused maximal Ca²⁺ release in the MII eggs while 5 nM was close to the lowest concentration required to cause sufficient Ca²⁺ release from the InsP₃-sensitive stores (data not shown here). To our surprise, the results in Fig. 5 show that at M-phase the amplitude of InsP₃-induced Ca²⁺ release in the parthenogenetic eggs is significantly lower (P<0.01) than that in the zygotes (both at 5 nM and at 10 nM), despite the fact that parthenogenetic eggs can generate Ca²⁺ oscillations while the zygotes cannot. In addition, it was found that, at interphase, the two types of cells did not show significant differences in their amplitude of Ca²⁺ release (P>0.05). These results therefore indicate that the sensitivity of InsP₃ receptor to InsP₃ stimulation is unlikely to be the determinant that restricts sperm-factor-induced Ca²⁺ oscillations in the zygotes.

Next, we examined whether there is any difference in the Ca²⁺ pool size between zygotes and parthenogenetic eggs. The results in Fig. 6 show that 10 μM ionomycin treatment, which is sufficient to deplete intracellular Ca²⁺ stores (Jones et al., 1995b), released more Ca²⁺ from zygotes than from parthenogenetic eggs. This suggests that Ca²⁺ pool size is also unlikely to be a determinant to restrict sperm-factor- induced Ca²⁺ oscillations in the zygotes.

The present study was undertaken to determine the specificity of action of the sperm factor in triggering Ca²⁺ oscillations in mammalian eggs. Our major finding is that the ability of mammalian eggs to generate sperm- factor-induced Ca²⁺ oscillations depends on their previous history: once exposed to the sperm extracts, no Ca²⁺ oscillations are induced when the eggs are given a second injection of the sperm extracts during meiosis or first mitosis. Our results suggest that the sperm-factor- induced Ca²⁺ oscillations are mediated by a maternal machinery that is abolished once the oocytes and eggs are activated by the sperm or sperm extract injection.

There are two lines of evidence supporting the idea that the putative maternal machinery is functional only once in mouse oocytes and eggs. First, we have shown that once the egg is fertilized or injected with the sperm extracts, its Ca²⁺ oscillation-generating ability to response to sperm extract injection is diminished. Second, the MII eggs would fail to generate sperm-factor-induced Ca²⁺ oscillations if they had already been stimulated by the sperm extracts during oocyte maturation.

An interesting question is that, if the maternal machinery indeed functions only once during meiosis and first mitosis, why do endogenous Ca²⁺ transients, as shown by Kono et al. (1996), occasionally occur in mouse zygotes at NEBD and during the first mitosis? It has been reported that, in four B6CBF1 mouse fertilized eggs detected from just after NEBD, three eggs showed at least one Ca²⁺ transient while in one of them four spikes were generated with a frequency of 30-40 minutes (Kono et al., 1996). However, Tombes and colleagues (1992) showed that only a single Ca²⁺ transient was detected in ICR mouse zygotes at NEBD and during the mitosis (Tombes et al., 1992). We have also examined Ca²⁺ signaling at NEBD and during first mitosis and found that our ICR mouse zygotes also generated only a single Ca²⁺ spike at NEBD and no additional Ca²⁺ transients were detected after NEBD and during the first mitosis (our unpublished data). Our result is consistent with that of Tombes et al. (1992), but not that of Kono et al. (1996).

At present, the signal and precise mechanism underlying Ca²⁺ signaling at NEBD in mammalian zygotes is unknown (Kono et al., 1996). We speculate that the various patterns of Ca²⁺ signaling at NEBD and during the first mitosis in mouse zygotes may reflect the inactivation states of the maternal machinery. We propose that, if the machinery is fully inactivated, the zygotes may generate only a single Ca²⁺ transient; if it is partially inactivated, then multiple Ca²⁺ transients could be generated. Moreover, if the machinery is not inactivated at all after egg activation, the M-phase eggs, upon stimulation with the sperm extracts during the first mitosis, will undergo fertilization-like Ca²⁺ oscillations (as shown in Figs 2D, 3B).

Our finding that the machinery is functional only once during meiosis/first mitosis suggests that inactivation of this machinery by the sperm extracts occurs in mammalian oocytes and eggs in a non-regenerative manner. It appears that this feature of inactivation is very similar to NAADP-induced desensitization of Ca²⁺ release in sea urchin eggs and homogenates (Aarhus et al., 1996; Genazzani et al., 1996, 1997). NAADP, nicotinic acid adenine dinucleotide phosphate, has been identified as an endogenous molecule that releases Ca²⁺ via a mechanism independent of InsP₃ receptors and ryanodine receptors in sea urchin egg and homogenates (Lee and Aarhus, 1995; Chini et al., 1995; Chini and Dousa, 1995; Genazzani et al., 1996), and in ascidian oocytes (Albrieux et al., 1998). The most striking feature of NAADP-induced Ca²⁺ release is that it can be selectively inactivated by subthreshold concentrations of this molecule, which per se do not cause Ca²⁺ release (Aarhus et al., 1996; Genazzani et al., 1996). In a recent report, it has been shown that pretreatment of ascidian oocytes with NAADP inhibited the fertilization-induced Ca²⁺ oscillations (Albrieux et al., 1998), suggesting that NAADP-induced Ca²⁺ release mechanism plays an essential role in determining sperm-induced Ca²⁺ oscillations. So far, there is only one report showing that NAADP can mobilize Ca²⁺ release in mammalian cells and affect the patterns of hormone-induced Ca²⁺ signaling (Cancela et al., 1999). It will be interesting to find out in future experiments whether the NAADP-sensitive Ca²⁺ signaling pathway is present and functional in mammalian oocytes and eggs, and whether pretreatment of mammalian eggs with NAADP inhibits fertilization-induced Ca²⁺ oscillations in a manner similar to that seen in ascidian oocytes. It is possible that inactivation of the putative maternal machinery by the sperm extracts reflects functional changes in NAADP-induced Ca²⁺ releasing mechanism after fertilization. Future studies should test this interesting possibility.

The present study has revealed a number of them. (1) The machinery is already present and functional in the maturing oocytes (Fig. 4A). (2) The function of the machinery is not only fertilization-status dependent, but also cell cycle dependent (Table 1). This is shown by our finding that the machinery is active in the metaphase but not in the interphase parthenogenetic eggs. This observation is in agreement with the finding that interphase marks a period of insensitivity to sperm-induced Ca²⁺ changes (Jones et al., 1995a). (3) Parthenogenetic agents, although able to trigger mammalian egg activation and embryonic development (Kline and Kline, 1992; Jones and Whittingham, 1996), is unable to inactivate this machinery, suggesting that the machinery is probably inactivated selectively by the sperm. (4) The machinery can be inactivated by the sperm extracts in a cell-cycle-independent manner, because Ca²⁺ oscillatory response to the second injection is not obtained whenever the first sperm extract injection was conducted at metaphase or interphase stage (Figs 2A, 3A, 4A). (5) Inactivation of the machinery in the zygotes is not due to changes in either InsP₃ receptor sensitivity to InsP₃ or Ca²⁺ pool size (Figs 5, 6).

The results of the present study suggest that the factor responsible for inactivating the machinery is of sperm origin. It seems to be protein-based, because injection of the heat- inactivated extracts into mouse oocytes and eggs did not abolish the cells’ ability to generate Ca²⁺ oscillations when microinjected with the functional sperm extracts at the metaphase (Figs 3B, 4B). Since the sperm contains a cytosolic protein factor that can triggers Ca²⁺ oscillations in mammalian eggs (Swann and Lai, 1997), it is possible that this factor itself may also serve as the trigger for inactivating the maternal machinery. However, because at present the functional sperm factor responsible for inducing Ca²⁺ oscillations has not yet been purified, we cannot distinguish between the sperm factor itself inactivating the machinery and the sperm providing an additional factor(s) for the inactivation.

In conclusion, our results of the present study suggest that the orderly sequence of Ca²⁺ oscillations induced by sperm or sperm extract injection require a functional maternal machinery. During meiosis and first mitosis, this machinery functions only once in mammalian oocytes and eggs and is inactivated by the sperm extracts in a non-regenerative manner. The nature of this putative machinery, its precise mechanism of inactivation, and the sperm-derived factor(s) responsible for this inactivation remain to be determined. We consider that answers to these questions are essential for understanding the precise molecular mechanism of sperm-induced Ca²⁺ oscillations in mammalian species.

This work was support by the Chinese Academy of Sciences, the National Natural Science Foundation of China and the Rockefeller Foundation. We thank members of our laboratory for helpful discussion. Tang and Dong are PhD student trainees supported by the ‘Contraception-21’ Program of the Rockefeller Foundation. F.-Z. S. and X.-Y. H. devote this paper to Dr Robert M Moor, FRS, who retired from his scientific post at the Babraham Institute, Cambridge in September 1999.