Article
Open access
Published: 25 June 2024

Shape of the first mitotic spindles impacts multinucleation in human embryos

Nature Communications volume 15, Article number: 5381 (2024) Cite this article

1484 Accesses
8 Altmetric
Metrics details

Subjects

Abstract

During human embryonic development, early cleavage-stage embryos are more susceptible to errors. Studies have shown that many problems occur during the first mitosis, such as direct cleavage, chromosome segregation errors, and multinucleation. However, the mechanisms whereby these errors occur during the first mitosis in human embryos remain unknown. To clarify this aspect, in the present study, we image discarded living human two-pronuclear stage zygotes using fluorescent labeling and confocal microscopy without microinjection of DNA or mRNA and investigate the association between spindle shape and nuclear abnormality during the first mitosis. We observe that the first mitotic spindles vary, and low-aspect-ratio-shaped spindles tend to lead to the formation of multiple nuclei at the 2-cell stage. Moreover, we observe defocusing poles in many of the first mitotic spindles, which are strongly associated with multinucleation. Additionally, we show that differences in the positions of the centrosomes cause spindle abnormality in the first mitosis. Furthermore, many multinuclei are modified to form mononuclei after the second mitosis because the occurrence of pole defocusing is firmly reduced. Our study will contribute markedly to research on the occurrence of mitotic errors during the early cleavage of human embryos.

The first mitotic division of human embryos is highly error prone

Article Open access 08 November 2022

Ran-GTP assembles a specialized spindle structure for accurate chromosome segregation in medaka early embryos

Article Open access 01 February 2024

Chromosome segregation error during early cleavage in mouse pre-implantation embryo does not necessarily cause developmental failure after blastocyst stage

Article Open access 21 January 2020

Introduction

During human embryonic development, many errors, such as chromosome mis-segregation, could occur that lead to aneuploidy^1,2,3, which originates during early mitosis^1,4,5,6. The first mitosis is different from the subsequent mitosis because it includes the fusion of paternal and maternal genomes. The parental genome separates each pronuclear envelope, and the separated genome is fused and arrayed in the spindle after pronuclear envelope breakdown. An accurate first mitosis is essential to prevent genetic defects; however, studies have shown that several errors could occur during the first mitosis, such as direct cleavage, chromosome segregation errors, and multinucleation^6,7,8. A live-cell imaging study using time-lapse confocal microscopy showed that the frequency of multinucleation at the 2-cell stage was 78% in human embryos⁹. Although multinucleated embryos transferring could lead to birth of babies, embryos with multiple nuclei at the 2-cell stage have lower blastocyst developmental rate and live birth rates than those with normal nuclei¹⁰; thus, nuclear errors can be used as a deselection biomarker for embryo transfer^{9,11,12,13,14}.

Many studies have revealed the abnormalities of DNA and cytoskeleton in the early cleavage stage of human embryos using fixed and immunofluorescence methods. Spindle abnormalities, such as multipolar spindles, are often observed in human zygotes^15,16. Observation of spindle and nuclear abnormalities on days 3 and 5 in fixed human embryos showed that the frequency of abnormalities was higher in cleavage-stage embryos than in blastocyst-stage embryos¹⁷. However, the mechanism by which nuclear and spindle abnormalities occur remains largely unknown. Therefore, live-cell imaging of early cleavage-stage embryos is required for further investigation.

Microinjections of DNA or mRNA into mammalian embryos were widely used for live imaging during early mitosis^{18,19,20,21,22}. Mouse embryos have dual spindles during the first mitosis, and the parental chromosomes remain apart until the second mitosis²³. However, there are differences between mouse and human zygotes; therefore, the results obtained from studies on mouse zygotes do not always agree with those of studies on human zygotes. For example, only 5% of mouse embryos have abnormal nuclei at the first cell division²⁴, and centrioles of sperm are not inherited because sperm centrioles degenerate during spermiogenesis²⁵. In mouse zygotes, acentriolar cytoplasmic microtubule-organizing centers are formed during the first divisions^23,26. Moreover, microinjections of DNA or mRNA into human embryos pose the risk of ethical issues²⁷. In addition, most human embryos used in research are donated surplus embryos by patients treated at fertility clinics. Surplus human embryos are mostly blastocysts, and it is difficult to obtain a sufficient number of human zygotes. Therefore, these factors hinder the progress of cell biology research on early human embryonic development.

Some recent studies show live imaging of human zygotes. These studies showed that chromosome mis-segregation often occurs during the first mitosis stage in human embryos¹⁶ and inappropriate pronuclei migration and chromosome clustering at the interface can cause lagging chromosomes leading to chromosome loss²². Moreover, a study using live imaging without genetic manipulation for human blastocysts showed that blastocyst mechanical stress from blastocyst expansion or biopsy triggers nuclear DNA loss²⁸. These studies on live-cell imaging of human embryos can reveal the mechanisms by which errors frequently occur in human embryos.

The present study focused on the movement of chromosomes and spindles during the first mitosis; we aimed to establish a method used for real-time imaging of DNA and microtubules in human two-pronuclear stage zygotes using chemical fluorescent labeling. We discovered that spindle shape was associated with blastomere nucleation status, spindle pole defocusing caused multinucleation, and differences in the positions of centrosomes cause spindle abnormality during the first mitosis in human embryos. Furthermore, we observed that many multinuclei are modified to form mononuclei after the second mitosis because pole defocusing was firmly reduced. Our study contributes to the identification of the causes of nuclear errors and developmental processes in human embryos.

Results

Live-cell imaging during the first mitosis in human embryos

We imaged human zygotes to observe the dynamics of DNA and microtubules during first mitosis. Figure 1a and Supplementary Movie 1 show representative confocal microscopy images of human embryos using fluorescent labeling. SPY-505DNA visualized DNA in the parental pronuclei, and DNA condensed into chromosomes before pronuclear envelope breakdown (PNBD) (Fig. 1a, SPY505-DNA (i)). Before the chromosomes were aligned, the spindle poles were visualized using SPY650-Tubulin (Fig. 1a, SPY650-Tubulin (ii)). In all imaged zygotes, the sperm tail was attached to the spindle poles (Fig. 1a, SPY650-Tubulin (iii), white arrowhead). When chromosomes segregate, the spindle elongates and forms the midbody (Fig. 1a, SPY650-Tubulin (iv), (v), Merge (vi)). Many granules labeled with SPY-505DNA and SPY-650 tubulin were imaged in all embryos and were observed under bright field (Fig. 1a, Bright field, white arrow). Cytoplasmic granules are slightly ovoid organelles formed in oocytes and zygotes^29,30, and a previous study showed that human embryos store proteins such as chromatin modification factor and tubulin in cytoplasmic lattices³¹. Figure 1b shows multipolar segregation (Supplementary Movie 2). A multipolar spindle was formed from three poles (Fig. 1b (ii) white arrow). The chromosomes did not array linearly but had a ‘Y’ shape and were segregated in multipolar (Fig. 1b (iii), (iv)).

**Fig. 1: Live-cell imaging using fluorescent labeling during the first mitosis in human embryos.**

To ensure that this method did not inhibit the dynamics of DNA and tubulin during the first mitosis, we compared the time from PNBD to furrow ingression between embryos of a control, an imaged, and a dyed group (Fig. 1c, Supplementary Fig. 1). Embryos in a control group were incubated in medium not containing chemical dye like SPY-505DNA and SPY-650tubulin; those in an imaged group were incubated in those with dye and imaged using a confocal microscopy; and those in a dyed group were incubated in medium with SPY dyes but not imaged. The mean times for the control, imaged, and dyed groups were 149.2, 155.6, and 155.1 min, respectively. No significant differences were observed between the groups, suggesting that visualizing chromosomes and tubulin using our method does not significantly disturb the progression of the first mitosis in human embryos.

Figure 1d shows the frequency of anaphase errors during the first mitosis. We observed five embryos undergoing multipolar segregation, and only one of them showed direct cleavage, a type of abnormal cleavage wherein one embryo divides into three or more blastomeres (Fig. 1d, second bar).

Figure 1e shows the nuclear status at the 2-cell stages. The mononuclear type has one large nucleus, and if there are other nuclei, their diameters are usually smaller than 10 µm. The multinuclear type does not have one large nucleus but has nuclei of various sizes in either or both of the cells in the 2-cell stage. Figure 1f shows the frequency of the nuclear status at the 2-cell stage. The frequency of the multinuclear type was higher than that of the mononuclear type (81% (n = 25/31), 19% (n = 6/31)). All embryos undergoing multipolar segregation demonstrated a multinuclear status (Fig. 1f, third bar). The rate of multinucleation for the control and dyed groups were 75% (15/20) and 78% (25/32), respectively (Fig. 1g). No significant differences were observed between the groups; the frequency of the multinuclear type was 81% (n = 25/31) in an imaged group, suggesting that the chemical dye and laser scanning used in our method does not significantly cause multinucleation at the two-cell stage in human embryos. Therefore, these suggests that anaphase and nuclear errors often occur in 2-cell stage human embryos.

Spindle shapes are firmly associated with multinucleation

Various types of spindles were observed in this study (Supplementary Figs. 2–4). We analyzed spindles at the end of metaphase from an angle where the chromosomes appear to be aligned linearly and where the spindle appears to have the maximum area and calculated the aspect ratio of the spindles (Fig. 2a, b (i)–(iii)). Figures 1a and 2a show representative images of high- and low-aspect-ratio (AR) spindles, respectively. The high-AR spindle was sharp, elongated, had focused poles, and led to a mononuclear type at the 2-cell stage (Fig. 2a (iv)–(vi), Supplementary Movie 3). The low-AR spindle appeared as dual spindles with defocused poles and chromosomes that could not unify during anaphase and became multinucleated (Fig. 2b (iv)–(vi), Supplementary Movie 4). Figure 2c showed that there was a significant difference between the AR of the spindles leading to the mononuclear and multinuclear types at the 2-cell stage. The results, including only bipolar segregation, did not change (Fig. 2d). Figure 2e shows the AR of all imaged spindles (n = 31). These results suggest that nuclear formation in 2-cell-stage embryos is strongly associated with spindle shape.

Spindle pole defocusing causes multinucleation

Figure 3a shows a spindle with unilaterally defocused poles (Supplementary Movie 5). This spindle started to form from two centers of cytoplasmic microtubule nucleation and formed two polar spindles (Fig. 3a (i) and (ii)), but the unilateral pole became defocused in the last half of metaphase (Fig. 3a (iii), white arrow), leading to the development of a multinuclear-type 2-cell embryo. Figure 3b shows a spindle with bilateral defocused poles (Supplementary Movie 6). This spindle also started to form from two spindle poles; a sperm tail was attached at the center of the pole (Fig. 3b (i) and (ii), white arrow), but the bilateral pole became defocused, and the sperm tail moved to the edge of the defocused pole in the last half of metaphase (Fig. 3b (iii), white arrow), making it a multinuclear-type 2-cell embryo. Figure 3c shows the significant difference in the AR of the spindles between focusing and defocusing poles in an embryo undergoing bipolar segregation. We often observed polar defocusing in the first mitotic spindle during metaphase (42%, n = 11/26) in the embryo undergoing bipolar segregation (Fig. 3d). Moreover, the frequency of defocused poles was significantly lower in mononuclear-type embryos than in multipolar-type embryos (0% (n = 0/6), 55% (n = 11/20, P < 0.05)) (Fig. 3d). Figure 3e shows the frequency of the spindles with defocused poles in an embryo undergoing bipolar segregation. Among 11 spindles with defocused poles, the spindles with unilateral defocused poles and bilateral defocused poles were observed at similar frequencies (45% (n = 6/11), 54% (n = 5/11)), and the defocused pole side was more often the pole not attached to the sperm tail (n = 4/6). Subsequently, we investigated spindles leading multinuclear type and having focused poles, not defocused poles (Fig. 3d, first bar, n = 9). Figure 3f shows that even excluding embryos that exhibited defocused poles, there was a significant difference between the AR of the spindle leading to mononucleated and multinucleated types. These results indicated that low AR spindles often induce multinucleation regardless of whether they have focusing or defocusing poles. In addition, poles of high-AR spindles rarely become defocused, whereas those of low-AR spindles often become those with defocused poles., and this leads to multinucleation at the 2-cell stage.

Multipolar spindles and defocused spindles differ

Low-AR spindles including defocused spindles and multipolar spindles cause multinucleation at the two-cell stage (Figs. 1–3). Both spindles are similar in that they do not have two focused poles. We investigated the formation process to examine their differences. Spindle formation began 10 min (IQR = 10 min) before PNBD and it took 30 min (IQR = 10.8 min) for the embryo to complete spindle formation. Moreover, the duration of spindle elongation from PNBD was 130 min (IQR = 11.0 min). A blue dashed line in Fig. 4a indicates the onset of spindle poles defocusing. Spindles with defocused poles were formed from two poles and became bipolar spindles with two focused poles during the first half of metaphase. The onset of spindle pole defocusing occurred 80 min (IQR = 19.0 min) from PNBD. Spindle morphology changed from two poles to defocused poles in the last half of the metaphase. Meanwhile, multipolar spindles with multipolar segregation were formed from three poles and appeared much earlier than those with defocused poles (Fig. 4a, black dashed line at 15.0 min (IQR = 12.5 min) vs 94.5 mins (IQR = 32.75 min), from the beginning of spindle formation). There were no significant differences in the duration of phases in the first stage of mitosis between the spindle types and nuclear type (Tables S1–2). Next, we investigated the position of centrosomes in different morphologic spindles using immunofluorescence. We used γ-tubulin to determine where centrosomes exist in spindle poles. γ-tubulin is an essential centrosome protein that is required for the formation of nucleation sites at MTOCs^32,33. We imaged human zygotes using the same method, and fixed embryo approximately one hour after the onset of PNBD. High-aspect ratio spindle with focused poles had two γ-tubulin positive poles, and the two poles lined up on a line perpendicular to the chromosomes (Fig. 4b (i)). On the other hand, spindles with defocused poles had two γ-tubulin positive poles, but the position was offset from the line perpendicular to the chromosomes (Fig. 4b (ii)). Multipolar spindles appeared to have three poles, but actually only have two γ-tubulin positive poles and the other pole was a γ-tubulin negative pole (Fig. 4b (iii). Therefore, spindles with defocused poles are significantly different from multipolar spindles owing to the formation process and the location of the centrosome. These results show that multipolar spindles and defocused spindles causing multinucleation have a different formation process and the position of centrosomes are different.

**Fig. 4: Multipolar spindles and defocused spindles do not have two focused poles and cause multinucleation, although the formation processes differ.**

Stable second mitotic spindles modify multinuclei

Next, we compared the frequency of the nuclear status between the 2- and 4-cell stages. To reduce phototoxicity to embryos, we limited the duration of confocal microscopic observation to 8.4 ± 3.1 [standard deviation (SD)] h. We imaged the embryos consecutively during the first mitosis, after which the imaged embryos were moved into a time-lapse incubator. After the second division, the embryos were imaged using a confocal microscope (Supplementary Fig. 1). The frequency of the multinuclear type was 81% and 35% at the 2- and 4-cell stages, respectively (Fig. 5a).

We measured the diameter of nuclei in each blastomere in the 2- and 4-cell stage embryos (Fig. 5b) and investigated how the nuclear diameter changed after the second mitoses. Figure 5c, d shows the same embryos and the nuclear diameters of all embryos that were imaged during the first mitosis and after the second mitosis (n = 11). There was a wide range of variations in plots of the diameter at the 2-cell stages, while nuclear diameters of approximately 20 µm were highly frequent at the 4-cell stage. The pink bar under 10 µm shows micronuclei caused by a lagging chromosome during the first mitosis. The lagging chromosome occurring during anaphase did not merge with the mono nucleus becoming a micro nucleus (Fig. S5a, Supplementary movies 7 and 8). There were no significant differences in spindle pole status, nuclear type, and spindle AR whether the embryos had lagging chromosomes or not (Figure S5b–d). Therefore, it is suggested that lagging chromosomes occur independent of spindle shape.

Figure 5c, d shows that medium-sized nuclei (7–18 µm) of the 2-cell stage mostly disappeared and small-sized nuclei (under 10 µm) were left at the 4-cell stage. To investigate how the phenomenon occur, we also imaged human embryos during the second mitosis (n = 9). Figure 5e–g shows how often the second mitotic mis-segregation and spindles defocusing occurs. Interestingly, no spindle pole defocusing was observed in the second mitosis and the occurrence of spindle poles defocusing during the second mitosis was significantly lower than those during the first mitosis (Fig. 5f, 42% (11/26) vs. 0% (n = 0/7)). The occurrence of multipolar segregation and a lagging chromosome was not significantly different between the first mitosis and the second mitosis (Fig. 5e, 16% (n = 5/31) vs. 22% (n = 2/9), Fig. 5g, 16% (5/31) vs 19% (n = 1/9)). Figure 5h shows how the multinuclear type at the two-cell stage becomes mononuclear during the second mitosis (Supplementary Movie 9). The chromatin of each nucleus in the multinucleated cell condensed and formed chromosomes (Fig. 5h (iii)), and the chromosomes gathered into a second mitotic spindle (Fig. 5h (v)) and segregated into two separate areas (Fig. 5h (viii)). These results suggest that the second mitotic spindles were stable, and most of the multinuclei gathered and arrayed in the spindle, leading to the formation of mononuclei.

Discussion

The purpose of the present study was to image live human embryos using fluorescent labeling, without using the conventional microinjection strategy, for studying the mechanisms underlying the occurrence of nuclear abnormalities during the first mitosis. We used the fluorescent labels SPY505-DNA and SPY650-Tubulin and established a method to image the dynamics of DNA and microtubules during first mitosis. We demonstrated that there was variation in the first mitotic spindles and that low-AR spindles were much more likely to lead to multinucleation at 2-cell stage (Fig. 2c, d). In addition, the low-AR spindles were unstable and often had defocused poles, leading to multinucleation at the 2-cell stage (Fig. 3c, d). Moreover, we also discovered that there was difference in centrosome position on the poles between each spindles (Fig. 4b). Figure 6 shows the model of spindle instability causing multinucleation. In addition, we discovered the reduction in frequency of the multinuclear type was 81% and 35% from the 2-cell to 4-cell stages (Fig. 5a). The occurrence of spindle poles defocusing during the second mitosis was significantly lower than those during the first mitosis (Fig. 5f, 42% (11/26) vs. 0% (n = 0/7)), which may contribute to the stability of the second spindle (Fig. 5).

**Fig. 6: Model of how spindle instability causes multinucleation during the first mitosis.**

Our study showed that the first mitotic spindles varied and were strongly associated with multinucleation. Spindle instability rarely occurs in normal somatic cells³⁴. Conventionally, mouse embryos are used as a substitute for human embryos to study the dynamics of DNA and the cytoskeleton during the cleavage stage. However, 80% of the first mitotic spindles in mouse embryos are barrel-shaped³⁵ and rarely lead to nuclear errors at the 2-cell stage (<5%)²⁴. In bovine embryos, which show over 60% multinucleation at the 2-cell stage, a study showed that the first mitotic spindles that were formed after juxtaposed pronuclear envelope breakdown were stable, and over 80% of the spindles were normal spindle³⁶. Spindle instability similar to the abnormal spindles in our study was observed in a study on human oocytes^37,38, and this meiotic spindle instability caused segregation errors in meiosis and led to aneuploidy in human eggs. The same study identified a molecular motor KIFC1 (kinesin superfamily protein C1) as a spindle-stabilizing protein; unlike mouse and bovine oocytes, human oocytes lack this protein³⁸. Interestingly, Human zygotes exhibit low levels of KIFC1 mRNA³⁸. It is possible that the first mitotic spindle instability in human embryos is inherited from oocytes.

In our study, we showed that there are two kinds of spindle that induce multinucleation at the two-cell stage: low-AR spindles which often have defocused poles and multipolar spindles. They were divided depend on the formation time. Although low-AR spindles with defocused poles were formed from two poles and became bipolar spindles with two focused poles during the first half of metaphase, the spindle morphology changed from two poles to defocused poles in the last half of metaphase. Meanwhile, multipolar spindles with multipolar segregation were formed from three poles and appeared much earlier than those with defocused poles (Fig. 4a, black dashed line, 15.0 min (IQR = 12.5 min) vs 94.5 min (IQR = 32.75 min), from the beginning of spindle formation). We also observed the sperm tail moving toward the edge of the defocused pole from the center during the last half of metaphase (Fig. 3b(ii), (iii) blue arrowhead, Supplementary Movie 6). A previous study showed that a sperm had a typical proximal centriole and atypical distal centriole, and the distal centriole exists at the edge of the sperm tail and is inherited by zygotes after fertilization³⁹. Furthermore, sperm centrioles devoid of pericentriolar matrix (PCM) recruit maternal PCM proteins upon reaching the egg during fertilization, forming a functional centrosome for embryo development In most species⁴⁰. In bovine zygotes, the centrosomes did not function as centers of cytoplasmic microtubule nucleation but are only loosely connected to spindles³⁶. However, it remains unknown how inherited sperm centrosome and maternal PCM form centrosomes during the first mitosis in humans. It is possible that the movements of the defocused poles and sperm tail indicate that the human zygote tends to have dysfunction of the centrosome inherited from the sperm during the first mitosis.

Moreover, the two kinds of spindles had different positions of γ-tubulin positive MTOCs, respectively. Multipolar spindles had three spindle poles but had only two γ-tubulin positive poles and the other pole was γ-tubulin negative (Fig. 4b(iii)). This result means that human zygotes can make MTOCs without centrosomes inherited from sperm like human oocytes. Meanwhile, spindles with defocused poles had two γ-tubulin positive poles, but the position was offset from the line perpendicular to the chromosomes (Fig. 4b(ii)). The long term of mitotic division may cause defocusing poles, because pole defocusing occurs during the last half of metaphase. The duration from PNBD to anaphase during the first mitosis is significantly longer than the second mitosis in human embryos¹⁶. Therefore, the centrosomes consisted of inherited sperm centriole, and maternal PCM may not be functionally maintained until the anaphase as MTOC. In fact, there were significantly reduced rates of poles defocusing from the first to second mitosis in our study. In other words, spindle stability may change as cell division progresses. Further investigation is necessary to determine whether improvements in spindle instability occur after the second mitosis.

This study has several limitations. First, human embryos were donated from a fertility clinic, and detailed patient information, such as underlying disease and method of ovarian stimulation, was not available. It is possible that the human zygotes used in this study had a bias in patient characteristics. Second, although we reported several phenomena like varying first mitotic spindles and defocused poles, we could not uncover all the underlying mechanism. Our study suggested that spindles with poles defocusing and multipolar spindles had different positions of γ-tubulin positive MTOCs compared to high-AR spindles. However, we did not reveal why low-AR spindles with focused poles tend to form multinuclei compared to high-AR spindles. In our study, four γ-tubulin spots during anaphase were observed (Fig. S6, whiteheads). This suggested that zygotic centrosomes are unstable and tend to divide or form other MTOC during anaphase even if spindles have focused poles during metaphase. These phenomena were not previously reported in other mammalian zygotes and may be specific to humans. To clarify the details, many human zygotes and further experiments are necessary.

Figure 6 shows the multinucleation model. Our study showed that there was heterogeneity in the first mitotic spindle, which was strongly associated with multinucleation. High-AR spindles were stable and led to a mononuclear type. On the other hand, low-AR spindles often lead to multinucleation. In low-AR spindles, the poles often defocused during the last half of metaphase. Many multinuclei are modified to form mononuclei after the second mitosis since pole defocusing was firmly reduced. Thus, spindle stability is important for nuclear formation. Further investigations are necessary to uncover the mechanism underlying these phenomena. Revealing the mechanism underlying spindle formation in cleavage-stage embryos will contribute to improving the success rate of assisted reproduction technologies in humans.

Methods

Ethical approval

This study was approved by the Ethics Committee of Akita University (approval number: 1090-2), Japan Society of Obstetrics and Gynecology (Registry No. 75), and Asada Ladies Clinic (approval number: 2022-13). Participation in this study was entirely voluntary and no financial inducements were given for embryo donation. We obtained written informed consent from patients who donated human embryos. Donated embryos were determined as “Not Human Subjects Research.” The patients were informed of the conditions of the donation, objectives, and methodology of human embryo research. They were offered counselling and alternative options, including discarding embryos and continued cryopreservation. Embryo donors were also informed that their donation would not affect their IVF cycle. The experiments were performed in accordance with the relevant regulations on human sperm/ovum/fertilized eggs set forth by the Japan Society of Obstetrics and Gynecology.

Human embryos

Frozen human two-pronuclear stage embryos were donated for research by couples who had completed fertility treatment. The mean age of patients at embryo freezing was 32.00 ± 3.03 (SD) years. All surplus embryos were fertilized by IVF or ICSI and frozen on day 1, 19.7 h ± 0.92 h (SD) after insemination. All samples were de-identified after freezing. All human embryos were thawed using Cryotop Safety Kit (Kitazato Corporation, Shizuoka, Japan) according to the manufacturer’s protocol and cultured for 30 min in HiGROW OVIT Plus (Fuso Pharmaceutical Industries, Ltd., Osaka, Japan). Medium was covered with mineral oil in an incubator at 37 °C, with 5% O₂, 5% CO₂, and 90% N₂.

Reagents for the detection of DNA and tubulin

Thirty-one human embryos were cultured in CultureCoin (Esco Medical, Egaa, Denmark) containing the live-cell stain probes SPY505-DNA (1:1000, Spirochrome, Thurgau, Switzerland) and SPY650-Tubulin (1:2000, Spirochrome) diluted in prewarmed HiGROW OVIT Plus (Fuso Pharmaceutical Industries) in a time-lapse incubator (Miri® Time-Lapse incubator; Esco Medical) at 37 °C, with 5% O₂, 6% CO₂, and 89% N₂. The embryos were captured every 5 min for at least 2 h to ensure that they were well dyed.

Live imaging

Thirty-one human embryos were imaged during the first mitosis. After the cytoplasmic halo in human zygotes disappears, PNBD occurs⁴¹. Thus, immediately after the cytoplasmic halo started to disappear, the embryos were transferred to a 20-μL drop of observation culture medium that was placed in a 35-mm glass-bottom dish (Matsunami Glass Industry, Osaka, Japan) under mineral oil (Supplementary Fig. 1). Human embryos were imaged using an LSM980 laser scanning confocal microscope (Zeiss Japan, Tokyo, Japan) equipped with an objective C-Apochromat 40x/NA 1.20 W Korr and controlled by the ZEN Blue software (Zeiss). Images were acquired at a resolution of 0.414 × 0.414 μm per pixel, every 10–20 min (60 × 1.25 µm optical sections) for a period of 8.4 ± 3.1 (SD) h and a depth of 16 bits. Fluorescent images were acquired using 488 and 639 nm lasers and bright-field images. Imaging was performed in an incubation chamber (Tokai Hit, Shizuoka, Japan) with the following parameters: temperature, 37 °C, with 5% O₂, 6% CO₂, and 89% N₂. After live-cell imaging, the embryos were placed in CultureCoin® containing the stain again. At least 3 h after the embryo reached the 4-cell stage, it was imaged again using the same microscope, not continuously but only once (Supplementary Fig. 1).

nine embryo was imaged in the second mitotic cycle. The embryos were cultured in CultureCoin until they reached the 2-cell stage. Images were acquired every 10–20 min (80 × 1.25 µm optical sections) for a 5.0 h (IQR = 5.7 h) period. The other conditions and imaging settings were the same as those used for imaging performed during the first mitosis.

Time-lapse imaging

Sixty-two human pronuclear stage embryos were transferred into CultureCoin® in a time-lapse incubator (Miri® Time-Lapse incubator, Esco Medical) after thawing. The wells containing 22 embryos (control group) were filled with 25 μL of HIGROW OVIT per embryo, and those containing 40 embryos (dyed group) were filled with 25 μL of the live-cell stain probes (at the same concentration used for imaging the embryos under mineral oil; Supplementary Fig. 1). The videos of the embryos were captured every 5 min during the time-lapse period. All videos were analyzed with the Miri® TL Viewer software (Esco Medical).

Image analysis

three-dimensional (3D) visualizations of human embryos were performed using the Imaris software (version 10.0; Oxford Instruments). A median filter was applied prior to analysis. Spindle images for analysis were captured in TIFF format (500 × 500 pixels) with 500% zoom using a Snapshot in Imaris10.0 (Oxford Instruments). This image was captured at the end of metaphase from an angle where the chromosomes appeared to be aligned in a straight line and where the spindle appeared to have the maximum area. Each captured image was processed using the Pixel Classification algorithm embedded in Ilastik version 1.4.0, image recognition software^42,43 and only the shape of the spindle was extracted from the background as a region of interest. Each processed image was binarized, and the aspect ratio, a ratio of major and minor axis, of each image was calculated automatically using Fiji⁴⁴. This “aspect ratio” is one of the basic amounts of image characteristics; the range is >1, with 1 meaning close to a true circle.

The morphology of the nuclei was analyzed at least 3 h after the first or second division. Only spherical nuclei were measured, and nuclei that were difficult to recognize and did not exhibit spherical shapes were not measured.

Immunofluorescence after live imaging

Four human 2PN zygotes were live-imaged using the same method. After PNBD, the zygotes were imaged for 30.0 ± 21.6 [SD]mins and fixed after 66.5 ± 29.6 [SD] mins from PNBD.

Human 2PN zygotes were fixed for 60 min at 37 °C in 100 mM HEPES (pH 7, titrated with KOH), 50 mM EGTA (pH 7, titrated with KOH), 10 mM MgSO₄, 2% formaldehyde (MeOH free) and 0.2% Triton X-100, based on previously published methods³⁷. Human zygotes were left in PBS containing 0.1% Triton X-100 overnight at 4 °C. All antibody incubations were performed in PBS, 3% BSA, and 0.1% Triton X-100, either overnight at 4 °C (for primary antibodies) or for 2 h at room temperature (for secondary antibodies). The primary antibody was mouse anti-γ-tubulin (1:500, GTU88, T5326, lot number:0000216733, Sigma-Aldrich). The secondary antibody was Alexa-Fluor-555 labeled anti-mouse (1:500, ab150118, lot number:1062774-2, Abcam). DNA was stained with 0.5 µg/ml Hoechst 33342 (Molecular Probes). Samples were imaged using LSM980 laser scanning confocal equipped with an Airyscan detector (Zeiss, Japan) using equipped with an objective C-Apochromat 40x/NA 1.20 W Korr.

Statistical analysis

All statistical analyses were performed using EZR⁴⁵, a modified version of R, with statistical functions frequently used in biostatistics. An unpaired two-tailed Student’s t-test was used for two groups (Figs. 2c, d and 3f) after assessing their distribution with the Kolmogorov-Smirnoff test. Two-tailed Mann–Whitney U-test (Fig. 3c, S5d), or a two-sided Fisher’s exact test (Figs. 1g, 3d, 4a, 4e–g, S5b, S5c), and Kruskal-Wallis tests were used to analyze more than two groups (Fig. 1c).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The raw data from this study are available from the corresponding author upon request. The primary confocal microscopy data were not uploaded to a data repository because of their large size but are available from the authors upon request. Source data are provided with this paper.

References

McCoy, R. C. et al. Common variants spanning PLK4 are associated with mitotic-origin aneuploidy in human embryos. Science 348, 235–238 (2015).
Article ADS CAS PubMed PubMed Central Google Scholar
Wells, D. & Delhanty, J. D. Comprehensive chromosomal analysis of human preimplantation embryos using whole genome amplification and single cell comparative genomic hybridization. Mol. Hum. Reprod. 6, 1055–1062 (2000).
Article CAS PubMed Google Scholar
van Echten-Arends, J. et al. Chromosomal mosaicism in human preimplantation embryos: a systematic review. Hum. Reprod. Update 17, 620–627 (2011).
Article PubMed Google Scholar
McCoy, R. C. Mosaicism in preimplantation human embryos: when chromosomal abnormalities are the norm. Trends Genet 33, 448–463 (2017).
Article CAS PubMed PubMed Central Google Scholar
Kort, D. H. et al. Human embryos commonly form abnormal nuclei during development: a mechanism of DNA damage, embryonic aneuploidy, and developmental arrest. Hum. Reprod. 31, 312–323 (2016).
CAS PubMed Google Scholar
Ottolini, C. S. et al. Tripolar mitosis and partitioning of the genome arrests human preimplantation development in vitro. Sci. Rep. 7, 9744 (2017).
Article ADS PubMed PubMed Central Google Scholar
Zhan, Q., Ye, Z., Clarke, R., Rosenwaks, Z. & Zaninovic, N. Direct unequal cleavages: embryo developmental competence, genetic constitution and clinical outcome. PLoS ONE 11, e0166398 (2016).
Article PubMed PubMed Central Google Scholar
Ezoe, K. et al. Human 1PN and 3PN zygotes recapitulate all morphokinetic events of normal fertilization but reveal novel developmental errors. Hum. Reprod. 37, 2307–2319 (2022).
Article PubMed Google Scholar
Hashimoto, S. et al. Multinucleation per se is not always sufficient as a marker of abnormality to decide against transferring human embryos. Fertil. Steril. 106, 133–139 (2016).
Article PubMed Google Scholar
Sayed, S. et al. Nucleation status of Day 2 pre-implantation embryos, acquired by time-lapse imaging during IVF, is associated with live birth. PLoS ONE 17, e0274502 (2022).
Article CAS PubMed PubMed Central Google Scholar
Del Carmen Nogales, M. et al. Type of chromosome abnormality affects embryo morphology dynamics. Fertil. Steril. 107, 229–235 (2017).
Article PubMed Google Scholar
Desai, N., Goldberg, J. M., Austin, C. & Falcone, T. Are cleavage anomalies, multinucleation, or specific cell cycle kinetics observed with time-lapse imaging predictive of embryo developmental capacity or ploidy? Fertil. Steril. 109, 665–674 (2018).
Article PubMed Google Scholar
Desch, L. et al. Embryo multinucleation at the two-cell stage is an independent predictor of intracytoplasmic sperm injection outcomes. Fertil. Steril. 107, 97–103 (2017).
Article PubMed Google Scholar
Setti, A. S., Figueira, R. C. S., Braga, D. P. A. F., Iaconelli, A. & Borges, E. Jr Blastomere nucleation: Predictive factors and influence of blastomere with no apparent nuclei on blastocyst development and implantation. JBRA Assist Reprod. 22, 102–107 (2018).
PubMed PubMed Central Google Scholar
Kai, Y., Moriwaki, H., Yumoto, K., Iwata, K. & Mio, Y. Assessment of developmental potential of human single pronucleated zygotes derived from conventional in vitro fertilization. J. Assist Reprod. Genet 35, 1377–1384 (2018).
Article PubMed PubMed Central Google Scholar
Currie, C. E. et al. The first mitotic division of human embryos is highly error prone. Nat. Commun. 13, 6755 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Chatzimeletiou, K., Morrison, E. E., Prapas, N., Prapas, Y. & Handyside, A. H. Spindle abnormalities in normally developing and arrested human preimplantation embryos in vitro identified by confocal laser scanning microscopy. Hum. Reprod. 20, 672–682 (2005).
Article PubMed Google Scholar
Ye, Y. & Homer, H. A. Two-step nuclear centring by competing microtubule- and actin-based mechanisms in 2-cell mouse embryos. EMBO Rep. 23, e55251 (2022).
Article CAS PubMed PubMed Central Google Scholar
Allais, A. & FitzHarris, G. Absence of a robust mitotic timer mechanism in early preimplantation mouse embryos leads to chromosome instability. Development 149, dev200391 (2022).
Article CAS PubMed Google Scholar
Vázquez-Diez, C., Paim, L. M. G. & FitzHarris, G. Cell-size-independent spindle checkpoint failure underlies chromosome segregation error in mouse embryos. Curr. Biol. 29, 865–873 e863 (2019).
Article PubMed Google Scholar
Yamagata, K. & FitzHarris, G. 4D imaging reveals a shift in chromosome segregation dynamics during mouse pre-implantation development. Cell Cycle 12, 157–165 (2013).
Article CAS PubMed PubMed Central Google Scholar
Cavazza, T. et al. Parental genome unification is highly error-prone in mammalian embryos. Cell 184, 2860–2877 e2822 (2021).
Article CAS PubMed PubMed Central Google Scholar
Reichmann, J. et al. Dual-spindle formation in zygotes keeps parental genomes apart in early mammalian embryos. Science 361, 189–193 (2018).
Article ADS CAS PubMed Google Scholar
Vázquez-Diez, C., Yamagata, K., Trivedi, S., Haverfield, J. & FitzHarris, G. Micronucleus formation causes perpetual unilateral chromosome inheritance in mouse embryos. Proc. Natl Acad. Sci. USA 113, 626–631 (2016).
Article ADS PubMed PubMed Central Google Scholar
Manandhar, G., Sutovsky, P., Joshi, H. C., Stearns, T. & Schatten, G. Centrosome reduction during mouse spermiogenesis. Dev. Biol. 203, 424–434 (1998).
Article CAS PubMed Google Scholar
Courtois, A., Schuh, M., Ellenberg, J. & Hiiragi, T. The transition from meiotic to mitotic spindle assembly is gradual during early mammalian development. J. Cell Biol. 198, 357–370 (2012).
Article CAS PubMed PubMed Central Google Scholar
Adashi, E. Y. & Cohen, I. G. Who will oversee the ethical limits of human embryo research? Nat. Biotechnol. 40, 463–464 (2022).
Article CAS PubMed Google Scholar
Domingo-Muelas, A. et al. Human embryo live imaging reveals nuclear DNA shedding during blastocyst expansion and biopsy. Cell 186, 3166–3181 e3118 (2023).
Article CAS PubMed Google Scholar
Ghetler, Y. et al. Human oocyte cryopreservation and the fate of cortical granules. Fertil. Steril. 86, 210–216 (2006).
Article PubMed Google Scholar
Papale, L., Fiorentino, A., Montag, M. & Tomasi, G. The zygote. Hum. Reprod. 27, i22–i49 (2012).
Article PubMed Google Scholar
Jentoft, I. M. A. et al. Mammalian oocytes store proteins for the early embryo on cytoplasmic lattices. Cell 186, 5308–5327.e25 (2023).
Article CAS PubMed Google Scholar
Oakley, C. E. & Oakley, B. R. Identification of γ-tubulin, a new member of the tubulin superfamily encoded by mipA gene of Aspergillus nidulans. Nature 338, 662–664 (1989).
Article ADS CAS PubMed Google Scholar
Moritz, M., Braunfeld, M. B., Guenebaut, V., Heuser, J. & Agard, D. A. Structure of the gamma-tubulin ring complex: a template for microtubule nucleation. Nat. Cell Biol. 2, 365–370 (2000).
Article CAS PubMed Google Scholar
Santaguida, S. & Amon, A. Short- and long-term effects of chromosome mis-segregation and aneuploidy. Nat. Rev. Mol. Cell Biol. 16, 473–485 (2015).
Article CAS PubMed Google Scholar
Li, X. H. et al. CHK2 is essential for spindle assembly and DNA repair during the first cleavage of mouse embryos. Aging 12, 10415–10426 (2020).
Article CAS PubMed PubMed Central Google Scholar
Schneider, I., de Ruijter-Villani, M., Hossain, M. J., Stout, T. A. E. & Ellenberg, J. Dual spindles assemble in bovine zygotes despite the presence of paternal centrosomes. J. Cell Biol. 220, E202010106 (2021).
Article CAS PubMed PubMed Central Google Scholar
Holubcová, Z., Blayney, M., Elder, K. & Schuh, M. Error-prone chromosome-mediated spindle assembly favors chromosome segregation defects in human oocytes. Science 348, 1143–1147 (2015).
Article ADS PubMed PubMed Central Google Scholar
So, C. et al. Mechanism of spindle pole organization and instability in human oocytes. Science 375, eabj3944 (2022).
Article ADS CAS PubMed Google Scholar
Fishman, E. L. et al. A novel atypical sperm centriole is functional during human fertilization. Nat. Commun. 9, 2210 (2018).
Article ADS PubMed PubMed Central Google Scholar
Schatten, H. & Sun, Q. Y. New insights into the role of centrosomes in mammalian fertilization and implications for ART. Reproduction 142, 793–801 (2011).
Article CAS PubMed Google Scholar
Ezoe, K. et al. Characteristics of the cytoplasmic halo during fertilisation correlate with the live birth rate after fresh cleaved embryo transfer on day 2 in minimal ovarian stimulation cycles: a retrospective observational study. Reprod. Biol. Endocrinol. 19, 172 (2021).
Article CAS PubMed PubMed Central Google Scholar
Shirahata, M. et al. Novel, improved grading system(s) for IDH-mutant astrocytic gliomas. Acta Neuropathol. 136, 153–166 (2018).
Article CAS PubMed Google Scholar
Sommer, C., Straehle, C., Köthe, U. & Hamprecht, F. A. Ilastik: Interactive learning and segmentation toolkit. In IEEE International Symposium on Biomedical Imaging: From Nano to Macro (2011).
Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012).
Article CAS PubMed Google Scholar
Kanda, Y. Investigation of the freely available easy-to-use software “EZR” for medical statistics. Bone Marrow Transpl. 48, 452–458 (2013).
Article CAS Google Scholar

Download references

Acknowledgements

We gratefully acknowledge Akihiro Ichikawa of Carl Zeiss Co. Ltd. and Soichi Koyota of Molecular Medicine Laboratory, Bioscience Education and Research Support Center, Akita University, for their support and assistance with image analysis in this work. We also thank Kota Saito and Miharu Maeda of Department of Biological Informatics and Experimental Therapeutics, Graduate School of Medicine, Akita University for advice about cell incubation. We would also like to thank Editage (www.editage.com) for their English language editing services. Y.O. is supported by the Imai Foundation. Y.T. is supported by the JSPS KAKENHI (https://www.jsps.go.jp/english/index.html; grant numbers 21H03073 and 21K19549).

Author information

Authors and Affiliations

Department of Obstetrics and Gynecology, Akita University Graduate School of Medicine, Akita, Japan
Yuki Ono, Hiromitsu Shirasawa, Kazumasa Takahashi, Mayumi Goto, Taichi Sakaguchi, Motonari Okabe, Takeo Hirakawa, Takuya Iwasawa, Akiko Fujishima, Tae Sugawara, Kenichi Makino, Hiroshi Miura, Yukiyo Kumazawa & Yukihiro Terada
Department of Neurosurgery, Akita University Graduate School of Medicine, Akita, Japan
Takahiro Ono
Asada Institute for Reproductive Medicine, Asada Ladies Clinic, Nagoya, Aichi, Japan
Noritaka Fukunaga & Yoshimasa Asada

Authors

Yuki Ono
View author publications
You can also search for this author in PubMed Google Scholar
Hiromitsu Shirasawa
View author publications
You can also search for this author in PubMed Google Scholar
Kazumasa Takahashi
View author publications
You can also search for this author in PubMed Google Scholar
Mayumi Goto
View author publications
You can also search for this author in PubMed Google Scholar
Takahiro Ono
View author publications
You can also search for this author in PubMed Google Scholar
Taichi Sakaguchi
View author publications
You can also search for this author in PubMed Google Scholar
Motonari Okabe
View author publications
You can also search for this author in PubMed Google Scholar
Takeo Hirakawa
View author publications
You can also search for this author in PubMed Google Scholar
Takuya Iwasawa
View author publications
You can also search for this author in PubMed Google Scholar
Akiko Fujishima
View author publications
You can also search for this author in PubMed Google Scholar
Tae Sugawara
View author publications
You can also search for this author in PubMed Google Scholar
Kenichi Makino
View author publications
You can also search for this author in PubMed Google Scholar
Hiroshi Miura
View author publications
You can also search for this author in PubMed Google Scholar
Noritaka Fukunaga
View author publications
You can also search for this author in PubMed Google Scholar
Yoshimasa Asada
View author publications
You can also search for this author in PubMed Google Scholar
Yukiyo Kumazawa
View author publications
You can also search for this author in PubMed Google Scholar
Yukihiro Terada
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

Y.O., K.T., and Y.T. designed the study. Y.O. and M.G. conducted the experiments. N.F. and Y.A. collected the human embryos and obtained ethical approval for the study. Y.O. performed the data analysis and interpretation. Particle analysis was performed by T.O.; Y.O. wrote the manuscript. The method of particle analysis was developed by T.O., H.S., K.T., M.O., Taichi Sakaguchi, T.H., T.I., A.F., Tae Sugawara, K.M., H.M., Y.K., and Y.T. All authors reviewed the final manuscript.

Corresponding author

Correspondence to Yuki Ono.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Cecilia Blengini, and the other, anonymous, reviewer for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Description of Additional Supplementary Files

Supplementary Movie 1

Supplementary Movie 2

Supplementary Movie 3

Supplementary Movie 4

Supplementary Movie 5

Supplementary Movie 6

Supplementary Movie 7

Supplementary Movie 8

Supplementary Movie 9

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Ono, Y., Shirasawa, H., Takahashi, K. et al. Shape of the first mitotic spindles impacts multinucleation in human embryos. Nat Commun 15, 5381 (2024). https://doi.org/10.1038/s41467-024-49815-8

Download citation

Received: 18 January 2024
Accepted: 18 June 2024
Published: 25 June 2024
DOI: https://doi.org/10.1038/s41467-024-49815-8

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.