Atmospheric methane levels increased over geological formations in Northern Siberia following a heat wave in 2020, suggesting that permafrost thaw releases methane from reservoirs.
Image credit: Wikimedia Commons.
Edited by Tetsuya Higashiyama, Nagoya University, Nagoya, Japan, and accepted by Editorial Board Member June B. Nasrallah November 9, 2020 (received for review July 23, 2020)
Significance
Flowering plants have evolved a unique double-fertilization process along with an actin filament (F-actin)-based gamete nuclear migration mechanism. However, how dynamic F-actin movement is controlled in the female gametophytic cells remains unclear. We identified that the movement of F-actin is promoted via an ARP2/3-independent WAVE/SCAR-signaling pathway. We also discovered that the plant class XI myosin XI-G has a function involved in the active movement of F-actin required for sperm nuclear migration, which is different from the canonical myosin function as a cargo transporter. These breakthroughs also provide us with opportunities to further understand how flowering plants control double fertilization and plant cytoskeleton dynamics.
Abstract
After eukaryotic fertilization, gamete nuclei migrate to fuse parental genomes in order to initiate development of the next generation. In most animals, microtubules control female and male pronuclear migration in the zygote. Flowering plants, on the other hand, have evolved actin filament (F-actin)-based sperm nuclear migration systems for karyogamy. Flowering plants have also evolved a unique double-fertilization process: two female gametophytic cells, the egg and central cells, are each fertilized by a sperm cell. The molecular and cellular mechanisms of how flowering plants utilize and control F-actin for double-fertilization events are largely unknown. Using confocal microscopy live-cell imaging with a combination of pharmacological and genetic approaches, we identified factors involved in F-actin dynamics and sperm nuclear migration in Arabidopsis thaliana (Arabidopsis) and Nicotiana tabacum (tobacco). We demonstrate that the F-actin regulator, SCAR2, but not the ARP2/3 protein complex, controls the coordinated active F-actin movement. These results imply that an ARP2/3-independent WAVE/SCAR-signaling pathway regulates F-actin dynamics in female gametophytic cells for fertilization. We also identify that the class XI myosin XI-G controls active F-actin movement in the Arabidopsis central cell. XI-G is not a simple transporter, moving cargos along F-actin, but can generate forces that control the dynamic movement of F-actin for fertilization. Our results provide insights into the mechanisms that control gamete nuclear migration and reveal regulatory pathways for dynamic F-actin movement in flowering plants.
Flowering plants have evolved a unique double-fertilization process. Two sperm cells fuse with two female gametophytic cells, the egg and central cells within the ovule (Fig. 1A), giving rise to the embryo and endosperm, respectively (1, 2). Sperm cells in flowering plants are nonmotile and delivered in close proximity to the egg and central cells by the pollen tube. After gamete cell fusion, in most animals, both male and female pronuclei move toward each other within the fertilized egg for gamete nuclear fusion, or karyogamy. Pronuclei movement is regulated by microtubules that assemble the sperm aster from the centrosome (3, 4). By contrast, flowering plants have lost the centrosome, and have established actin filament (F-actin)-based sperm nuclear migration systems for successful double fertilization (5⇓–7). Prior to fertilization, both the egg and central cells form a mesh-like structure of F-actin that shows constant inward movement from the plasma membrane to the center of the cell, where the nucleus resides (Fig. 1A) (5, 8, 9). This meshwork movement begins prior to gamete cell fusion and continues until the completion of karyogamy (5). In Arabidopsis thaliana (Arabidopsis) and Oryza sativa (rice), F-actin aster-like structures are formed surrounding the sperm nuclei just after the sperm nuclei enter into the female gametophytic cells. The transfer of the actin aster-sperm nucleus complex coincides with the constant F-actin meshwork movement (5, 9). The movement of the sperm nucleus by F-actin is consistent in Nicotiana tabacum (tobacco) and Zea mays (maize), and disruption of F-actin arrests sperm nuclear migration (5⇓–7). These results demonstrate that coordinated F-actin meshwork inward movement plays an essential role in sperm nuclear migration of flowering plants.
Fig. 1.
F-actin dynamics in the central cell is WAVE/SCAR dependent, but ARP2/3 complex independent. (A, Top) Z-projected confocal image of the central cell F-actin (cyan, proFWA::Lifeact:Venus), central cell nucleus (yellow, proFWA::H2B:mRuby2), and autofluorescence (magenta) marking the central cell border. (A, Bottom) Schematic diagram of the mature Arabidopsis ovule. Arrows indicate the direction of central cell F-actin movement from the plasma membrane periphery to the nucleus. (B–D, F, and G) Time-lapse (1-min interval, marked by five different colors) stacks of Z-projected central cell F-actin images of the mock treatment (B), wiskostatin (10 µM for 1h incubation) treatment (C), scar2-1 mutant (D), CK-666 (200 µM for 1 h incubation) treatment (F), and the arp2-1 mutant (G). Dashed circles indicate the position of the central cell nucleus. F-actin marked by different colors denotes F-actin movement, whereas white color resulting from overlapping of all colors represents less or no movement. (E) The transcriptional activity of the Arabidopsis ARP2 promoter is visualized by proARP2::H2B:Clover (green). Arrow points to the central cell nucleus and autofluorescence marks the central cell border. (H) Mean velocity of F-actin dynamics in the central cell (**P < 0.001; ns, not significant; Tukey-Kramer HSD test). The box spans first and third quartiles, and the line inside the box shows the median. Bars on the top and bottom represent the maximum and minimum values. (Scale bar, 20 µm.)
F-actin meshwork movement in the female gametophytic cell requires the constant formation of F-actin at the plasma membrane. The rate of F-actin formation can be controlled by regulators that nucleate actin monomers to initiate new filaments, control polymerization during elongation, and prevent disassembly of F-actin (10). In plants, membrane-associating small GTPase-signaling proteins, Rho-GTPase of Plants (ROPs), facilitate cell morphogenesis by controlling actin polymerization (11, 12). The Wiskott–Aldrich syndrome protein family verprolin-homologous/suppressor of the cAMP receptor (WAVE/SCAR) family are effector proteins that directly interact with ROPs and promote actin nucleation (13, 14). The WAVE/SCAR complex is the main activator of the F-actin regulatory ACTIN RELATED PROTEIN 2/3 (ARP2/3) protein complex (15, 16). ARP2/3 directly promotes polymerization of branched actin filaments from the sides of preexisting actin filaments and forms a highly dense branched F-actin network (17, 18). Phenotypes of functionally null single or combinatorial mutants of any of these WAVE/SCAR-ARP2/3 components have been intensively studied in trichomes (15, 19⇓–21), and the lack of intermediate phenotypes indicates that the WAVE/SCAR and ARP2/3 protein complexes constitute the sole pathway that controls the trichome morphology via F-actin regulation (14, 19). The involvement of WAVE/SCAR in sperm nuclear migration has been suggested in the egg and central cells of tobacco and maize using wiskostatin, an inhibitor of WAVE/SCAR activity (7). However, it still remained open whether the WAVE/SCAR activity is mediated through the ARP2/3 pathway, and the genetic data confirming the role of WAVE/SCAR-ARP2/3 in fertilization were still missing.
The Arabidopsis central cell is more than five times larger than the egg cell. This large cell size allows us to visualize F-actin dynamics in detail (4, 5), thus providing an excellent platform to understand the dynamics of plant fertilization. ROP8 is specifically expressed in the central cell and promotes the assembly of F-actin at the plasma membrane, maintaining the constant F-actin meshwork movement (5). Several pharmacological analyses identified putative factors, such as the F-actin motor proteins, myosins, that can control the dynamic F-actin meshwork movement in the female gametophytic cells (5, 7, 9). However, apart from the genetically confirmed ROP8 involvement in the Arabidopsis central cell, it remains largely unknown how F-actin movement for sperm nuclear migration is regulated (5). To identify what factors and pathways are involved in F-actin movement for sperm nuclear migration in flowering plants, we performed genetic and pharmacological analyses with live-cell confocal microscopy imaging in Arabidopsis. We show that WAVE/SCAR and class XI myosin, but not the ARP2/3 complex, play important roles in sperm nuclear migration. Insights into the ARP2/3-independent WAVE/SCAR pathway as well as the role of class XI myosin in the female gametophyte will facilitate further understanding of the mechanism of plant fertilization and cytoskeleton dynamics in flowering plants.
Results
F-Actin Meshwork Movement in the Female Gametophytic Cell Is SCAR2 Dependent, but ARP2/3 Independent.
F-actin meshwork in the Arabidopsis central cell is initiated at the plasma membrane by ROP8 and moves to the nucleus for sperm nuclear migration (5). In plants, ROPs promote F-actin nucleation by interacting with WAVE/SCAR (13, 14). To investigate the involvement of WAVE/SCAR, we applied wiskostatin (22), a small molecule that inhibits WAVE/SCAR activity, to dissected Arabidopsis ovules and examined F-actin meshwork movement in the central cell. Wiskostatin stabilizes the native auto-inhibitory interaction between the GTPase-binding domain and the VCA (verprolin homology) domain of WAVE/SCAR proteins (22, 23). The movement of the central cell F-actin meshwork in the F-actin marker line (proFWA::Lifeact:Venus) (5) was impaired when treated with 10 µM wiskostatin (Fig. 1 B, C, and H, and Movie S1), suggesting that the WAVE/SCAR activity is required for F-actin dynamics in the central cell.
Among Arabidopsis SCARs, SCAR2, and SCAR4 appear to be expressed in the central cell with SCAR2 expression being higher (24, 25). We observed sperm nuclear migration phenotypes in the wild-type (WT), scar2-1 (26), scar4-1 (13), and scar2-1;4–1 double-mutant plants by pollinating them with a sperm-specific histone marker line (proHTR10::HTR10:mRFP1) (Fig. 2 and SI Appendix, Fig. S1) (27). After sperm release to the ovule (Fig. 2A), sperm nuclei, containing condensed chromatin, start migrating toward the nucleus of the egg and central cells (Fig. 2B). Sperm nuclear migration is a rapid event that is completed within 5 min (5) and then followed by karyogamy and sperm chromatin decondensation (Fig. 2C). Sperm chromatin decondensation starts only after successful karyogamy (5, 28, 29). In scar2-1 and scar2-1;4–1, 33.9 and 35.8% of ovules showed no or partially decondensed sperm chromatin in the central cell, respectively (Fig. 2 D and E and SI Appendix, Fig. S1). We did not observe this decondensation-delay phenotype in either WT or scar4-1 (Fig. 2E). Consistently, the central cell F-actin meshwork movement in scar2-1 and scar2-1;4–1 was significantly slower than that of WT or scar4-1 (Fig. 1 B, D, and H; SI Appendix, Fig. S2; and Movie S1). These results show that SCAR2 is the main SCAR factor that controls the central cell F-actin meshwork movement important for sperm nuclear migration.
Fig. 2.
Sperm nuclear migration is delayed in scar2-1 and xi-g central cells. (A–D) Representative images of sperm chromatin dynamics: (A) Sperm cells just released from the pollen tube into the ovule. (B) Sperm nuclei starting to move toward the central cell and egg cell nuclei. (C) Sperm chromatin decondensed in both the central cell and egg-cell nuclei. (D) Delayed karyogamy and sperm chromatin decondensation observed in the scar2-1 central cell. Note that sperm chromatin became fully decondensed in the egg cell nucleus while sperm chromatin remained condensed in the central cell. This delayed phenotype was not observed in WT. Sperm chromatin was visualized by the sperm-specific histone marker proHTR10::HTR10:mRFP1. Arrows and arrowheads point to the condensed and decondensed sperm chromatin, respectively. Dashed circles indicate the position of the central cell nucleus. Autofluorescence of the central cell border was also visualized. (E) Status of sperm chromatin 9 h after pollination. Stages A–D are shown in A–D, respectively. (Scale bar, 20 µm.)
The activation of the ARP2/3 complex through the WAVE/SCAR pathway is required to induce cellular actin nucleation (14⇓–16). To investigate the involvement of the ARP2/3 complex, the effect of an ARP2/3 complex inhibitor, CK-666 (30), on the central cell F-actin meshwork movement was examined. CK-666 (40 to 200 µM) stabilizes the inactive state of the complex, blocking the movement of the ARP2 and ARP3 subunits into the activated filament-like conformation (30, 31). The 200-µM CK-666 application altered the F-actin orientation to perpendicular to the long axis of the cell in the cotyledon pavement cell (SI Appendix, Fig. S3 A and B), which is typical of ARP2/3 mutants (SI Appendix, Fig. S4) (20). Surprisingly, 200 µM CK-666 did not affect F-actin meshwork movement in the central cell even after a 1-h incubation (Fig. 1 B, F, and H, and Movie S1). These results suggest that the ARP2/3 complex is not involved in F-actin meshwork movement in the central cell. To genetically confirm that ARP2/3 is dispensable for sperm nuclear migration, we investigated the fertilization phenotype in the arp2-1 (20), dis2-1 (arpc2) (32), and arpc4-t2 (33) mutants. None of these mutants showed significant difference in either F-actin meshwork movement (Fig. 1 B, G, and H; SI Appendix, Fig. S2; and Movie S1) or sperm nuclear migration (Fig. 2E and SI Appendix, Fig. S1) compared to WT. Gene expression data indicate that ARP2 is expressed in the central cell (24, 25), and, indeed, the ARP2 promoter activated expression in the central cell (Fig. 1E). These results show that, although ARP2 is expressed in the central cell, ARP2/3 is not involved in fertilization and provide genetic evidence that SCAR2 regulates F-actin dynamics in the central cell through an ARP2/3-independent pathway.
The Class XI Myosin XI-G Plays a Major Role in the Active Movement of F-Actin Meshwork in the Female Gametophytic Cell for Fertilization.
The 2, 3-butanedione monoxime (BDM) inhibits myosin activity by blocking an ATPase activity of the myosin superfamily (5, 7, 9). The application of BDM to Arabidopsis ovules arrests F-actin meshwork movement (5), indicative of myosin involvement. The class XI myosin gene XI-G shows relatively high expression compared to other class XI myosins in the Arabidopsis central cell (25, 24), and the XI-G promoter activated transcription in the central cell and synergid cells of the female gametophyte (Fig. 3E). The xi-g knockout mutant (34) displayed significantly slower F-actin meshwork movement in the central cell compared to WT (Fig. 3 A, B, and F, and Movie S2). Similar to scar2-1 (Fig. 2E), 40% of xi-g ovules had no or partially decondensed sperm chromatin in the central cell (Fig. 2E and SI Appendix, Fig. S1). These data genetically show that the class XI myosin XI-G plays a major role in F-actin dynamics for sperm nuclear migration in the central cell.
Fig. 3.
The class XI myosin XI-G is involved in F-actin meshwork movement in the Arabidopsis central cell. (A–D) Time-lapse (1-min interval, marked by five different colors) stacks of Z-projected central cell F-actin images of the mock treatment (A), the xi-g mutant (B), 20 mM BDM treatment (C), and 50 mM BDM treatment (D). Dashed circles indicate the position of the central cell nucleus. F-actin marked by different colors denotes F-actin movement, whereas white color resulting from overlapping of all colors, represents less or no movement. (E) The transcriptional activity of the Arabidopsis XI-G promoter is visualized by proXI-G::H2B:Clover (green). Autofluorescence marks the central cell wall; the arrow and arrowheads point to the central cell nucleus and synergid nuclei, respectively. (F) Mean velocity of F-actin dynamics in the central cell. Levels not connected by the same letter (a–c) are significantly different (P < 0.01, Tukey-Kramer HSD test). The box spans first and third quartiles, and the line inside the box shows the median. Bars on the top and bottom represent the maximum and minimum values. (G) The orientation of F-actin in the central cell was evaluated by measuring the angles of F-actin cables (shown in red) made with a line radiating from the center of the central cell nucleus (shown as dashed lines). Black dots represent individual angle data and violin shapes show the kernel probability densities (**P < 0.001; Tukey-Kramer HSD test). (Scale bar, 20 µm.)
F-Actin Meshwork Movement Is Controlled by a Noncanonical Function of the Myosin.
The application of 20 mM BDM did not cause any change in F-actin meshwork movement (Fig. 3 A, C, and F, and Movie S2); however, a 50-mM BDM application immediately stopped F-actin meshwork movement (Fig. 3 A, D, and F, and Movie S2). Myosin XI generates the power for filament buckling and straightening by sliding antiparallel actin filaments and/or translocating actin filaments along membranes (35, 36). We also checked F-actin structures in the presence of BDM and observed that actin filaments were straightened and parallel with each other in a 50-mM BDM, but not in a 20-mM treatment (Fig. 3 A, C, D, and G; SI Appendix, Fig. S5; and Movie S2). These data further support that 50 mM BDM is required for the clear observation of myosin inhibitory function in F-actin meshwork movement in the Arabidopsis central cell.
In contrast to the lack of effect on F-actin meshwork movement with 20 mM BDM in the central cell (Fig. 3 A, C, and F, and Movie S2), 20 mM or less BDM can inhibit the myosin function as a cargo transporter in tobacco leaf and maize root apex cells (7, 37). To investigate the effect of BDM on organelle movement in the Arabidopsis central cell, we monitored mitochondrial dynamics using the Arabidopsis central cell mitochondrial marker line proDD65::coxlV:GFP (38). Mitochondrial motility in the central cell was reduced immediately with a 20-mM BDM application compared to the mock (Fig. 4 and Movie S3). At 20 mM, BDM does not affect F-actin meshwork movement (Fig. 3F), showing that F-actin meshwork moves without organelle movement in the central cell. We also found that mitochondrial motility was not affected in the xi-g, scar2-1, or arp2-1 central cell (Fig. 4, SI Appendix, Fig. S6, and Movie S3). The central cell F-actin meshwork movements in xi-g and scar2-1 are greatly reduced (Figs. 1H and 3F). Taken together, these results indicate that F-actin meshwork movement in the central cell is independent of organelle movement and that XI-G controls F-actin meshwork movement through a noncargo transport function of the myosin.
Fig. 4.
The 20 mM BDM affects mitochondrial movement in the Arabidopsis central cell. (A–D) Time-lapse (1-min interval, marked by five different colors) stacks of Z-projected central cell mitochondrial movement images of the mock treatment (A), 20 mM BDM treatment (B), 50 mM BDM treatment (C), and in the xi-g mutant (D). Mitochondria marked by different colors denote movement, whereas white color resulting from overlapping of all colors represents less or no movement. Dashed circles indicate the position of the central cell nucleus. (E) Average velocity of mitochondrial movement in the central cell. Error bars represent SEM. Levels not connected by the same letter (A and B) are significantly different (P < 0.01, Tukey-Kramer HSD test). (Scale bar, 20 µm.)
F-Actin Meshwork Movement Controlled by Myosin Is Conserved across the Plant Kingdom.
In tobacco, as low as 2 mM of a BDM application stops mitochondrial movement in leaf cells (7). On the other hand, as in Arabidopsis (Fig. 3 A, C, and F, and Movie S2), 20 mM BDM does not impair sperm nuclear migration in the tobacco central cell (7). To verify the involvement of myosins in sperm nuclear migration in tobacco, 50 mM BDM was applied to tobacco central cells, and sperm nuclear migration was monitored. Sperm nuclei incorporated into tobacco central cells dissected out from the ovule did not move toward the polar nuclei with 50 mM BDM (SI Appendix, Fig. S7, and Movie S4). These results show that the active role of myosins in F-actin meshwork movement for sperm nuclear migration is conserved among flowering plants. Furthermore, the differences in the BDM sensitivity between mitochondrial movement and F-actin meshwork movement in both Arabidopsis (Figs. 3C and 4B) and tobacco central cells (SI Appendix, Fig. S7 and Movie S4) (7) show that the myosin function for F-actin meshwork movement is distinct from the canonical organelle transport function.
Discussion
In flowering plants, F-actin controls sperm nuclear migration for successful fertilization. This work has revealed that F-actin regulatory pathways involving the class XI myosin XI-G (Fig. 3 B and F) and SCAR2 (Fig. 1H), but not the ARP2/3 complex (Fig. 1H and SI Appendix, Fig. S2F), regulate F-actin dynamics in the female gametophytic cell that are important for sperm nuclear migration. In trichomes and cotyledon pavement cells, WAVE/SCAR solely relays the signal to ARP2/3 for F-actin organization (15, 16, 19). Our results indicate that an ARP2/3-independent WAVE/SCAR F-actin regulatory pathway exists in the Arabidopsis female gametophytic cell (Fig. 5). We also revealed that the plant-specific class XI myosin XI-G is critical for F-actin meshwork movement required for sperm nuclear migration in the Arabidopsis central cell. Class XI myosins are known to control the organization of F-actin and generate force for filament buckling and straightening in somatic cells (36, 39). However, the movement of F-actin from the plasma membrane to the nucleus in the female gametophytic cell occurs at the whole-cell level, and the involvement of XI-G on such a large scale as occurs in the dynamic F-actin inward movement is unusual. Furthermore, this newly identified myosin function is distinct from the myosin canonical function as the organelle transporter (Figs. 3 and 4).
Fig. 5.
Model of F-actin dynamics in the female gamete for sperm nuclear migration. (A and B) Schematic image of F-actin meshwork movement (A) and the pathway controlling F-actin movement in the Arabidopsis central cell (B). ROP8 localizes to the plasma membrane and interacts with SCAR2. The ROP8-SCAR2–signaling pathway positively regulates the meshwork F-actin movement in a gametophyte-specific ARP2/3-independent manner. Myosins including XI-G also regulate the meshwork F-actin movement through myosin functions distinguishable from the movement of organelles.
The plant Rho-GTPase gene, ROP8, and SCAR2 are both expressed in the Arabidopsis central cell (5, 24, 25), and they directly interact with each other (13). The mutants show both reduced F-actin meshwork inward movement (Fig. 1 D and H) (5) and delay in sperm nuclear migration (Fig. 2E and SI Appendix, Fig. S1) (5), indicating that a ROP8-SCAR2–mediated signaling pathway controls F-actin dynamics in the Arabidopsis central cell (Fig. 5). In SCARs, besides highly conserved SHD and WA domains, there are plant-specific domains with unknown functions (14). It is possible that these domains control the ARP2/3-independent WAVE/SCAR pathway in the female gametophytic cell for fertilization. Functional domain dissections of SCAR2 and the identification of other WAVE/SCAR pathway gene (15, 40⇓–42) involvement will be awaited to further understand this ARP2/3-independent WAVE/SCAR pathway in the female gametophyte. SCAR4 is also expressed in the central cell (24, 25, 43). However, fertilization is not affected in scar4-1 (Fig. 2E and SI Appendix, Fig. S2F), and the scar2-1;4–1 double mutant shows neither additive nor synergistic effects (SI Appendix, Fig. S2F). These results indicate that the SCAR4 pathway is distinct from the SCAR2 pathway in the central cell. Similar to SCAR4, ARP2 is also expressed in the central cell (Fig. 1E) and is dispensable for proper F-actin meshwork movement (Figs. 1H and 2E). SCAR4 may play a role in the canonical WAVE/SCAR-ARP2/3 pathway, possibly controlling F-actin branching in the central cell for processes other than gamete nuclear migration.
Our myosin studies (Figs. 3 and 4) indicate that the myosin-driven organelle movement does not play a role in F-actin meshwork movement and that there must exist a different mechanism by which XI-G controls F-actin meshwork movement in the Arabidopsis central cell. In vitro, myosins can cross-link actin filaments, and myosin motor activity generates contractile forces that result in directional F-actin movement (44, 45). These F-actin dynamics are similar to F-actin meshwork inward movement observed in the female gametophytic cell (5, 9). Any mutants displaying a defect of F-actin meshwork movement do not completely arrest sperm nuclear migration in the central cell (Fig. 2E). Therefore, in parallel to SCAR2, other pathways such as class XI myosins and/or other actin nucleators such as formins (46) likely control F-actin meshwork movement. In the case of class XI myosins, gene functional redundancy (47) may partially complement the XI-G function for maintaining F-actin meshwork in the central cell. XI-I interacts with the nucleus, linking the nuclear envelope with F-actin to maintain the position and shape of the nucleus; however, unlike XI-G, XI-I itself does not regulate F-actin movement (48). Although XI-I appears to be less expressed in the female gametophytic cells, including the central cell (24, 25), XI-I may help the incorporated sperm nucleus associate rapidly with the F-actin meshwork in the female gametophytic cells for fertilization. Further analyses will reveal how class XI myosins play their roles in the unique F-actin dynamics in the central cell.
In flowering plants, the double-fertilization events of the egg and central cells are regulated independently, but the essential role of F-actin in sperm nuclear migration is conserved in both cells (5, 7). The involvement of WAVE/SCAR and myosins in fertilization has also been shown in both the egg and central cells (5, 7, 9). However, ROP8 is expressed only in the Arabidopsis central cell (5), and it still remains unknown what counterpart of ROP8 controls F-actin dynamics for fertilization in the egg cell. This question is also the case with XI-G. Interestingly, SCAR2 appears to be expressed in the Arabidopsis egg cell as the highest among SCAR genes as well, with SCAR3 as the second highest (49). We did not observe sperm nuclear migration delay in scar2-1 egg cells (SI Appendix, Fig. S1). Due to the cell size, the sperm nuclear migration distance within the fertilized egg cell is shorter than that of the fertilized central cell, and indeed, the sperm nuclear migration time is shorter in the egg cell (50). Therefore, it is possible that we simply could not detect sperm nuclear migration phenotypes in the egg cell with our fertilization assay system and that SCAR2 may also play a role in F-actin meshwork movement in the egg cell. Another possibility is that SCAR3 may control the egg F-actin dynamics for fertilization. The effect of scar on sperm nuclear migration in the egg cell should be amplified in the yet-to-be-identified egg rop and/or myosin xi mutant; thus, investigation of these genes together will facilitate the identification of egg fertilization factors.
In Arabidopsis, rapid sperm chromatin decondensation in the fertilized central cell is required prior to its first mitotic division of the fused triploid nucleus for proper endosperm development (29). One of the reasons why the constant F-actin meshwork movement is already initiated even before pollen-tube arrival to the female gametophyte is to ensure the rapid movement of the sperm nucleus immediately after plasmogamy for successful karyogamy and completion of sperm chromatin decondensation prior to the first mitotic division (51). Nevertheless, knowledge of detailed molecular and cellular mechanisms of flowering-plant double fertilization are still lacking. The large central cell enables further understanding of the basis of flowering-plant fertilization. Our work uncovers a female gametophyte-specific regulatory pathway for F-actin meshwork inward movement and essential roles of SCAR2 and XI-G for sperm nuclear migration in the Arabidopsis central cell. Further investigation will reveal differences and similarities between not only the fertilization processes of the egg and central cells, but also the mechanisms of F-actin dynamics in somatic and reproductive cells.
Materials and Methods
Plant Material and Growth Conditions.
All Arabidopsis plant lines used in this work were Columbia-0 (Col-0) ecotype. Seeds were first germinated, and then seedlings were kept for 2 wk under short-day conditions (8 h light, 22 °C and 16 h dark, 18 °C). Plants were then shifted to 22 °C with continuous light. The proFWA::Lifeact:Venus (5), scar2-1 (26), arp2-1 (20), dis2-1 (arpc2) (32), arpc4-t2 (33), and proDD65::coxlV:GFP (38) lines have been described previously. The proFWA::H2B:mRuby2 and pro2 × 35S::Lifeact:Venus constructs were generated in the multisite gateway binary vectors pAlligatorG43 and pAlligatorR43, respectively (5). The xi-g (34) (SALK018032C), scar2-1 (SALK039449) (26), and scar4-1 (SALK_116410C) (13) mutant lines were obtained from the Arabidopsis Biological Resource Center. Seeds of dis2-1 (arpc2) (32), and arpc4-t2 (33) mutants were a gift from Daniel B. Szymanski, Purdue University, West Lafayette, IN. The homozygosity of all mutant lines was confirmed by PCR using gene-specific primers flanking the T-DNA insertion site and T-DNA–specific primers (SI Appendix, Table S1). The scar2-1, scar4-1, arp2-1, dis2-1, arpc4-t2, and xi-g mutant lines were crossed with the proFWA::Lifeact:Venus line, and homozygous lines were obtained from the F2 population. The scar2-1 was crossed with the scar4-1–harbored proFWA::Lifeact:Venus marker line to generate double mutants and homozygous line was obtained from the F2 population.
Fertilization Assay.
Arabidopsis WT, xi-g, scar2-1, scar4-1, scar2-1;scar4-1, arp2-1, dis2-1, and arpc4-t2 pistils were emasculated 2 d prior to pollination. Pistils were pollinated with the proHTR10::HTR10:mRFP1 sperm-specific histone marker line (27). Ovules were then dissected 9 h after pollination, and sperm nuclear migration was observed by confocal microscopy.
Plasmid Construction and Transformation.
The promoter sequences of Arabidopsis XI-G, ARP2, and DD65 genes were amplified by PCR using the KOD-plus ver. 2 PCR kit (Toyobo). Primer sequences for PCR are listed in SI Appendix, Table S1, and all constructs were generated by the Multisite Gateway Technology (Invitrogen). The multisite gateway binary vectors pAlligatorR43 and pAlligatorG43 and the ENTRY clone plasmids pENTR221-Histone2B and pENTRP2rP3-Clover have been described previously (5). All constructs were transformed into Arabidopsis Col-0 using the floral dip method (52).
F-Actin Dynamics Assay and Chemical Preparation.
Arabidopsis pistils were emasculated 2 d prior to imaging. Pistils were dissected out by a sharp knife, and mature ovules were collected into the assay medium (2.1 g/L Nitsch basal salt mixture, 5% wt/vol trehalose dehydrate, 0.05% wt/vol 2-[N-morpholino]ethanesulfonic acid-KOH [pH 5.8], and 1x Gamborg vitamin) in a glass-bottom dish as described previously (53). For each experiment, ovules from two pistils were collected into a 200-µL assay medium. Myosin inhibitor, BDM (stock, 500 mM in the assay buffer; Sigma-Aldrich) was prepared before each experiment, and 20 mM (no incubation) and 50 mM (no incubation) were used as working concentrations. WAVE/SCAR inhibitor, wiskostatin (stock, 10 mM in dimethylsulfoxide [DMSO]; Sigma-Aldrich); ARP2/3 inhibitor CK-666 (stock, 10 mM in DMSO; Sigma-Aldrich) was prepared and kept in −80 °C. Freshly prepared working concentrations of wiskostatin (10 µM, 1 h incubation) and CK-666 (200 µM, 1 h incubation) were used.
Imaging.
An FV1200 laser scanning confocal system (Olympus) equipped with 488-, 515-, and 559-nm laser lines and the GaAsP detection filter were used to illuminate Clover, coxlV:GFP, Alexa Fluor 488 (488 nm), Lifeact:Venus (515 nm), and H2B:mRuby2 and HTR10:mRFP1 (559 nm). Snapshot or time-lapse (5-s to 1-min interval) images with z-planes (15 to 20 µm total, 3 to 4 µm for each slice) were acquired using FV10-ASW 4.2 software. Laser 3 to 4%, HV 500–550, gain 1.25 and Kalman 2 options were applied to capture images. All confocal images were analyzed and processed using Fiji (ImageJ) software.
F-Actin Dynamics Quantification.
F-actin velocity was measured in two steps. First, kymographs of the F-actin movement were generated with Fiji. Z-projected images of the central cell were processed in Fiji by the following sequence of actions: adjust brightness and contrast; Process > Filters > Gaussian Blur; Process > Background subtraction > Rolling ball radius; adjust brightness and contrast; obtain kymograph with installed macro in Fiji. Macro was installed from the following link: http://dev.mri.cnrs.fr/projects/imagej-macros/wiki/Velocity_Measurement_Tool. Second, in the kymograph, segmented lines were drawn to track the movement of actin cables, and velocities were obtained based on the installed macro in Fiji.
F-Actin Angle and Orientation Measurement.
Z-projected actin cable images, that were processed by setting all pixel values less than 300–300 to mask the background noise, were converted to skeletonized images using the LPX imageJ plug-in lineExtract (54) with the default values (SI Appendix, Fig. S8A). Using a depth-first traversal algorithm, all pixel pairs, of pixels adjacent to each other, in the skeletonized images were first identified as lines. Then, angles of each pixel pair line relative to the center of the central cell nucleus were determined. Of the pixel pair, whichever pixel was identified second by the depth-first traversal algorithm was used as the vertex of the angle of the line relative to the center of the nucleus. The angle was calculated using the law of cosines from a triangle generated by the pixel pair and the pixel at the center of the nucleus as vertices. The angle was determined as a 0° to 90° angle, subtracting from 180° if necessary. Only lines present between the chalazal end and the top nucleus edge of the central cell were measured due to the quality of the original image.
Mitochondrial Velocity Measurement.
To calculate mitochondrial velocities, we used the TrackMate plugin (55) in Fiji. All image stacks were cropped to contain only the area above 52.8 μm on the y axis. TrackMate ran with an estimated mitochondrial velocity of 1.2 μm and a threshold of 1. The auto function on initial thresholding removed low-quality mitochondrial predictions. TrackMate was run with the simple LAP tracker to match mitochondrial predictions through time (SI Appendix, Fig. S8 B–D). The linking maximum distance and gap-closing maximum distance were set to 3 μm. The gap-closing maximum frame was set to 2 frames.
Data Availability.
All study data are included in the article and supporting information.
Acknowledgments
We thank Drs. Robert B. Goldberg and Anthony Clark for their critical comments on this manuscript; Dr. Yukinosuke Ohnishi for image processing; and Dr. Daniel B. Szymanski (Purdue University) for dis2-1 and arpc4-t2 seeds. This work was supported by NSF Grant IOS-1928836 (to T.K.); National Institute of Food and Agriculture, US Department of Agriculture Hatch Program Grant 1014280 (to T.K.); the National Natural Science Foundation of China Grants 31570317 and 31270362 (to X.P. and M.-X.S.); and the Ministry of Education, Culture, Sports, Science and Technology of Japan Grants-in-Aid for Scientific Research on Innovative Areas Grants 17H05846 and 19H04869 (to D.M.). M.F.A., U.F., and T.K. were supported by a start-up fund from the Department of Plant and Soil Sciences and the College of Agriculture, Food and Environment, University of Kentucky.
Footnotes
- ↵1To whom correspondence may be addressed. Email: tomo.k{at}uky.edu.
Author contributions: M.F.A., X.P., M.-X.S., and T.K. designed research; M.F.A., U.F., X.P., and S.W.H. performed research; D.M. contributed new reagents/analytic tools; M.F.A. and T.K. analyzed data; and M.F.A. and T.K. wrote the paper.
The authors declare no competing interest.
This article is a PNAS Direct Submission. T.H. is a guest editor invited by the Editorial Board.
This article contains supporting information online at https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.2015550117/-/DCSupplemental.
Published under the PNAS license.
References
ARP2/3-independent WAVE/SCAR pathway and class XI myosin control sperm nuclear migration in flowering plants
Article Classifications
- Biological Sciences
- Plant Biology
Sign up for Article Alerts
Jump to section
You May Also be Interested in
Pottery, burnt clay, and burnt flint from Jordan dated from 7752 BCE to 5069 BCE yielded a record of variations in Earth’s magnetic field strength during the Neolithic period.
Image credit: Thomas E. Levy.
Mechanics of ant tunneling could inspire the development of improved robotic mining techniques.
Image credit: Jose E. Andrade and David R. Miller (California Institute of Technology, Pasadena, CA).
When it comes to potential geoengineering initiatives, researchers and policymakers need to know what to study—and when to stop.
Image credit: Shutterstock/Venera Salman.
Researchers found a gene that could pave the way toward modifying crops so that they produce stable hybrid seed lines with desirable traits.
Image credit: Aleksandr Ozerov.