Sperm-triggered calcium oscillations during meiosis in ascidian oocytes first pause, restart, then stop: correlations with cell cycle kinase activity

Sperm-triggered calcium (Ca²⁺) oscillations occur during meiosis in many species of oocyte: mammal, ascidian, starfish, mollusc, nemertean and annelid (Cuthbertson and Cobbold, 1985; Speksnijder et al., 1989; Stricker, 1995, 1996; Deguchi et al., 1996; Eckberg and Miller, 1995). In both mammalian and ascidian oocytes, Ca²⁺ oscillations are the trigger that initiates embryonic development (Miyazaki et al., 1993; Swann and Ozil, 1994; McDougall and Sardet, 1995). The culmination of evidence suggests that there are at least two factors that trigger and maintain these Ca²⁺ oscillations. Sperm trigger these Ca²⁺ oscillations but they are maintained by the metaphase-like cytoplasm of the oocyte.

The identity and mode of action of the sperm factor that triggers these Ca²⁺ oscillations is not known with certainty. There are however three hypotheses. These are the Ca²⁺ bomb hypothesis (Jaffe, 1991), the receptor hypothesis (Jaffe, 1990) and the soluble sperm factor hypothesis (Whitaker and Swann, 1993). The observation that sperm cytosolic extracts faithfully mimic the sperm-triggered Ca²⁺ oscillations argues in favour of a soluble sperm factor model of fertilisation (Swann, 1990).

Likewise the exact nature of the maternal factor that maintains the sperm-triggered Ca²⁺ oscillations is not known with certainty. It is known, however, that the Ca²⁺ oscillations in mouse oocytes are maintained during the meiotic metaphase-like state and terminate as the fertilised oocyte enters interphase (Jones et al., 1995). In addition, when the metaphase-like state is prolonged artificially the sperm-triggered Ca²⁺ oscillations are also prolonged for as long as the metaphase-like state persists (Jones et al., 1995).

The metaphase state is established by the kinase activity of maturation/mitosis promoting factor (MPF) and is characterised by nuclear envelope breakdown, chromatin condensation and the formation of a mitotic spindle (reviewed by Doree and Galas, 1994). MPF is composed of a kinase (Cdk1: Cell division kinase 1) and a regulatory cyclin B subunit (Nurse and Thuriaux, 1980; Lohka et al., 1988; Gautier et al., 1988; Dunphy et al., 1988; Evans et al., 1983). Metaphase ends when several proteins, including cyclin B, are targeted via ubiquitination for proteolytic destruction by the anaphase promoting complex (APC) (Murray et al., 1989; King et al., 1995; Irninger et al., 1995; Shirayama et al., 1998).

The meiotic cell cycle of oocytes is different from the mitotic cell cycles in that the metaphase-like organisation of the microtubules and chromosomes does not follow precisely variations in MPF activity as it does in mitosis (Verlhac et al., 1994; reviewed by Murray, 1998). Instead, the metaphase-like organisation of the chromatin, the microtubules and the nuclear envelope follows variations in mitogen-activated protein (MAP) kinase activity (Verhac et al., 1994; Moos et al., 1996). MAP kinase is the family name for a number of Ser/Thr protein kinases (Cobb et al., 1991; Payne et al., 1991; Seger et al., 1991). During meiotic maturation in clam (Shibuya et al., 1992), Xenopus (Ferrel et al., 1991) and mouse (Verlhac et al., 1994) MAP kinase is stimulated. With the exception of clam, where the MAP kinase activity is only transient, MAP kinase activity persists during the completion of meiosis (Ferrel et al., 1991; Verlhac et al., 1994). These persistently elevated MAP kinase levels are responsible for the CSF (cytostatic factor) activity that arrests mammalian oocytes at metaphase II (MII) by reducing the rate of cyclin B destruction (Colledge et al., 1994; Hashimoto et al., 1994; Verlhac et al., 1996). In Xenopus oocytes, the sperm-triggered Ca²⁺ signal overrides the CSF activity of the MAP kinase pathway and initiates cyclin B destruction followed by the loss of the upstream activator of MAP kinase Mos (Lorca et al., 1993). It is not yet known how CSF or the sperm-triggered Ca²⁺ signals interact with the APC to affect cyclin B ubiquitination and cyclin B destruction. However, Ca²⁺ signals have been shown to increase the activity of the 26 S proteasome that degrades ubiquitinated cyclin B at two points during the meiotic cell cycle in fertilised ascidian oocytes (Kawahara and Yokasawa, 1994).

Fertilisation of metaphase I-arrested (MI) ascidian oocytes triggers two phases of Ca²⁺ oscillations that accompany meiosis (Speksnijder et al., 1989). Recently it has been shown that the second phase of Ca²⁺ oscillations in ascidian oocytes is required to reactivate the MPF activity that precedes extrusion of the second polar body (Russo et al., 1996). However, the finding that a monotonic Ca²⁺ signal leads to the formation of a MII spindle suggests that the MPF activity rises again even in the absence of further Ca²⁺ signals (Sensui and Morisawa, 1996). To resolve this issue, we have directly measured the MPF activity in oocytes that are activated by triggering a monotonic Ca²⁺ signal to determine whether subsequent Ca²⁺ signals are required to reactivate MPF during meiosis. In addition to this, we have tested the hypothesis that the sperm-triggered Ca²⁺ oscillations are regulated by either MAP kinase or MPF activity during meiosis.

The tunicate Ascidiella aspersa was collected locally from North East England. The animals were kept at 10°C in the laboratory. Oocytes and sperm were obtained by removing the animals from their outer tunic and puncturing first the gonoduct then the sperm duct. The oocytes were transferred to 4.5 ml Millipore-filtered natural sea water at 19°C and the sperm were stored dry at 4°C. All further experiments were performed with this sea water. The oocytes were partially dechorionated by incubating them with 0.1-0.2% trypsin at room temperature for 30-45 minutes. Following incubation with trypsin, the oocytes were aspirated gently once or twice to remove their chorion mechanically. Once the chorion was removed the oocytes were kept in 1% gelatin/formaldehyde-coated Petri dishes. All further glass and plastic ware that the oocytes come into contact with was coated with this mixture of gelatin/formaldehyde.

Oocytes were bathed in the vital DNA dye Hoechst 33342 (10 μg/ml) (Sigma) for 10-15 minutes then returned to sea water. The microtubules were visualised by microinjection of Rhodamine-labelled tubulin (Cytoskeleton). The chromatin and microtubules were visualised using a Leica confocal microscope (40×/n.a. 1.0 oil) and by using an epifluorescence microscope (Olympus IX70, 60×/n.a. 1.2 water objective).

Oocytes were transferred to a wedge based on the design by Keihart (1982). Micropipettes were pulled on a Kopf 720 puller from GC100-T10 glass (Clarke Electromedical). The micropipettes were advanced towards the oocytes using a Narishige hydraulic three-way micromanipulator. The micropipettes were tip filled then inserted into the oocytes. The oocytes were pressure injected using a Narishige IM-300.

Intracellular Ca2+ levels were recorded by microinjecting fura 2 dextran (10 kDa; Molecular Probes) to give an intracellular concentration of between 10 and 20 μM. About 30 minutes after injection, the oocytes were observed with an epifluorescence Olympus IX70 microscope (40×/n.a. 0.9 air objective). The exciting light passed through a filter wheel (Newcastle Photonics) housing 340 nm and 380 nm band-pass filters and a 450 nm dichroic before reaching the injected oocytes. The emitted light collected by the objective lens was filtered through a 510 nm band-pass filter and entered a photomultiplier housed on the microscope (9124 A, Thorn EMI). The analogue signal was digitised and counted using a photon-counting card (Newcastle Photonics).

The MBP (Myelin Basic Protein) kinase activity and histone H1 kinase activity were assayed as described previously (Verlhac et al., 1994). Samples of 5 oocytes were collected by first washing the oocytes through 1 M glycine three times to remove the sea water (this does not alter meiotic progression). For one series of experiments, the oocytes were first microinjected with InsP3 then harvested in the same manner. The oocytes were then removed in a volume of 2 μl and transferred to 8 μl reaction buffer (25 mM Hepes, 80 mM β-glycerophosphate, 5 mM EGTA, 10 mM MgCl2, 1 mM DTT, 10 μg/ml leupeptin/pepstatin/aprotinin, 0.2 mM AEBSF, 1 mM benzamidine, 100 μM NaVO4, 5 mM NaF, pH 7.2). At this point, the oocytes were snap-frozen in liquid nitrogen. After defrosting the samples on ice, which is sufficient to lyse the dechorionated oocytes, 2 μl 6× reaction mixture was added to the lid of each Eppendorph (0.9 mg/ml myelin basic protein (Sigma) or histone H1 (Sigma type III from calf thymus), 0.6 mM ATP, 0.5 mCi/ml[32P]ATP, 60 μM cAMP-dependent protein kinase inhibitor in 1× Reaction Buffer at pH 7.2). The reaction was started synchronously by spinning the Eppendorphs. They were then transferred to a water bath at 30oC for 10 minutes. Following this, the reaction was stopped synchronously by adding 2× sample buffer to the lid of each Eppendorph and spinning in a microfuge. The samples were then heated to 95°C for 3 minutes and resolved on 15% polyacrylamide gels. The resolved proteins on the gel were placed in a Phosphorimager (Fujix; 1500 Bas Reader) and the incorporation of [32P] measured quantitatively. The in gel kinase assay was based on the method of Shibuya and colleagues (1992). Briefly, 30 oocytes were collected at each time point in sample buffer and subjected to 15% PAGE. The kinase reaction, renaturation and washing steps were carried out as previously described (Shibuya et al., 1992). The incorporated radioactivity was measured using a Phosphorimager. These experiments were carried out in duplicate where one gel was prepared without the substrate myelin basic protein present. As a positive control and as a marker of molecular weight, we dissolved 2 μl active p42 rat ERK2 (Upstate Biotechnology) in sample buffer and ran this alongside the ascidian oocyte samples.

The histone H1 kinase activity is elevated in the unfertilised oocyte and falls each time a polar body is extruded. This is in agreement with previous data from a different species of ascidian (Russo et al., 1996). 5 minutes after fertilisation, the histone H1 kinase activity is low and remains at these low levels for a minimum period of 5 minutes (Fig. 1A). The histone H1 kinase activity is elevated again 15 minutes after fertilisation and finally falls to low levels 25 minutes after fertilisation (Fig. 1A).

The MBP kinase activity is also elevated in the unfertilised oocyte (Fig. 1B). Unlike the histone H1 kinase activity, the MBP kinase activity increases and reaches peak levels 10 minutes after fertilisation (Fig. 1B). The MBP kinase activity subsequently declines and reaches low levels 25 minutes after fertilisation (Fig. 1B). To determine the identity of the MBP kinase, we performed an ‘in gel’ kinase assay. The MBP kinase activity migrates at around 42 kDa (Fig. 1C). The autophosphorylation was less than 5% of the signal. The overall pattern of activity matches the pattern obtained in cell lysates (Fig. 1C).

The state of the chromatin correlates with the activity of either MPF or MAP kinase: it remains in a metaphase-like configuration during the entire period of meiosis after fertilisation (Fig. 2A). The chromatin finally decondenses at the end of meiosis 25-30 minutes after fertilisation (Fig. 2A). The decondensed chromatin constrained by the pronuclear membrane is visible below the strongly stained polar bodies 30 minutes after fertilisation (Fig. 2A).

Likewise, the state of the microtubules correlates with the activity of either MPF or MAP kinase: they remain in a metaphase-like configuration throughout the entire period of meiosis following fertilisation (Fig. 2B). Extrusion of both polar bodies is depicted (Fig. 2B). These events correlate to the times that MPF activity is low. When the first polar body is being extruded, the sperm aster becomes visible but does not reach an interphase size (Fig. 2B). The sperm aster increases dramatically in size minutes after extrusion of the second polar body (Fig. 2B).

Sperm trigger two series of Ca²⁺ oscillations in A. aspersa oocytes that correlate to the period when MPF activity is elevated.

The first series of Ca²⁺ oscillations lasts for approximately 5 minutes and terminates 2-3 minutes before extrusion of the first polar body (Fig. 3A). After the first series of Ca²⁺ transients normally a clear gap phase of between 4 and 6 minutes ensues(Fig. 3A). After the gap a second series of Ca²⁺ oscillations begins that lasts for between 10 and 15 minutes (Fig. 3A).

Ascidians display three easily recognisable characteristics that we have used to achieve cell cycle synchrony. The peak of the first Ca²⁺ transient triggers the first visible morphological event – the cortical contraction. The cortical contraction begins when the first Ca²⁺ signal has peaked in Phallusia mammillata oocytes (Roegiers et al., 1995). This is also the case in A. aspersa oocytes. For this reason, this time point is taken as zero and is subsequently referred to as the point of fertilisation (Fig. 3A). In this way, we have synchronised the Ca²⁺ data and the kinase data with the initiation of the cortical contraction. We have used the two other morphological characteristics to check cell cycle synchrony. These are extrusion of the first polar body (7-9 minutes postfertilisation) and extrusion of the second polar body (24-28 minutes postfertilisation (Fig. 3B)). A few minutes after this time, the male pronucleus becomes discernible near the centre of the enlarging sperm aster (Fig. 3B).

It is therefore possible to directly check cell cycle synchrony by observing these three morphological indicators and thus compare the pattern of sperm-triggered Ca²⁺ oscillations to the activities of both MPF and MAP kinase with a high degree of fidelity.

The second series of Ca²⁺ signals are proposed to reactivate the MPF activity that accompanies formation of the MII spindle (Russo et al., 1996). However, the observation that a single Ca²⁺signal can result in the formation of a MII spindle suggests that further Ca²⁺ signals are not necessary to form a MII spindle (Sensui and Morisawa, 1996). In addition, it is not known whether a single Ca²⁺ signal results in the reactivation of MPF. For this reason, we have directly measured the MPF activity following a single Ca²⁺ signal triggered by microinjection of the Ca²⁺-releasing second messenger InsP₃. This has the same effect as microinjection of Ca²⁺ buffers that transiently raise intracellular Ca²⁺ to around 1 μM (Sensui and Morisawa, 1996). Microinjection of inositol (1,4,5)-trisphosphate (InsP₃) at pipette concentrations that ranged from 25 to 200 μM consistently activated the oocytes and triggered extrusion of the first polar body, but did not trigger extrusion of the second polar body (n=47). Following extrusion of the first polar body, the chromatin remains condensed in a metaphase-like state 30 minutes after injection of InsP₃ when it would normally be decondensed following fertilisation (Fig. 4A). The microtubules also remain in a metaphase-like state 37 minutes after injection: a horizontal MII spindle forms adjacent to the plasma membrane (Fig. 4B). The MPF activity falls following injection of InsP₃ and the first polar body is extruded (Fig. 4C). The MPF activity then increases to elevated levels 20 minutes after microinjection (Fig. 4C). Interestingly, the MPF remains at elevated levels 30 minutes after microinjection when the MPF activity in control oocytes has fallen (Fig. 4C). These data show that MPF will reactivate following a single Ca²⁺ signal. In addition, they reveal that the MPF activity remains at elevated levels for a prolonged period of time. These oocytes seem to be arrested at MII.

Sperm-triggered Ca²⁺ oscillations are maintained by the metaphase-like cytoplasm of the oocyte (Jones et al., 1995). In ascidians, these Ca²⁺ oscillations are temporally associated with MPF activity (Figs 1A, 3A). We therefore tested whether we could experimentally modulate the sperm-triggered Ca²⁺oscillations by modulating the MPF activity. To do this, we exploited the finding that MII-arrested oocytes can be fertilised. To create MII oocytes, we microinjected InsP₃ (50μM pipette concentration, 0.1-0.2% injected). This triggers a monotonic Ca²⁺ signal (Fig. 4Di).

Fertilisation of these MII-arrested oocytes triggered only one series of Ca²⁺ oscillations followed by extrusion of the second polar body (Fig. 4Dii). Fertilisation of control oocytes from the same batch displayed the two phases of Ca²⁺ oscillations that normally accompany fertilisation (Fig. 4Diii). These data indicate that the sperm-triggered Ca²⁺ oscillations can be entrained by the oocyte cytoplasm and again reveal that the sperm-triggered Ca²⁺ oscillations correlate temporally with elevated MPF activity.

We were next interested to determine the nature of the mechanism responsible for inducing the pause that separates the two series of Ca²⁺ oscillations. During this period, the first polar body is extruded. Two possibilities are either that the oocyte calcium release system becomes desensitised or that the signal generating the Ca²⁺ oscillations becomes desensitised or depleted (in the case of the substrate). The results indicate that the Ca²⁺ signalling system is not desensitised in mid meiosis when the Ca²⁺ oscillations pause. The threshold pipette concentration of InsP₃ that triggers a Ca²⁺ signal is 2 μM (Fig. 5A). InsP₃ at pipette concentrations of 2 μM also triggered a significant Ca²⁺ release 7-9 minutes after fertilisation (Fig. 5B). These data suggest that the Ca²⁺ oscillations are likely to pause because the signal generating them becomes inactive or the substrate depleted during the period when the first polar body is extruded.

In addition to this the Ca²⁺ signalling system does eventually become desensitised at the end of meiosis when the Ca²⁺ oscillations stop. Higher concentrations of 10 μM InsP₃ are required to trigger a Ca²⁺ signal following exit from meiosis (Fig. 5C).

The sperm-triggered Ca²⁺ oscillations during fertilisation of mouse oocytes correlate temporally with the metaphase-like state and terminate just before pronuclear formation (Jones et al., 1995). This metaphase-like state is maintained by MAP kinase activity and not MPF activity (Verlhac et al., 1994). These observations would suggest that the MAP kinase-induced metaphase-like state entrains the sperm-triggered Ca²⁺ oscillations. In addition, pronuclei from fertilised mouse oocytes are capable of triggering Ca²⁺ oscillations when microinjected into unfertilised mouse oocytes (Kono et al., 1995). If the soluble sperm factor hypothesis is the correct view of how sperm trigger Ca²⁺ oscillations at fertilisation (Whitaker and Swann, 1993), these data suggest the possibility that this soluble factor is either sequestered in or is associated with the pronuclei. This may account for the termination of the sperm-triggered Ca²⁺ oscillations when the oocyte exits meiosis and pronuclei form. We have used the ascidian oocyte to determine how the sperm-triggered Ca²⁺ oscillations are modulated during the meiotic cell cycle.

Ascidian oocytes, like the oocytes from numerous other species, display Ca²⁺ oscillations at fertilisation (Sardet et al., 1998). Unlike mammalian oocytes, the Ca²⁺ oscillations triggered at fertilisation of A. aspersa oocytes can be separated into two distinct series; the first series of Ca²⁺oscillations precede extrusion of the first polar body and the second series of Ca²⁺ oscillations precede extrusion of the second polar body (Fig. 3A). The frequency of the Ca²⁺ oscillations varies from batch to batch. For this reason, we will not discuss the obvious variation in spike frequency that sometimes distinguishes phase I Ca²⁺ oscillations from phase II ones. The most obvious distinguishing feature is the gap.

The sperm-triggered Ca²⁺ oscillations are present when theMPF activity is elevated and pause when the MPF activity is low (see model 1). However the sperm-triggered Ca²⁺ oscillations do not seem to bear a temporal correlation with the MAP kinase activity which is elevated when the Ca²⁺ oscillations pause and start again (see model 1). In addition to this, the Ca²⁺ oscillations pause during interkinesis when the oocyte remains in a metaphase-like state: pronuclei are not present.

It is worth noting that the oocyte remains in a metaphase-like state during the period that the MPF activity is low – monitored by direct visualisation of the state of the chromatin, microtubules and pronuclei. These data add support to the idea that elevated MAP kinase activity maintains the metaphase-like configuration of the oocyte cytoplasm when the MPF activity is low during interkinesis (Verlhac et al., 1994). Interestingly, in addition to the chromatin and the microtubules, we find that the sensitivity of the Ca²⁺ signalling system also correlates with the metaphase-like state (Fig. 5). It is not yet clear why this should be the case, but it is intriguing to note that the morphology of the endoplasmic reticulum (ER) changes dramatically during meiotic progression – large islands of ER form throughout the completion of meiosis (Speksnijder et al., 1993).

By inhibiting the second phase of sperm-triggered Ca²⁺ oscillations in Ciona intestinalis oocytes with the Ca²⁺ chelator BAPTA, the reactivation of MPF that accompanies formation of the MII spindle is inhibited (Russo et al., 1996). These data imply that the second phase of Ca²⁺ oscillations is required to reactivate MPF prior to extrusion of the second polar body. However, a single increase in Ca²⁺ triggers extrusion of the first polar body followed by formation of a MII spindle (Sensui and Morisawa, 1996). These data suggest that the MPF activity initially decreases then increases even in the absence of further Ca2+ signals. We have triggered a single Ca²⁺ signal in order to determine whether MPF reactivates following extrusion of the first polar body. Our data show that a monotonic Ca²⁺ signal leads first to the inactivation then the reactivation of MPF activity (Fig. 4C). This is consistent with the finding that a monotonic Ca²⁺ signal leads to the formation of a MII spindle (Sensui and Morisawa, 1996). These data may be reconciled with the previous finding (Russo et al., 1996) when it is noted that high concentrations of BAPTA (circa 5 mM) have nonspecific effects that are not associated with its Ca²⁺ -chelating ability (Pethig et al., 1989).

To determine whether the metaphase cytoplasm of the oocyte can entrain the sperm-triggered Ca²⁺ oscillations, we compared the pattern of Ca²⁺ oscillations between two types of oocyte: those released from MI and those released from MII at fertilisation. If the Ca²⁺ oscillations are entrained by the cell cycle position, we would predict that the second phase of Ca²⁺ oscillations would be absent when MII-arrested oocytes are fertilised. Fertilisation of MII-arrested oocytes would not allow sufficient time to generate the second phase of Ca²⁺ oscillations during meiosis since the second polar body is extruded 7-9 minutes after fertilisation. Moreover, if the second phase of Ca²⁺ oscillations occurred they would be present during interphase.

Our data reveal that fertilisation of MII-arrested ascidian oocytes triggers the first phase of Ca²⁺ oscillations but not the second (Fig. 4Dii). These data are depicted in model 2 below.

Having established that the Ca²⁺ oscillations bear a close temporal correlation with the MPF activity, we tested two hypotheses. One, that the MPF activity sensitises the generation of the Ca²⁺ oscillations and two, that the MPF activity modulates the sensitivity of the Ca²⁺ signalling system. To distinguish between these two possibilities, we tested the sensitivity of the Ca²⁺ signalling system during the time that the Ca²⁺ oscillations pause. The second of these two possibilities seemed more likely since Ca²⁺ influx is required to sustain the second phase of Ca²⁺ oscillations in fertilised ascidian oocytes (Arnoult et al., 1996). We might therefore have predicted that the pause represents the time required to refill the depleted Ca²⁺ stores. Indeed Ca²⁺ store depletion has been demonstrated to desensitise the affinity of the InsP₃ receptor for its agonist InsP₃ (Iino and Endo, 1992; Nunn and Taylor, 1992). We therefore tested the sensitivity of the Ca²⁺-releasing second messenger InsP₃. It should be noted that only the InsP₃ receptor is involved in generating the second phase of Ca²⁺ oscillations in fertilised ascidian oocytes (McDougall and Sardet, 1995; Russo et al., 1996; Albrieux et al., 1997).

To determine whether the sperm-triggered Ca²⁺ oscillations pause because the Ca²⁺ signalling system becomes desensitised, we injected InsP₃ into ascidian oocytes before, during and after meiosis. Our data reveal that the Ca²⁺ signalling system remains equally sensitive during the period that the sperm-triggered Ca²⁺ oscillations pause (Fig. 5B), and finally becomes desensitised when the Ca²⁺ oscillations have stopped and the oocyte exits meiosis (Fig. 5C).

These data indicate that the Ca²⁺ oscillations pause even when the Ca²⁺ release system remains sensitive. The most likely explanation of these data is that the metaphase cytoplasm (containing active MPF) of the oocyte sensitises either the paternally supplied trigger that initiates the Ca²⁺ oscillations or modulates the level of the substrate (phosphatidyl inositol bis phosphate) that is cleaved to generate InsP₃. In addition to this, the Ca²⁺ signalling system is sensitised during the entire metaphase-like state. These two events together would help generate and sustain a Ca²⁺ oscillation during both meiotic metaphases in ascidian oocytes and so ensure rapid meiotic exit.

We would like to thank Daniel Perez-Monogiovi and Evelyn Houliston for help with the kinase assays. We would also like to thank Christian Sardet, Evelyn Houliston, Rada Philipova and Michael Whitaker for useful discussions. This work was supported by a grant from the Wellcome Trust (Ref. No. 051540/Z/97/PMG/LB).