Downregulation of store-operated Ca²⁺ entry during mammalian meiosis is required for the egg-to-embryo transition

Sperm–egg fusion marks the initiation of multicellular organism development and is triggered by a specialized Ca²⁺ transient at fertilization in all sexually reproducing species investigated to date (Antoine et al., 2000; Miyazaki et al., 1993; Runft et al., 2002; Stricker, 1999). This fertilization-specific Ca²⁺ signal is required to mediate the egg-to-embryo transition leading to embryonic development. Ca²⁺ is well suited as the egg activation messenger throughout phylogeny, because as a second messenger it spans an impressive spatial and temporal spectrum endowing it with the ability to encode multiple cellular processes (Bootman et al., 2002). Importantly eggs acquire the ability to produce the fertilization-specific Ca²⁺ transient during oocyte maturation, which is a cellular differentiation pathway that prepares the egg for the egg-to-embryo transition (Eppig et al., 2004).

Throughout oogenesis growing vertebrate oocytes are arrested in prophase of the first meiotic division in a G2-like state with an intact nuclear envelope, referred to as the germinal vesicle (GV) stage in mammalian oocytes (Eppig et al., 2004). This is a stable meiotic cell cycle arrest that can last for several decades in humans (Hassold and Hunt, 2001). Oocyte maturation is initiated at ovulation when oocytes are released from their follicle. This marks their commitment to the completion of meiosis I and progression to the second meiotic metaphase. Mature eggs remain arrested at metaphase II (MII) through a cytostatic-factor (CSF)-mediated arrest until fertilization (Miyazaki, 1995; Smith, 1989). At fertilization PLC-ζ is introduced by the sperm into the egg cytoplasm and initiates the characteristic repetitive Ca²⁺ oscillations that are required for the egg-to-embryo transition (Saunders et al., 2002). Ca²⁺ oscillations release the CSF-mediated MII arrest resulting in the completion of meiosis II with the extrusion of the second polar body, and mediate the transition to the mitotic embryonic cell cycles. In addition, the Ca²⁺ transient at fertilization mediates another essential cellular function, that is the block to polyspermy to maintain the ploidy of the embryo, which is required for embryonic development (Runft et al., 2002; Stricker, 1999; Whitaker, 2006).

The Xenopus laevis model has provided important insights into the remodeling of Ca²⁺ signaling pathways during oocyte maturation in preparation for fertilization (Machaca, 2007). IP₃-dependent Ca²⁺ release is sensitized due in part to ER remodeling and the enrichment of IP₃ receptors in large ER patches (El-Jouni et al., 2005; Machaca, 2004). A role of IP₃ receptor phosphorylation is also likely in this cell cycle-dependent sensitization (Sun et al., 2009). The plasma membrane Ca²⁺-ATPase is internalized (El-Jouni et al., 2008), and the store-operated Ca²⁺ entry (SOCE) pathway is completely inactivated (Machaca and Haun, 2000; Machaca and Haun, 2002). SOCE is a Ca²⁺ influx pathway activated in response to intracellular Ca²⁺ store depletion and represents the predominant Ca²⁺ entry pathway in vertebrate oocytes (Hartzell, 1996; Kline and Kline, 1992b; Machaca and Haun, 2000). In a similar fashion IP₃-dependent Ca²⁺ release is sensitized during mammalian oocyte maturation (Lee et al., 2006; Mehlmann et al., 1996; Xu et al., 2003). In mammals IP₃-dependent Ca²⁺ release is required for egg activation at fertilization (Kline and Kline, 1994; Miyazaki et al., 1992), whereas the role of the other major Ca²⁺ release channel, the ryanodine receptor is contentious (Jones et al., 1995; Kline and Kline, 1994; Miyazaki et al., 1992; Miyazaki et al., 1993; Swann, 1992).

Fertilization results in either a single Ca²⁺ transient in species such as Xenopus and sea urchin or multiple Ca²⁺ oscillations in mammals and ascidians for example (Stricker, 1999). This species-specific Ca²⁺ signal at fertilization is essential for the egg-to-embryo transition. Ca²⁺ oscillations in mammals have been shown to be required for egg activation (Cuthbertson and Cobbold, 1985; Ducibella et al., 2002) and Ca²⁺ influx through the SOCE pathway was shown to be important for maintaining Ca²⁺ oscillations for several hours, presumably through refilling of Ca²⁺ stores (Igusa and Miyazaki, 1983; Kline and Kline, 1992b; McGuinness et al., 1996; Mohri et al., 2001). SOCE is mediated by two primary molecules, the Orai1 channel at the cell membrane and the ER Ca²⁺ sensor STIM1, which contains a lumenal EF-hand that binds Ca²⁺ and monitor Ca²⁺ store content (Lewis, 2007). Upon store depletion STIM1 clusters into so called puncta and translocates to a cortical ER domain that localizes ∼20 nm below the cell membrane allowing it to physically recruit Orai1 into coincident puncta and gate the Orai1 channel open, thus allowing Ca²⁺ influx into the cell (Cahalan, 2009; Hogan et al., 2010).

Therefore, in contrast to IP₃ dependent Ca²⁺ release, SOCE is regulated differentially in species with a single Ca²⁺ spike at fertilization (Xenopus) as compared to those with multiple Ca²⁺ oscillations (mouse). In a similar fashion to what is observed during Xenopus meiosis, SOCE inactivates fully during mitosis of mammalian cells (Arredouani et al., 2010; Preston et al., 1991; Tani et al., 2007). Therefore, SOCE inactivates during both meiosis and mitosis, with the only exception being mammalian oocyte meiosis during which SOCE activity can still be detected. Furthermore, SOCE has been implicated in defining the oscillatory Ca²⁺ signal in mammalian eggs (Kline and Kline, 1992b; Mohri et al., 2001). Based on these findings we have previously argued that this differential SOCE regulation during oocyte maturation may be a determinant of the dynamics of the fertilization-specific Ca²⁺ signal (Machaca, 2007): single spike in species where SOCE is completely inactivated versus multiple oscillations in species where SOCE remains active after oocyte maturation.

Here we investigate the mechanisms controlling SOCE during mammalian oocyte maturation and show that SOCE is downregulated during oocyte maturation. Molecularly this SOCE inhibition is mediated by internalization of Orai1 and a diminution in the ability of STIM1 to translocate to the cortical ER. SOCE in mature MII eggs is mediated by STIM1 and Orai1 and overexpression of these molecules disrupts the dynamics of the Ca²⁺ oscillations at egg activation and alters the egg-to-embryo transition resulting in defects in early embryonic development.

We measured SOCE during mouse oocyte maturation after store depletion in Ca²⁺-free medium followed by Ca²⁺ addition to separate Ca²⁺ release from Ca²⁺ influx. Stores were depleted with the endoplasmic reticulum Ca²⁺-ATPase (SERCA) blocker thapsigargin, which prevents store refilling. Because of an endogenous Ca²⁺ leak pathway from the ER, this results in store depletion. Oocytes at the germinal vesicle stage (GV) exhibit a robust SOCE signal that is dramatically decreased in MII eggs following oocyte maturation (Fig. 1A), with a relative SOCE in immature oocytes of 1.71±0.32 as compared to 0.68±0.18 in mature MII eggs (Fig. 1C), a statistically significant decrease (P<0.0019). SOCE inhibition during oocyte maturation was confirmed using TPEN to chelate store Ca²⁺, thus simulating store depletion in the absence of any Ca²⁺ release from stores (supplementary material Fig. S1A).

As has been previously reported, Ca²⁺ store content increases dramatically during oocyte maturation (Jones et al., 1995; Tombes et al., 1992), resulting in elevated Ca²⁺ release in MII eggs (Fig. 1B). Relative Ca²⁺ store content based on thapsigargin stimulated Ca²⁺ release was calculated at 0.45±0.17 in GV oocytes compared to 2.05±0.32 in MII eggs, a statistically significant increase (P<1.7×10⁻⁴) (Fig. 1B). However, measuring Ca²⁺ store content using thapsigargin is problematic because the endogenous leak pathway is slow, thus as stores are being emptied, Ca²⁺ is buffered and extruded out of the cell. As such the dynamics of the Ca²⁺ dye and Ca²⁺ extrusion/buffering could incorporate significant inaccuracies in measuring Ca²⁺ store content. A more direct approach to measure Ca²⁺ store content is to empty Ca²⁺ stores using the Ca²⁺ ionophore, ionomycin, in Ca²⁺-free media. This experimental paradigm shows 5-fold higher store Ca²⁺ content in MII eggs when compared to GV oocytes (supplementary material Fig. S1B).

We then sought to empty Ca²⁺ stores during oocyte maturation through direct activation of the IP₃ receptor, the primary Ca²⁺ release channel in mouse oocytes. Toward that goal we loaded oocytes with caged-IP₃ and uncaged it repeatedly during the imaging time course with UV light, and measured Ca²⁺ levels using Fluo-3 (Fig. 2A). Surprisingly, IP₃ uncaging did not produce any Ca²⁺ release in GV oocytes incubated in Ca²⁺-free medium. Nonetheless, when the cells were switched to Ca²⁺-containing medium a large SOCE is readily recorded, followed by Ca²⁺ oscillations after several minutes of Ca²⁺ influx (Fig. 2A). These data argue that Ca²⁺ stores are already empty using this experimental paradigm, thus preventing any Ca²⁺ release following IP₃ uncaging. When Ca²⁺ flows into the GV oocyte through SOCE following Ca²⁺ addition to the extracellular space, Ca²⁺ stores refill leading to Ca²⁺ oscillations as IP₃ levels are high in the cell due to the uncaging pulses. If this interpretation is correct then simply incubating GV oocytes in Ca²⁺-free medium for the same duration would lead to Ca²⁺ store depletion. This is indeed the case as shown in Fig. 2B. GV oocytes were loaded with only Fluo-3 (without any caged-IP₃) and incubated in Ca²⁺-free medium for over 10 minutes, then switched to Ca²⁺ containing medium (Fig. 2B, GV). This leads to the activation of a robust SOCE that is similar to that observed following thapsigargin-mediated store depletion or when cells were loaded with caged-IP₃ (Fig. 2B, GV). However, in this case, in the absence of caged-IP₃, no Ca²⁺ oscillations are observed following store refilling (Fig. 2B). This shows that the leak pathway in GV oocytes is highly active and is sufficient to fully empty Ca²⁺ stores when cells are simply incubated in Ca²⁺-free medium for several minutes.

In contrast, incubating MII eggs in Ca²⁺-free medium for over 10 minutes does not lead to Ca²⁺ store depletion or the activation of SOCE (Fig. 2B, MII). This argues that during oocyte maturation the ER Ca²⁺ leak pathway is inhibited. When MII cells are loaded with caged-IP₃ and uncaged for several minutes to produce IP₃, two response patterns are observed (Fig. 2A, MII). In both cases only Ca²⁺ oscillations are observed but they either occur as IP₃ is uncaged in Ca²⁺-free medium or only after Ca²⁺ addition to the extracellular medium. These data suggest that Ca²⁺ influx through SOCE potentiates IP₃-dependent Ca²⁺ release most likely through Ca²⁺-induced Ca²⁺ release.

Together these results show that during mouse oocyte maturation Ca²⁺ signaling remodels at multiple levels: (1) Ca²⁺ store content increases; (2) SOCE is downregulated but not completely inactivated; and (3) the ER Ca²⁺ leak pathways is significantly inhibited during maturation.

During Xenopus oocyte maturation SOCE completely inactivates due to the internalization of Orai1 and the inhibition of STIM1 clustering (Machaca and Haun, 2000; Machaca and Haun, 2002; Yu et al., 2009; Yu et al., 2010). To determine whether a similar regulation of STIM1 and Orai1 occurs in mouse oocyte meiosis, we expressed GFP-tagged Orai1 and mCherry-tagged STIM1 and observed their distribution during meiosis. As expected, Orai1 is enriched at the cell membrane, and STIM1 is diffusely distributed to the ER (Fig. 3A). Store depletion does not result in any noticeable change in Orai1 distribution; however STIM1 becomes enriched in the cortical ER below the plasma membrane (Fig. 3A). STIM1 redistribution in response to store depletion is illustrated by a line scan showing cortical ER enrichment following store depletion. Higher magnification imaging close to the cell membrane reveals patches of colocalized STIM1 and Orai1 that are reminiscent of the STIM1 puncta described in mammalian cells (Liou et al., 2007; Luik et al., 2006; Luik et al., 2008), however these STIM1–Orai1 patches appear more diffuse and less well defined than the typical STIM1 puncta observed in mammalian cells or Xenopus oocytes.

In MII eggs after oocyte maturation, Orai1 is removed from the plasma membrane and becomes enriched intracellularly (Fig. 3B,C). Therefore, Orai1 is internalized during mouse oocyte maturation. Staining oocytes expressing GFP–Orai1 with the early endosomal marker EEA1 shows that during oocyte maturation Orai1 is internalized into an early endosomal compartment as it colocalizes with EEA1 in MII eggs but not in GV oocytes (Fig. 3D). Therefore, as is the case in Xenopus (Yu et al., 2010), Orai1 trafficking in GV and MII cells is distinct.

The subcellular distribution of STIM1 during oocyte maturation reflects the previously described ER remodeling (Kline et al., 1999). STIM1 shows increased density in the egg cortex (Fig. 3B), reminiscent of ER remodeling. Upon store depletion there is a moderate enrichment of Orai1 at the cell membrane coupled to an increase in STIM1 clusters (Fig. 3B), arguing that STIM1 retains some ability to form clusters following store depletion in MII eggs. The membrane enrichment of Orai1 in response to store depletion was measured in the zoomed in regions using a series of multiple line scans across the cell membrane (Fig. 3B). Averaging these line scans and normalizing to the maximal signal reveals the enrichment of Orai1 at the plasma membrane in MII eggs after thapsigargin (TG) treatment but not in control cells (Fig. 3B). It is important to note that the increase in STIM1 clustering was variable from cell to cell and sometimes undetectable (∼20% of the cells); this is also the case for Orai1 cell membrane enrichment.

Together these results show that during oocyte maturation Orai1 is internalized into an intracellular early endosomal compartment yet retains the ability to translocate back to the cell membrane although with much reduced efficiency. STIM1 loses its ability to translocate to the cortical ER domain following store depletion in MII eggs. However, STIM1 retains the capacity to form clusters in MII eggs. As is the case in Xenopus oocytes, STIM1 mobility on SDS-PAGE was shown to shift in MII mouse eggs strongly arguing that it is phosphorylated (Gómez-Fernández et al., 2009). The subcellular distribution of STIM1 and Orai1 before and after store depletion in MII eggs is consistent with the greatly reduced SOCE detected functionally (Fig. 1).

Another formal possibility that we investigated in terms of the residual SOCE recorded in MII eggs is an Orai isoform switch during oocyte maturation. Mammalian genomes encode three Orai isoforms that are differentially expressed. The role of Orai2 is not clear, but Orai3 has been implicated in the arachidonic acid channel (ARC) (Mignen et al., 2008; Thompson et al., 2010) and during the progression of breast cancer (Faouzi et al., 2011; Motiani et al., 2010). In fact Orai3 seems to be the predominant Orai in estrogen positive breast cancer cells, whereas Orai1 assumes this function in other breast cancers (Motiani et al., 2010). Given this precedent we wondered whether a similar isoform switch occurs during oocyte maturation where Orai3 becomes the SOCE channel in the mature MII egg. Staining for endogenous Orai3 shows that it is expressed in both the GV and MII stages of maturation, and that unlike Orai1 its membrane residence is not modulated during oocyte maturation, as it remains enriched at the cell membrane in MII eggs (Fig. 4A).

To determine whether Orai3 is functional in both cell types we used 2-APB, which has been shown to directly open Orai3 channels independent of store depletion (DeHaven et al., 2008; Peinelt et al., 2008; Zhang et al., 2008). We therefore tested the effect of 2-APB on Ca²⁺ influx in GV and MII cells with (Fig. 4B) or without (Fig. 4C) store depletion. In both cases 2-APB potentiates Ca²⁺ influx confirming the membrane residence and functionality of Orai3 in GV and MII cells.

Orai1 internalizes during oocyte maturation yet a small percentage of internalized Orai1 translocates to the cell membrane following store depletion. In addition, Orai3 is present at the cell membrane throughout maturation. Hence both channels could associate with STIM1 to mediate the residual SOCE in the MII egg. To differentiate between those two possibilities we attempted to knock down Orai1, STIM1 and Orai3 using various approaches, including morpholinos and siRNA. However, we were unable to achieve significant and consistent knockdown to allow for a conclusive determination of the composition of the MII SOCE channel. We therefore overexpressed these proteins in various combinations and studied the effect on SOCE in both GV and MII (Fig. 5).

Overexpression of Orai1, STIM1 or Orai3 alone in GV oocytes has no effect on the amplitude of the ensuing SOCE (Fig. 5A,B). In contrast, SOCE was potentiated when STIM1 was co-expressed with either Orai1 or Orai3 (Fig. 5A,B), showing that STIM1 can associate with either channel to mediate Ca²⁺ influx in the GV oocyte. In MII eggs expression of Orai1 or STIM1 alone increases SOCE (Fig. 5C,D), but co-expression of both Orai1 and STIM1 results in a large significant SOCE similar to that observed in GV oocytes (Fig. 5C,D). Therefore, co-expression of STIM1 and Orai1 reverses the downregulation of SOCE observed during oocyte maturation.

Expression of Orai3 alone or co-expression of Orai3 with STIM1 did not potentiate SOCE in MII cells (Fig. 5C,D); showing that for some reason STIM1 cannot associate with and activate Orai3 in MII eggs. Together these results strongly argue that the SOCE channel in MII cells is composed of Orai1 and STIM1, and that both proteins maintain a level of residual activity that is able to induce the small SOCE (Fig. 1). Orai1 can translocate back to the cell membrane and STIM1 sustains a reduced ability to form clusters and recruit Orai1. However, in this case, and in contrast to cells in interphase, Orai1 recruitment is from an intracellular pool and not through the classical clustering of Orai1 channels in the plane of the cell membrane. Unfortunately, because we were unable to knock down STIM1 we could not formally test this conclusion.

Overexpression of STIM1 and Orai1 in MII eggs reverses SOCE downregulation and results in MII eggs that produce a significant SOCE in response to store depletion (Fig. 5). This provides a unique opportunity to test the physiological role of SOCE inhibition during oocyte maturation. As discussed above the egg activates at fertilization in response to a specific Ca²⁺ transient, which in mammalian oocytes consists of repetitive Ca²⁺ spikes that last for a couple of hours. SrCl₂ has been used extensively as a fertilization surrogate to activate MII eggs (Gordo et al., 2002; Kline and Kline, 1992a; Knott et al., 2006). Indeed SrCl₂ leads to the typical repetitive, periodic pattern of Ca²⁺ oscillations in control MII eggs (Fig. 6A). Interestingly, expression of STIM1 and Orai1 results in broader Ca²⁺ spikes, a decreased inter-spike interval and spikes that often fuse into an extended Ca²⁺ plateau that could last for tens of minutes (Fig. 6A). The Ca²⁺ oscillation frequency was calculated from the number of Ca²⁺ spikes over the 120 minutes recording period and was significantly different (P<0.00162) between the control (5.57±0.75, n = 14) and the STIM1–Orai1 overexpressing eggs (11.66±1.66, n = 11) (Fig. 6B). The broader Ca²⁺ spikes and their increased frequency combine to produce on average a more sustained Ca²⁺ rise over the duration of the oscillations (Fig. 6A). To quantify the duration of elevated Ca²⁺ signal, we calculated the percent time that Ca²⁺ is at high levels (defined as Ca²⁺ levels 20% above baseline) for the duration of the 120 minutes recording (Fig. 6C). This shows that cytoplasmic Ca²⁺ levels are elevated for 50.58±6.53% of the time in STIM1+Orai1 overexpressing cells as compared to only 17.43±2.8% of the time in control cells (Fig. 6C). Thus overexpression of STIM1 and Orai1 results in elevated Ca²⁺ levels for significantly longer times (P<2.92×10⁻⁵) when compared to control cells. These data show that SOCE levels modulate the oscillatory pacemaker and the integrated duration of the Ca²⁺ signal during egg activation.

As discussed the spatial and temporal properties of the Ca²⁺ signal at fertilization represent the code that orchestrates egg activation and the ensuing egg-to-embryo transition. Overexpression of STIM1 and Orai1 disrupts this critical code by altering the dynamics of the Ca²⁺ oscillations that direct egg activation. We therefore tested the effect of STIM1 and Orai1 overexpression on the egg-to-embryo transition and the early embryonic divisions over a 24-hour time course following SrCl₂ activation. We assessed the transition of MII eggs to the 2-polar bodies (2-PB) stage marking the completion of meiosis, to the pronuclear (PN) stage, and to the 2-cell stage. Interestingly, when compared with control untreated MII eggs, cells overexpressing STIM1 and Orai1 activate faster, where a significantly larger percentage of the cell population reaches the 2-PB stage within the first 2 hours after SrCl₂ activation (Fig. 6D) (60.67±9.49% STIM1–Orai1 vs 37±9.86% Control). This faster activation starts to fade at the pronuclear stage (PN), where although a higher percentage of the STIM1–Orai1 expressing cells reach the PN stage at 4 hours as compared to control cells, this difference is not significant. However, a small percentage of the STIM1–Orai1 overexpressing MII eggs population reaches the 2-cell stage (9±4.58% as compared to 35.33±14.11% for control cells); rather the majority of MII eggs overexpressing STIM1–Orai1 either do not progress to the 2-cell stage or fragment (39.33±15.88%) during the 24-hour incubation period. Examples of the different stages of egg activation and early embryonic divisions are shown in Fig. 6E. We could not detect any morphological differences at the 2-PB or PN stages between control and STIM1–Orai1 injected cells, so the examples shown represent both the experimental and control groups. Examples of the cellular fragmentation in STIM1–Orai1-injected cells are also shown (Fig. 6E).

Our results show that the physiological downregulation of SOCE that occurs during oocyte maturation is critical for the ensuing egg-to-embryo transition. SrCl₂ activation data argue that SOCE plays a critical role in modulating the activity of the oscillatory pacemaker in the egg thus defining the periodicity and duration of the repetitive Ca²⁺ spikes that mediate egg activation. This conclusion is consistent with the findings of McGuinness et al. (McGuinness et al., 1996) who observed a correlation between Ca²⁺ influx and spiking intervals in mouse oocytes and have argued for SOCE as the regulator of spiking interval through a combination of store refilling and Ca²⁺-induced Ca²⁺ release.

Importantly, we show that increasing the amplitude of SOCE in MII eggs leads to prolonged spikes at the expense of a decreased inter-spike interval. The decrease in inter-spike interval is dramatic enough where it can lead to a plateau of high Ca²⁺ levels as in the example in Fig. 6A. These alterations to Ca²⁺ oscillations due to enhanced SOCE in MII eggs combine to produce a prolonged integrated Ca²⁺ rise over the duration of the oscillations (Fig. 6C). Therefore, the cellular machinery in MII eggs overexpressing STIM1 and Orai1 experience a prolonged Ca²⁺ rise as compared to control cells. Importantly we show that this alteration in the dynamics of Ca²⁺ oscillations during egg activation disrupts the egg-to-embryo transition. Specifically it leads to more rapid egg activation initially followed by cellular fragmentation and the demise of the embryo at later time points. This is consistent with the fact that cellular events at activation are Ca²⁺-dependent, and shows that SOCE downregulation during oocyte maturation is a critical determinant of the Ca²⁺ oscillations dynamics at egg activation and as such the egg-to-embryo transition. Furthermore, these results support and extend recent studies showing that the timing and pattern of Ca²⁺ oscillations affect embryo quality (Ajduk et al., 2011).

The distribution of STIM1 and Orai1 we observe is distinct from that previously reported in mouse oocytes (Gómez-Fernández et al., 2009; Gómez-Fernández et al., 2012) where STIM1 was described to localize to distinct vesicular-like ER structures before store depletion and to large ER areas after store depletion. Orai1 distribution was described as patchy membrane localized in MII eggs. We do not understand the reasons for these discrepancies, however the unusual distribution of STIM1 described by Gomez-Fernandez et al. is quite distinct from the spatial distribution of the ER viewed with traditional ER markers such as the lipophilic dicarbocyanine dye DiI (FitzHarris et al., 2003; Mehlmann et al., 1995), or after direct staining for the IP₃ receptor, a resident ER protein (Wakai et al., 2012). These studies show a diffuse ER similar to what we observe in STIM1 expressing cells (Fig. 3). Furthermore, in pig oocytes STIM1 exhibits a more diffuse ER-like distribution and clustering following store depletion (Wakai et al., 2012). This brings into question the physiological significance of the unusual distribution of STIM1 observed by Gomez-Fernandez et al. (Gómez-Fernández et al., 2009).

Consistent with our results manipulating STIM1 or Orai1 expression levels in pig oocytes results in defective Ca²⁺ oscillations at fertilization and embryonic development (Lee et al., 2012; Wang et al., 2012). Surprisingly however in both mouse and pig oocytes previous reports detected little or no SOCE in GV oocytes (Gómez-Fernández et al., 2012; Wang et al., 2012). In contrast we detect a robust SOCE in GV oocytes in the mouse that is significantly larger than that measured in MII cells. The reasons for these differences are unclear to us.

In summary, SOCE is a ubiquitous Ca²⁺ influx pathway that regulates many physiological functions including activation of immune cells and secretion (Parekh and Putney, 2005). SOCE has been shown to be consistently downregulated during cell division (Arredouani et al., 2010), however the significance of this inhibition remained unclear. Here we show that SOCE levels in the mature MII egg modulate the periodicity and duration of Ca²⁺ oscillations and the duration of the Ca²⁺ signal at egg activation. The specific dynamics of the Ca²⁺ oscillations represent the code that instructs the egg-to-embryo transition. Accordingly, altering SOCE regulation during meiosis disrupts the egg-to-embryo transition resulting in the death of the zygote. Therefore, SOCE inhibition during mammalian meiosis is essential for egg activation and early embryonic development.

Female CD-1 mice between 6 and 12 weeks of age (Charles River Laboratories, Wilmington, MA) were used for all experiments. All animal experiments were performed according to approved guidelines. To obtain germinal vesicle (GV) oocytes, mice were super-stimulated with injection of 5 IU pregnant mare serum gonadotropin (PMSG; Sigma, St. Louis, MO; all chemicals from Sigma unless otherwise specified) and oocytes were recovered from ovaries 40 hours post-PMSG in HEPES-buffered tyrode's lactate solution (TL-HEPES) [in mM: 114 NaCl, 3.2 KCl, 2 NaHCO₃, 0.4 NaH₂PO₄, 10 Na lactate, 2 CaCl₂, 0.5 MgCl₂, 10 HEPES, 0.2 Na pyruvate] supplemented with 4% BSA and 10 µM milrinone to prevent spontaneous resumption of meiosis. Metaphase II (MII) eggs were obtained by super-stimulating females with 5 IU PMSG followed by stimulation of ovulation with injection of 5 IU human chorionic gonadotropin (hCG) 44–46 hours post-PMSG injection. Cumulus oocyte complexes (COC) were recovered from oviducts 14 hours post-hCG into TL-HEPES supplemented with 4% BSA. MII eggs were stripped of cumulus cells by a brief incubation in 0.1% bovine testis hyaluronidase in TL-HEPES and placed in potassium simplex optimized medium (KSOM; Specialty Media, Phillipsburg, NJ) under paraffin oil at 36.5°C in a humidified atmosphere containing 5% CO₂. For experiments requiring in vitro maturation of oocytes, GV oocytes were thoroughly washed of milrinone and cultured in Chatot, Ziomek, Bavister (CZB) medium (in mM: 81.62 NaCl, 4.83 KCl, 1.18 KH₂PO₄, 1.18 MgSO₄, 1.7 CaCl₂, 25.12 NaHCO₃, 31.3 Na lactate, 0.27 Na pyruvate, 1 L-glutamine, 0.11 EDTA) supplemented with 4 mg/ml BSA under paraffin oil at 36.5°C and 5% CO₂ for 14–16 hours.

Eggs were placed in 50 µl drops of TL-HEPES supplemented with 20% FCS under paraffin oil. While being held with a holding pipette, eggs were injected using the PLI-188 picoinjector (Harvard Apparatus, Cambridge, MA) attached to a glass micropipette containing the reagent. Pneumatic pressure was applied to deliver the reagent into the cytoplasm of the eggs.

Oocytes were fixed in freshly prepared 4% paraformaldehyde (PFA) for 30 minutes, washed with immunofluorescence (IF) buffer made up of phosphate-buffered saline (PBS) supplemented with 1% BSA and 5% goat serum, then permeabilized with 1% Triton X-100 for 30 minutes. Oocytes were stripped of zona pellucida by a brief incubation in acidic tyrode's solution. Following several washes in IF buffer, oocytes were incubated with primary antibody (1∶1000 dilution) overnight at 4°C. After primary antibody incubation, oocytes were washed in IF buffer before incubation with secondary antibody (1∶2000), goat anti-rabbit IgG labeled with Alexa Fluor 594 for 1 hour at room temperature.

Live cell imaging was performed on a Zeiss LSM510 confocal microscope using Plan Apo 63× /1.4 oil DIC II objective and LSM510 AIM software. Green fluorescent protein (GFP) signal was excited with 488 nm laser and emissions collected through a bandwidth of 492–558 nm. mCherry signal was excited at 543 nm and emission collected through a bandwidth of 572–699 nm. Images were analyzed using LSM510 AIM software and figures compiled using Adobe Photoshop.

Eggs were incubated in 1 µM Fura-2/acetoxymethyl ester (Life Technologies, Grand Island, NY) supplemented with 0.02% Pluronic F-127 (Life Technologies) in TL-HEPES for 15 minutes at room temperature. Following incubation, eggs were washed in Ca²⁺-free TL-HEPES and attached on the bottom of glass-bottomed dish in a 50 µl drop of medium. The glass-bottomed dish was placed on an inverted microscope (Nikon, Tokyo, Japan) fitted for fluorescence measurements with the InCyt Standard Im Imaging System (Intracellular Imaging, Cincinnati, OH). The system was fitted with a DVC 340M Monochrome Integrating 12-bit CCD camera with binning capabilities. Excitation light source was a 175 watt Xenon arc lamp for both UV and visible light excitation. A variable intensity filter wheel (Sutter Instrument, Novato, CA) rotated between 340 and 380 nm excitation wavelengths and the fluorescence ratios of 340/380 nm were recorded every 20 seconds. Caged IP₃ was injected into eggs previously loaded with Fluo3 am (1 µM; Life Technologies). Fluo-3 was excited by exposure to 488 nm wavelength, and IP₃ was uncaged by exposure to pulses of UV light of 360 nm for 100 milliseconds.

To induce parthenogenetic activation, MII eggs were placed in Ca²⁺-free CZB medium supplemented with 10 mM SrCl₂ under paraffin oil at 36.5°C and 5% CO₂ for 2 hours. At the completion of SrCl₂ incubation, eggs were washed in TL-HEPES and transferred to drops of KSOM under paraffin oil at 36.5°C and 5% CO₂ and examined for signs of activation and development periodically. Extrusion of second polar bodies (2PB), pronuclei (PN) formation, and cell divisions were monitored.

pCDNA3-GFP-Orai1 was linearized with XhoI and pCMV-XL5-mCherry-STIM1 with XbaI downstream of T7 promoter. pCDNA3.1-Orai3 was linearized with AvrII downstream of T7 promoter. In vitro transcription/translation was performed using the T7 mMessage mMachine kit (Life Technologies) and polyA tails were added to the 3′ end of the coding sequence using a polyA tailing kit (Life Technologies).

The statements made herein are solely the responsibility of the authors.