In vertebrates, a rise in intracellular free Ca2+ (Ca2+i) levels during fertilization initiates second metaphase (mII) exit and the developmental programme. The Ca2+ rise has long been considered to be crucial for development, but verifying this contribution would benefit from defining its role during fertilization. Here, we delineate the role of Ca2+ release during mII exit in wild-type mouse eggs and show that it is dispensable for full-term development. Exit from mII can be induced by Zn2+-specific sequestration without Ca2+ release, eliciting Cyclin B degradation in a manner dependent upon the proteasome pathway and intact microtubules, but not accompanied by degradation of the meiotic regulator Emi2. Parthenogenotes generated by Zn2+ sequestration developed in vitro with normal expression of Ca2+-sensitive genes. Meiotic exit induced by either Ca2+ oscillations or a single Ca2+ rise in oocytes containing a signaling-deficient sperm resulted in comparable developmental rates. In the absence of Ca2+ release, full-term development occurred ∼50% less efficiently, but at readily detectable rates, with the birth of 27 offspring. These results show in intact mouse oocytes that Zn2+ is essential for mII arrest and suggest that triggering meiotic exit is the sole indispensable developmental role of Ca2+ signaling in mammalian fertilization.
Fertilizable vertebrate oocytes are typically arrested in the second meiotic metaphase (mII) by the cytostatic factor Emi2, which sustains mII until fertilization by preventing the anaphase promoting complex (APC) E3 ubiquitin ligase from associating productively with its co-activator, Cdc20 (Stricker, 1999; Schmidt et al., 2005; Peters, 2006; Shoji et al., 2006). The principal effect of this association is to impede the ubiquitylation and consequent proteasomal destruction of the separase inhibitor securin and of cyclin B (CycB), an essential component of maturation promoting factor (MPF), which is responsible for mII arrest (Gautier et al., 1990; Peters, 2002; Peters, 2006). Thus, Emi2 sustains mII by stabilizing CycB, and, in Xenopus, is itself removed following Ca2+-dependent ubiquitylation by the Skp-Cullin-Fbox E3 ubiquitin ligase SCFTrcpb (Tung et al., 2005). A model of the Emi2 regulatory network is presented elsewhere (Perry and Verlhac, 2008).
Emi2 possesses a functional paralogue, Emi1, that prevents premature APC activation in early mitosis (Reimann et al., 2001a). Emi1 inhibits APC activity via separable APC- and IBR/TRIAD/C6HC Zn2+-binding domains resembling the Emi2 C terminus; mutation of the Emi1 zinc-binding region (ZBR) converts Emi1 into an APC substrate (Schmidt et al., 2005; Miller et al., 2006). A single ZBR mutant of Xenopus Emi2 lacks cytostatic activity in cell-free extracts, but its stability has not been reported (Schmidt et al., 2005).
Fertilization in metazoans is characterized by Ca2+ release from inositol 1,4,5-trisphosphate [IP3; Ins(1,4,5)P3] receptor-sensitive oocyte stores to enable cytoplasmic signaling (Mazia, 1937; Whitaker and Irvine, 1984; Miyazaki et al., 1992) (for a review, see Runft et al., 2002). The resultant increase in intracytoplasmic `free' Ca2+ (Ca2+i) concentration is considered to be a universal requisite among animals for initiating both meiotic exit and the events of oocyte activation that presage full-term development (Runft et al., 2002; Yanagimachi, 1994; Ducibella and Fissore, 2008). Sperm-oocyte union at fertilization in Xenopus induces a single [Ca2+i] rise lasting 5 minutes (Kline and Nuccitelli, 1985; Runft et al., 2002). In the mouse, a relatively large initial [Ca2+i] increase is followed by oscillatory increases that spike every 5-15 minutes until pronucleus formation ∼4 hours later (Igusa et al., 1983; Cuthbertson and Cobbold, 1985; Runft et al., 2002).
Ca2+ signaling in Xenopus fertilization results in phosphorylation of Emi2 to target it for Plk1-mediated phosphorylation and proteolysis, and hence relieve mII arrest (Lorca et al., 1993; Runft et al., 2002; Rauh et al., 2005; Schmidt et al., 2005). This signal is relayed in diverse species via the Ca2+-dependent kinase, calmodulin kinase II (CaMKII) (Lorca et al., 1993); injection of mouse mII oocytes with cRNA encoding constitutively active CaMKII results in meiotic progression (Knott et al., 2006) and native CaMKII activity in newly fertilized mouse eggs shadows [Ca2+i] oscillations (Markoulaki et al., 2004). Moreover, the events of oocyte activation are impaired by depleting the γ isoform of CaMKII (CaMKIIγ) or prevented in animals carrying a targeted deletion of the CaMKIIγ gene (Camk2g) (Chang et al., 2009; Backs et al., 2010). The roles of Ca2+ release and CaMKIIγ are different in oocyte activation in that CaMKIIγ is not directly required for cortical granule exocytosis or maternal mRNA recruitment. However, both these processes do require Ca2+ release and/or meiotic progression (Backs et al., 2010).
A large and growing body of evidence links the [Ca2+i] change at fertilization to early mammalian development (Runft et al., 2002; Ducibella and Fissore, 2008). Blocking [Ca2+i] increase prevents mouse mII exit in response to oocyte-activating stimuli (Kline and Kline, 1992) and discrete activation events are driven by differential responses to [Ca2+i] oscillation number and the sum of their duration (Ducibella et al., 2002; Ozil et al., 2005). Fertilization-induced [Ca2+i] elevation also regulates key downstream processes (Rogers et al., 2006; Ozil et al., 2006; Ducibella and Fissore, 2008). The absence or premature termination of [Ca2+i] oscillations causes marked downregulation of genes expressed at the eight-cell and blastocyst stages (Rogers et al., 2006), while stimulating [Ca2+i] oscillation frequency by electrical hyperstimulation causes upregulation of (fewer) genes; both affected peri- and post-implantation development (Ozil et al., 2006). Thus, there is evidence of short- and long-term developmental roles for Ca2+i mobilization during mammalian fertilization and it is possible that this reflects a requirement for CaMKIIγ activity. Activation of the Ca2+-dependent protein phosphatase calcineurin (PP2B) is essential for mII exit in Xenopus (Mochida and Hunt, 2007; Nishiyama et al., 2007), raising the possibility of additional mouse oocyte activation pathways with downstream developmental roles. The study of these pathways would benefit from isolating the role of Ca2+ release per se during fertilization.
We explored this possibility as part of our interest in signaling in mII exit and here report that Zn2+ is essential for mouse mII arrest: Zn2+ depletion induces Ca2+-independent mII exit without Ca2+ release. We harnessed this observation to explore the developmental role of fertilization-induced Ca2+ release and show that, in the presence of a paternal genome, oscillatory, monotonic or zero Ca2+ rises can support healthy full-term development. This work introduces a pivotal cellular role for Zn2+ in meiotic homeostasis and suggests that the sole indispensable developmental role of fertilization-induced Ca2+i release is to induce mII exit.
MATERIALS AND METHODS
Collection, culture and activation of oocytes
Mice were supplied by SLC (Shizuoka-ken, Japan) and handled according to institutional guidelines. Eight- to 12-week-old B6D2F1 females were superovulated using standard serial intraperitoneal injections of pregnant mare serum gonadotropin (PMSG) followed 48 hours later by human chorionic gonadotropin (hCG). Oviductal metaphase II (mII) oocytes were collected typically 12 to 15 hours post-hCG injection and cumulus cells removed following hyaluronidase treatment as previously described (Yoshida et al., 2007a).
`Conventional' activation of mII oocytes was with SrCl2 or ethanol. For SrCl2 activation, oocytes were incubated in Ca2+-free CZB (Chatot et al., 1989) supplemented with 10 mM SrCl2 in humidified CO2 [5% (v/v) in air] at 37°C for 1 to 6 hours. Oocytes were washed in KSOM and incubation continued at 37°C. Activation with 7% (v/v) ethanol in HEPES-buffered CZB (CZBH) was at room temperature for 5 minutes, after which oocytes were washed in KSOM and incubated in humidified CO2 [5% (v/v) in air] at 37°C. Activation by exposure to N,N,N′,N′-tetrakis-(2-pyridylmethyl)-ethylenediamine (TPEN; Sigma-Aldrich, MO, USA) was for 45 minutes in KSOM containing TPEN at the concentration indicated (typically 100 μM), in humidified CO2 [5% (v/v) in air] at 37°C. Following TPEN treatment, oocytes were washed in KSOM (three or four times), then either incubated in KSOM or subjected to imaging to determine relative [Ca2+i]; the period between TPEN wash-out and the initiation of imaging was 60 to 120 seconds. In the experiments shown in Fig. 3D, cRNA-injected mII oocytes were exposed to TPEN for 45 minutes and then to ethanol, as described above. Where appropriate, oocytes were incubated in nocodazole (Sigma) for 1.5 hours before exposure to 10 mM SrCl2 or 100 μM TPEN for 1 hour, and washing and continued incubation in KSOM for 6 hours.
To obtain diploid parthenogenotes, activation media and KSOM for subsequent incubations were supplemented with 5 μg/ml cytochalasin B (MP Biomedicals LLC, OH, USA) for 6 hours from the initiation of activation. Parthenogenotes were then washed and incubated in KSOM lacking cytochalasin B.
Transition metal ion chelation and rescue assays
To assess the effect on mII oocytes of treatment with TPEN, we selected those whose Pb1 had degraded so that Pb2 extrusion could be visualized more clearly. Oocytes collected 12-15 hours post-hCG were placed in KSOM containing TPEN. TPEN exhibits the following Kds for divalent transition metal ions specifically assessed in this work: Zn2+, 1015.58 M–1; Fe2+, 1014.61 M–1; Cu2+, 1020 M–1; Mn2+, 1010.27 M–1. Binding constants for TPEN and other transition metal ion-binding reagents used here are shown in Table S1 in the supplementary material. TPEN was stored in aliquots at –20°C as a 10 mM stock solution in DMSO and freshly diluted in KSOM immediately prior to incubation. Oocytes were typically examined for Pb2 extrusion 90 minutes after TPEN exposure. In some experiments, oocytes had been injected with cRNAs 4-5 hours prior to KSOM/TPEN incubation. Where appropriate, TPEN-containing media was supplemented with 10 μM epoxomycin (Biomol International, PA, USA) or 20 μM Z-LLL-CHO (Peptide Institute, Saito, Japan), which are (respectively) irreversible and reversible proteasome inhibitors. Transition metal ion rescue assays were performed either in parallel with, or sequential to, TPEN incubation. For rescue in parallel, oocytes lacking a Pb1 were incubated for 1.5 hours in KSOM containing 100 μM TPEN together with 100 μM ZnSO4 (Zn2+), MnSO4 (Mn2+), CuSO4 (Cu2+) or FeSO4 (Fe2+), as appropriate. Oocytes were then washed in KSOM and scored for Pb2 extrusion 1.5 hours later. For sequential rescue, oocytes were cultured for 2 hours in KSOM containing 100 μM TPEN plus 20 μM Z-LLL-CHO, washed in KSOM/20 μM Z-LLL-CHO alone (to remove TPEN) and transferred to KSOM/20 μM Z-LLL-CHO supplemented with 100 μM ZnSO4 (Zn2+), MnSO4 (Mn2+), CuSO4 (Cu2+), FeSO4 (Fe2+), NiSO4 (Ni2+) or CoSO4 (Co2+), as appropriate and incubated for 1 hour before washing and continuing the culture for 3 hours in KSOM alone. Scoring was of Pb2 extrusion. To determine the effect of chelating non-Zn2+ transition metal ions, mII oocytes were incubated for 1.5 to 3.0 hours in KSOM containing 100 μM p-aminosalicylic acid (MP Biomedicals, CA, USA) for Mn2+ chelation, or ammonium tetrathiomolybdate (Sigma) for Cu2+ chelation or 2,2′-bipyridine (2,2′-dipyridyl; Sigma) for Fe2+ chelation. Meiotic progression was scored by the appearance of a Pb2.
Preparation and injection of cRNA and siRNA
Cloning Emi2 to generate cRNA-encoded mCherry fusions has been described elsewhere (Shoji et al., 2006). Mouse Cyclin B1 with a C-terminal Venus fusion was generated by inserting an XhoI-XbaI Venus fragment into plasmid pCI-neo (Promega Corp., WI, USA). A Cyclin B1 PCR product was inserted into this construct [using primers (5′ to 3′): CTAGCTAGCACCATGGCGCTCAGGGTCAC and CTGCTCGAGCCATGCCTTTGTCACGGCC] as an NheI-XhoI fragment.
cRNAs were synthesized in vitro from linear plasmid DNA template and 5′-capped and polyadenylated in the same reaction using an mSCRIPT™ mRNA Production System (Epicentre Biotechnologies, WI, USA) according to the manufacturer's instructions. cRNAs were dissolved in nuclease-free distilled water, quantified and stored in aliquots at –80°C. Double-stranded siRNAs (iGENE Therapeutics, Tsukuba, Japan) were designed as described previously (Shoji et al., 2006) and stored in aliquots at –80°C.
RNA solutions were diluted with sterile PBS to the desired concentration and injected (typically at concentrations of 0.5 to 1 mg/ml for cRNA and 25 μM for siRNA) within 1 hour of thawing via a piezo-actuated micropipette (tip inner diameter 6∼7 μm) into mII oocytes in M2 medium. Oocytes were cultured for 4 or 7 hours following injection of cRNA or siRNA respectively, and, where appropriate, then transferred to KSOM containing TPEN.
Imaging relative [Ca2+i] and available [Zn2+]
Indirect imaging to determine relative [Ca2+i] within mII oocytes and following 500 μM IP3 injection, ICSI, ethanol treatment, SrCl2 treatment and/or TPEN treatment was as described previously (Yoshida et al., 2007a). Injection and analysis of mII oocytes was ∼18-20 hours after hCG injection. Oocytes were loaded for 30 minutes with 5 μM fura 2 acetoxymethyl ester (Fura 2-AM; Molecular Probes, CA, USA) before exposing them to the activating stimulus. Fluorescence recordings were then initiated immediately, as previously described (Yoshida et al., 2007a), and processed with AQUA COSMOS ratio imaging application software (Hamamatsu Photonics, Japan). For experiments with TPEN, recordings typically started 60-120 seconds after TPEN wash out. In some experiments, TPEN-treated oocytes (n=17) were quickly washed and transferred to a juxtaposed drop on the microscope stage for immediate imaging, with an interval of ∼30 seconds between washing and recording.
To determine the effect of Ca2+i chelation on activation stimuli, oocytes were incubated in the plasma membrane permeant Ca2+ sponge 1,2-bis-(o-aminophenoxy)-ethane-N,N,N′,N′-tetraacetic acid tetraacetoxymethyl ester (BAPTA-AM) (Calbiochem, EMD Chemicals, Gibbstown, NJ, USA) at 50 μM in KSOM for 20 minutes in humidified CO2 [5% (v/v) in air] at 37°C. Following BAPTA loading, Pb1-lacking oocytes were either injected with sperm heads (ICSI) or challenged to activation by SrCl2, ethanol or TPEN as described elsewhere. Meiotic progression was confirmed after 2 hours of continued incubation in KSOM in humidified CO2 at 37°C by the presence of a Pb2 with or without Tuba/DNA fluorescence imaging.
The membrane permeant AM ester forms of Zn2+-sensitive dyes, FluoZin-3 (Invitrogen, CA, USA) or RhodZin-3 (Invitrogen), were used to measure relative ooplasmic [Zn2+]c. Oocytes were incubated in KSOM containing 10 μM FluoZin-3 or RhodZin-3 for 30 minutes either prior to, or following, ICSI or exposure to ethanol, SrCl2 or TPEN. Fluorescence was monitored following excitation respectively at 470/40 nm (FluoZin-3) or 540/25 nm (RhodZin-3) and emission detection at 535/50 (FluoZin-3) or 605/55 nm (RhodZin-3) using a BioZero-8000 microscope/detector (Keyence, Osaka, Japan) and analyzed with BZ-Analyzer software (Keyence).
Protein fluorescence imaging
Immunocytochemistry, differential interference contrast microscopy (DIC) and epifluorescence imaging were essentially as described previously (Yoshida et al., 2007a). Cortical granule staining after sample fixation [4% (w/v) paraformaldehyde] was with 100 μg/ml fluorescein isothiocyanate (FITC)-conjugated lens culinaris agglutinin (Sigma). Images of live oocytes following cRNA injection were captured via a BZ-8000 (Keyence) and analyzed using BZ-Analyzer software (Keyence). Excitation at 540/25 nm was used with a TRITC (red) filter system for mCherry fluorescence detection and at 480/30 nm with a GFP (green) filter system to detect Venus epifluorescence.
The response of mII oocytes to TPEN was visualized by time-lapse microscopy of oocytes containing a Venus-tubulin-α (Tuba) fusion protein whose expression was driven by the ZP3 transgene promoter on a C57BL/6×C3H background (subsequently back-crossed to C57BL/6). The 4.5 kbp BciVI-MluI pZP3→Venus-Tuba transgene fragment was generated by inserting a 2019 bp pZP3-containing XhoI-KpnI genomic DNA fragment upstream of a 710 bp BamHI-BsrGI Venus fragment linked to a 1633 bp BsrGI-MluI fragment from pEGFP-Tuba, which encodes human tubulin-α (Clontech Laboratories, CA, USA). Oocytes were placed in a KSOM droplet under mineral oil on a glass-bottomed dish containing 100 μM TPEN on the stage of a TE2000 inverted microscope (Nikon, Japan) equipped with a CSU10 confocal scanning unit (Yokogawa, Japan) and a humidified chamber [5% (v/v) CO2 in air] at 37°C. DIC images and fluorescent (488 nm) images (typically 13 focal planes, step size=2 μm) were captured at 5-minute intervals by a C9100-13 ImagEM EM-CCD camera (Hamamatsu Photonics, Shizuoka, Japan) driven by MetaMorph (Molecular Devices, CA, USA) image analysis software.
Sperm preparation, microinjection and nuclear transfer
Sperm preparations were as described previously, with minor modifications (Yoshida et al., 2007b). Briefly, cauda epididymidal spermatozoa from 12- to 30-week-old male B6D2F1 mice were triturated in nuclear isolation medium [NIM: 125 mM KCl, 2.6 mM NaCl, 7.8 mM Na2HPO4, 1.4 mM KH2PO4, 3.0 mM EDTA (pH 7.45)] containing 1.0% (w/v) 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) at room temperature (25°C) for 1 minute. Sperm were washed twice in ambient temperature NIM to give control de-membranated heads, cHds. Where appropriate, sperm suspensions (or part thereof) were incubated at 48°C for 30 minutes, with trituration after 15 minutes, to generate `inactivated', iHd preparations. (Sperm preparations contained all sperm-derived debris, including tails, mid-pieces and other fragments, but only heads were injected.) Sperm were mixed with one to two volumes of 15% (w/v) PVP360 (average Mr=360,000; Kanto Chemical, Tokyo, Japan) and microinjected as described (Yoshida and Perry, 2007), typically within 30 minutes of PVP mixing. Batches of oocytes were typically injected within 15 minutes and activation immediately induced by exposure to SrCl2 (Perry et al., 2000), ethanol or TPEN followed by washing in KSOM and continuing KSOM incubation in 5% (v/v) CO2 humidified air at 37°C. Parthenogenetic activation of iHd-injected oocytes was as described above, within 30 minutes of injection. Cumulus cell nuclear transfer into enucleated mII oocytes was essentially as described previously (Wakayama et al., 1998) with TPEN activation one to two hours post nuclear transfer, where appropriate.
Ratiometric quantification of mRNAs (qPCR) in day 4.5 (E4.5, 96 hours) embryos was essentially as described previously (Amanai et al., 2006; Shoji et al., 2006) using the following primer pairs (5′ to 3′): H2afz, GCGTATCACCCCTCGTCACTTG and TCTTCTGTTGTCCTTTCTTCCCG; Pou5f1 (Oct4), CGTGAAGTTGGAGAAGGTGGAACC and GCAGCTTGGCAAACTGTTCTAGCTC; Sox2, GGAAAAAAACCACCAATCCCATCC andTTTGCGAACTCCCTGCGAAG;Nanog, GCAAGCGGTGGCAGAAAAAC and GCAATGGATGCTGGG ATACTCC; Cdx2, GGAAGCCAAGTGAAAACCAG and CTTGGCT CTGCGGTTCTG; Socs3, GCAGATCAACAGATGAGCCA and TGG GACAGAGGGCATTTAAG; Eif3s10, AAGGGGTGATGATGCAAGAC and AGGTGGACCCCAACTCTCTT. Eight samples per group were analyzed in triplicate per primer pair and data normalized with respect to H2afz.
Standard immunoblotting (IB) for Fig. 2B was with rabbit polyclonal anti-CycB1 [Santa Cruz Biotechnology, CA, USA; 1:500 (v/v)], anti-Tubg (γ-tubulin; Abcam, MA, USA; 1:100 (v/v)], rabbit monoclonal anti-phosphoErk1/2 [Cell Signaling Technology, MA, USA; 1:1000 (v/v)], rabbit polyclonal anti-Erk1/2 [Cell Signaling; 1:1000 (v/v)], or rabbit polyclonal anti-Tuba [α-tubulin; Abcam; 1:1000 (v/v)] primary antibodies, and anti-rabbit IgG [Invitrogen, CA, USA; 1:10,000 (v/v)] secondary antibody.
Blastocyst cell counting
E4.5 embryos were fixed [4% (w/v) paraformaldehyde] and subjected to standard incubation at 4°C overnight in rabbit anti-Oct4 (1:10,000; a kind gift from Dr H. Niwa) and for 1 hour at 37°C in mouse anti-Cdx2 (1:100; MU392-UC, BioGenex Laboratories, CA, USA), followed by a 1 hour incubation at 37°C in Alexa 488-conjugated anti-rabbit IgG (Invitrogen) and TRITC-conjugated anti-mouse IgG (Sigma). Cells exclusively stained with Alexa were scored as Oct4-positive pluriblasts, and those exclusively TRITC staining, as Cdx2-positive trophoblasts.
Unless stated otherwise, Student's t-tests were respectively applied to comparative unpaired analyses of Ca2+i oscillation and developmental frequencies. Pearson's chi-squared (χ2) tests were performed where indicated. Data for each experiment were collected on at least 2 days.
Zn2+ is required to maintain mII arrest in mouse oocytes
To determine whether Zn2+ played a role in mouse mII arrest, we exposed intact oocytes to the selective Zn2+ chelator, N,N,N′,N′-tetrakis-(2-pyridylmethyl)-ethylenediamine (TPEN), which exhibits a low affinity for Ca2+ (kd[Ca]=4.0×10–5 M, kd[Zn]=2.6×10–16 M) (Arslan et al., 1985). All underwent dose-dependent mII exit, evidenced by spindle movement and Pb2 extrusion, with kinetics for 100 μM TPEN (100 μM is used throughout unless stated otherwise) similar to those of the parthenogenetic agent, SrCl2 at 10 mM (80.6±1.3 and 73.4±1.4 minutes, respectively, n=25 and n=22) (Fig. 1A-C; see Fig. S1A in the supplementary material). Chelating agents selective for Mn2+, Cu2+ or Fe2+ did not induce oocyte activation (see Fig. S1B in the supplementary material). Meiotic release by TPEN was inhibited by parallel or sequential incubation in 100 μM Zn2+ but not Mn2+, Cu2+ or Fe2+ (Fig. 1D,E; see Fig. S1C in the supplementary material). To ascertain the effect of TPEN on levels of exchangeable cytoplasmic Zn2+ (Zn2+)c, oocytes were loaded with the Zn2+-sensitive fluorescent probes, FluoZin-3 or RhodZin-3. These have kds for Zn2+ at least seven orders of magnitude higher than that of TPEN (see Table S1 in the supplementary material) and did not induce mII exit. FluoZin-3 and RhodZin-3 fluorescence decreased in response to TPEN (Fig. 1F,G), correlating TPEN treatment, mII exit and [Zn2+]c. The TPEN-induced decrease in FluoZin-3 fluorescence was reversed by 100 μM Zn2+ (see Fig. S1D in the supplementary material). Exposure of FluoZin-3-loaded mII oocytes to Sr2+ (which induces mII exit) but not Mn2+ or Fe2+ (which do not) elicited a 50% decrease in fluorescence within 15 minutes (see Fig. S2 in the supplementary material), but induction of mII exit by intracytoplasmic sperm injection (ICSI) did not induce analogous Zn2+c redistribution.
Meiotic exit Induced by Zn2+ depletion neither requires nor mobilizes Ca2+i
The mobilization of Ca2+i is ubiquitous in fertilization (Stricker, 1999). Sequestration of Ca2+i with BAPTA (Kline and Kline, 1992) prevented the induction of mII exit (Pb2 extrusion) by sperm, ethanol or SrCl2, but had no effect on TPEN-induced meiotic progression (Fig. 2A,B). Exposure of oocytes to TPEN did not induce a [Ca2+i] increase (0/37), unlike sperm (n=26) or SrCl2 (n=15) (Fig. 2C; see Fig. S3A in the supplementary material). TPEN was washed out after treatment and [Ca2+i] recording was initiated within ∼30-120 seconds, but no [Ca2+i] change was observed (n=40) over 6 hours.
TPEN also inhibited Ca2+i mobilization, as pre-incubation with TPEN prevented [Ca2+i] oscillations in response to Sr2+ and reduced the frequencies of both ICSI- and inositol 1,4,5-trisphosphate-induced [Ca2+i] oscillations (see Fig. S3B,C in the supplementary material) (Lawrence et al., 1998). The reduction of ICSI-induced oscillations was significantly (P<0.0001) rescued by 100 μM Zn2+, although Zn2+ had no effect on ICSI-induced oscillations per se (see Fig. S3C in the supplementary material). These findings indicate potent inhibition of IP3 signaling and Ca2+ release by TPEN and that Zn2+ depletion induces meiotic exit independently of Ca2+i.
TPEN delineates Ca2+i-dependent oocyte activation events
Mammalian fertilization induces oocyte APC activation, degradation of proteasomal targets, including Emi2 and CycB, and cortical granule (CG) exocytosis (a vesicle fusion event) (Runft et al., 2002; Schmidt et al., 2005; Madgwick et al., 2006; Perry and Verlhac, 2008). We investigated whether TPEN also induced these events, first estimating changes in relative Emi2 levels. When mII oocytes were injected with complementary RNA (cRNA) encoding Emi2-mCherry and challenged 4 hours later with either sperm or SrCl2, Emi2 degradation was rapid; sperm induced a 50% fluorescence decline in 14.1 minutes (Fig. 3A). TPEN treatment did not effect Emi2-mCherry levels (Fig. 3A) but induced depletion of endogenous phospho-MAPK and CycB1 (Fig. 3B), and CycB1-Venus (in oocytes injected with CycB1-Venus cRNA; Fig. 3C), with similar kinetics to those induced by sperm or SrCl2. To determine whether Emi2 degradation required Zn2+c, oocytes expressing Emi2-mCherry were pre-incubated with TPEN and then exposed to ethanol. Controls treated with ethanol alone (n=36) all underwent [Ca2+i] release and marked loss of Emi2-mCherry fluorescence, although most oocytes treated with TPEN and then ethanol (30/38) underwent a Ca2+i rise without Emi2-mCherry degradation (Fig. 3D; not shown). This suggests that a Ca2+ signal is insufficient to ensure Emi2 degradation in the absence of Zn2+c.
TPEN-induced mII exit and recombinant CycB1 degradation were prevented by the 26S proteasome inhibitors epoxomicin or Z-LLL-CHO (Fig. 3E,F). When the APC activator Cdc20 was reduced to less than 10% of control levels by RNAi (Shoji et al., 2006; Amanai et al., 2006), TPEN treatment failed to induce mII exit (Fig. 3G). Mouse mII exit requires an intact spindle (Kubiak et al., 1993). Consistent with this, microtubule disruption by nocodazole completely prevented activation by either SrCl2 (n=16) or TPEN (n=27). TPEN treatment also failed to induce CG exocytosis (Fig. 3H). Thus, Zn2+ chelation destabilizes mII arrest in a Cdc20- and microtubule-dependent, but Ca2+i-independent, manner that requires proteasomal activity.
Developmental consequences of oocyte Zn2+ sequestration
The finding that TPEN induces [Ca2+i]-static meiotic resumption raised the issue of whether resultant embryos could develop and caused us to examine the developmental role of fertilization-induced [Ca2+i] changes. Oocytes activated by continual TPEN exposure rarely developed beyond two cells, suggesting embryotoxicity. Those exposed for only 45 minutes (in the presence of the microfilament inhibitor, cytochalasin B, to prevent cytokinesis and therefore loss of maternal chromosomes) produced 27.3±13.7% diploid parthenogenetic blastocysts (Fig. 4A) and when supplemented with 100 μM ZnSO4 for 3 hours post-TPEN, expanded blastocyst development improved to 75.3±3.7% by embryonic day 4.5 (E4.5, 96 hours) (Fig. 4A,B). The transition metal cations Mn2+, Cu2+, Fe2+, Ni2+ or Co2+ did not rescue TPEN toxicity (Fig. 4C), indicating that embryotoxicity had been specifically due to Zn2+ deprivation. We also applied the TPEN activation regimen to cumulus cell nuclear transfer (Fig. 4D). Development in vitro to the expanded blastocyst stage at E4.5 following the transfer of cumulus cell nuclei and activation with SrCl2 (19.6% of pronuclear zygotes) was markedly reduced when activation was with TPEN (7.6%, P<0.0001) (Fig. 4D). Activation with SrCl2 followed by treatment with TPEN also resulted in low developmental rates of nuclear transfer embryos (1.7%; Fig. 4D).
Diploid E4.5 parthenogenotes arising from TPEN or SrCl2 activation contained similar levels of Oct4 (Pou5f1), Sox2, Nanog and Cdx2 (Fig. 4E), although the TPEN group had fewer Oct4- and Cdx2-expressing cells (Fig. 4F,G). Transcripts whose levels are reportedly sensitive to [Ca2+i] oscillations (Ozil et al., 2006; Rogers et al., 2006), including Eif3s10, Socs3, Rpl9 and Sfrs2, were present at similar levels in TPEN and SrCl2 groups at E2.5 (Fig. 4H).
The sperm-borne signal for the [Ca2+i] rise can be inactivated by heating, so that sperm heads (`inactivated heads', iHds) do not elicit mII exit (Fig. 5A) (Perry et al., 1999; Perry et al., 2000; Yoshida et al., 2007b). Exposure of iHd-injected oocytes to SrCl2 or ethanol, respectively, induced [Ca2+i] oscillations or a prolonged monotonic [Ca2+i] rise (Rickords and White, 1993), whereas 194/195 (99.49%) exposed to TPEN exhibited no [Ca2+i] change (Fig. 5A), with one case undergoing a single small bell-shaped release lasting 3.16 minutes, which was insufficient to induce mII exit (Kline and Kline, 1992). Embryos in SrCl2 and ethanol groups developed comparably in vitro with good, if poorer, development in the TPEN group (Fig. 5B,C). Following two-cell embryo transfer, SrCl2 and ethanol groups yielded equivalent full-term development (respectively, 9.0 and 9.2% of transfers; P=0.758); TPEN-induced development produced 24 offspring (4.7%) that grew into healthy fertile adults (Fig. 5D,E). This lower rate probably reflects TPEN embryotoxicity rather than lack of [Ca2+i] mobilization, as iHd-injected oocyte exposure to TPEN, or to SrCl2 (i.e. activation with [Ca2+i] oscillations) followed by TPEN, yielded similar rates (P=0.380) of full-term development (Fig. 5D). Recording Ca2+i levels during iHd injection in standard medium showed a small transient rise (n=4), but no rise in Ca2+-free medium (n=4; Fig. 5A). TPEN activation of oocytes injected with iHds in Ca2+-free medium (eliminating external Ca2+ influx during injection; Fig. 5A) caused pronounced oocyte trauma that frequently resulted in death, yet we obtained three offspring from 175 embryos (1.7%). Full-term development is therefore not conditioned upon Ca2+i release during mII exit and [Ca2+i] oscillations do not assure an altered rate of full-term development compared with a single [Ca2+i] rise.
This work delineates roles of Zn2+ and Ca2+ during mammalian mII and mII exit. From the results obtained, it can be argued that Zn2+ is required for mII arrest and that Ca2+i release during fertilization is not essential for full-term development. The work provides evidence that Zn2+ is required for the Emi2-mediated regulation of meiotic arrest in mouse mII oocytes. Meiotic resumption after Zn2+ depletion is not accompanied either by Ca2+ release or Emi2 degradation, both of which are induced by Ca2+-dependent oocyte activation (Fig. 2A-C; Fig. 3A). Furthermore, events of mII exit that depend on the APC-proteasome pathway, including cyclin B degradation and chromosome separation (Peters, 2006), occur with similar kinetics whether or not Ca2+ is mobilized (Fig. 1A-C; Fig. 3B,C). This observation implies that Zn2+ depletion activates or unmasks the APC-proteasome pathway, even in the presence of Emi2. Indeed, TPEN-induced mII exit is prevented by proteasome inhibitors (Fig. 3E,F) or removal of the APC activator Cdc20 by RNAi (Fig. 3G). Finally, Emi2 contains a putative zinc-binding region and, unlike Emi2, other cytostatic factor candidates have been shown to be dispensable for mouse mII arrest, including Emi1, Mos (Shoji et al., 2006) and spindle assembly checkpoint proteins Bub1, Mad2 and BubR1 (Tsurumi et al., 2004).
Induction of mII exit by sperm, SrCl2 or ethanol – but not TPEN – is accompanied by Emi2 disappearance (Fig. 3A), suggesting that Emi2 degradation requires Ca2+ release. As Emi1 undergoes SCFTrcpb-mediated destruction in response to phosphorylation by Plk1 (Hansen et al., 2004), this requirement could be due to Ca2+-dependent kinases that become activated during fertilization and phosphorylate Emi2 to target it for proteolysis. Sequential phosphorylation by CaMKII and Plk1 mediates the Ca2+-sensitive degradation of Xenopus Emi2 (Rauh et al., 2005; Schmidt et al., 2005). Recent work employing RNAi or gene targeting suggests that CaMKIIγ plays an important role in physiological mII exit in the mouse (Chang et al., 2009; Backs et al., 2010). However, it remains to be seen whether Emi2 degradation is CaMKIIγ-dependent as it is in Xenopus (Rauh et al., 2005). Our result that depletion of Zn2+ inhibits ethanol-induced Emi2 degradation (Fig. 3D) implies that Zn2+ might also be involved in Emi2 degradation during mII exit. It remains to be seen precisely how mouse Emi2 degradation is controlled at meiotic resumption and why Emi1 and Emi2 – unlike Mad2 – have evolved a Zn2+-dependent mechanism for APC inhibition (Reimann et al., 2001b; Herzog et al., 2009).
Emi2 depletion induces meiotic exit (Shoji et al., 2006) and its degradation is caused by Ca2+i release. Our results show that neither Ca2+i release nor Emi2 degradation at the time of mII exit is essential for full-term development. It may be inferred that Emi2 destruction is the sole indispensable role of Ca2+ release during meiotic exit. We did not observe Ca2+ release at any time before, during or after the addition of TPEN to, or release of TPEN from, oocytes in procedures that produced offspring, including iHd injection in Ca2+-free medium (Fig. 5A). It has previously been reported that the characteristics of Ca2+ signaling at fertilization are important for multiple events of oocyte activation and later development in vitro (Ducibella et al., 2002; Ozil et al., 2005; Rogers et al., 2006) and in vivo (Ozil et al., 2006). These studies modulated Ca2+i dynamics in Ca2+-free media, or with electropermeabilization or the eukaryotic protein synthesis inhibitor cycloheximide. Electrical pulses and protracted Ca2+ deprivation may influence development owing to variables other than Ca2+i mobilization, reflecting limited specificity or secondary effects. The negative developmental effects of cycloheximide (Rogers et al., 2006) were rescued by co-incubation with ethanol or Sr2+, but only when the co-incubation followed relatively brief cycloheximide exposure – too little for any demonstrable effect – and without allowing for the possibility that Ca2+ or Sr2+ treatment is antagonistic to cycloheximide activity.
When mII exit was induced by oscillations (Sr2+) or a monotonic Ca2+ rise (ethanol) just after iHd injection, embryos developed at indistinguishable rates, arguing against developmental enhancement by Ca2+ oscillations. Given the clear implication that Ca2+i oscillations are dispensable during mouse fertilization, the issue arises as to their function. The regulation of APCCdc20 by Emi2 to sustain mII arrest and the precipitous degradation of Emi2 following activation (Fig. 3A) argue against cumulative CycB degradation with each Ca2+i oscillation (Shoji et al., 2006; Ducibella and Fissore, 2008). Rather, it is possible that the main function of oscillations is as a fail-safe to peg the [Ca2+i] to sub-cytotoxic levels while periodically enabling Ca2+ signaling; enduringly elevated [Ca2+i] causes oocyte death and persistence until pronucleus formation may ensure that [Emi2] remains low until Emi2 mRNA levels decline (Berridge, 1987; Perry et al., 1999; Perry et al., 2000; Fujimoto et al., 2004).
This is perhaps the first definitive role for Zn2+ in the oocyte, but it is unlikely to be the last given the range of biological functions that Zn2+ controls and the subtle regulatory complexity of the oocyte-to-embryo transition. Development of embryos is poor following Zn2+ sequestration (relative to non-sequestration), even when accompanied by Ca2+ release (Fig. 5D). Potential physiological roles for Zn2+ in ensuring the gamete-to-embryo transition include transcription, signaling, histone modification and chromatin remodeling (Yamasaki et al., 2007; Bottomley et al., 2008; Maret, 2009; Viiri et al., 2009). These roles could account for the observation that nuclear transfer TPEN embryos arrest at the two-cell stage (Fig. 4D). Two-cell arrest also occurs in embryos lacking Brg1, which binds to zinc-finger proteins through a unique N-terminal domain and is essential in mice for the general activation of zygotic transcription and oocyte-to-zygote transition (Kadam and Emerson, 2003; Bultman et al., 2006). Although Brg1 is not known to bind Zn2+, disruption of Zn2+ homeostasis could perturb the Brg1 regulatory network and others involving Zn2+. We cannot exclude the possibility that technical aspects of Zn2+ depletion and nuclear transfer have a synergistically negative effect, but development was initially normal and this is a topic of ongoing investigation. Manipulating oocyte [Zn2+] promises to pave the way for new approaches to understanding mechanisms that underlie early embryogenesis in mammals.
The authors declare no competing financial interests.
The authors are grateful to Drs Maki Asami, Christoph Klein, Karen Lee and Matthew VerMilyea for their constructive criticisms, and to LARGE for embryo transfers. This work was funded by RIKEN.
Competing interests statement