Yolk granule fusion and microtubule aster formation regulate cortical granule translocation and exocytosis in zebrafish oocytes

Dynamic reorganization of the cytoplasm is key to many core cellular processes, such as cell division, cell migration, and cell polarization. Cytoskeletal rearrangements are thought to constitute the main drivers of cytoplasmic flows and reorganization. In contrast, remarkably little is known about how dynamic changes in size and shape of cell organelles affect cytoplasmic organization. Here, we show that within the maturing zebrafish oocyte, the surface localization of exocytosis-competent cortical granules (Cgs) upon germinal vesicle breakdown (GVBD) is achieved by the combined activities of yolk granule (Yg) fusion and microtubule aster formation and translocation. We find that Cgs are moved towards the oocyte surface through radially outward cytoplasmic flows induced by Ygs fusing and compacting towards the oocyte center in response to GVBD. We further show that vesicles decorated with the small Rab GTPase Rab11, a master regulator of vesicular trafficking and exocytosis, accumulate together with Cgs at the oocyte surface. This accumulation is achieved by Rab11-positive vesicles being transported by acentrosomal microtubule asters, the formation of which is induced by the release of CyclinB/Cdk1 upon GVBD, and which display a net movement towards the oocyte surface by preferentially binding to the oocyte actin cortex. We finally demonstrate that the decoration of Cgs by Rab11 at the oocyte surface is needed for Cg exocytosis and subsequent chorion elevation, a process central in egg activation. Collectively, these findings unravel a yet unrecognized role of organelle fusion, functioning together with cytoskeletal rearrangements, in orchestrating cytoplasmic organization during oocyte maturation.


Introduction
Oogenesis marks the very first step in development, establishing the maternal blueprint for embryonic patterning. During this process, the oocyte grows in size by acquiring maternally provided material and completes its first meiosis to eventually become arrested in the metaphase of meiosis II until fertilization occurs [1]. Central to oogenesis is the accurate positioning of large organelles, such as the oocyte nucleus (germinal vesicle (GV)) and meiotic spindle, but also small fate-determining molecules, such as mRNAs and proteins, within the oocyte, a process fundamental for embryonic axis formation and cell fate specification [2][3][4][5][6][7][8][9][10][11][12].
Yet, how such positioning of cytoplasmic components is orchestrated in space and time is still only poorly understood. Cytoplasmic organization can occur in the absence of external cues, suggesting that the cytoplasm is capable of self-organization [13,14]. Previous research has highlighted an important role for the cell cytoskeleton, and especially the microtubule and actin networks, in driving such cytoplasmic self-organization [15][16][17]. For instance, microtubules and the movement of motors along microtubule tracks can power cytoplasmic flows and the repositioning of microtubule asters by generating viscous drag forces to the surrounding cytoplasm [15,[18][19][20][21]. Likewise, myosin II-dependent contractions of both cortical and bulk actin networks can result in large-scale actomyosin network flows, which, in turn, drag the adjacent cytoplasm via friction forces acting at their interface [16,22,23]. In addition to these motor-dependent processes, actin polymerization on the surface of organelles can drive organelle motility, thereby generating active diffusion within the bulk of the cytoplasm [23][24][25]. However, to what extent cellular processes other than cytoskeletal rearrangements, such as dynamic fusion and splitting of organelles, also function in cytoplasmic reorganization, remains unclear.
To tackle this question, we turned to the last stage of zebrafish oogenesis (oocyte maturation), during which cytoplasmic reorganizations are accompanied by changes in organelle shape, size, and position, preparing the oocyte for fertilization and embryonic development [26]. Zebrafish oogenesis constitutes a 5-stage process: During stages I and II, the oocyte animal-vegetal (AV) axis becomes determined through the vegetal pole localization of the Balbiani body, a membrane-less structure rich in mitochondria, proteins, and mRNAs required for the vegetal pole establishment [7,27]. Stages II and III mark the formation of both cortical granules (Cgs) and yolk granules (Ygs). Cgs are exocytosed upon fertilization to induce chorion elevation, thus preventing lethal polyspermy and protecting it against physical damage [28], while Ygs function as energy reservoirs for subsequent embryonic development [26]. Stage III oocytes remain arrested in prophase I until oocyte maturation begins. During oocyte maturation (stage IV), a multitude of transitions take place within the oocyte. For instance, the GV migrates to and breaks down at the animal pole of the oocyte, which is then followed by ooplasm, the oocyte cytoplasm, accumulating at the animal pole, forming a yolk-free blastodisc. Concomitantly, Cgs obtain exocytosis competency, thus preparing the mature oocyte/egg for fertilization [26,[29][30][31][32]. How these different aspects of cytoplasmic reorganization are orchestrated and spatiotemporally controlled is still largely unknown.
Here, we show that the accumulation of Cgs at the oocyte surface and their acquisition of exocytosis competency upon GV breakdown (GVBD) are driven by the concerted activities of Yg fusion and compaction and microtubule network rearrangements, respectively. Yg fusion and compaction towards the oocyte center function in this process by inducing radially outward cytoplasmic flows that lead to the translocation and accumulation of Cgs at the oocyte surface. The microtubule network, in contrast, triggers accumulation of Rab11-positive vesicles at the oocyte surface by reorganizing into acentrosomal aster-like structures that collectively translocate towards the oocyte surface and take along Rab11-positive vesicles. Finally, the decoration of Cgs by Rab11 at the oocyte surface confers competency to Cgs to be exocytosed during mature oocyte/egg activation.

GV breakdown triggers blastodisc formation, Yg fusion and compaction, and Cg outward flow
To unravel the molecular, cellular, and biophysical mechanisms underlying ooplasmic reorganization during zebrafish oocyte maturation, we first analyzed how these processes occur in space and time. To this end, ovaries of female zebrafish were harvested, and stage III oocytes were isolated according to their size and ooplasmic opacity and exposed to the steroid hormone DHP triggering their maturation [26,31]. To determine which processes take place during oocyte maturation, we monitored ooplasmic reorganization in oocytes from Tg(hsp:Clip170-eGFP) females, in which the ooplasm is ubiquitously labeled by GFP (Fig 1A and S1 Movie). We found that following GV translocation to and breakdown at the animal pole of the oocyte, ooplasm accumulated at the animal pole, indicative of blastodisc formation (Fig 1A-1A"). In order to visualize which other ooplasmic rearrangements occur upon GV breakdown, we additionally marked both Ygs and Cgs within the ooplasm by exposing Tg(hsp:Clip170-eGFP) oocytes to Lysotracker dye, which exclusively labels Ygs but not Cgs, allowing us to distinguish between these different granule types (Fig 1B and S2 and S3 Movies). We further confirmed that granules not labeled by Lysotracker were indeed Cgs by showing that within the mature oocyte/egg, they colocalized with Rab11 and underwent exocytosis upon egg activation (Fig 4B), features typically associated with Cgs [28]. We found that Ygs underwent multiple fusion events during oocyte maturation, increasing their average cross-sectional area by a factor of approximately 2 (Fig 1B-1C' and S2 Movie). To monitor how such fusion events are followed up by changes in the subcellular distribution of Ygs and Cgs within the oocyte, we focused our analysis on the second half of oocyte maturation (2 to 4.5 h after the DHP hormone addition), when the oocyte volume remained largely unchanged, thereby also removing the possible contribution of volume changes to our analysis (S1A-S1A' Fig). By simultaneously labeling, segmenting, and following different ooplasmic components and their phase fractions during this time window, we found Ygs to further compact to the oocyte center, while both ooplasm and Cgs displayed short-range movements towards the oocyte cortex, resulting in their accumulation there (Fig 1D-1D"' and S3 Movie). Of note, such ooplasmic rearrangements were not occurring if oocytes remained arrested at stage III in the absence of the DHP hormone (S1C Fig). Collectively, these findings suggest that oocyte maturation is temporally correlated with blastodisc formation at the AP, Yg fusion and compaction to the oocyte center, and Cg translocation towards the oocyte cortex (Figs 1 and S1D).

Bulk actomyosin drives blastodisc expansion
Our finding that blastodisc formation is spatiotemporally linked to GV breakdown at the animal pole of the oocyte (Fig 1A) suggests that these processes might be functionally linked. Given the large size of the GV, accounting for nearly approximately 1.5% of the total stage III oocyte volume, we hypothesized that the nucleoplasm released after GV breakdown might directly result in blastodisc formation. To test this possibility, we labeled the nucleoplasm stored within the GV with fluorescently-labeled Dextran and followed its subcellular distribution after GV breakdown (S1B Fig). This showed that despite the apparent rapid diffusion of the nucleoplasm from the point of GV breakdown at the animal pole towards the vegetal pole of the oocyte, the majority of the nucleoplasm still remained at the animal pole, thereby initiating blastodisc formation (S1B-S1B" Fig). Moreover, the size of the blastodisc continued to grow after the completion of GV breakdown, rather than shrink as expected for continuous diffusion of the nucleoplasm away from the animal pole (S1B' Fig), suggesting that mechanisms other than passive GV nucleoplasm release must be involved in blastodisc formation. Bulk actomyosin network contraction and flows have previously been implicated in blastodisc expansion within the fertilized egg [23]. To determine whether actomyosin network contraction is also involved in the initial blastodisc formation during oocyte maturation, we exposed immature stage III oocytes to inhibitors specifically interfering with actin polymerization (Cytochalasin B (Cyto B)) and myosin II activity (para-Nitroblebbistatin (PBb)), and monitored how such treatment affects blastodisc formation. We found that in oocytes treated with 30 μg/ml Cyto B or 100 μM PBb, blastodisc expansion was strongly diminished (S2A-S2A ' Fig and S4 Movie), indicating that actomyosin network contraction is required for this process.
To determine how actomyosin network contraction functions in blastodisc formation, we analyzed dynamic changes in the intensity of F-actin during oocyte maturation using Tg (actb1:Utr-GFP) oocytes labeling F-actin. This analysis revealed that actin became enriched within the GV and was released to the surrounding ooplasm upon GV breakdown ( To examine whether these processes establish an actin gradient along the AV axis of the oocyte, we visualized F-actin distribution in sections of stage IV zebrafish oocytes that had just undergone GV breakdown using Phalloidin staining. This revealed an animal-to-vegetal bulk actin gradient with peak levels close to the animal pole of the oocyte (Fig 2B and 2B'). Such a bulk actin gradient, analogous to the situation in fertilized eggs after fertilization [23], might lead to bulk actomyosin flows directed towards the animal pole of the oocyte, which, by dragging along the ooplasm, then triggers the accumulation of ooplasm at the animal pole leading to blastodisc formation. As visualizing bulk actin flows within the opaque maturing oocyte turned out to be challenging, we prepared ooplasmic extracts from stage IV oocytes, where bulk actin and Ygs were clearly recognizable. We found that bulk actin flows within these extracts were accompanied by ooplasm flowing along and accumulating where actin aggregates were formed (Fig 2C and 2C' and S6 Movie), suggesting that ooplasmic flows within the intact oocyte might also be driven by such bulk actin flows. Further supporting this assumption, measuring ooplasmic flows along the AV oocyte axis revealed that they occur predominantly close to the animal pole of the oocyte (Fig 2D and 2D' and S7 Movie), where also the bulk actin gradient was detectable (Fig 2B').
Kymograph acquired along the AV axis of the oocyte shown in Fig 1A as a function of time. The yellow dashed lines outline the GV contour and the dashed white line marks the time point of GVBD. The blue zone demarcates the blastodisc region. (A") Blastodisc clearance, measured as the height of blastodisc at the end of the maturation process as shown in Fig 1A and 1A', for WT oocytes (N = 2 experiments, n = 16 oocytes). See Table A in S1 Data for underlying data. (B) Fluorescence images of stage III Tg(hsp:clip170-GFP) oocytes labeling ooplasm (cyan) and exposed to Lysotracker to mark Yg (magenta) and Cg (black, identified by their exclusion of both Clip-170-GFP and Lysotracker) before maturation (stage III) and 60 and 120 min after maturation onset. (B') Average Yg area (N = 3, n = 20) over time. Yellow box indicates the period during which GVBD takes place. See Table B in S1 Data for underlying data. (C) Fluorescence images of stage III oocytes exposed to Lysotracker to label Yg (magenta) at 60-120 min after maturation onset. Dashed lines with the same color demarcate the outline of Yg that will undergo fusion. Asterisks mark the time point of fusion. (C') Histogram of Yg fusion events during oocyte maturation (N = 3, n = 8). See Table C in S1 Data for underlying data. (D) Fluorescence images of stage III Tg(hsp:clip170-GFP) oocytes labeling ooplasm (cyan) and exposed to Lysotracker to mark Yg (magenta) and Cg (black) at 120, 180, and 270 min after maturation onset (first row). Images in the bottom rows show segmented Yg and Cg obtained from the images in the first row. White dashed lines mark the oocyte outline. Yellow dashed lines denote the initial distribution of Yg in the second row and the final distribution of Cg in the third row. (D') Yg radial velocity between 120 and 270 min after maturation onset (N = 2, n = 8). See Table D in S1 Data for underlying data. (D") Temporal projection of segmented Yg (top) and Cg (bottom) of the oocyte shown in (D) between 120 and 270 min after maturation onset, illustrating the inward motion of Yg and the outward movement of Cg. (D"') Changes in phase fractions for ooplasm (top, green), Yg (middle, magenta), and Cg (bottom, cyan) between 120 and 270 min after maturation onset. Normalized (norm) radii of 0.5 and 1 correspond to the oocyte interior and cortex, respectively (N = 3, n = 6). See Table E in S1 Data for underlying data. Schematics demarcate the imaging plane used for obtaining the images in each panel. Note that in panel (A), processes deeper within the oocyte are captured, while in panels (B-D), more superficial parts of the oocyte are captured. Error bars, SEM. AP, animal pole; AV, animal-vegetal; Cgs, cortical granules; GV, germinal vesicle; GVBD, germinal vesicle breakdown; VP, vegetal pole; WT, wild-type; Ygs, yolk granules.
https://doi.org/10.1371/journal.pbio.3002146.g001  Table A in S2 Data for underlying data. (C) Fluorescence images of Notably, bulk actin levels ceased once the first meiosis was completed (S3A and S3A' Fig  and S8 Movie), raising questions as to the mechanisms by which the size of blastodisc is maintained during the second meiosis where no clear bulk actin was detectable anymore within the oocyte ooplasm (S3B Fig). Strikingly, we observed actin comet-like structures forming with no preferential orientation on the surface of granules located close to the blastodisc interface ( Fig 2E; note that we were unable to distinguish between Ygs and Cgs in these experiments as colabelling actin and Ygs turned out to be challenging due to difficulties in detecting actin in the bulk of the ooplasm). This observation is highly reminiscent of the role of actin comet-like structures on Ygs in fertilized eggs, which promote ooplasm-Yg segregation by preventing Ygs from returning into the blastodisc region [23]. Together, our results suggest that-similar to the situation in fertilized eggs-the combined function of actin flows towards the animal pole and actin comets on the surface of granules are responsible for blastodisc formation and maintenance during oocyte maturation.
Finally, given that the changes in bulk actin dynamics were concomitant with cell cycle progression (S3A and S3A' Fig), we asked whether CyclinB/Cdk1 complex, the key cell cycle regulator previously suggested to induce bulk actin polymerization and flows within the fertilized egg [23], might also function as an effector by which GV breakdown triggers bulk actin polymerization and flows and consequently blastodisc formation during oocyte maturation. To this end, we injected CyclinB-GFP mRNA into stage III oocytes ( Fig 2F) and detected endogenous phosphorylated CyclinB levels by immunofluorescence (S3D Fig) to visualize its localization as a proxy for CyclinB/Cdk1 activity within the maturing oocyte [33]. We found CyclinB to be highly enriched within the GV of stage III oocytes and then released to the surrounding ooplasm upon GV breakdown (Figs 2F and S3D and S9 Movie). Importantly, the release of CyclinB led to a spatially restricted gradient of CyclinB at the oocyte animal pole, which persisted for more than 1 h after GV breakdown (Fig 2F'), highly reminiscent of the gradients observed for the nucleoplasm (S1B-S1B" Fig) and bulk actin (Fig 2B and 2B') following GV breakdown. In line with the role of CyclinB in bulk actin polymerization, their dynamics were also temporally coordinated, both peaking at the GVBD onset (S3B and S3C Fig). Moreover, inhibiting CyclinB synthesis by exposing stage IV oocytes undergoing maturation to 700 μM Cycloheximide resulted in decreased ooplasmic flows and blastodisc formation (S2A and S2A' ooplasmic extract obtained from stage III Tg(actb1:Utr-GFP) oocytes labeling F-actin (gray) and exposed to Lysotracker to mark Ygs  Table B in S2 Data for underlying data. (E) Fluorescence images of oocytes injected with 200 pg Utrophin-GFP mRNA to label F-actin during first and second meiosis corresponding to 135 and 170 min after maturation induction with the DHP hormone, respectively. White arrowheads mark actin comets forming within the blastodisc region on the surface of granules. (F) Fluorescence images of stage III oocytes injected with CyclinB-GFP mRNA before (stage III) and 30, 60, and 90 min after maturation onset. Arrowhead denotes the GVBD onset. The yellow line along the oocyte circumference (Circum) was used to acquire the intensity profiles plotted in (F'). (F') Change in CyclinB signal normalized to its distribution at the time prior to GVBD, measured at 30 min (cyan) and 60 min (magenta) after GVBD along the oocyte circumference (the yellow line in F). Circumference of 0 and 800 μm correspond to the AP and VP, respectively (N = 2, n = 5). See Table C in S2 Data for underlying data. (G) Schematic summarizing the role of actin in ooplasmic flows and blastodisc formation. Bulk actin, initially stored within the GV, is released at the AP of the oocyte upon GVBD, thereby generating a local actin gradient. This actin gradient triggers bulk actomyosin flows towards the AP, which drags the ooplasm along, resulting in blastodisc formation. In addition, the blastodisc interface is maintained during first and second meiosis by actin comet-like structures forming on the surface of granules and-in analogy to previous observations in fertilized eggs [23]-preventing them from diffusing into the blastodisc region. Schematics in each panel demarcate the imaging plane used for obtaining the images in that panel. . This suggests that the activation and release of CyclinB/Cdk1 at the animal pole of the oocyte drives bulk actin polymerization, which, in turn, triggers bulk actin and ooplasmic flows leading to blastodisc formation ( Fig 2G).

Microtubule network forms asters upon GV breakdown
To determine whether other cytoskeletal elements, and in particular microtubules, might also have a function in ooplasmic reorganization during oocyte maturation, we exposed immature stage III oocytes to inhibitors specifically interfering with microtubule assembly/disassembly (Colchicine and Taxol) and monitored how such treatment affects ooplasmic reorganization and oocyte maturation. We found that in oocytes treated with 200 μM of Colchicine, blastodisc formation was largely unaffected (S2A and S2A ' Fig and S4 Movie), suggesting that microtubules might be dispensable for this process. Interestingly, however, both Colchicine-and Taxol-treated oocytes displayed strongly reduced chorion elevation, a process previously shown to be mediated by the release of Cgs at the oocyte surface ( [28,32]; S2B and S2B' Fig).
To understand whether and how microtubules might be involved in Cg relocalization and/or exocytosis during oocyte maturation, we first monitored how the microtubule cytoskeleton changes during oocyte maturation in Tg(XlEef1a1:dclk2-GFP) oocytes labeling microtubules. We found microtubules to be uniformly distributed throughout the ooplasm of stage III oocytes ( Fig 3A). This microtubule network, however, transformed drastically upon GV breakdown with numerous bright foci of microtubules appearing in the bulk of the ooplasm (Fig 3A and S10 Movie). Closer examination of these foci revealed that they resembled microtubule aster-like structures undergoing dynamic fusion and splitting over time (Fig 3B and S11 Movie), the number of which increased shortly after GV breakdown and then dropped again (Fig 3B'). These consecutive phases of aster formation and disappearance gave rise to a microtubule transformation wave propagating from the animal to the vegetal pole of the oocyte, where asters were forming at the leading edge and dissolving again at the trailing edge ( Fig 3A'-3A").
We next asked what signals might trigger this microtubule transformation wave. Given that the microtubule transformation wave was initiated upon GV breakdown and that the cell cycle regulator CyclinB/Cdk1 complex, stored within GV (Figs 2F and S3D), has previously been found to trigger microtubule reorganization [34][35][36][37], we hypothesized that the release of CyclinB/Cdk1 at the animal pole upon GV breakdown and its diffusion towards the vegetal pole of the oocyte might be involved in this process. To test this possibility, we exposed stage IV oocytes undergoing maturation to 700 μM Cycloheximide inhibiting the synthesis of CyclinB or to 250 μM of the specific Cdk1 inhibitor Dinaciclib (Figs 3C and S4A). Strikingly, microtubule asters prematurely disappeared in oocytes exposed to Cycloheximide or Dinaciclib, leading to a more homogeneous distribution of microtubules reminiscent of the situation in immature oocytes before GV breakdown (Figs 3C and 3C' and S4A and S12 Movie). This suggests that CyclinB/Cdk1 activation and release upon GV breakdown drives the observed microtubule transformation wave. Importantly, blocking F-actin polymerization by treating oocytes with 20 μg/ml Cyto B did not interfere with microtubule aster formation, suggesting that these mechanisms are distinct at the molecular level (S4B Fig).

Partial microtubule depolymerization drives microtubule aster formation during oocyte maturation
The formation of aster-like acentrosomal microtubule structures in Xenopus egg extracts has previously been shown to rely on the activity of microtubule motor dynein, clustering microtubule minus ends to the aster center [38]. Hence, we asked whether dynein motors might also To this end, we exposed stage III oocytes to 75 μM of the dynein inhibitor Ciliobrevin D [15]. Microtubule asters of Ciliobrevin D-treated oocytes failed to fully contract and generated abnormal asters and/or dense networks remaining connected across large distances within the oocyte (Figs 3D and S4C and S13 Movie), suggesting that dynein motor activity is needed for proper microtubule aster formation.
To further determine whether dynein motor activity needs to be high for microtubules forming asters, as previously suggested [36], we estimated the motor force regime underlying this microtubule network contraction/aster formation process during oocyte maturation. To this end, we measured microtubule contraction length scales by identifying the domain that gives rise to each microtubule cluster detectable at the final time point of contraction and determined the size of the first and second biggest domains (ξ 1 and ξ 2 ), as used in percolation theory to describe the mode of network contraction [39]. In the local contraction regime, ξ 1 and ξ 2 are of similar magnitude, while for large-scale contractions, ξ 1 becomes close to the system size at the expense of ξ 2 shrinking in size. To explore the ξ 1 versus ξ 2 space more systematically, we performed a network contraction analysis in oocytes exposed to DMSO (control) or 75 μM Ciliobrevin D to inhibit dynein motor activity. This analysis indicated that control oocytes exhibit local contractions (ξ 1 �ξ 2 ; S4D and S4D' Fig), while 40% of maturing oocytes exposed to 75 μM Ciliobrevin D exhibited large-scale network contractions (ξ 1 >ξ 2 ; S4D Fig), where the size of the biggest cluster reaches close to the system size (ξ 1 � 200 μm). Based on predictions from active gel contractility [39], such local network contractions, and ultimately the formation of small-sized asters, in control, but not dynein-inhibited oocytes, suggests that the dynein motor activity levels are high within maturing oocytes.
Given that microtubule aster formation is thought to be driven by increasing the ratio of microtubule motors to microtubule number [40], we further asked whether a reduction in microtubule number upon GV breakdown, and thus an increase in the ratio of microtubule motors to microtubule number, might lead to the observed aster formation of the microtubule network. To this end, we analyzed whether the total amount of polymerized microtubules kymograph in (A'). (A') Kymograph acquired along the circumference of the oocyte shown in (A) as a function of time. Dashed lines mark the leading and trailing edges of the microtubule aster formation wave. (A") Speed of microtubule aster formation wave along the oocyte circumference, measured from kymographs as shown in (A'); N = 9 experiments, n = 16 oocytes. See Table A in S3 Data for underlying data. (B) Maximum fluorescence intensity projection of high-resolution images of stage III Tg(Xla.Eef1a1:dclk2a-GFP) oocytes labeling microtubules during consecutive stages before (stage III) and 70, 85, 100, and 140 min after maturation onset. (B') Microtubule aster number as a function of time. Aster assembly during the second hour after maturation onset is followed by initially rapid and then slower disassembly phases (N = 2, n = 8). See Table B in S3 Data for underlying data. (C) Kymographs acquired along the AV axis of Tg(Xla.Eef1a1:dclk2a-GFP) oocytes exposed to DMSO (control, left) or 700 μM CHX (right) as a function of time. Arrowheads mark the time point of oocyte exposure to DMSO/CHX. Dashed lines mark the leading and trailing edges of the microtubule aster formation wave. (C') Microtubule aster number as a function of time for Tg(Xla.Eef1a1:dclk2a-GFP) oocytes exposed to DMSO (cyan, N = 2, n = 8) or CHX (magenta, N = 2, n = 9). The yellow box indicates the time window of oocyte exposure to DMSO or CHX in the respective experiments. See Table C in S3 Data for underlying data. (D) Kymographs acquired along the lateral (left) and AV (right) axis of exemplary Tg(Xla.Eef1a1:dclk2a-GFP) oocytes exposed to Cili D as a function of time demonstrating large-scale contractions and abnormal aster formation, respectively. (E) Microtubule intensity measured in a 50 pix-wide region at the blastodisc in the vicinity of the GV over time normalized to intensity values at GVBD onset (N = 2, n = 9). See Table D in S3 Data for underlying data. (F) Fluorescent images of stage III Tg(Xla.Eef1a1:dclk2a-GFP) oocytes exposed to DMSO (control, left) or Colchi (right) for 240 min in the absence of DHP (immature oocyte). The orange dashed boxes denote the lateral regions used for obtaining the kymographs shown on the right. The panels on the right indicate the kymographs of microtubule intensity along the lateral axis of the oocytes shown on the left as a function of time. The green arrowheads denote the time point of oocyte exposure to DMSO or Colchi. (F') Microtubule aster number as a function of time for oocytes exposed to DMSO (blue, N = 3, n = 14) or Colchi (green, N = 3, n = 13) without DHP (immature oocytes). The arrowhead denotes the time point of oocyte exposure to DMSO or Colchi. See Table E  changes upon GV breakdown by monitoring intensity changes in the vicinity of the GV in Tg (XlEef1a1:dclk2-GFP) oocytes labeling microtubules (Figs 3E and S4E). This analysis showed that the total amount of polymerized microtubules decreased up to approximately 30% of its initial levels just as the first microtubule asters appeared in the ooplasm (Fig 3E). To determine whether such depolymerization of microtubules would be sufficient to trigger this transformation, we treated immature stage III oocytes, still displaying uniform microtubule distribution, to 300 μM of the microtubule depolymerizing drug Colchicine or DMSO as control ( Fig 3F) and analyzed resultant changes in microtubule network organization. Remarkably, partial depolymerization of microtubules in Colchicine-treated stage III oocytes led to the premature formation of numerous microtubule asters displaying extensive fusion and splitting dynamics, while the microtubule network of DMSO-treated control oocytes remained unchanged ( Fig 3F  and 3F' and S14 Movie). Interestingly, the network contraction analysis of the microtubule asters forming in Colchicine-treated immature oocytes revealed small contraction length scales, indicative of the high dynein activity present already within the immature oocytes before GV breakdown (S4D-S4D' Fig). This suggests that the partial depolymerization of microtubules, rather than increased dynein motor activity, upon GV breakdown is responsible for the microtubule network transformations observed during oocyte maturation.
Collectively, these findings suggest a model where the microtubule network in early stage III oocytes is in a "jammed" configuration that cannot reorganize despite its high motor activity. The partial disassembly of this network upon GV breakdown will, in turn, "unjam" the network by increasing the ratio of microtubule motors to microtubules, eventually leading to microtubule aster formation ( Fig 3G).

Microtubule asters ensure proper chorion elevation upon activation
To determine whether and how the transformation of the microtubule network into acentrosomal asters is linked to its apparent requirement for chorion elevation, as suggested by our microtubule interference experiments (S2B and S2B' Fig), we analyzed the spatiotemporal dynamics of microtubule aster formation relative to the reorganization of Ygs, Cgs, and ooplasm. This analysis revealed that microtubule asters, while travelling as a wave from the animal to the vegetal pole of the oocyte, exhibited a radially outward-directed flow from the center to the surface of the oocyte, leading to microtubule aster accumulation in cortical regions of the oocyte (Fig 4A-4A" and S15 Movie). Notably, the outward flow of microtubule asters was also observed in immature oocytes treated with Colchicine to trigger premature microtubule depolymerization-mediated aster formation (Fig 3F), suggesting that this outward flow is due to some inherent radial polarity of the oocyte. Consistent with this possibility, the subcellular localization of the microtubule-cortex anchoring proteins Dynein and Numa [41], visualized by immunofluorescence, displayed strong cortical accumulation all around the oocyte periphery (S4F and S4F' Fig), pointing at the possibility that higher levels of dynein at the cortex might trigger the outward movement of microtubule asters by preferentially anchoring microtubules to the oocyte cortex [42]. In line with this assumption, in oocytes exposed to Ciliobrevin D, abnormal aster formation was accompanied by reduced movement of these aster-like structures towards the oocyte cortex (S4G and S4G' Fig).
Given the similarity in the radially outward-directed movement of both microtubule asters and Cgs during oocyte maturation (Figs 1D" and 4A'), we asked whether microtubules might drag Cgs towards the cortex where their exocytosis is needed for chorion elevation. However, interfering with microtubule aster formation by treating stage III oocytes with 200 μM Colchicine did not affect Cg accumulation at the oocyte cortex (S5A and S5A' Fig), suggesting that microtubules do not function in chorion elevation by transporting Cgs towards the oocyte cortex.  Table A in S4 Data for underlying data. (B) Fluorescence images of Tg(actb2:Rab11a-NeonGreen) oocytes marking Rab11-positive vesicles (green) and exposed to Lysotracker to label Ygs (magenta) at stage V (mature oocyte/egg, left) and 5 min after activation (mpa) with E3 medium (right). Cgs (black) are identified by their exclusion of Lysotracker and the ooplasmic signal. The yellow line indicates the region used for displaying the orthogonal view (bottom images). Asterisks mark exemplary Cg undergoing exocytosis, and arrowheads denote the localization of Rab11 on Cg prior to/during their exocytosis. (C) Left: Fluorescence images of stage III Tg(actb2:Rab11a-NeonGreen) oocytes marking Rab11-positive vesicles (magenta, top rows) and injected with 400 pg of DCLK-mKO2 mRNA to label microtubules (green, middle rows) 120, 150, 180, and 215 min after maturation onset in control oocytes (WT, top panels) or oocytes exposed to 50 μM Taxol (bottom panels). Overlaid images are shown in the bottom rows. Dashed lines indicate the regions used for acquiring kymographs on the right. Right: Kymographs acquired along the marked area of the oocytes shown on the left as a function of time. Arrowheads point at exemplary Rab11-positive vesicles or microtubule Alternatively, microtubule aster formation and translocation towards the oocyte cortex might be required for chorion elevation by regulating Cg exocytosis. The Rab family of proteins regulates various aspects of cellular trafficking and has been implicated in Cg exocytosis in various animal species [43][44][45]. In particular, Rab11 has previously been shown to localize to the surface of Cgs in both zebrafish and C. elegans oocytes [28,44] and to be required for the synchronous secretion of Cgs upon fertilization in C.elegans [44]. We thus asked whether the outward moving microtubule asters might be involved in chorion elevation by transporting Rab11-positive vesicles towards the oocyte cortex, where Cgs reside, thereby facilitating their exocytosis by decorating Cgs with Rab11. To test this possibility, we generated Tg(actb1: Rab11a-NeonGreen) animals, allowing us to visualize Rab11 dynamics within the maturing oocyte. Consistent with previous observations, we found Rab11 to colocalize with Cgs at the oocyte surface upon egg activation (Fig 4B and S16 Movie). Interestingly, we also found that Rab11-positive vesicles displayed outward flows together with microtubule asters during oocyte maturation and eventually colocalized with Cgs at the cortex of mature oocytes/eggs (Fig 4C and S17 Movie). To determine whether microtubules are involved in this translocation of Rab11-positive vesicles to the oocyte cortex, we analyzed whether Rab11-positive vesicle distribution changes when microtubule aster formation is altered in oocytes. To interfere with microtubule dynamics, we treated oocytes with 200 μM of the microtubule depolymerizing drug Colchicine and found a strongly reduced density of Rab11-positive vesicles at the cortex of treated mature oocytes/eggs (S6A and S6A' Fig; [36]). In contrast, stabilizing microtubules with 50 μM Taxol led to the formation of enlarged microtubule aster-like structures, accompanied by Rab11-positive vesicles aggregating on and moving together with these enlarged microtubule asters towards the oocyte surface ( Fig 4C and 4C' and S18 Movie). In line with a critical function of microtubules in Rab11 translocation to the oocyte surface and decoration of Cgs, we found both Cg exocytosis and chorion elevation to be compromised in both Colchicine-or Taxoltreated mature oocytes/eggs (S6B, S6B' and S2B-S2B' Figs). Finally, to investigate whether Rab11 is indeed required for Cg exocytosis, and thus chorion elevation, we expressed a dominant-negative variant of Rab11, Rab11 S25N [46], to block Rab11 activity during oocyte maturation ( Fig 4D). Strikingly, we found that overexpression of Rab11 S25N led to strongly reduced chorion elevation in activated oocytes/eggs (Fig 4D'), suggesting that Rab11 activity is required for this process. In contrast, the overexpression of Rab11 S25N had no clearly recognizable effects on microtubule aster formation and/or Cg relocalization (S6C and S6D Fig). Taken together, these results indicate that microtubule asters, by moving towards the cortex upon GV breakdown, take along Rab11-positive vesicles, and that this cortical translocation of Rab11-positive vesicles is required for the decoration of Cgs at the cortex with Rab11 and, thus, Cg exocytosis.

Yolk granule fusion triggers cytoplasmic flows transporting cortical granules towards the oocyte cortex
Questions remain as to the mechanisms underlying the relocalization of Cgs towards the oocyte cortex since neither the inhibition of myosin II activity nor the depolymerization of asters moving towards the cortex. (C') Rab11 radial velocity during maturation onset for control oocytes (left, N = 2, n = 10) and oocytes exposed to 50 μM Taxol (right, N = 3, n = 11). See Table B in S4 Data for underlying data. (D) Brightfield images of oocytes injected with 350 pg of Rab11-mcherry (left) or Rab11S25N-mCherry (right; DN) mRNA, induced to undergo oocyte maturation for 270 min and activated consequently by exposure to E3 medium for 30 min. Black lines demarcate the distance between the egg and its overlaying chorion. (D') Chorion elevation, measured as chorion diameter normalized to the oocyte diameter, of oocytes injected with 350 pg of Rab11-mcherry (blue, control, N = 3, n = 42) or Rab11S25N-mCherry (magenta, N = 3, n = 46) mRNA. See Table C Fig and S19 Movie), arguing against a direct involvement of these cytoskeletal networks (or at least the specific components tested in our experiments) in this process. To identify such mechanism(s), we performed a detailed spatiotemporal analysis of Cg translocation towards the oocyte surface. This showed that in immature stage III oocytes, Cgs were distributed in between Ygs (Fig 1D). However, as oocyte maturation proceeded and GV breakdown occurred at the oocyte animal pole, Ygs underwent extensive fusion and compaction concomitant with Cgs translocation towards the oocyte cortex ( Fig 1D" and 1D"'). Generally, the compaction of a compressible material embedded within an incompressible fluid is expected to drive outward fluid flows due to volume conservation. Therefore, we postulated that Yg fusion and compaction might result in outward ooplasmic flows, which, in turn, take along Cgs, thereby translocating them to the oocyte surface. To directly test this possibility, we treated oocytes with 100 μM Ouabain, a known inhibitor of Na + /K + ATPase pumps [47], as an increase in intracellular K + levels has previously been suggested to induce Yg fusion in other teleosts [1,48]. In oocytes treated with Ouabain, the oocyte remained opaque and Ygs largely failed to fuse and compact towards the oocyte center (Figs 5A, 5A', S7A and S7B and S20 Movie; [48]). Moreover, Cg translocation to the cortex was strongly reduced (Fig 5B and 5B' and S19 and S21 Movies), suggesting that Yg fusion and compaction are required for Cg translocation to the oocyte surface. Notably, we also observed that not only Cg translocation but also blastodisc formation was reduced in Ouabain-treated oocytes (S7C and S7C' Fig), suggesting that Yg fusion and resultant outward cytoplasmic flows also contribute to blastodisc formation. In line with such a function, we found that Yg fusion was more pronounced at the animal pole of the oocyte, where the blastodisc is forming (S7D and S7D ' Fig). Importantly, these different effects of Ouabain treatment were not due to Ouabain affecting GV breakdown and, thus, cell cycle progression and its associated actin and microtubule cytoskeletal rearrangements, or the deposition of Rab11-positive vesicles at the surface of mature oocytes/eggs, as all of these processes still occurred in the presence of Ouabain treatment (S7E- S7G Fig). Taken together, these results suggest that Yg fusion and its associated radially outward-directed ooplasmic flows drive the translocation of Cg to the oocyte cortex and contribute to blastodisc formation at the oocyte animal pole.
Finally, we asked what signals might trigger Yg fusion. Given that an increase in intracellular K + levels has previously been proposed to trigger YG fusion in Sea Bass oocytes [48], we monitored dynamic changes in intracellular K + levels during oocyte maturation by injecting a K + indicator (K + -Green) into the immature stage III oocyte. Strikingly, we found K + to become enriched in small vesicles, which increased in number upon GV breakdown and underwent extensive fusion with each other and Ygs, ultimately increasing K + levels inside Ygs undergoing fusion (Fig 5C and 5C'). In contrast, no such spatiotemporal changes in K + levels upon GV breakdown were detected in Ouabain-treated oocytes defective in Yg fusion (S7H and S7H '  Fig and S22 Movie). Given that Ouabain blocks Na + /K + ATPase pumps, this suggests that GV breakdown leads to Yg fusion by triggering a Na + /K + ATPase-dependent increase in the concentration of K + within Ygs. This Yg fusion and compaction to the oocyte center, in turn, triggers radially outward-directed ooplasmic flows, which carry along and position Cgs at the oocyte cortex (Fig 5D).

Discussion
Our study provides novel insight into both the processes underlying oocyte maturation and the general mechanisms by which cytoplasmic reorganization is achieved within cells. Cgs are Golgi-derived secretory vesicles, which localize to the cortex of the mature oocytes/eggs to undergo exocytosis upon fertilization and induce chorion elevation/modification, thus preventing the entry of additional sperm into the oocyte [49]. Small Rab GTPase family of proteins have been found to associate with Cgs and regulate their transport and exocytosis in oocytes of diverse organisms [43,44,50]. In mouse oocytes, for instance, positioning of Cgs to the oocyte cortex has been proposed to rely on both myosin Va motors localizing to these granules and inducing their movement on an intrinsically polarized bulk actin network, as well as "hitchhiking" on outward moving Rab11a-vesicles [43]. Our findings that Yg fusion and compaction to the oocyte center drive Cg movements towards the oocyte circumference, and that microtubule aster formation and translocation to the cortex lead to the concomitant accumulation of Rab11-positive vesicle at the circumference, identifies a yet unknown mechanism of Cg translocation and exocytosis during oocyte maturation. Why the mechanisms regulating Cg translocation and exocytosis differ between mice and zebrafish is still unknown, but likely the presence of Ygs in zebrafish but not mouse oocytes has led to different functional adaptations of cytoplasmic components.
Beyond the specific regulation of Cg localization within the maturing oocyte, our findings also shed light on the general mechanisms underlying cytoplasmic reorganization and the role of the cell cytoskeleton therein. Cytoskeletal networks, such as the actin and microtubule cytoskeleton, have been implicated as the main driving force underlying cytoplasmic organization in oocytes and embryos. In particular, myosin II-mediated actin network flows dragging the adjacent cytoplasm have been shown to generate cytoplasmic streaming in early Drosophila and zebrafish embryos [22,23], and the motion of kinesin I motors along microtubule arrays anchored to the cortex to drive cytoplasmic flows in Drosophila and C. elegans oocytes by exerting viscous drag forces to the surrounding ooplasm [18,20]. In addition, the assembly and interaction of actin comets and centrosomal microtubule asters with organelles such as Ygs, the nucleus, or the meiotic spindle can position those organelles to specific locations within oocytes and embryos [23][24][25]. Our findings that the partial disassembly of a prestressed microtubule network results in its transformation into acentrosomal microtubule asters, which move towards the oocyte cortex in a dynein-dependent manner, thereby carrying along and enriching Rab11-positive vesicles at the oocyte surface, suggest a novel mechanism by which the microtubule network can affect vesicle localization within cells. Moreover, our observation after maturation onset. (A') Average Yg cross-sectional area normalized to its value at stage III as a function of time during oocyte maturation for oocytes exposed to DMSO (green, N = 3 experiments, n = 15 oocytes) or Ouabain (red, N = 3, n = 14). See Table A in S5 Data for underlying data. (B) Fluorescence images of DMSO-treated (left panels) and Ouabain-treated (right panels) stage III Tg(hsp:clip170-GFP) oocytes labeling ooplasm (cyan) and exposed to Lysotracker to mark Yg (magenta) and Cgs (black, identified by their exclusion of both Clip-170-GFP and Lysotracker) at 120, 180, and 270 min after maturation onset (top rows). Images in the middle and bottom rows show segmented Yg and Cg, respectively, obtained from the images in the first rows. White dashed lines mark the oocyte outline. Yellow dashed lines denote the initial distribution of Yg in the middle rows and the final distribution of Cg in the bottom rows. (B') Changes in phase fractions of Yg (magenta) and Cg (cyan) for oocytes exposed to DMSO (top, N = 3, n = 9) or Ouabain (bottom, N = 2, n = 8) between 120 and 270 min after maturation onset. Normalized (norm) radii of 0.5 and 1 correspond to the oocyte interior and cortex, respectively. See Table B in S5 Data for underlying data. (C) Top row: Fluorescence images of stage III oocytes injected with K + indicator (K + -Green, green) and Dextran Alexa Fluor 647 to mark ooplasm (magenta) before (stage III) and 75, 150, and 270 min after maturation onset. Cgs (black) are identified by their exclusion of both Dextran and K + -Green. The vesicles enriched with K + fused with each other and with Yg, thereby increasing their internal K + concentration and hence becoming dark blue (in Green-Fire-Blue Lookup Table). that such transformation of the microtubule network is driven by the cell cycle regulators Cdk1/CyclinB partially disassembling the microtubule network, mechanistically links cell cycle progression to cytoskeletal reorganization in cells.
One of the key findings of our study is that the regulated fusion and compaction of Ygs to the oocyte center can trigger outward cytoplasmic flows, which, in turn, transport Cgs towards the oocyte circumference. Importantly, the effect of Ygs fusion and compaction is independent of the role of bulk actomyosin flows in generating ooplasmic flows towards the oocyte AP, as in the PBb-treated oocytes, Yg fusion remains largely unchanged compared to control oocytes (S7I and S7I' Fig). This suggests that beyond ATP/GTP hydrolysis-dependent actin/microtubule polymerization and the activity of their associated motor proteins, cytoplasmic flows and reorganization can also be achieved by coordinated changes in the size and shape of organelles. Interestingly, these are processes that-differently from the aforementioned cytoskeletal rearrangements-do not directly depend on the activity of motor proteins and polymerases but instead are regulated by changes in the propensity of organelles to deform and/or undergo fusion and compaction. Our observation that Yg fusion and compaction are mediated by a rise in the level of K + within Ygs, which depends on the activity of Na + /K + ATPase pumps and is triggered by GV breakdown, points at intracellular ion concentration as a critical regulator of organelle fusion and resultant cytoplasmic reorganization. Notably, an increase of intracellular K + within the oocyte coinciding with the fusion of Ygs has previously been observed in oocytes of various teleosts, suggesting that K + -driven Yg fusion might represent an evolutionarily conserved mechanism in Yg-containing oocytes. How GV breakdown leads to an increase in K + levels within Ygs and how such increase promotes Yg fusion is not yet understood, but it is conceivable that electrostatic interactions between the plasma membrane surrounding Ygs and K + molecules promote the capacity of their plasma membrane to undergo fusion.
The different mechanisms identified in this study regulating blastodisc formation, and Cg translocation and decoration with Rab11, all depend on cell cycle progression and GV breakdown. In contrast, the molecular realization of these mechanisms does not seem to depend strictly on each other, as interference with one of these mechanisms does not necessarily lead to a complete failure in the progression of the remaining mechanisms. For instance, abrogating actomyosin flows does not affect microtubule aster formation and contraction or Yg fusion/ compaction and, vice versa, interfering with the latter two processes does not prevent the release of actin into the ooplasm upon GV breakdown. However, these distinct mechanisms functionally interact in reorganizing ooplasmic components, with Yg fusion/compaction and actomyosin flow together triggering blastodisc formation, and Yg fusion/compaction and microtubules aster translocation together regulating exocytosis competency of Cgs and, thus, chorion elevation. These distinct molecular realizations of the contributing mechanisms, combined with their partially overlapping function, might not only confer robustness to the cytoplasmic reorganization, but also allow the independent use of these mechanisms in separate developmental processes.
Cytoplasmic reorganization relies on the inherent self-organizing properties of the cytoplasm and a combination of prepatterning of the cell and external signals modulating this selforganizing activity. While the external signals and prepatterning can vary depending on the specific organismal context, cytoplasmic self-organization represents a generic propensity of the cytoplasm that emerges as a result of a certain composition or state of the cytoplasm. Our finding of Yg fusion eliciting large-scale cytoplasmic streaming reveals a novel feature of such inherent self-organizing capacity of the cytoplasm that plays an important role in reorganizing the cytoplasm during zebrafish oocyte maturation. Given that Ygs constitute a major cytoplasmic component in oocytes from birds, reptiles, worms, and fish, Yg fusion might represent a common principle of regulating cytoplasmic streaming during oogenesis.

Ethics statement
All animal procedures and protocols were performed in accordance with protocols (#66.018/ 0010-WF/II/3b/2014 incl. amendments 1-7) that were approved by the national Animal Experimentation Commission at the Federal Ministry of Austria in line with EU and national legislation, which includes an ethical evaluation of the performed experiments.

Ovarian follicle isolation
Methods of ovarian follicle isolation and culture were adapted from [5,52,53]. Female fish were anesthetized in 0.02% Tricaine and killed by decapitation. Ovaries were harvested in culture medium [90% Leibovitz's L-15 medium with L-glutamine (Thermo Fisher) (pH 9.0), Penicillin-Streptomycin 50 U/ml, and 0.5% bovine serum albumin (BSA, Sigma-Aldrich)]. Follicles were isolated from ovaries by gentle pipetting with a glass Pasteur pipette and dissection with forceps. Stage III oocytes were identified by their size (500 to 690 μm), opaque cytoplasmic density, intact GV, and undetectable blastodisc [31] and kept in culture medium for up to 12 h at 25 to 26˚C.

Blastodisc clearance analysis
Tg(hsp:Clip170-eGFP) oocytes were imaged during oocyte maturation using an inverted Leica SP5 or SP8 confocal microscope equipped with a Leica 20× objective. Oocytes with clear AV orientation were selected for further analysis. A kymograph was acquired along an 80 pix-wide line covering the oocyte AV axis. A line was then fitted to the blastodisc-Ygs interface in the kymograph over time until the end of oocyte maturation. Blastodisc clearance was measured from the tangent of the slope of the fitted line.

Yolk granules fusion analysis
To determine the number of Ygs fusion events, the Lysotracker-exposed oocytes were imaged using an inverted Leica SP5 or SP8 confocal microscope equipped with a Leica 20× objective. Maximum intensity projection images of Ygs were then used to track the fusion dynamics of Ygs. Fusion events of 10 to 20 Ygs per oocyte were counted during oocyte maturation using Fiji. These results were then plotted as a histogram.

Yolk granules, cortical granules, and ooplasm segmentation
To measure average Yg size over time, the Lysotracker-exposed oocytes were imaged using an inverted Leica SP5 or SP8 confocal microscope equipped with a Leica 20× objective. Maximum intensity projection images of Ygs were used to segment Yg (Lysotracker-signal positive) using Ilastik software. The segmented images were then analyzed in Fiji to measure the 2D cross-sectional area of Ygs over time.
For Cgs density at the oocyte surface, surface images (30 to 50 μm beneath the cortex) of Tg (hsp:Clip170-eGFP) or Tg(actb2:Rab11a-NeonGreen) oocytes exposed to Lysotracker were acquired using an inverted Leica SP5 or Sp8 confocal microscope equipped with a Leica 20× objective. Cgs were identified as ooplasm-as well as Lysotracker-negative granules and segmented using Ilastik software. The cross-sectional area of the imaging plane was also segmented with Ilastik, and the sum of the Cg area present on the imaging plane normalized to the area of the plane was measured as the Cg density at the oocyte surface over time using a custom-designed MATLAB script.
To measure the Cg distribution along the oocyte radius, a similar approach was employed on images taken at 50 to 75 μm beneath the cortex. Using the segmented images, oocyte center of mass and radius were determined, and the distance of each Cg to the oocyte center was measured and normalized to the oocyte radius. These results were then plotted as histogram with bin size of 0.05 along the normalized radial axis. Of note, the local decrease of CGs close to the cortex is the result of the segmentation procedure, not being able to precisely outline the position of the cortex relative to the small CGs present in this area. However, as this trend was observed in all of experimental conditions, it is unlikely to affect the interpretation of these experiments.
For performing phase fraction analysis, surface images of Tg(hsp:Clip170-eGFP) oocytes exposed to Lysotracker during 120 to 270 min after oocyte maturation induction were segmented using Ilastik software as described above to obtain Yg, Cg, ooplasm, and whole oocyte segmentations. Oocyte radius for angles between 0 and 360 degrees (with increments of 5 degrees) was then measured from oocyte segmentation. With the oocyte radius in hand, the distribution of binarized Yg, Cg, and ooplasm phases across all angles and over the oocyte radius was measured. These distributions were normalized to the total phase, the sum of all 3 phases, to obtain their relative phase distributions. The changes in relative phases of Yg, Cg, and ooplasm between 120 and 270 min were then calculated and plotted in the corresponding condition.
For Rab11 density at the surface of mature oocytes/eggs, Tg(actb2:Rab11a-NeonGreen) oocytes were incubated in 1 μg/ml DHP-containing medium for 270 min to complete maturation. Surface images (30 to 50 μm beneath the cortex) of the mature oocytes/eggs were acquired using an inverted Leica SP5 or Sp8 confocal microscope equipped with a Leica 20× objective. Rab11 vesicles were then segmented using Ilastik software. The cross-sectional area of the imaging plane was also segmented with Ilastik, and the sum of the Rab11 area present on the imaging plane, normalized to the area of the plane, was measured as the Rab11 density at the oocyte surface using a custom-designed MATLAB script.

Yolk granules flow measurement
To measure the Yg flows, oocytes were exposed to Lysotracker for labeling Ygs and imaged with an inverted Leica SP8 confocal microscope equipped with a Leica 20×/40× objective. Multiple kymographs were acquired using Fiji along 10 pix-wide lines covering different oocyte axes. Yg flow speeds were then measured from the slopes of lines fitted to the outermost Ygs flowing towards the oocyte center in the kymographs.

Oocyte volume measurement
To calculate changes in oocyte volume during oocyte maturation, Tg(hsp:Clip170-eGFP) oocytes were imaged in 3D with 125 Z slices, each 2 μm apart, covering 250 μm depth of the oocyte using an inverted Leica SP5 confocal microscope equipped with a Leica 10×/20× objective. The images were then segmented with Ilastik software to capture the complete oocyte shape, and the oocyte volume was measured as the sum of the number of pixels in segmented and binarized images and normalized to oocyte volume at the onset of the maturation process.

Nucleoplasm distribution upon germinal vesicle breakdown
To measure nucleoplasm distribution upon GVBD, 10 KDa Dextran Alexa Fluor 647 was injected into or in the vicinity of the GV, resulting in the enrichment of Dextran in the GV. Next, a line scan was acquired using Fiji along a 30 pix-wide line covering the oocyte circumference, and Dextran intensity was measured along this line during the maturation process. The Dextran intensity profile was then subtracted from the profile prior to GVBD and normalized to its highest value.

Chorion elevation assay
Stage III oocytes were injected with 350 pg of Rab11-mCherry or Rab11S25N-mCherry mRNA, incubated in culture medium for 3 h as described above, and then induced to undergo maturation by the addition of the DHP hormone for 270 min. In the last 30 min of oocyte maturation, the follicle membrane was removed manually using forceps [53]. The defolliculated fully mature oocytes/eggs were then transferred to a dish containing E3, triggering Cg exocytosis and, thereby, chorion elevation [28]. After 30 min in the E3 medium, the activated eggs were imaged with a Stereo-microscope, and the images were then used to calculate the extent of chorion elevation, measured as chorion diameter normalized to the oocyte diameter, using Fiji software [61].

Phalloidin and immunohistochemistry
Stage III oocytes were harvested from wild-type female fish and induced to undergo maturation by the addition of DHP to the culture medium. After GVBD (1 to 1.5 h after maturation onset), the oocytes were fixed in a glass vial containing 2% paraformaldehyde at 4˚C overnight.
Fixed oocytes were then mounted in 4% LMP agarose and sectioned using a vibratome (VT1200S, Leica) into 200 μm thick slices. Sections were then washed 3× for 10 min in PBS with 0.1% Triton X-100, permeabilized by PBS with 0.5% Triton X-100 for 1 h at room temperature, and blocked in PBS containing 0.1% Triton X-100, 1% DMSO, and 10% goat serum (blocking buffer) for 3 h at room temperature. To visualize Dynein, Numa, or phosphorylated CyclinB levels/localization, the samples were incubated in anti-Dynein (clone 70.1, Sigma, dissolved 1:1,000 in blocking buffer), anti-Numa (Thermo Fisher, dissolved 1:100 in blocking buffer), and anti-Cyclin B1 (phospho S147, Abcam, dissolved 1:100 in blocking buffer) primary antibody at 4˚C overnight, washed 3× for 10 min in PBS with 0.1% Triton X-100 at room temperature and incubated with secondary antibody (goat anti-mouse/rabbit conjugated to Alexa Fluor 488 (Molecular Probes), 1:250 in blocking buffer) for 3 h at room temperature. For visualizing the bulk actin gradient, the oocytes were fixed, sectioned, and permeabilized as described above, and the slices were then incubated in Phalloidin 488 (Invitrogen, dissolved 1:200 in blocking buffer) at 4˚C overnight. The stained sections were imaged using an inverted Leica SP5 confocal microscope equipped with a Leica 20× objective.

Bulk actin gradient analysis
Phalloidin-labelled and sectioned stage IV oocytes, which had just undergone GVBD, were imaged using an inverted Leica SP5 or Sp8 confocal microscope equipped with a Leica 20× objective. Oocytes with clear AV orientation were selected for further analysis. Maximum intensity projection images were used to segment bulk actin-containing ooplasmic pockets with Ilastik. The mean bulk actin intensity within these pockets was measured in Fiji and averaged over a 350-μm wide window centered along the oocyte AV axis.

Oocyte extract preparation
To generate extracts of maturing oocytes, approximately 50 to 100 Tg(actb1:Utr-GFP) oocytes labelling F-actin were incubated in 3.5 ml L15 medium containing 1 μM Lysotracker and 1 μg/ ml DHP for 60 to 90 min until GV broke down (oocytes were screened for GVBD under the stereomicroscope). Oocyte extracts were prepared by puncturing oocytes using a spike-micropipette with an inner diameter of 50 to 100 μm (BioMedical Instruments) attached to a syringe allowing for the aspiration of ooplasm-Ygs mixture. The pressure was manually controlled to prevent the aspiration of the medium into the pipette and thus the dilution of the oocyte extract. The ooplasm-Yg mixture was then promptly released into mineral oil (Sigma) inside a glass bottom Petri dish (MatTek) and imaged as described above 15 to 30 min after the extraction procedure had started.

F-actin intensity measurement
To measure nuclear F-actin intensity within the GV shortly before and after GVBD in Tg (actb1:Utr-GFP) oocytes, 35 μm × 35 μm-shaped ROIs were defined within the GV, and the averaged intensities over time were obtained using Fiji. To measure F-actin intensity within the blastodisc of Tg(actb1:Utr-GFP) during oocyte maturation, a kymograph was acquired along a 50 pix-wide line covering the oocyte AV axis. A line scan was then performed at the blastodisc region to measure the F-actin intensity within the blastodisc region.

Ooplasmic flow measurement
To measure ooplasmic flows, oocytes were injected with Dextran Alexa 647 to label ooplasm and imaged with an inverted Leica SP8 confocal microscope equipped with a Leica 20× objective. A kymograph was acquired using Fiji along a 50 pix-wide line covering the oocyte AV axis. Ooplasmic flow speeds were then measured from the slopes of lines fitted to the ooplasm pockets flowing towards the animal pole on the kymograph. The initial positions of those pockets were normalized to the length of the AV axis and used for obtaining the ooplasmic flow profile.

CyclinB intensity measurement
To measure CyclinB intensity within the blastodisc of CyclinB1-GFP mRNA injected oocytes during oocyte maturation, a kymograph was acquired using Fiji along a 30 pix-wide line covering the oocyte circumference. A line scan was then performed at the blastodisc region to measure the CyclinB intensity there.

Microtubule intensity measurement
To measure microtubule intensity within the blastodisc of Tg(Xla.Eef1a1:dclk2a-GFP) oocytes during oocyte maturation, a kymograph was acquired using Fiji along a 50 pix-wide line covering the oocyte AV axis. A line scan was then performed at the blastodisc region to measure the microtubule intensity there.

Microtubule aster number and density
Microtubule asters were followed over time using Imaris 9.1.2 (using spot detection algorithm), and their tracking data were imported to MATLAB R2019b for obtaining their number over time. In cases where oocyte volume turned out to be different between different experimental conditions, microtubule aster density was calculated by normalizing the total aster number to the oocyte area.

Microtubule cluster size analysis
The surface of Tg(Xla.Eef1a1:dclk2a-GFP) oocytes was imaged with an inverted Leica SP5 or SP8 confocal microscope equipped with a Leica 20× objective. Maximum intensity projections of these images (along the z slices) were used to obtain temporal intensity projections of microtubules during aster formation using the Temporal-Color Code plugin in Fiji. The temporally projected images were segmented with the Ilastik software to outline the microtubule structures undergoing contractions. Cluster sizes were then measured by applying the Voronoi algorithm plugin in Fiji to the segmented images. The size of the first and second biggest clusters was determined by a custom-designed MATLAB script.

Microtubule flow measurement
To measure microtubule aster flow speed, Tg(Xla.Eef1a1:dclk2a-GFP) oocytes labeling microtubules were imaged with an inverted Leica SP8 confocal microscope equipped with a Leica 20×/40× objective. Multiple kymographs were acquired using Fiji along 10 pix-wide lines covering different oocyte axes. Microtubule aster flow speeds were then measured from the slopes of lines fitted to the microtubule asters flowing towards the oocyte surface in the kymographs.

Rab11 flow measurement
To measure the speed of Rab11 vesicles, Tg(actb2:Rab11a-NeonGreen) oocytes (wild-type or exposed to 50 μM Taxol for 1 h) labeling Rab11 vesicles were imaged with an inverted Leica SP8 confocal microscope equipped with a Leica 40× objective. Multiple kymographs were acquired using Fiji along 10 pix-wide lines covering different oocyte axes. Rab11 flow speeds were then measured from the slopes of lines fitted to the Rab11 vesicles flowing towards the oocyte surface in the kymographs.

Cortical granules depletion ratio analysis
To calculate the ratio of Cgs depletion upon egg activation, Tg(actb2:Rab11a-NeonGreen) oocytes labeling Rab11 were exposed to 1 μM Lysotracker marking Ygs and treated with DMSO (control), 200 μM Colchicine, 50 μM Taxol, or 100 μM Ouabain for 60 min. Consecutively, 1 μg/ml DHP was added to the solution initiating oocyte maturation, which was completed 270 min after the DHP addition. As described above, the culture medium was exchanged during the incubation process to improve oocyte quality and survival rate. Mature oocytes/eggs, identified by their cytoplasmic clarity (except for oocytes exposed to Ouabain) and their completion of GVBD, were mounted in 0.7% LMP agarose. Next, 6.5 ml E3 medium was added on top of the solidified agarose to initiate egg activation/Cg exocytosis, during which mature oocytes/eggs were imaged as detailed above. To measure the Cg depletion rate upon egg activation, Cgs in the last frame before and the first frame after the completion of egg activation were segmented using Ilastik software [62]. The depletion rate (in percentage) was then calculated as the ratio of Cgs that underwent exocytosis during activation and thus were absent in the final image compared to the stage before activation onset.

Brightfield intensity measurement
To measure brightfield intensity during oocyte maturation, a 225 μm × 225 μm region was defined within the oocyte center, and the averaged intensities over time were obtained using Fiji.

K + vesicle number analysis
Superficial stacks of oocytes injected with 0.5 nl of 0.7 mg/ml ION Potassium Green-2 TMA + Salt, K + indicator (Abcam), and 2 mg/ml of 10 KDa Dextran Alexa Fluor 647 (to label ooplasm) diluted in nuclease-free water were imaged with an inverted Leica SP8 confocal microscope equipped with a Leica 40× objective. K + vesicles were then segmented using Ilastik software, and their number was counted over time with a custom-designed MATLAB script.

Quantification and statistical analysis
Statistical analysis was performed using Prism 8 (GraphPad Software). Data are represented as mean ± SEM and analyzed with the Mann-Whitney test. A P value < 0.05 was considered statistically significant. Sample size and P values are mentioned within the figure legends. to or in the vicinity of GV to label nucleoplasm, normalized to its distribution at the time point before GVBD, at 30 (cyan), 60 (magenta), and 120 min (green) after GVBD measured over the oocyte circumference (the yellow line in B, N = 1, n = 6). Circumferences of 0 and 800 μm correspond to the oocyte AP and VP, respectively. Asterisks track peak values for each curve. See Table G in S1 Data for underlying data. (C) Changes in phase fractions for ooplasm (left, green), Yg (middle, magenta), and Cg (right, cyan) between 120 and 270 min after imaging start in the absence of the maturation hormone DHP. Normalized (norm) radii of 0.5 and 1 correspond to the oocyte interior and cortex, respectively (N = 2, n = 6). See Table H Table D in S2 Data for underlying data. (B) Brightfield images of oocytes exposed to DMSO, 250 μM Colchi, or 25 μM Taxol (stabilizing microtubules), which were induced to undergo oocyte maturation for 270 min and subsequently activated by exposure to E3 medium for 30 min. (B') Chorion elevation, measured as chorion diameter normalized to the oocyte diameter, of oocytes exposed to DMSO (blue, control, N = 3, n = 44), Colchi (dark green, N = 3, n = 36), or Taxol (light green, N = 3, n = 34). See Table E Table F in S2 Data for underlying data. (C) CyclinB intensity normalized to its value at GVBD onset and measured at the AP, first 400-500 μm along the circumference line shown in Fig 2F and 2F', of oocytes injected with CyclinB-GFP mRNA over time (N = 2, n = 7). Meiosis stages in (B) and (C) are indicated according to CyclinB dynamics. See Table G Table F in S3 Data for underlying data. (C) Microtubule contraction type distribution (1 for normal asters, 0.5 for large-scale contractions, and 0 for abnormal asters) in Tg(Xla.Eef1a1:dclk2a-GFP) oocytes labelling microtubules exposed to DMSO (blue, N = 2, n = 9) or Cili D (red, N = 2, n = 37). See Table G Table E in S4 Data for underlying data. (B) Fluorescence images of Tg(actb2:Rab11a-NeonGreen) oocytes marking Rab11positive vesicles (green) and exposed to Lysotracker to label Ygs (magenta) pre-(left), during (middle), and post-activation (right) with E3 medium and treated with DMSO (control, top), 200 μM Colchicine (Colchi, middle), or 50 μM Taxol (bottom). Cgs (black) are identified by their exclusion of Lysotracker and Rab11 signal. On the right of the fluorescence images (for pre-and post-activation) segmented Cg are shown. On the right of the fluorescence images (for activation) zoomed-in images are shown of the ROI demarcated by the white dashed boxes in the fluorescence images. Arrowheads mark exemplary Cg undergoing exocytosis. (B') Cg depletion ratio upon egg activation for oocytes exposed to DMSO (blue, N = 3, n = 12), 200 μM Colchi (green, N = 3, n = 14) or 50 μM Taxol (black, N = 3, n = 10). See Table F in S4 Data for underlying data. (C) Microtubule aster density (count in 10 5 μm 2 ) for oocytes injected with 350 pg Rab11-mcherry (blue, N = 3, n = 9) or Rab11S25N-mCherry (magenta; dominant negative, N = 3, n = 9) mRNA during oocyte maturation. See Table G in S4 Data for underlying data. (D) Changes in phase fractions of Yg (magenta) and Cg (cyan) for oocytes injected with 350 pg Rab11-mcherry (left, N = 2, n = 6) or Rab11S25N-mCherry (right, N = 2, n = 6) between 120 and 270 min after maturation onset. Normalized (norm) radii of 0.5 and 1 correspond to the oocyte interior and cortex, respectively. See Table H Table D in S5 Data for underlying data. (B) Temporal projection of segmented Ygs in oocytes exposed to DMSO (control, left) or Ouabain (right) (same oocytes as in Fig 5B) between 120 and 270 min after maturation onset. (C) Fluorescence images of stage V Tg(hsp:clip170-GFP) oocytes labeling the ooplasm exposed to DMSO (control, left) or 100 μM Ouabain (right). Ygs and Cgs are depicted by their exclusion of ooplasmic signal. Blue lines mark the blastodisc height as measured in (C'). (C') Blastodisc clearance, measured as the height of blastodisc at the end of the maturation process, for oocytes exposed to DMSO (control, blue, N = 4, n = 29) or 100 μM Ouabain (magenta, N = 3, n = 13). See Table E in S5 Data for underlying data. (D) Maximum fluorescence intensity projection of oocytes exposed to Lysotracker to mark Yg 120 min after maturation onset. (D') Yg area averaged for the first, second, and third hour (h) after maturation onset along the oocyte AV axis (N = 3, n = 8). See Table F in S5 Data for underlying data. (E) Fluorescence images of stage III Tg(actb1:Utr-GFP) oocytes labeling F-actin exposed to 100 μM Ouabain before (stage III) and 60, 120, and 270 min after maturation onset. The dashed line indicates the GV region. (F) Fluorescence images of stage III Tg(Xla.Eef1a1: dclk2a-GFP) oocytes labeling microtubules exposed to DMSO (top) or 100 μM Ouabain (bottom) at 60, 120, 180, 240, and 270 min after maturation onset. (F') Microtubule aster density for oocytes exposed to DMSO (blue, N = 3, n = 9) or Ouabain (magenta, N = 2, n = 7) during oocyte maturation. See Table G in S5 Data for underlying data. (G) Fluorescence images of Tg (actb2:Rab11a-NeonGreen) oocytes marking Rab11-positive vesicles (green) and exposed to Lysotracker to label Yg (magenta) pre-(left) and during (right) activation with E3 medium for oocytes exposed to DMSO (control, top) or 100 μM Ouabain (bottom). Cg are identified by their exclusion of Lysotracker and the ooplasmic signal. The images on the right of the fluorescence images (for pre-activation) show segmented Cg. The images on the right of the fluorescence images (for activation) are zoomed-in images of the regions outlined by the white dashed boxes in the fluorescence images. Arrowheads mark the colocalization of Rab11-positive vesicles with Cg. (H) Fluorescence images (top row) of stage III oocytes injected with K + indicator (K + -Green, green) and Dextran Alexa Fluor 647 to mark ooplasm (magenta) exposed to 100 μM Ouabain before (stage III) and 150 and 270 min after maturation onset. Cg (black) are identified by their exclusion of both Dextran and K + -Green. Images in bottom row show segmented K + vesicles identified from the images in the top row. Dashed lines mark the oocyte outline. (H') Number of K + vesicles in superficial stacks of WT oocytes (blue, N = 2, n = 8, same data as in Fig 5C') or oocytes exposed to 100 μM Ouabain (magenta, N = 2, n = 12) during maturation. See Table H in S5 Data for underlying data. (I) Fluorescence images of stage III Tg(hsp:clip170-GFP) oocytes labelling the ooplasm (cyan) exposed to Lysotracker to mark Yg (magenta) for oocytes and treated with 100 μM PBb before (stage III) and 120, 180, and 270 min after maturation onset. Cg (black) were identified by their exclusion of both Clip-170-GFP and Lysotracker. (I') Increase in average Yg area at the end of oocyte maturation normalized to its value at stage III for oocytes exposed to DMSO (green, N = 3, n = 15) or PBb (orange, N = 3, n = 7). See Table I Figs 1A", 1B', 1C', 1D', 1D"', S1A', S1B" and S1C.