Constitutive IP₃R1-mediated Ca²⁺ release reduces Ca²⁺ store content and stimulates mitochondrial metabolism in mouse GV oocytes

ABSTRACT

In mammals, fertilization initiates Ca²⁺ oscillations in metaphase II oocytes, which are required for the activation of embryo development. Germinal vesicle (GV) oocytes also display Ca²⁺ oscillations, although these unfold spontaneously in the absence of any known agonist(s) and their function remains unclear. We found that the main intracellular store of Ca²⁺ in GV oocytes, the endoplasmic reticulum ([Ca²⁺]_ER), constitutively ‘leaks’ Ca²⁺ through the type 1 inositol 1,4,5-trisphosphate receptor. The [Ca²⁺]_ER leak ceases around the resumption of meiosis, the GV breakdown (GVBD) stage, which coincides with the first noticeable accumulation of Ca²⁺ in the stores. It also concurs with downregulation of the Ca²⁺ influx and termination of the oscillations, which seemed underpinned by the inactivation of the putative plasma membrane Ca²⁺ channels. Lastly, we demonstrate that mitochondria take up Ca²⁺ during the Ca²⁺ oscillations, mounting their own oscillations that stimulate the mitochondrial redox state and increase the ATP levels of GV oocytes. These distinct features of Ca²⁺ homeostasis in GV oocytes are likely to underpin the acquisition of both maturation and developmental competence, as well as fulfill stage-specific cellular functions during oocyte maturation.

Summary: Mouse GV oocytes display spontaneous Ca²⁺ oscillations that reduce intracellular Ca²⁺ stores and stimulate mitochondrial metabolic output, maintaining cellular homeostasis during prolonged meiotic arrest.

INTRODUCTION

Prior to ovulation and following a systemic luteinizing hormone (LH) surge, prophase-arrested germinal vesicle (GV) oocytes resume meiosis and progress to the metaphase stage of the second meiosis (MII), completing a process that is commonly referred to as oocyte maturation. The arrest that ensues in MII oocytes is relieved by fertilization, as following gamete fusion the sperm promotes periodical increases in the intracellular concentration of free Ca²⁺, which are termed Ca²⁺ oscillations (Stricker, 1999). The ability of MII oocytes to mount precise spatiotemporal patterns of Ca²⁺ oscillations in response to fertilization is progressively acquired during oocyte maturation (Wakai et al., 2011). Other parameters of Ca²⁺ homeostasis also undergo marked changes during maturation, which are likely to impact how oscillations unfold (Wakai and Fissore, 2013). For example, the content of the main intracellular Ca²⁺ store within the endoplasmic reticulum ([Ca²⁺]_ER) (Jones et al., 1995; Mehlmann and Kline, 1994; Wakai et al., 2012), markedly increases during maturation and there is a pronounced downregulation of Ca²⁺ influx as maturation progresses (Cheon et al., 2013; Lee et al., 2013). Remarkably, the molecules and regulatory mechanisms that underlie this optimization of Ca²⁺ oscillations and changes in Ca²⁺ homeostasis during oocyte maturation are largely unknown and may experience dynamic modifications such that mechanisms active at the GV stage may not be so at the MII stage, and vice versa.

GV oocytes also exhibit oscillatory Ca²⁺ responses (Carroll and Swann, 1992; Carroll et al., 1994). In contrast to sperm-induced Ca²⁺ oscillations, the spontaneous, repetitive Ca²⁺ rises in GV oocytes are of small amplitude and occur approximately every 1 to 3 min. These Ca²⁺ oscillations appear to be agonist-independent and are constitutive, although the mechanisms that underpin them and their functions remain poorly investigated. It is noteworthy that these spontaneous Ca²⁺ oscillations persist for a few hours and cease around the time of GV breakdown (GVBD), which is when oocytes simultaneously undergo the most drastic increase in [Ca²⁺]_ER store content and experience a sharp downregulation in Ca²⁺ entry (Cheon et al., 2013; Wakai and Fissore, 2013). The temporal coincidence of these phenomena implicates [Ca²⁺]_ER-associated mechanisms in the occurrence of spontaneous Ca²⁺ oscillations.

Inositol 1,4,5-trisphosphate receptor family (IP₃R)-mediated Ca²⁺ release from the [Ca²⁺]_ER is primarily responsible for the Ca²⁺ oscillations during fertilization (Miyazaki et al., 1992). The spontaneous Ca²⁺ oscillations at the GV stage are also mediated by IP₃R1 (also known as ITPR1), as inhibition of IP₃R1 function by heparin blocks the oscillations (Carroll and Swann, 1992). Nevertheless, the regulation of these Ca²⁺ oscillations is likely to be more complex, as given that they unfold in the absence of a specific agonist, they must be regulated by other factors of cellular Ca²⁺ homeostasis such as Ca²⁺ influx and Ca²⁺ clearing mechanisms. To this end, continued Ca²⁺ oscillations require replenishment of the [Ca²⁺]_ER, a function that is mostly carried out by the sarcoplasmic/endoplasmic reticulum Ca²⁺ ATPase (SERCA) pump protein family (Wakai et al., 2013). In this regard, the content of the [Ca²⁺]_ER is vital to maintain the oscillations because it is the main source of Ca²⁺. The mitochondria also take up cytosolic Ca²⁺, thereby contributing to shaping of the spatiotemporal patterns of Ca²⁺ responses, while simultaneously promoting a number of events that sustain ATP levels in cells (Hajnóczky et al., 1995). In fact, the Ca²⁺-driven ATP output is likely to be the mitochondria's most critical contribution towards Ca²⁺ homeostasis in MII oocytes, as ATP production maintains SERCA activity, which is required for maintaining [Ca²⁺]_ER levels and sustaining sperm-triggered Ca²⁺ oscillations (Dumollard et al., 2004; Wakai et al., 2013). It is nevertheless unknown whether the spontaneous oscillations observed in GV oocytes play a similar role in mitochondrial function. In addition, the mechanisms and underlying molecules involved in the regulation of Ca²⁺ homeostasis in GV oocytes remain poorly investigated.

In the present study, we uncover a unique regulatory mechanism of [Ca²⁺]_ER content in GV oocytes, which is reversed in MII oocytes; namely, in GV oocytes [Ca²⁺]_ER levels are maintained persistently low by a constitutive Ca²⁺ ‘leak’ through IP₃R1 channels that is manifested in the form of spontaneous Ca²⁺ oscillations. The Ca²⁺ oscillations are actuated by nearly constitutive Ca²⁺ influx, and downregulation of this influx, which terminates the oscillations, appears linked to the progressive increase in the content of the [Ca²⁺]_ER that occurs during oocyte maturation. Further, we demonstrate a novel role for the spontaneous Ca²⁺ oscillations, as part of the cytosolic Ca²⁺ is transferred into the mitochondria where it stimulates mitochondrial metabolism, thereby increasing the levels of ATP in GV oocytes. These distinct regulatory mechanisms of organellar Ca²⁺ homeostasis in GV oocytes are likely to underpin the acquisition of both maturation and developmental competence, as well as fulfill stage-specific cellular function during oocyte maturation.

RESULTS

GV oocytes constitutively leak Ca²⁺ from [Ca²⁺]_ER stores

It is a well-known phenomenon that GV-arrested mouse oocytes show reduced intracellular Ca²⁺ stores content (Jones et al., 1995; Mehlmann and Kline, 1994; Wakai et al., 2012). This was demonstrated using ionomycin to empty the ER stores, and measure Ca²⁺ responses in the presence of Ca²⁺-free conditions containing EGTA. Further, those studies showed that the Ca²⁺ store content, as assessed by the same procedure, sharply increased between the GV and GVBD stages and less so thereafter until the MII stage (Fig. S1).

The underlying reason(s) by which Ca²⁺ stores have less Ca²⁺ at the GV stage is not known, although the occurrence of spontaneous Ca²⁺ oscillations suggests a possible molecular mechanism. Hence, we propose that a Ca²⁺ leak out of the ER, which lowers its Ca²⁺ content, induces transient Ca²⁺ rises by stimulating Ca²⁺-induced Ca²⁺ release (CICR) through IP₃R1 channels, and supports oscillations by promoting Ca²⁺ influx. To ascertain the presence of a constitutive Ca²⁺ leak in GV oocytes, we took advantage of the knowledge that the plasma membrane Ca²⁺ ATPase (PMCA, also known as ATP2B) family of proteins, which remove Ca²⁺ from the cell, are inhibited by millimolar (mM) concentrations of lanthanum (La³⁺). Further, at these concentrations La³⁺ generates a Ca²⁺ insulation system, because both Ca²⁺ influx and efflux are inhibited, allowing the leaking Ca²⁺ to build up in the cytosol and induce CICR (Bird and Putney, 2005). Using this experimental system, we found that, indeed, GV oocytes (n=21/21) cultured in the absence of external Ca²⁺ but in the presence of 1 mM La³⁺ show a sharp, single Ca²⁺ transient a few minutes after addition of La³⁺ (Fig. 1A). Significantly, the same procedure in GVBD oocytes (n=4/10) generates a smaller Ca²⁺ response with a shallow sloping rise (Fig. 1B), and Ca²⁺ responses are no longer induced in MII oocytes (n=0/14) (Fig. 1C). Quantification of the response showed that the amplitude of the Ca²⁺ peak induced by La³⁺ progressively decreased during maturation (Fig. 1D). We next examined the remaining Ca²⁺ content in the ER, as the expectation would be that the leaked Ca²⁺ would empty the [Ca²⁺]_ER store, which should then display a decreased Ca²⁺ response following addition of the SERCA inhibitor thapsigargin (10 μM). This was the case, as whereas addition of thapsigargin failed to cause a Ca²⁺ rise in GV oocytes, it induced greater responses in GVBD oocytes and even greater in MII oocytes (Fig. 1E). Taken together, these results show that GV oocytes spontaneously leak Ca²⁺ out of the [Ca²⁺]_ER store, and this leak is progressively inactivated during maturation starting at around the GVBD stage.

Fig. 1.

GV oocytes display a constitutive Ca²⁺ leak out of the ER store. (A–C) Representative traces of Ca²⁺ responses induced by La³⁺ (La) Ca²⁺ insulation in GV (A), GVBD (B) and MII (C) oocytes (GV, n=21; GVBD, n=10; MII, n=14). Ca²⁺ measurements were performed in Ca²⁺-free medium containing 1 mM LaCl₃. When the Ca²⁺ increase returned to baseline, 10 μM thapsigargin (Tg) was applied to confirm the remaining Ca²⁺ in the ER. (D,E) Comparisons of the amplitude of the Ca²⁺ peak induced by La³⁺ (D) and the area under each curve (arbitrary units, A.U.) of thapsigargin-induced Ca²⁺ rise (E) between the different stages of oocyte maturation. Error bars represent s.d., and bars with different superscripts are significantly different from each other (P<0.05).

IP₃R acts as the [Ca²⁺]_ER leak channel in GV oocytes

To gain insight into the mechanism(s) underlying the constitutive Ca²⁺ leak from the [Ca²⁺]_ER store in GV oocytes, we assessed the role of IP₃R1 channels. To accomplish this we used a ligand-induced knockdown approach. The type 1 IP₃R (ITPR1) is the most abundantly expressed IP₃R isoform in mammalian oocytes (Fissore et al., 1999; Parrington et al., 1998), and adenophostin A (AdA) is a potent agonist of IP₃Rs capable of promoting long-lasting Ca²⁺ release through this receptor, while at the time inducing powerful IP₃R1 degradation (Brind et al., 2000; He et al., 1999). Consistent with these previous results, we observed that injection of 10 μM AdA into GV oocytes stimulated Ca²⁺ release and oscillations (Fig. 2A), and reduced IP₃R1 protein to nearly undetectable levels by 6 h post-injection (Fig. 2B). Using the Ca²⁺ insulation approach we examined whether the Ca²⁺ leak was present in AdA-injected GV oocytes. Addition of La³⁺ failed to induce a Ca²⁺ rise in just over half of the oocytes (16/30), and the Ca²⁺ rise was severely disturbed in the remaining oocytes (14/30) (Fig. 2B). Moreover, the spontaneous Ca²⁺ oscillations present in GV oocytes at physiological concentrations of extracellular Ca²⁺ (1.7 mM) were inhibited in AdA-injected oocytes (Fig. 2C). Consistent with above results, the [Ca²⁺]_ER content estimated from the Ca²⁺ response to ionomycin was greatly increased in AdA-injected GV oocytes, reaching levels similar to those observed in MII oocytes (Fig. 2D). These results demonstrate that IP₃R1 channels are largely responsible for the constitutive Ca²⁺ leak that occurs in GV oocytes and maintains low levels of Ca²⁺ in [Ca²⁺]_ER. Consistent with these results, inhibition of the oocyte family of PLC proteins with 10 μM of the PLC inhibitor U73122 terminated the oscillations of GV oocytes (Fig. S3A), whereas addition of the same concentration of inactive analog U73433 did not block oscillations (Fig. S3B).

Fig. 2.

IP₃R1 channels are responsible for the low [Ca²⁺]_ER content of GV oocytes. (A) Representative Ca²⁺ responses induced by the potent agonist of IP₃R channels, adenophostin A (AdA), in GV oocytes is shown. Loss of IP₃R1 expression was confirmed by western blot analysis using antibodies specific to IP₃R1 and α-tubulin 6 h after injection of 10 μM AdA. Oocytes undergoing microinjection of AdA (GV-AdA) or buffer (GV) were cultured in the presence of IBMX to maintain the arrest at the GV stage. (B) Representative traces of Ca²⁺ responses induced by La³⁺ (La) Ca²⁺ insulation in AdA-injecting GV oocytes. Ca²⁺ measurements were performed in Ca²⁺-free medium containing 1 mM LaCl₃. The number of oocytes responding to lanthanide is shown. (C) Representative traces of Ca²⁺ responses in GV oocytes (black trace) and AdA-injected GV oocytes (red trace) are shown. Ca²⁺ measurement were performed under normal external Ca²⁺ concentration (1.7 mM). The percentages of oocytes showing spontaneous Ca²⁺ oscillations (oocytes showing more than 3 repetitive Ca²⁺ rises) were compared. Numbers of oocytes showing oscillations/total number of oocytes are indicated. (D) ER Ca²⁺ store content in in GV oocytes (black trace; n=20), AdA-injecting GV oocytes (red trace; n=14) and MII oocytes (blue trace; n=10), which was estimated by the mean fluorescent (Fura-2) peak after addition of 1 μM ionomycin. Amplitudes of the Ca²⁺ peak were compared. Error bars represent s.d., and bars with different superscripts are significantly different from each other (P<0.05).

Ca²⁺ influx is required for the spontaneous Ca²⁺ oscillations in GV oocytes

Although the constitutive Ca²⁺ release and oscillations in GV oocytes are seemingly autonomous and independent of a known agonist, the oscillatory Ca²⁺ responses suggest participation of the IP₃R1 channels. We hypothesize that in GV oocytes the incoming Ca²⁺ modulates the function of IP₃R1 channels. GV oocytes express several divalent cation permeable plasma membrane channels for Ca²⁺ influx, although the individual contribution of each of these channels to the influx and oscillations has not yet been fully ascertained. One possible mechanism of Ca²⁺ influx is the store-operated Ca²⁺ entry (SOCE) mechanism, which is ubiquitous in non-excitable cells and is activated by the depletion of [Ca²⁺]_ER (Parekh and Putney, 2005). We and others have previously demonstrated that Ca²⁺ influx, and SOCE activity following over-expression of SOCE mechanism components, is highest in GV oocytes and progressively decreases during oocyte maturation (Cheon et al., 2013; Lee et al., 2013). Thus, this temporal coincidence of downregulation of Ca²⁺ influx and termination of Ca²⁺ oscillations at around the GVBD stage strengthens the association of Ca²⁺ influx and the presence of oscillations in GV oocytes. To decipher the role of Ca²⁺ influx on the oscillations at this stage, we modulated the strength of Ca²⁺ influx by overexpressing STIM1 and ORAI1, which are the molecular components of SOCE. We observed that under the regular external Ca²⁺ concentrations of 1.7 mM, control GV oocytes displayed the customary Ca²⁺ oscillations, whereas oocytes overexpressing STIM1 and ORAI1 showed permanently increased basal Ca²⁺ levels, indicative of persistent Ca²⁺ influx (Fig. 3A,D). We also found that these changes in intracellular Ca²⁺ were due to Ca²⁺ influx, as addition of 100 µM 2-APB, a broad-spectrum Ca²⁺ channel inhibitor (DeHaven et al., 2008), terminated the oscillations and/or returned the levels of Ca²⁺ to basal values. We next exposed oocytes to medium containing lower external Ca²⁺ concentrations, 0.5 mM, which prevented oscillations in control oocytes, although oocytes overexpressing STIM1 and ORAI1 now displayed oscillations rather than elevated basal Ca²⁺ levels, confirming the association of influx and oscillations (Fig. 3B,D). Finally, under nominal Ca²⁺-free conditions, both groups of oocytes were no longer able to mount changes in intracellular Ca²⁺ levels (Fig. 3C,D). Collectively, these results suggest that in GV oocytes, Ca²⁺ influx regulates intracellular Ca²⁺ responses, and therefore Ca²⁺ influx not only replenishes the [Ca²⁺]_ER content but also actuates the spontaneous oscillations at this stage.

Fig. 3.

The magnitude of Ca²⁺ influx impacts the strength of spontaneous Ca²⁺ oscillations in GV oocytes. (A–C) Representative traces of Ca²⁺ responses in control GV oocytes (black traces) and in GV oocytes overexpressing STIM1 and ORAI1 (Stim1/Orai1-OVE, red traces) are shown. Oocytes were arrested at GV stage throughout the imaging by the addition of 100 mM IBMX. Ca²⁺ measurements were performed under 1.7 mM (A), 0.5 mM (B) and 0 mM (C) external Ca²⁺ concentrations. Under normal external Ca²⁺ concentrations, 1.7 mM, 2-APB, a broad-spectrum SOCE inhibitor, was added at concentrations of 100 µM. (D) The percentages of oocytes showing spontaneous Ca²⁺ oscillations (oocytes showing more than 3 repetitive Ca²⁺ rises were counted) were compared. Numbers of oocytes showing oscillations/total number of oocytes are indicated.

Decline in the number of channels at the plasma membrane coincides with downregulation of spontaneous Ca²⁺ oscillations

Given that in most oocytes the spontaneous Ca²⁺ oscillations cease 2 to 4 h after the initiation of maturation, at the GVBD stage (Carroll and Swann, 1992), the downregulation of Ca²⁺ influx is likely to occur during this time period, which curiously precedes the increase in [Ca²⁺]_ER store content that takes place mostly after GVBD. Previous studies have noticed downregulation of influx and SOCE during mouse oocyte maturation (Cheon et al., 2013; Lee et al., 2013), although the precise timing and mechanism(s) remain unknown. Because expression of ORAI1 in mouse GV oocytes results in specific plasma membrane localization, and when expressed together with STIM1 they stimulate Ca²⁺ influx in these cells, we expressed fluorescently labeled human ORAI1 to examine its fate during maturation. Following injection of ORAI1 cRNA, we examined the distribution of ORAI1–RFP at 0, 2, 4, 8 and 12 h after the initiation of in vitro maturation, which broadly corresponded with GV, early GVBD, late GVBD, MI and MII stages of meiotic progression, respectively. We found that there is an abrupt decrease in the presence of ORAI1–RFP in the plasma membrane at 2 h of maturation (Fig. 4A–C), suggesting that internalization of the putative active channel(s) might be one of the mechanisms that bring about the inactivation of Ca²⁺ influx and termination of the Ca²⁺ oscillations.

Fig. 4.

Progression of oocyte maturation causes reduced presence of ORAI1 at the plasma membrane. (A) The subcellular distribution of ORAI1 was analyzed using an mRFP-tagged (top panel) version of the protein. The observations were performed at 0, 2, 4, 8 and 12 h after initiation of in vitro maturation, which corresponded with GV, early GVBD, late GVBD, MI and MII stages, respectively. Square area on the top panels indicate areas magnified to observe ORAI1 plasma membrane presence at the different stages of maturation. Higher magnification views of the selected area (inset panels) are shown to the left of each image where differences in ORAI1 distribution between different maturation stages can be observed. Corresponding DIC images are shown in the bottom panel. Scale bar: 20 µm. (B) Intensity profiles of the line scans drawn in oocytes in A (white lines in insets), representing the distribution of ORAI1–RFP fluorescence from the cytoplasm (Cyto) and across the plasma membrane (PM) at different stages of maturation. (C) Mean±s.d. relative intensity of ORAI1 signal between plasma membrane and cytoplasm is shown.

To further demonstrate the association between the expression of Ca²⁺ channels in the plasma membrane and the spontaneous Ca²⁺ oscillations, we interfered with the expected distribution of ORAI1–RFP in GV oocytes. Synaptosomal-associated protein 25 (SNAP25) is a t-SNARE protein component of the trans-SNARE complex involved in membrane fusion between vesicles and plasma membrane. In Xenopus oocytes, expression of a dominant-negative SNAP25 mutant (SNAP25^Δ20) was shown to effectively block exocytosis and inhibit ORAI1 trafficking between endosomes and the plasma membrane (Yu et al., 2009). Expression of SNAP25^Δ20 or wild-type SNAP25 (SNAP25^WT) in mouse GV oocytes (Fig. S2A) did not relieve meiotic arrest, contrary to what is observed in Xenopus oocytes, although SNAP25^Δ20 compromised the distribution of ORAI1, which appeared discontinuous, in contrast to the continuous distribution displayed in control oocytes (Fig. S2B). Furthermore, the spontaneous Ca²⁺ oscillations in oocytes expressing SNAP25^Δ20 were severely disturbed (Fig. S2C). Taken together, these results support the notion that the presence of active channels in the plasma membrane is required for spontaneous Ca²⁺ oscillations in GV oocytes.

Spontaneous cytoplasmic Ca²⁺ oscillations cause Ca²⁺ oscillations and stimulate metabolism in mitochondria

The physiological significance of the spontaneous Ca²⁺ oscillations in GV oocytes is poorly understood. A recent study in somatic cells uncovered a novel function for this constitutive, IP₃R-mediated Ca²⁺ release in maintaining cellular bioenergetics by transferring Ca²⁺ to the mitochondria (Cárdenas et al., 2010). The role of Ca²⁺ transfer between the ER and mitochondria has also been documented in mammalian MII oocytes, as the sperm-triggered Ca²⁺ oscillations during fertilization were shown to be sustained by Ca²⁺-driven ATP production (Dumollard et al., 2004; Wakai et al., 2013), which very likely was required for refilling of the [Ca²⁺]_ER by SERCA proteins. To address the roles of the spontaneous Ca²⁺ oscillations, we measured mitochondrial Ca²⁺ ([Ca²⁺]_mt) levels in GV oocytes. Radiometric chimeric fluorescent protein pericam harboring a mitochondrial targeting sequence (pericam-mito) has been successfully used to detect Ca²⁺ changes in the mitochondria of somatic cells (Nagai et al., 2001) and to show Ca²⁺ increases in mitochondria in fertilized mouse eggs (Dumollard et al., 2008). Pericam-mito cRNA was injected into GV oocytes and confocal images of pericam-mito fluorescence demonstrated it was successfully targeted to the mitochondria of GV oocytes, as it co-localized with mitochondrial staining provided by Mitotracker (Fig. 5A). To determine whether pericam-mito in GV oocytes could detect intra-mitochondrial oscillations, we examined Ca²⁺ responses in oocytes displaying spontaneous oscillations. As expected, the fluorescence intensities of the excitation wavelengths shifted in opposite directions with fluorescence intensities increasing at 480 nm (F480) with Ca²⁺ rises, whereas at the same time intensities at 410 nm (F410) decreasing (Fig. 5B). Importantly, these ratiometric measurements of [Ca²⁺]_mt (F480/F410) revealed that in GV oocytes, mitochondrial Ca²⁺ responses were oscillatory, which is consistent with the spontaneous cytoplasmic Ca²⁺ oscillations observed at this stage. In MII-arrested oocytes, contrarily, changes in [Ca²⁺]_mt levels were not observed, which is consistent with the lack of Ca²⁺ oscillations in unfertilized oocytes (Fig. 5C).

Fig. 5.

Mitochondrial Ca²⁺ uptake of IP₃R1-mediated Ca²⁺ release stimulates mitochondrial metabolism in GV oocytes. (A) Confocal images of radiometric pericam-mito fluorescence at 480 nm (F480, green) in GV oocytes, to detect mitochondrial Ca²⁺. To visualize colocalization with mitochondria, oocytes were stained with MitoTracker (red). Brightfield, BF. Scale bar: 20 µm. (B) The fluorescence intensities of pericam-mito in GV oocytes shifted in opposite directions between the excitation at 405 nm (left axis) and 480 nm (right axis) coinciding with Ca²⁺ oscillations. (C) Emission ratios (F480:F405) of pericam-mito, which allows estimation of the relative changes in mitochondrial Ca²⁺ ([Ca²⁺]_mt), were analyzed in GV (black trace; n=9/12) and MII oocytes (red trace; n=0/10) under normal external Ca²⁺ concentrations (1.7 mM). (D,E) Changes in NADH (left axis) and FAD (right axis) autofluorecence in GV (D; n=8/10) and MII oocytes (E; n=0/9) are reported. The intensities of NADH (360 nm wavelengths) and FAD (480 nm wavelengths) autofluorescence shifted in opposite directions in GV oocytes but not in MII oocytes, reflecting the stimulation of mitochondrial metabolism by the Ca²⁺ oscillations. (F) Representative traces of [Ca²⁺]_mt in GV oocytes (black trace; n=8/9) and adenophostin A (AdA)-injected GV oocytes (red trace; n=0/9) with reduced IP₃R1 expression (Fig. 2) are shown. [Ca²⁺]_mt measurements were performed under normal external Ca²⁺ concentrations, 1.7 mM, and 6 h after injection of AdA. (G) Comparison of ATP levels between GV oocytes and AdA-injected GV oocytes. The levels of ATP were estimated using the emission ratio of AT1.03 (YFP:CFP). Error bars represent s.d. Oocytes were injected at the GV stage and remained at this stage in media supplemented with IBMX.

A method to demonstrate how [Ca²⁺]_mt oscillations affect mitochondrial function is to evaluate changes in the oxidation status of nicotinamide adenine dinucleotide (NAD) and flavin adenine dinucleotide (FAD). The reduced form of NAD, NADH, and oxidized FAD display autofluorescence, which can be used to estimate the mitochondrial redox state (Duchen et al., 2003; Dumollard et al., 2004). NADH and FADH₂ are products of the TCA cycle that undergo oxidation during the process of oxidative phosphorylation, which occurs in the mitochondria and is stimulated by Ca²⁺ uptake. Time-lapse imaging of NADH (360 nm wavelengths) and FAD (480 nm wavelengths) demonstrates that, consistent with the oscillatory response in [Ca²⁺]_mt, oscillations in autofluorescence are detected in GV oocytes, whereas changes in autofluorescence are not observed in MII-arrested oocytes (Fig. 5D,E). These results suggest that at the steady state in GV oocytes, the Ca²⁺-sensitive mitochondrial dehydrogenases are activated by spontaneous, cytoplasmic and mitochondrial Ca²⁺ oscillations.

To directly assess the contribution of [Ca²⁺]_mt oscillations to cellular ATP levels, we expressed a fluorescence resonance energy transfer (FRET)-based ATP biosensor, ATeam AT1.03, which reports intracellular ATP concentrations (Imamura et al., 2009), with the emission ratio of AT1.03 fluorescence (YFP:CFP) used to estimate ATP levels. A conventional approach for real-time imaging of intracellular ATP levels in a single live cell is luminescence by luciferase, which has been used in attempts to show the dynamic changes in ATP levels during mouse oocyte maturation (Yu et al., 2010). However, the luciferase luminescence depends not only on the ATP level but also on multiple other parameters including oxygen, pH and luciferin. The expression of AT1.03 persists stably in the oocyte and has been successfully used to report intracellular ATP levels during maturation (Dalton et al., 2014; Wakai et al., 2015).

We compared ATP concentrations in uninjected oocytes vs oocytes injected with 10 μM AdA, which do not display Ca²⁺ oscillations. Accordingly, 6 h after injection of AdA and AT1.03 cRNA, we first monitored Ca²⁺ responses and confirmed that [Ca²⁺]_mt oscillations were inhibited in AdA-injected oocytes, whereas they were undisturbed in uninjected oocytes (Fig. 5F). We then found that in AdA-injected oocytes and without [Ca²⁺]_mt oscillations the ATP levels were significantly decreased (Fig. 5G). Taken together, we interpret these results to mean that spontaneous Ca²⁺ oscillations induce corresponding Ca²⁺ changes in the mitochondria, which subsequently stimulate ATP synthesis.

DISCUSSION

During maturation, mammalian oocytes undergo marked changes in Ca²⁺ homeostasis. The underlying mechanism(s) and function of these changes are not known. In the present study we report unique features of Ca²⁺ regulation in mouse GV oocytes: (1) they display constitutive Ca²⁺ release through IP₃R1 that mediates spontaneous Ca²⁺ oscillations and maintains the [Ca²⁺]_ER levels persistently low; (2) they show continuous Ca²⁺ influx from extracellular media that sustains the oscillations; (3) the spontaneous Ca²⁺ oscillations propagate into the neighboring mitochondria, activate mitochondrial metabolism and increase the production of ATP and ATP content. These results suggest that the Ca²⁺ changes observed in GV oocytes are associated with cellular homeostasis and may be a requirement for the acquisition of maturation and developmental competence.

[Ca²⁺]_ER levels and the ER Ca²⁺ leak through IP₃R1

This spatiotemporal complexity of intracellular Ca²⁺ responses is in part a consequence of the requirement for Ca²⁺ from two sources, extracellular medium and intracellular stores, to be integrated to produce stereotypical responses. To accomplish this, cells possess a variety of pumps, channels and buffering mechanisms in the plasma membrane and intracellular organelles, the ‘Ca²⁺ toolkit’, which make possible the generation of precise Ca²⁺ rises (Berridge et al., 2003). The ER is the main intracellular Ca²⁺ store ([Ca²⁺]_ER), and serves as the main source of Ca²⁺ in oocytes of most species. The majority of the uptake of Ca²⁺ into the ER is mediated by SERCA proteins, and we have previously demonstrated that mouse oocytes constitutively express the housekeeping SERCA2b protein (also known as ATP2A2) (Wakai et al., 2013). Therefore, given the steady expression SERCA2b in oocytes, it is unlikely that functional changes in this molecule can explain the increase in [Ca²⁺]_ER levels that begin around the GVBD stage and continue during maturation. We found instead that GV oocytes display a constitutive Ca²⁺ leak out of the ER whose regulation may more effectively impact [Ca²⁺]_ER levels. To uncover this leak, we used high concentrations of La³⁺ to insulate Ca²⁺ traffic in and out of the cell. We found that the active Ca²⁺ leak out of the ER in GV oocytes ceases around the GVBD stage, which is when [Ca²⁺]_ER starts to increase. We thus propose that regulation of this leak may control [Ca²⁺]_ER levels during oocyte maturation.

Although the molecular mechanism(s) underlying this Ca²⁺ leak remain undetermined, IP₃R1 may act as an ER leak channel in GV oocytes. Under basal conditions, several regulatory mechanisms can sensitize IP₃R1 including phosphorylation and thiol group modifications (Ivanova et al., 2014). With regards to phosphorylation, a candidate kinase is the protein kinase A (PKA) enzyme family, which phosphorylate IP₃R1 in somatic cells and increases the activity of the channel (DeSouza et al., 2002). A similar mechanism may operate in oocytes, as we have previously described PKA phosphorylation of IP₃R1 in GV oocytes, which shows a consistent and significant decrease at the time of GVBD (Wakai et al., 2012) coinciding with termination of the oscillations. Further, high levels of cAMP and PKA activity are needed to maintain arrest at the GV stage, and their reduction by the LH surge that induces the resumption of meiosis is also well established (Norris et al., 2009). It is therefore possible that IP₃R1 phosphorylation by PKA might be a mechanism involved in regulation of the ER Ca²⁺ leak and Ca²⁺ homeostasis during oocyte maturation. Whether modifications of IP₃R1 phosphorylation associated with anti-apoptotic effects by some members of the Bcl-2 family of proteins (Oakes et al., 2005) play any role in GV oocytes is unknown.

Changes in oxidation/reduction status during maturation may also modulate the ER Ca²⁺ leak. Thimerosal, a thiol oxidizing agent, is known to induce oscillations in mammalian GV and MII oocytes (Swann, 1991; Wakai et al., 2012). It is proposed that thimerosal accomplishes this by thiol modification and sensitization of IP₃Rs1. In fact, intracellular levels of reduced glutathione, which is the major antioxidant in the cell, increase during maturation (Dumollard et al., 2007), which may render GV oocytes more susceptible to oxidative stimuli. Lastly, besides IP₃R1, cells have at their disposal other passive Ca²⁺ leak mechanisms out of the ER, although their identity and regulation remain largely unknown (Ivanova et al., 2014).

The mechanism(s) of spontaneous Ca²⁺ oscillations in GV oocytes

In most cell types, especially in non-excitable cells, Ca²⁺ release from stores activates Ca²⁺ influx through plasma membrane channels, an observation that led to the discovery of SOCE (Parekh and Putney, 2005). We propose that persistently low [Ca²⁺]_ER levels and spontaneous Ca²⁺ oscillations in GV oocytes serve as the natural trigger for Ca²⁺ influx, which is required to sustain the oscillations. To confirm the role of Ca²⁺ influx, we modulated it by overexpressing STIM1 and ORAI1, the components of SOCE, in the presence of different concentrations of extracellular Ca²⁺. We found that high rates of Ca²⁺ influx persistently elevate basal Ca²⁺, whereas moderate levels of Ca²⁺ influx lead to generation of oscillations, and the absence of influx caused by a lack of Ca²⁺ in the extracellular media terminates the oscillations. These results demonstrate that Ca²⁺ influx is sufficient to stimulate oscillations in GV oocytes.

The underlying signaling pathway(s) whereby Ca²⁺ influx promotes these Ca²⁺ oscillations is not established. Oocytes are known to express PLC family proteins, especially PLCB1, which has been associated with oscillations in GV oocytes (Avazeri et al., 2000; Igarashi et al., 2007). Addition of the PLC inhibitor U73122 to oscillating GV oocytes terminated oscillations (Fig. S3A). Conversely, addition of the same concentration of U73433, which is the inactive analogue of U73122, did not block oscillations (Fig. S3B). It could be argued that increased cytosolic levels of Ca²⁺ caused by the enhanced influx stimulate inositol 1,4,5-trisphosphate (IP₃) synthesis, which induces Ca²⁺ release through IP₃R1. In the presence of basal IP₃ levels and given the biphasic regulation of the open probability of IP₃R1 by Ca²⁺ concentrations, Ca²⁺ release is amplified through CICR, maintaining spontaneous oscillations (Peres et al., 1991).

The channel(s) that underlie Ca²⁺ influx in these cells remains to be fully established. We promoted Ca²⁺ influx by overexpressing STIM1 and ORAI1, although recent genetic studies seem to suggest that SOCE is not involved in Ca²⁺ influx in mouse GV oocytes (Bernhardt et al., 2015). Importantly, a number of other plasma membrane permeable Ca²⁺ channels are expressed in mammalian GV oocytes, including the T-type voltage-gated Ca²⁺ channels CaV_3.2, which might underlie the increase in [Ca²⁺]_ER levels during maturation (Bernhardt et al., 2015). In this context, Ca²⁺ influx through voltage-gated Ca²⁺ channels is known to be the main trigger of CICR in excitable cells (Bers, 2002), and it could play a similar role in GV oocytes. Other channels expressed in GV oocytes are members of the family of transient receptor potential (TRP) channels, which are widely distributed in mammalian tissues. Mouse MII oocytes functionally express TRPV3 and although its expression is reduced in GV oocytes, it becomes progressively active during maturation (Carvacho et al., 2013). Another TRP channel, a TRPM7-like channel, is active in GV oocytes and spontaneous Ca²⁺ oscillations were reduced by the TRPM7 antagonist NS8593 (Carvacho et al., 2016). Therefore, there are several channels capable of mediating Ca²⁺ influx in GV oocytes. Future studies should determine which one(s) are active during the oscillations and how they are regulated.

The roles of mitochondrial Ca²⁺ oscillations in GV oocytes

The functional significance of Ca²⁺ oscillations in many systems remains to be fully elucidated, although it is generally accepted that their periodical behavior provides a digital signal to downstream effectors (Dupont et al., 2011). Owing to their proximity to the ER and IP₃R1 channels, mitochondria are a common downstream target of Ca²⁺ increases (Rizzuto et al., 1998; Rizzuto and Pozzan, 2006). Cytosolic Ca²⁺ reaches the mitochondrial matrix by permeating the outer mitochondrial membrane through the voltage-dependent anion-selective channel (VDAC) (Blachly-Dyson et al., 1993) and the inner mitochondrial membrane (IMM) via the Ca²⁺ selective mitochondrial calcium uniporter (MCU) (Baughman et al., 2011; De Stefani et al., 2011); the contact sites between the ER and mitochondria create localized sites of high Ca²⁺ concentration required for influx into the matrix. Besides contributing to Ca²⁺ homeostasis, the propagation of Ca²⁺ increases into the mitochondria is important for a variety of cell functions, ranging from ATP production to cell death (Rizzuto and Pozzan, 2006). In particular, basal uptake of Ca²⁺ by the mitochondria in resting cells is important to maintain NADH production to support oxidative phosphorylation (Cárdenas et al., 2010). Our data here are in line with this concept, as ligand-mediated degradation of IP₃R1, which inhibits the constitutive [Ca²⁺]_ER leak and eliminates Ca²⁺ uptake into the mitochondria, also significantly decreases ATP levels. Future studies should identify the molecular presence of these mitochondrial Ca²⁺ transporters and how they are regulated in mammalian GV oocytes.

Distinct regulation of Ca²⁺ homeostasis during oocyte maturation

The roles and regulation of Ca²⁺ homeostasis and mitochondrial output are seemingly different in GV and MII oocytes (Fig. 6). After fertilization, the function of the mitochondria is indispensable for Ca²⁺ oscillations because it supplies the ATP necessary for the Ca²⁺ pumps required to maintain Ca²⁺ levels in both the cytosol and in the ER (Dumollard et al., 2004; Wakai et al., 2013); a high [Ca²⁺]_ER level in MII oocytes is important for robust and long-lasting Ca²⁺ oscillations responsible for completing all events of egg activation and initiation of embryo development. Conversely, the spontaneous Ca²⁺ oscillations in GV oocytes and their counterpart oscillations in the mitochondria seem mostly important for cell bioenergetics to maintain the steady state that the meiotic arrest represents. Further, the low levels of [Ca²⁺]_ER in GV oocytes may lower the risk of apoptosis, as high [Ca²⁺]_ER levels render cells prone to apoptosis (Rong and Distelhorst, 2008). Namely, loss of the high [Ca²⁺]_ER content in ovulated MII oocytes appears to underlie the time-dependent aging process that sets in, which compromises these oocytes' developmental competence (Gordo et al., 2002). Lastly, it cannot be discounted that Ca²⁺ from the surrounding cumulus and granulosa cells enter GV oocytes through gap junctions, especially after stimulation with gonadotropins, and modify Ca²⁺ homeostasis in these cells (Homa, 1995; Flores et al., 1990), although the contributions of this regulation in GVs requires additional examination.

Fig. 6.

Schematic diagrams of the distinct regulation of Ca²⁺ homeostasis between GV and MII oocytes. In GV oocytes (left), Ca²⁺ influx stimulates Ca²⁺ oscillations by stimulating IP₃R1 and triggering CICR and/or activating PLC family proteins, leading to IP₃ synthesis. Note that a variety of Ca²⁺ channels in the plasma membrane including Cav3.2 T-type, TRP and/or ORAI1 channels are proposed to mediate the majority of this influx, although the plasma membrane channels that mediate Ca²⁺ influx during maturation and fertilization remain to be fully identified. The persistently low [Ca²⁺]_ER levels and spontaneous Ca²⁺ oscillations in GV oocytes serve as the natural trigger for Ca²⁺ influx through a store-operated Ca²⁺ influx pathway that is normally associated with expression of STIM1 and ORAI1, which is required to sustain the oscillations. Spontaneous Ca²⁺ oscillations are propagated into the neighboring mitochondria, activating mitochondrial metabolism and increasing the production of ATP. During oocyte maturation, around the GVBD stage, the putative active channel(s) are internalized, which causes downregulation of Ca²⁺ influx and termination of spontaneous Ca²⁺ oscillations, thereby allowing the increase in [Ca²⁺]_ER content during maturation. A high [Ca²⁺]_ER level in unfertilized (arrested) MII oocytes (middle) is important for robust and long-lasting Ca²⁺ oscillations responsible for the completion of all events of egg activation. After fertilization, the sperm-derived PLCζ (PLCZ1) triggers Ca²⁺ oscillations in MII oocytes (right). Mitochondria take up part of IP₃R1-mediated Ca²⁺ release, which is required to maintain [Ca²⁺]_ER levels and long-lasting Ca²⁺ oscillations to maintain ATP output that is necessary to support SERCA activity that in turn replenishes [Ca²⁺]_ER and sustains the oscillations.

In summary, the present study provides new insights into the mechanisms of Ca²⁺ homeostasis in GV oocytes. We found that cytosolic Ca²⁺ oscillations are propagated into the mitochondria and contribute to their metabolism. We also show that finely controlled Ca²⁺ influx regulates oscillations in GV oocytes. Lastly, we point to the distinct regulatory mechanisms of Ca²⁺ homeostasis in GV and MII oocytes that may fulfill stage-specific cellular functions. Future studies should define the channel(s) and the regulatory mechanisms that underpin Ca²⁺ oscillations in GVs, as well as the mitochondrial Ca²⁺ transporter(s) and their regulation during maturation and fertilization. Understanding and modulation of these molecular processes may lead to improvements in in vitro oocyte maturation and embryo development.

MATERIALS AND METHODS

Chemical reagents

Ionomycin and thapsigargin were purchased from Calbiochem (San Diego, CA). Other all chemicals were from Sigma-Aldrich (St Louis, MO) unless otherwise specified.

Collection of oocytes

GV oocytes were collected from the ovaries of 8- to 12-week-old CD-1 female mice. Females were injected with 5 IU of pregnant mare serum gonadotrophin (PMSG). Cumulus cell-enclosed GV oocytes were recovered 42–46 h post-PMSG administration, and the cumulus cells were removed by repeated pipetting. GV arrest was maintained where necessary by the addition of 100 mM 3-isobutyl-1-methylxanthine (IBMX) to the medium. Oocyte maturation was induced by removing IBMX, and oocytes were matured in vitro for 12–15 h in CZB medium under paraffin oil, at 37°C in a humidified atmosphere containing 5% CO₂. All animal procedures were performed according to research animal protocols approved by the University of Massachusetts Institutional Animal Care and Use Committee.

Plasmids

Human STIM1–YFP and ORAI1 were generously provided by Tobias Meyer (Stanford University, Stanford, CA) and Mohamed Trebak (Albany Medical College, Albany, NY), respectively. STIM1–YFP was subcloned into a pcDNA6/Myc-His B vector (Invitrogen) between the restriction sites AgeI and XbaI. The ORAI1 insert was amplified by PCR and ligated to the N-terminus of the mRFP-bearing pcDNA6/Myc-His B vector (Dominique Alfandari, University of Massachusetts, Amherst, MA) between EcoRI and XhoI restriction sites. The SNAP25 mutant (SNAP25^Δ20) sequence was amplified by PCR from the mouse SNAP25 expression vector (Origene) and subcloned into a pcDNA6/Myc-His B vector between the EcoRI and XhoI restriction sites. The sequences of the forward and reverse primers were, respectively, 5′-CGGAATTCGCCACCATGGCCGAAGACGCAGACATG-3′, 5′-GCTCGAGTTAATCAGCCTTCTCCATGAT-3′. ATP biosensor, AT1.03 and mitochondrial Ca²⁺ probe, pericam-mito, in pcDNA3.1 vector were kindly provided by Dr Hiromi Imamura (Kyoto University, Japan) and Dr Atsushi Miyawaki (RIKEN Center for Brain Science, Japan).

Preparation and microinjection of cRNA

Plasmids were linearized with a restriction enzyme downstream of the insert to be transcribed. Capped DNA was transcribed in vitro using the T7 mMESSAGE mMACHINE Kit (Themo Fisher Scientific, MA) using the promoter that was contained in the constructs. A poly(A)-tail was added to the transcribed RNAs using a Tailing Kit (Themo Fisher Scientific), and poly(A)-tailed RNAs were eluted with RNase-free water and stored in aliquots at −20°C. For microinjection, cRNA solution was loaded into glass micropipettes at a concentration of 500 ng/l and delivered into oocytes by pneumatic pressure (PLI-100, Harvard Apparatus, Cambridge, MA). The volumes injected typically ranged from 2 to 10 pl, which is 1–5% of oocyte volume (250–300 pl).

Live cell imaging

To measure cytoplasmic Ca²⁺, oocytes were incubated with 1.25 mM Fura-2 (Themo Fisher Scientific) supplemented with 0.02% pluronic acid (Themo Fisher Scientific) for 20 min at room temperature. Fura-2-loaded oocytes were attached to glass-bottomed dishes (MatTek Corp., Ashland, MA) and placed on the stage of an inverted microscope. Fura-2 fluorescence was excited with 340 nm and 380 nm wavelengths and emitted light was collected at wavelengths above 510 nm. Radiometric measurement of pericam-mito is efficiently used as a mitochondrial-specific Ca²⁺ probe (Nagai et al., 2001). To estimate [Ca²⁺]_mt levels, emitted light was collected after dual excitation at 410 nm and 480 nm wavelengths. The confocal fluorescence images of pericam-mito were obtained using a laser-scanning confocal microscope (LSM 510 META, Carl Zeiss Microimaging Inc., Germany) fitted with a 63×1.4 NA oil-immersion objective lens. Images were acquired with LSM software (Carl Zeiss). ATeam AT1.03, a fluorescence resonance energy transfer-based ATP indicator, has been used successfully to measure cellular ATP levels in live somatic cells (Imamura et al., 2009). To estimate the relative changes in ATP levels, the emission ratio of AT1.03 (YFP:CFP) was imaged using a CFP excitation filter, dichroic beam splitter, and CFP and YFP emission filters. CFP and YFP intensities were collected every 20 s by a cooled Photometrics SenSys CCD camera (Roper Scientific, Tucson, AZ).

Western blot analysis

Cell lysates from mouse oocytes were prepared by adding 2× sample buffer [0.125 M Tris-HCl, pH 6.8; 4% (w/v) SDS; 20% (w/v) glycerol; 0.01% (w/v) Bromophenol Blue; 10% (w/v) β-mercaptoethanol]. Proteins were separated by SDS-PAGE and transferred to PVDF membrane (Millipore, Bedford, MA). After blocking, membranes were probed with the rabbit polyclonal antibody specific to IP₃R1 (1:1000; a generous gift from Dr Jan Baptiste Parys, Katholieke Universiteit, Leuven, Belgium; Parys et al., 1995). Goat anti-rabbit antibody conjugated to horseradish peroxidase (HRP) was used as a secondary antibody (1:2000; STAR124P; Bio-Rad, Hercules, CA) for detection of chemiluminescence using a Western Lightning Plus-ECL kit (NEN Life Science Products, Boston, MA) according to the manufacturer's instructions. The signal was digitally captured using a Kodak 440 Imaging Station (Rochester, NY). The same membranes were stripped at 50°C for 30 min (62.5 mM Tris, 2% SDS and 100 mM 2-beta mercaptoethanol) and re-probed with anti-α-tubulin monoclonal antibody (1:1000; T-9026; Sigma-Aldrich, St Louis, MO) to detect tubulin.

Statistical analysis

Values from three or more experiments performed on different batches of oocytes were analyzed by Student's t-test or one-way ANOVA followed by Fisher's protected least significant difference test as appropriate. Differences were considered significant at P<0.05. Significance among groups or treatments is denoted in bar graphs by different superscripts or by the presence of asterisks.