Fertilisation and thimerosal stimulate similar calcium spiking patterns in mouse oocytes but by separate mechanisms

The fertilisation of the mammalian oocyte is followed by a series of repetitive Ca²⁺ spikes (Cuthbertson et al., 1981; Igusa and Miyazaki, 1986; Miyazaki et al., 1986; Kline and Kline, 1992a). These Ca²⁺ spikes cause the release of cortical granules to induce the block to polyspermy, the reactivation of meiosis to initiate development (Kline and Kline, 1992b; Vincent et al., 1992) and, in the hamster, a series of corresponding hyperpolarising transients due to increases in K⁺ conductance (Miyazaki and Igusa, 1981).

Two hypotheses have been advanced to explain how the spermatozoon might stimulate Ca²⁺ release. A sperm-associated molecule may interact with a receptor coupled to a G-protein in the oocyte membrane to activate phospholipase C, generate inositol 1,4,5-trisphosphate (InsP₃) and thereby mobilise Ca²⁺ (Miyazaki, 1990; Jaffe, 1990). Activation of a G-protein in hamster oocytes by addition of 5-hydroxytryptamine (5-HT) or guanosine-5′-O-(3-thiotriphosphate) (GTPγS), or by the injection of InsP₃, does induce a series of Ca²⁺ pulses, although their form differs from that reported at fertilisation (Miyazaki, 1988; Swann et al., 1989; Miyazaki et al., 1990). Moreover, both sperm and GTPγSinduced Ca²⁺ transients are blocked by preinjection of guanosine-5′-O-(2-thiodiphosphate) (GDP_βS; Miyazaki, 1988). However, this mechanism is unlikely to play a role in fertilisation because it is inhibited by phorbol ester, which has little effect on fertilisation (Swann et al., 1989; Miyazaki et al., 1990). Recently, a sperm ligand that is likely to bind to an oocyte integrin-like receptor has been described (Blobel et al., 1992; Tarone et al., 1993), but it is not clear that such a system would function through G-protein activation. Alternatively, the fertilising spermatozoon may introduce a soluble Ca²⁺-releasing messenger into the oocyte cytoplasm after sperm:oocyte fusion (Dale et al., 1985; Stice and Robl, 1990; Swann, 1990).

The Ca²⁺ transients seem to be derived from internal Ca²⁺ stores since they can occur in the absence of external Ca²⁺ (Igusa and Miyazaki, 1983; Miyazaki, 1988), as can parthenogenetic activation via Ca²⁺ ionophore (Vincent et al., 1992). Depletion of Ca²⁺ stores by treatment with thapsigargin leads to reduction or inhibition of spikes (Kline and Kline, 1992a). However, refilling of internal pools by external Ca2+ is required; extended periods in Ca2+-free medium lead to loss of Ca²⁺ transients. Transients may result from the sensitisation of a Ca²⁺-induced Ca²⁺ release (CICR) system (Miyazaki, 1991; Swann, 1991). Recently, the selective sulphydryl reagent thimerosal has been shown to increase the sensitivity of CICR in unfertilised hamster (Swann, 1991) and mouse (Swann, 1992) oocytes, and its effects are blocked or reversed by dithiothreitol (DTT; Swann, 1992).

In this paper, we compare the Ca²⁺ transients induced in the mouse oocyte at fertilisation with those induced by thimerosal. We propose that, while the latter may activate Ca²⁺ spiking by enhancing the sensitivity of InsP₃ receptors through a redox mechanism that is inhibited by DTT, the spiking that occurs after fertilisation is initiated by a separate mechanism, which is maintained or even accelerated by DTT. The origin of the DTT sensitivity of the spiking mechanism is investigated. The results show that DTT is able to increase the permeability of the oolemma to divalent cations in both fertilised and unfertilised oocytes and is also able to enhance known activators of Ca²⁺ spiking in oocytes. We suggest that DTT might enhance the transduction event that normally occurs at fertilisation to initiate both the influx of external Ca²⁺ and the onset of CICR. The nature of this transduction event and its potential constituent components are discussed.

MF1 female mice (3-4 weeks; OLAC, Bicester, UK) and F₁ female mice (C57BL/10ScSn/Ola female × CBA/Ca/Ola male, bred in the laboratory) were superovulated by intraperitoneal injection of 5 or 10 i.u. of pregnant mare’s serum gonadotrophin (PMS; Intervet, Cambridge, UK) followed 48 hours later by 5 or 10 i.u. of human chorionic gonadotrophin (hCG; Intervet). Unfertilised oocytes (12-13 hours post-hCG) were released from the oviduct into warmed medium H6+4 mg/ml bovine serum albumin (BSA; a Hepes buffered form of modified T6 medium, Nasr-Esfahani et al., 1990). Cumulus cells were removed by brief exposure to hyaluronidase (0.1 mM; Sigma) and zonae pellucidae removed by exposure to alpha-chymotrypsin (0.001%; Sigma type II; Vincent et al., 1992). Oocytes were held in drops of H6+BSA under paraffin oil (FSA Laboratories, Loughborough, UK) in Falcon tissue culture dishes. All manipulations were carried out at 37°C on heated stages, pads or in incubators.

Spermatozoa were expelled from the vas efferentia and cauda epididymides of male CFLP mice into 1 ml of Whittingham’s medium (Whittingham, 1971) containing 30 mg/ml BSA and incubated under oil for 1-3 hours at 37°C and 5% CO₂ to capacitate. Aliquots of 40-100 μl were taken for insemination of oocytes. Oocytes were fertilised either by insemination of zona-free oocytes attached to the base of chambers on the warmed microscope stage (see below) or, when a longer interval between fertilisation and analysis of Ca2+ spiking was desired, in drops of T6+BSA medium under oil. These latter oocytes were then transferred to chambers.

Activation of oocytes parthenogenetically was achieved in three ways. (i) After a 5 to 15 minute incubation in H6 + BSA free of Ca2+, oocytes were exposed to Ca2+ ionophore (ionomycin; 5 μM) in Ca2+-free H6+BSA for 2 minutes. Full activation was confirmed by observation of polar body extrusion (Vincent et al., 1992). As DMSO was used as a solvent in the ionophore stock solution, an equivalent dilution of the highest concentration (0.1 or 0.25% DMSO) was used in control groups. After each treatment, oocytes were washed 5 times over a 30 minute period. The first three washes were performed in Ca2+-free medium. (ii) Alternatively, after a 30 minute incubation in H6 + BSA containing 1.2 mM Ca2+, oocytes were exposed to a solution of H6+BSA free of Ca2+ and phosphate but containing SrCl2 ± DTT (at various concentrations as recorded under Results). (iii) In some experiments, after a 30 minute incubation in a solution of H6 + BSA containing 1.2 mM Ca2+, oocytes were exposed to 50 μM acetylcholine in the same solution.

Microtubules and chromosomes in oocytes were stained in specially designed chambers as described in Maro et al. (1984). Cells were fixed at 37°C for 30 minutes in 3.7% formaldehyde in PBS in the presence of 0.5% Triton X-100 (Sigma) for extraction and were washed in phosphate-buffered saline (PBS). β-Tubulin was visualized with a rat monoclonal anti-tubulin antibody (Kilmartin et al., 1982) followed by rhodamine-labelled anti-rat IgG. Chromosomes were stained by incubation in Hoechst dye 33342 (10 μg ml-1 in PBS) for 30 minutes.

Cortical granule release was visualised by a procedure derived from that of Cherr et al. (1988). Oocytes were fixed in 3% paraformaldehyde in PBS for 30 minutes and then washed extensively in a blocking solution of 1 mg ml-1 BSA, 100 mM glycine and 0.2% sodium azide in PBS. To visualise exclusively the content of the cortical granules after extrusion, oocytes were not permeabilized. Oocytes were incubated in 10 μg ml-1 Lens culnaris agglutinin conjugated to fluorescein isothiocyanate (FITC-LCA; United States Biochemical Corporation) in blocking solution for 15 minutes and then washed extensively in the blocking buffer.

8-10 zona-free oocytes were washed and transferred to H6 + polyvinylpyrrolidone (PVP; 6 mg ml-1) on a coverslip that had been precoated with concanavalin A (Con A; 0.2 mg ml-1 in PBS) and which formed the base of a metallic perfusion chamber (Moreton, 1991). Oocytes were then loaded with fura-2 ace toxymethlyester (2 μM; Molecular Probes) for 20 to 30 minutes and washed extensively with H6+PVP. The chamber was then placed in a well on the stage of a Nikon Diaphot TMD inverted epi-fluorescence microscope for imaging. Solutions were introduced via a system of continuous perfusion through the chamber maintained at 37°C.

Intracellular free divalent cation activity was imaged through a Nikon CF-Fluor 20× objective and intensified CCD camera (Extended ISIS, Photonic Science, Robertsbridge, UK), by calculating the ratio of fura-2 fluorescence at 510 nm, excited by UV light alternately at 340 and 380 nm from twin Xenon arc lamps and grating monochromators. Excitation wavelengths were alternated by a rotating chopper mirror attached to a stepper-motor, which was driven in synchrony with the video signal from the camera, to switch wavelengths at the end of each video frame. The resulting video signals were combined by an ‘Imagine’ digital image processor (Synoptics Ltd., Cambridge, UK) using a lookup table to implement the formula of Grynkiewicz et al. (1985). The calculation was done in real time, to give a ‘live’ image of intracellular Ca2+ concentration ([Ca2+]_i) which was updated every 80 msec, and smoothed by recursive filtering with a 200 msec time-constant to reduce the noise (for further details see O’Sullivan, 1989; Moreton, 1991). When both Ca2+ and Sr2+ were present in the incubation, the relative contribution of the two ions to changes in fluorescence was not established.

Mn2+ entry was followed by imaging the resulting quench in fura-2 fluorescence at 510 nm at a constant excitation wavelength of 360 nm. At this wavelength, fura-2 fluorescence is independent of [Ca²⁺]i but fluorescense is still quenched by Mn²⁺ (Hallam et al., 1988).

In all cases, the live image was recorded continuously on video tape, and subsequently played back and re-digitised into a framestore, using software written in the semper language (Synoptics Ltd.) to sample selected oocytes and to record and plot either mean [Ca²⁺]i or fluorescence readings at regular time intervals. In most cases, data were sampled at 4 or 5 seconds intervals, but for more detailed analysis sampling could be speeded up to 0.8 seconds (eg. Fig. 2).

To determine whether fura-2 loaded oocytes had been fertilised, the coverslip was examined in the chamber under bright field for evidence of polar body extrusion. The coverslip was then removed, fixed in 4% formaldehyde for 20-40 minutes, rinsed in PBS, stained with Hoechst 33258 (5 mg ml-1; Sigma) for 10 minutes and mounted in Citifluor for assessment of sperm entry.

Acetylcholine, dithiothreitol, ionophore (ionomycin), thimerosal, CaCl₂, MnCl₂ and SrCl₂ were all from Sigma. Fura-2 AM was from Molecular Probes.

The Ca²⁺ responses of mouse oocytes to sperm and thimerosal are illustrated in Fig. 1. Following the insemination of spermatozoa into the chamber, the initiation of Ca²⁺ spiking in the first oocyte began after a delay of 7-15 minutes (Fig. 1A; n=63). The subsequent pattern of Ca²⁺ spikes varied considerably between oocytes and, in most experiments, some oocytes remained unfertilised; the Ca²⁺ levels in these served as controls for those in fertilised oocytes. In general, the first transient was broader and had a larger amplitude than the subsequent spikes as has been described by others (Kline and Kline, 1992a; Swann, 1992). In some oocytes, there was a gradual reduction in spike amplitude with time, as recorded in previous experiments on mouse (Cuthbertson et al., 1981) and hamster (Igusa et al., 1983) oocytes.

Addition of thimerosal to unfertilised oocytes produced a series of Ca²⁺ spikes which also usually began with a much broader spike (Fig. 1B,C). The threshold for the stimulatory action of thimerosal was approximately 1 μM (Fig. 1D), and both the amplitude and initial frequency of Ca²⁺ spiking were maximal between 10 and 20 μM (Fig. 1C). Further raising the thimerosal concentration to 100 μM had no marked effect on these initial parameters, but did accentuate the tendency for spikes to accelerate and the baseline to drift upwards (Fig. 1B). A characteristic feature of each spike is the gradual pacemaker rise in Ca²⁺ that precedes the rapid rising phase (see asterisk in Fig. 1C).

The characteristics of the individual spikes induced by fertilisation and thimerosal were remarkably similar with regard to their amplitude, width, rate of rise and rate of recovery (Table 1). In both cases, the rate of rise was always faster than the rate of recovery and, on average, each spike lasted for about 50 seconds with a relatively stable plateau which ended abruptly with the sudden onset of the recovery phase. As has been recorded previously for hamster oocytes (Miyazaki et al., 1986), the first few transients following fertilisation appeared as waves in that the Ca²⁺ signal initiated at a focal point on one side of the oocyte and then spread to the opposite side at a rate of 47.5±6.06 μm sec^-1 (n=10). This spatial organization is shown in Fig. 2A for the onset of the first Ca²⁺ transient following fertilisation. Subsequent chromosomal staining revealed that these waves initiated from the site of sperm entry. The first Ca²⁺ transient following thimerosal stimulation also appeared as a wave originating from a focal point but, with the resolution of our imaging system (5 ratio images sec^-1; see methods), the Ca²⁺ gradients did not appear as steep as those observed following fertilisation (Fig. 2B). For example, as the Ca²⁺ wave spreads after fertilisation, an approximate 10-fold difference in [Ca²⁺]_i between the initiation site and the opposing pole of the oocyte can be detected, as judged by the false-colour (Fig. 2A, 5.6 seconds). In contrast, thimerosal-evoked Ca²⁺ waves appeared more diffuse and such large intracellular Ca²⁺ gradients were not observed (Fig. 2B). Thimerosal-evoked Ca²⁺ waves may reflect regional variations in the sensitivity of second messenger mediated Ca²⁺ release (Rooney et al., 1991).

Ca²⁺ spiking in response to either fertilisation or thimerosal was very sensitive to changes in the concentration of external Ca²⁺. Removal of Ca²⁺ abolished spiking whereas an increase of Ca²⁺ from the normal level of 1.2 mM to 2.4 or 12 mM resulted in acceleration (data not shown).

In most oocytes, the spiking behaviour induced by fertilisation continued over the periods recorded (up to 3.5 hours post-insemination) regardless of whether spermatozoa were removed. By contrast, the thimerosal-induced spikes could be reversed if thimerosal was washed out of the bathing solution. In the experiment shown in Fig. 3A, removal of thimerosal (100 μM) at the end of the first spike resulted in the appearance of one additional spike, which was a consistent feature of the six oocytes analysed, followed by a complete cessation of the Ca²⁺ spikes observed normally in the continuous presence of this agent (cf. Fig. 1B with Fig. 3A). Upon the re-addition of thimerosal, a regular spiking pattern was restored rapidly (Fig. 3A), with no evidence of the initial, very broad spike observed normally when oocytes were first stimulated with thimerosal.

Thimerosal is a thiol reagent which probably acts by oxidizing sulphydryl groups, a process that can be reversed by dithiothreitol (DTT). The addition of DTT (8 mM) to oocytes stimulated in 20 μM thimerosal stopped the Ca²⁺ spiking immediately (Fig. 3C; n=10). Upon removal of DTT, regular spiking was restored after a delay. Note that the addition of DTT caused a small reversible depression of the baseline level of Ca²⁺ (Fig. 3C). In contrast to this inhibitory effect on thimerosal-induced Ca²⁺ spiking, DTT was found consistently to accelerate the spikes that develop following fertilisation (Fig. 3B; n=17). During the treatment with DTT, the period between spikes was approximately 100 seconds, which is very much shorter than the normal spike interval recorded following fertilisation (Fig. 1A). The acceleratory effect of DTT was reversible (Fig. 3B).

Fertilisation is associated with a number of changes to the organisation of the oocyte, any or all of which might influence the DTT responsiveness of Ca²⁺ spikes. The incorporation of the fusing spermatozoal membrane into the oolemma and the fusion of the cortical granules with the oolemma both result in a mosaic membrane structure with a composition differing from that prefertilisation (Wolf and Ziomek, 1983; Shapiro et al., 1981). The activation of the arrested meiotic state through the completion of metaphase to interphase is associated with changes in membrane properties, including the profile of functional ion channels in the oolemma (Day et al., 1990, 1991). DTT might, for example, stimulate Ca²⁺ influx through newly recruited ion channels. In an attempt to determine why DTT has a different effect on oocytes depending on whether they have been fertilised or treated with thimerosal, we examined whether or not thimerosal could induce some of the fertilisation events such as cortical granule release and oocyte activation.

Two techniques were used to assess whether or not cortical granule release occurred. A functional assay was provided by the acquisition of resistance by the zona pellucida to digestion by chymotrypsin so called zona hardening. The data in Table 2 show that exposure to concentrations of thimerosal adequate to stimulate a train of Ca²⁺ spikes also causes zona hardening. A direct test for evidence of cortical granule discharge is to stain the exocytosed contents of the granules on the surface of the thimerosal-treated oocyte. Fig. 4 confirms that cortical granules are indeed discharged. Of 24 control oocytes, only 3 (12%) showed evidence of cortical granule exocytosis, whilst of 40 oocytes exposed to 100 μM thimerosal for 5 minutes followed by a 10 to 30 minute wash, 39 (98%) showed exocytosis. These results make it unlikely that cortical granule exocytosis is responsible for the stimulatiory effect of DTT on the fertilisation-induced train of spikes.

Although thimerosal induced cortical granule release, it did not activate oocytes. Thus, no polar body extrusion was observed in any of 34 control oocytes or any of 68 oocytes exposed to thimerosal (20-100 μM) for 30 minutes followed by a recovery period of 2 hours. Examination of oocytes immunocytochemically for tubulin and chromatin after exposure to 100 μM thimerosal for 11 to 40 minutes revealed that the spindle had been destroyed completely in 46/50 oocytes and severely disrupted in the remaining 4 oocytes. The oxidation of tubulin thiol groups is known to interfere with tubulin polymerisation and to be reversed by DTT (Kuriyama and Sakai, 1974). On neutralisation of the thimerosal with DTT, followed by an incubation in control medium, 30/41 oocytes had regenerated typical metaphase spindles (Fig. 5A) although in 11 of them the chromosomes were slightly dispersed from the metaphase plate. Of the remaining 11 oocytes, 10 had slightly abnormal spindles (Fig. 5C) and one lacked a spindle (Fig. 5B). An intact spindle is required for transition from M-phase to interphase (Maro et al., 1986), which therefore cannot occur during the period of exposure to thimerosal. Is it possible then that the passage from M-phase to interphase, causes fertilised oocytes to become sensitive to the stimulatory effect of DTT? In an attempt to separate the reactivation of meiosis from other effects of sperm entry, oocytes were exposed to either spermatozoa (n=10) or Ca²⁺ ionophore (n=10), which activates development parthenogenetically by a single phasic release of Ca²⁺ from internal stores as well as inducing cortical granule release (Vincent et al., 1992). After culture for 1-2 hours to allow polar body extrusion and passage to interphase, the oocytes were loaded with fura-2. Both groups of oocytes, together with unfertilised oocytes, were then exposed to thimerosal to induce a train of spikes (Fig. 6). Superimposition of DTT resulted in the arrest of the spiking pattern in the unfertilised (Fig. 6C) and ionophore-activated (Fig. 6A) oocytes, but not in the fertilised oocytes (Fig. 6B). Thus, some change had apparently occurred in fertilised activated oocytes which was not evident in parthenogenetically activated oocytes, suggesting that the transition to interphase was not responsible for the changed response to DTT. It seemed possible that the fertilising sperm itself might introduce a component that conferred DTT sensitivity. Alter natively, a mechanism responsive to DTT might reside within the oocyte and be sensitised by sperm entry.

Activation by ionophore does not elicit the periodic Ca²⁺ spiking seen after sperm activation. An alternative approach to oocyte activation and the transition to interphase is to expose oocytes to Sr²⁺ (Fraser, 1987; Kline and Kline, 1992a), which yields spikes of divalent cations and thus more closely resembles fertilisation. We found that 8 mM Sr²⁺ did not induce a train of spikes in unfertilised oocytes (n=18). However, superimposition of 8 mM DTT resulted in an immediate prolonged elevation of fluorescent signal (Fig. 7A; n=10). Conversely, oocytes that failed to respond to 8 mM DTT always responded to a subsequent addition of Sr²⁺, again giving a maintained elevation of fluorescence (Fig. 7B; n=10). Sr²⁺ and DTT were acting in strict synergy and, by reducing their concentrations, it was possible to find a combination (4 mM Sr²⁺ and 1 mM DTT) that could reproduce a train of oscillations that resembled that seen at fertilisation in that the first spike was broader than subsequent spikes, and the spike frequency (and sometimes the amplitude) tended to fall with time (Fig 7C; n=61). All oocytes that showed a pulsatile response to Sr²⁺ or a prolonged elevation in their fluorescence also activated. In contrast, oocytes that were exposed to DTT in the presence of external Ca²⁺ concentrations up to 12 mM showed no pulsatile response and did not activate (Fig. 8A,B). In each instance, DTT elicited only a slight elevation of the basal intracellular Ca²⁺ concentration.

An alternative way of activating mammalian oocytes is to stimulate Ca²⁺ release via a receptor coupled G-protein (Miyazaki, 1988; Miyazaki et al., 1990). Acetylcholine (50 μM) triggered a train of Ca²⁺ spikes in mouse oocytes (Fig. 9; n=20). Superimposition of DTT consistently caused an acceleration in the train of spikes, reminiscent of its effect on the spikes following fertilisation (n=10).

Thus, in these experiments, the DTT sensitisation was evident when Sr²⁺ or acetylcholine substituted for a fertilising spermatozoon. These results suggest that a capacity to respond to DTT resides in the oocyte and does not require a sperm for it to be revealed. What is the nature of this capacity and where in the oocyte does it reside?

One possible effect of DTT might be to stimulate the influx of external Ca²⁺ across the oolemma. A more rapid refilling of internal stores would be expected to increase the frequency of Ca²⁺ spikes. It has been proposed but not demonstrated that fertilisation is accompanied by an increased influx of Ca²⁺ (Igusa and Miyazaki, 1983). To test the possibility that DTT increases divalent cation influx, unfertilised oocytes were loaded with fura-2 and placed in Mn²⁺ (0.1 and 0.5 mM), which also enters via divalent cation channels but quenches the fura-2 signal (Grynkiewicz et al., 1985). The fluorescence due to Mn²⁺ was recorded at a continuous excitation wavelength of 360 nm at which fura 2 emission is independent of [Ca²⁺] (Hallam et al., 1988).

After establishing the basal rate of quench, DTT was added and the percentage increase in the Mn²⁺ quench rate calculated as 49±2 % (mean ± s.e.m. for 0.1 mM Mn²⁺; n=5) and 50±10 % (mean ± s.e.m. for 0.5 mM Mn²⁺; n=5) (Fig. 10A,B). Thus, DTT was indeed able to increase the rate of divalent cation influx into the oocytes.

The possibility that increased cation influx stimulated by DTT is responsible for the accleration in the fertilisation induced train of spikes was examined more directly (Fig. 10C,D). Oocytes initially fertilised in Ca²⁺ showed the familiar train of spikes when Ca²⁺ was monitored by ratioing the fura-2 fluorescence at 340/380 nm (Fig. 10C; n=10). In a continuation of the same experiment, Ca²⁺ was replaced with 0.1 mM Mn²⁺ and the influx of Mn²⁺ was followed by monitoring the quench of fura-2 fluorescence at 360 nm. There was a rapid basal rate of quench (Fig. 10D). Superimposition of DTT resulted in a further increase in the rate of quench and the percentage increase was calculated as 76±37 % (mean±s.e.m. for 0.1 mM Mn; Fig. 10D). The DTT-induced acceleration of the fertilisation train might thus be explained by an enhanced influx of Ca²⁺ across the oolemma.

The response of mouse oocytes to sulphydryl reagents is paradoxical in that stimulation of Ca²⁺ spiking was observed following either oxidation (by thimerosal) or reduction (by dithiothreitol). The stimulatory effect of thimerosal on mouse oocytes was reversible and could also be inhibited by the reducing agent DTT. In contrast, when DTT was combined with other stimuli, it enhanced spiking, accelerating the spiking in oocytes activated by a spermatozoon or acetylcholine and acting synergistically with Sr²⁺ to induce spiking in quiescent oocytes. Unlike the other stimuli, thimerosal arrests oocytes in metaphase preventing completion of meiosis and progression into interphase. However, we have shown the arrested cell cycle does not account for the differential sensitivity of the oocytes to DTT since spikes induced by thimerosal in ionophore-activated, interphase oocytes are not accelerated by DTT. Moreover, DTT has stimulatory effects on cation influx in both fertilised (interphase) and unfertilised (meiotic) oocytes.

We have attempted to resolve the paradoxical response to sulphydryl reagents by proposing that thimerosal and DTT may be acting at two separate sites, which might correspond to the two processes that are thought to be activated at fertilisation. These two processes are, firstly, an increase in Ca²⁺ entry across the plasma membrane, which has been inferred from the observation that fertilisation caused the membrane potential to hyperpolarise (Igusa and Miyazaki, 1983). The second process, a ten-fold increase in the sensitivity of the internal Ca²⁺ stores to Ca²⁺-induced Ca²⁺ release (CICR), was established by injecting pulses of Ca²⁺ of increasing amplitude to measure the triggering thresholds required in unfertilised and fertilised oocytes (Igusa and Miyazaki, 1983). Thus, the model proposed originally by Igusa and Miyazaki (1983) stressed the importance of the linkage of continuous Ca²⁺ influx to periodic release of intracellular Ca²⁺ via a process of CICR (Igusa and Miyazaki, 1983, 1986; Swann, 1991).

We argue below that thimerosal probably acts by mimicking the sensitisation of the internal release mechanism, whereas the stimulatory effects of DTT seem to depend, at least in part, upon the stimulation of Ca²⁺ influx events occurring at the plasma membrane.

We have shown that the spatiotemporal characteristics of the thimerosal-induced Ca²⁺ signal in mouse oocytes are remarkably similar to those of the signal at fertilisation, suggesting that this thiol reagent is activating the same internal release mechanism by increasing the sensitivity of CICR (Swann, 1991; Miyazaki et al., 1992a; Carrol and Swann, 1992). The nature of the internal stores responsible for this regenerative release of Ca²⁺ has yet to be established, likely candidates for the receptors being those for ryanodine (Swann, 1992) or InsP₃ (Miyazaki et al., 1992b). The evidence for ryanodine receptors has come from studies on mouse oocytes, where ryanodine was found to release Ca²⁺ and appeared to enhance the sensitivity of oocytes to Ca²⁺ (Swann, 1992). The sensitivity of ryanodine receptors might be regulated by cyclic ADP ribose (cADPR) since the latter was able to potentiate Ca²⁺ release due to ryanodine in sea urchin eggs (Galione et al., 1991; McDougall et al., 1993). The Ca²⁺ spiking induced by thimerosal may result from a similar sensitisation of ryanodine receptors which have been shown to be potentiated by thiol reagents (Salama et al., 1992a,b).

Alternatively, an InsP₃ receptor may be involved in the sensitisation of hamster oocytes to Ca²⁺ injections (Swann, 1991; Carrol and Swann, 1992). Injection of InsP₃ into mammalian eggs triggers an all-or-none regenerative release of Ca²⁺ (Miyazaki, 1988; Peres, 1990; Peres et al., 1991), which might depend on the process of CICR as first proposed by Igusa and Miyazaki (1983). Studies on other cells have revealed that Ca²⁺ functions as a co-agonist on the InsP₃ receptor and thus contributes to an all-or-none release process through a positive feedback effect (Iino, 1990; Finch et al., 1991; Bezprozvanny et al., 1991; Iino and Endo, 1992). Moreover, studies on several cell types have shown that thiol oxidation can enhance Ca²⁺ release (Missiaen et al., 1991; Rooney et al., 1991; Bootman et al., 1992; Miyazaki et al., 1992a) and stimulate Ca²⁺ waves (Rooney et al., 1991). Since this sensitisation can be blocked by the InsP₃ receptor antibody 18A10, it has been argued that the regenerative release of Ca²⁺ by thimerosal can be brought about by InsP₃ receptors (Miyazaki et al., 1992b). However, although it seems likely that thimerosal acts by oxidation of critical sulphydryl groups in either InsP₃ or ryanodine receptors to mimick the Ca²⁺ spiking that occurs at fertilisation, an oxidation-dependent mechanism cannot be involved normally, as fertilisation-induced spikes continue and are accelerated in the presence of DTT.

Unlike thimerosal, DTT is not able to induce spiking by itself but it does have the remarkable capacity of enhancing all the known activators of Ca²⁺ spiking in mouse oocytes except thimerosal. Thus, addition of Sr²⁺ to mouse oocytes initiates repetitive spiking after a long delay (approximately 20 minutes; Kline and Kline, 1992b). The mode of action of Sr²⁺ is unknown but it may act as a surrogate in triggering CICR (Cognard and Raymond, 1985). Studies on frog skeletal muscle suggest that “the relative affinity of the internal Ca²⁺releasing stores appears to be Ba²⁺>Sr²⁺>Ca²⁺” (Cognard and Raymond, 1985). The ability of Sr²⁺ to induce spiking was greatly potentiated by DTT through a mechanism that is associated with an enhanced divalent entry component as monitored by Mn²⁺ quenching of fura-2.

Indeed, stimulation of divalent cation entry is a consistent feature of the action of DTT both in unfertilised and fertilised oocytes. This increased rate of divalent cation entry would also account for the acceleration seen when DTT was superimposed on the train of spikes induced by either acetyl choline or a spermatozoon. Whether the stimulation of divalent entry is due to some direct action of DTT on a plasma membrane Ca²⁺ channel or due to the production of a second messenger, such as InsP₃, which in turn stimulates influx is unclear. However, the fact that DTT can act synergistically with Sr²⁺ to initiate a train of spikes indicates that this reducing agent can enhance the transduction event that leads to divalent cation influx and the onset of CICR.

Mechanisms linking sperm-oocyte interaction to the activation of Ca²⁺ spiking might also be sensitive to DTT. For example, the sperm may act via integrin receptors (Blobel et al., 1992; Tarone et al., 1993) or CD4-like molecules (Mori et al., 1992) on the oocyte to generate second messengers such as InsP₃. InsP₃ formation via integrin receptors may involve tyrosine kinase-linked mechanisms (Kornberg et al., 1991) coupled to phospholipase C-γ1 (Whitaker and Swann, 1993), a process known to induce Ca²⁺ oscillations in human neutrophils (Jaconi et al., 1991). Murine oocytes have a CD4/p56lck transducing system very similar to that found in T-cells (Mori et al., 1992) in which the oxidation of sulphydryls by N-ethylmaleimide inhibited the association between CD4 and pp56lck and the tyrosine phosphorylation of PLC-γ1 (Kanner et al., 1992). Thus, DTT may act by keeping critical cysteine residues reduced so that these proteins can interact with each other (Turner et al., 1990; Shaw et al., 1990). Inhibition of v-src by herbimycin A can be reversed by DTT, again suggesting that free SH-groups are essential for signal transduction through these non receptor tyrosine kinases (Vehara et al., 1989). Although the oocyte contains high levels of reducing power in the form of glutathione (10-20 mM; Calvin et al., 1986; Boerjan and de Boer, 1990; Nasr-Esfahani and Johnson, 1992), its use may require the activity of specific reducing enzymes, as this requirement has been found for the reduction of critical SH-groups involved in transcription factor-DNA and steroid hormone-receptor interactions (Xanthoudakis et al., 1992; Grippo et al., 1985). DTT may mimick the action of such enzymic factors to increase reducing conditions nonspecifically and thereby enhance the signalling events responsible for generating second messengers such as InsP₃.

The experiments in this paper are consistent with the hypothesis originally proposed by Igusa and Miyazaki (1983) that Ca²⁺ spiking in oocytes depends upon the activation of two processes. Firstly, the case for an obligatory sensitisation of CICR for successful oocyte spiking is overwhelming, and this sensitisation can be achieved in a number of ways. Under normal conditions (or following acetylcholine stimulation), this sensitisation is probably mediated by a second messenger, of which InsP₃ or cADPR are likely candidates. Alternatively, the release channels (whether ryanodine or InsP₃) can be sensitised by oxidation of critical cysteine residues as occurs during the action of thimerosal or by the action of Sr²⁺ with its higher affinity for releasing internal Ca²⁺ stores. The second process is an increased influx of divalent cations, which may occur either in parallel with the increased sensitisation of CICR or may be linked to the latter, as might occur if the emptying of Ca²⁺ stores stimulates a secondary capacitative influx mechanism (Putney, 1986). Depletion of internal stores by thapsigargin enhances Ca²⁺ entry into mouse oocytes (Kline and Kline, 1992a), suggesting that such a capacitative mechanism exists.

The paradoxical action of the reagents DTT and thimerosal can be accommodated on the basis of the hypothesis that InsP₃ functions as a second messenger during fertilisation (Miyazaki et al., 1992b). The interaction between sperm and oocyte may initiate the production of InsP₃ through a mechanism that is enhanced by DTT. An increase in InsP₃ is then responsible for activating the two processes necessary to initiate Ca²⁺ spiking. It may act at the cell surface to promote the entry of external Ca²⁺ through a mechanism that remains to be defined. This enhanced entry of Ca²⁺ might then serve to sensitise the internal stores by virtue of increasing their Ca²⁺ content. However, such a sensitisation by Ca²⁺ seems unlikely, because DTT failed to induce spiking in unfertilised oocytes even when exposed to high concentrations of external Ca²⁺. Alternatively the sensitisation of CICR may be achieved by InsP₃ diffusing in from the cell surface. Whatever the mechanism is, we expect that the signalling system will depend upon reduced sulphydryl groups to account for the stimulatory action of DTT.

Many thanks are due to Martin George and Brendan Doe for technical assistance. This work was supported by an AFRC grant to M. J. B., and a programme grant from the MRC to M. H. J. and Professor P. Braude. T. R. C. is a Royal Society University Research Fellow.