Polarity establishment in the plant zygote at a glance Free

Body axis formation, which begins with cell division of the zygote, constitutes the first developmental events in multicellular organisms. In most flowering plants, the zygote undergoes an asymmetric cell division that produces two daughter cells along an apical-basal axis. Although subsequent cell division patterns vary in many species, in the Brassicaceae family, which includes the model plant Arabidopsis thaliana (hereafter referred to as Arabidopsis), cell division in the zygote and developing embryo is regular and shows a simple pattern. Therefore, many insights into plant embryogenesis have been gained through observations using Arabidopsis.

Arabidopsis egg cells are polar, with the nucleus located at the apical end and large vacuoles concentrated at the basal end. After fertilization, the nucleus moves to the center of the cell and the vacuoles disperse throughout the cell (Faure et al., 2002). The zygote then begins to elongate and polarize, and the nucleus migrates apically (Ueda et al., 2011). Once polarization is complete, the zygote divides asymmetrically, with the site of division plane formation determined by the apically positioned nucleus (Mansfield and Briarty, 1991). In case of Arabidopsis, the apical cell divides in a stereotypical pattern to form a spherical embryo, which gives rise to most of the plant body, including the shoot apex. The basal cell divides only a few times in a single orientation, forming a line of cells that becomes the suspensor (see poster). This tissue is responsible for supplying nutrients from maternal tissues to the embryo. Most suspensor cells die during seed maturation, but a few cells next to the embryo survive as part of the root apex (Laux et al., 2004). Thus, first asymmetric division of the zygote clearly defines the primary apical-basal body axis of a multicellular organism, which is essential for subsequent patterning and morphogenesis.

Despite the importance of this first division in plant zygotes, the dynamics and polarization mechanisms underlying this process have long resisted investigation (Kimata and Ueda, 2020; Ueda and Berger, 2018; Ueda and Laux, 2012). This is because plant zygotes are embedded in maternal tissue (ovules), making it difficult to observe them in situ. Recently, however, we have developed a live-cell imaging method for Arabidopsis zygotes, enabling observation of the intracellular changes in zygotes from fertilization to the first division (see Box 1) (Gooh et al., 2015; Kimata et al., 2016, 2019). This Cell Science at a Glance article and the accompanying poster describe the latest findings on the mechanisms of zygote polarization based on live-cell imaging in Arabidopsis. We briefly introduce the molecular mechanisms of Arabidopsis fertilization and zygote development, summarize what has been observed regarding polarization of intracellular components including cytoskeleton and organelles, and conclude by discussing key open questions about these dynamics.

In angiosperms, the multicellular female gametophyte (embryo sac) forms after meiosis of the megaspore mother cell inside an ovule (Dresselhaus et al., 2016). In Arabidopsis, the female gametophyte consists of an egg cell, two synergid cells, a central cell and three antipodal cells. The male gametophyte (pollen) consists of two sperm cells and a vegetative cell, which are produced via meiosis of a microspore mother cell in the anther. After pollination, the two sperm cells are transported to the ovule via a pollen tube formed by the vegetative cell (see poster). One sperm cell then fuses with the egg cell and the other with the central cell, a process known as double fertilization. During fertilization, one sperm cell contacts and adheres to the apex of the egg cell, after which the plasma membranes (PMs) of the two cells fuse. Several factors are known to be involved in this process (see poster). GAMETE EXPRESSED 2 (GEX2), a transmembrane protein expressed in sperm cells, regulates attachment of the sperm cell to the egg cell surface (Mori et al., 2014). GENERATIVE CELL SPECIFIC 1 (GCS1), also known as HAPLESS 2 (HAP2), is a transmembrane protein in sperm cells that controls female–male membrane fusion (Mori et al., 2006; von Besser et al., 2006). After fusion, remnants of the sperm cell membrane remain at the tip of the zygote for several hours during zygote elongation, providing a useful marker of the elongation site (Kang et al., 2023).

Female zygote fusion factors include a family of small cysteine-rich EGG CELL 1 (EC1) peptides that ensure the proper localization of GCS1 during fertilization (Sprunck et al., 2012). When the sperm cell arrives, the egg cell secretes EC1, causing GCS1 to move from the endocytic system to the PM of the sperm cell (see poster). This sperm activation process is necessary for membrane fusion and fertilization, which fail to occur in ec1 mutants (Sprunck et al., 2012). The mechanism by which egg cells recognize sperm cell arrival and secrete EC1 is not well understood. After membrane fusion, the sperm nucleus is released into the egg cell and fuses with the egg nucleus, forming the zygote nucleus. Incidentally, in Arabidopsis, the sperm fusion site and the direction of elongation coincide, but it is unclear whether the sperm entry point determinates the direction of polarity of the zygote, as in animals. At least in rice, in vitro fertilization experiments have shown that the sperm fusion site is not involved in zygote polarity (Nakajima et al., 2010).

The molecular mechanisms underlying zygote polarization remain largely unknown. However, the signaling pathways involved in this process have gradually emerged from forward genetic studies using mutants with defects in zygote polarization, elongation, and asymmetric division (see poster).

SHORT SUSPENSOR (SSP) encodes an interleukin-1 receptor-associated kinase (IRAK), a Pelle-type pseudokinase that activates the signaling pathway responsible for inducing zygote polarization after fertilization (Bayer et al., 2009). SSP is transcribed in sperm cells; after cell fusion, SSP transcripts enter the egg cell and are transiently translated in the zygote. SSP activates a mitogen-activated protein kinase (MAPK) cascade that includes the YODA (YDA) MAPK kinase kinase, the MAPK kinase MKK4 and MKK5, and the MAPKs MPK3 and MPK6. Zygotes lacking these proteins are shorter and fail to divide asymmetrically, resulting in embryos with short suspensors (Lukowitz et al., 2004; Wang et al., 2007; Zhang et al., 2017).

BRASSINOSTEROID SIGNALING KINASE 1 (BSK1) and BSK2, which belong to the same BSK family as SSP, are also required for YDA activation in the zygote (Neu et al., 2019). BSK family proteins often form complexes with somatic embryogenesis receptor kinase (SERK) type receptor-like kinases (Majhi et al., 2019; Shi et al., 2013; Zhao et al., 2019), but SSP, a protein unique to the Brassicaceae, can activate YDA alone without a receptor (Neu et al., 2019). In stomatal development, YDA is activated by the receptor kinase ERECTA family (Shpak et al., 2005). Similarly, in zygote polarization, the maternal ERECTA proteins, ERECTA-LIKE 1 (ERL1) and ERL2 activate the YDA pathway, possibly functioning upstream of BSK1 (Wang et al., 2021).

The WRKY2 transcription factor is a phosphorylation target of MPK3 and MPK6 in the zygote. Activated WRKY2 induces de novo transcription of WUSCHEL HOMEOBOX 8 (WOX8) by binding directly to its promoter, along with the HOMEODOMAIN GLABROUS 11 (HDG11) and HDG12 maternal transcription factors (Ueda et al., 2011, 2017). The zygotes of wrky2, hdg11 and hdg12 mutants, similar to those of MAPK pathway-deficient lines, fail to polarize, but the expression of WOX8 can rescue this phenotype. Thus, the MAPK pathway, activated by SSP, BSK1 and BSK2, and its downstream transcriptional activation of WOX8 play a crucial role in zygote polarization, demonstrating the need for cooperation between paternal and maternal factors during zygote polarization and early embryogenesis.

WOX8 and WOX9 (which are functionally redundant for each other) act as transcriptional activators during embryogenesis (Breuninger et al., 2008; Dolzblasz et al., 2016). WOX8 is already expressed in egg cells and zygotes; however, given that the wox8 wox9 double mutant does not show obvious abnormalities in the first asymmetric division, it is likely that there are other genes downstream of HDG genes and WRKY2 that control zygote polarization (Haecker et al., 2004; Breuninger et al., 2008). It also remains unclear which genes, besides WOX8 and WOX9, are activated by WRKY2, HDG11 and HDG12, and which downstream genes are transcriptionally activated by WOX8 and WOX9.

The live-cell imaging system described in Box 1 has been utilized to observe the dynamics of several intracellular structures during Arabidopsis zygote polarization. Here, we will summarize the spatiotemporal dynamics of cytoskeletal components and organelles, including microtubules, F-actin, the nucleus, mitochondria and vacuoles, as well as discuss potential functions underlying these patterns.

One intracellular feature of zygotes that can be tracked in detail using the live-cell imaging system is the alignment of microtubules (MTs) and actin filaments (F-actin), the main components of the plant cytoskeleton (Kimata et al., 2016). To visualize the cytoskeleton before and after fertilization, plants have been transformed with fluorescently tagged F-actin and tubulin expressed under the control of the EC1 promoter (Sprunck et al., 2012), which acts specifically in egg cells and zygotes. A nuclear marker, the histone protein H2B fused to red fluorescent protein, has been expressed simultaneously to monitor the zygote stage based on nuclear size and position (Kimata et al., 2016).

In an egg cell, the nucleus is located at the cell apex and the MTs are aligned longitudinally along the apical-basal axis. After fertilization, the cell immediately shrinks, the nucleus moves to the center of the cell and the longitudinal microtubular arrays disappear. When zygote elongation starts, the MTs form transverse rings in the subapical region (MT rings). This ring structure is maintained during elongation but then disappears (see poster). The basal end of the zygote contains helical cortical microtubules, which are maintained from the elongation stage to at least the one-cell embryo (Kimata et al., 2016).

The egg cell has a mesh-like pattern of F-actin that also transiently disappears after fertilization. As the zygote elongates, F-actin accumulates in a cap-like pattern at the apical tip, and the filaments gradually become oriented longitudinally along the apical-basal axis (see poster). Thus, the polarity of the cytoskeleton in the egg cell is first disrupted by fertilization and then reorients dramatically with the onset of zygote elongation (Kimata et al., 2016).

What are the roles of these cytoskeletal components in zygote elongation? Live imaging of zygotes treated with the microtubule polymerization inhibitor oryzalin has shown that the apex becomes swollen and rounded and directional elongation was inhibited (Kimata et al., 2016). This indicates that the apical transverse MT rings that appear during elongation normally restrict the direction of cell growth. In zygotes treated with the actin polymerization inhibitor latrunculin B, the nucleus fails to migrate toward the apical end and remains near the center of the cell, resulting in a symmetric cell division. Thus, MTs and F-actin regulate zygote elongation and nuclear migration, respectively, thereby both contributing to essential steps in zygote polarization (Kimata et al., 2016).

Live-cell imaging has also been used to analyze the dynamics of several organelles. Prior to fertilization, the basal region of the egg cell is occupied by very large vacuoles. After fertilization, these vacuoles shrink markedly along with cell size (Faure et al., 2002). During subsequent zygote elongation, the vacuoles form tubular structures around the nucleus and gradually migrate basally along the F-actin array (Kimata et al., 2019) (see poster). In elongating zygotes, the mitochondria fuse into long filamentous structures along the F-actin array and then accumulate in the apical region in mature zygotes. Interestingly, the mitochondria temporarily fragment and show a rounded rather than filamentous morphology during the asymmetric division of the zygote, resulting in the distribution of more mitochondria to the apical cell than the basal cell (see poster) (Kimata et al., 2020). The nucleus, vacuoles and mitochondria are all aligned with F-actin structures, suggesting that F-actin determines the positions of organelles in the zygote.

Few genes are known to regulate cytoskeletal or organelle dynamics during zygote polarization, except for several genes that regulate vacuole migration (Matsumoto et al., 2021). In mutants of these genes, the vacuoles fail to migrate to the basal end of the zygote, resulting in a more symmetric distribution of organelles during the division (see poster). In a mutant of SHOOT GRAVITROPISM2 (SGR2), which encodes a vacuolar membrane protein, the vacuoles are spherical and fail to form tubules during elongation. A mutant of VESICLE TRANSPORT THROUGH INTERACTION WITH T-SNARES11 (VTI11), which encodes a putative vacuolar fusion factor, also exhibits spherical vacuoles. In vti11, tubular vacuoles form but are not properly separated, resulting in a tangled vacuolar network. In both sgr2 and vti11 mutants, vacuoles cannot migrate to the basal end of the zygote. Instead, they accumulate in the apical region, resulting in a failure of the nucleus to migrate apically, and thus producing a symmetric cell division.

To confirm that vacuole contents could not migrate from the apical to the basal end in these mutants, a vacuolar lumen marker was labeled with Kaede, a photoconvertible fluorescent protein whose color changes from green to red upon UV irradiation (Matsumoto et al., 2021). In these experiments, the apical end of the zygote was exposed to UV so that the apical vacuoles were red (photoconverted) and the basal vacuoles were green. Photoconversion was followed by observations of changes in vacuole color throughout the cell. In the wild-type zygotes, all the vacuoles eventually showed overlapping red and green fluorescence, demonstrating the exchange of materials between photoconverted and non-photoconverted vacuoles via the merging of green and red vacuoles. In sgr2 and vti11, the photoconverted apical vacuoles remained red. Therefore, SGR2 and VTI11 regulate the initial formation of the tubular vacuole structures and the movement of vacuolar contents into tubular structures, respectively. Furthermore, a yda mutant fails to unequally distribute the vacuoles, even though formation of the vacuolar network proceeds normally, suggesting that YDA contributes to the directional migration of the vacuoles (Matsumoto et al., 2021).

The importance of the vacuole in zygote polarization is also supported by a comprehensive and quantitative evaluation, which used hierarchical clustering of live-cell imaging data to reveal that the apical and basal regions of the zygote are defined by microtubules and vacuoles, respectively (Hiromoto et al., 2023). This compartmentalization is already evident at the onset of zygote elongation and is maintained through asymmetric cell division, suggesting a polarization mechanism that sets the apical and basal boundaries based on the position of the MT rings and vacuole from the early zygote stage.

As mentioned above, the Arabidopsis zygote also undergoes polarized growth and elongation after fertilization. Plant cell elongation can occur by diffuse growth or by tip growth. Diffuse-growing cells expand along their entire length, whereas tip-growing cells expand at only one end of the cell. Most plant cells are diffuse growing; only some, such as pollen tubes and root hairs, are tip growing (see poster). To determine whether the zygote elongates via tip or diffuse growth, a fluorescently tagged marker of sperm cell membrane has been used to locate the apical tip (Kang et al., 2023; Shiba et al., 2023). As mentioned above, after the PM of the sperm cell fuses with that of the egg cell at fertilization, it remains localized to the apical end of the zygote. Live-cell imaging of zygotes of a plant carrying different markers for egg PM and sperm PM shows that sperm PM remnants at the very tip of the zygote disappear after several hours of elongation, whereas subapical and lateral sperm PM remnants do not change their positions (Kang et al., 2023).

To quantify the movements of the sperm PM remnants on the zygote surface, we recently used an image-based coordinate normalization (ICN) method that was developed by modifying the particle image velocimetry (PIV) technique, which detects the mobility of a particular object from one time point (T) to the next (T+Δt) through correlation matching. The ICN method revealed that most apical sperm PM remnants move forward with the growing tip of the zygote, whereas the position of lateral remnants does not change during elongation. These observations indicate that zygote growth occurs only in the apical tip region, a characteristic of tip-growing cells (Kang et al., 2023).

During the live imaging of cultured ovules, the position and orientation of each ovule fluctuate as it floats in the liquid culture medium and undergoes cell division. Because these ovule movements are greater in magnitude than the elongation of the zygote, it has been difficult to track changes in the zygote during long-term live imaging. The contour-based coordinate normalization (CCN) method was developed to solve these problems (see poster). In the CCN method, the cell contour at each time point is extracted from time-lapse images, and cell surface points with characteristic curvatures are identified. Next, the movement of these characteristic points is tracked based on correlation matching. Finally, rotational and horizontal shifts are calculated based on principal component analysis (PCA) and are used to normalize the position coordinates of zygotes that had been displaced during time-lapse imaging. The CCN method allows one to trace changes in cell morphology globally in the zygote, from elongation to division. Using this method, we found that the growth rate and tip width of the zygote transiently increase before cell division (Kang et al., 2023). By contrast, the growth rate and thickness of tip-growing root hairs remain constant during elongation (see poster). This means that the mode of zygote cell growth depends on the precise developmental stage – zygotes might have a specific temporal regulatory mechanism that shifts the cell from a tip-growing elongation phase to a preparatory phase immediately prior to asymmetric cell division (Kang et al., 2023).

The cytoskeletons of zygotes and root hairs have a similar F-actin pattern, but very different microtubule orientations. Although microtubules are aligned longitudinally in root hairs, zygotes contain transverse MT rings in the subapical region, which, as discussed above, are important for zygote elongation (Kimata et al., 2016; Hirano et al., 2018). The fern protonema is another type of tip-growing cell, which, like Arabidopsis zygotes but unlike root hairs, undergoes cell division. The protonema also contains transverse MT rings in the subapical region (Murata et al., 1987; Murata and Wada, 1989), suggesting that whether a cell divides or not is related to the presence of MT rings. Alternatively, the differences in microtubular orientation might be related to differences in growth rate and tip shape.

Live-cell imaging of Arabidopsis zygotes has provided new insights into zygote polarization and elongation as well as the dynamics of intracellular components such as the cytoskeleton and organelles. However, various key questions remain unanswered. For example, what is the relationship between the time course of zygote polarization and changes in gene expression? Given that cell cycle progression is associated with the temporal control of various aspects of plant reproduction, it is important to investigate the role of cell cycle regulation in zygote development (Maruyama et al., 2015; Voichek et al., 2023 preprint; Yin et al., 2014). In addition, the molecular mechanisms that regulate polarization remain largely unknown. The regulatory factors involved could be identified using reverse genetic screens based on gene expression profiles before and after fertilization. Furthermore, even though zygotes divide transversely in a wide variety of plant species, suggesting conservation of factors regulating zygote polarization, studies on the establishment of the body axis are still limited to a few model plants. Live-cell imaging and quantitative mathematical analysis, combined with a phylogenetic approach, will help answer these fundamental questions.