PLK-1 suppresses centrosome maturation and microtubule polymerization to ensure faithful oocyte meiosis

In most cell types, centrosomes mediate cell division by providing organizational cues for the formation of a microtubule-based bipolar spindle. Centrosomes are comprised of a centriole pair and surrounding pericentriolar material (PCM) that helps to both assemble the spindle by nucleating microtubules and maintain bipolarity by organizing the microtubule minus ends into two poles. In contrast, oocyte meiosis is a specialized form of cell division in which centrosomes are typically absent. Nonetheless, bipolar spindles still form, demonstrating that alternate mechanisms must be utilized in oocytes to ensure faithful meiosis.

In the absence of centrosomes, mouse oocytes form multiple transient structures called acentriolar microtubule-organizing centers (aMTOCs) that nucleate microtubules and facilitate bipolar spindle assembly (Schuh and Ellenberg, 2007). Interestingly, aMTOCs are not essential for spindle assembly, as spindles can still form in mouse oocytes depleted of these structures (So et al., 2022). Furthermore, aMTOCs have not been reported in human, bovine, or porcine oocyte spindles (So et al., 2022; Wu et al., 2022), further suggesting the dispensability of these structures in assembling a functional bipolar spindle. However, the mechanisms that guide acentrosomal spindle formation in the absence of aMTOCs remain unclear.

We utilized Caenorhabditiselegans oocytes as a model to reveal such mechanisms, since this system also does not rely on distinct MTOC structures for spindle formation (Connolly et al., 2015; Wolff et al., 2016). Instead, microtubules nucleate in the vicinity of chromosomes upon nuclear envelope breakdown (NEBD) and form a cage-like structure. The microtubule minus ends are then sorted outward to form multiple nascent poles that subsequently coalesce to form a bipolar spindle (Wolff et al., 2016). While this process is not centrosome mediated, many factors that are centrosome associated during mitosis are utilized during meiosis. The kinase AIR-1^{Aurora A} facilitates PCM recruitment and centrosome maturation during mitosis (Hannak et al., 2001) and is also required for spindle regulation during oocyte meiosis (Sumiyoshi et al., 2015). Similarly, microtubule-associated proteins TAC-1^TACC and ZYG-9^XMAP215 regulate microtubule assembly at centrosomes (Bellanger and Gönczy, 2003; Le Bot et al., 2003; Matthews et al., 1998) and also stabilize acentrosomal poles in oocytes (Cavin-Meza et al., 2022; Harvey et al., 2023). Thus, proteins found at the centrosome during mitosis can be repurposed during acentrosomal cell division to perform alternate functions.

Polo-like kinase 1 (PLK-1) is a highly conserved kinase that plays essential roles during mitosis in various organisms, including mediating bipolar spindle assembly (Golsteyn et al., 1995; Hartwell et al., 1973; Llamazares et al., 1991; Sunkel and Glover, 1988) and PCM recruitment (Cabral et al., 2019; Ohta et al., 2021). Several targets of PLK-1 during mitosis have been identified, including the microtubule-nucleating factor γ-tubulin and SPD-5, a scaffolding protein that helps build the PCM (Haren et al., 2009; Lane and Nigg, 1996; Rios et al., 2024; Woodruff et al., 2015). Additionally, PLK-1 mediates NEBD during C. elegans mitosis by phosphorylating the nuclear lamina (Chase et al., 2000; Martino et al., 2017; Rahman et al., 2015; Velez-Aguilera et al., 2020). While PLK-1’s roles during mitosis have been extensively studied, its functions in oocyte meiosis are less well characterized.

Work in mouse oocytes demonstrated that PLK1 is required for recruitment of centrosomal proteins to aMTOCs and for the subsequent assembly of a bipolar spindle (Clift and Schuh, 2015; Little and Jordan, 2020; Solc et al., 2015), thereby implicating PLK1 in aMTOC-driven spindle formation. A recent study of C. elegans oocytes reported spindle defects following PLK-1 inhibition, suggesting that this kinase contributes to aMTOC-independent spindle organization as well (Taylor et al., 2023). However, the specific roles that PLK-1 plays in spindle assembly and whether it is also necessary after spindle formation to ensure faithful meiosis remain unknown.

Here, we address these questions by utilizing the auxin-inducible degradation (AID) method to rapidly deplete PLK-1 from C. elegans oocytes and employing an analog-sensitive plk-1 allele to specifically inhibit kinase activity. Using these approaches, we find that PLK-1 is required to maintain bipolar spindle stability, prevent excess microtubule nucleation throughout the oocyte, and inhibit PCM recruitment to the sperm-provided centrioles, thus revealing novel roles for this conserved kinase in oocyte meiosis that are distinct from its functions in mitosis.

We first sought to test whether PLK-1 plays a role in assembling and stabilizing the acentrosomal spindle. A previous study demonstrated that PLK-1 depletion via RNAi inhibits NEBD in oocytes, thereby preventing spindle formation (Chase et al., 2000). Therefore, we utilized the AID system to achieve rapid protein depletion (Divekar et al., 2021; Zhang et al., 2015); we reasoned that this would allow us to deplete PLK-1 after the nuclear envelope had already broken down to assess subsequent roles for PLK-1 during oocyte meiosis. Using CRISPR/Cas9, we generated a strain in which PLK-1 is tagged at the endogenous locus with both a GFP and degron tag on the N terminus, while the ubiquitin ligase TIR1 is expressed by a germline-specific promoter (Fig. 1 A). Notably, this strain exhibited low embryonic lethality (1.2% lethality; Table 1), indicating that the tags did not have a major effect on PLK-1 function and the TIR1 transgene did not negatively contribute to fitness. Additionally, we assessed the localization of the tagged PLK-1 protein and found that it localized to the centrosomes and holocentric kinetochores during mitosis, as well as to the chromosomes, spindle poles, and midzone during various stages of oocyte meiosis (Fig. S1), consistent with existing literature (Chase et al., 2000; Schmucker and Sumara, 2014; Taylor et al., 2023). Thus, the tags do not affect overall worm fitness or impair protein localization.

To validate that we could recapitulate the known PLK-1 depletion phenotype using this strain, we incubated worms on plates containing 1 mM auxin for 18 h (“long-term AID,” Fig. 1 A). This treatment resulted in reduced PLK-1 staining and 100% embryonic lethality (Fig. 1 C and Table 1). We also observed a large number of oocytes with a hollow area of tubulin staining around the condensed chromosomes, consistent with the size of the nucleus (Fig. 1 C, row 2). Since previous work has shown that PLK-1 promotes NEBD (Chase et al., 2000; Martino et al., 2017; Velez-Aguilera et al., 2020), we assessed the localization of the inner nuclear membrane protein emerin (EMR-1) (Lee et al., 2000) and confirmed that the nuclear envelope was indeed present in oocytes with these hollow areas of tubulin (Fig. S2 A). To ensure that these were not premature oocytes that had not yet triggered NEBD, we evaluated the oocytes using a number of criteria (Fig. 1 B). Maturing oocytes in the C. elegans germline progress in an assembly like fashion toward the spermatheca, where they are fertilized. As oocytes mature, the nuclear envelope breaks down and the meiotic spindle begins to form (McCarter et al., 1999). Since oocytes undergo morphological changes during this process, mature oocytes in the +1 position can be identified based on the position of the chromosomes relative to the cortex, the shape of the oocyte, and the presence of sperm (Fig. 1 B). We found that following long-term PLK-1 AID, the majority of mature oocytes lacked a bipolar spindle and instead displayed either an empty region of tubulin surrounding chromosomes (“NE intact”; 44/81 oocytes, 54%) or a microtubule array where some tubulin density is present but no bipolar spindle has assembled (“MT cage”; 18/81 oocytes, 22%) (Fig. 1 D). In both cases we observed that EMR-1 was still present in long-term auxin-treated oocytes, while it was absent in controls (Fig. S2; and Videos 1, 2, and 3). Our findings thus support previously published work implicating PLK-1 in promoting NEBD and validate the use of our AID strain to investigate PLK-1 function.

In our long-term AID experiments, we observed a subset of PLK-1–depleted oocytes where NEBD had occurred, yet yielded disorganized spindles (17/81 oocytes; 21%) (Fig. 1 D), suggesting that PLK-1 may be required for normal spindle assembly. This is in line with previous studies that reported spindle defects using different methods of PLK-1 inhibition (Chase et al., 2000; Taylor et al., 2023).

To better understand how PLK-1 might promote bipolar spindle assembly, we next attempted to remove PLK-1 from oocytes after NEBD and assess its effects on spindle organization. Soaking worms in a 5 mM auxin solution for 45 min (“short-term AID,” Fig. 1 A) was sufficient to achieve a consistent loss of PLK-1 staining, and spindles that were able to form in these conditions were usually aberrant (Fig. 2, A and B), thus implicating PLK-1 in spindle assembly. However, this short-term AID treatment also resulted in a large percentage of oocytes with intact nuclear envelopes (42/68 oocytes; 61.8%, Fig. 2 B), similar to long-term AID (Fig. 1 D), suggesting that PLK-1 was being depleted prior to NEBD in many oocytes. Because we were unable to achieve robust PLK-1 depletion with shorter soaking times, we instead used RNAi to enrich for oocytes in which the nuclear envelope had already broken down by depleting the anaphase-promoting complex component EMB-30. This condition arrests oocytes in metaphase I, allowing us to enrich for oocytes with assembled bipolar spindles (Furuta et al., 2000). Under these conditions, we found that the majority of auxin-treated spindles were disorganized (74/89; 83.1%, Fig. 2, A and B), supporting a role for PLK-1 in promoting proper spindle architecture.

Given that we are using metaphase arrest in these experiments, we infer that PLK-1 was depleted from most oocytes after the spindle had already achieved bipolarity, suggesting that PLK-1 is required to maintain spindle stability. To directly test this hypothesis, we generated a PLK-1 AID strain that also expressed mCherry::histone and GFP::tubulin to visualize the effects of PLK-1 removal from preformed spindles in real time. We performed ex utero live imaging of PLK-1 depletion by dissecting oocytes into a 500 µM auxin solution and immediately filming the spindles (“acute AID,” Fig. 1 A). Live imaging of metaphase-arrested (Fig. 2 C; and Videos 4 and 5) and unarrested oocytes (Fig. 2 D; and Videos 6 and 7) revealed that spindles began to exhibit defects soon after PLK-1 depletion was initiated. In control conditions, metaphase-arrested spindles (7/7 spindles) as well as unarrested spindles (8/8 spindles) retained their structure. In contrast, when oocytes were dissected into auxin, the spindle poles moved apart while remaining relatively focused, and the midspindle region lost its integrity. This resulted in a disorganized structure during metaphase arrest (6/6 spindles) that in unarrested conditions was unable to progress to anaphase (7/7 spindles). Because we observed similar spindle defects in both arrested and unarrested oocytes, these phenotypes are likely not a byproduct of the metaphase arrest. Altogether, our results support a role for PLK-1 in both assembling and stabilizing the oocyte spindle.

In characterizing the aberrant spindles that formed following PLK-1 AID, we assessed the localization of the microtubule minus end–binding protein ASPM-1, which marks spindle poles. This analysis confirmed that PLK-1–depleted spindles were disorganized, as they lacked the two defined ASPM-1–marked poles present in control spindles (Fig. 3 A). Moreover, this analysis revealed an unexpected phenotype not previously reported using other depletion methods (Chase et al., 2000; Taylor et al., 2023). Control-treated oocytes typically do not have distinct concentrations of tubulin density or ASPM-1 in the cell other than at the meiotic spindle. However, after short-term (Fig. 3, A and C) and long-term (Fig. 3, B and C) PLK-1 depletion, we observed ectopic microtubule asters that were marked by distinct populations of ASPM-1.

Curiously, we did not observe the formation of these microtubule asters following acute PLK-1 AID (Fig. 2, C and D; and Fig. 4 B) when oocytes were dissected directly into auxin and then filmed for 15 min to assess immediate effects on the spindle. However, residual GFP::PLK-1 signal was often still apparent on the chromosomes in those depletion experiments (Fig. 2 C), indicating that we were not achieving complete depletion of PLK-1 in that timeframe. This is consistent with the fact that we needed to soak worms in auxin for 45 min in our short-term AID experiments to observe complete loss of PLK-1 staining (Fig. 2 A). Thus, we modified our acute AID method to achieve more robust depletion in line with short-term AID treatment. To this end, we soaked the worms in auxin for 30 min prior to dissection and then performed ex utero live imaging for 15 min (“acute AID + pre-treatment,” Fig. 1 A). In control conditions, no asters or other significant tubulin densities were present throughout the oocyte (7/7 oocytes, Fig. 4 A and Video 8). Conversely, following auxin treatment, ectopic asters formed throughout the oocyte and converged with each other over time, resulting in large tubulin-rich structures (7/7 oocytes, Fig. 4 A and Video 9). The lack of ectopic asters present after immediate administration of auxin suggests that this phenotype only arises after robust PLK-1 depletion, as compared with the spindle defects that were apparent even when PLK-1 was partially depleted from the oocyte (Fig. 2 D).

Previous work has shown that depletion of the microtubule depolymerase KLP-7^MCAK via RNAi results in the formation of ectopic microtubule asters (Gigant et al., 2017), reminiscent of the PLK-1 depletion phenotype. Moreover, there is evidence that mammalian PLK1 phosphorylates MCAK during mitosis to promote its depolymerase activity (Zhang et al., 2011). Therefore, we hypothesized that the ectopic asters seen upon PLK-1 depletion might be caused by misregulation of KLP-7. If this were the sole mechanism by which ectopic aster formation was occurring following PLK-1 AID, we would expect co-depletion of PLK-1 and KLP-7 to exhibit a similar phenotype as either single depletion alone. As expected, RNAi-mediated depletion of KLP-7 resulted in the formation of some ectopic microtubule asters (8/8 oocytes, Video 10), while acute PLK-1 AID alone (without auxin pre-treatment) did not result in ectopic aster formation (Fig. 4, B and C; and Video 11). Strikingly, when we combined klp-7(RNAi) with acute PLK-1 AID, there was a significant increase in microtubule polymerization compared with either individual depletion condition alone (8/8 oocytes, Fig. 4, B and C; and Video 12). These co-depleted oocytes had an increased number of ectopic microtubule asters, and the asters that formed had increased tubulin intensity compared with those that formed following klp-7(RNAi) alone (Fig. 4 C). This enhanced phenotype suggests that the asters observed following PLK-1 depletion are not solely caused by misregulation of KLP-7. However, we cannot rule out the possibility that KLP-7 is not fully depleted in our RNAi conditions; if this is the case then PLK-1 depletion could result in an enhanced phenotype by suppressing residual KLP-7. While we favor the first interpretation, in either scenario our results are consistent with the possibility that PLK-1 regulates KLP-7 in oocytes.

Interestingly, when analyzing microtubule polymerization in the short-term and long-term PLK-1 depletion conditions, we often noticed excess microtubule density adjacent to the sperm DNA (Fig. 5, A–C). These asters were marked by clusters of ASPM-1, suggesting that the minus ends are organized into distinct foci at these sites (Fig. 5, A and B). This ectopic microtubule polymerization was striking given what is known about the C. elegans germline. During oogenesis, the maternal centrioles are eliminated prior to the turn of the gonad arm (Mikeladze-Dvali et al., 2012). Upon fertilization, the oocyte inherits a centriole pair from the sperm, but the sperm-provided centrioles do not recruit maternal PCM or nucleate microtubules until the completion of the meiotic divisions (McNally et al., 2012). This led us to speculate that the ectopic polymerization we observed near the sperm may be caused by premature centrosome maturation, implicating PLK-1 in regulating this process. Consistent with this hypothesis, we found that PLK-1 localizes to the sperm centrioles, as demonstrated by its co-localization with ZYG-1, a kinase regulating centriole replication (O’Connell, 2002) (Fig. 5 D).

To determine if PLK-1 normally inhibits centrosome maturation in oocytes, we assessed whether PCM components were recruited to the vicinity of the sperm DNA after PLK-1 AID. We first investigated the localization of SPD-5, a central structural scaffold protein involved in recruiting downstream PCM components (Hamill et al., 2002). Although SPD-5 is typically not visible at sperm centrioles until the completion of the meiotic divisions (McNally et al., 2012), we observed distinct SPD-5 foci adjacent to the sperm DNA following both short-term and long-term PLK-1 depletion (Fig. 6, A and B). We quantified the number of SPD-5–positive foci per oocyte and found that these foci are present in over 89% of oocytes in both AID conditions (Fig. 6, A and B). We speculate that instances of two foci represent SPD-5 localizing to both the sperm-derived mother and daughter centrioles and that one focus represents cases where we cannot resolve the two centrioles from each other.

We next assessed the localization of TAC-1, a PCM component associated with microtubule assembly in the C. elegans embryo (Le Bot et al., 2003). While TAC-1 does not typically concentrate near the sperm (McNally et al., 2012), in both short-term and long-term PLK-1 AID conditions, we observed distinct localization of TAC-1 near the sperm DNA (Fig. 6, C and D). Both the TAC-1 and SPD-5 concentrations were often coincident with sperm-adjacent microtubule asters (Fig. 6, A–D). Taken together, these results demonstrate that depletion of PLK-1 leads to premature recruitment of PCM components to the sperm centrioles, indicative of premature centrosome maturation.

Notably, we did not observe SPD-5 or TAC-1 concentrated at the microtubule asters located in other regions of the oocyte (n = 0/33 oocytes; n = 1/16 oocytes, respectively) (Fig. S3). These asters are therefore compositionally different from the sperm-adjacent asters, suggesting that they do not arise from PCM recruitment. Thus, PLK-1 appears to regulate microtubule assembly in the oocyte via multiple distinct mechanisms.

Although the mechanisms suppressing centrosome maturation in C. elegans oocytes are poorly understood, a previous study found that depletion of either kinesin-1 or its cargo adaptor KCA-1 causes premature recruitment of PCM to the sperm centrioles (McNally et al., 2012). Therefore, the effects of PLK-1 depletion on the sperm centrioles could be related to misregulation of these proteins. Consistent with this hypothesis, we found that short-term and long-term depletion of PLK-1 resulted in altered localization of KCA-1 (Fig. 7, A–D). In control conditions, KCA-1 either forms a cloud around the sperm DNA, which has been proposed to directly inhibit PCM recruitment to the sperm centrioles, or does not display localization near the sperm, depending on the stage of cell division (McNally et al., 2012) (Fig. 7, A and B). Following short-term (Fig. 7, A and C) and long-term (Fig. 7, B and D) PLK-1 depletion, KCA-1 instead appeared as either 1 or 2 distinct puncta near the sperm in most oocytes. Thus, PLK-1 appears to regulate KCA-1 and, presumably, kinesin-1 localization in oocytes.

To further explore a connection between PLK-1 and kinesin-1/KCA-1, we assessed the distance between the sperm DNA and the oocyte spindle in PLK-1–depleted oocytes. A previous study demonstrated that depletion of KCA-1 caused these structures to move closer together, sometimes resulting in capture of the meiotic spindle by the sperm aster (McNally et al., 2012). Similarly, we found that PLK-1 depletion shortened the average distance between the sperm DNA and the oocyte chromosomes in both long-term (29.8 µm) and short-term AID conditions (21.5 µm) compared with controls (36.6 and 35.2 µm, respectively; Fig. 7 E). We also found cases where the sperm aster was adjacent to the oocyte chromosomes (Fig. 6 A), suggesting a role for PLK-1 in ensuring that the meiotic spindle and sperm DNA remain positionally separate within the oocyte. These findings support the interpretation that by preventing premature centrosome maturation, PLK-1 allows the meiotic divisions to be completed without interference.

Finally, we sought to understand whether the phenotypes observed following PLK-1 AID required PLK-1’s catalytic activity. To do this, we utilized an analog-sensitive strain, in which PLK-1 has been mutated at two amino acids (plk-1^as, Fig. 8 A) so that its kinase activity can be inhibited by addition of the ATP analog 1-NM-PP1 (Bishop et al., 2000; Gómez-Cavazos et al., 2020). Remarkably, treatment of this plk-1^as strain with 1-NM-PP1 for 12 min was sufficient to recapitulate all of the phenotypes we observed using short- and long-term AID. First, treated oocytes had disorganized spindles (48/50 oocytes; 96%, Fig. 8 B), consistent with defects previously reported using this strain (Taylor et al., 2023). Moreover, we observed the formation of ectopic microtubule asters near the oocyte chromosomes and sperm DNA (Fig. 8, C–E); the frequencies of these asters throughout the oocytes were comparable with PLK-1 AID (Fig. S4). Finally, 1-NM-PP1 treatment resulted in decreased sperm-to-spindle distances (Fig. 8 F) and SPD-5 localization near the sperm DNA (Fig. 8, G and H). Taken together, these results indicate that PLK-1’s kinase activity is required to promote acentrosomal spindle assembly and stability, to suppress excessive microtubule polymerization, and to prevent premature centrosome maturation in oocytes.

Taken together, our work reveals important new roles for PLK-1 during oocyte meiosis (Fig. 9). PLK-1 dynamically localizes to various regions of the oocyte spindle, where it is required for both bipolar spindle assembly and stability. Moreover, we found that PLK-1 prevents excess microtubule polymerization throughout the entire oocyte and also localizes to the sperm-provided centrioles, where it prevents premature centrosome maturation. As these functions are all dependent on kinase activity, we propose that PLK-1 phosphorylates downstream target proteins in oocytes to regulate these important aspects of the meiotic divisions.

In mouse oocytes, aMTOCs comprised of PCM proteins nucleate microtubules and form asters that assist with spindle assembly; aMTOCs collect along the nuclear envelope and fragment into smaller structures upon NEBD that distribute to the two spindle poles (Luksza et al., 2013; Schuh and Ellenberg, 2007). It has been shown that PLK-1 is required for aMTOC fragmentation, thus facilitating proper spindle formation (Clift and Schuh, 2015; Little and Jordan, 2020; Solc et al., 2015). Now, we report that PLK-1 also is required for spindle assembly in C. elegans oocytes, a system that is not thought to rely on aMTOCs (Connolly et al., 2015; Wolff et al., 2016). Thus, PLK-1 contributes to acentrosomal spindle assembly in multiple ways.

Notably, although aMTOCs have not been previously reported in C. elegans oocytes, we observed microtubule asters throughout the oocyte after PLK-1 depletion. Thus, we considered the possibility that these asters might be transient aMTOCs that are normally difficult to image but become large and/or stable enough to visualize upon PLK-1 depletion. To investigate this hypothesis, we assessed whether known aMTOC components localize to the ectopic asters. We found that SPD-5, whose homolog CDK5RAP2 localizes to aMTOCs in mouse oocytes (So et al., 2019), was not detectable in these ectopic asters. Similarly, the centrosome-associated protein TAC-1, whose homolog TACC3 localizes to a singular aMTOC during the early stages of spindle assembly in human oocytes (Wu et al., 2022), was also not present on the microtubule asters. Thus, if the asters observed upon PLK-1 depletion are stabilized aMTOCs, they are comprised of alternate proteins than the canonical proteins associated with aMTOCs in other systems.

While spindle defects were apparent soon after oocytes were dissected into auxin, visualizing ectopic microtubule aster formation required pre-treating worms prior to live imaging. This requirement could be for several reasons. One possibility we considered is that it takes time for PLK-1 depletion to influence downstream targets that in turn trigger excess microtubule polymerization. However, since we were able to see ectopic aster formation in the analog-sensitive plk-1 strain after a short 1-NM-PP1 incubation, we think that this explanation is unlikely. Instead, we hypothesize that robust depletion of PLK-1 is necessary to have such a dramatic effect on the oocyte, necessitating the long auxin incubation time.

Microtubule growth has been linked to the amount of free tubulin in the cell (which should be the same in the oocyte regardless of PLK-1 depletion/inhibition) as well as the amount of ZYG-9 and TAC-1 (Srayko et al., 2005), which are associated with microtubule assembly in the C. elegans embryo (Bellanger and Gönczy, 2003). Since TAC-1 does not localize to the ectopic asters throughout the oocyte, we infer the same for ZYG-9 because these proteins co-localize in mitosis and meiosis (Cavin-Meza et al., 2022; Le Bot et al., 2003). Thus, we speculate that the asters throughout the oocyte may be due to altered activity of other microtubule regulatory factors. A plausible candidate for such a factor is the microtubule depolymerase KLP-7^MCAK, as a previous study reported ectopic asters upon KLP-7 depletion (Gigant et al., 2017). In our experiments, we found that co-depletion of PLK-1 and KLP-7 had a more extreme phenotype than either single depletion alone, which suggests that the increased microtubule polymerization seen upon PLK-1 depletion may not be solely due to KLP-7 misregulation. However, our results do not exclude the possibility that PLK-1 regulates KLP-7; indeed, since PLK1 regulates MCAK activity in mitosis (Zhang et al., 2011), we think that it is a likely target. Other known PLK-1 targets include canonical microtubule nucleators γ-tubulin and TPXL-1^TPX2 (Roostalu et al., 2015; Strome et al., 2001; Woodruff et al., 2017). Misregulation of these proteins could also contribute to the PLK-1 depletion phenotype, though depletion of these proteins has not been reported to cause meiotic spindle defects in past studies (Bobinnec et al., 2000; Ozlü et al., 2005). Future work aimed at identifying PLK-1 phosphorylation targets that regulate microtubule polymerization in the oocyte may elucidate how ectopic aster formation is normally suppressed.

Another question raised by our findings is how PLK-1 suppresses centrosome maturation until the completion of the meiotic divisions. This is especially puzzling since PLK-1 plays the opposite role during mitosis, where it facilitates the recruitment of SPD-5 and other PCM components to centrosomes (Lane and Nigg, 1996; Lee and Rhee, 2011). Conversely, in oocytes, we find that PLK-1 localizes to the sperm-derived centrioles and appears to block PCM accumulation.

How PLK-1 performs these two opposite functions is a fascinating question for future study. Previous work demonstrated that kinesin-1 and its cargo adaptor KCA-1 prevent the accumulation of SPD-5 and other PCM components on the sperm centrioles prior to the completion of meiosis (McNally et al., 2012). Our finding that PLK-1 depletion results in altered KCA-1 localization raises the possibility that redistribution of these proteins triggers premature PCM recruitment. It has been proposed that kinesin-1/KCA-1 may act by forming a shell around the sperm DNA that blocks PCM assembly until the completion of the meiotic divisions (McNally et al., 2012). If this model is correct, then disruption of this shell upon PLK-1 depletion would initiate premature centrosome maturation. In this scenario, PLK-1 could potentially act by phosphorylating kinesin-1 and/or KCA-1 to control their localization. Alternatively, PLK-1 could phosphorylate PCM components to block their accumulation during meiosis or phosphorylate other regulatory factors. Further experiments examining known PLK-1 consensus motifs (Nakajima et al., 2003) and manipulating putative phosphorylation sites in candidate proteins may help elucidate the mechanism of how PLK-1 depletion is affecting premature centrosome maturation.

Another fascinating question relates to the mechanism by which SPD-5 accumulates at the sperm-derived centrioles in the absence of PLK-1. During centrosome maturation in C. elegans mitosis, SPD-5 oligomerizes and builds a scaffold for the recruitment of other PCM components (Ohta et al., 2021; Woodruff et al., 2015; Wueseke et al., 2016). Since PLK-1–mediated phosphorylation is thought to drive SPD-5 multimerization (Nakajo et al., 2022; Rios et al., 2024), how does SPD-5 accumulate at centrioles and drive the recruitment of other PCM components if PLK-1 is not present? One possibility is that SPD-5 accumulates without being phosphorylated under these conditions, which would suggest that phosphorylation is not always required for SPD-5 multimerization. Alternatively, it is possible that another kinase may be able to phosphorylate SPD-5 when PLK-1 is depleted. Future experiments distinguishing between these possibilities will enhance our understanding of SPD-5 multimerization and PCM assembly.

Interestingly, another known function of PLK-1 is to promote disengagement of the centriole pair prior to mitosis, a process that is required for subsequent PCM recruitment (Tsou et al., 2009). Consistent with PLK-1’s alternate role in our system, our data suggest that during oocyte meiosis, PLK-1 may prevent premature separation of the two centrioles. Typically, the mother and daughter centrioles disengage after the completion of meiosis II (Cabral et al., 2013); however, we often observed two distinct foci of SPD-5 and KCA-1 adjacent to the sperm DNA in PLK-1–depleted oocytes. While centrioles were not labeled in these conditions, we reason that these two distinct foci mark the mother and daughter centrioles as they are prematurely separating and recruiting PCM. This contrasts with the single focus of ZYG-1/PLK-1 observed in untreated oocytes, which we infer marks the closely associated centriole pair prior to disengagement.

In human cells, Plk1 and separase function at different times to activate centriole disengagement during mitosis (Kim et al., 2015; Tsou et al., 2009). The C. elegans ortholog of separase, SEP-1, has also been shown to promote separation of the sperm-derived centrioles after the completion of female meiosis (Cabral et al., 2013). Interestingly, while sep-1(RNAi) results in defective centriole separation during mitosis, co-depletion of SEP-1 and KCA-1 via RNAi rescues this phenotype (Cabral et al., 2013), indicating a potential link between centriole separation and KCA-1. Given that we observed altered localization of KCA-1 following PLK-1 depletion as well as centriole disengagement, we postulate that premature centriole disengagement may be caused by KCA-1 mislocalization.

In summary, reported in this study are multiple novel roles for PLK-1 in oocyte meiosis that are distinctly different from the canonical roles of this kinase in mitosis. These alternate roles for PLK-1 during oocyte meiosis provide abundant opportunities for follow-up investigations on meiotic spindle dynamics, microtubule polymerization, and centrosome maturation.

All strains used in this study are listed in Table 2. Strains PHX3354 and PHX7793 were generated by SunyBiotech via CRISPR/Cas9 editing of the endogenous plk-1 locus in the CA1199 (Zhang et al., 2015) and SMW44 background strains, respectively. OC1002 was a gift from Kevin O’Connell, Laboratory of Biochemistry and Genetics, National Institute of Diabetes and Digestive and Kidney Diseases, NIH, Bethesda, MD, USA, and OD3696 (Gómez-Cavazos et al., 2020) was obtained from the C. elegans Genetics Center. All strains were maintained at 15°C.

Adult worms were picked into a 10-μl drop of meiosis medium (0.5 mg/ml inulin, 25 mM HEPES, and 20% FBS in Leibovitz’s L-15 Media [11415-048; Gibco]) (Laband et al., 2018) on poly-L-lysine slides and dissected to remove oocytes. Slides were covered with a coverslip and slowly lowered into liquid nitrogen for 5–10 min, after which the coverslip was rapidly removed via razor blade and the slide was submerged in −20°C MeOH for 45 min. Samples were rehydrated in PBS, blocked in AbDil (PBS with 4% BSA, 0.1% Triton-X-100, and 0.02% Na-Azide) overnight at 4°C, and then incubated in primary antibodies overnight at 4°C. The following day, slides were washed three times with PBS with 0.1% Triton-X-100 (PBST) and incubated with secondary antibodies for 2 h at room temperature. After three PBST washes, samples were incubated with mouse anti–α-tubulin–FITC for 2 h at room temperature and washed again. Samples were then incubated with Hoechst (1:1,000 in PBST) for 15 min and washed twice with PBST. Finally, samples were mounted in 0.5% p-phenylenediamine, 20 mM Tris-Cl, pH 8.8, and 90% glycerol, sealed with nail polish, and stored at 4°C. Further details are available in Wolff et al. (2022).

The primary antibodies used in this study were rat–α–EMR-1 (1:50), mouse–α-Tubulin–FITC (1:500, F2168; Sigma-Aldrich [clone DM1α]), mouse–α-GFP (1:250, A11120, clone 3E6; Invitrogen), rabbit–α–ASPM-1 (1:5,000, gift from Arshad Desai, Department of Cellular and Molecular Medicine, University of California, San Diego, CA, USA [Wignall and Villeneuve, 2009]), rabbit–α–TAC-1 (1:50 [Cavin-Meza et al., 2022]), rabbit–α–SPD-5 (1:1,500, gift from Bruce Bowerman, Institute of Molecular Biology, University of Oregon, Eugene, OR, USA [Hamill et al., 2002]), and rabbit–α–KCA-1 (1:200, gift from Frank McNally, Department of Molecular and Cellular Biology, University of California, Davis, CA, USA [McNally et al., 2010]). Rat anti–EMR-1 was generated by Pocono Rabbit Farm and Laboratory Inc. using the last 14 residues of EMR-1 (C-QLKLVAETNPEDTI) and then was affinity purified. ChromoTek Spot-Label Alexa Fluor 568 (1:800, ebAF568; Proteintech) was used to detect ZYG-1 in strain OC1002. Alexa Fluor goat anti-rabbit 555 (A21428; Invitrogen) and Alexa Fluor goat anti-mouse 647 (A21235; Invitrogen) secondary antibodies were diluted 1:500 in PBST.

RNAi cultures were grown by picking individual clones from an RNAi library (Fraser et al., 2000; Kamath et al., 2003) at 37°C in LB supplemented with 100 μg/ml ampicillin. Cultures were grown overnight, centrifuged, and plated on nematode growth medium (NGM) plates containing 100 μg/ml ampicillin and 1 mM IPTG. Plates were dried overnight at room temperature in the dark. Worms were synchronized in preparation for experimentation by bleaching gravid adults, collecting resulting embryos, and incubating them on plates lacking food overnight. The following day, hatched L1s were transferred to RNAi plates and grown to adulthood at 15°C for 6 days. Further details are available in Wolff et al. (2022).

Multiple modes of auxin treatment were utilized in this study, as summarized in Fig. 1 A and briefly described here. By varying the duration of auxin treatments, we aimed to investigate different roles of PLK-1 on the meiotic spindle. Long-term AID was performed to validate the PLK-1 AID strain (i.e., to ensure that long-term depletion recapitulated previously published plk-1(RNAi) phenotypes). Short-term AID was performed to investigate the role of PLK-1 in spindle assembly, and short-term AID combined with emb-30(RNAi) was used to test whether PLK-1 was required for spindle maintenance. Acute AID was performed to assess the consequences of PLK-1 depletion in real time; this enabled us to remove PLK-1 from preformed spindles and confirm a role for PLK-1 in stabilizing the spindle. Note that for some acute AID experiments there was an additional auxin incubation to achieve better depletion (acute AID + pre-treatment). Further details are available in Divekar et al. (2021).

Long-term auxin was administered by incubating worms on NGM plates containing auxin for 18 h. Plates were prepared as normal NGM plates except for the addition of auxin dissolved in 100% EtOH for a final concentration of 1 mM auxin. Plates were stored in the dark at 4°C for no >2 mo. For experimentation, synchronized worms (described earlier) were grown to the L4 stage on standard NGM plates. Worms were then transferred onto 1 mM auxin-containing plates and incubated at 15°C for 18 h prior to dissection and immunofluorescence.

Short-term auxin treatment was performed by soaking intact worms in meiosis media containing 5 mM auxin. This solution was prepared by diluting a stock of 400 mM auxin dissolved in 100% EtOH that was protected from light. For experimentation, adult worms were picked into 10 μl of 5 mM auxin solution and incubated for 40 min in a humidity chamber to prevent evaporation of the solution. For vehicle control treatment, adults were picked into 10 μl of meiosis media with the equivalent volume of 100% ethanol and incubated identically. Following 40 min of incubation, worms were dissected (∼5 min) and then progressed through the standard immunofluorescence protocol as described above for a total of ∼45 min of auxin treatment.

Acute auxin treatment was performed by picking 10–15 worms into 10 μl of meiosis media containing 500 µM auxin. This solution was prepared using a stock of 400 mM auxin dissolved in 100% EtOH. The control solution consisted of an equivalent volume of 100% ethanol diluted in meiosis media. Adult worms were picked into a 10-μl drop of auxin or control solution and immediately dissected and mounted for ex utero live imaging as described below.

Worms were set up identically to the acute AID protocol, except for one additional incubation step. After worms were picked into 10 μl of solution, the entire slide was incubated in a humidity chamber for 30 min at room temperature to prevent evaporation of the liquid. This incubation time allowed for a more complete PLK-1 depletion as compared with acute AID. After this pre-treatment step, worms were immediately dissected and mounted for ex utero live imaging.

For experiments investigating PLK-1 catalytic activity (Fig. 8), the ATP analog 1-NM-PP1 was used to inhibit PLK-1 in the analog-sensitive plk-1 strain OD3696. Adult worms were picked into meiosis media containing 20 μM 1-NM-PP1 (Cat#13330; Cayman chemical company), dissected, and incubated for 12 min prior to submersion in liquid nitrogen and subsequent fixation. Control oocytes were soaked in meiosis media containing the equivalent volume of DMSO.

Worms were taken from desired experimental plates and dissected into 5 μl of meiosis media containing either auxin or ethanol (control) in the center of a live-imaging apparatus (Divekar et al., 2021; Laband et al., 2018). A Vaseline ring was made using a syringe to contain the sample, and an 18 × 18-mm coverslip was laid on top. The slide was inverted and immediately imaged.

Fixed imaging was performed at room temperature on a DeltaVision Core deconvolution microscope (Applied Precision) with a Photometrics CoolSnap HQ2 camera using either a 40× (NA = 1.3) or 100× (NA = 1.4) Olympus oil-immersion objective lens. Z-stacks were obtained at 0.2-μm increments. Images were acquired and deconvolved (ratio method, 15 cycles) using SoftWoRx (Applied Precision) and subsequently processed using ImageJ. Images are shown as maximum intensity projections unless otherwise indicated. This microscope is housed in the Northwestern University Biological Imaging Facility, supported by the NU Office for Research.

Live imaging was performed using a Nikon SoRa spinning disk confocal microscope with an oil-immersion 60× (1.42 NA) objective lens at room temperature. Images were acquired using a Yokogawa CSU-W1 dual-disk spinning disk unit with a 50-μm pinhole and a Hamamatsu ORCA-Fusion Digital CMOS Camera. The microscope was controlled by the Nikon SoRa imaging software NIS-Elements AR. Fourteen z-stacks at 0.5-μm increments were taken every 30 s at room temperature. Videos were processed using ImageJ to create maximum intensity projections. The Nikon SoRa microscope is housed in the Northwestern University Biological Imaging Facility, supported by the NU Office for Research.

When quantifying our results, similar phenotypes were observed in MI and MII. Therefore, all quantifications of phenotypic categories represent pooled results from MI and MII oocytes (Fig. 1 D; Fig. 2 B; Fig. 3 C; Fig. 5, A and B; Fig. 6, A–D; Fig. 7, C and D; Fig. 8, B–D and H; and Fig. S2 B). At minimum three biological replicates were performed for all experiments.

To distinguish between asters near the oocyte chromosomes and asters near the sperm DNA, a line was drawn across the center of the oocyte to bisect it into two halves. Asters were categorized based on if they were in the half containing the oocyte chromosomes versus the half containing the sperm DNA.

Sum projections were made of all z slices in the tubulin channel at t = 12 min using ImageJ. Mean gray values ranged from 0 to 255, so a threshold intensity value was set at 64 (corresponds to dark blue in Fig. 4 B). The number of asters within the oocyte that exceeded this threshold value were counted. Additionally, the fluorescence intensity value of any area within the oocyte above the threshold was measured using an ROI set to reflect the average aster size (∼4 × 4 μm). The same ROI was used to analyze all samples.

Using 40X images of the entire oocyte, sum projections of tubulin intensity were made in ImageJ from 1.4-μm above and below the center of the sperm DNA (2.8-μm stack total, 14 z-slices). The fluorescence intensity was measured surrounding the sperm DNA using an ROI of ∼4 × 4 μm. The same ROI was used to measure three other areas within the oocyte cytoplasm (away from the meiotic spindle and sperm DNA) for background subtraction of cortical tubulin intensity. The average tubulin intensity of the background was subtracted from the tubulin intensity surrounding the sperm DNA, and the resulting values were graphed.

Fluorescence intensities for DNA, PLK-1, and ZYG-1 were measured using ImageJ on single-slice projections using the same 2-μM long line, drawn from one end of the sperm DNA to just past the center of the sperm centriole, as determined by ZYG-1 staining. Directionality of the line is indicated by the white arrow.

Using 40X images of the entire cell, oocyte chromosomes and sperm DNA were rendered into 3D surfaces using the “Surfaces” tool in the 3D imaging software Imaris (Bitplane). For each cell analyzed, the center of both the oocyte chromosomes and sperm DNA was determined using the DAPI channel, and the distance between them was measured in 3D, as depicted in Fig. 7 E.

All statistics were performed using GraphPad Prism 10. The Shapiro–Wilk normality was first used to test for normality of data distribution. Subsequently, normally distributed data were analyzed using the two-tailed t test, while non-normally distributed data were analyzed using the Mann–Whitney U nonparametric test. Statistical methodology, significance values, sample sizes, and number of biological replicates are indicated in corresponding figures and/or legends.

Fig. S1 shows the localization of PLK-1 during various stages of oocyte meiosis. Fig. S2 shows the persistence of nuclear envelope marker EMR-1 after long-term auxin treatment. Fig. S3 shows that the microtubule asters near oocyte chromosomes arising from PLK-1 depletion do not recruit PCM components, unlike sperm asters. Fig. S4 shows the frequency of total ectopic microtubule asters per oocyte after short-term AID, long-term AID, and 1-NM-PP1 treatment. Videos 1, 2, and 3 step through the individual z-slices of the max projected images shown in Fig. S2 A. Video 1 shows the control treatment and Videos 2 and 3 show the long-term auxin treatment (showing an intact nuclear envelope and the microtubule cage phenotype, respectively). Video 4 shows the live imaging of a metaphase-arrested oocyte spindle without auxin, as shown in Fig. 2 C. Video 5 shows the live imaging of a metaphase-arrested oocyte spindle after acute auxin treatment, as shown in Fig. 2 C. Video 6 shows the live imaging of an unarrested oocyte spindle without auxin, as shown in Fig. 2 D. Video 7 shows the live imaging of an unarrested oocyte after acute auxin treatment, as shown in Fig. 2 D. Video 8 shows the live imaging of a pre-treated oocyte incubated in vehicle solution, as shown in Fig. 4 A. Video 9 shows the live imaging of an oocyte pre-treated with auxin solution to fully deplete PLK-1, as shown in Fig. 4 A. Videos 10, 11, and 12 show the live imaging of oocytes in which tubulin is pseudocolored to visualize intensity, corresponding to Fig. 4 B; shown are a klp-7(RNAi) oocyte without auxin (Video 10), an oocyte after auxin treatment (Video 11), and a klp-7(RNAi) oocyte treated with auxin (Video 12).

All data in this manuscript are available from the corresponding author upon reasonable request.

We would like to thank members of the Wignall lab for support and Gabriel Cavin-Meza, Emily Czajkowski, Hannah Horton, Elizabeth Lopez, Ilka Lorenzo, and Jordy Martinez in particular for critical reading of the manuscript. We are also grateful to Bruce Bowerman, Arshad Desai, Frank McNally, and Kevin O’Connell for reagents.

This work was supported by the National Institutes of Health (NIH) R01GM124354 and R01GM141386 (to S.M. Wignall) and by the NIH Reproductive Science, Medicine, and Technology training grant T32HD094699 (to J.G. Narula). Microscopy was performed at the Biological Imaging Facility at Northwestern University, supported by the NU Office of Research and the Department of Molecular Biosciences.

Author contributions: J.G. Narula: conceptualization, formal analysis, investigation, methodology, validation, visualization, and writing—original draft, review, and editing. S.M. Wignall: conceptualization, funding acquisition, methodology, project administration, supervision, validation, visualization, and writing—original draft, review, and editing.