Multiple motors cooperate to establish and maintain acentrosomal spindle bipolarity in C. elegans oocyte meiosis
Abstract
While centrosomes organize spindle poles during mitosis, oocyte meiosis can occur in their absence. Spindles in human oocytes frequently fail to maintain bipolarity and consequently undergo chromosome segregation errors, making it important to understand the mechanisms that promote acentrosomal spindle stability. To this end, we have optimized the auxin-inducible degron system in Caenorhabditis elegans to remove the factors from pre-formed oocyte spindles within minutes and assess the effects on spindle structure. This approach revealed that dynein is required to maintain the integrity of acentrosomal poles; removal of dynein from bipolar spindles caused pole splaying, and when coupled with a monopolar spindle induced by depletion of the kinesin-12 motor KLP-18, dynein depletion led to a complete dissolution of the monopole. Surprisingly, we went on to discover that following monopole disruption, individual chromosomes were able to reorganize local microtubules and re-establish a miniature bipolar spindle that mediated chromosome segregation. This revealed the existence of redundant microtubule sorting forces that are undetectable when KLP-18 and dynein are active. We found that the kinesin-5 family motor BMK-1 provides this force, uncovering the first evidence that kinesin-5 contributes to C. elegans meiotic spindle organization. Altogether, our studies have revealed how multiple motors are working synchronously to establish and maintain bipolarity in the absence of centrosomes.
Editor's evaluation
Oocytes in most species contain acentrosomal spindles and how these are formed and maintained is not entirely clear. Cavin-Meza et al. find that dynein is required for the integrity of bipolar spindle poles as well as spindle monopoles induced by depletion of the kinesin-12 motor KLP-18. They further show that in the absence of dynein motors, and functional spindle poles, microtubules are still able to organize into spindle-like structures via the action of chromosomes and the kinesin-5 motor BMK-1. They also found that chromosomes can still be segregated in an anaphase-like motion in these oocytes.
https://doi.org/10.7554/eLife.72872.sa0eLife digest
Meiosis is a specialized form of cell division that produces the gametes required for sexual reproduction, such as egg and sperm cells. Before the cell splits, it copies its genome so that it has four sets of chromosomes. Genetic information is then shuffled between the chromosomes, and the cell undergoes two rounds of division, resulting in four gametes that are genetically distinct.
Prior to division, the duplicated chromosomes are separated by rope-like protein polymers called microtubules. In most cells, structures called centrosomes organize these fibers into a spindle shape that emanates from two ‘poles’ on opposite ends of the cell: the microtubules then attach to the chromosomes and pull them apart. Despite not having centrosomes, egg cells, or ‘oocytes’, are still able to arrange their microtubules into a similar bipolar shape. However, how oocytes form these ‘acentrosomal’ spindles is poorly understood.
Centrosomes do not organize the spindle alone, and receive help from various motor proteins such as dynein. Previous work showed that dynein is involved in arranging acentrosomal poles, but it was not known if it was required to hold the poles together after they initially formed. To investigate, Cavin-Meza et al. developed a strategy that can rapidly remove dynein from oocytes of the roundworm Caenorhabditis elegans.
The experiment showed that dynein is required both to assemble and stabilize acentrosomal spindles in C. elegans. When dynein and an additional motor protein, KLP-18, were both removed from oocytes simultaneously, the poles blew apart, completely disrupting spindle organization. Surprisingly, Cavin-Meza et al. found that the spindles were able to reform and separate the chromosomes. Further probing revealed, for the first time, that a third motor protein (called BMK-1) also helps to organize the spindle into its bipolar structure.
These findings reveal the important role motor proteins play in stabilizing spindles and separating chromosomes in oocytes. Meiosis is prone to mistakes, and these errors are a major cause of miscarriages and birth defects in humans. Therefore, understanding the underlying mechanisms of how oocyte spindles form and remain stable could shed light on why chromosomes sometimes fail to segregate. This may eventually lead to new strategies for combating infertility.
Introduction
Centrosomes serve as prominent microtubule organizing centers (MTOCs) during mitosis and provide a clear structural blueprint for the formation of a bipolar spindle. However, oocytes of many species lack centrosomes. In human oocytes, it has been shown that acentrosomal spindles are sometimes highly unstable; poles go through an extended period where they can split apart and then come back together, and spindles that display this instability have a high incidence of chromosome segregation errors (Holubcová et al., 2015). The causes of this instability and the mechanisms by which acentrosomal spindles are stabilized in the absence of centrosomes remain poorly understood.
We use Caenorhabditis elegans as a model to identify factors required for acentrosomal pole stability. In C. elegans oocytes, microtubules are nucleated in proximity to the disassembling nuclear envelope in a spherical cage-like structure. Microtubule minus ends are then pushed outward to the periphery of the array by the kinesin-12 family motor KLP-18, where they form multiple poles with clearly defined clusters of minus ends. Spindle poles then coalesce, eventually forming a bipolar structure with aligned chromosomes (Wolff et al., 2016; Gigant et al., 2017). Shortly after establishing a metaphase plate, the spindle shortens and rotates perpendicular to the cortex and chromosomes begin to segregate in anaphase (Albertson and Thomson, 1993). The cortical set of homologous chromosomes are discarded as a polar body, and the remaining chromosomes undergo a second round of meiosis to generate the final set of maternal DNA.
While these stages of acentrosomal spindle assembly have been documented in multiple studies, the mechanisms by which microtubules organize into stable acentrosomal poles in this organism are not well understood. It would stand to reason that mitotic pole proteins are ideal candidates for also promoting meiotic pole stability; many proteins that serve important functions in C. elegans mitotic spindle assembly can be found within acentrosomal meiotic spindles (reviewed in Severson et al., 2016; Mullen et al., 2019). Various microtubule-associated proteins (MAPs) and microtubule motors have been demonstrated to concentrate at poles and contribute to acentrosomal spindle assembly, such as ASPM-1 (Connolly et al., 2014), KLP-7MCAK (Connolly et al., 2015; Han et al., 2015; Gigant et al., 2017), and MEI-1/2katanin (Srayko et al., 2000). A study using a fast-acting temperature-sensitive mei-1 mutant demonstrated that katanin is essential to maintain spindle structure (McNally et al., 2014), but whether any other pole-associated-proteins are required for spindle maintenance is not known.
Another factor that has been implicated in spindle pole organization is the minus-end-directed microtubule motor dynein. Studies of mitotically dividing cells in multiple organisms have shown that dynein is required for spindle pole focusing and for centrosome anchoring to poles (reviewed in Borgal and Wakefield, 2018) and dynein also promotes poleward flux of newly formed microtubule minus ends (Elting et al., 2014; Hueschen et al., 2017). Moreover, there is evidence that dynein is involved in the organization of acentrosomal poles. Inhibition of dynein in Xenopus egg extracts creates broad, splayed acentrosomal spindle poles (Heald et al., 1996), and injection of dynein antibodies into Xenopus oocytes disrupts overall meiotic spindle organization (Becker et al., 2003). In mouse oocytes, dynein is involved in the early stages of spindle assembly, so dynein inhibition prevents spindle (and thus pole) formation entirely (Luksza et al., 2013; Clift and Schuh, 2015).
In C. elegans, dynein is crucial for the proper positioning of both mitotic and meiotic spindles, and for spindle rotation during meiotic anaphase (Ellefson and McNally, 2009; van der Voet et al., 2009; Ellefson and McNally, 2011; Crowder et al., 2015). Moreover, a number of studies have shown that depletion of dynein or its cofactor dynactin causes spindle defects in oocytes, such as longer spindles with pole defects (Yang et al., 2005; Ellefson and McNally, 2009; Ellefson and McNally, 2011; Crowder et al., 2015; Muscat et al., 2015). However, whether dynein is required to maintain acentrosomal pole integrity after bipolarity has been established is not known. Thus, there is a need for an approach that can rapidly and efficiently remove dynein function from oocytes in a conditional manner.
Here, we have addressed this challenge using the auxin-inducible degron (AID) system (Zhang et al., 2015). We found that this method can efficiently remove dynein from oocytes within minutes, enabling us to remove this protein either prior to spindle assembly or after bipolarity has already been established. This approach revealed that dynein is essential for both the formation and stabilization of acentrosomal spindle poles. Removing dynein from monopolar spindles caused catastrophic breakdown of the monopole, further supporting a role for dynein in pole integrity. Moreover, this experiment also revealed a striking phenotype that led to additional insights into acentrosomal spindle function. Specifically, we found that following disruption of the monopole, individual chromosomes were able to reorganize local microtubules into a bipolar spindle that could mediate anaphase-like segregation. This phenotype provided an opportunity to identify proteins contributing to spindle bipolarity that are masked under normal conditions. Through this assay, we identified the kinesin-5 family motor BMK-1 as an outward sorting force on microtubules, providing the first evidence that kinesin-5 contributes to spindle assembly in C. elegans. Altogether, these studies have furthered our understanding of the motor forces that contribute to acentrosomal spindle assembly and maintenance during meiosis.
Results
Dynein is localized to acentrosomal spindles and is required for pole focusing
Since dynein is required for multiple cellular processes in C. elegans, full RNAi-mediated depletion of this protein causes developmental defects. Thus, previous studies of dynein in oocytes have relied on partial depletion (Yang et al., 2005; Ellefson and McNally, 2009; Ellefson and McNally, 2011; Muscat et al., 2015; McNally et al., 2016; Laband et al., 2017). In an attempt to achieve more complete depletion without affecting worm development, we sought to rapidly deplete dynein from adult worms using the AID system. For this, we utilized a strain in which the heavy chain of dynein, DHC-1, was tagged at the endogenous locus with both GFP and a degron tag; this strain also expressed the ubiquitin ligase TIR1 using the sun-1 promoter to enable auxin-mediated degradation of DHC-1 specifically in the germline (Zhang et al., 2015; Figure 1A; we refer to this strain as ‘Dynein AID’). This strain appears superficially wild type, with no noticeable effects on worm development, brood size, or percent of male progeny, suggesting that the GFP and degron tags are not affecting dynein function in the absence of auxin (Zhang et al., 2015). With this system, we are able to deplete dynein on a shorter timescale than RNAi, allowing us to assess the effects of dynein depletion on spindle assembly and maintenance without affecting prior processes. To validate that this strain worked for efficient DHC-1 depletion, we first performed ‘long-term’ auxin-mediated depletion, growing adult worms on auxin-containing plates for 4 hr (Figure 1A). We then imaged one-cell-stage mitotic embryos using immunofluorescence (IF) and looked for canonical dynein depletion phenotypes, such as failure to separate centrosomes and improper mitotic spindle positioning (Gönczy et al., 1999; Nguyen-Ngoc et al., 2007; Kiyomitsu and Cheeseman, 2013). After a 4 hr incubation of adult Dynein AID worms on auxin-containing plates, we observed clear defects in centrosome separation and spindle positioning along the A-P axis and dynein was undetectable (22/22 mitotic embryos) (Figure 1B), confirming strong depletion in embryos under conditions where germline and worm development are not affected.

Dynein is required for acentrosomal pole focusing.
(A) Schematic representation of the Dynein auxin-inducible degron (AID) system for DHC-1 depletion in the C. elegans germ line. Methodologies for all three auxin treatments used in this study are depicted. (B) Immunofluorescence (IF) imaging of one-cell mitotically dividing embryos shows that auxin treatment causes efficient dynein depletion and canonical mitotic spindle defects (n = 22). Shown are tubulin (green), DNA (blue), ASPM-1 (red), and dynein (not shown in merge). (C) IF imaging of oocyte spindles in the Dynein AID strain shows that dynein is localized to the spindle, with increasing enrichment at acentrosomal poles at the anaphase transition; shown are tubulin (green), DNA (blue), dynein (red), and ASPM-1 (not shown in merge). Cortex is represented by the dashed line. (D) Representative images of oocyte spindles (GFP::tubulin and GFP::histone) in germline counting and corresponding quantifications; auxin treatment leads to splayed poles and spindle rotation defects. Cortex is represented by the dashed line. (E) IF imaging of Dynein AID conditions showing effects of 4 hr auxin treatment on metaphase (second row) and anaphase (bottom row); shown are tubulin (green), DNA (blue), ASPM-1 (red), and dynein (not shown in merge). ASPM-1 labeling supports initial observations of splayed poles seen in germline counting. (F) Quantifications of IF imaging shown in (E); meiotic spindles have significantly splayed poles upon auxin treatment. All scale bars = 2.5 µm. Controls for exposure to auxin plates can be found in Figure 1—figure supplement 1. Further experimentation with dynein mislocalization through aspm-1(RNAi) or lin-5(RNAi) can be found in Figure 1—figure supplement 2.
To further validate the Dynein AID strain, we assessed the localization of tagged dynein throughout oocyte meiosis and found that the pattern was consistent with previous studies (Figure 1C; Ellefson and McNally, 2009; van der Voet et al., 2009; Ellefson and McNally, 2011; Crowder et al., 2015). Dynein was weakly associated with forming and bipolar spindles with a slight enrichment at acentrosomal poles, labeled by the microtubule minus end marker ASPM-1, and then became strongly enriched at poles during early anaphase; this enrichment decreased by late anaphase, with a population that was retained near the cortex. Live imaging corroborated our fixed imaging (Video 1, n = 3) and also made clearer a population of dynein at kinetochores, as has been shown in other live imaging studies (McNally et al., 2016; Danlasky et al., 2020).
Localization of dynein during oocyte meiosis.
Shows an oocyte expressing mCherry::tubulin, dissected into a control Meiosis Medium solution. DHC-1::degron::GFP localization can be seen at spindle poles during metaphase, alongside faint localization to kinetochores. Localization is consistent across all oocytes filmed (n = 3). Time elapsed shown in min:s. Scale bar = 5 μm.
Given this initial validation, we created a Dynein AID strain that also expressed GFP::tubulin and GFP::histone and then assessed the effects of dynein depletion in oocytes by quantifying spindle morphology in intact worms (Figure 1D, Table 1). This ‘germline counting’ method takes advantage of the fact that the C. elegans germline is arranged in a production-line fashion (Wolff et al., 2022). Oocytes are fertilized in the –1 position of the germline, and then oocyte spindles assemble as the newly fertilized embryo moves through the spermatheca and into the +1 position (Wolff et al., 2016). Under control conditions, bipolar oocyte spindles with focused poles are prevalent by the time the embryo reaches the +1 position (Figure 1D; Wolff et al., 2022). However, following 4 hr of auxin treatment, most spindles had splayed poles; a majority of spindles displayed microtubule bundles that did not come together at a single focused point on either side of the chromosomes (84/121, 69.4%), similar to what was observed following depletion of dynactin component DNC-1 in a previous study (Crowder et al., 2015). Imaging of ASPM-1 confirmed this phenotype at a percentage similar to that observed using germline counting (39/54, 72.2%); following a 4 hr auxin treatment, microtubule minus ends lacked coalesced points between the ends of microtubule bundles (Figure 1E and F, Table 2). Moreover, we also saw defects in spindle rotation in anaphase (Figure 1D and E), consistent with previous studies (Ellefson and McNally, 2009; van der Voet et al., 2009; Ellefson and McNally, 2011; Crowder et al., 2015). To confirm that these defects were not caused by nonspecific effects of auxin, we performed germline counting in a ‘TIR1 control’ strain that contained all of the same transgenes but lacked degron-tagged DHC-1. In this strain, there was no significant difference between control and auxin-treated worms at any position in the germline, and all stages of spindle assembly appeared identical (Figure 1—figure supplement 1, Table 3). These findings confirm the prediction that dynein is required for pole focusing during spindle assembly in oocytes and validate the use of the Dynein AID strain to study dynein in oocytes and embryos.
Exact numbers and percentages from germline counting experiment, to assess effects of dynein depletion on oocyte spindle assembly (corresponds to Figure 1D).
Location | Treatment | MT cage | Unfocused | Focused | Anaphase |
---|---|---|---|---|---|
–1 position | No auxin | 60/70 (85.7%) | 10/70 (14.3%) | N/A | N/A |
With auxin | 47/54 (87.0%) | 7/54 (13.0%) | N/A | N/A | |
Spermatheca | No auxin | N/A | 54/60 (90.0%) | 6/60 (10.0%) | N/A |
With auxin | N/A | 63/66 (95.5%) | 3/66 (4.5%) | N/A | |
+ 1 position | No auxin | N/A | 8/160 (5.0%) | 97/160 (60.6%) | 55/160 (34.4%) |
With auxin | N/A | 84/121 (69.4%) | 8/121 (6.6%) | 29/121 (24.0%) |
Exact numbers and percentages from quantification of IF images, to validate the phenotypes observed in germline counting (corresponds to Figure 1F).
Treatment | Focused poles | Unfocused poles | Anaphases |
---|---|---|---|
No Auxin | 35/60 (58.3%) | 3/60 (5.0%) | 22/60 (36.7%) |
With Auxin | 3/54 (5.6%) | 39/54 (72.2%) | 12/54 (22.2%) |
Exact numbers and percentages from germline counting experiment, to assess effects of auxin treatment on oocyte spindle assembly in a strain that does not contain degron-tagged dynein (corresponds to Figure 1—figure supplement 1).
Location | Treatment | MT cage | Unfocused | Focused | Anaphase |
---|---|---|---|---|---|
–1 position | No auxin | 135/146 (92.5%) | 11/146 (7.5%) | N/A | N/A |
With auxin | 137/152 (90.1%) | 15/152 (9.9%) | N/A | N/A | |
Spermatheca | No auxin | N/A | 204/222 (91.9%) | 18/222 (8.1%) | N/A |
With auxin | N/A | 226/245 (92.2%) | 19/245 (7.8%) | N/A | |
+ 1 position | No auxin | N/A | 22/452 (4.9%) | 284/452 (62.8%) | 146/452 (32.3%) |
With auxin | N/A | 26/435 (6.0%) | 271/435 (62.3%) | 138/435 (31.7%) |
Delocalization of dynein from spindle poles has similar effects as Dynein AID depletion
The fact that dynein normally localizes to acentrosomal poles and depletion caused pole splaying suggests that dynein’s localization is critical to its function in spindle pole formation. To test this hypothesis, we sought to delocalize dynein from poles by depleting two known dynein regulators, ASPM-1 and LIN-5 (NuMA). The MAP NuMA has been shown to direct cytoplasmic dynein to microtubule minus ends to promote pole focusing during mitosis (Elting et al., 2014; Hueschen et al., 2017), and the C. elegans homolog of NuMA, LIN-5, has been shown to target dynein to spindle poles in oocytes to promote spindle rotation (van der Voet et al., 2009). Additionally, ASPM-1 is required to properly target LIN-5 to acentrosomal poles, placing it upstream of both LIN-5 and DHC-1 (van der Voet et al., 2009). Therefore, we depleted either LIN-5 or ASPM-1 and utilized germline counting to compare oocyte spindle morphologies to Dynein AID (Figure 1—figure supplement 2A, Table 4). The frequency of pole splaying was nearly identical between these two RNAi conditions and long-term Dynein AID (167/261, 64%, and 364/521, 69.9%, for the +1 position in lin-5 and aspm-1 RNAi, respectively, compared to 84/121, 69.4% for Dynein AID); this quantification was in agreement with previous observations of spindle morphology following aspm-1(RNAi) or lin-5(RNAi) (van der Voet et al., 2009; Wignall and Villeneuve, 2009; Connolly et al., 2014; Laband et al., 2017).
Exact numbers and percentages from germline counting experiment, to compare spindle morphologies of lin-5/aspm-1(RNAi) to dynein AID experiments (corresponds to Figure 1—figure supplement 2A).
Location | Condition | Focused poles | Unfocused poles | Anaphases |
---|---|---|---|---|
Spermatheca | control (RNAi) | 7/103 (6.8%) | 96/103 (93.2%) | N/A |
lin-5 (RNAi) | 7/135 (5.2%) | 128/135 (94.8%) | N/A | |
aspm-1 (RNAi) | 9/202 (4.5%) | 193/202 (95.5%) | N/A | |
+ 1 position | control (RNAi) | 124/191 (64.9%) | 9/191 (4.7%) | 58/191 (30.4%) |
lin-5 (RNAi) | 12/261 (4.6%) | 167/261 (64.0%) | 82/261 (31.4%) | |
aspm-1(RNAi) | 18/521 (3.5%) | 364/521 (69.9%) | 139/521 (26.6%) |
To determine if the pole splaying observed following LIN-5 and ASPM-1 depletion was mostly a consequence of dynein mislocalization, we combined these RNAi conditions with Dynein AID depletion (Figure 1—figure supplement 2B). Treatment with auxin to remove dynein did not exacerbate the pole defects of either aspm-1(RNAi) or lin-5(RNAi), nor present any new phenotypes that were not present in the RNAi conditions alone, except for removing the kinetochore population of dynein (pole defects were consistent in 37/37 lin-5(RNAi) and 56/56 aspm-1(RNAi) oocyte spindles treated with auxin). These results confirm that dynein’s localization to poles is necessary for its role in pole coalescence during spindle assembly.
Dynein is required throughout meiosis to maintain focused poles
After corroborating that dynein was required to focus acentrosomal poles during spindle assembly, we next sought to determine whether dynein was required to maintain focused poles by performing more rapid depletion (‘short-term AID,’ Figure 1A). We therefore arrested oocytes in metaphase I to enrich for pre-formed spindles (by depleting anaphase-promoting complex component EMB-30; Furuta et al., 2000), soaked worms in auxin for 25–30 min to deplete dynein, and then performed IF. Under these conditions, nearly all spindles had severely splayed poles (Figure 2A and B); we measured the widths of the spindle equator and the poles and found an approximately 50% increase in the pole to equator ratio upon auxin treatment. Despite this pole splaying, microtubule bundles retained lateral associations with the chromosomes, resulting in a spindle that was still bipolar but more rectangular than wild-type spindles. Notably, we observed these same phenotypes in the absence of emb-30(RNAi), demonstrating that they are not dependent on the metaphase arrest (Figure 2A and B).

Dynein is required to maintain focused spindle poles.
(A) Immunofluorescence (IF) imaging of oocyte spindles in control or metaphase I-arrest (emb-30(RNAi)) conditions; shown are tubulin (green), DNA (blue), ASPM-1 (red), and dynein (not shown in merge). Acentrosomal poles become unfocused upon acute dynein depletion. Scale bars = 2.5 µm. (B) Quantification of the ratio between the width of the spindle equator and the width of the spindle poles shows that addition of auxin causes nearly all spindle poles to unfocus and splay, both with and without the metaphase arrest. Statistical significance between means of pole splaying was determined via a two-tailed t-test for unarrested and arrested conditions, with a p-value of <2.2 × 10–16 for both. Further quantification of changes in spindle length and microtubule minus end distribution upon auxin treatment can be seen in Figure 2—figure supplement 1A and B. (C) Ex utero live imaging of metaphase-arrested spindles; vehicle-treated control oocytes are shown on the left, auxin-treated oocytes are shown on the right. Addition of auxin causes rapid depletion of dynein (labeled with GFP, bottom row) and dynamic splaying of both spindle poles (shown using mCherry::tubulin, top row; arrowheads). Time elapsed shown in min:s. Scale bars = 5 µm. (D) Ex utero live imaging of metaphase-arrested spindles, shown using GFP::tubulin and GFP::histone; addition of auxin (bottom row) causes dynamic splaying of poles (arrowheads) but does not grossly disrupt chromosome alignment; vehicle-treated control oocytes are shown on the top row. Time elapsed shown in min:s. Scale bars = 5 µm. Controls for effects of auxin alone in ex utero imaging can be seen in Figure 2—figure supplement 2. Additional ex utero imaging of unarrested oocytes treated with auxin can be seen in Figure 2—figure supplement 3.
-
Figure 2—source data 1
The source data for Figure 2B is provided.
The equator and pole widths are listed for each spindle, as well as the calculated pole/equator ratio; data for unarrested (control(RNAi)) and metaphase I-arrested (emb-30(RNAi)) spindles are listed in separate tabs.
- https://cdn.elifesciences.org/articles/72872/elife-72872-fig2-data1-v3.xlsx
During this analysis, we also noticed that while ASPM-1 remained localized to dynein-depleted poles it appeared more dispersed throughout the spindle; this suggests a decrease in poleward transport of microtubule minus ends, as has been reported in dynein-depleted mitotic spindles (Hueschen et al., 2017). To quantify this phenotype, we performed linescans to measure ASPM-1 levels across the spindle. While ASPM-1 localizes to two major peaks that define the poles of wild-type spindles, it was more diffusely localized across the entire length of the spindle following dynein depletion (Figure 2—figure supplement 1). Moreover, we also measured the pole-to-pole lengths of these spindles and found that both unarrested and metaphase-arrested spindles were significantly longer (~3.5 μm) after short-term Dynein AID depletion (Figure 2—figure supplement 1). These data suggest that dynein not only promotes focusing of microtubule minus ends into discrete poles, but also provides an inward force that maintains spindle length.
To confirm these results and to assess the dynamics of pole splaying, we coupled rapid dynein depletion with live imaging by dissecting metaphase-arrested oocytes directly into auxin solution and then immediately filming (‘acute AID’; Figure 1A). We generated a Dynein AID strain expressing mCherry::tubulin so that we could visualize the spindle while dynein, tagged with GFP, was being depleted. In the absence of auxin, there was no noticeable change in dynein localization or brightness throughout the imaging time frame (Figure 2C, Video 2, n = 4). In contrast, when oocytes were dissected into auxin, dynein appeared depleted from meiotic spindles within roughly 3 min and acentrosomal poles began to splay (Figure 2C, Video 3, n = 4). Splaying occurred at both poles (Figure 2C, arrowheads) and spindle length increased over time. To visualize the movement and alignment of chromosomes during this process, we repeated this experiment in our GFP::tubulin and GFP::histone Dynein AID strain (Figure 2D, Video 4, n = 7). As before, metaphase-arrested spindles were quickly disrupted upon dissection into auxin solution (Video 5, n = 8); poles began to splay (Figure 2D, arrowheads) and spindles again lengthened. Despite these changes, microtubule bundles were able to remain associated with chromosomes and the spindles maintained bipolarity.
Dynein localizes to spindle poles in the Dynein AID strain in the absence of auxin.
Shows a metaphase-arrested (emb-30(RNAi)) oocyte expressing mCherry::tubulin (top), dissected into control Meiosis Medium solution. DHC-1::degron::GFP (bottom) is clearly localized to spindle poles and faintly seen at kinetochores and spindle morphology does not change during the imaging timeframe in the absence of auxin. Corresponds to Figure 2C. Results are consistent across all oocytes filmed (n = 4). Time elapsed shown in min:s. Scale bar = 5 μm.
Auxin treatment of Dynein AID worms rapidly depletes dynein and unfocuses poles.
Shows a metaphase-arrested (emb-30(RNAi)) oocyte expressing mCherry::tubulin (top), dissected into Meiosis Medium containing 100 μM auxin to deplete dynein. Once dissected into auxin solution, rapid depletion of DHC-1::degron::GFP (bottom) is evident within 3 min, at which point spindle poles begin to unfocus and splay apart. Corresponds to Figure 2C. Phenotypes are consistent across all oocytes filmed (n = 4). Time elapsed shown in min:s. Scale bar = 5 μm.
Metaphase-arrested spindles in the Dynein AID strain remain stable and maintain chromosome alignment in the absence of auxin.
Shows a metaphase-arrested oocyte expressing GFP::tubulin and GFP::histone in the Dynein AID strain (where DHC-1 is degron-tagged and TIR1 is expressed in the germ line), dissected into control Meiosis Medium solution. No major changes in spindle length or shape occur, and chromosomes are stably aligned in the spindle center. Corresponds to Figure 2D. Phenotypes are consistent across all oocytes filmed (n = 7). Time elapsed shown in min:s. Scale bar = 5 μm.
Auxin treatment unfocuses poles, but microtubules largely remain aligned along a single axis.
Shows a metaphase-arrested oocyte expressing GFP::tubulin and GFP::histone in the Dynein AID strain (where DHC-1 is degron-tagged and TIR1 is expressed in the germ line), dissected into Meiosis Medium containing 100 μM auxin to deplete dynein. Upon dissection into auxin solution, rapid and dynamic splaying of acentrosomal poles occurs alongside a notable increase in spindle length, but microtubule bundles and chromosomes stay associated and aligned along a single axis. Corresponds to Figure 2D. Phenotypes are consistent across all oocytes filmed (n = 8). Time elapsed shown in min:s. Scale bar = 5 μm.
To ensure that these phenotypes were not caused by auxin itself having nonspecific effects on dissected oocytes, we performed analogous live imaging experiments in our TIR1 control strain (lacking degron-tagged DHC-1) to assess the effects of auxin treatment in the absence of dynein depletion. Neither untreated nor auxin-treated oocytes exhibited significant changes to either spindle length or pole to equator ratio (n = 5 for each condition); we compared both individual traces and average change from baseline for both of these parameters (Figure 2—figure supplement 2, Videos 6 and 7). Additionally, we repeated these experiments without emb-30(RNAi) and saw the same pole defects in the absence of the metaphase arrest (Figure 2—figure supplement 3, Videos 8–11). These auxin-treated oocytes progressed through anaphase despite pole splaying, albeit with lagging chromosomes, as has been observed previously following dynein depletion (Muscat et al., 2015; McNally et al., 2016). Moreover, there were defects in spindle rotation (n = 9), consistent with prior studies (Ellefson and McNally, 2009; van der Voet et al., 2009; Ellefson and McNally, 2011; Crowder et al., 2015). Altogether, these data demonstrate that dynein is not only required to focus poles during meiotic spindle assembly, but is also required to maintain the integrity of acentrosomal poles once they have been established.
Metaphase-arrested spindles in the TIR1 control strain are stable and maintain chromosome alignment in the absence of auxin.
Shows a metaphase-arrested oocyte expressing GFP::tubulin and GFP::histone in the TIR1 control strain (that contains TIR1 but lacks a degron tag on DHC-1). Dissection into control Meiosis Medium yields no significant change in spindle length or organization. Notably, the spindle appears identical to those observed in control oocytes containing GFP-tagged DHC-1. Corresponds to Figure 2—figure supplement 2A. Phenotypes are consistent across all oocytes filmed (n = 5). Time elapsed shown in min:s. Scale bar = 5 μm.
Addition of auxin to metaphase-arrested oocytes in the TIR1 control strain does not perturb spindle architecture.
Shows a metaphase-arrested oocyte expressing GFP::tubulin and GFP::histone in the TIR1 control strain (that contains TIR1 but lacks a degron tag on DHC-1). Dissection into Meiosis Medium containing 100 μM auxin yields no significant change in spindle length or organization, demonstrating that auxin addition itself is not responsible for phenotypes observed in our acute Dynein auxin-inducible degron (AID) imaging. Corresponds to Figure 2—figure supplement 2A. Phenotypes are consistent across all oocytes filmed (n = 5). Time elapsed shown in min:s. Scale bar = 5 μm.
Normal oocytes undergo two rounds of bipolar spindle formation and anaphase segregation.
Shows an unarrested oocyte expressing GFP::tubulin and GFP::histone, dissected into control Meiosis Medium solution. After one round of segregation, a second bipolar spindle is formed, rotates perpendicular to the cortex, and successfully extrudes a second polar body. Corresponds to the movie in row 1 of Figure 2—figure supplement 3. Phenotypes are consistent across all oocytes filmed (n = 8). t = 0 is set to onset of meiosis II. Time elapsed shown in min:s. Scale bar = 5μm.
Normal oocytes undergo two rounds of bipolar spindle formation and anaphase segregation.
Another example of an unarrested oocyte expressing GFP::tubulin and GFP::histone, dissected into control Meiosis Medium solution. After one round of segregation, a second bipolar spindle is formed, rotates perpendicular to the cortex, and successfully extrudes a second polar body. Corresponds to the movie in row 2 of Figure 2—figure supplement 3. Phenotypes are consistent across all oocytes filmed (n = 8). t = 0 is set to onset of meiosis II. Time elapsed shown in (min):(sec). Scale bar = 5 μm.
Splaying of acentrosomal poles occurs in unarrested oocytes depleted of dynein.
Shows an unarrested oocyte expressing GFP::tubulin and GFP::histone, dissected into Meiosis Medium containing 100μM auxin to deplete dynein. A spindle tries to form in the absence of dynein, leading to dynamic and unfocused poles. Anaphase segregation still occurs, albeit with a failure in spindle rotation. Corresponds to the movie in row 3 of Figure 2—figure supplement 3. Phenotypes are consistent across all oocytes filmed (n = 9). t = 0 is set to onset of meiosis II. Time elapsed shown in min:s. Scale bar = 5μm.
Splaying of acentrosomal poles occurs in unarrested oocytes depleted of dynein.
Another example of an unarrested oocyte expressing GFP::tubulin and GFP::histone, dissected into Meiosis Medium containing 100 μM auxin to deplete dynein. A spindle tries to form in the absence of dynein, leading to dynamic and unfocused poles. Anaphase segregation still occurs, albeit with a failure in spindle rotation. Corresponds to the movie in row 4 of Figure 2—figure supplement 3. Phenotypes are consistent across all oocytes filmed (n = 9). t = 0 is set to onset of meiosis II. Time elapsed shown in (min):(sec). Scale bar = 5 μm
Removal of dynein from monopolar spindles causes catastrophic breakdown of spindle architecture
Despite the pole splaying observed following dynein depletion, spindle microtubules remained aligned along the same axis and bipolarity was not lost. Therefore, we hypothesized that microtubule bundles in the central region imparted stability to the entire spindle. To test this directly, we sought to deplete dynein from a monopolar spindle, generated by depleting the kinesin-12 motor KLP-18; these structures have a single focused pole and therefore no area of microtubule overlap that could theoretically stabilize the structure if the single pole is disrupted (Wignall and Villeneuve, 2009). Previously, we found that partial dynein RNAi on a monopolar spindle led to chromosomes moving a substantial distance away from the center of intact monopoles (Muscat et al., 2015); the AID approach now allows us to revisit these results without limiting the amount of dynein depletion to avoid developmental defects.
First, we performed long-term Dynein AID in worms treated with klp-18(RNAi) and found that oocytes from these worms rarely displayed a monopolar spindle (Figure 3A, Table 5). Instead, oocyte chromosomes were dispersed, retaining lateral associations to microtubule bundles (116/125 in the +1 position, 92.8%). IF imaging following short-term AID treatment confirmed this result; ASPM-1 labeling revealed that monopoles had begun breaking apart, ejecting chromosomes and associated microtubule bundles into the cytoplasm (Figure 3B). In some cases, the entirety of the monopole disappeared, and each chromosome retained its own associated microtubule bundles. Many of these structures had a distinct cone shape, with microtubule ends focused on one side of the chromosome and splayed on the other side (Figure 3B, zooms); we speculate that the microtubule minus ends are coalesced together at the pointy end of this structure, retaining some association even after they were released from the monopole upon spindle breakdown.

Monopolar spindles break down upon dynein depletion.
(A) Representative images of klp-18(RNAi) oocyte spindles (GFP::tubulin and GFP::histone) in germline counting and corresponding quantifications; dynein depletion leads to dissolution of monopoles, releasing chromosomes with associated microtubule bundles into the cytoplasm. Scale bars = 2.5 μm (B) Immunofluorescence (IF) imaging of monopole breakdown after dynein depletion; shown are tubulin (green), DNA (blue), ASPM-1 (red), and dynein (not shown in merge). Chromosomes released into the cytoplasm retain lateral microtubule associations (zooms). Scale bars = 2.5 μm; note that the +Auxin images are zoomed out more since chromosomes are more dispersed following monopole breakdown and a larger region needed to be imaged.
Exact numbers and percentages from germline counting experiment, to assess effects of dynein depletion on monopolar spindles (corresponds to Figure 3A).
Condition | Treatment | Monopolar | Breakdown |
---|---|---|---|
klp-18 (RNAi) | No auxin | 119/119 (100.0%) | 0/119 (0.0%) |
With auxin | 9/125 (7.2%) | 116/125 (92.8%) |
We observed the same phenotype using acute dynein depletion coupled with live ex utero imaging; when oocytes containing monopolar spindles were dissected into auxin, we observed rapid breakdown of the monopole on a timescale similar to that required for dynein depletion from bipolar spindles (Figure 4A and B, Video 12 and Video 13, n = 5). As the monopolar spindle began to fragment, microtubule bundles remained associated with individual chromosomes via lateral associations (Figure 4A, arrowheads), as seen in fixed imaging (Figure 3B). These results demonstrate an essential role for dynein in maintaining pole integrity and suggest that in the case of dynein depletion from bipolar spindles the intact overlap zone stabilizes the spindle even though poles are completely disrupted.

Individualized chromosomes undergo anaphase-like segregation in the absence of KLP-18 and dynein.
(A) Ex utero live imaging of oocytes expressing GFP::tubulin and GFP::histone in control and klp-18(RNAi) conditions. In the control (top row), a monopolar spindle forms and then chromosomes move back towards the monopole in anaphase, as previously described (Muscat et al., 2015). Addition of auxin to acutely deplete dynein (row 2) leads to rapid breakdown of monopolar spindles. Intriguingly, individualized chromosomes undergo an anaphase-like segregation post-breakdown (zooms, row 3). Microtubule bundles remain laterally associated with chromosomes (arrowheads) prior to anaphase-like segregation. Time elapsed shown in min:s. Scale bars = 5 µm. (B) Another example of acute auxin treatment, to remove dynein, from an oocyte expressing GFP::tubulin and GFP::histone in klp-18(RNAi) conditions; after breakdown of the monopole and reorganization of microtubules (arrowheads), individual chromosomes are able to undergo synchronized anaphase-like segregation. Time elapsed shown in min:s. Scale bars = 5 µm. (C) Quantification of timelapses from control anaphases (the Dynein auxin-inducible degron [AID] strain without auxin), anaphases following dynein depletion (the Dynein AID strain with auxin), and miniature anaphases. Each timelapse was synchronized to initiation of anaphase A, and the distance between the center of each chromosome was measured at each 15 s timepoint. For each condition, the number of timelapses used is represented by n, and the shaded area around each average line represents the SEM. Note that for the mini anaphase condition, though we used five movies, we measured multiple mini anaphases in each (total anaphases = 35). Miniature anaphases do not exhibit significant differences in segregation rates or distances to wild-type anaphase spindles, but do not reach distances of dynein depletions alone.
-
Figure 4—source data 1
The source data for Figure 4C is provided.
The chromosome segregation distances (in μm) for each timepoint are shown for control spindles (control(RNAi) without auxin), dynein-depleted spindles (control(RNAi) with auxin), and mini spindles (klp-18(RNAi) with auxin). The compiled data is in the first tab, and raw data for each condition is listed in separate tabs.
- https://cdn.elifesciences.org/articles/72872/elife-72872-fig4-data1-v3.xlsx
Monopolar spindles break down and individual chromosomes segregate following dynein depletion.
Shown is an oocyte from the Dynein AID strain expressing GFP::tubulin and GFP::histone following klp-18(RNAi) to generate monopolar spindles, dissected into Meiosis Medium containing 100 μM auxin to deplete dynein. Upon dissection into auxin solution, the monopole quickly dissolves, leading to ejection of individual chromosomes (with laterally associated microtubule bundles) into the cytoplasm, followed by a synchronous anaphase-like segregation event. Corresponds to Figure 4A. Phenotypes are consistent across all oocytes filmed (n = 5). Time elapsed shown in min:s. Scale bar = 5 μm.
Monopolar spindles break down and individual chromosomes segregate following dynein depletion.
Another example of a Dynein AID oocyte expressing GFP::tubulin and GFP::histone following klp-18(RNAi), dissected into Meiosis Medium containing 100 μM auxin to deplete dynein. In this example, reorganization of local microtubule bundles around individual chromosomes can be observed after breakdown of the monopolar spindle. Shortly after reorganization of these microtubules, anaphase-like segregation occurs synchronously across all chromosomes. Corresponds to Figure 4B. Phenotypes are consistent across all oocytes filmed (n = 5). Time elapsed shown in min:s. Scale bar = 5 μm.
Individualized chromosomes can undergo an anaphase-like segregation in the absence of KLP-18 and dynein
In addition to demonstrating the dynamics of monopole breakdown, our live imaging of acute dynein depletion from monopolar spindles revealed an intriguing phenotype. These experiments were not done using metaphase arrest, so in the absence of auxin, chromosomes extended outward from the monopole towards microtubule plus ends, and then moved back towards the monopole during anaphase (Figure 4A, Video 14, n = 6), as has been described previously (Muscat et al., 2015). In auxin-treated oocytes, as filming continued past the full breakdown of the monopole, we were surprised to see that each individual chromosome began to segregate (Figure 4A and B, Videos 12 and 13, n = 5). This segregation was incredibly striking as it occurred synchronously across the oocyte and each set of separating chromosomes exhibited distinct bidirectional movement, reminiscent of normal chromosome segregation. The fact that we observed these anaphase-like movements in the absence of both KLP-18 and dynein suggests that neither motor is absolutely required for chromosome segregation in oocytes.
Dynamics of a normal monopolar spindle.
Shows an oocyte expressing GFP::tubulin and GFP::histone following klp-18(RNAi), dissected into a Meiosis Medium control solution. Chromosomes extend outward towards microtubule plus ends forming a monopolar spindle during metaphase, but retract towards the minus ends during anaphase. Corresponds to Figure 4A. Phenotypes are consistent across all oocytes filmed (n = 6). Time elapsed shown in min:s. Scale bar = 5 μm.
To assess whether these segregations were similar to normal anaphases, we first quantified the segregation distances of individual chromosomes and compared them to both wild-type and dynein-depleted anaphases (Figure 4C). Chromosomes on dynein-depleted klp-18(RNAi) mini anaphase spindles segregated to a similar final distance as wild-type spindles (~5 µm for both wild-type and mini anaphases). However, mini anaphases did not appear to segregate as far as dynein-depleted bipolar anaphases, where KLP-18 was not depleted (~6.5 µm). This observation raises the possibility that, even though chromosomes can segregate without KLP-18, this motor may normally contribute to anaphase spindle elongation through microtubule sliding. This contribution has been difficult to test in previous studies due to the monopolar spindle phenotype that is quickly generated upon removal of KLP-18 (Wolff et al., 2021), which prevents normal anaphase segregation from occurring.
To further characterize these segregations, we performed fixed imaging to generate higher-resolution images. This analysis revealed that the mini anaphases have key hallmarks of normal anaphase spindles; when chromosomes were still close together, we observed lateral microtubule bundles running alongside the separating chromosomes (Figure 5, ‘early anaphase’), and as chromosomes segregated further and spindle length increased, microtubules were largely localized between chromosomes (Figure 5, ‘late anaphase’). This suggests that, like normal anaphase spindles, these mini spindles undergo morphological changes and anaphase-B-like spindle elongation as they drive chromosomes apart.

Miniature anaphase spindles recapitulate key features of normal anaphase progression.
Immunofluorescence (IF) imaging of miniature anaphases at multiple chromosome segregation distances, representing various stages of anaphase; shown are tubulin (green), DNA (blue), ASPM-1 (red), and dynein (not shown in merge). For each stage, a zoomed-out view is shown (top row), as well as a zoom of particular mini anaphase spindles (bottom row). Segregation events closely resemble different stages of anaphase A and anaphase B. Scale bars = 2.5 µm. Further characterization of miniature anaphases via localization of SUMO and separase can be found in Figure 5—figure supplement 1 and Figure 5—figure supplement 2, respectively.
Additionally, we used fixed imaging to assess the localization of other proteins that are present during normal anaphase. First, we imaged the ring complex (RC), a collection of proteins that forms a ring around the center of each chromosome in oocytes. RCs aid in chromosome congression (Wignall and Villeneuve, 2009; Muscat et al., 2015; Pelisch et al., 2017; Hollis et al., 2020) and then are released from chromosomes during anaphase; they remain as a ring between segregating chromosomes and then are disassembled as anaphase progresses (Dumont et al., 2010; Muscat et al., 2015; Davis-Roca et al., 2017; Pelisch et al., 2017; Davis-Roca et al., 2018; Pelisch et al., 2019). To assess RC behavior, we imaged SUMO, a post-translational modification found in the RC (Pelisch et al., 2017). In control and dynein depletion conditions, SUMO-marked RCs were removed from chromosomes during anaphase and remained between segregating chromosomes, as expected (Figure 5—figure supplement 1). During monopolar spindle breakdown, SUMO localized to RCs as individual chromosomes detached from the dissolving monopole. In mini anaphases, SUMO marked a singular disassembling RC between each segregating chromosome pair (93/93 oocytes observed), congruent with the normal behavior of RCs in anaphase.
Next, we assessed SEP-1 (separase), a protease that cleaves a component of the cohesin complex, enabling homologous chromosomes (meiosis I) and sister chromatids (meiosis II) to separate (Siomos et al., 2001; Kudo et al., 2006). During wild-type meiosis, SEP-1 localizes to kinetochores in metaphase and relocalizes to RCs at anaphase onset; by the end of anaphase, SEP-1 is no longer detectable on the anaphase spindle (Muscat et al., 2015; Davis-Roca et al., 2017). We observed this same pattern when assessing monopolar breakdowns following klp-18(RNAi) and short-term Dynein AID. After monopole breakdown, SEP-1 relocalized from kinetochores to RCs, and then was absent from each mini spindle in late anaphase (81/81 oocytes observed) (Figure 5—figure supplement 2). Taken together, these data demonstrate that miniature spindles recapitulate the key aspects of normal anaphase progression.
Microtubules can reorganize into mini bipolar spindles in the absence of dynein and KLP-18
The fact that we observed chromosome segregation with hallmarks of normal anaphase led us to hypothesize that some local microtubule reorganization must be occurring after monopole breakdown; without a bipolar distribution of microtubule minus ends, it should not be possible for chromosomes to move away from each other in the manner we observed. Therefore, while all microtubule minus ends are together in the monopole before dynein depletion, we speculated that these microtubules may reorganize into a bipolar structure around each chromosome following monopole breakdown. In order to better characterize this possible reorganization, we assessed the localization of ASPM-1 as a marker of microtubule minus ends. On a few occasions (n = 3 oocytes), we were able to capture what appeared to be a transitional period, indicating that reorganization was occurring; ASPM-1 could be seen on both sides of these mini spindles, with minor enrichment towards areas resembling spindle poles (Figure 6A). Note that while ASPM-1 also appeared to localize on chromosomes in these experiments, this has been previously shown to be nonspecific cross-reactivity of this particular antibody (Wignall and Villeneuve, 2009).

Microtubules reorganize into miniature bipolar spindles that can segregate chromosomes.
(A) Immunofluorescence (IF) imaging of microtubules (green), DNA (blue), ASPM-1 (red), and dynein (not shown in merge) in monopolar spindle breakdown conditions (klp-18(RNAi) + short-term Dynein auxin-inducible degron [AID]). Microtubule bundles appear to reorganize around individual chromosomes, seen through ASPM-1 flanking either side of the chromosome; note that in these images ASPM-1 also appears to be on chromosomes, but that is background staining that sometimes occurs with this antibody (Wignall and Villeneuve, 2009) and is not real signal. (B) IF imaging of SPD-1 localization in the Dynein AID strain in control RNAi conditions (rows 1–4) or following klp-18(RNAi) (rows 5–7); shown are tubulin (green), DNA (blue), SPD-1 (red), and dynein (not shown in merge). SPD-1 does not localize to spindles in metaphase (rows 1, 3; 18/18 metaphases), but localizes to overlapping microtubules in anaphase spindles in the presence or absence of dynein (rows 2, 4; 17/17 anaphases). SPD-1 is not localized to monopolar spindles either before or after monopole breakdown (rows 5, 6; 60/60 monopoles and breakdowns), but can clearly be seen localized to miniature anaphases (row 7; 27/27 mini anaphases). Cortex is represented by the dashed line. All scale bars = 2.5 µm.
This imaging suggested that microtubules might reorganize into a mini bipolar spindle with overlapping antiparallel microtubules in the center, which could facilitate chromosome segregation. To test this idea, we assessed the localization of a well-studied spindle protein, the PRC1 homolog SPD-1. SPD-1 localizes to the anaphase spindle midzone in oocytes, marking a region of antiparallel microtubule overlap (Hattersley et al., 2016; Gigant et al., 2017; Mullen et al., 2017; Figure 6B, row 2), and we found that SPD-1 was also at this location in the absence of dynein (Figure 6B, row 4). In intact and disassembling monopoles, SPD-1 was not localized to microtubules as it only associates with the spindle in anaphase. However, once chromosomes had been ejected into the cytoplasm, SPD-1 could be clearly observed between segregating chromosomes (Figure 6B, bottom row, 122/122 oocytes observed), demonstrating that the mini anaphase spindles have a region of antiparallel microtubule overlap in the center.
The kinesin-5 family motor BMK-1 provides an outward sorting force that allows spindle reorganization in the absence of KLP-18 and dynein
Our discovery that microtubules were able to reorganize around dispersed chromosomes and establish a region of antiparallel overlap at the center of each mini spindle suggested the presence of an activity that can sort microtubules and re-establish bipolarity in the absence of KLP-18 and dynein. We saw this condition as a unique opportunity to probe for redundant motor forces that may be present in a normal meiotic spindle, but would otherwise be masked by the presence of the forces provided by KLP-18 and dynein. One strong candidate was the sole kinesin-5 family motor in C. elegans, BMK-1 (Bishop et al., 2005). Kinesin-5 motors have essential roles in establishing spindle bipolarity in many organisms (reviewed in Mann and Wadsworth, 2019), but loss of BMK-1 has not been reported to have major phenotypes in C. elegans; BMK-1 depletion has no effects on spindle morphology in either oocytes or embryos (Bishop et al., 2005) and only minor effects on chromosome segregation rates (Saunders et al., 2007; Laband et al., 2017). We therefore hypothesized that BMK-1 could be providing a supplementary outward sorting force that is normally masked by the contributions of KLP-18.
To probe this hypothesis, we first sought to confirm the localization of BMK-1 on meiotic spindles and to test if this motor was localized to microtubules under monopole breakdown conditions. BMK-1 was broadly associated with bipolar metaphase and anaphase spindles in both control and dynein-depletion conditions (Figure 7A). In intact monopolar spindles, BMK-1 also localized to microtubules. Importantly, when we depleted dynein and induced monopole breakdown, BMK-1 still localized to microtubules (Figure 7A, zooms), placing it in a location where it could contribute to microtubule reorganization (150/150 oocytes observed).

BMK-1 localizes to the meiotic spindle and is required for microtubule reorganization and the formation of miniature anaphases.
(A) Immunofluorescence (IF) imaging of oocytes in either control or klp-18(RNAi) conditions in the Dynein auxin-inducible degron (AID) strain in the presence and absence of short-term auxin treatment; shown are tubulin (green), DNA (blue), BMK-1 (red), and dynein (not shown in merge). BMK-1 is localized to spindle microtubules in all conditions (52/52 metaphases, 43/43 anaphases), including following monopolar spindle breakdown (40/40 monopoles and breakdowns) and in mini anaphases (15/15 mini anaphases). Cortex is represented by the dashed line. Scale bars = 2.5 µm. (B) IF imaging of embryos following klp-18(RNAi) in the Dynein AID strain lacking functional BMK-1 (bmk-1(ok391)); shown are tubulin (green), DNA (blue), ASPM-1 (red), and dynein (not shown in merge). Following monopolar spindle breakdown in the presence of auxin (rows 2, 3), embryos do not contain miniature anaphases and lack chromosome segregation. Scale bars = 2.5 µm. (C) Quantifications of images shown in (B) compared to wild-type (WT) embryos; monopolar spindles still break down, but no miniature anaphases are observed in embryos lacking BMK-1 function. (D) Ex utero live imaging of GFP::tubulin and GFP::histone following acute auxin treatment to remove dynein in klp-18(RNAi); bmk-1(syb3914) conditions; miniature anaphases do not form in the absence of BMK-1. Time elapsed shown in min:s. Scale bars = 5 µm. Validation of bmk-1 mutants via IF imaging can be seen in Figure 7—figure supplement 1.
To test if BMK-1 was necessary for the formation of miniature anaphases, we utilized two bmk-1 mutants: (1) a previously characterized allele, bmk-1(ok391), that introduces a premature stop codon in the motor domain (Bishop et al., 2005), and (2) a new deletion of the entire bmk-1 locus generated using CRISPR-Cas9 (bmk-1(syb3914)). To validate these deletions, we utilized IF imaging with an α-BMK-1 antibody and confirmed that BMK-1 was no longer present on the meiotic spindle in either mutant (no BMK-1 staining in 71/71 oocytes observed) (Figure 7—figure supplement 1). We then generated monopolar spindles, performed short-term Dynein AID, and performed IF in the bmk-1(ok391) background. Under these conditions, we were unable to observe any mini anaphase spindles (0/164 oocytes), even though monopole breakdown still occurred (Figure 7B and C, Table 6). This suggested that removal of BMK-1 function abolished microtubule reorganization and prevented chromosome segregation. To confirm this result, we generated a version of our Dynein AID live imaging strain (expressing GFP::tubulin and GFP::histone to visualize the spindle) containing the bmk-1(syb3914) deletion and performed acute dynein depletion coupled with ex utero imaging to determine if mini anaphases could still form. In klp-18(RNAi), bmk-1(syb3914) worms in the absence of auxin, monopolar spindles formed and then chromosomes moved back towards the pole in a manner indistinguishable from normal monopolar anaphase (Video 15, n = 4). When auxin was added to deplete dynein, monopolar spindles broke down as expected and individual chromosomes remained associated with microtubule bundles. However, as time elapsed, there were no signs of segregation, microtubule density decreased, and chromosomes remained in the cytoplasm with no discernible anaphase spindles forming (Figure 7D, Video 16, n = 3). These data support the hypothesis that BMK-1 provides outward sorting force on microtubules during oocyte meiosis, redundant to the forces produced by KLP-18.
Exact numbers and percentages from quantification of IF images, to determine if the removal of functional BMK-1 prevents the formation of miniature anaphases in KLP-18/dynein-depleted oocytes (corresponds to Figure 7C).
Condition | Treatment | Breakdown | Mini anaphases | No segregation |
---|---|---|---|---|
klp-18 (RNAi) | bmk-1 (WT) | 82/130 (63.1%) | 43/130 (33.1%) | 5/130 (3.8%) |
bmk-1 (ok391) | 111/164 (67.7%) | 0/164 (0.0%) | 53/164 (32.3%) |
Monopolar spindle dynamics are normal in the absence of BMK-1.
Shows a Dynein AID; bmk-1(syb3914) oocyte expressing GFP::tubulin and GFP::histone following klp-18(RNAi) to generate monopolar spindles, dissected into control Meiosis Medium solution. Without functional BMK-1 or KLP-18, monopolar spindles can form and monopolar anaphase still occurs, with chromosomes moving outwards in metaphase and back towards the monopole in anaphase. Phenotypes are consistent across all oocytes filmed (n = 4). Time elapsed shown in min:s. Scale bar = 5 μm.
Loss of BMK-1 function prevents the formation of miniature anaphases.
Shows a Dynein AID; bmk-1(syb3914) oocyte expressing GFP::tubulin and GFP::histone following klp-18(RNAi) dissected into Meiosis Medium containing 100 μM auxin to deplete dynein. Without functional BMK-1, formation of miniature bipolar anaphases does not occur after the breakdown of the monopolar spindle. Corresponds to Figure 6D. Phenotypes are consistent across all oocytes filmed (n = 3). Time elapsed shown in min:s. Scale bar = 5 μm.
Discussion
Contributions of dynein to oocyte meiosis
Collectively, these data have contributed to a more complete model for how acentrosomal spindles are assembled and stabilized during oocyte meiosis (Figure 8). Dynein activity is required throughout the meiotic divisions to establish and maintain focused poles, and its removal via acute AID generated spindle defects within minutes. Dynein depletion increased the length of bipolar oocyte spindles and led to increased dispersion of microtubule minus ends across the spindle. When performing these depletions on a monopolar spindle, the monopole completely broke apart and microtubule bundles were ejected into the cytoplasm. Together, these findings demonstrate that dynein is required to stitch minus ends together into a stable pole structure.

Dynein, KLP-18, and BMK-1 work in concert to establish and maintain spindle bipolarity in C. elegans oocyte meiosis.
Chromosomes (blue), microtubules (green), and microtubule minus ends (red). Dynein is required throughout the meiotic divisions to maintain focused acentrosomal poles. If removed from stable bipolar spindles (top) using short-term depletions, poles rapidly unfocus and splay, leading to the same phenotype as long-term depletions. Dynein depletion from monopolar spindles (bottom) ejects individual chromosomes and associated microtubule bundles into the cytoplasm. BMK-1 is able to provide an outward sorting force in the absence of KLP-18 and dynein, enabling reorganization of microtubules into a miniature anaphase spindle, promoting chromosome segregation. In the absence of BMK-1, these miniature anaphases cannot form and anaphase-like segregations no longer occur.
Remarkably, we also found that after the monopole disassembled, microtubules were able to reorganize into a ‘miniature’ spindle; chromosomes then segregated on these spindles in the absence of dynein. This finding seemingly contradicts one of our previous studies, where we proposed that dynein could drive chromosomes towards poles along laterally associated microtubule bundles, representing a type of ‘anaphase A’ segregation mechanism (Muscat et al., 2015). Dynein was a good candidate for such a role since we found that it provided chromosomes with a minus-end-directed force prior to anaphase, and we observed lagging chromosomes in anaphase following dynein inhibition (Muscat et al., 2015); lagging chromosomes upon dynein depletion have also been reported by others (McNally et al., 2016) and were observed in the current study (Figure 2—figure supplement 3). However, since chromosome movement was not completely blocked following dynein inhibition, when we proposed our original model we speculated that other mechanisms, such as spindle elongation, might also contribute to segregation (Muscat et al., 2015). Indeed, in subsequent years it has become clear that this anaphase-B-like mechanism is the major driver of chromosome segregation in C. elegans oocytes, and that anaphase A chromosome-to-pole movement provides only a minor contribution (McNally et al., 2016; Laband et al., 2017). Given these follow-up studies, it is now clear that if dynein plays a role in chromosome segregation, its contribution is redundant with other, more dominant, mechanisms; this would explain why most chromosomes were able to move poleward following dynein inhibition in our original study and in subsequent studies by other labs (Muscat et al., 2015; McNally et al., 2016; Laband et al., 2017; Danlasky et al., 2020). This updated view is also consistent with the results of our current study; in our Dynein AID depletions, unarrested bipolar spindles progress through anaphase and mini spindles formed following klp-18(RNAi) also undergo anaphase-like segregation, reaching comparable segregation distances to that of normal anaphase spindles. Interestingly, our studies also revealed that anaphase-B-like spindle elongation can occur in the absence of KLP-18. Although it is possible that KLP-18 may contribute to outward sliding of microtubules during wild-type anaphase, our work shows that this motor is not absolutely required for anaphase spindle elongation and demonstrates that there must be other factors that can perform this function.
Multiple motors cooperate in C. elegans oocyte meiosis to effectively form and stabilize an acentrosomal bipolar spindle
Our work also accentuates the importance of balanced forces within a bipolar meiotic spindle; we found that loss of dynein activity had strong phenotypes that manifested rapidly, within 2 min of auxin treatment. Notably, acute dynein depletion caused spindle lengthening, suggesting that dynein may normally provide an inward force on the spindle, as has been demonstrated in previous studies of mitosis (Ferenz et al., 2009; van Heesbeen et al., 2014). Recently, our lab also demonstrated that inactivation of KLP-18 in stable, bipolar spindles caused spindle shortening followed by collapse of microtubule minus ends into a monopolar spindle (Wolff et al., 2021). This collapse is due to a loss of the outward force provided by KLP-18, theoretically enabling the inward force provided by dynein to dominate and highlighting how quickly an imbalance of these motor forces can lead to gross defects in spindle organization. In the current study, removing dynein from oocytes that already lacked KLP-18 resulted in a catastrophic breakdown of the monopolar spindle, releasing individual chromosomes into the cytoplasm. Going forward, it would be interesting to acutely remove these proteins at the same time to see whether oocyte spindles are able to maintain bipolarity and directly test whether these motors antagonize each other in C. elegans.
While KLP-18/kinesin-12 provides the major outward force in C. elegans oocyte meiosis (Wolff et al., 2016), this critical function is performed by kinesin-5 in mouse oocytes (Schuh and Ellenberg, 2007) and in mitosis in many organisms (reviewed in Mann and Wadsworth, 2019). Previous studies have shown that dynein and kinesin-5 antagonize each other in these organisms; inhibition of both motors simultaneously enables bipolar spindle formation (Mitchison et al., 2005; Tanenbaum et al., 2008; Ferenz et al., 2009; van Heesbeen et al., 2014). Interestingly, it has been shown that Kif15/kinesin-12 is capable of providing a supplemental outward force that can support bipolarity when kinesin-5 is inhibited (Tanenbaum et al., 2009; Vanneste et al., 2009; Raaijmakers et al., 2012; Sturgill and Ohi, 2013; Sturgill et al., 2016). Here, we have provided evidence that these roles have been reversed in C. elegans oocyte meiosis; BMK-1/kinesin-5 appears to be providing a redundant outward sorting force that was only detectable once we had removed KLP-18 and dynein from the meiotic spindle. Why the roles of kinesin-5 and kinesin-12 have been seemingly switched between different organisms is a fascinating open question. In the future, it would be interesting to perform KLP-18/DHC-1 double depletions in a strain lacking BMK-1 function to further probe the relationship between these three motors during acentrosomal spindle formation in C. elegans.
While this work has expanded our understanding of the role BMK-1 plays in C. elegans oocyte meiosis, further experimentation will be valuable for understanding the exact mechanism of BMK-1 function in the context of a normal, bipolar meiotic spindle. Biophysical assays to determine motor walking speed and force generation would help frame how much BMK-1 actually contributes in comparison to other meiotic motors such as KLP-18. Also, since kinesin-5 activity is known to be regulated in other systems by kinases and protein-protein interactions (reviewed in Mann and Wadsworth, 2019), it would be beneficial to determine if BMK-1 has any interacting partners that provide some regulation of function. Aurora B kinase (AIR-2) has been shown to be required for BMK-1 spindle localization and AIR-2 can phosphorylate BMK-1 in vitro (Bishop et al., 2005) implicating AIR-2 in BMK-1 regulation. Inhibition of AIR-2’s kinase activity also leads to collapsed oocyte spindles (Divekar et al., 2021a), consistent with a role for AIR-2 in regulating spindle force generation. However, it is likely that other factors also regulate BMK-1. Finally, another interesting question is whether BMK-1 acts to reorganize the existing microtubule bundles that remain chromosome-associated following monopole breakdown or whether chromatin-mediated microtubule nucleation also plays a role in miniature spindle formation, as it does in many systems during acentrosomal spindle assembly (Meunier and Vernos, 2016; Radford et al., 2017).
Different acentrosomal pole proteins have distinct roles in pole coalescence and stability
Our analysis of ASPM-1 and LIN-5 provides further evidence that these proteins are directly influencing acentrosomal pole organization and stability by directing cytoplasmic dynein localization. In either aspm-1 or lin-5 RNAi, spindle phenotypes were nearly identical to those seen in Dynein AID experimentation, and co-depletion of dynein did not worsen those phenotypes. The shared phenotype across depletions of these proteins suggests that these pole proteins function in a single pathway; this provides some credence to the concept that acentrosomal poles could contain a cross-linked network of interacting pole proteins that provide cohesion and stability. However, recent work from our lab has made it clear that not all pole protein depletions yield the same defects in spindle bipolarity. Acute depletion of ZYG-9, a homolog of the microtubule polymerase XMAP215, disrupts spindle integrity in a manner that does not resemble Dynein AID depletions. While dynein depletion splays spindle poles, microtubule bundles appear to remain stable and unperturbed, generating a rectangular spindle that can still progress through meiosis (albeit with spindle rotation defects). However, acute ZYG-9 depletion leads to a unstable multipolar spindle with defects in the middle of the spindle as well as at the poles, suggesting a broader role of ZYG-9 (Mullen et al., 2022).
From these recent studies, it is clear that multiple groups of proteins are required to establish and maintain acentrosomal poles. However, despite having similar localizations in the meiotic spindle, pole proteins have separable functions that impact pole stability in unique ways. Rather than all pole proteins interacting as a cohesive, cross-linked network around microtubule minus ends, it seems more plausible that different groups of proteins work in parallel to form and stabilize acentrosomal poles. These degron-based depletions with essential pole proteins have allowed for further mechanistic understanding of how each pole protein is contributing to spindle stability. There remain numerous pole proteins that could be subjected to rapid depletion via the AID system; the ability to probe the roles of essential proteins in both spindle assembly and maintenance is vital to building a more robust model of how meiotic spindles are able to achieve bipolarity in the absence of centrosomes to faithfully congress and segregate chromosomes.
Materials and methods
C. elegans strains used
Request a detailed protocolName | Description | Genotype |
---|---|---|
CA1215 | dhc-1::degron::GFP; Psun-1::TIR1::mRuby | dhc-1(ie28[dhc-1::degron::GFP]) I; ieSi38 IV |
PHX3914 | dhc-1::degron::GFP; Psun-1::TIR1::mRuby; GFP::H2B; GFP::tubulin; Δbmk-1 | dhc-1(ie28[dhc-1::degron::GFP]) I; unc-119(ed3) III; ruIs32[unc-119(+) pie-1::gfp::h2b] III; ruIs57[unc-119(+) pie-1::GFP::tubulin]; ieSi38 IV; bmk-1(syb3914) V |
SMW22 | Psun-1::TIR1::mRuby; mCherry::tubulin | unc-119(ed3) III; weIs21 [pJA138 (pie-1::mCherry::tubulin::pie-1)]; ieSi38 IV |
SMW31 | Psun-1::TIR1::mRuby; GFP::H2B; GFP::tubulin | unc-119(ed3) III; ruIs32[unc-119(+) pie-1::gfp::h2b] III; ruIs57[unc-119(+) pie-1::GFP::tubulin]; ieSi38 IV |
SMW46 | dhc-1::degron::GFP; Psun-1::TIR1::mRuby; GFP::H2B; GFP::tubulin | dhc-1(ie28[dhc-1::degron::GFP]) I; unc-119(ed3) III; ruIs32[unc-119(+) pie-1::gfp::h2b] III; ruIs57[unc-119(+) pie-1::GFP::tubulin]; ieSi38 IV |
SMW47 | dhc-1::degron::GFP; Psun-1::TIR1::mRuby; bmk-1(ok391) | dhc-1(ie28[dhc-1::degron::GFP]) I; ieSi38 IV; bmk-1(ok391) V |
SMW48 | dhc-1::degron::GFP; Psun-1::TIR1::mRuby mCherry::tubulin | dhc-1(ie28[dhc-1::degron::GFP]) I; weIs21 [pJA138 (pie-1::mCherry::tub::pie-1)]; ieSi38 IV |
SV1005 | bmk-1(ok391) | bmk-1(ok391) V |
Generation of C. elegans strains
Request a detailed protocolCA1215 was generated via CRISPR/Cas9 editing of the endogenous dhc-1 locus (Zhang et al., 2015).
PHX3914 was generated via CRISPR/Cas9 editing of the endogenous bmk-1 locus by SunyBiotech.
SV1005 was a knockout strain generated by the C. elegans Deletion Mutant Consortium.
SMW46-SMW48 were generated by crossing two strains and screening multiple generations of progeny to ensure homozygosity of all desired traits.
SMW46: Crossed males of SMW31 with CA1215 hermaphrodites
SMW47: Crossed males of CA1215 with SV1005 hermaphrodites
SMW48: Crossed males of CA1215 with SMW22 hermaphrodites
RNAi feeding
Request a detailed protocolOur RNAi protocol is described in detail in Wolff et al., 2022. Briefly, from an RNAi library (Kamath et al., 2003), individual RNAi clones were picked and grown overnight at 37°C in LB with 100 μg/mL ampicillin. Overnight cultures were spun down, resuspended, and plated on nematode growth medium (NGM) plates containing 100 μg/mL ampicillin and 1 mM IPTG. Plates were dried in a dark space overnight at 22°C; at the same time, worms were synchronized for experimentation via bleaching of gravid adults, collecting remaining embryos, and plating on foodless plates to hatch overnight. The next day, newly hatched L1 worms were transferred to appropriate RNAi plates and grown to adulthood at 15°C for 5–6 days. In specific experiments utilizing lin-5(RNAi) or aspm-1(RNAi), L1 worms were plated on empty vector (EV) L440 control RNAi plates for 2–3 days, then transferred to lin-5(RNAi) or aspm-1(RNAi) plates 72 hr prior to fixation.
Immunofluorescence and antibodies
Request a detailed protocolOur IF protocol is described in detail in Wolff et al., 2022. Briefly, adult worms, grown on either EV control RNAi or experimental RNAi, 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) (Laband et al., 2018) in the center of a poly-L-lysine-coated glass slide. Worms were dissected, then fixed via freeze cracking and plunging into –20°C MeOH as described in Oegema et al., 2001. Oocytes and embryos were fixed for 40–45 min, rehydrated in PBS, and blocked in AbDil (PBS with 4% BSA, 0.1% Triton-X-100, 0.02% NaN3) overnight at 4°C. Primary antibodies were diluted in AbDil and incubated with sample overnight at 4°C. The following day, the samples were moved to room temperature and rinsed three times in PBST (PBS with 0.1% Triton-X-100), then incubated with secondary antibodies (diluted in PBST) for 2 hr. Next, the samples were washed three times in PBST again and incubated with mouse anti-α-Tubulin-FITC (diluted in PBST) for 2 hr. Again, the samples were washed three times in PBST, then incubated with Hoescht (1:1000 in PBST) for 15 min. Finally, the samples were washed two times in PBST, mounted in 0.5% p-phenylenediamine, 20 mM Tris-Cl, pH 8.8, 90% glycerol, then sealed with nail polish and stored at 4°C.
The primary antibodies used in this study were rabbit-α-ASPM-1 (1:5000, gift from Arshad Desai), rabbit-α-BMK-1 (1:250, gift from Jill Schumacher), mouse-α-Tubulin-FITC (1:500, DM1α, Sigma), mouse-α-GFP (1:250, 3E6, Invitrogen), mouse-α-SUMO (1:500, gift from Federico Pelisch), rabbit anti-SEP-1 (1:250, gift of Andy Golden), and rabbit-α-SPD-1 (Mullen et al., 2017). All rabbit and mouse Alexa Fluor secondary antibodies (Invitrogen) were used at 1:500.
Ex utero live imaging
Request a detailed protocolFifteen adult worms, grown on either control RNAi or experimental RNAi, were picked into a 10 μL drop of Meiosis Medium (described above) in the center of a custom-made apparatus for live imaging (Laband et al., 2018; Divekar et al., 2021b). All worms were quickly dissected, and an eyelash pick was used to push remaining worm bodies to the outside of the drop, leaving only the oocytes and embryos in the center to avoid disruption from worm movement during the imaging process. Vaseline was laid in a ring around the drop through a syringe, and a 18 × 18 mm #1 coverslip was laid on top of the Vaseline ring, sealing the drop. This sealed slide was moved immediately to the spinning disk stage and inverted, allowing oocytes and embryos to float down to the surface of the coverslip and be subsequently imaged. For movies of unarrested spindles, we looked for embryos that had extruded the first polar body but where the meiosis II spindle had not yet formed; thus t = 0 was synchronized to the onset of meiosis II. For experiments involving emb-30(RNAi), t = 0 was defined as the first frame that we were able to capture after finding a bipolar spindle to image. For these experiments, we did not image spindles that exhibited hallmarks of prolonged metaphase arrest when they were dissected (e.g., stretching of chromosomes and widening of the spindle midzone) to avoid artifacts arising from the metaphase-arrest condition.
Microscopy
Request a detailed protocolAll fixed imaging was performed on a DeltaVision Core deconvolution microscope with a 100x objective (NA = 1.4) (Applied Precision). This microscope is housed in the Northwestern University Biological Imaging Facility supported by the NU Office for Research. Image stacks were obtained at 0.2 μm z-steps and deconvolved using SoftWoRx (Applied Precision). All IF images in this study were deconvolved and displayed as full maximum intensity projections of data stacks encompassing the entire spindle structure (typically ~4–6 μm).
All live imaging was performed using a spinning disk confocal microscope with a 63x HC PL APO 1.40 NA objective lens. A spinning disk confocal unit (CSU-X1; Yokogawa Electric Corporation) attached to an inverted microscope (Leica DMI6000 SD) and a Spectral Applied Imaging laser merge ILE3030 and a back-thinned electron-multiplying charge-coupled device (EMCCD) camera (Photometrics Evolve 521 Delta) were used for image acquisition. The microscope and attached devices were controlled using Metamorph Image Series Environment software (Molecular Devices). Typically, 10–15 z-stacks at 1 μm increments were taken every 15–30 s at room temperature. Images were processed using ImageJ; images are shown as maximum intensity projections of the entire spindle structure. The spinning disk microscope is housed in the Northwestern University Biological Imaging Facility supported by the NU Office for Research.
Auxin treatment
Auxin treatments were performed using three different methods (long term, short term, and acute); further details and diagrams of these methods can be found in Divekar et al., 2021b.
Long-term AID
Request a detailed protocolStandard NGM RNAi plates were prepared, but 500 mM auxin (dissolved in 100% EtOH) was added to a final concentration of 1 mM prior to pouring 6 cm plates. These plates were kept in a dark room to solidify and dry. For long-term auxin experiments, auxin plates were seeded with the desired RNAi culture, incubated overnight to induce dsRNA production, and put at 4°C until needed. In order to achieve a 4 hr auxin treatment, adult worms were grown on separate RNAi plates that did not contain auxin (poured and seeded in parallel with the auxin plates), then transferred to auxin-containing plates 4 hr prior to EtOH fixation or dissection and IF.
Short-term AID
Request a detailed protocolA solution of Meiosis Medium containing 1 mM auxin was prepared from a 200 mM auxin stock (dissolved in 100% EtOH) and kept on ice. Adult worms were picked into 10 μL drops of Medium + auxin, and then the slides were placed inside a custom-made humidity chamber to avoid evaporation of the drop. Following incubation for the desired time (30–45 min), worms were dissected and subjected to the standard IF protocol described above.
Acute AID
Request a detailed protocolFor acute auxin treatments coupled with ex utero live imaging, a solution of Meiosis Medium containing 100 μM auxin was made from a 200 mM auxin stock (dissolved in 100% EtOH). Adult worms were picked into 10 μL drops of Medium + auxin and immediately dissected on a custom-made live imaging slide apparatus (as described in the ex utero imaging section). Slides were moved quickly to the spinning disk microscope, and acquisition was started as soon as an oocyte could be found.
Ethanol fixation for germline counting
Request a detailed protocolEthanol fixation and germline counting methods are described in more detail in Wolff et al., 2022. Briefly, SMW46 worms were subjected to EV control RNAi (or klp-18(RNAi)), utilizing methods described above, and were grown to adulthood over 5 days. 4 hr prior to fixation, worms were either transferred to EV control RNAi (with 1 mM auxin) or left on their original RNAi plates. For each biological replicate, ~40 adults were picked off their respective plates into a 10 μL drop of M9 on a standard glass slide, and a small piece of Whatman paper was used to absorb excess M9, with care being used to avoid pulling worms onto filter paper. Once worms had formed a tight cluster with little residual M9, a 10 μL drop of 100% EtOH was quickly pipetted onto the worms. Within seconds, worms became rigid and straight; once the first EtOH drop had dried, another 10 μL drop was applied and allowed to completely dry. After the third drop was added and dried, a 10 μL drop of 50% Vectashield Mounting Media (Vector Laboratories H-1000) and 50% M9 was placed onto the worms, and a 18 × 18 mm coverslip was gently placed on top. Excess media was aspirated away, slides were sealed with nail polish, and stored at 4°C (typically imaged within a week of fixation).
Data analysis
Request a detailed protocolFigure 1D: Quantifications were made by viewing whole Dynein AID worms expressing GFP::tubulin and GFP::histone and observing three positions in each worm (–1 position, Spermatheca, + 1 position) in both gonad arms. When spindle morphology was clear enough to confidently categorize, spindles were then counted into one of four categories (Microtubule Cage, Unfocused, Focused, Anaphase). The number of spindles counted across all slides (n) is placed above each respective location and condition. Each bar comprises counts from at least four biological replicates. Exact numbers used to generate the graph are given in Table 1.
Figure 1E and F: Categorization of acentrosomal poles into either focused or unfocused/splayed was done by eye, looking at both microtubule and ASPM-1 channels. In any case where microtubule bundles were clearly separated from one another at an acentrosomal pole, with multiple ASPM-1 foci, that spindle was considered unfocused/splayed. Quantifications of acentrosomal pole splaying were made by categorizing multiple fixed images; spindles counted across all slides (n) are placed above each respective condition. Each bar comprises spindles imaged across three biological replicates. Exact numbers used to generate the graph in Figure 1F are given in Table 2.
Figure 1—figure supplement 1: Quantifications were made by viewing whole worms expressing GFP::tubulin and GFP::histone and TIR1, but lacking a degron-tagged version of DHC-1 (labeled as ‘TIR1 control’), and observing three positions in each worm (–1 position, Spermatheca, and +1 position) in both gonad arms. When spindle morphology was clear enough to confidently categorize, spindles were then counted into one of four categories (Microtubule Cage, Unfocused, Focused, Anaphase). The number of spindles counted across all slides (n) is placed above each respective location and condition. Each bar comprises counts from at least three biological replicates. Exact numbers used to generate the graph are given in Table 3.
Figure 1—figure supplement 2A: Quantifications were made by viewing whole Dynein AID worms expressing GFP::tubulin and GFP::histone and observing two positions in each worm (Spermatheca and +1 position) in both gonad arms. When spindle morphology was clear enough to confidently categorize, spindles were then counted into one of three categories (Unfocused, Focused, Anaphase). The number of spindles counted across all slides (n) is placed above each respective location and condition. Each bar comprises counts from at least three biological replicates. Exact numbers used to generate the graph are given in Table 4.
Figure 2B: To determine the degree of splaying at acentrosomal poles, we employed a quantification that measures the ratio of spindle widths between the equator of the spindle (as defined by the metaphase-aligned chromosomes) with the average width of both poles (guided by ASPM-1 staining). Using this pole splaying ratio (r), we displayed these data as boxplots using RStudio; boxplots represent first quartile, median, and third quartile. Statistical significance between the mean r value of control and auxin-treated spindles was determined via a two-tailed t-test for both unarrested and metaphase-arrested conditions.
Figure 2—figure supplement 1A and B: The same data set of control and auxin-treated spindles measured in Figure 2B were further analyzed here. Using Fiji, a rectangular ROI (12 μm or 14 μm wide × 4 μm tall) was used on all images of a single condition to measure the average pixel intensity of ASPM-1 across the whole width of the ROI via the Plot Profile Tool. These values were then averaged from all images in that condition to produce a single value for ASPM-1 intensity at each pixel across the standardized 12 μm/14 μm length. This was plotted as a solid line for each condition, with shaded areas above and below the line representing the SEM. All images were taken with the same exposures for all channels. Spindles were sum projected over 20 slices (0.2 μm step size) for each image prior to measurement. Each condition was averaged using images from at least three biological replicates. Spindle length measurements were done in Imaris. Poles were projected into 3D volumes using the Surfaces Tool (based on fluorescence intensity), and the center of each volume was determined. The exact micron distance between the designated center of the two volumes was measured, and this distance was averaged across images of the same condition. Quantifications were arranged as boxplots, and statistical significance was determined via a two-tailed t-test. Each condition was averaged using images from at least three biological replicates.
Figure 2—figure supplement 2B and C: Timelapses of the TIR1 control strain were opened in Imaris to measure both spindle length and the ratio of pole splaying (r), as described for Figure 2B. Quantifications of spindle length and pole splaying (r) were performed as described above for Figure 2B and Figure 2—figure supplement 1. For display of this data, we employed two methodologies: to avoid averaging out individual variations, we first plotted all 10 timelapses as single traces (green hues for control, violet hues for auxin-treated). However, to get a better visualization of the trends over the course of the timelapse, we also generated averages of the change in length/pole ratio (r) over time (when compared to baseline). For average change plots, shaded area represents the standard deviation amongst all five timelapses in that condition.
Figure 3A: Quantifications were made by viewing whole Dynein AID worms expressing GFP::tubulin and GFP::histone and observing all +1 spindles. When spindle morphology was clear enough to confidently categorize, spindles were then counted as either monopolar (a single structure with chromosomes fanned away from the center) or undergoing breakdown (noticeable separation of microtubules and chromosomes). The number of structures counted across all slides (n) is placed above each respective location and condition. Each bar comprises counts from three biological replicates. Exact numbers used to generate the graph are given in Table 5.
Figure 4C: Timepoints from live imaging videos were analyzed utilizing Imaris. Chromosomes were rendered into 3D surfaces using the ‘Surfaces’ tool, and the center of each volume was determined. For each chromosome that was within the z-stack throughout the entirety of the timelapse, measurements of the distance between the center of segregating chromosomes were taken. These distances were averaged at each 15 s timepoint, and each timelapse was standardized to the same starting point (onset of anaphase) to allow for further averaging between separate videos. The average segregation distance across each condition was plotted (number of videos in each condition represented by n), and shaded area represents SEM of each average.
Figure 7B and C: Classification of monopolar spindles into one of three categories (Breakdown, Mini Anaphases, No Segregation) was done by eye, looking at DNA, microtubule, and ASPM-1 channels. In any case where some bivalents were clearly separated from the monopolar spindle, yet a distinct ASPM-1 monopole remained with some number of attached bivalents, was classified as ‘Breakdown.’ Whenever all six bivalents could be seen dispersed in the cytoplasm with no remaining ASPM-1 monopole, and appeared to have some segregation between chromosomes, that was considered ‘Mini Anaphase.’ Any oocytes containing dispersed chromosomes without any indication of segregation or an anaphase spindle were classified as ‘No Segregation.’ The number of structures counted across all slides (n) is placed above each respective condition. Each bar comprises counts from five biological replicates. Exact numbers used to generate the graph are given in Table 6.
Data availability
All data generated or analyzed in this study are included in the manuscript and supporting files. Source data files have been provided for Figure 2B, Figure 2 - figure supplement 1, Figure 2 - figure supplement 2, and Figure 4C.
References
-
Context-dependent spindle pole focusingEssays in Biochemistry 62:803–813.https://doi.org/10.1042/EBC20180034
-
KLP-7 acts through the Ndc80 complex to limit pole number in C. elegans oocyte meiotic spindle assemblyJournal of Cell Biology 210:917–932.https://doi.org/10.1083/jcb.201412010
-
Evidence for anaphase pulling forces during C. elegans meiosisJournal of Cell Biology 219:5179.https://doi.org/10.1083/jcb.202005179
-
Caenorhabditis elegans oocytes detect meiotic errors in the absence of canonical end-on kinetochore attachmentsThe Journal of Cell Biology 216:1243–1253.https://doi.org/10.1083/jcb.201608042
-
Kinesin-1 and cytoplasmic dynein act sequentially to move the meiotic spindle to the oocyte cortex in Caenorhabditis elegansMolecular Biology of the Cell 20:2722–2730.https://doi.org/10.1091/mbc.E08-12-1253
-
CDK-1 inhibits meiotic spindle shortening and dynein-dependent spindle rotation in C. elegansThe Journal of Cell Biology 193:1229–1244.https://doi.org/10.1083/jcb.201104008
-
Force on spindle microtubule minus ends moves chromosomesThe Journal of Cell Biology 206:245–256.https://doi.org/10.1083/jcb.201401091
-
EMB-30: an APC4 homologue required for metaphase-to-anaphase transitions during meiosis and mitosis in Caenorhabditis elegansMolecular Biology of the Cell 11:1401–1419.https://doi.org/10.1091/mbc.11.4.1401
-
Inhibition of ectopic microtubule assembly by the kinesin-13 KLP-7 prevents chromosome segregation and cytokinesis defects in oocytesDevelopment (Cambridge, England) 144:1674–1686.https://doi.org/10.1242/dev.147504
-
Chromosome segregation occurs by microtubule pushing in oocytesNature Communications 8:1499.https://doi.org/10.1038/s41467-017-01539-8
-
Live imaging of C. elegans oocytes and early embryosMethods in Cell Biology 145:217–236.https://doi.org/10.1016/bs.mcb.2018.03.025
-
Rebuilding MTOCs upon centriole loss during mouse oogenesisDevelopmental Biology 382:48–56.https://doi.org/10.1016/j.ydbio.2013.07.029
-
Kinesin-5 Regulation and Function in MitosisTrends in Cell Biology 29:66–79.https://doi.org/10.1016/j.tcb.2018.08.004
-
A novel chromosome segregation mechanism during female meiosisMolecular Biology of the Cell 27:2576–2589.https://doi.org/10.1091/mbc.E16-05-0331
-
Acentrosomal Microtubule Assembly in Mitosis: The Where, When, and HowTrends in Cell Biology 26:80–87.https://doi.org/10.1016/j.tcb.2015.09.001
-
Spindle assembly and chromosome dynamics during oocyte meiosisCurrent Opinion in Cell Biology 60:53–59.https://doi.org/10.1016/j.ceb.2019.03.014
-
Functional Analysis of Kinetochore Assembly in Caenorhabditis elegansJournal of Cell Biology 153:1209–1226.https://doi.org/10.1083/jcb.153.6.1209
-
Kinesin-5 acts as a brake in anaphase spindle elongationCurrent Biology 17:R453–R454.https://doi.org/10.1016/j.cub.2007.05.001
-
Oocyte Meiotic Spindle Assembly and FunctionCurrent Topics in Developmental Biology 116:65–98.https://doi.org/10.1016/bs.ctdb.2015.11.031
-
MEI-1/MEI-2 katanin-like microtubule severing activity is required for Caenorhabditis elegans meiosisGenes & Development 14:1072–1084.
-
Kinesin-5 inhibitor resistance is driven by kinesin-12The Journal of Cell Biology 213:213–227.https://doi.org/10.1083/jcb.201507036
-
Kif15 cooperates with eg5 to promote bipolar spindle assemblyCurrent Biology 19:1703–1711.https://doi.org/10.1016/j.cub.2009.08.027
-
Methods for Investigating Cell Division Mechanisms in C. elegansMethods in Molecular Biology (Clifton, N.J.) 2415:19–35.https://doi.org/10.1007/978-1-0716-1904-9_2
-
Kinesin-1 mediates translocation of the meiotic spindle to the oocyte cortex through KCA-1, a novel cargo adapterThe Journal of Cell Biology 169:447–457.https://doi.org/10.1083/jcb.200411132
-
The auxin-inducible degradation (AID) system enables versatile conditional protein depletion in C. elegansDevelopment (Cambridge, England) 142:4374–4384.https://doi.org/10.1242/dev.129635
Decision letter
-
Federico PelischReviewing Editor; University of Dundee, United Kingdom
-
Jessica K TylerSenior Editor; Weill Cornell Medicine, United States
-
Federico PelischReviewer; University of Dundee, United Kingdom
-
Thomas Müller-ReichertReviewer; Technische Universität Dresden, Germany
In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.
Decision letter after peer review:
Thank you for submitting your article "Multiple motors cooperate to establish and maintain acentrosomal spindle bipolarity in C. elegans oocyte meiosis" for consideration by eLife. Your article has been reviewed by 4 peer reviewers, including Federico Pelisch as Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Jessica Tyler as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Thomas Müller-Reichert (Reviewer #2).
The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.
Essential revisions:
While all reviewers found your work very interesting, there was agreement that a number of important issues need to be addressed in a revised version. The following is a summary of the main concerns that have to be addressed and you will find the detailed reviews below.
1) The phenotypes should be fully analysed and described quantitatively as much as possible. For each statement, the measured values, the number of measurements and the statistical data should be clearly provided also in the text.
2) The newly generated AID strain(s) need to be more thoroughly characterised:
a) Quantitative information on dynein depletion and its comparison with previously used methods (RNAi and ts alleles) is needed;
b) it would also be important to make sure none of the observed effects depend on the AID or AID::GFP tag which can frequently result in a partially defective protein.
c) control that auxin alone (without any AID) does not cause splayed poles in exposed isolated oocytes.
3) Assessment of protein depletions should be presented, more so in cases where a 'negative' result is shown (see the examples mentioned in the review reports regarding lin-5 and aspm-1).
4) Data in the present manuscript, as it stands, shows that the lab's previous suggested model of Dynein action on chromosome movement is compromised. Also, Dynein localisation itself is different in some cases when compared to the Muscat et al., (2015) paper. Without proper discussion of this issue including plausible explanations for this discrepancy, the reader is left wondering what the actual functions of Dynein are.
5) Results should be based on live imaging experiments whenever possible and the quality of the videos needs to be improved. For example, Figure 2 – supplement 2 is a key Figure and these videos should start with properly assembled Metaphase I spindles and from there, pole disassembly should be assessed. Also, as presented, the difference in intensity between – and + auxin makes this result not entirely convincing to me.
6) Unpublished data should be documented within this manuscript or not discussed at all.
Reviewer #1:
Oocytes in most species contain acentrosomal spindles and the understanding of how these spindles are assembled is limited. The manuscript by Cavin-Meza et al., therefore addresses an important question. The manuscript describes a set of different phenotypes after depletion of the motor proteins Dynein, KLP-18, and BMK-1. While the role of these proteins has been addressed before with variable depth, the current manuscript shows novel phenotypes revealing potential novel functions for these proteins on acentrosomal spindle assembly/stability.
The first part of the manuscript focuses on the role of dynein in female spindle stability. While the use of the AID::dynein allele is an elegant way to address spindle stability, there is insufficient evidence to fully back the authors' claims. A more detailed comparison between this AID allele and the methods used before to assess dynein depletion/loss of function would be critical. Particularly, a comparison with a ts allele, which can also be (and indeed was) used for fast dynein inactivation would increase the confidence in the observed results. This is linked to the apparent lack of reproducibility of some data included in the Muscat et al., (2015) paper from the same lab. It is not clear what the authors' model is for the two different roles (pole stability and chromosome segregation) they propose for Dynein and whether different dynein populations control these roles (spindle and kinetochore).
The authors take advantage of arresting oocytes at metaphase I to allow spindles to form before the acute depletion of Dynein. This is a nice system because it would allow a proper dynamic analysis of spindle pole disassembly in vivo. However, it appears that 'anecdotal' information is taken from the few live imaging experiments and most of the conclusions are taken from experiments involving fixed samples.
Overall, the manuscript includes a set of (very) interesting phenotypes, but it does not go a long way in their characterisation or mechanisms involved. Therefore, in order to thoroughly characterise the presented phenotypes and to put them in context with existing data, the manuscript would benefit from extensive revisions, including new experiments and quantification of all the phenotypes described.
Recommendations for the authors
– The phenotypes should be fully analysed and described quantitatively as much as possible: tubulin levels, spindle length, spindle width, etc.
– The need for live imaging as the core basis for the authors' claims is exemplified in this figure. While the authors do include data from live imaging experiments which seem promising, there are several issues with these experiments that require attention.
1) The authors have an extremely useful system to analyse disassembly but need to make full use of it. They need make sure they image the newly fertilised oocyte (first one after the spermatheca), therefore avoiding side effects due to prolonged arrest.
2) It is not clear how the authors define t = 0. This is not a trivial issue, and it is vital to know whether the two oocytes being analysed have the same 'history' (i.e. been arrested for ~ same time). For example, if one oocyte has just undergone fertilisation whereas the other has been arrested for longer, susceptibility to Dynein depletion could be drastically altered.
3) The lack of consistency in the signal intensity is concerning (see Figure 2D, GFP::histone signal differs massively between '- Auxin' and '+ Auxin').
4) In Figure 2 – supplement 2, the issue highlighted in 3) is even more concerning, but additionally, and equally or more important, the authors should be able to start the imaging experiment from a 'normal' metaphase in both conditions. Without this careful comparison (in addition to imaging GFP::ASPM-1 as described below) the claims about the pole stability defects are not entirely clear.
5) Why is there no Dynein signal on kinetochores as reported previously? Is this related to fixation? In fact, In Figure 2C a clear kinetochore signal is observed, highlighting the issues related to fixed imaging. Also, Dynein localisation during anaphase (mostly on the inward face of chromosomes) seems different from the Muscat et al., (2015) paper, where Dynein localised in the outer face of the chromosomes.
6) Does the AID allele give the same phenotype as the dhc-1(ct76ts) allele previously used by the authors or the dhc-1(or195)? This would also be important to make sure the effects do not depend on the GFP tag which can frequently result in a partially defective protein. These alleles have both been described as ‘fast acting’ and hence would be ideal for the comparison.
7) In Figure 2 – supplement 1 B it appears that the spindle in ‘- Auxin’ has began shortening. A proper comparison is needed using GFP::ASPM-1 in live imaging experiments.
8) There seems to be an enormous defect in segregation in Figure 2 – supplement 2. Why is this?
9) In Figure 3 how can the authors be sure they are comparing all the different conditions in Meiosis I (and not Meiosis II). In some cases, it appears that sisters, rather than homologues, are being observed. Also, the lack of consistency in the magnification for each panel does not help the reader.
10) The comparison of the monopole breakdown phenotype in Figure 3 is difficult to judge without an indication of the depletion efficiencies for lin-5 and aspm-1 RNAi. This makes the paragraph starting in line 216 highly speculative.
11) In Figure 3 – supplement 1, the lin-5(RNAi) phenotype is scarcely described in the text. For example, in monopolar spindles depleted of LIN-5 tubulin and ASPM-1 surround the chromosome surface, with Dynein exhibiting a similar localisation. Additionally, it seems again that Dynein depletion (in combination with lin-5(RNAi)) has no effect on chromosome segregation.
12) Figures 4A and B are very difficult to analyse. Also, if anything they seem to show that Dynein is not required for anaphase chromosome movement, contradicting their previous report.
13) In Figure 5 (and Figure 5 – supplement 1), SPD-1 should be analysed in live oocytes and while the analysis of SUMO localisation is informative, imaging of GFP::separase would contribute important information to the manuscript in terms of how anaphase progresses in the different conditions. This could also provide a good reference point for the unarrested videos.
14) In Figure 5A the authors claim that ‘ASPM-1 labeling could be clearly seen on both sides of these mini spindles, with some enrichment towards areas resembling spindle poles’. This is not clear to me. What is clear is the kinetochore-like signal (which the authors do not mention). Some sort of quantification for the localisation analysis should be provided. This would also benefit from live analysis to see the dynamics of ASPM-1 localisation.
15) The authors could consider depleting kinetochore Dynein recruiters to dissect the role of different populations. Depletion of RZZ components or SPDL-1 could therefore be a useful comparison in the different backgrounds.
Reviewer #2:
Cavin-Meza et al., investigated the role of motor proteins and their coordinated action in maintaining spindle bipolarity under acentrosomal conditions. An auxin-inducible degron system was used to rapidly remove proteins from pre-formed spindles in C. elegans oocytes. In addition, this approach was used to disrupt monopoles and study subsequent microtubule re-arrangement for chromosome segregation in formed miniature spindles. Specifically, the authors find that removal of dynein causes a splaying of onopolaral spindle poles. In combination with depletion of KLP-18, removal of dynein causes a disruption of monopolar spindles. This is followed by a formation of locally dispersed small spindles, which show arrangement of microtubules in between the segregated chromosomes and localization of midzone proteins, such as SPD-1. Interestingly, the authors present evidence for the kinesin-5 family motor protein BMK-1 for providing the force for this chromosome movement in these mini spindles.
The major strength of this paper is the application of the degron system to systematically manipulate oocyte spindle assembly after protein depletion. I don’t see any major weakness. This is a very interesting paper with significantly new information extending our current knowledge on acentrosomal spindles. In general, I am very positive about this manuscript. Most of the conclusions of this paper are supported by data, but some points need to be clarified and some aspects of data presentation could be worked out better.
1) In general, the text reads very well, but real numbers always have to be extracted from figures, figure legends and/or tables. The manuscript reads as a text without numbers. For each statement, the measured values, the number of measurements and the statistical data should be clearly provided also in the text. For instance, it is not clear for each presented experiment how many oocytes/samples have been analyzed.
2) The authors claim a ‚complete dynein depletion’ (in comparison to partial depletions that have been performed by others) but any quantitative information on the level of depletion is missing. Just showing LM images of oocytes with depleted GFP-dynein is not a convincing argument about a full depletion. This should be clarified.
3) I think a general control experiment is completely missing here. Have the authors ever analyzed spindles in oocytes with GFP-tubulin and without the auxin-inducible degron system after exposure to auxin? This is an absolutely necessary experiment to do. Hopefully, auxin alone does not cause splayed poles in exposed isolated oocytes. It has to be kept in mind that oocytes (without the fully developed eggshell) are extremely sensitive. Please add data on such control experiments as supplementary information.
4) There is a lot of ‚jumping’ between Figures 3 and 4 in the text (see paragraph “Dynein acts at spindle poles to promote pole formation and integrity”). The authors might re-think the order of call outs.
5) Figure 4A-B shows key information. I think the panels are way too small, and I really encourage the authors to present these interesting findings in a better way.
6) As for Figure 5 —figure supplement 1, I have a hard time to see the “onopolar persisted on what appeared to be a singular disassembling RC between each segregating chromosome pair”. A clearer image should be presented.
7) Line 384, add ‚monopolar’ to read: “resulted in a catastrophic breakdown of the onopolar spindle…”.
8) Lines 430-438, it is problematic that the authors are discussing the results of the presented manuscript in the context of unpublished data. This unpublished information should be documented somehow within this manuscript or deleted.
9) Line 479 and so forth (line 499, 529 etc.), did the authors use several ‚meiosis media’? In case only one buffer was used, change to ‚medium’. In any case, I would give the experimental conditions here and not fully rely on the details given in a reference.
10) Lines, 539-40, mention the exact auxin treatment protocols that were used within this study. Other, not applied protocols are irrelevant in the context of this paper.
The authors have to take care of these issues before this paper can be accepted for publication in eLife. I would be happy to review the revised version of this manuscript.
Reviewer #3:
Cavin-Meza et al., have parsed out the importance of different motor proteins during oocyte spindle assembly in C. elegans. The authors developed an auxin-inducible dynein degradation system to remove dynein from the spindle. Using this system, they addressed the question of whether dynein is needed to maintain a bipolar spindle. With the depletion of dynein after spindle assembly, the spindles form parallel arrays but still maintain chromosomes in the middle. These results suggested that dynein was involved in pole focusing. To determine further explore the role of dynein in pole focusing, they depleted KLP-18, which normally gives monopolar spindle formation. With depletion of dynein, they find that the monopoles disappear and microtubule bundles form around the chromatin. Intriguingly, the microtubule bundles can allow chromosomes to segregate at anaphase. They then address the question of what motor protein is allowing anaphase onset. They knock out BMK-1 and find that bipolar spindle assembly and chromosome segregation no longer occurs. These results show that BMK-1 contributes to spindle assembly, but somewhat redundantly with other motors. Overall, the experiments were carefully planned and executed. The development of the degradable dynein is a useful tool for the fild. The novel finding that BMK-1 does contribute to spindle assembly is important for understanding the overlapping forces on the spindle. Furthermore, the results shows that there are differences between how the motors function between mitosis and meiosis, which is intriguing and important. Only an additional control experiments is needed.
Overall, the manuscript details an important contribution of how the motor proteins are functioning together for spindle assembly. The manuscript is well-written, with clear rationales. I only have one concern:
1) The aspm-1(RNAi) and lin-5(RNAi) and strains also contain the dynein degron background with dynein tagged. Although the controls show that without auxin, dynein is clearly working, sometimes the degron tags can cause a sensitized background due to a reduction in protein levels or activity. Although the phenotypes for the single depletions have been confirmed with other studies, the aspm-1(RNAi); klp1-18(RNAi) double mutant has not. Can the authors perform the same experiment with a strain without the tagged dynein?
Reviewer #4:
Cavin-Meza et al., investigated the role of dynein motors in meiotic spindle pole organization and spindle function using C. elegans oocytes as a model. They optimized the auxin-inducible degron system to degrade endogenous AID-tagged dynein acutely, rapidly, and more completely in oocytes. They find that dynein is required for integrity of bipolar spindle poles as well as spindle monopoles induced by depletion of the kinesin-12 motor KLP-18. They further show that in the absence of dynein motors, and functional spindle poles, microtubules are still able to organize into spindle-like structures via the action of chromosomes and the kinesin-5 motor BMK-1. They also found that chromosomes can still be segregated in anaphase-like motion in these oocytes.
These results are consistent with many previous studies in mitotic and meiotic cells that investigated the function of dynein on spindle poles and chromosome segregation and showed spindle pole focusing function of this motor. Furthermore, it is known that upon removal of acentrosomal spindle pole components that assemble meiotic spindles in mouse oocytes, chromatin-mediated microtubule assembly can build spindles that can segregate chromosomes. This study is thus generally consistent with data from other models.
The authors have optimized the AID-degron system beautifully and this gives them as well as the field new opportunities to study protein function in C. elegans oocytes in a much cleaner and more complete depletion background. The timescale in which they can deplete endogenous proteins is impressive.
1. The authors previously published a paper where they found dynein inhibition causes chromosome segregation errors and that in this organism oocyte chromosomes are segregated independently of kinetochores in a pathway that involves dynein. While this is not in agreement with a more accepted model where inter-chromosomal microtubule arrays push chromosomes apart, it had so far provided an alternative model for oocyte chromosome segregation in the field. In this current manuscript, the authors show that dynein inhibition, presumably more complete and cleaner than other methods, in fact does not lead to any chromosome segregation errors. This is in direct conflict with their previously published work, but this major discrepancy is not addressed in the current work.
2. The authors base their hypothesis on a previously published study in human oocytes reporting that spindles fail to maintain bipolarity and mis-segregate chromosomes. This work was performed on oocytes that were unhealthy and unsuitable for IVF treatment, and it is now clear that such atretic oocytes do not represent a general oocyte form. While the maternal age effect is indeed of clinical interest, given the strong interest in many areas of cell biology in understanding the fundamental principles of chromosome segregation in oocytes (of many species), it is not necessary to base experiments on observations made using human oocytes.
3. The contribution of chromatin-mediated mitotic and meiotic spindle assembly pathways was largely ignored in the discussion although these are likely to majorly contribute to miniature spindle assembly in dynein AID oocytes. Similarly, the contribution of dynein to spindle pole focusing in other systems needs more coverage.
4. While the images are of high quality and very accessible, the lack of appropriate image analysis and quantification means the manuscript is largely descriptive. It would be important to support several conclusions made throughout the paper with robust quantification and statistical analyses.
5. While the writing is quite accessible to readers working on the C. elegans model, it is much less clear for those who work on other models. It would be very helpful to re-write technical bits such as ‘embryos dissected into auxin’. It is also important to include citations from other models when summarizing what is known about certain spindle features in oocytes. There is generally ambiguity in the writing that leaves the reader wondering about the precise point that is being made. E.g. microtubule bundles were able to remain relatively aligned’ – this and several other statements in the paper are very descriptive and without quantitative image analysis to back them up, it is difficult to grasp what they mean.
This manuscript can be significantly improved by providing a detailed discussion concerning the conflict between the data within and the lab's previously published work that shows dynein inhibition causes chromosome segregation errors. Furthermore, robust image quantification and statistical analyses are needed to make the paper much less descriptive cell biology than it is now. It would also be good to include some controls to show that 4 hour treatment of non-AID strain wild-type oocytes has no effect on several aspects of spindle assembly and function studied here.
https://doi.org/10.7554/eLife.72872.sa1Author response
Essential revisions:
While all reviewers found your work very interesting, there was agreement that a number of important issues need to be addressed in a revised version. The following is a summary of the main concerns that have to be addressed and you will find the detailed reviews below.
1) The phenotypes should be fully analysed and described quantitatively as much as possible. For each statement, the measured values, the number of measurements and the statistical data should be clearly provided also in the text.
To better support our conclusions, we have made efforts to provide more rigorous quantification of numerous aspects of our data. First, we performed new pole splaying quantifications. Instead of judging by eye whether a pole was “focused” or “splayed” (as we did in the original submission), we measured the widths of the poles and of the center of the spindle; we then calculated a ratio of the pole to equator width as a measurement of how splayed the poles are (this is shown in Figure 2B). This new analysis confirmed the finding that we reported in our original manuscript – that the poles are substantially splayed upon auxin treatment. Additionally, we have added rigorous controls to demonstrate that the effects we see are not caused by auxin treatment (Figure 2 —figure supplement 2). We dissected oocytes lacking degron-tagged dynein into auxin and then imaged spindles over time, demonstrating that neither the spindle length nor the pole-to-equator ratio changed. Finally, we now report the number of images analyzed for all figures, we have added quantitative information into the manuscript when discussing results, and we have tried to emphasize the more quantitative aspects of our paper in our revised manuscript, such as our measurements of chromosome segregation rates and distance traces in live imaging.
2) The newly generated AID strain(s) need to be more thoroughly characterised:
a) Quantitative information on dynein depletion and its comparison with previously used methods (RNAi and ts alleles) is needed;
b) it would also be important to make sure none of the observed effects depend on the AID or AID::GFP tag which can frequently result in a partially defective protein.
c) control that auxin alone (without any AID) does not cause splayed poles in exposed isolated oocytes.
a) In the original version of the manuscript we made a few statements implying that we were achieving “complete” dynein depletion using our AID method, contrasting this to previous studies by ourselves and others that used partial RNAi or temperature-sensitive mutants. However, the reviewers rightly pointed out that since we had not quantified the level of depletion, we could not make this statement. Therefore, for the revision, we attempted to quantitatively assess dynein depletion using western blotting. Unfortunately, we had difficulties transferring DHC-1, given the large size of this protein (especially with the GFP and degron tags added); we extensively tried to optimize this DHC-1 western but we were unsuccessful, despite getting westerns to work for other AID strains in our lab (for one example, see our recent BioRxiv preprint, Mullen and Cavin-Meza, et.al., 2022). We therefore ask that this experiment not be required for acceptance of the manuscript, as the data we present is consistent with substantial dynein depletion. We show a drastic reduction of DHC-1::GFP signal upon dissection of oocytes into auxin, undetectable dynein via immunofluorescence following auxin treatment, and the phenotypes that we report are highly reproducible, suggesting that there is not a lot of variability in the efficiency of protein depletion between experiments. However, since we recognize that we have been unable to demonstrate that our depletion is in fact complete (or to compare the level of depletion to the other methods, due to the problems with western blotting noted above), we have revised our manuscript to soften our language, and we no longer claim that we have achieved complete depletion. The revised wording therefore more accurately describes our results.
b) The phenotypes observed in our dynein depletion conditions do not appear to depend on the GFP tag. The pole splaying phenotypes that we observe following dynein depletion are very similar to what has been reported for partial dynein or dynactin depletion via RNAi by another lab, so pole splaying on bipolar spindles does not appear to depend on the GFP tag. Moreover, we previously reported the results of partial dynein inhibition on monopolar spindles (Muscat, et.al. 2015); our goal when we performed those original experiments was to ask whether dynein provided chromosomes with a minus-end directed force (i.e., would chromosomes be found further out towards the plus ends on a monopolar spindle following dynein depletion?). When we optimized those experiments we tried multiple conditions for partial dynein RNAi (i.e., different amounts of time on the RNAi plates) and we noticed that with longer incubations on the RNAi plates, we could not find any monopolar spindles – chromosomes appeared dispersed throughout the oocyte (see Author response image 1).
At the time we did not fully understand this phenotype, so we only analyzed oocytes where monopolar spindles were intact (with more partial dynein depletion conditions) to address our question about forces on chromosomes. However, given the results presented in the current manuscript, we now understand that we could not find monopolar spindles with more complete dynein depletion because the monopole had blown apart, just like what we report in the current manuscript using the degron strain. This past observation suggests that the monopole dissolution phenotype does not depend on the GFP tag either.To speak to the functionality of DHC-1::degron::GFP, we have never observed defects in any aspect of the meiotic divisions in the Dynein AID strain in the absence of auxin. In the original Zhang et al., paper from the Dernburg lab, where this strain was first reported, no significant changes in brood size or percentage of male progeny were detected (this information is found in their supplemental tables); we have now added this information to our Results section. In our own hands, we similarly have not observed anything in worm upkeep (development, viability, brood size, etc.) that would suggest that there are unintended consequences of tagging DHC-1, and we have not perceived any changes in spindle morphology in the absence of auxin in this strain, when compared to wild-type spindles (see various control images and videos throughout our manuscript). Therefore, while we cannot be 100% certain that the tagged DHC-1 is fully wild-type, we don’t think that there are major issues with this strain that would affect the conclusions of our manuscript.
c) We have now performed this important control experiment utilizing a worm strain expressing TIR1 and same fluorophores (GFP::tubulin and GFP::histone), but lacking the degron-tagged version of DHC-1 (this experiment is now presented in Figure 2 —figure supplement 2). In this strain, exposure of isolated oocytes to auxin has no noticeable effect on spindle morphology; spindle length remained constant and the ratio between the pole and equator widths (a measurement of pole splaying, as described in response to point #1) also did not change.
3) Assessment of protein depletions should be presented, more so in cases where a 'negative' result is shown (see the examples mentioned in the review reports regarding lin-5 and aspm-1).
In the original manuscript, we presented double RNAi of klp-18(RNAi) with either lin-5(RNAi) or aspm-1(RNAi), and we observed that some monopolar spindles broke down. This phenotype was similar to the dynein depletion phenotype on monopolar spindles, but was less penetrant since some monopoles were intact; this is the negative result noted above by the reviewer. In the revision, we re-evaluated the inclusion of this partial breakdown result, given the difficulties in interpreting these data. Therefore, this negative result is no longer presented in the revised manuscript. For an explanation of why we have not added quantification of dynein depletion, please see point #2 above.
4) Data in the present manuscript, as it stands, shows that the lab's previous suggested model of Dynein action on chromosome movement is compromised. Also, Dynein localisation itself is different in some cases when compared to the Muscat et al., (2015) paper. Without proper discussion of this issue including plausible explanations for this discrepancy, the reader is left wondering what the actual functions of Dynein are.
It is true that we previously hypothesized in Muscat et al., (2015) that dynein contributed to chromosome segregation; this was based on evidence that (1) chromosomes are subjected to minus-end forces prior to anaphase, (2) dynein was observed near the outside surface of segregating chromosomes, and (3) there were lagging anaphase chromosomes in a dynein temperature-sensitive mutant following a shift to the restrictive temperature. However, in that paper we noted that the dynein ts mutant phenotype was not completely penetrant – although we saw lagging chromosomes, chromosome segregation was not completely blocked. Therefore, in the discussion we were open about the fact that, even if our model that dynein facilitated chromosome segregation was correct, that other mechanisms may contribute. In particular, we stated that since the spindle elongates in anaphase, that it was “likely that spindle elongation also contributes to chromosome segregation”. In the years since our paper was published, it has become clear based on the work of multiple labs that this anaphase B-like spindle elongation is the major driver of chromosome segregation, and that, if dynein does play a role in segregation, it is redundant with other mechanisms. However, since we never held the viewpoint that dynein was the only driver of chromosome segregation, we do not see a major contradiction between our 2015 study and the conclusions of the current study. The findings in our 2015 paper and current manuscript are also consistent; dynein depletion does cause lagging chromosomes (seen in our original study and also in Figure 2 —figure supplement 3 of the current study), and dynein does localize near the outer surface of chromosomes (seen in our original study and also in Figure 1C of the current study, although it is now recognized based on the work of other labs that this likely represents pole, rather than chromosomal staining). Therefore, we simply see this as an example of the field moving forward in the 7 years since our first study was published, changing the interpretation of some of our original experiments. Our new study reinforces the current view that chromosome segregation does not require dynein, by revealing another experimental condition where chromosomes can segregate following dynein depletion (in this case also in the absence of KLP-18); chromosome segregation under these conditions is likely driven by anaphase B-like spindle elongation, since the miniature spindles lengthen as chromosomes segregate. We have now included a discussion of these issues related to our 2015 paper in our Discussion section.
5) Results should be based on live imaging experiments whenever possible and the quality of the videos needs to be improved. For example, Figure 2 – supplement 2 is a key Figure and these videos should start with properly assembled Metaphase I spindles and from there, pole disassembly should be assessed. Also, as presented, the difference in intensity between – and + auxin makes this result not entirely convincing to me.
In our manuscript, all of the major conclusions are supported by both live and fixed imaging. Live imaging is used to show that (1) acentrosomal spindle poles splay following acute dynein depletion, (2) monopoles break apart following dynein depletion, (3) miniature spindles form that support chromosome segregation following monopole breakdown, and (4) that the formation of these miniature spindles requires BMK-1. We think that the live and fixed imaging methods used in the manuscript are complementary, as live imaging provides essential dynamic information, while fixed imaging provides higher resolution images and the ability to easily assess multiple spindle markers, to better understand the phenotypes.
In the revision, we have made efforts to generate additional live imaging data and to improve the quality of our videos as requested. In the original submission, we had a few videos, such as the ones mentioned in Figure 2 —figure supplement 2 (now Figure 2 —figure supplement 3), where the intensity of the GFP::histone signal was different when comparing the control (auxin) and experimental (+auxin) conditions. We do acknowledge that we frequently observe some variation in fluorescence intensity between videos. This is because we utilize the same laser power, EM Gain, and exposure timings across every video for a given worm strain. However, the embryo is a three-dimensional object, and the depth of the spindle in the Z-plane (i.e., how close the spindle is to the coverslip) can significantly impact the clarity and brightness of the timelapse. We do not think that these differences in signal intensity are indicative of atypical variance in expression levels, and we also do not think that they reflect poorly on the quality of our data, as the phenotypes that we observe are very consistent, despite the differences in the GFP signal. However, to make the videos easier to evaluate, we have generated new videos, and we have swapped out the timelapses highlighted in our figures with better matched control and experimental examples (i.e., where the intensity of the fluorophores are more similar between the two conditions). The phenotypes observed in the new videos are identical to those reported in the original version of the manuscript.
We also agree with the reviewer that we would ideally start the timelapses in Figure 2 —figure supplement 2 (now Figure 2 —figure supplement 3) with fully assembled Metaphase I spindles, to show that the poles of pre-formed unarrested spindles splay upon dynein depletion (as we show in our metaphase-arrest experiments). However, when we start with unarrested Metaphase I spindles and begin filming, these spindles have usually shortened and transitioned into anaphase before dynein is fully depleted, complicating analysis. Therefore, we instead scan the slide for embryos where the first polar body has just been extruded (where the Meiosis II spindle has not yet formed) and we immediately start filming. In these videos, spindles begin to form as dynein is being depleted, by the time these spindles reach metaphase, the poles are clearly splayed. Although this is not a perfect experiment, it is the best we can do, given how long unarrested oocytes typically stay in metaphase (not long), and the timing of dynein depletion.
We think that these improvements to the manuscript better support our conclusions, and hope that they will be acceptable to the reviewers.
6) Unpublished data should be documented within this manuscript or not discussed at all.
In our discussion, we have a few sentences describing results from another study performed in our lab; when we submitted our original manuscript this other study was unpublished. However, we recently uploaded this work to BioRxiv (Mullen and Cavin-Meza, et al., 2022), so we now cite this preprint in the current manuscript.
Reviewer #1:
Recommendations for the authors
1) The authors have an extremely useful system to analyse disassembly but need to make full use of it. They need make sure they image the newly fertilised oocyte (first one after the spermatheca), therefore avoiding side effects due to prolonged arrest.
We agree that avoiding embryos that have been arrested for long periods of time is critical. However, it is difficult to implement this specific suggestion of the reviewer for technical reasons. For our two major assays (live imaging and fixed immunofluorescence), we need to dissect the oocytes out of the worm to image them, so we lose the positional information of the germ line and we cannot tell which oocyte is the most recently fertilized. Although in principle we could soak whole worms in auxin and then image depletion of dynein in an intact worm (so we can make sure to image the most newly fertilized oocyte, as the reviewer suggests), we have found that it takes substantially longer to deplete proteins when the oocytes are not dissected, likely because it takes additional time for the auxin to reach the oocytes (we discussed these various methods in more detail in Divekar, et.al. 2021, Current Protocols). Therefore, it would not be possible to do the type of rapid depletion experiments that we present in the paper (depleting dynein within minutes), in a situation where we could retain the positional information of the germ line.
However, after looking at many hundreds of metaphase-arrested embryos in both live and fixed imaging conditions, we are confident that we can accurately identify the hallmarks of prolonged arrest (reduced dynamics of spindle and chromosomes, stretching of chromosomes, gradual widening of spindle midzone). For the experiments in the manuscript, we did not analyze any oocytes that displayed these phenotypes, which we hope minimizes this concern of the reviewer. We have now added a sentence to the Materials and methods (section on live imaging) to make it clear that we excluded some oocytes from imaging for this reason. In addition, for the major experiments of the paper, we also show similar phenotypes in unarrested conditions, so we do not believe that the metaphase arrest influences the conclusions of our manuscript.
2) It is not clear how the authors define t = 0. This is not a trivial issue, and it is vital to know whether the two oocytes being analysed have the same 'history' (i.e. been arrested for ~ same time). For example, if one oocyte has just undergone fertilisation whereas the other has been arrested for longer, susceptibility to Dynein depletion could be drastically altered.
This is a good point, and we have now described how we define t=0 in in the manuscript, in the “Ex utero live imaging” section of the Materials and methods. For embryos that are not arrested, we have acquired a set of new videos where t=0 is the onset of Meiosis II. We believe that this is a reasonable point for synchronization because Metaphase I is difficult to catch immediately after dissection (it requires considerable luck to find an embryo at metaphase I shortly after mounting the slide). Additionally, when we catch Metaphase I spindles that are unarrested, by the time DHC-1 is depleted from embryos, these spindles have usually shortened and transitioned into anaphase, complicating analysis. Therefore, we changed our strategy to instead scan the slide for embryos where the first polar body has just been extruded (where the meiosis II spindle has not yet formed); this is t=0.
For our metaphase-arrested videos, t=0 is defined as the onset of filming (roughly 5 minutes post-dissection, depending on the speed of mounting the slide and how quickly we are able to find a suitable embryo). We acknowledge that it is possible that the duration of arrest could have unforeseen consequences on depletion phenotypes. However, we exclude embryos that appear to be exhibiting prolonged metaphase arrest (see point #1 above), and the phenotypes from the embryos that we observe are extremely consistent despite the lack of perfect synchronization -- particularly the pole splaying phenotypes we observe in both arrested and unarrested conditions (i.e., Figure 2D versus Figure 2 —figure supplement 3). Therefore, we do not think that the duration of the arrest, if it indeed has an effect on the susceptibility to dynein depletion, is having a big enough effect to alter our conclusions; if this were the case then we think that we would see more variability in our phenotypes.
3) The lack of consistency in the signal intensity is concerning (see Figure 2D, GFP::histone signal differs massively between '- Auxin' and '+ Auxin').
We agree that there were differences in the intensity of the GFP::histone signal between the -Auxin and +Auxin conditions of Figure 2D in the original manuscript; to address this concern we have replaced the original -Auxin video with a newly acquired one where the GFP::histone signal is more similar to the +Auxin video. We do acknowledge that we frequently observe some variation in fluorescence intensity between videos. This is because we utilize the same laser power, EM Gain, and exposure timings across every video for a given worm strain. However, the embryo is a three-dimensional object, and the depth of the spindle in the Z-plane (i.e., how close the spindle is to the coverslip) can significantly impact the clarity and brightness of the timelapse. We do not think that these differences in signal intensity are indicative of atypical variance in expression levels, and we also do not think that they reflect poorly on the quality of our data, as the phenotypes we observe are very consistent, despite the differences in the GFP signal.
4) In Figure 2 – supplement 2, the issue highlighted in 3) is even more concerning, but additionally, and equally or more important, the authors should be able to start the imaging experiment from a 'normal' metaphase in both conditions. Without this careful comparison (in addition to imaging GFP::ASPM-1 as described below) the claims about the pole stability defects are not entirely clear.
We agree with this point. We have now taken additional videos of unarrested embryos so that the videos in Figure 2 – supplement 2 (now Figure 2 – supplement 3) have been replaced; we also now show two videos for both the +Auxin and -Auxin condition. For these new videos we synchronize t=0 to the onset of Meiosis II (please revisit point #2 for an explanation of why we chose this timepoint, and why we cannot start imaging at metaphase, as suggested by the reviewer). See point #7 below for in-depth discussion of GFP::ASPM-1 imaging.
5) Why is there no Dynein signal on kinetochores as reported previously? Is this related to fixation? In fact, In Figure 2C a clear kinetochore signal is observed, highlighting the issues related to fixed imaging. Also, Dynein localisation during anaphase (mostly on the inward face of chromosomes) seems different from the Muscat et al., (2015) paper, where Dynein localised in the outer face of the chromosomes.
It is true that we see kinetochore staining more clearly in live imaging with the mCherry::Tub, DHC-1::deg::GFP strain. However, we do often see kinetochore staining in fixed imaging as well. Kinetochore staining is present in many of our images (see Figure 1C, metaphase, Figure 1E, -Auxin control; you can see faint cup-like staining adjacent to the bivalents); although we realize that this is hard to see in the projection image in the figure, when we look through z-stacks the kinetochore staining is more obvious in single slice images. Therefore, we think that dynein on the spindle obscures the kinetochore pattern, making it less obvious in fixed images of bipolar spindles. Supporting this interpretation, kinetochore staining is much clearer when spindle-associated dynein is reduced (see Figure 1 —figure supplement 2B; aspm-1(RNAi) image) or when the chromosomes are more spread out, reducing interference from spindle-localized dynein (see the various fixed monopolar spindle images throughout the manuscript). Therefore, we do not think that our data conflicts with published work. In terms of anaphase, while we sometimes observe some DHC-1 staining near the inside face of chromosomes, we do not consider this stronger than the localization observed on the poleward side of chromosomes (very common across all of our images).
In addition to this specific point, we would also like to address the concerns raised about our fixed imaging experiments more generally. While we certainly agree that fixed imaging has caveats (no tracking of dynamics, potential artifacts from fixation, non-specific binding of antibodies), we also wish to emphasize that fixed imaging can have significant benefits over live imaging (improved resolution of signal, ability to visualize numerous channels simultaneously, can obtain data quickly without months of strain generation and crosses that be genetically difficult). We believe that both fixed and live imaging have specific contexts in which they are the better-suited method of acquisition. Therefore, we have strived to utilize both methods in our study and we think that these imaging modalities are complementary; although there may be some slight differences in our data depending on the method by which they were acquired (e.g. the different levels of dynein detected at kinetochores, as noted by the reviewer), we don’t think that any of the major conclusions of our manuscript are compromised by these caveats.
6) Does the AID allele give the same phenotype as the dhc-1(ct76ts) allele previously used by the authors or the dhc-1(or195)? This would also be important to make sure the effects do not depend on the GFP tag which can frequently result in a partially defective protein. These alleles have both been described as ‘fast acting’ and hence would be ideal for the comparison.
We have not performed direct comparisons between our Dynein AID strain and either of the mentioned fast-acting ts alleles; since we do not have a microscope stage that can be quickly heated, assessing whether a shift to the restrictive temperature affects the organization of preformed spindles would be technically challenging for us with our available equipment. Therefore, we respectfully ask that these experiments not be required for acceptance.
Regardless, we do not think that the phenotypes seen in our dynein depletion conditions depend on the GFP tag. The pole splaying phenotypes that we observe following dynein depletion are very similar to what has been reported for partial dynein or dynactin depletion via RNAi by the McNally lab, so pole splaying on bipolar spindles does not appear to depend on the GFP tag. Moreover, we previously reported the results of partial dynein inhibition on monopolar spindles (Muscat, et.al. 2015); our goal when we performed those original experiments was to ask whether dynein provided chromosomes with a minus-end directed force (i.e., would chromosomes be found further out towards the plus ends on a monopolar spindle following dynein depletion?). When we optimized those experiments we tried multiple conditions for partial dynein RNAi (i.e., different amounts of time on the RNAi plates) and we noticed that with longer incubations on the RNAi plates, we often could not find any monopolar spindles – chromosomes appeared dispersed throughout the oocyte with associated microtubules (see Author response image 2). At the time we did not fully understand that phenotype, so we only analyzed oocytes where monopolar spindles were intact (with more partial dynein depletion) to address our question about forces on chromosomes. However, given the results presented in the current manuscript, we now understand that we could not find monopolar spindles with more complete dynein depletion because the monopole had blown apart, just like what we report in the current manuscript using the degron strain. This observation using RNAi suggests that the monopole dissolution phenotype does not depends on the GFP tag either. Since we made these observations during experimental optimization, we did not quantify the phenotype at the time and so we are not able to add these data to the manuscript without re-doing those partial depletion experiments to generate more n’s. Therefore, in the interest of time, we ask that this not be required for acceptance (though if the reviewer considers this an essential experiment, we certainly could do it again).
To speak to the functionality of DHC-1::degron::GFP, we have never observed defects in any aspect of the meiotic divisions in the Dynein AID strain (CA1215) in the absence of auxin. To our eyes, all of our control experiments with this strain present normally when compared to spindles from other common lab strains such as N2, EU1067 (GFP::tubulin; GFP::histone), or OD868 (GFP::tubulin; mCherry::histone). While we understand that adding any tag to a protein has the possibility of interfering with functionality, we have not observed anything in worm upkeep (development, viability, brood size, etc.) or experimentally that would suggest that there are unintended consequences of tagging DHC-1. In the original Zhang et al., paper from the Dernburg lab, where this strain was first reported, no significant changes in brood size or % of male progeny were detected (found in their supplemental tables); we have now added this information to our Results section and we have cited their paper in the relevant section of the manuscript.
7) In Figure 2 – supplement 1 B it appears that the spindle in '- Auxin' has began shortening. A proper comparison is needed using GFP::ASPM-1 in live imaging experiments.
We went back to the original image file for this particular control spindle, and we noticed that it is slightly tilted in the z-plane, which we think gave the appearance that it is shortening. However, the measured pole to pole length of the image in question is 6.13 microns, very close to the mean spindle length calculated in our data (6.07 microns). If this is truly a point of contention, we will happily remove this particular image and replace it with a different representative image.
We agree that imaging ASPM-1 live would complement the fixed imaging analysis in Figure 2 – supplement 1, and in principle this is a good idea. However, since it would require a substantial amount of work (due to some issues with strain construction), we are of the opinion that it may not be worth the effort. In principle, we could cross an existing GFP::ASPM-1 strain into the Dynein-AID strain. However, since Dynein is also tagged with GFP in this strain, we believe that this will make it difficult to interpret GFP::ASPM-1 localization during our videos. Since DHC-1 is on poles and the spindle itself, it would be hard to determine what signal was truly ASPM-1 until DHC-1 was fully depleted, which somewhat defeats the purpose of filming ASPM-1 localization in the first place. An alternative would be to use CRISPR/Cas9 to tag ASPM-1 with mCherry in our Dynein AID strain, or vice versa (DHC-1::degron::mCherry into EU2876, the GFP::ASPM-1 strain). While this is a feasible experimental setup, we do not believe that the work and time investment of generating either of these strains would be worth it. The data that we present in our linescan analysis of fixed images is consistent and reproducible, and the pole splaying phenotype that we observe in both live and fixed imaging is extremely clear (with and without being able to see ASPM-1 localization).
While we do agree that observing ASPM-1 using live imaging could further support some of our observations, we anticipate that we might also lose some of the resolution that we took advantage of for this linescan quantification (please note the tightness of our averages and SEM error bars). Therefore, due to the complications (and time investment) of generating a new strain using CRISPR, and our perception that it would not significantly improve the quality of our data, we have opted to not perform this experiment for this revision and we respectfully ask that it not be required for acceptance of our manuscript.
8) There seems to be an enormous defect in segregation in Figure 2 – supplement 2. Why is this?
This is correct; in our original representative video for this supplement, the spindle in question had failed to extrude the first polar body in Meiosis I, resulting in additional chromosomes being present during Meiosis II. We think that this is one cause of the lagging chromosomes in anaphase. Moreover, this phenotype is also consistent with previous studies that reported lagging chromosomes following dynein depletion in oocytes (one example is our own previous work in Muscat, et.al. 2015, and this same observation was subsequently made by McNally et.al. 2016). Although we originally thought that this would be acceptable to use as the representative video even though the polar body had failed to extrude (since it showed the pole splaying phenotype), after considering point #2 above about synchronization of videos to a standardized t=0, we have replaced this +Auxin video with two newly acquired videos, both synchronized to start filming at the onset of Meiosis II (with clear evidence of a successful polar body extrusion in Meiosis I).
9) In Figure 3 how can the authors be sure they are comparing all the different conditions in Meiosis I (and not Meiosis II). In some cases, it appears that sisters, rather than homologues, are being observed. Also, the lack of consistency in the magnification for each panel does not help the reader.
It is true that our original Figure 3 contained examples of both Meiosis I and Meiosis II spindles. We found that depletion of dynein from either MI or MII monopolar spindles caused the monopoles to blow apart; there was no difference in the phenotype, so we showed examples of both. While the chromosomes are different between Meiosis I and II (bivalents vs. individual chromosomes), we are not aware of any literature that has demonstrated major differences in the organization of acentrosomal spindles when comparing the two meiotic divisions in C. elegans. Therefore, we do not see any inherent complication in assessing both MI and MII spindles and showing examples of both (since the phenotypes that we report are the same for both divisions). However, if there is a reason we have not thought of for why we should not treat these spindles as equal, we would appreciate that information, so that we can reconsider this point and adjust this figure.
Regarding the differences in magnification, we agree that it would in theory be better to crop everything the same for consistency. However, we cropped images the way we did to maximize the clarity of each image. Specifically, in cases where the images have been considerably zoomed out, we felt that this better captured the sense of monopole breakdown in relation to the distance across the cytoplasm; the reader can get a better sense of just how far these chromosomes have drifted apart when we present a more zoomed out image. We understand that this creates inconsistency in the figure, since the control monopolar spindles are zoomed in more. However, if we cropped these control images so that they were zoomed out a similar amount, the spindle would be much smaller (and harder to see), so we would prefer not to do this. Instead, we put a sentence in the Figure 3 legend, explaining to the reader that the +Auxin images are zoomed out more (to make sure they notice this when evaluating our images). However, if the reviewer feels strongly about this point, we would gladly reconsider redoing our cropping.
10) The comparison of the monopole breakdown phenotype in Figure 3 is difficult to judge without an indication of the depletion efficiencies for lin-5 and aspm-1 RNAi. This makes the paragraph starting in line 216 highly speculative.
This is a fair criticism and we have re-evaluated the inclusion of our partial breakdown of monopolar spindles using either lin-5 or aspm-1 RNAi, given the difficulties in interpreting these data. Upon reflection, we have decided to restructure our lin-5 and aspm-1 RNAi data, removing the data in question (See Figure 1 —figure supplement 2), and we have adjusted the manuscript accordingly.
11) In Figure 3 – supplement 1, the lin-5(RNAi) phenotype is scarcely described in the text. For example, in monopolar spindles depleted of LIN-5 tubulin and ASPM-1 surround the chromosome surface, with Dynein exhibiting a similar localisation. Additionally, it seems again that Dynein depletion (in combination with lin-5(RNAi)) has no effect on chromosome segregation.
As noted in response to point #10, we have restructured our lin-5 and aspm-1 RNAi data and have adjusted the manuscript to discuss our findings in a different manner (Please see Figure 1 —figure supplement 2 and corresponding text in the Results section describing this figure). This reorganization has removed the data highlighted in this comment. We have discussed dynein depletion and chromosome segregation in the next point (#12).
12) Figures 4A and B are very difficult to analyse. Also, if anything they seem to show that Dynein is not required for anaphase chromosome movement, contradicting their previous report.
In response to this and a similar comment from Reviewer 2 (point #5), we have reorganized Figure 4 so that we could make the images larger and easier to see; we also added zooms that highlight the key findings – that microtubules are able to reorganize and chromosomes are able to segregate on these miniature spindles. We hope that these improvements will make the data easier to evaluate.
To speak to the comment of how this data may contradict our previous 2015 paper in eLife (also raised by reviewer 3, point #1), the reviewer is correct that we previously hypothesized that dynein contributed to chromosome segregation; this was based on evidence that chromosomes are subjected to minus-end forces prior to anaphase, and our observation that there were lagging chromosomes in a dynein temperature sensitive mutant following a shift to the restrictive temperature. However, in that paper we noted that the phenotype was not completely penetrant – although we saw lagging chromosomes, chromosome segregation was not completely blocked. Therefore, in the discussion we were open about the fact that, even if our model that dynein facilitated chromosome segregation was correct, that other mechanisms may contribute. In particular, we stated that since the spindle elongates in anaphase, that it was “likely that spindle elongation also contributes to chromosome segregation”. In the years since our paper was published, it has become clear based on the work of multiple labs that this anaphase-B like spindle elongation is the major driver of chromosome segregation, and that, if dynein does play a role in segregation, it is redundant with other mechanisms. However, since we never held the viewpoint that dynein was the only driver of chromosome segregation, we do not see a major contradiction between our 2015 study and the current study. The findings in our 2015 paper and current manuscript are also consistent; dynein depletion does cause lagging chromosomes (seen in our original study and also in Figure 2 —figure supplement 3 of the current study). Therefore, we simply see this as an example of the field moving forward in the 7 years since our first study was published, changing the interpretation of our original experiments. Our current study reinforces this updated view, showing another experimental condition where chromosomes can segregate following dynein depletion (in this case also in the absence of KLP-18). We have now included a more thorough discussion of this issue in our Discussion section.
13) In Figure 5 (and Figure 5 – supplement 1), SPD-1 should be analysed in live oocytes and while the analysis of SUMO localisation is informative, imaging of GFP::separase would contribute important information to the manuscript in terms of how anaphase progresses in the different conditions. This could also provide a good reference point for the unarrested videos.
Although in principle it would be nice to have live imaging of SPD-1 to complement our fixed imaging, we do not believe that the perceived benefits of utilizing live instead of fixed imaging are large enough to warrant the time it would take to do this experiment. SPD-1 localization is only apparent during a short window of time (anaphase spindles that exist for ~5 minutes), and we think that it would be hard to catch this window (and to interpret the resulting videos) if the chromosomes were not also labeled in this strain. Therefore, we would have to cross both SPD-1::GFP and mCherry::histone into our Dynein degron strain, but the main mCherry::histone transgene that people use (made by the Desai lab in strain OD56), is on chromosome IV, the same chromosome as TIR1. Therefore, this would be a very complicated cross and it require a lot of effort to generate the necessary strain (and then of course there would be additional time to acquire the actual videos). Since we do not think that our fixed imaging is missing any crucial localization information, we don’t think that this suggested experiment would be worth potentially months of work. As an alternative, we have repeated our fixed SPD-1 staining again (imaged an additional ~50 embryos in auxin treatment) to strengthen our SPD-1 data, and we have replaced the original representative images with clearer examples; we hope that this is satisfactory to the reviewer.
We greatly appreciate the suggestion of imaging separase. Due to the difficulties in performing live imaging (which would involve the same strain construction issues noted above, since we would need to use mCherry::histone to label chromosomes), we have chosen to perform fixed imaging to assess separase. We have generated a brand new supplemental figure (Figure 5 —figure supplement 2) to demonstrate that separase localization on miniature anaphases is consistent with what has been seen on wild-type anaphase spindles.
14) In Figure 5A the authors claim that 'ASPM-1 labeling could be clearly seen on both sides of these mini spindles, with some enrichment towards areas resembling spindle poles'. This is not clear to me. What is clear is the kinetochore-like signal (which the authors do not mention). Some sort of quantification for the localisation analysis should be provided. This would also benefit from live analysis to see the dynamics of ASPM-1 localisation.
We have updated our manuscript to mention the off-target kinetochore staining that sometimes occurs when utilizing this ASPM-1 antibody (originally documented in Wignall and Villeneuve 2009); we had originally pointed out this background staining only in the figure legend, but now we have also added this information to the Results section so that it will be easier for the reader to find. While this image unfortunately does have more ASPM-1 background staining than most of our other images, we believe that our observation of ASPM-1 signal on either side of these individual, seemingly bipolar spindles around individual chromosomes is real and indicative of some form of microtubule reorganization occurring following monopole breakdown (singular polarity) to mini anaphases (bipolarity). To address the reviewer’s comment, we have softened our language describing this data, removing the word “clearly” from the quoted sentence and changing the text from “some enrichment” to “minor enrichment” when referring to the localization of ASPM-1 to pole-like structures. For our thoughts on assessing GFP::ASPM-1 localization, please revisit point #7.
15) The authors could consider depleting kinetochore Dynein recruiters to dissect the role of different populations. Depletion of RZZ components or SPDL-1 could therefore be a useful comparison in the different backgrounds.
This a good suggestion. However, we are not sure that performing these experiments would provide substantial new insights since two other publications have already examined the role of kinetochore-localized dynein for chromosome segregation (Laband et al., 2017 and Danlasky et al., 2020). Between the two studies, they depleted ROD-1, ZWL-1, and SPDL-1; none of these depletions yielded a change in chromosome segregation rate/distance during anaphase. Therefore, we are not sure what new information we would learn if we performed similar experiments. Notably, Danlasky et al., also utilized Dynein AID and noted that Anaphase B velocity increased from wild-type anaphase spindles. This finding is in strong agreement with our chromosome segregation traces (Figure 4C) and suggests that dynein activity on the spindle, not at kinetochores, has an effect on anaphase spindle elongation.
Reviewer #2:
Cavin-Meza et al., investigated the role of motor proteins and their coordinated action in maintaining spindle bipolarity under acentrosomal conditions. An auxin-inducible degron system was used to rapidly remove proteins from pre-formed spindles in C. elegans oocytes. In addition, this approach was used to disrupt monopoles and study subsequent microtubule re-arrangement for chromosome segregation in formed miniature spindles. Specifically, the authors find that removal of dynein causes a splaying of onopolaral spindle poles. In combination with depletion of KLP-18, removal of dynein causes a disruption of monopolar spindles. This is followed by a formation of locally dispersed small spindles, which show arrangement of microtubules in between the segregated chromosomes and localization of midzone proteins, such as SPD-1. Interestingly, the authors present evidence for the kinesin-5 family motor protein BMK-1 for providing the force for this chromosome movement in these mini spindles.
The major strength of this paper is the application of the degron system to systematically manipulate oocyte spindle assembly after protein depletion. I don’t see any major weakness. This is a very interesting paper with significantly new information extending our current knowledge on acentrosomal spindles. In general, I am very positive about this manuscript. Most of the conclusions of this paper are supported by data, but some points need to be clarified and some aspects of data presentation could be worked out better.
1) In general, the text reads very well, but real numbers always have to be extracted from figures, figure legends and/or tables. The manuscript reads as a text without numbers. For each statement, the measured values, the number of measurements and the statistical data should be clearly provided also in the text. For instance, it is not clear for each presented experiment how many oocytes/samples have been analyzed.
This is a good suggestion and we have updated numerous areas in the Results section to address this point.
2) The authors claim a ‚complete dynein depletion’ (in comparison to partial depletions that have been performed by others) but any quantitative information on the level of depletion is missing. Just showing LM images of oocytes with depleted GFP-dynein is not a convincing argument about a full depletion. This should be clarified.
We completely agree with the reviewer that we would ideally present stronger evidence for the conclusion that we are completely depleting dynein. We therefore tried to perform a western blot to more quantitatively assess the level of DHC-1 depletion. However, we were unsuccessful in our attempts, despite being able to successfully perform westerns on other degron-GFP-tagged strains. We think that this is because dynein is very large when tagged (around 500kD) and this makes it difficult to transfer. Therefore, since we were not able to do this experiment, we have removed all sentences from the manuscript that previously stated that we were completely depleting dynein, to soften our conclusions. We now mostly emphasize that the auxin system is powerful because it allows rapid depletion (rather than emphasizing that it allows “complete” depletion). The only remaining sentence that claims that our depletion may be better than that achieved in previous studies is in the first paragraph of the Results section, where we note that most previous studies used partial RNAi and we state that we are “attempting” to achieve “more complete” depletion. We think that this is fair, since past studies using partial RNAi have intentionally scaled back the amount of depletion (to avoid developmental defects), and we do not have to do that with the auxin system; therefore, it is likely that our depletion is more complete (even if we cannot show, and therefore don’t state, that it is fully complete). We hope this addresses the reviewer’s concern.
3) I think a general control experiment is completely missing here. Have the authors ever analyzed spindles in oocytes with GFP-tubulin and without the auxin-inducible degron system after exposure to auxin? This is an absolutely necessary experiment to do. Hopefully, auxin alone does not cause splayed poles in exposed isolated oocytes. It has to be kept in mind that oocytes (without the fully developed eggshell) are extremely sensitive. Please add data on such control experiments as supplementary information.
We agree that this is a crucial control. When we first started using the degron system in our lab, we performed multiple controls and we generated concrete data showing that there are no side effects of auxin treatment. However, we completely forgot that this was unpublished and that we needed to show these important controls in the current paper – we apologize for this oversight! Please see our new supplemental figure (Figure 2 —figure supplement 2) that thoroughly addresses the impact of Auxin treatment on live oocytes. We observed no significant changes in either spindle length or pole to equator width (a ratio) when comparing untreated and auxin-treated oocytes.
4) There is a lot of ‚jumping’ between Figures 3 and 4 in the text (see paragraph “Dynein acts at spindle poles to promote pole formation and integrity”). The authors might re-think the order of call outs.
We have significantly restructured this section of the manuscript by moving the ASPM-1 and LIN-5 data earlier in the manuscript. We believe this has improved the clarity and general flow of the manuscript (since there is less jumping around), and we hope that the reviewer agrees.
5) Figure 4A-B shows key information. I think the panels are way too small, and I really encourage the authors to present these interesting findings in a better way.
We thank the reviewer for this excellent suggestion. We have restructured Figure 4, making the original Figure 4C its own new Figure (the new Figure 5), to give us room to increase the size of panels 4A and 4B. This also gave us enough room to add zoomed images to 4A, to emphasize key features from the video stills. We think that these changes make it easier for readers to evaluate our findings.
6) As for Figure 5 —figure supplement 1, I have a hard time to see the “labeling persisted on what appeared to be a singular disassembling RC between each segregating chromosome pair”. A clearer image should be presented.
We have switched out the previous image for a new image of SUMO labeling on mini anaphases; we think that the new image better shows the singular RC between separating chromosomes.
7) Line 384, add ‚monopolar’ to read: “resulted in a catastrophic breakdown of the monopolar spindle…”.
This is a good suggestion and we have made this change.
8) Lines 430-438, it is problematic that the authors are discussing the results of the presented manuscript in the context of unpublished data. This unpublished information should be documented somehow within this manuscript or deleted.
We understand the reviewer’s concern. Since the time of our original submission, we have posted a preprint to bioRxiv reporting the findings mentioned in the discussion . Therefore, we were able to add a reference to this preprint, which hopefully addresses this concern.
9) Line 479 and so forth (line 499, 529 etc.), did the authors use several ‚meiosis media’? In case only one buffer was used, change to ‚medium’. In any case, I would give the experimental conditions here and not fully rely on the details given in a reference.
We agree that it should be listed as “medium” so we have made this change. In addition, we have included the full recipe of this Meiosis Medium in the Methods section, as suggested.
10) Lines, 539-40, mention the exact auxin treatment protocols that were used within this study. Other, not applied protocols are irrelevant in the context of this paper.
We thank the reviewer for this suggestion. We have updated this section of the Methods (“Auxin treatment of C. elegans”) to only discuss the exact protocols presented in this study.
Reviewer #3:
Cavin-Meza et al., have parsed out the importance of different motor proteins during oocyte spindle assembly in C. elegans. The authors developed an auxin-inducible dynein degradation system to remove dynein from the spindle. Using this system, they addressed the question of whether dynein is needed to maintain a bipolar spindle. With the depletion of dynein after spindle assembly, the spindles form parallel arrays but still maintain chromosomes in the middle. These results suggested that dynein was involved in pole focusing. To determine further explore the role of dynein in pole focusing, they depleted KLP-18, which normally gives monopolar spindle formation. With depletion of dynein, they find that the monopoles disappear and microtubule bundles form around the chromatin. Intriguingly, the microtubule bundles can allow chromosomes to segregate at anaphase. They then address the question of what motor protein is allowing anaphase onset. They knock out BMK-1 and find that bipolar spindle assembly and chromosome segregation no longer occurs. These results show that BMK-1 contributes to spindle assembly, but somewhat redundantly with other motors. Overall, the experiments were carefully planned and executed. The development of the degradable dynein is a useful tool for the fild. The novel finding that BMK-1 does contribute to spindle assembly is important for understanding the overlapping forces on the spindle. Furthermore, the results shows that there are differences between how the motors function between mitosis and meiosis, which is intriguing and important. Only an additional control experiments is needed.
Overall, the manuscript details an important contribution of how the motor proteins are functioning together for spindle assembly. The manuscript is well-written, with clear rationales. I only have one concern:
1) The aspm-1(RNAi) and lin-5(RNAi) and strains also contain the dynein degron background with dynein tagged. Although the controls show that without auxin, dynein is clearly working, sometimes the degron tags can cause a sensitized background due to a reduction in protein levels or activity. Although the phenotypes for the single depletions have been confirmed with other studies, the aspm-1(RNAi); klp1-18(RNAi) double mutant has not. Can the authors perform the same experiment with a strain without the tagged dynein?
We thank the reviewer for raising this important point. We agree that this would be a better version of the double RNAi experiment to perform. However, in response to comments from other reviewers, we have restructured the presentation of our ASPM-1 and LIN-5 experiments, and we have chosen to remove the double RNAi partial monopolar breakdown experiments from our paper. Thus, all the depletions in this revised manuscript are single depletion phenotypes that are corroborated by previous studies (such as Van der Voet et al., and Laband et al.,). Nevertheless, we appreciate the suggestion and we certainly would have performed these experiments in our TIR1 control strain if we were still including the experiments in question.
Reviewer #4:
Cavin-Meza et al., investigated the role of dynein motors in meiotic spindle pole organization and spindle function using C. elegans oocytes as a model. They optimized the auxin-inducible degron system to degrade endogenous AID-tagged dynein acutely, rapidly, and more completely in oocytes. They find that dynein is required for integrity of bipolar spindle poles as well as spindle monopoles induced by depletion of the kinesin-12 motor KLP-18. They further show that in the absence of dynein motors, and functional spindle poles, microtubules are still able to organize into spindle-like structures via the action of chromosomes and the kinesin-5 motor BMK-1. They also found that chromosomes can still be segregated in anaphase-like motion in these oocytes.
These results are consistent with many previous studies in mitotic and meiotic cells that investigated the function of dynein on spindle poles and chromosome segregation and showed spindle pole focusing function of this motor. Furthermore, it is known that upon removal of acentrosomal spindle pole components that assemble meiotic spindles in mouse oocytes, chromatin-mediated microtubule assembly can build spindles that can segregate chromosomes. This study is thus generally consistent with data from other models.
The authors have optimized the AID-degron system beautifully and this gives them as well as the field new opportunities to study protein function in C. elegans oocytes in a much cleaner and more complete depletion background. The timescale in which they can deplete endogenous proteins is impressive.
1. The authors previously published a paper where they found dynein inhibition causes chromosome segregation errors and that in this organism oocyte chromosomes are segregated independently of kinetochores in a pathway that involves dynein. While this is not in agreement with a more accepted model where inter-chromosomal microtubule arrays push chromosomes apart, it had so far provided an alternative model for oocyte chromosome segregation in the field. In this current manuscript, the authors show that dynein inhibition, presumably more complete and cleaner than other methods, in fact does not lead to any chromosome segregation errors. This is in direct conflict with their previously published work, but this major discrepancy is not addressed in the current work.
In response to this and a similar comment by Reviewer 1, we have expanded our Discussion to address this issue. The data presented in the current manuscript do not conflict with any of the data in our 2015 paper. Rather, new information has emerged in the years since we published our original paper that has changed our interpretation of the data in our 2015 paper, and has led to an updated view of chromosome segregation. Please see our response to comment #12 from Reviewer 1 for additional explanation.
2. The authors base their hypothesis on a previously published study in human oocytes reporting that spindles fail to maintain bipolarity and mis-segregate chromosomes. This work was performed on oocytes that were unhealthy and unsuitable for IVF treatment, and it is now clear that such atretic oocytes do not represent a general oocyte form. While the maternal age effect is indeed of clinical interest, given the strong interest in many areas of cell biology in understanding the fundamental principles of chromosome segregation in oocytes (of many species), it is not necessary to base experiments on observations made using human oocytes.
We did not intend for our introduction to give the impression that we are solely using Holubcova et al., 2015 as a basis for doing the experimentation in this paper; while we still think it is worthwhile to cite this study, we have updated the wording of the Introduction so it does not sound like this was the sole motivation for our study. We completely agree with this reviewer that the pursuit of fundamental knowledge about acentrosomal spindles and how they facilitate proper chromosome segregation is intriguing and motivating in of itself.
3. The contribution of chromatin-mediated mitotic and meiotic spindle assembly pathways was largely ignored in the discussion although these are likely to majorly contribute to miniature spindle assembly in dynein AID oocytes. Similarly, the contribution of dynein to spindle pole focusing in other systems needs more coverage.
We thank the reviewer for these great suggestions. We agree that chromatin-mediated microtubule nucleation may play a role in the formation of the miniature anaphase spindles. We have now added text to the Discussion to raise this interesting idea and we have cited several reviews on this topic. We also appreciate the suggestion to add information about dynein’s role in spindle pole focusing in other systems. Most of our references involving pole focusing driven by dynein were originally centered in mitosis, so we have added some additional background on dynein’s roles in acentrosomal spindle assembly to our Introduction, citing studies using Xenopus egg extract (Heald et al., 1996), Xenopus oocytes (Becker et al., 2003), and mouse oocytes (Luksza et al., 2013, Clift and Schuh 2015).
4. While the images are of high quality and very accessible, the lack of appropriate image analysis and quantification means the manuscript is largely descriptive. It would be important to support several conclusions made throughout the paper with robust quantification and statistical analyses.
We have addressed this concern in a number of ways (many of these changes are also detailed in our previous responses to the other reviewers, who made specific suggestions along these lines). To better support our conclusions, we have made efforts to provide more rigorous quantification of numerous aspects of our data. First, we performed new pole splaying quantifications. Instead of judging whether a pole was “focused” or “splayed” by eye (as we did in the original submission), we measured the widths of the poles and of the center of the spindle; we then calculated a ratio of the pole to equator width as a measurement of how splayed the poles are (this is shown in Figure 2B). This new analysis confirmed the finding that we reported in our original manuscript – that the poles are substantially splayed upon auxin treatment. Additionally, we have added rigorous controls to demonstrate that the effects we see are not caused by auxin treatment (Figure 2 —figure supplement 2). We dissected oocytes lacking degron-tagged dynein into auxin and then imaged spindles over time, demonstrating that neither the spindle length nor the pole-to-equator ratio changed. Finally, we now report the number of images analyzed for all figures, we have added quantitative information into the manuscript when discussing results, and we have tried to emphasize the more quantitative aspects of our paper in our revised manuscript, such as our measurements of chromosome segregation rates and distance traces in live imaging.
5. While the writing is quite accessible to readers working on the C. elegans model, it is much less clear for those who work on other models. It would be very helpful to re-write technical bits such as 'embryos dissected into auxin'. It is also important to include citations from other models when summarizing what is known about certain spindle features in oocytes. There is generally ambiguity in the writing that leaves the reader wondering about the precise point that is being made. E.g. microtubule bundles were able to remain relatively aligned' – this and several other statements in the paper are very descriptive and without quantitative image analysis to back them up, it is difficult to grasp what they mean.
We definitely want to make the manuscript accessible to readers working on different models so we appreciate this feedback. In response, we have tried to edit some of the worm-specific language to make it less confusing, we have added a schematic to Figure 1A to better explain our auxin treatment methods (“long-term”, “short-term”, and “acute”), and we have changed the figures and text to be consistent when referring to these methods. We have also gone through the manuscript and have tried to remove ambiguity (e.g. removing phrases such as “relatively aligned” and replacing them with more precise descriptions). We think that these changes will help understanding. In addition, we understand the desire for additional references detailing work in other systems. We were originally very brief in our introduction in an effort to be succinct. However, in response to this concern, we have attempted to elaborate further on important points (as mentioned in comment #3) and we hope that this reviewer will now feel that we have provided adequate information for all readers to find the results of this study clear and accessible.
This manuscript can be significantly improved by providing a detailed discussion concerning the conflict between the data within and the lab's previously published work that shows dynein inhibition causes chromosome segregation errors. Furthermore, robust image quantification and statistical analyses are needed to make the paper much less descriptive cell biology than it is now. It would also be good to include some controls to show that 4 hour treatment of non-AID strain wild-type oocytes has no effect on several aspects of spindle assembly and function studied here.
We have now included a discussion of our previous work (see the response to comment #1 above) and more quantitative analysis (see the response to comment #4 above). We thank the reviewer for the excellent suggestion of an additional control experiment, and we agree that it is important to show that long-term AID (4 hours on auxin-containing plates) does not have adverse effects on acentrosomal spindle assembly. We have now included a new supplemental figure (Figure 1 —figure supplement 1) reporting this analysis. We repeated our germline counting to assess proper spindle formation in a strain that contains all the same fluorophores and transgenes, but that is lacking the degron::GFP tag on DHC-1. We found no discernable difference between untreated and treated oocytes across a large sample size.
https://doi.org/10.7554/eLife.72872.sa2Article and author information
Author details
Funding
National Institute of General Medical Sciences (R01GM124354)
- Sarah M Wignall
National Cancer Institute (T32CA009560)
- Gabriel Cavin-Meza
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Acknowledgements
We thank all members of the Wignall lab and the WiLa ICB for support and discussion, and Emily Czajkowski, Nikita Divekar, Hannah Horton, Juhi Narula, and Ian Wolff for critical reading of the manuscript. Additionally, we thank Karlin Compton for their help in piloting certain experimental conditions. We thank SunyBiotech for their service in generating PHX3914 for live imaging experiments. Finally, we thank Liangyu Zhang and the Dernburg Lab for providing CA1215 (the original dhc-1::degron::GFP worm strain) and for corresponding with us to provide advice, and Arshad Desai, Andy Golden, Federico Pelisch, and Jill Schumacher for antibodies. Some strains were provided by the Caenorhabditis Genetics Center (CGC), which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). This work was funded by NIH R01GM124354 (to SMW) and by NIH/NCI training grant T32 CA009560 (to GCM).
Senior Editor
- Jessica K Tyler, Weill Cornell Medicine, United States
Reviewing Editor
- Federico Pelisch, University of Dundee, United Kingdom
Reviewers
- Federico Pelisch, University of Dundee, United Kingdom
- Thomas Müller-Reichert, Technische Universität Dresden, Germany
Version history
- Received: August 6, 2021
- Preprint posted: September 9, 2021 (view preprint)
- Accepted: February 10, 2022
- Accepted Manuscript published: February 11, 2022 (version 1)
- Accepted Manuscript updated: February 17, 2022 (version 2)
- Version of Record published: March 29, 2022 (version 3)
Copyright
© 2022, Cavin-Meza et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.
Metrics
-
- 1,522
- Page views
-
- 229
- Downloads
-
- 6
- Citations
Article citation count generated by polling the highest count across the following sources: PubMed Central, Crossref, Scopus.
Download links
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Further reading
-
- Biochemistry and Chemical Biology
- Cell Biology
The ATPase p97 (also known as VCP, Cdc48) has crucial functions in a variety of important cellular processes such as protein quality control, organellar homeostasis, and DNA damage repair, and its de-regulation is linked to neuromuscular diseases and cancer. p97 is tightly controlled by numerous regulatory cofactors, but the full range and function of the p97–cofactor network is unknown. Here, we identify the hitherto uncharacterized FAM104 proteins as a conserved family of p97 interactors. The two human family members VCP nuclear cofactor family member 1 and 2 (VCF1/2) bind p97 directly via a novel, alpha-helical motif and associate with p97-UFD1-NPL4 and p97-UBXN2B complexes in cells. VCF1/2 localize to the nucleus and promote the nuclear import of p97. Loss of VCF1/2 results in reduced nuclear p97 levels, slow growth, and hypersensitivity to chemical inhibition of p97 in the absence and presence of DNA damage, suggesting that FAM104 proteins are critical regulators of nuclear p97 functions.
-
- Cell Biology
- Neuroscience
The amyloid beta (Aβ) plaques found in Alzheimer’s disease (AD) patients’ brains contain collagens and are embedded extracellularly. Several collagens have been proposed to influence Aβ aggregate formation, yet their role in clearance is unknown. To investigate the potential role of collagens in forming and clearance of extracellular aggregates in vivo, we created a transgenic Caenorhabditis elegans strain that expresses and secretes human Aβ1-42. This secreted Aβ forms aggregates in two distinct places within the extracellular matrix. In a screen for extracellular human Aβ aggregation regulators, we identified different collagens to ameliorate or potentiate Aβ aggregation. We show that a disintegrin and metalloprotease a disintegrin and metalloprotease 2 (ADM-2), an ortholog of ADAM9, reduces the load of extracellular Aβ aggregates. ADM-2 is required and sufficient to remove the extracellular Aβ aggregates. Thus, we provide in vivo evidence of collagens essential for aggregate formation and metalloprotease participating in extracellular Aβ aggregate removal.