https://orcid.org/0000-0001-8046-2608
Abstract
In most flowering plants, the asymmetric cell division of zygotes is the initial step that establishes the apical–basal axis. In the Arabidopsis zygote, vacuolar accumulation at the basal cell end is crucial to ensure zygotic division asymmetry. Despite the importance, it was unclear whether this polar vacuolar distribution was achieved by predominant biogenesis at the basal region or by directional movement after biogenesis. Here, we found that apical and basal vacuolar contents are dynamically exchanged via a tubular vacuolar network and the vacuoles gradually migrate toward the basal end. The mutant of a vacuolar membrane protein, SHOOT GRAVITROPISM2 (SGR2), failed to form tubular vacuoles, and the mutant of a putative vacuolar fusion factor, VESICLE TRANSPORT THROUGH INTERACTION WITH T-SOLUBLE N-ETHYLMALEIMIDE-SENSITIVE FUSION PROTEIN ATTACHMENT PROTEIN RECEPTORS (SNARES) 11 (VTI11), could not flexibly rearrange the vacuolar network. Both mutants failed to exchange the apical and basal vacuolar contents and to polarly migrate the vacuoles, resulting in a more symmetric division of zygotes. Additionally, we observed that in contrast to sgr2, the zygotic defects of vti11 were rescued by the pharmacological depletion of phosphatidylinositol 3-phosphate (PI3P), a distinct phospholipid in the vacuolar membrane. Thus, SGR2 and VTI11 have individual sites of action in zygotic vacuolar membrane processes. Further, a mutant of YODA (YDA) mitogen-activated protein kinase kinase kinase, a core component of the embryonic axis formation pathway, generated the proper vacuolar network; however, it failed to migrate the vacuoles toward the basal region, which suggests impaired directional cues. Overall, we conclude that SGR2- and VTI11-dependent vacuolar exchange and YDA-mediated directional migration are necessary to achieve polar vacuolar distribution in the zygote.
Introduction
Body axis formation is an initial event for a single-celled zygote to produce a multicellular organism. In various flowering plants, the apical–basal axis is already evident owing to zygote polarity, which is marked by the nucleus and large vacuoles in the apical and basal regions, respectively (Jensen 1968, Mansfield et al. 1991, Suzuki et al. 1992). Based on live-cell imaging analysis of Arabidopsis thaliana, we recently revealed that the vacuoles form tubular strands and accumulate at the basal cell end in the zygote (Kimata et al. 2019). This tubular vacuole formation is impaired in the mutant of SHOOT GRAVITROPISM2 (SGR2), which encodes a phospholipase A1-like protein localizing on the vacuolar membrane (Kato et al. 2002, Morita et al. 2002, Kimata et al. 2019). The sgr2 mutant generates large spherical vacuoles, which occupy the apical region; therefore, nuclear migration toward the apical cell tip is disturbed, which results in a more symmetric division of the zygote (Kimata et al. 2019). Moreover, the apical vacuoles in sgr2 were detected at the later embryo stages and disrupted pattern formation was observed; this reflects the importance of polar vacuolar distribution in the zygote in order to complete embryogenesis (Kimata et al. 2019). In spite of these crucial roles, the mechanisms underlying polar vacuole positioning are elusive. It is still unknown whether the unequal distribution in the zygote is predetermined by a predominant vacuole biogenesis at the basal region, e.g. via active production of vacuolar precursors, provacuoles, from the basal endoplasmic reticulum (ER), or by a polar vacuolar migration toward the basal end after biogenesis.
Owing to its putative properties of hydrolyzing phospholipids, SGR2 has been predicted to produce and/or degrade some specific phospholipids on the vacuolar membrane, although these phospholipids are still unidentified (Kato et al. 2002). Various phospholipids are associated with distinct membrane components, such as vesicle docking sites, and play important roles in vacuolar organization (Munnik and Nielsen 2011). For example, phosphatidylinositol 3-phosphate (PI3P) prevents vacuolar membrane fusion; therefore, large connected vacuoles are formed in the presence of Wortmannin (Wm) reagent, which blocks PI3P synthesis (Lee et al. 2008, Zheng et al. 2014b). Moreover, Wm treatment rescues vacuolar defects in the mutant of VESICLE TRANSPORT THROUGH INTERACTION WITH T-SOLUBLE N-ETHYLMALEIMIDE-SENSITIVE FUSION PROTEIN ATTACHMENT PROTEIN RECEPTORS (SNARES) 11 (VTI11), which encodes the vacuolar/prevacuolar compartment Qb-SNARE (Zheng et al. 2014a, 2014b). SNARE proteins facilitate recognition between specific transport vesicles and target membranes, and VTI11 is predicted to promote vacuolar membrane fusion (Sanderfoot et al. 2001, Ebine et al. 2008, Zheng et al. 2014b). The vti11 mutant produces spherical vacuoles in various tissues and exhibits impaired gravitropic response similar to that of sgr2 (Kato et al. 2002, Saito et al. 2011). With this background, in view of the phospholipid-dependent effect, VTI11 might function in the vacuolar shape changes via the SGR2-related pathway.
In addition to the mechanisms underlying vacuolar dynamics, the directional cue to set the vacuolar accumulation site remains obscure. Fertilization activates a mitogen-activated protein kinase (MAPK) signaling pathway in the zygote to trigger the transcriptional cascade, which is crucial to set zygote division asymmetry and contribute to embryo patterning (Bayer et al. 2009, Jeong et al. 2011, Ueda et al. 2017). A mutant of the core component, the mitogen-activated protein kinase kinase (MAPKK) kinase YODA (YDA), exhibits a more symmetric zygotic division (Lukowitz et al. 2004). However, it is still unknown whether and how YDA-mediated signaling directs the intracellular events in the zygote.
In this study, we showed that both vti11 and yda affected the polar vacuolar positioning in the zygote. All the mutants (i.e. sgr2, vti11 and yda) failed to undergo asymmetric zygotic division; however, their vacuolar behaviors differed. These differences were confirmed by various analyses such as the photoconversion of target vacuoles and pharmacological depletion of PI3P. Overall, our results showed that the polar vacuole distribution in the zygote requires SGR2-dependent tubular vacuole formation, VTI11-mediated dynamic rearrangement of vacuolar network and YDA-induced directional vacuolar migration.
Results
vti11 and sgr2 affect the zygotic asymmetric division in a manner different to that of yda
To investigate whether VTI11 contributes to the zygote polarization similar to SGR2 and YDA, we first compared the respective mutant phenotypes corresponding to the asymmetric division of the zygote (Fig. 1A). As the mutants of VTI11 and SGR2, we used well-known respective alleles, sgr2-1 and vti11, both of which were in the Columbia (Col-0) background (Kato et al. 2002). Because the original yda-1 allele was generated in the Landsberg erecta (Ler) background, we used yda-2991, which was isolated among an ethyl methanesulfonate–mutagenized Col-0 strain, Targeting Induced Local Lesions IN Genomes (TILLING) collection, and harbored the mutation at the 644th Arg that changed it to Cys (CGT to TGT) (Till et al. 2003, Lukowitz et al. 2004). The homozygous yda-2991 plants were not segregated from self-pollinated heterozygous plants, while the heterozygous and wild-type (WT) plants were segregated at close to 2:1 ratio [0% (−/−), 67% (±) and 33% (+/+), N = 116], showing that yda-2991 is a strong allele, similar to the original yda-1. Indeed, as reported in yda-1, heterozygous yda-2991 plants produced short zygotes, which divided in a more symmetrical manner, and dwarf embryos, whose suspensors were not recognizable at the globular stage (Fig. 1A and Supplementary Fig. S1a) (Lukowitz et al. 2004).
Comparison of mutant phenotypes associated with zygote elongation and its asymmetric division. (A) Differential interference contrast (DIC) microscopic images of cleared one-cell stage embryos of the WT, sgr2-1, vti11 and yda-2991. Embryos are outlined, and magenta arrowheads indicate the cell division planes. The apically remaining large vacuole is colored in the enlarged images (insets). (B) The ratio of the apical cells divided by basal cell lengths in one-cell stage embryos, denoted as the asymmetric cell division of the zygote. (C) The total length of one-cell stage embryos (sum of apical and basal cell lengths), denoted as zygote elongation (N = 80 for each genotype). Each point represents the individual sample, and the red points represent the 20 embryos, whose ratio of apical to basal daughter cell lengths was lowest among the total 80 embryos produced in yda-2991 heterozygous plants. The letters on the graph indicate significant differences determined by Tukey–Kramer test; P < 0.01. Scale bar: 10 µm.
The one-cell stage embryos of the WT contained small apical cells, whereas sgr2-1, vti11 and yda-2991 exhibited longer apical cells harboring the large vacuoles (Fig. 1A). Therefore, we measured the ratio of apical to basal daughter cell lengths as an inverse proxy for the asymmetry of zygotic division (Fig. 1B) and the total length of one-cell stage embryos as a proxy for zygote elongation (Fig. 1C). Because a quarter of the total embryos in the heterozygous yda-2991 plants seemed to divide in a more symmetric manner (25.9%, N = 437), we estimated that 25% of embryos showing the low ratio of apical to basal daughter cell lengths would be yda-2991 homozygous and used them for the statistical analysis to compare with WT, sgr2-1 and vti11 (Fig. 1B, C). We found that vti11 divided less asymmetrically than the WT (Fig. 1B); however, zygote elongation remained unaffected (Fig. 1C). This feature was strikingly similar to that observed in sgr2-1; however, it was different from that observed in yda-2991, whose zygote was shorter and divided more symmetrically (Fig. 1A–C) (Lukowitz et al. 2004, Kimata et al. 2019). This suggested that the role and/or contribution of VTI11, SGR2 and YDA in zygote polarization are not the same.
sgr2 and vti11 failed to form tubular vacuoles, and yda affected vacuolar distribution polarity
Owing to the presence of large vacuoles in apical cells of sgr2-1, vti11 and yda-2991, we analyzed vacuolar distribution dynamics in the zygotes. We performed time-lapse imaging of the dual-color marker that labels the vacuolar membrane and nucleus (vacuolar membrane/nucleus) (Fig. 2 and Supplementary Movie S1) (Kimata et al. 2019). As reported previously, an unfertilized egg cell of the WT had the nucleus at the apical pole and huge vacuoles in the basal region (Fig. 2A) (Kimata et al. 2019). This polar organization was not affected in sgr2-1, although its vacuoles were more spherical (Fig. 2B) (Kimata et al. 2019). It was similar in vti11, where vacuoles were spherical as reported in the other tissues (Fig. 2C) (Saito et al. 2011). The vacuolar shape and position were unaffected in yda-2991 (Fig. 2D), owing to the gene function associated with the fertilization-triggered signaling pathway (Bayer et al. 2009).
Live-cell imaging of vacuolar dynamics during zygote polarization. (A–H) Two-photon excitation microscopy (2PEM) image of the egg cell (A–D) and the time-lapse observation of the zygote in in vitro-cultivated ovules (E–H) expressing the vacuolar membrane/nuclear marker. WT (A, E), sgr2-1 (B, F), vti11 (C, G) and yda-2991 (D, H) are shown. MIP images are shown, and numbers indicate the time (h:min) from the first frame. Images were obtained at 20-min intervals. Arrowheads indicate nuclei, and brackets indicate the lengths of apical and basal cells. The inset shows an enlarged image of the perinuclear region. Illustrations in the right show a summary of elongating zygotes and one-cell stage embryos of each strain. Scale bars: 10 µm.
After fertilization, the WT and mutants exhibited temporal disorganization, which included cell shrinkage, nuclear localization at cell center, and vacuolar dispersion in the apical and basal cell regions (Fig. 2E–H and Supplementary Movie S1) (Kimata et al. 2016, 2019). Subsequently, the WT formed tubular vacuoles around the nucleus, and the nucleus migrated toward the apical tip during cell elongation along the apical–basal axis (Fig. 2E) (Kimata et al. 2019). Finally, the zygote divided asymmetrically with unequal inheritance of most vacuoles in the basal cell. In sgr2-1, large spherical vacuoles were formed without tubular vacuoles as reported previously (Fig. 2F) (Kimata et al. 2019). The sgr2-1 also generated spotted vacuoles, which were small and isolated from the other vacuoles. The large apical vacuoles prevented the nucleus from reaching the apical tip, which resulted in a more symmetric division. In vti11, the zygote formed spherical and spotted vacuoles as in the sgr2-1; however, vti11 also exhibited elongated vacuoles, which were apparently thinner than those observed in the WT (Fig. 2G). In some cases, these thin tubular vacuoles lumped and formed a tangled structure. In extreme cases, they caused vacuolar membrane accumulation and exhibited brighter fluorescent signals. In the mature zygote, large vacuoles remained in the apical region and blocked nuclear migration toward the apical cell tip, causing a more symmetric division of the zygote as with sgr2-1.
By contrast, yda-2991 formed tubular vacuoles as in the WT; however, cell elongation was impaired and the vacuoles did not polarly position but accumulated at both the apical and basal cell ends (Fig. 2H). Furthermore, the nucleus remained at the cell center, which resulted in a more symmetric division. These different phenotypes suggested that VTI11, SGR2 and YDA might act in specific steps of polar vacuolar positioning.
SGR2 and VTI11 function in tubular vacuole formation and vacuolar rearrangement, respectively
In order to distinguish between the roles of VTI11, SGR2 and YDA, we tracked vacuolar dynamics in these mutants with a shorter time interval and higher spatial resolution. We initially performed conventional time-lapse imaging at a long interval (20 min) to monitor the young zygote until cell division without causing severe laser damage (Supplementary Movie S1) (Kurihara et al. 2017). However, it was not sufficient to track the rapid vacuolar dynamics. Therefore, we used a 1-min interval and focused on the perinuclear region with higher magnification (1.7 times the xy resolution) (Fig. 3A–D and Supplementary Movie S2). In the WT, tubular vacuoles were elongated from apical vacuoles, fused with basal vacuoles and/or other tubules, and separated from each other (Fig. 3A). In sgr2-1, the apical spherical vacuoles did not form tubular vacuoles and only small spotted vacuoles were detected in the perinuclear region (Fig. 3B). In contrast, in vti11, the apical vacuoles formed tubular vacuoles, and they were connected with each other (Fig. 3C). But these tubular vacuoles were apparently thinner, and were not properly separated, resulting in a tangled structure. Three-dimensional (3D)-reconstructed images of vacuolar lumen marker showed almost no signals in the perinuclear region of sgr2-1 and vti11 (Fig. 3E, and Supplementary Fig. 1b) (Kimata et al. 2019), which showed that the spotted vacuoles in sgr2-1 and thin vacuoles in vti11 contained less lumen space. Taken together, these results showed that SGR2 and VTI11 function in tubular vacuole formation and its dynamic rearrangement, respectively, and that both are necessary to form vacuolar network spanning from the apical to basal regions.
Vacuolar dynamics in perinuclear regions of the zygote. (A–D) Time-lapse observation of the zygote in in vitro cultivated ovules expressing the vacuolar membrane/nuclear marker. The dynamics of perinuclear vacuoles in WT (A), sgr2-1 (B), vti11 (C) and yda-2991 (D) are shown. MIP images are shown, and numbers indicate the time (h:min) from the first frame. Images were obtained at 1-min intervals. Orange arrows and yellow arrowheads indicate elongating and fused vacuoles, respectively. Brackets show the gaps between separated vacuoles. (E) 3D images of vacuoles in mature zygote of WT, sgr2-1 and vti11. Images were generated using the serial optical sections of 2PEM images of vacuolar lumen marker. Open brackets show the perinuclear area. Scale bars: 10 µm.
By contrast, yda-2991 displayed a connected vacuolar network, which was flexibly rearranged with elongation, fusion and separation of tubular vacuoles, as in the WT (Fig. 3D). This suggested a role for YDA in general cell polarity and not in the vacuolar shape change, which is in agreement with the impaired cell elongation in yda-2991 (Fig. 1C).
Exchange of apical and basal vacuolar contents was impaired in sgr2 and vti11
In the WT, the tubular networks connecting the apical and basal vacuoles were dynamically rearranged (Fig. 3) and the vacuoles were positioned at the basal cell end (Fig. 2). This was not observed in sgr2-1 and vti11, which raised the idea that vacuolar rearrangement might dynamically exchange the lumenal contents of apical and basal vacuoles to gradually move the vacuoles toward the basal end. To test this hypothesis, we generated a photoconvertible marker of the vacuolar lumen and combined it with a nuclear marker (photoconvertible vacuolar lumen/nucleus) and analyzed the exchange of the photoconverted apical vacuoles and intact basal vacuoles (Fig. 4). In order to specifically label the apical vacuole, a small area (0.5 × 0.5 µm) was irradiated with ultraviolet (UV) spotlight to convert the fluorescence from green to red (Fig. 4A). We observed the zygotes immediately before and at 0, 20 and 40 min after photoconversion (Fig. 4B–D). Foremost, the success of photoconvertion was confirmed by the reduced ratio of the green to red fluorescence of the target vacuoles in the WT [1.21 (before) to 0.38 (0 min)] (Fig. 4B). Next, the vacuoles in the whole cell were observed as white (merging of green and red), and the ratio was gradually increased to the value comparable before irradiation [0.96 (20 min) and 1.26 (40 min) compared to 1.21 (before)], which demonstrated the exchange of the photoconverted and non-converted vacuoles. In sgr2-1 and vti11, the photoconverted apical vacuoles remained red and the green/red ratios were not restored [0.22 (0 min) to 0.19 (40 min) in sgr2-1 and 0.60 (0 min) to 0.60 (40 min) in vti11] (Fig. 4C, D).
Photoconversion of the apical vacuoles and time-course observation. (A) The scheme for photoconverting the apical vacuole. (B–D) 2PEM images of the zygotes expressing the photoconvertible vacuolar lumen/nucleus marker before and 0, 20 and 40 min after the pinpoint UV irradiation to the apical vacuoles. Purple lightning marks show irradiated sites. WT (B), sgr2-1 (C) and vti11 (D) are shown. Center plane images are shown. Scale bar: 10 µm.
These results revealed active exchange of the apical and basal vacuolar contents in the WT and their separation in sgr2-1 and vti11, which was in agreement with the presence and absence of the perinuclear vacuolar tubules. It reflects the importance of the tubular the network in connecting apical and basal vacuoles. Furthermore, the exchange of existing vacuoles demonstrated that polar vacuole distribution is not determined by biogenesis but by the directional vacuolar migration after biogenesis.
Wm treatment rescued vacuolar defects of vti11, but not of sgr2 and yda
The vti11 generated tangled vacuoles in the zygote, which is likely due to the impaired vacuolar separation. It seemed to be at odds with the reports that VTI11 is required for vacuolar membrane fusion in various cell types (Zheng et al. 2014b). Therefore, we assessed whether the zygotic defects of vti11 depend on vacuolar membrane processes. We used Wm reagent, which depletes PI3P from vacuolar membrane, and thus, it rescues vti11 defects in hypocotyl (Vermeer et al. 2006, Zheng et al. 2014b). In the presence of Wm, vti11 zygotes exhibited proper asymmetric division without inheriting large apical vacuoles (Fig. 5A, B). This ‘rescue effect’ was reversed upon exogenous application of PI3P and its carrier (‘Wm-PI3P’ in Fig. 5C, D), whereas the combination of Wm and empty carrier exhibited restoration (‘Wm-carrier’ in Fig. 5C, D). This antagonistic rescue effect of vti11 was strikingly similar to that observed in the hypocotyl (Vermeer et al. 2006, Zheng et al. 2014b); this suggests that VTI11 acts in the zygote via its SNARE activity, which is balanced with PI3P on the vacuolar membrane. Furthermore, Wm treatment did not restore the defects of sgr2-1 and yda-2991 (Fig. 5A, B), which confirms their specific roles in polar vacuolar distribution.
Effects of Wm-induced PI3P depletion on zygote division asymmetry. (A) 2PEM images of the one-cell stage embryos of the indicated strains expressing the vacuolar membrane/nuclear marker after exposure to DMSO and Wm for 24 h. MIP images are shown. Brackets indicate the lengths of the apical and basal cells. (B) The ratio of apical cells divided by basal cell lengths in one-cell stage embryos of DMSO- and Wm-treated zygotes for 24 h. Each point represents the individual sample, and red points represent the 25% embryos, whose ratio was highest among the total embryos produced in yda-2991 heterozygous plants (N ≥ 10 for each genotype). (C) 2PEM images of the one-cell stage embryos of vti11 expressing the vacuolar membrane/nuclear marker after exposure to DMSO and Wm in combination with PI3P and its carrier or only the carrier for 24 h. (D) The ratio of apical cells divided by basal cell lengths in one-cell stage embryos of DMSO- and Wm-treated vti11 zygotes in combination with the carrier and PI3P for 24 h (N ≥ 23 for each condition). The letters on the graph indicate significant differences determined by Tukey–Kramer test; P < 0.05. Scale bars: 10 µm.
Discussion
Our findings demonstrated the detailed dynamics of polar vacuolar distribution in the zygote. Further, we identified that SGR2, VTI11 and YDA are required for specific processes, i.e. tubular vacuole formation, vacuolar rearrangement and polarity setting, which are all necessary for the directional vacuolar migration and the consequent asymmetric division of the zygote. Our findings provide insights into the mechanisms underlying vacuolar shape change and highlight the important role of vacuolar dynamics in cellular asymmetry in plant development.
In WT zygotes, tubular vacuoles emerged and flexibly rearranged to form a vacuolar network. In contrast, sgr2 and vti11 generated spotted and spherical vacuoles and failed to form tubular networks. A previous study showed that the total surface area of zygotic vacuoles is reduced in sgr2; however, the vacuolar lumen volume is not affected (Kimata et al. 2019). Therefore, SGR2 would be required to fuse small empty vacuoles to large ones and subsequently increase the vacuolar membrane amount to elongate tubules. As VTI11 has been predicted to promote membrane fusion (Zheng et al. 2014a, 2014b), it is likely that it contributes to this membrane supply. However, SGR2 and VTI11 would have different sites of action as only vti11 defects were restored by Wm-induced PI3P depletion, which suggests that PI3P is not targeted by the phospholipase A1-like protein encoded by SGR2 (Morita et al. 2002).
Furthermore, in contrast to sgr2, vti11 formed thinner vacuolar tubules, which contained insufficient lumen and excess membrane. What are the causes of these vti11-specific structures? A possible explanation includes one of the vacuole biogenesis routes for vacuolar precursors, provacuoles, which originate from ER and maturate with various cargo proteins supplied via fusion with Golgi-derived vesicles and/or already developed vacuoles (Uemura and Ueda 2014, Viotti 2014). For example, VACUOLAR PROTEIN SORTING45 (VPS45), which interacts with SNARE complex at the trans-Golgi network (TGN), is required to supply various vacuolar cargos, such as soluble storage proteins, and thus VPS45 RNAi strain fails to generate large vacuoles (Zouhar et al. 2009). A similar pathway acts in the formation of the connected tubular vacuoles in the root meristematic cells, in which an empty loop emerges as a compartment in the tubular vacuoles and is rapidly filled that results in tubular vacuole extension (Viotti et al. 2013). The VTI11 encodes Qb-SNARE, which localizes in vacuoles and TGN (Uemura et al. 2004). Therefore, VTI11 might mediate membrane fusion between immature vacuoles and Golgi-derived vesicles to fill the lumen; thus, the failure in this process could generate thinner vacuoles with less content in the vti11 zygote.
The vti11 zygote also generated tangled vacuoles, which occurred apparently due to the reduced separation of tubular vacuoles. At present, we cannot comment on how this phenotype was caused, because the exact molecular function of VTI11 and the vacuolar separation mechanism in plants are both obscure. The antagonistic effect of Wm and PI3P on vti11 implies that this phenotype is a result of impaired vacuolar membrane fusion. If so, we can draw hints from the regulatory model of yeast, where vacuolar fusion is a prerequisite for proper tubular vacuole formation, which is per se necessary for the subsequent fission (Röthlisberger et al. 2009). In this model, vacuolar fusion triggers the association of dynamin-related protein, Vps1, to vacuolar docking sites. Next, Vps1 constricts the vacuoles to adjust tubular strands at appropriate diameters for Dnm1, which executes vacuolar fission (Röthlisberger et al. 2009). As an alternative, simpler hypothesis, VTI11 might mediate vesicle fusion to the vacuoles and provide some vacuolar separation factors, such as yeast Vps10, which recruit retromer and Vps1 to split small vacuolar compartments into recycling vesicles (Arlt et al. 2015). Therefore, it would be important to identify the factors that mediate vacuolar separation in plants and track their localization during vacuolar rearrangement in Arabidopsis zygotes.
Additionally, we found that the apical and basal vacuolar contents were dynamically exchanged via tubular network and gradually accumulated at the basal cell end. We previously showed that a longitudinal F-actin array is necessary for both tubular vacuole formation and polar vacuole distribution in the zygote (Kimata et al. 2019). Therefore, we inferred that fused vacuoles migrate toward basal end along F-actin cables. It is likely that YDA-mediated signaling would set this direction because yda zygotes failed to unequally distribute the vacuoles into two daughter cells without any detectable defects in vacuolar network formation. The zygote elongation along apical–basal axis was also affected in yda, which suggests that YDA impacts cell polarity in general. In addition, YDA might have other roles in vacuolar behavior, e.g. in vacuolar expansion, because the vacuole volume seemed reduced in the short zygote of yda.
Further, YDA drives the MAPK cascade under the paternally inherited pseudokinase, SHORT SUSPENSOR (SSP) and activates several transcription factors, such as WRKY2, which induce de novo transcription to support apical–basal axis formation (Bayer et al. 2009, Lukowitz et al. 2004, Ueda et al. 2017). Similar to yda, the ssp and wrky2 mutants did not exhibit zygote elongation and asymmetric cell division, which suggests that cell polarity is determined by unidentified downstream factors; e.g., actin-binding proteins and small GTPases, which direct F-actin assembly in polar growth of root hairs and pollen tubes (Szymanski and Staiger 2018). Because the polarity regulators of the zygote are rarely known, future identification of the downstream genes will be necessary to link fertilization signals to polar vacuolar distribution. Furthermore, the live-cell imaging of these factors with various vacuolar trafficking components such as the other SNARE proteins and diverse phospholipids (Simon et al. 2014, Uemura and Ueda 2014) will demonstrate how YDA, SGR2 and VTI11 achieve polar vacuole migration for the asymmetric division of the zygote.
Materials and Methods
Plant strains and growth conditions
All Arabidopsis lines were in a Col-0 background. The sgr2-1 and vti11 mutants were previously described (Kato et al. 2002). The yda-2991 mutant was isolated from a TILLING collection in the Col-0 background by W. Lukowitz, and the erecta-105 mutation, which was included in the TILLING plants, was removed from yda-2991 (Torii et al. 1996, Till et al. 2003). For genotyping of yda-2991, a 342-bp fragment was PCR-amplified using gene-specific primers (5ʹ-GAGTTCTTTACAGGTTCCTG-3ʹ and 5ʹ-CTTGCATGTCCAATGTCCTG-3ʹ) and digested with MluCI enzyme (R0538; New England Biolabs). The WT generates three fragments (274, 61 and 7 bp), whereas yda-2991 produces four fragments (150, 124, 61 and 7 bp) owing to an additional cut site at the mutated position. Plants were grown at 18–22°C under continuous light or long-day conditions (16-h light/8-h dark).
Plasmid construction and transgenic lines
The vacuolar membrane marker (EC1p::VHP1-mGFP), nuclear marker (EC1p::H2B-tdTomato and DD22p::H2B-mCherry) and vacuolar lumen marker (EC1p::SP-mTur2-CTPP) were previously described (Kimata et al. 2019).
The photoconvertible vacuolar lumen marker was EC1p::SP-Kaede-CTPP (coded as MU2420), containing the 463-bp EGG CELL1 (EC1) promoter (Sprunck et al. 2012), photoconvertible Kaede that fused to the signal peptide (SP) sequence of endoxyloglucan transferase-A1 (EXGT-A1; AT2G06850) at the N-terminus and to a vacuolar sorting signal COOH-terminal propeptide (CTPP) (Vitale and Raikhel 1999) at the C-terminus, and nopaline synthase terminator in a pMDC123 binary vector (Curtis and Grossniklaus 2003). This EC1p::SP-Kaede-CTPP was combined with the nuclear marker (EC1p::H2B-tdTomato and DD22p::H2B-mCherry) (Kimata et al. 2019).
The constructs were transformed into Arabidopsis using the floral dip method (Clough and Bent 1998), and at least 10 independent plants were observed in the T1 generation to check the fluorescent pattern. Then at least two independent lines were established for the observation using two-photon excitation microscopy (2PEM) in T2 or T3 generation.
Zygote imaging and inhibitor treatment
Cleared embryos were observed with an upright microscope (AxioImager A2; Zeiss) as previously described (Ueda et al. 2017). The zygote live-cell imaging was performed using a laser-scanning inverted microscope (A1R MP; Nikon) equipped with a Ti:sapphire femtosecond pulse laser (Mai Tai DeepSee; Spectra-Physics) as previously described (Gooh et al. 2015, Kurihara et al. 2017, Ueda et al. 2020). Time-lapse images were taken every 20 min at 31 z-stacks (Fig. 2) or 1 min at 3 z-stacks (Fig. 3) with 1-µm intervals. For photoconversion, 402.4 nm laser was irradiated to a small square region (0.5 µm × 0.5 µm).
For chemical treatment, Wm (W1628; Sigma) was dissolved in dimethyl sulfoxide (DMSO) and added to the ovule cultivation medium at 0.1 μM in 0.1% DMSO, as previously described (Nambo et al. 2016). The diC8-PI(3)P (P-3008; Echelon Biosciences) and Unlabeled Shuttle PIP Carrier 3 (carrier) (P-9C3; Echelon Biosciences) were dissolved in distilled water at 2 mM and mixed at equivalent volume following manufacturer’s instruction (Echelon Biosciences). The mixture of PI3P and carrier was added to the ovule cultivation media at 1 μM PI3P.
Image processing and quantitative analysis
Maximum Intensity Projection (MIP) and image processing were performed using NIS-Elements AR 4.10 software (Nikon) or Fiji (https://fiji.sc/). Fiji was also used to measure the apical and basal cell length. To evaluate connectivity of the vacuoles, we time-sequentially measured the ratio of the maximum intensity in green channel to the maximum intensity in red channel in the targeted vacuole using Fiji. Three-dimensional reconstruction of the vacuolar lumen was performed using Imaris (version 9.7.2; Bitplane, Belfast, UK).
Supplementary Data
Supplementary data are available at PCP online.
Data Availability
The data underlying this article are available in the article and in its online supplementary material.
Funding
Japan Society for the Promotion of Science {JSPS; Grant-in-Aid for JSPS Research Fellow (JP19J30006 to Y.K.), Grant-in-Aid for Young Scientists (JP21K15117 to Y.K.), Grant-in-Aid for Scientific Research on Innovative Areas [19H04859, 19H05670 and 19H05676 to M.U.; JP16H06465, JP16H06464 and JP16K21727 to T. Higashiyama; JP18H05492 and JP16H06280 (Advanced Bioimaging Support) to T. Higaki], Grant-in-Aid for Scientific Research (B) (JP19H03243 to M.U. and JP20H03289 to T. Higaki) and Grant-in-Aid for Challenging Exploratory Research (JP19K22421 to M.U.)}, Suntory Rising Stars Encouragement Program in Life Sciences (SunRiSE) to M.U. and Toray Science Foundation to M.U. (20-6102).
Acknowledgements
We thank Tomomi Yamada, Yumi Kuwabara and Terumi Nishii for technical support and Daisuke Kurihara and Yoko Mizuta for helpful discussion. We appreciate Wolfgang Lukowitz, Miyo Terao Morita and Takehide Kato for providing mutant seeds and Shoji Segami and Masayoshi Maeshima for providing VHP1 marker. The microscopy was supported by the Live Imaging Centre at the Institute of Transformative Bio-Molecules (WPI-ITbM) of Nagoya University. We would like to thank Editage (www.editage.com) for English language editing.
Disclosure
The authors have no conflicts of interest to declare.
References
Author notes
Hikari Matsumoto and Yusuke Kimata contributed equally to the article.
