Physcomitrium patens SUN2 Mediates MTOC Association with the Nuclear Envelope and Facilitates Chromosome Alignment during Spindle Assembly

Abstract

Plant cells lack centrosomes and instead utilize acentrosomal microtubule organizing centers (MTOCs) to rapidly increase the number of microtubules at the onset of spindle assembly. Although several proteins required for MTOC formation have been identified, how the MTOC is positioned at the right place is not known. Here, we show that the inner nuclear membrane protein SUN2 is required for MTOC association with the nuclear envelope (NE) during mitotic prophase in the moss Physcomitrium patens. In actively dividing protonemal cells, microtubules accumulate around the NE during prophase. In particular, regional MTOC is formed at the apical surface of the nucleus. However, microtubule accumulation around the NE was impaired and apical MTOCs were mislocalized in sun2 knockout cells. Upon NE breakdown, the mitotic spindle was assembled with mislocalized MTOCs. However, completion of chromosome alignment in the spindle was delayed; in severe cases, the chromosome was transiently detached from the spindle body. SUN2 tended to localize to the apical surface of the nucleus during prophase in a microtubule-dependent manner. Based on these results, we propose that SUN2 facilitates the attachment of microtubules to chromosomes during spindle assembly by localizing microtubules to the NE. MTOC mispositioning was also observed during the first division of the gametophore tissue. Thus, this study suggests that microtubule–nucleus linking, a well-known function of SUN in animals and yeast, is conserved in plants.

Introduction

The linking of the nucleus to the cytoskeleton is a common feature of eukaryotic cells. The nuclear-associated cytoskeleton determines the nuclear position, which is involved in cell physiology and fate, and applies force to the nucleus, which affects gene expression (Gundersen and Worman 2013, Almonacid et al. 2019). Linkage between the nucleus and the microtubule organizing center (MTOC) has been reported in animals and fungi; centrosomes in certain animal cell types and the spindle pole body (SPB) in yeast are associated with the nuclear envelope (NE) (Mejat and Misteli 2010). The key conserved factors that link the nucleus to the cytoskeleton in animals and yeasts is the linkers of the nucleoskeleton to the cytoskeleton (LINC) complex, which comprises the inner nuclear membrane protein SUN and the outer nuclear membrane protein (Klarsicht/ANC-1/Syn homology (KASH); Mejat and Misteli 2010, Meier 2016, Jahed et al. 2021). SUN has three recognizable domains: a transmembrane region, multimerization domain, and C-terminal SUN domain. The SUN domain binds to the C-terminus of KASH in the nuclear intermembrane region, whereas the N-terminus of SUN interacts with chromosomes. The N-terminus of the KASH protein interacts with microtubules and/or actin filaments in the cytoplasm either directly or via motor proteins (myosin, kinesin and dynein). Mutations in SUN in animals cause a variety of cellular defects, such as nuclear deformation and mispositioning, and sometimes cause diseases in humans (Mejat and Misteli 2010, Meinke et al. 2014). In yeast, mutations in SUN lead to lethality owing to the failure of SPB separation and bipolar spindle formation during mitosis (Hagan and Yanagida 1995, Jaspersen et al. 2006). The role of SUN in telomere anchorage during meiosis is also conserved in yeast and animals (Mejat and Misteli 2010).

In plants, the SUN family is classified as Cter-SUN or mid-SUN depending on whether the SUN domain is located near the C-terminus, similar to animal and fungal SUNs, or in the middle of the protein (Graumann et al. 2014). Similar to animal/yeast ortholog, Cter-SUN is an inner nuclear membrane protein required for nuclear morphology, movement and telomere anchorage during meiosis (Meier et al. 2017). In Arabidopsis thaliana sun1 mutant, the shape of the nucleus is circular (Oda and Fukuda 2011, Zhou et al. 2012). In contrast, the plant-unique mid-SUN is localized not only in the NE but also in the ER (Graumann et al. 2014). Arabidopsis mid-SUNs (SUN3, 4 and 5) are redundantly essential for early seed development and are involved in nuclear morphology (Graumann et al. 2014) and ER-related activities (Xue et al. 2022). Regarding the link to the cytoskeleton, nuclear migration in Arabidopsis is driven by actin and myosin XI–i, which bind to the WPP domain-interacting tail-anchored protein (WIT) and WPP domain-interacting protein (WIP) (WIT–WIP) complex, i.e. the functional homolog of KASH, which interacts with SUN (Tamura et al. 2013). However, whether the SUN-KASH bridge is linked to the microtubule cytoskeleton in plants remains unclear.

The moss Physcomitrium patens is a model system suitable for studying cytoskeletal and nuclear dynamics, owing to its amenability to high-resolution live microscopy and gene editing techniques. Recent studies have revealed that microtubules and two kinesin family proteins drive nuclear migration (kinesin with calponin homology domain (KCH) for retrograde migration and armadillo repeat-containing kinesin (ARK) for anterograde migration) (Yamada and Goshima 2018, Yoshida et al. 2023). However, the mechanisms by which KCH and ARK recognize the nucleus remain unclear. We initiated the current study to test the hypothesis that the SUN-KASH-kinesin axis is responsible for nuclear motility. First, we attempted to conduct a loss-of-function study of SUN genes, which are more easily identifiable via sequence homology searches than KASH/WIP/WIT genes. In addition to nuclear motility, we observed a defect in MTOC positioning in prophase following the deletion of SUN2, one of two Cter-SUNs. Moreover, the sun2 knockout (KO) line showed delayed chromosome congression during prometaphase. Thus, this study showed that SUN is involved in MTOC positioning and attachment to the NE, revealing the functional conservation of SUN in the three major kingdoms. Furthermore, our data suggested that microtubule attachment to chromosomes during spindle assembly was facilitated by SUN2-dependent microtubule accumulation around the nucleus.

Results

Nuclear migration in subapical cells is suppressed in the absence of SUN2

Physcomitrium patens possesses two Cter-SUNs (SUN1 and SUN2) and two mid-SUNs (SUN3 and SUN4) (Fig. 1A). To investigate the function of Cter-SUN in P. patens, we aimed to delete nearly the entire open reading frame of SUN1 and SUN2 using CRISPR/Cas9 in a line expressing GFP-tubulin and histoneH2B-mCherry. We successfully obtained KO lines for SUN2 (Supplementary Fig. S1A, B). The sun2 KO line grew indistinguishably from the parental line on culture plates (Fig. 1B, C).

To identify the phenotypes of the sun2 KO line at the cellular level, we conducted long-term time-lapse imaging of the microtubules and chromosomes using low-resolution microscopy (Fig. 2A; Movie 1). Changes in nuclear morphology were not convincingly detected by this microscopy. In contrast, abnormal nuclear movement was observed. In the wild-type protonemata, the daughter nuclei moved in the apical direction in the apical daughter cell and in the basal direction in the subapical daughter cell after apical cell division (Fig. 2A, B). In the sun2 KO line, nuclear movement of apical cells was comparable to that of control cells. However, the rate of basal motility significantly decreased in the subapical cells from 145 min after the anaphase onset (Fig. 2B, C). In addition, abnormal apical movement was occasionally observed (Fig. 2C, negative values). This phenotype was suppressed by the ectopic SUN2-mCerlean expression, indicating that the observed motility defects were caused by SUN2 depletion. We conclude that directional nuclear migration during interphase is partially impaired by the loss of SUN2.

Mitosis is delayed in the absence of SUN2

In addition to nuclear motility, we identified mitotic defects in the sun2 KO line. Physcomitrium patens protonemal apical cells undergo highly precise mitotic cell division under laboratory culture conditions (Nakaoka et al. 2012). We confirmed this by observing 55 mitotic events in the control line: the duration from NE breakdown (NEBD) to anaphase onset was 13.9 ± 1.32 min (± SD). This duration was significantly increased in the sun2 KO line (16.3 ± 1.76 min; n = 50) (Fig. 2D). This phenotype was suppressed when SUN2-mCerlean was ectopically expressed (14.6 ± 1.41 min, n = 55).

To test whether the N-terminal (putative chromatin-binding) or C-terminal (putative KASH/WIP-binding) domain of SUN2 is responsible for mitotic progression, two truncated constructs were constructed and individually transformed into the sun2 KO line. Both truncated constructs were localized along the NE (Supplementary Fig. S2); however, neither construct restored mitotic duration (Fig. 2D). These results indicated that both termini are required for rapid mitotic progression.

For unknown reasons, we were unable to obtain a moss line with complete deletion of the SUN1 gene. Therefore, we generated SUN1 loss-of-function mutants using CRISPR/Cas9 with different guide-RNA sets. We obtained three alleles from the background of sun2 KO lines. In one allele, a 298-bp deletion was detected in exon2 and exon3 of SUN1 (sun1-1/sun2 KO allele) (Supplementary Fig. S1C). This is likely a strong loss-of-function allele of Cter-SUN. However, the mitotic duration was not further extended compared to single sun2 KO (Fig. 2F). In addition, we obtained an allele, termed sun1–4, in the background of the sun2+ line, where a 716-bp deletion was detected in exon1/2 (Supplementary Fig. S1C). Mitosis was delayed in this mutant, similar to the sun2 KO (Fig. 2E). These results suggest that SUN1 also plays a role in controlling cell division in the protonemata, although it cannot be ruled out that the mutant SUN1 protein acts on SUN2 in a dominant-negative fashion. Because the sun2 KO robustly showed mitotic phenotypes and the null allele could not be obtained for sun1, we used the sun2 single KO line for most subsequent analyses.

SUN2 is required for proper positioning of mitotic MTOC in protonemal cells

To examine the nuclear morphology and spindle/chromosome dynamics, we used high-resolution live microscopy. In the control line, the nucleus became ellipsoidal or diamond-shaped 10–90 min before NEBD in protonemal apical cells, concomitant with the observation of the surrounding microtubule bundles (Fig. 3A, B). The microtubules applied force to the nucleus: no change in shape was observed when the microtubules were depolymerized with oryzalin (Fig. 6B). In the sun2 KO line, the nucleus remained round-shaped and microtubule bundles around the NE were less prominent (Fig. 3A, B). In control subapical cells, the nucleus was ellipsoidal during interphase; however, it became round in the absence of SUN2 (Fig. 3C, D). The nuclear phenotype was rescued by ectopic expression of SUN2-mCerulean. Thus, SUN2 is required for microtubule-dependent nuclear morphogenesis, which is consistent with previous observations in Arabidopsis (Oda and Fukuda 2011, Zhou et al. 2012).

Just before NEBD (<10 min), the nucleus in the control apical cells transformed again to round shape, accompanied by the emergence of the microtubule ‘apical cap’, which refers to the accumulation of microtubules at the apical side of the nuclear surface (Fig. 4A, Movie 2). These microtubules, as MTOC, are thought to offer a force to change nuclear morphology, move the late prophase nucleus and serve as the initial source of spindle microtubules after NEBD (Nakaoka et al. 2012). The apical cap was infrequently observed in the sun2 KO line (Fig. 4A, Movie 2). Instead, MTOCs were detected at different positions in 21 of the 25 cells (Fig. 4B). The reduction of microtubules on the apical side of the nucleus was confirmed by signal quantification (Fig. 4C, D). The microtubules on the basal side were also reduced (Fig. 4C, E; Movie 2). These phenotypes were suppressed by SUN2-mCerlean expression, indicating that the failure of apical cap formation and basal microtubule accumulation was caused by SUN2 protein depletion. The truncated SUN2 constructs, which lacked either the N-terminal region or the C-terminal SUN domain, did not rescue the phenotype, suggesting that both regions are required for the proper positioning of microtubule bundles (Fig. 4A, B). From these results, we conclude that SUN2 is required for microtubule association with NE in late prophase.

SUN2 is required for efficient chromosome alignment in the spindle

In addition to the MTOC position, high-resolution imaging revealed differences in chromosomal dynamics during spindle assembly (Fig. 5A, Movie 2). In control cells, the nucleus rapidly migrated apically in late prophase, followed by NEBD [Fig. 5B, C (kymographs)]. The apical motility of the chromosomes persisted for a few minutes after NEBD. Chromosomes on the basal side of the nucleus traveled more rapidly and over longer distances than those on the apical side, leading to chromosome congression at the spindle equator within ∼10 min (Fig. 5C). In contrast, rapid apical movement of the nucleus in late prophase and prometaphase was largely suppressed in the sun2 KO line. Furthermore, the histone signal in the kymograph tended to extend basally after NEBD; consequently, the initiation of apical migration was delayed (arrowheads in Fig. 5B, C). In two cases, we observed a clearly misaligned chromosome detached from the spindle body (Fig. 5A, arrowheads). However, misaligned chromosomes were eventually captured by spindle microtubules during prolonged prometaphase, which was different from kinetochore deficiency, in which chromosome congression is never achieved (Kozgunova et al. 2019). We conclude that SUN2 facilitates chromosome alignment in the spindle.

Asymmetric SUN2 distribution during apical MTOC assembly

As expected, endogenous SUN2 tagged with mNeonGreen (mNG) was uniformly localized to the NE during interphase (Fig. 6A). In contrast, time-lapse microscopy and signal quantification indicated that SUN2 localization was asymmetric in late prophase; SUN2-mNG was more enriched on the apical side, partially overlapping the microtubule apical cap (Fig. 6B, C).

To test whether the asymmetric distribution was dependent on microtubules, we depolymerized the cytoplasmic microtubules with oryzalin, followed by time-lapse microscopy. Quantification of signal intensity showed no apical accumulation of SUN2 in late prophase under these conditions (Fig. 6B, C). In contrast, the depolymerization of actin filaments with latrunculin A did not disrupt the asymmetric distribution of SUN2 (Fig. 6B, C). These results indicate that apical enrichment of SUN2 and microtubules are mutually dependent.

During spindle assembly, SUN2-mNG initially showed punctate signals on the spindle (Fig. 6D). The number of signals gradually decreased, and the spindle at metaphase was cleared.

SUN2 controls the MTOC position and microtubule–NE interaction in the gametophore initial cell

The gametophore in Physcomitrella is the leafy shoot, which develops from protonemal filaments. In the first stage of gametophore development, stem cells undergo a type of asymmetric division distinct from that of the protonemata (Harrison et al. 2009, Kofuji and Hasebe 2014). We examined the role of SUN2 in these cells. In this system, the microtubule cloud, or also called regional MTOC ‘gametosome’, emerges at the apical cytoplasm and functions as the dominant microtubule nucleation site (Kosetsu et al. 2017) (Supplementary Fig. S3 and Fig. 7A, arrowhead). In control cells expressing GFP-tubulin, gametosomes appeared in the apical cytoplasm in prophase, and the line connecting the nuclear and gametosome centers was nearly parallel to the long axis of the cell (Fig. 7B, C). In the sun2 KO line, gametosomes were formed. However, its position relative to the nucleus was more variable (Fig. 7C). The gametosome dictates spindle orientation, and consequently, the division plane in gametophore initial cells (Kosetsu et al. 2017). Consistent with the variable positions of the gametosome, the orientation of the metaphase spindle (Fig. 7D–F) and cell plate (Fig. 7G–I) was also variable in the sun2 KO line. Finally, we observed that perinuclear microtubules decreased in the absence of SUN2 using the moss line expressing mCherry-tubulin, which visualizes these microtubules better than GFP-tubulin, owing to lower background signals (Supplementary Fig. S3). Thus, SUN2 plays a similar role in the gametophore initial cell in terms of microtubule–NE association and MTOC positioning.

Discussion

The contribution of nuclear membrane proteins to cellular events in plants remains poorly understood. In this study, we found that the SUN2 protein in P. patens not only plays a well-established role in nuclear shaping and positioning but also facilitates chromosome alignment during mitosis. A series of live imaging supports a model in which SUN2 mediates the interaction between MTOC and the nucleus during mitotic prophase, enabling the efficient association of microtubules with chromosomes during spindle assembly.

Physcomitrella SUN2 couples NE with microtubules

The loss of centrosomes is a striking event in plant lineages (Buschmann and Zachgo 2016). Several types of acentrosomal MTOCs have been developed as centrosome substitutes (Lloyd and Chan 2006, Buschmann et al. 2016, Yi and Goshima 2018, Naramoto et al. 2022). In some cases, the proteins required for MTOC formation have been identified (Liu and Lee 2022). However, little is known about the spatial control of acentrosomal MTOCs in plants.

The structure and location of prophase MTOC differ between protonemal and gametophore tissues in P. patens. The apical cap is a unique form of protonemal cell. In the absence of SUN2, this structure was scarcely observed. However, MTOC per se remained observed either near the non-apical surface of the NE or in the apical cytoplasm (i.e. detached from the NE). In contrast, gametosome MTOC, which is essential for spindle orientation, is assembled in the small apical cytoplasm, not around the NE, in the gametophore initial cell in the wild type; gametosomal microtubules extend and interact with the NE [Supplementary Fig. S3; (Kosetsu et al. 2017)]. The position of gametosomes in the cytoplasm was more variable in the absence of SUN2. Thus, while mislocalization of MTOC in the sun2 KO line was consistent between protonemal and gametophore initial cells, different MTOC phenotypes were observed in these two cell types. However, the perinuclear microtubules that did not participate in the apical cap or gametosome were reduced in both cell types. Based on these observations, we conclude that SUN2 has a common function of linking microtubules to the prophase nucleus in these two cell types and suggest that the loss of this linking function affects the location of distinct types of MTOCs (apical cap and gametosome) in sun2 KO.

The microtubule–NE-linking function of SUN2 may not be limited to the prophase; the interphase nucleus in subapical cells was deformed concomitant with a reduction in the surrounding microtubules in the sun2 KO line. In animals, the SUN protein forms a central part of the LINC complex, which connects the cytoskeleton, including microtubules and actin, with nuclear laminae and chromosomes across the NE (Mejat and Misteli 2010, Gundersen and Worman 2013). In Arabidopsis, SUN mediates actin–NE interactions via WIP/WIT and myosin XI–i (Tamura et al. 2013). Our results indicated that the microtubule-linking function of SUN is preserved in plants.

SUN2 prevents chromosome scattering and facilitates chromosome alignment during mitosis

Delayed chromosome alignment and anaphase onset in protonemal cells suggest that chromosome–microtubule interactions do not occur promptly during spindle assembly in the sun2 KO line. This is consistent with the observation that the dominant MTOC is mispositioned in the mutant; analogously, centrosome depletion extends the duration of prometaphase in animal cells (Moutinho-Pereira et al. 2013). In addition, we observed that chromosomes tended to scatter along the long axis of the cells after NEBD. Severe chromosome/kinetochore detachment from the spindle body was observed on the basal side of the cells. Thus, perinuclear microtubules that do not directly participate in apical MTOC may also be critical for efficient chromosome capture. One possibility is that these microtubules act as a physical barrier, restricting the scattering of chromosomes after NEBD, thereby facilitating their attachment to the microtubules at the kinetochore. However, it is not ruled out that SUN2 also actively participates in the spindle assembly process during early prometaphase, e.g. by removing nuclear membrane remnants in the spindle matrix (Turgay et al. 2014).

Limitations of the study

Compared with the known loss-of-function phenotype of SUN in animal cells, the phenotypes observed in this study were mild. For example, we did not detect defects in nuclear migration in the apical cells of sun2 KO or sun1-1/sun2 KO. Similarly, in Arabidopsis, the sun1 KO/sun2 knockdown line showed no dramatic developmental defects under laboratory conditions (Oda and Fukuda 2011, Zhou et al. 2012). Another allele, sun1-1/sun2-2, shows severe phenotypes during meiosis (Varas et al. 2015), suggesting that Cter-SUN may be more important for meiosis in Arabidopsis than for vegetative growth. We speculate that the lack of severe mitotic defects was partly due to the presence of an intact mid-SUN in the mutants. A recent study identified ER-related activities in Arabidopsis mid-SUN (Xue et al. 2022). However, when overexpressed in tobacco leaf epidermal cells, Arabidopsis mid-SUN is localized to both the NE and ER (Graumann et al. 2014, Xue et al. 2022). Considering the domain structure, the mid-SUN may also act as a linker between the cytoskeleton and NE (Graumann et al. 2014, Meier et al. 2017). Despite several attempts using different constructs, we could not obtain KO or mutant alleles for PpSUN3; SUN3 might be required for essential processes in cell proliferation. A comprehensive loss-of-function analysis of Cter- and mid-SUN would be an interesting topic for future research.

How SUN2 contributes to nuclear migration at the molecular level remains unclear. The identification of adapter proteins is key to understanding their molecular mechanisms. Meanwhile, a synthetic phenotype with ark or kch mutants would clarify whether SUN2 mediates anterograde or retrograde (or both) nuclear movement, as anterograde or retrograde movement is specifically and almost completely suppressed in the ark KO and kch KO lines, respectively (Yamada and Goshima 2018, Yoshida et al. 2023).

The types of plant MTOCs that require SUN for localization are another outstanding question. In seed plants, many cell types develop a specialized regional MTOC called ‘polar cap’ or ‘pro-spindle’ in late prophase, which caps both apical and basal sides of the NE (Smirnova and Bajer 1998, Liu and Lee 2022). AtSUN1 and AtSUN2 are abundantly localized at the polar cap (Oda and Fukuda 2011, Tatout et al. 2014). It would be interesting to revisit the sun mutants and examine whether SUN mediates the association between NE and this type of MTOC.

Materials and Methods

The majority of the methods used in this study were identical to those described in our recent studies (Ta et al. 2023, Yoshida et al. 2023).

Physcomitrium patens culture and transformation

All strains in this study were derived from the Gransden ecotype of Physcomitrium (Physcomitrella) patens (Ashton and Cove 1977). Physcomitrium patens culture and transformation protocols followed were as described by Yamada et al. (2016). Briefly, mosses were regularly cultured on BCDAT plates at 25°C under continuous light illumination. A standard polyethylene glycol–mediated method was exploited for transformation. Prior to transformation, sonicated protonemata were cultured on BCDAT agar medium for 5–6 d. Transgenic lines were selected using corresponding antibiotics. Line confirmation was conducted through visual inspection followed by genotyping PCR (Supplementary Fig. S1, Supplementary Table S4). Sequencing was performed to confirm the CRISPR mutant lines. The lines generated in this study are listed in Supplementary Table S1.

Plasmid construction

The plasmids and primers used in this study are listed in Supplementary Tables S2 and S3, respectively. CRISPR targets with high specificity were manually selected in the first three exons of SUN1 gene and the regions near the start or stop codons of SUN2 gene. All target sequences were synthesized and ligated into the BsaI site of pPY156, which is based on pCasGuide/pUC18 and contains a hygromycin-resistant cassette (Lopez-Obando et al. 2016, Yi and Goshima 2020). For endogenous tagging via homologous recombination, the plasmid was constructed using the In-Fusion HD Cloning Kit (Takara); 1- to 2-kb sequences of the 5´ and 3´ ends of the genes of interest flanked the fragment that consisted of an in-frame linker, mNG tagged with FLAG and G418- or blasticidin S–resistant cassette. The mNG codon was optimized for expression in Arabidopsis. For all the rescue experiment, the SUN2 coding sequence was amplified from the moss cDNA library (full-length, truncation, mutant) and ligated into the pENTR/d-TOPO vector containing the in-frame linker, mCerulean-coding sequence, followed by the Gateway LR reaction (Invitrogen, Waltham, MA) into a vector containing the P. patens EF1α promoter, nourseothricin resistance cassette and 1-kb sequences homologous to the PTA1 locus.

Moss growth assay

The 5- to 7-day-old sonicated protonemata with similar sizes were inoculated to the BCDAT plate. Two plates each containing 25 pieces of inoculated protonemata were made for each strain. After 20 d of incubation under the continuous light, images of overall moss or gametophores were acquired using a C-765 Ultra Zoom digital camera (Olympus) or SMZ800N, respectively.

Microscopy

Time-lapse microscopy was performed as described by Nakaoka et al. (2012). Briefly, in the long-term time-lapse imaging experiments for the observation of protonemal cells, the protonemata were cultured on thin layers of BCD agarose in 6-well glass-bottom dishes for 5–7 d. Wide-field, epifluorescence images were acquired with a Nikon Ti microscope (10 × 0.45 NA; numerical aperture lens, Zyla 4.2P CMOS camera [Andor, Belfast, Northern Ireland] and Nikon Intensilight Epi-fluorescence Illuminator) at intervals of 1 min (no z-stacks). For high-resolution imaging, protonemata were inoculated onto the agar pad in a 35-mm glass-bottom dish, followed by culturing for 5–7 d. Confocal imaging was performed with a Nikon Ti microscope attached to a CSU-X1 spinning-disk confocal scanner unit (Yokogawa, Tokyo, Japan), Electron-Multiplying Charge-Coupled Device camera (ImagEM, Hamamatsu, Hamamatsu City, Japan) and two laser lines (561 and 488 nm). Lens of NA 100 × 1.45 was used for most experiments related to protonemal and gametophore cell divisions. For the quantification of gametophore division plane angle, 40 × 1.30 NA lens was used. To induce the gametophore cells, protonemal cells were treated with 2-isopentenyladenine (2iP) for 5–10 min, 20–22 h before imaging (Kosetsu et al. 2017). Stock solution of oryzalin, latrunculin A, FM4-64 and 2iP in DMSO was diluted with distilled water to working concentrations of 10 µM oryzalin, 25 µM latrunculin A, 10 µM FM4-64 and 1 µM 2iP. Prior to drug addition, the protonemal tissue on the agarose pad was preincubated in water for 1 h for absorption. After water removal, 0.3 ml of the drug solution was added and image acquisition was started 5–10 min later. DMSO was used as a control. Imaging was performed at 22–25°C in the dark.

Image data analysis

All raw data processing and measurements were performed using the Fiji software.

Moss growth on the culture plate

The images of the moss on the culture plate were outlined automatically, and the area was measured using Fiji.

Mitotic duration

For caulonemal apical cells, time-lapse images were obtained every 1 min using an epifluorescence (wide-field) microscope and a 10 × 0.45 NA lens, and the duration between NEBD and the anaphase onset was measured.

Nuclear velocity

For caulonemal apical cells, time-lapse images were obtained every 1 min using an epifluorescence (wide-field) microscope and a 10 × 0.45 NA lens. A kymograph of chromosomes was generated along the dividing cell. To obtain nuclear velocity, the inclination of the nuclear signal in the subapical cell 145 min after anaphase onset was manually measured using Fiji.

Nuclear circularity

For caulonemal subapical cells in interphase, the images of the nuclei were obtained with z-stacks at 1-µm intervals for a range of 9 µm using a spinning-disk confocal microscope and a 100 × 1.45 NA lens. The z-stack images were processed by maximum z-projection. The nuclei were outlined automatically, and the circularity of the nucleus was measured using Fiji. For caulonemal apical cells in mitotic prophase, time-lapse images were obtained every 2 min with z-stacks at 1-µm intervals for a range of 7 µm using a spinning-disk confocal microscope and a 100 × 1.45 NA lens. The circularity of the nucleus at the best focal plane was measured manually using Fiji.

Microtubule intensity

For caulonemal apical cells, time-lapse images were obtained every 10 s with z-stacks at 2-µm intervals for a range of 4 µm using a spinning-disk confocal microscope and a 100 × 1.45 NA lens. The best focal plane was selected for the analysis. The mean intensity of the microtubules in a box of 10 × 30 pixels on both the basal and apical sides of the nucleus was measured immediately before NEBD, and the background intensity of each image was subtracted.

Chromosome dynamics during cell division

For caulonemal apical cells, time-lapse images were obtained every 10 s with z-stacks at 2-µm intervals for a range of 4 µm using a spinning-disk confocal microscope and a 100 × 1.45 NA lens. The best focal plane was selected for analysis. A kymograph of chromosome mass was generated along the spindle pole-to-pole axis. For quantification, the chromosome mass on the kymograph was manually outlined using Fiji and, for each timepoint, the distance from the basal edge of the nucleus (set at 11 min before NEBD) was calculated.

Distribution of SUN2-mNG along the nuclear membrane

For caulonemal apical cells, time-lapse images were obtained every 30 s with z-stacks at 1.5-µm intervals for a range of 3 µm using a spinning-disk confocal microscope and a 100 × 1.45 NA lens. The best focal plane was selected for analysis. The intensity of SUN2-mNG and the microtubules along the nucleus was measured just before NEBD, and the background intensity of each image was subtracted. The intensity of each pixel on the drawn line (3-pixel width) was divided by the mean intensity of the entire length of the line to get the relative intensity. The cells were divided into 10 sections, and the average relative intensity of each section is displayed.

Gametosome position

For gametophore initial cells, time-lapse images were obtained every 30 s with z-stacks at 1-µm intervals for a range of 6 µm using a spinning-disk confocal microscope and a 100 × 1.45 NA lens. The best focal plane was selected for analysis. The relative angle between the line connecting the nuclear center to gametosome center and the long axis of the cell was calculated.

Spindle orientation

For gametophore initial cells, time-lapse images were obtained every 30 s with z-stacks at 3-µm intervals for a range of 12 µm using a spinning-disk confocal microscope and a 100 × 1.45 NA lens. The best focal plane was selected for analysis. The angle between the long axis of the spindle and the long axis of the cell was calculated.

Division plane orientation

Gametophore initial cells were stained with FM4-64 prior to the imaging. The images of gametophore initial cells were obtained with z-stacks at 2.5-µm intervals for a range of 25 µm using a spinning-disk confocal microscope and a 40 × 1.30 NA lens. The z-stack images were processed by maximum z-projection using Fiji, and the relative angle between the division plane and the long axis of the cell was calculated.

Statistical analysis

The Shapiro–Wilk test was used for all samples to check for normality. If the sample was assumed to be normally distributed, the F-test (two groups) or Bartlett’s test (multiple groups) was conducted to test homoscedasticity. If the samples had a normal distribution and equal variance, Student’s t-test (two groups) or Tukey’s multiple comparison test (multiple groups) was used. If the samples had a normal distribution but not equal variance, Welch’s two-sample t-test (two groups) or the Games–Howell test (multiple groups) was used. If the samples did not have a normal distribution, Mann–Whitney U test (two groups) or Steel–Dwass test (multiple groups) was used. All statistical analyses were performed using R software. Obtained P-values are denoted as follows: *, P < 0.05; **, P < 0.01; ***, P < 0.001 and ****, P < 0.0001. Data from multiple experiments were combined because of insufficient sample numbers in a single experiment unless otherwise stated.

Supplementary Data

Supplementary data are available at PCPonline.

Data Availability

All materials and data are available from the corresponding author upon request. The gene sequences used in this study are available in Phytozome under the following accession numbers: SUN1 (Pp3c7_4170), SUN2 (Pp3c11_22530), SUN3 (Pp3c21_2240) and SUN4 (Pp3c18_19540).

Funding

The Japan Society for the Promotion of Science KAKENHI (18KK0202, 22H04717 and 22H02644 to G.G.).

Acknowledgments

We are grateful to Maya Hakozaki for providing moss lines and Chiemi Koketsu and Rie Indaba for media preparation. M.W.Y. is a recipient of the Japan Society for the Promotion of Science pre-doctoral fellowship.

Author Contributions

M.W.Y. and G.G. designed the research; M.W.Y. and N.O. performed the experiments. M.W.Y. analyzed the data; M.W.Y. and G.G. wrote the paper.

Disclosures

The authors have no conflicts of interest to declare.

References