A localized calcium transient and polar body abscission

ABSTRACT

Polar body emission is a special form of cytokinesis in oocyte meiosis that ensures the correct number of chromosomes in reproduction-competent eggs. The molecular mechanism of the last step, polar body abscission, is poorly understood. While it has been proposed that Ca²⁺ signaling plays important roles in embryonic cytokinesis, to date transient increases in intracellular free Ca²⁺ have been difficult to document in oocyte meiosis except for the global Ca²⁺ wave induced by sperm at fertilization. Here, we find that microinjection of the calcium chelator dibromo-BAPTA inhibits polar body abscission in Xenopus laevis oocytes. Using a novel, microtubule-targeted ratio-metric calcium sensor, we detected a calcium transient that is focused at the contractile ring-associated plasma membrane and which occurred after anaphase and constriction of the contractile ring but prior to abscission. This calcium transient was confirmed by mobile calcium probes. Further, the Ca²⁺-sensitive protein kinase Cβ C2 domain transiently translocated to the contractile ring-associated membrane simultaneously with the calcium transient. Collectively, these results demonstrate that a calcium transient, apparently originating at the contractile ring-associated plasma membrane, promotes polar body abscission.

KEYWORDS:

• Xenopus oocytes
• meiosis
• polar body
• abscission
• calcium

Introduction

In animal cells, cytokinesis begins with the formation of an actomyosin based-contractile ring tethered to the plasma membrane. The contractile ring constricts, pulling the associated plasma membrane inward to eventually form the intercellular bridge, a thin cell process connecting the two nascent daughter cells. Cytokinesis ends with abscission, the severing of the intercellular bridge. In the middle of the intercellular bridge is the midbody, a bundle of tightly packed microtubules associated with a variety of abscission regulators [1]. In human mitotic cells, it is currently thought that abscission is accomplished by secondary ingression that is not powered by the contractile ring. Rather, secondary ingression is thought to be catalyzed by the assembly and contraction of filaments built from the ESCRT-III (endosomal sorting complex required for transport-III) complex [2,3]. The midbody-localized ESCRT-III also provides the platform (via its CHMP1B subunit) to recruit microtubule severing protein spastin in an apparent coordination between this secondary constriction and midbody microtubule severing [4]. Other members of the ESCRT-III complex may serve to receive positive and negative signals for abscission execution (below).

Mitotic abscission is controlled by the NoCut abscission checkpoint, which senses midbody defects (e.g. presence of lagging chromosomes) and delays abscission until the defects are resolved [5]. Midbody-associated Aurora B plays a critical role in abscission checkpoint [6], by phosphorylating many abscission regulators including the mitotic kinesin-like protein 1 (Mklp1) to prevent furrow regression [7] and ESCRT-III’s CHMP4C subunit, to delay abscission [8,9]. The function of Aurora B in preventing furrow regression is conserved in oocyte meiosis, at least in worms [10] and ascidians [11].

On the other hand, phosphatidylinositol-3 kinase (PI3K) pathway appears to promote cytokinesis abscission though CHMP4B. A centrosomal protein complex of the PI3P binding protein FYVE-CENT and its partner TTC19 are translocated to the midbody by the kinesin KIF13A where TTC19 binds CHMP4B to effect abscission [12]. Interestingly, a recent study reported that depletion of Beclin-1, a component of class III PI3K, and chemical inhibition of PI3K inhibit polar body abscission in mouse oocytes [13].

In addition to the ESCRT-III constriction machinery and accessory proteins, Rab GTPase-mediated membrane vesicle trafficking and SNAREs (soluble N-ethylmaleimide sensitive factor attachment protein receptor)-mediated vesicle-plasma membrane fusion play critical roles in abscission [14,15]. Membrane fusion in other systems is inhibited by intracellular calcium chelation, suggesting the potential involvement of transient, local elevation of intracellular free calcium [16]. However, whether calcium transient has a direct role in mitotic abscission is unknown.

Relative to mitosis, abscission during vertebrate oocyte meiosis is very poorly understood [17]. That is, while the highly asymmetric divisions which produce the two polar bodies begin in a manner that resembles mitotic cytokinesis, with formation and closure of a Rho-dependent actomyosin ring [18], far less is known about abscission, in spite of the fact that failed abscission in oocyte meiosis results in production of eggs with supernumerary chromosomes, a condition that results in infertility [19]. Indeed, we currently lack such basic information as the timing of polar body abscission in mice or Xenopus, partly because a severed polar body remains trapped between the egg and the overlaying vitelline membranes so the status of abscission cannot be easily determined [20].

In this study, we have examined the timing of polar body abscission during Xenopus oocyte meiosis, based on the ability of fluorescent dextran to enter the polar body when deposited in the oocyte. Guided by the timing of abscission, we have discovered a short-lived but robust calcium transient that occurs after cytokinetic apparatus ingression but before polar body abscission. Further, we have determined that the calcium transient is focused at the plasma membrane in the approximate center of the intercellular bridge. Finally, we demonstrate that intracellular calcium buffering following anaphase inhibited polar body abscission.

Results

Progesterone treatment triggers meiotic maturation in Xenopus oocytes with the time to germinal vesicle breakdown (Figure 1(a), GVBD, time = 0 min) being relatively variable, but with the time from GVBD to the later steps of meiosis being more consistent [18,20]. Within 30 minutes of GVBD, a nascent meiosis I spindle with condensed chromosomes can be seen at the cortex in the animal pole. Approximately 120 minutes after GVBD, oocytes abruptly initiate anaphase with half the chromosome homologs being pulled into the protruding polar body enclosure (Figure 1(a) time = 120 min). Fifteen (15) minutes after anaphase initiation, the first polar body is fully protruded, and the contractile ring appears fully constricted [18,21]. Within 30 minutes of anaphase-I initiation, the metaphase II spindle forms and associates stably with the cortex, typically near the first polar body. It is unknown when first polar body abscission occurs since it is trapped between the plasma membrane and the overlaying vitelline membrane [20]. To develop an assay for first polar body abscission, we exploited the fact that impalement of metaphase II oocytes triggers artificial completion of meiosis such that second polar body is emitted as the metaphase arrest is relieved. Specifically, we microinjected fluorescent dextran into the oocytes, thereby introducing a fluorescent marker of cytoplasmic continuity into the oocytes while simultaneously triggering emission of the second polar body (Figure 1(a)). Because the dextran is present at the time the second polar body is formed, the second polar body will inevitably be labeled by it. For the first polar body, however, dextran will only be contained within it if its cytoplasm was continuous with that of the oocyte at the time of the injection, that is, if abscission had not yet occurred. Thus, if both polar bodies contain dextran, first polar body abscission had not yet occurred; if only one of the polar bodies contains dextran, first polar body abscission had occurred. Preliminary experiments using the dextran microinjection assay indicated that most oocytes had completed first polar body abscission within 3 hours following GVBD (not shown).

Figure 1. Buffering intracellular calcium in Xenopus oocytes inhibited polar body abscission.(a) Timeline of oocyte maturation and polar body abscission assay. Details described in the text. Chromosomes in green and dextran in red. The haploid chromosomes in activated eggs (far right) are typically too deep in the cytoplasm to be imaged. pb1: first polar body; pb2: second polar body. (b) Representative images depicting abscission assays of eGFP-H2B expressing oocytes injected with fluorescent dextran at the indicated time (minutes; above the images) from anaphase initiation. Dextran fluorescence ratio of the first polar body (1^st) over the second polar body (2^nd) is indicated below the images. Whereas first polar bodies always contained chromatin mass, individual chromosomes were often visible in second polar bodies. (c) Summary (means ± SD) of dextran fluorescence ratio of oocytes injected with dextran at the indicated time from anaphase initiation. P values represent One-way ANOVA and Tukey’s HSD test results. A horizontal line with two vertical ticks at the ends indicate comparison between the two end time points while a horizontal line covers all time points as indicated. Also indicated are the percentage of oocytes in each time point that had completed abscission. (d) Dextran fluorescence ratio (1^st polar body/2^nd polar body or metaphase II spindle; means ± SD) of oocytes injected with water (control) or DB-BAPTA 5 min after anaphase initiation. P = 1.2E-08 by t-test (two-tailed). The graph shows a DB-BAPTA-injected oocyte, depicting 1^st polar body containing dextran and metaphase II chromosomes scattered at the cortex.

Figure 1. (Continued).

To more precisely determine the timing of polar body abscission, we carried out dextran microinjection experiments at various time points following anaphase initiation, based on chromosome separation [18] (Figure 1(b)). To assess abscission quantitatively, we measured the ratio of dextran fluorescence of first polar body over that of the second polar body as an indicator of dextran permeability. Under our experimental conditions, a fluorescence ratio of less than 0.1 reflected background autofluorescence in the first polar body, as compared to oocytes not injected with fluorescence dextran. Based on this criterion, none of the oocytes (0/47) had completed abscission when dextran injection was performed 20 minutes or less after anaphase initiation (Figure 1(c)). Three oocytes out of 10 (30%) had completed abscission at 25 minutes after anaphase initiation. In contrast, 73% (11/15) had completed abscission at 30 minutes after anaphase initiation. Similar proportions of oocytes had completed abscission when examined 40 to 60 minutes after anaphase initiation (Figure 1(c)). Overall, 28/41 (68%) of the oocytes examined at 30 minutes or longer after anaphase initiation had completed abscission. The relatively large error bars, especially at 30 minutes or later, reflected the significant proportions of the oocytes that failed abscission. These results suggest that first polar body abscission occurred between 25 and 30 minutes after anaphase initiation.

To determine whether polar body abscission is regulated by Ca²⁺ signaling, we injected dibromo-BAPTA (DB-BAPTA) or water (as a control) into the oocytes 5 minutes after anaphase initiation (indicated by chromosome separation) [22], and then subjected eggs to the polar body abscission assay. Water-injected oocytes exhibited dextran fluorescence ratio indicative of normal abscission. In contrast, DB-BAPTA-injected oocytes clearly failed polar body abscission (Figure 1(d)). In addition, DB-BAPTA-injected oocytes failed to form a second polar body, but instead exhibited metaphase II chromosomes scattered at the cortex (Figure 1(d)), as result of DB-BAPTA disrupting metaphase II spindles, consistent with previous results [22].

With the exception of the fertilization-induced calcium increase [23], transient increases in Ca²⁺ during meiotic maturation in oocytes have proven elusive [22,24–26]. It has been speculated that this is due to shortcomings in the spatial or temporal resolution of the imaging regimes previously employed, an argument which seems plausible, particularly in the Xenopus system, where the oocyte is ~1 mm diameter and given the fact that meiotic maturation occurs over the course of several hours (see above). The remodeling and compartmentalization of the plasma membrane and cortex that accompany meiotic maturation in general and polar body emission in particular further complicate matters, in that the amount of yolk-free cytoplasm available for soluble calcium probes changes dramatically over space and time. Our previous attempts [22], using untagged GCaMP3 and membrane-anchored Lck-GCaMP3, have failed to detect any calcium transients during polar body formation, although we had mainly focused on spindle assembly and metaphase to anaphase initiation. GCaMP3 exhibits brighter fluorescence signal at the spindle, likely reflecting enrichment of the probe in the yolk-free spindle region. Lck-GCaMP3, on the other hand, exhibits bright fluorescence around the neck of the forming polar body. However, as the calcium-insensitive Lck-mCherry exhibits similarly increasing fluorescence, the increase of Lck-GCaMP3 fluorescence appears to reflect increase of membrane accumulation at the polar body neck [22]. To overcome these issues, we imaged the animal pole at high spatiotemporal resolution using a novel, ratio-metric probe for microtubule-proximal free calcium – TubeCaMP3 (Figure 2(a)). TubeCaMP3 comprises the microtubule binding domain of ensconsin (EMTB; [27]) flanked by an N-terminal mCherry and a C-terminal fluorescent calcium sensor, GCaMP3 [28]. EMTB has been used successfully to image microtubules in Xenopus oocytes [20] and embryos [29], while GCaMP3 has been employed to monitor calcium in a wide variety of cell types [28], including Xenopus oocytes [30]. The rationale for this probe is simple: by comparing the ratio of fluorescent signal from mCherry to that of the GCaMP3, a measure of microtubule-proximal intracellular free calcium is obtained. This is particularly useful for imaging intracellular free calcium in the regions of spindles, which are relatively rich in cytoplasm (and thus, more conventional cytoplasmic calcium probes) due to their tendency to exclude yolk platelets [31].

Figure 2. A local calcium transient during polar body emission detected by TubeCaMP3.

(a) Construct design and working principle of TubeCaMP3. Details are described in the text. (b) A time series exhibiting the response of spindle microtubule (red circle: spindle cross section) fluorescence to UV photolysis of caged-IP₃ in TubeCaMP3-expressing oocytes. Time in minute:second. (c) A line plot summarizes the TubeCaMP3 (TC3) fluorescence ratio (F_TC3-GCaMP3/F_TC3-mCherry) change in (b). Means ± S.E.M; n = 10. (d) TubeCaMP3 signals during polar body emission in Xenopus oocytes. Shown are selective still frames of a 3D-reconstructed confocal time series. A transient Ca²⁺ elevation is detected ~12 min after anaphase onset (arrows, depicting separating chromosomes). Green: TC3-GCaMP3; red: TC3-mCherry; blue: Hoechst (DNA). Time in minute:second from anaphase onset (00:00). (e) A line plot summarizes the fluorescence ratio changes of TubeCaMP3 subunits (F_TC3-GCaMP3/ F_TC3-mCherry) in (d). Means ± S.E.M; n = 9. (9 out of 16 TubeCaMP3-expressing oocytes exhibited the Ca²⁺ transient.)

To test whether TubeCaMP3 reports on microtubule-proximal increases in intracellular free calcium, oocytes expressing TubeCaMP3 were injected with caged-inositol trisphosphate (IP₃) [32], treated with progesterone to induce meiotic maturation, and then subjected to UV photolysis at the meiotic spindle. Following UV exposure, TubeCaMP3ʹs GCaMP3 fluorescence signal increased in response to IP₃ release while its mCherry fluorescence signal remained unchanged (Figure 2(b), time in minute:second). Hence the change of relative fluorescence of TubeCaMP3 (GCaMP3/mCherry ratio) at the spindle region reflects the fluctuation of microtubule proximal Ca²⁺ (Figure 2(c)).

To detect possible calcium transients during polar body emission, we carried out live imaging of Xenopus oocytes expressing TubeCaMP3 and labeled with Hoechst dye (to monitor chromosome dynamics). Imaging began at the onset of anaphase and continued for at least 20 minutes. Remarkably, in 9/16 oocytes, an abrupt and transient increase in GCaMP3 fluorescence was detected that coincided with the intercellular bridge and the regions flanking it, on both the polar body and the oocyte side (Figure 2(d), 12:00; The complete series of this and another oocyte are depicted in supplementary Fig. S1). This calcium transient commenced within ~12 minutes of anaphase onset and, a minute later, calcium signal spread to the polar body but was lost from the oocyte side (Figure 2(d), 13:00). Then, just as abruptly, intracellular free calcium returned to resting levels (Figure 2(d), 14:00). In the rest of the oocytes (7/16), we detected no increase of GCaMP3 fluorescence, even when imaging was extended to 60 minutes after anaphase onset.

The results obtained with TubeCaMP3 indicated the existence of a highly focal calcium transient associated with the intercellular bridge that forms prior to abscission of the first polar body. To obtain independent confirmation of this result, we employed two different chemical Ca²⁺ indicators, Oregon Green^TM 488 BAPTA-2 (OG-2; K_d for Ca²⁺ ~580 nM) and Oregon Green^TM 488 BAPTA-6 f (OG-6 f; K_d for Ca²⁺ ~3 µM) [33]. Oocytes injected with OG-2 exhibited a sudden rise of calcium ~12 minutes after the onset of anaphase (Figure 3(a), 19:50; the graph depicts quantification of the entire series, which is available in supplementary Fig. S2). Elevated calcium was seen in the protruded polar body and the oocyte cortex immediately adjacent to the polar body (Figure 3a, “near”) but not oocyte cortex further away from the polar body (Figure 3(a), “far”). Furthermore, whereas the increased OG-2 signal in the oocyte (‘near”) faded within a minute or so, elevated OG-2 signal in the polar body lasted 2–3 minutes (Figure 3(a)). Consistent with the results obtained using TubeCaMP3, we found that in 8/14 (OG-2) (Figure 3(b,c)) or 10/15 (OG-6 f) (Supplementary Fig. S3) this polar body associated calcium elevation could be detected. These results were obtained by imaging a single plane, at 3 sec intervals, encompassing both the polar body and a region of the oocyte cortex corresponding to “far” in Figure 3(a) (circle in Figure 3(b), 00:00). These data (Figure 3a, b, and c) clearly showed that a single calcium transient occurred and it was restricted to the immediate vicinity of the forming polar body, unlike the global calcium transient at fertilization [23].

Figure 3. The local calcium transient detected by Oregon Green^TM 488 BAPTA-2.

(a) Selective frames of a confocal time series (representative of three oocytes) depicting the transient rise of calcium in the polar body (pb; reconstructed xz-plane) and the immediately adjacent oocyte cortex (“near”) but not oocyte cortex further away (“far”). Anaphase onset was noted at 07:30 (minutes:seconds; inset depicting chromosome stretching, compared to that of 6:00). The xy-planes represent oocyte surface (directly beneath the polar body).The graph depicts fluorescence changes in the polar body (or the cortex directly above the spindle before polar body protrusion), the immediately adjacent oocyte cortex (“near”) and oocyte cortex about 30 µm away from the polar body (“far”). Arrow depicts the time of anaphase initiation. (b) Selective frames of a time series show specific rise of OG-2 signals in the polar body but not in the oocyte cortex away from the polar body, corresponding to “far” in Figure 3(a). Time in minute:second. The entire series of this oocyte is available as supplementary Movie S1, with time 0 corresponding to 03:57 in the movie. (c) A line plot summarizes the fluorescence change of OG-2 at the polar body region (red line) and oocyte cortex corresponding to “far” in Figure 3(a) (black line). Means ± S.E.M.; n = 8. (8 out of 14 OG-2-injected oocytes exhibit the Ca²⁺ transient.)

While the chemical dyes permit rapid imaging of intracellular increases in calcium, they produce diffuse signal, making it difficult to discern any potential localized pattern of calcium increase. We therefore imaged fluorescent versions of the C2 domain of protein kinase Cβ (PKC-β-C2). The C2 domain binds to phosphatidylserine (PS) in the presence of elevated calcium [34] and, because PS is most abundant in the plasma membrane, fusions of the C2 domain with fluorescent proteins have been used to monitor plasma membrane and calcium dynamics in processes ranging from exocytosis [35] to cell [30] and tissue [36] repair. As above, oocytes expressing eGFP-PKCβ-C2 and H2B-mRFP were subjected to time-lapse confocal imaging close to metaphase-to-anaphase transition. Consistent with the results obtained with TubeCaMP3 and the chemical calcium dyes, eGFP-PKCβ-C2 (or mRFP-PKCβ-C2) was transiently recruited, in 17/24 oocytes (71%), to what appeared to be the polar body neck ~12 minutes after anaphase onset (Figure 4(a); The complete series of this oocyte is depicted in supplementary Fig. S4). Furthermore, the eGFP-PKCβ-C2 revealed an additional feature of the calcium transient not detected by the other probes: concentration at the approximate middle of the intercellular bridge. To more precisely identify the site of eGFP-PKCβ-C2 concentration, we compared its distribution to that of a plasma membrane marker Lck-mCherry [22]. This revealed that C2 was indeed concentrated at the plasma membrane in the narrow neck separating the oocyte from the forming polar body (Figure 4(b)). As expected, injection of DB-BAPTA but not water 5 minutes after anaphase initiation, abolished the C2 recruitment to the polar body neck (Figure 4(c); p = 0.019).

Figure 4. PKCβ-C2 distribution during the Ca²⁺ transient.

(a) Selective frames of a representative oocyte expressing mRFP-H2B and eGFP-PKCβ-C2 depicting the transient rise of PKCβ-C2 signals between two sets of chromosomes during polar body emission. Chromosome separation was shown in insets (Time in minute:second). (b) Top:Selective frames of the horizontal cross section (xy-plane view) of the intercellular bridge (polar body “neck”) during eGFP-PKCβ-C2 signal (green) rise (arrows). Time in minute:second from the start of imaging. The entire series is available in Supplementary Movie S2 (n = 3).Bottom:The same time frames shown in vertical cross section (reconstructed xz-plane view). (c) Oocytes were injected with water or DB-BAPTA 5 minutes after anaphase initiation followed by live imaging to determine the presence (yes) or absence (no) of eGFP-PKCβ-C2 recruitment to the polar body neck. Summary of six independent experiments with the total oocyte numbers indicated. P = 0.019 by Chi-square (two-tailed). (d) Selective frames of a time series exhibiting the colocalization of Anillin-GFP (green, marking the contractile ring) mRFP-PKCβ-C2 (red) during the C2 recruitment. Time in minute:second from the start of image acquisition. The entire series is available in Supplementary movie S3 (n = 3).

Figure 4. (Continued).

The concentration of C2 at the neck separating the nascent polar body from the oocyte, combined with the results obtained with TubeCaMP3 and OG-2 (and OG-6 f), suggest that calcium may be focally elevated in a narrow ring of plasma membrane closely apposed to the contractile ring. To test this, we compared the spatiotemporal dynamics of C2 to anillin, a protein that serves as a link to the plasma membrane and the contractile ring [37] in both mitotic [38] and meiotic cytokinesis [39]. Comparison of the localization pattern of C2 to anillin showed that the C2 concentrated in the region where anillin was most abundant, as expected if calcium is focally elevated at the plasma membrane overlying the contractile ring (Figure 4(d)). A second finding emerged from this analysis: C2 was not recruited to the plasma membrane of the neck until after closure of the contractile ring had essentially stopped, as expected if the calcium is primarily involved in regulation of the late stages of cytokinesis.

Discussion

In this study we have demonstrated a calcium transient in Xenopus oocytes that is spatiotemporally correlated with polar body abscission, the final step in polar body emission. The calcium transient occurred at ~12 minutes after anaphase initiation and at a point when the contractile ring had fully constricted, but before abscission which occurs 25–30 minutes after anaphase initiation (Figure 1). Two lines of evidence indicate that this calcium transient is essential for abscission: first, the proportion of oocytes that undergo abscission (68%) roughly parallels the proportion that display the calcium transient (~60%; TubeCaMP3, OG-2 and OG6f) or PKCβ-C2 recruitment (71%); second, and more tellingly, microinjection of DB-BAPTA before the anticipated time of abscission prevents abscission. We were unable to determine directly if the oocytes showing calcium transient were the same oocytes that completed abscission. The experiment would involve confocal imaging of live oocytes for anaphase initiation (chromosome segregation) followed by time lapse calcium imaging, after which the oocyte needs to be transferred to a nearby apparatus for fluorescent dextran injection before transfer back for imaging fluorescence dextran. Oocytes become very fragile, especially after calcium imaging since it requires relatively high laser intensity and short intervals (therefore many timepoints) to capture the brief “burst”, and easily damaged during transfers. Furthermore, our lab is equipped with an aged MRC 1024 confocal system capable of imaging only two fluorophores, green and red fluorescence, making it impossible to do this experiment which would require three fluorophores: a chromosome probe, a calcium probe, and fluorescent dextran.

Where does the calcium come from? Based on the results obtained with the various probes, the transient itself is focused on the plasma membrane overlaying the contractile ring. That is, the GCaMP3/mCherry ratio of TubeCaMP3 increased on both sides of the contractile ring (restricted point of the spindle in Figure 2(d), 12:00) with signal gradients consistent with the calcium origin being at the level of the plasma membrane. Moreover, the mobile calcium probe OG-2 detected simultaneous elevation of calcium in the polar body and the oocyte cortex immediately adjacent to the polar body (within ~30 µm radius). The rapid rise and sharp decline of calcium at oocyte cortex near the polar body (Figure 3(a,b), “near”) and the equally rapid rise but slower decline of calcium in the polar body are consistent with a common calcium source (i.e. at the polar body neck). The asymmetric dissipation of free calcium, fast on the oocyte side and slower in the polar body, likely reflects the huge difference in cytoplasmic volumes, hence calcium diffusion rate/buffer capacity, between the two (125,000:1, based on diameters of 20 µm and 1 mm for polar body and oocyte respectively). Finally, and most importantly, the calcium sensitive C2 probe, the only one of the probes that can directly report on plasma membrane-associated calcium increase, was transiently recruited to the region of the intercellular bridge where the plasma membrane is closely apposed to the contractile ring. That this probe is reporting on the same calcium transient detected by the other probes is indicated by the similar timing of and duration of its recruitment.

It is unknown what triggers the highly restricted calcium transients detected here. Parry et al., have reported microdomain, inositol 1, 4, 5-trisphosphate (IP₃)-mediated calcium transients associated with each mitotic spindle in syncytial Drosophila embryos [40,41]. The close contact between spindles and the endoplasmic reticulum (ER) has been noted in many cell types [40,42,43] and in Xenopus oocytes (our unpublished data). It is therefore plausible that the spindle may be involved in triggering local calcium transients. We speculate that these calcium transients are dependent on the unique interaction between the spindle apparatus and the oocyte cortex/plasma membranes. For examples, the spindle including the segregating chromosomes may be associated with a non-membrane phospholipase C (PLC), hydrolyzing phosphatidylinositol trisphosphate (PIP₃) releasing IP₃, hence calcium transients, locally. Such non-membrane PLC has been reported to be associated with yeast kinetochores [44] and in mammalian sperms [45].

What might be the function of the calcium transient? The timing of the calcium transient, ~12 minutes after anaphase initiation when the contractile ring had fully constricted, suggests that it is unrelated to spindle assembly [22] or contractile ring constriction [46]. In contrast, the timing and localization of the transient are consistent with a role during abscission, as is the demonstration that abscission is blocked by DB-BAPTA injection. We therefore suggest that the calcium transient either plays a critical role in a process leading to abscission or in the abscission event itself. Previous results indicate that the spindle midbody forms approximately 14–16 minutes after anaphase initiation [18], within the window of BAPTA sensitivity shown in this study. Given that the midbody is essential for abscission during mitosis, and given the localization of the calcium transient, it is plausible that the target of the calcium increase is a component(s) of the midbody. In this regard, it has been reported that in HeLa cells calcineurin, a calcium- and calmodulin-dependent protein phosphatase, accumulates at the midbody. Inhibition of calcineurin or calcium chelation blocks cytokinesis abscission [47]. Another attractive (and nonexclusive) possibility is the ESCRT machinery. Specifically, it has been reported that experimental elevation of calcium promotes the recruitment of ALG2 and ESCRTIII to the plasma membrane [48]. Finally, abscission entails membrane-membrane fusion events [49] which in other contexts are triggered by intracellular calcium increases [16].

Method

Xenopus oocyte preparation

Sexually mature Xenopus laevis females were obtained from Nasco (Fort Atkinson, Wisconsin). The animals were housed in the animal care center of University of Ottawa, Faculty of Medicine. The experimental protocol involving the use of these animals has been approved by the Animal Care Committee of University of Ottawa (OHRI-2092). The females were primed with 50 units of pregnant mare gonadotropin (Sigma-Aldrich, P0130) 3–10 days prior to sacrifice. Stage VI oocytes were manually defolliculated with fine forceps in oocyte culture medium (OCM, 480 mL Leibovitz’s L-15 medium, 320 mL sterile water, 0.32 g bovine serum albumin, 0.5 mg/mL gentamycin, pH 7.6–7.8) and kept at 18°C until use [50]. Only stage VI oocytes with uniform coloration were chosen for the experiments, and randomly allocated among experimental groups since they are morphologically indistinguishable. For each experiment reported here, at least three independent repeats were performed each with a separate donor female.

Time-lapse confocal fluorescence imaging

To induce maturation, oocytes were incubated in OCM plus 1 μM progesterone in room temperature. Oocytes were monitored, every 10 minutes, for the appearance of a white “maturation spot” on the animal hemisphere, indicative of germinal vesicle breakdown (or GVBD). GVBD oocytes were transferred to fresh OCM. Anaphase I initiation commences approximately 120 min following GVBD [21]. Oocytes were accordingly placed in the imaging chamber (a glass bottom petri dish with a cover slip to “clamp” oocytes; [51]) prior to anaphase initiation. Anaphase initiation was monitored by time lapse imaging, based on chromosome separation (or “stretching”) as shown in the figures. All post-anaphase time points reported here were based on time lapse imaging of individual oocytes. Imaging was acquired with a 63x oil objective on a Zeiss Axiovert microscope with a Bio-Rad MRC-1024 laser-scanning confocal system in our lab (Figure 1(b,d); Figure 2(b); Figure 3(a,b); Figure 4(a, c, and d)) or a Leica DMI6000B microscope with a Quorum spinning disk confocal system at the imaging core of University of Ottawa (Figure 2(d); Figure 4(b)). The Bio-Rad MRC-1024 is equipped with two laser lines at 488 nm and 568 nm respectively. The fluorescence filters are 522/35 nm (for 488 nm excitation) and 605/32 nm (for 568 nm excitation). We used simultaneous image acquisition mode in experiments involving two fluorophores (Figure 1(b,d); Figure 2(b); Figure 3(a,b); Figure 4(a)). The Quorum spinning disk uses an electron multiplying CCD (EMCCD) camera to record fluorescence images so sequential imaging mode was employed (Figure 2(d) and Figure 4(b)), 405 nm excitation and 460/50 nm emission filter for DAPI; 490 nm excitation and 522/45 nm emission filter for GFP; 561 nm excitation and 610/80 nm emission filter for mCherry. Time-lapse imaging was taken at various time intervals and the z-sections are typically 1–2 μm thick. Images were processed and presented using Volocity 6.3 or ImageJ-Fiji. All images/frames in a time series or movie, often of multiple segments, were acquired with the same settings during time-lapse imaging of the same oocyte.

Polar body abscission assay

To assess abscission, the oocytes were injected with fluorescently labeled dextran (Texas Red^TM, 3000 MW; D3328, Invitrogen; 10 mg/mL; 20 nL per oocyte) at the indicated time. Since this injection also triggers second polar body emission [51], the second polar body was always labeled with dextran fluorescence, hence serving as an internal control. At least 15 minutes following dextran injection, oocytes were examined by confocal microscopy. The fluorescence status of the first polar body indicated if abscission occurred at the time of dextran injection. In rare occasions when dextran injection did not activate the oocytes, metaphase II spindle, instead of second polar body, was observed along with the first polar body. Fluorescence ratio was calculated as relative dextran fluorescence of the first polar body over that of the second polar body (or the cortical region near metaphase II spindle). To determine the effect of DB-BAPTA on polar body abscission, individual oocytes expressing eGFP-H2B were monitored for anaphase initiation. Five (5) minutes after the initiation of chromosome separation, the oocyte was injected with 10 nL of water (as control) or DB-BAPTA (25 mM in water). The injected oocytes were examined one hour later for polar body abscission as described above. Since DB-BAPTA disrupts metaphase II spindle and prevents second polar body emission [22], dextran fluorescence ratio was calculated as relative dextran fluorescence of the first polar body over that of the cortical region near metaphase II chromosomes.

Construction of TubeCaMP3

First, we inserted the microtubule-binding sequence, EMTB (amino acids 18–282 of ensconsin) [27] to pCS2-GCaMP3 [22]. The EMTB sequence was cloned from the EMTB-2x mCherry plasmid (Addgene #26742) [29] using the following two primers (forward and reverse respectively): 5’-TAT GAA TTC ACC ATG GCA GTG CGA AGC GAA ACA-3’ and 5’-TAT GAA TTC GAA GAG CCC TCA GGT GG −3’. The EMTB fragment was digested by EcoRI restriction endonuclease followed by inserting into pCS2-GCaMP3 [22], recosulting in pCS2-EMTB-GCaMP3. pCS2-EMTB-GCaMP3 was linearized by inverse PCR using the following primers: 5’- ATG GCA GTG CGA AGC GAA AC −3’ and 5’- GGT GAA TTC GAA TCG ATG GGA TCC TG −3’. The complete reading frame of mCherry (one copy) was amplified using the following PCR primers (5’- CGA TTC GAA TTC ACC GCC ACC ATG GTG AGC AAG −3’ and 5’- GCT TCG CAC TGC CAT CTT GTA CAG CTC GTC CAT G − 3’), resulting in a PCR product containing 15 bp homology to the linearized pCS2-EMTB-GCaMP3 plasmid on either side of the mCherry reading frame. pCS2-EMTB-GCaMP3 and an mCherry fragment is then combined by In-Fusion® cloning technique [52], resulting in pCS2-mCherry-EMTB-GCaMP3 (TubeCaMP3).

Uncaging of NPE-caged IP₃

NPE-caged inositol 1,4,5-trisphosphate or caged IP₃ (Invitrogen, I23580) was prepared at 0.5 mM in water and stored at −20°C. Caged IP₃ was injected into the oocytes at a final concentration of 1–1.5 μM before imaging. To “uncage” IP₃, the oocyte was exposed to UV excitation (filter set: Chroma 11000V3, 350/50 nm; light source: 100 W mercury bulb) through a 63x oil objective. UV exposure time was precisely controlled by an electronic shutter (LAMBDA SC, Sutter Instrument) via the manufacturer’s program. The oocyte was simultaneously subjected to confocal imaging through the same objective.

Fluorescent probes

The plasmids for eGFP-H2B, mRFP-H2B, Lck-mCherry, eGFP-PKCβ-C2, mRFP- and eGFP-PKCβ-C2 and Anillin-GFP have been described previously [22,29,35,53]. All cDNA constructs were linearized before transcribed in vitro using Ambion SP6 mMESSAGE kit. mRNAs were individually titrated for injection into immature oocytes. Typically, mRNA-injected oocytes were incubated for at least 12 hours to allow protein expression before the addition of progesterone to induce meiotic maturation. Organic calcium probes Oregon Green 488 BAPTA-2 or OG-2 (K_d = 580 nM, Invitrogen, O6808). Oregon Green 488 BAPTA-6 f or OG-6 f (K_d = 3 µM, Invitrogen, O23990) were dissolved in water (2–4 mM) and stored at −20°C. They were injected into oocytes at the final concentration of micromolar range (2.5–5 µM) before imaging.

Statistics

Mean values were compared with Student’s t-test (two tailed) or One-way ANOVA and Tukey’s HSD test. P < 0.05 is considered significant.

Acknowledgments

We thank Dr. Chloë van Oostende-Triplet and Ms. Skye McBride of University of Ottawa Image Core for advice and assistance during this study. GM is a recipient of a scholarship from China Scholarship Council, and a Designation 2020 Faculty of Medicine Scholarship of University of Ottawa.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data Availability Statement

The authors confirm that the data supporting the findings of this study are available within the article [and/or] its supplementary materials.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15384101.2022.2092815

Additional information

Funding

This work was supported by the National Institutes of Health [GM52932]; Canadian Institute of Health Research [PJT168891].