Cytoplasmic Ca²⁺ changes dynamically during the interaction of the pollen tube with synergid cells

Sexual reproduction in flowering plants starts when a pollen grain lands on the surface of the stigma, hydrates and germinates a pollen tube. The pollen tube enters the style and grows through the transmitting tract tissue to the ovary, where it is guided along the funiculus and enters the micropyle of the ovule to deliver two sperm cells to the embryo sac to effect double fertilization. This directional growth of the pollen tube from the stigma to the embryo sac is controlled by cellular interactions, and Ca²⁺ is thought to be a key player in regulating these processes (Dumas and Gaude, 2006; Chae and Lord, 2011).

Pollen tube guidance by the embryo sac involves at least two steps (Shimizu and Okada, 2000). First, funicular guidance directs the pollen tube from the placenta along the funiculus to the ovule. This is followed by micropylar guidance, which guides the pollen tube into the micropyle to reach the embryo sac (reviewed by Márton and Dresselhaus, 2010; Takeuchi and Higashiyama, 2011). Electron microscopic studies showed that the synergid cells and the filiform apparatus, a structure at their micropylar pole, contain high Ca²⁺ concentrations ([Ca²⁺]) (Chaubal and Reger, 1990; Chaubal and Reger, 1992a; Chaubal and Reger, 1992b; Chaubal and Reger, 1993). In Arabidopsis, the transcription factor MYB98 is expressed in synergid cells and required for development of the filiform apparatus and micropylar guidance, suggesting a role of the filiform apparatus in this process (Kasahara et al., 2005; Punwani et al., 2007). In addition, in vitro assays of pollen tube chemotropism towards excised pistil tissue and various compounds suggested that Ca²⁺ might be a potential attractant (Mascarenhas and Machlis, 1962; Reger et al., 1992). In Torenia fournieri, however, elevation of [Ca²⁺] did not affect pollen tube attraction in in vitro guidance assays (Higashiyama et al., 2006). In this system, it was unambiguously shown that the synergids produce pollen tube attractants (Higashiyama et al., 2001). Recently, the Torenia LURE proteins, secreted cysteine-rich peptides that accumulate in the filiform apparatus of the synergid cells, have been identified as pollen tube attractants (Okuda et al., 2009; Okuda and Higashiyama, 2010). In Arabidopsis, synergids and other embryo sac cells produce a diverse set of cysteine-rich proteins (Jones-Rhoades et al., 2007; Dresselhaus and Márton, 2009; Wuest et al., 2010). This indicates that similar attractants produced by the synergid cells also exist in Arabidopsis, although their molecular identity is not yet known.

However, it is not clear whether Ca²⁺ in the synergid cells is relevant to pollen tube guidance and reception in Arabidopsis. On the other hand, the auto-inhibited Ca²⁺ ATPase (ACA) ACA9, which is predicted to be activated by Ca²⁺/calmodulin, is expressed primarily in the plasma membrane of the Arabidopsis pollen tube, and disruption of ACA9 results in partial male sterility. In an aca9 mutant, pollen tubes normally reach the embryo sac and cease growth, but they fail to rupture and release the sperm cells (Schiøtt et al., 2004). This finding suggests the involvement of Ca²⁺ signaling in sperm discharge. The relationship between [Ca²⁺] and pollen tube growth has previously been examined in vitro, where pollen tubes germinate and grow in culture medium. Ratiometric ion imaging using fluorescent dyes has revealed that the apical domain of a pollen tube grown in vitro contains a tip-focused [Ca²⁺] gradient (Pierson et al., 1994; Pierson et al., 1996; Cheung and Wu, 2008). Genetically encoded Ca²⁺ indicators, yellow cameleons (YCs), have been developed to monitor [Ca²⁺] in living cells (Miyawaki et al., 1997; Nagai et al., 2004). Using plants expressing YC3.1 or YC3.60, Ca²⁺ dynamics in the pollen grain, pollen tub, and stigmatic papillar cell were monitored during pollination, as well as on pollen germination medium (Iwano et al., 2004; Watahiki et al., 2004; Iwano et al., 2009). These studies revealed that cytosolic Ca²⁺ concentrations ([Ca²⁺]_cyt) in the tip region of the growing pollen tube are maintained at submicromolar levels under all conditions.

Imaging of micropylar guidance has been performed under a semi-in vivo fertilization system in Torenia and Arabidopsis (Higashiyama et al., 1998; Palanivelu and Preuss, 2006; Higashiyama and Hamamura, 2008; Hamamura et al., 2011). In Arabidopsis, labeling the cytoplasm of the synergid cells and pollen tubes with GFP and/or RFP revealed that a pollen tube emerging from a cut style grows towards excised ovules, enters the ovule through the micropyle, ceases growth at the synergid cells and ruptures to release the sperm cells (Sandaklie-Nikolova et al., 2007). Furthermore, labeling of sperm cell nuclei with mRFP and labeling of the synergid, egg and central cells with GFP have enabled the successful study of double fertilization (Ingouff et al., 2007; Hamamura et al., 2011).

Here, to examine Ca²⁺ dynamics during micropylar guidance and pollen tube reception, we employed Arabidopsis plants expressing the Ca²⁺ sensor YC3.60 in pollen tubes and synergid cells and monitored Ca²⁺ dynamics in these cells during semi-in vivo fertilization. We set up a monitoring system using an EM-CCD camera attached to a wide-field microscope to follow Ca²⁺ dynamics and found remarkable changes in [Ca²⁺]_cyt both in the pollen tube and the synergid cells, suggesting an involvement of Ca²⁺ in the communication between these cells. We discuss our results in relation to the physiological relevance of [Ca²⁺]_cyt for the interactions between the pollen tube and synergid cell during micropylar pollen tube guidance and pollen tube reception.

Arabidopsis thaliana accession Columbia was used for generation of transgenic plants. Plants were grown in pots containing soil/perlite (3:1, by volume) in a growth chamber with a light intensity of 120-150 mmol m⁻² s⁻¹ during the daily 12-hour light period. The temperature was maintained at 22±2°C.

A DNA fragment encoding DsRed was amplified from pDsRed2 vector (Clontech) by PCR using the following primers: 5′-GCTCTAGATGGCCTCCTCCGAGAACGT-3′ and 5′-GTGCGAGCTCTACAGGAACAGGTGGTGGC-3′; the incorporated XbaI and SacI sites are underlined. The obtained DNA fragment was digested with XbaI and SacI, and then ligated to the XbaI and SacI sites of the pBI121-pAct1::YC4.60(ER) vector used previously (Iwano et al., 2009), yielding the pBI121-pAct1::RFP vector.

For the constructs expressing YC3.60 in the synergid cells, promoter regions of DD2 (At5g43510) and DD2-like (At5g43513) genes were amplified by PCR with primers (5′-CTTAAGCTTTCTCTGTTTCTTATCAG-3′ and 5′-CCATGGTTCTTGGTAGAAATTAACCAATATCAC-3′ for DD2 promoter; 5′-AAGCTTGAGTTTGGATTGATTCGAGGC-3′ and 5′-CCATGGTTCTTTGTAGAAATTAACCAATATCAC-3′ for DD2-like promoter; the incorporated HindIII and NcoI sites are underlined). Obtained fragments were digested with HindIII and NcoI and ligated to HindIII and NcoI sites of pLat52::YC3.60 previously constructed (Iwano et al., 2009) for yielding pDD2::YC3.60 and pDD2-like::YC3.60.

The pBI121-Act1::RFP plasmid, pDD2::YC3.60 plasmid and pDD2-like::YC3.60 plasmid were electroporated into Agrobacterium tumefaciens strain EHA105 (Hood et al., 1993). The Agrobacterium infiltration procedure was performed with unopened flower buds of Arabidopsis accession Columbia as previously described (Iwano et al., 2004). Transformed seeds were selected on 1/2 MS plates containing kanamycin (50 μg/ml) and were analyzed using PCR to test for the presence of the RFP (DsRed) gene.

For imaging under the semi-in vivo condition, wild-type Arabidopsis flowers excised before anther dehiscence were collected onto an agar plate and the anthers were removed. Pollen grains from freshly dehisced anthers of pAct1::YC3.60-expressing plants were loaded onto the wild-type stigma. Thirty minutes after manual pollination, the upper part of the pollinated pistil was excised and mounted in the germination medium (Boavida and McCormick, 2007) in a moistened glass-bottomed dish. Next, following previously reported methods (Palanivelu and Preuss, 2006), pAct1::YC3.60- or pDD2like::YC3.60-expressing ovules were excised from transgenic Arabidopsis plants. After 2 hours at 23°C, the [Ca²⁺]_cyt in pollen tubes growing through the style and towards the synergid cells during micropylar guidance was imaged with a Zeiss Axio Observer inverted microscope equipped with a Xenon lamp, an EM-CCD Evolve 512 camera (Photometrics) and a filter exchanging system (Prior). Imaging of the cameleon emission ratio was accomplished using two emission filters (480/30 for ECFP, 535/40 for Venus and 641/75 for DsRed). After background subtraction, the Venus/ECFP ratio was determined using MetaMorph software (Molecular Devices, USA). A Zeiss C Apo 40×/1.20 water immersion objective lens and a Plan-Apo 20×/0.8 objective lens were used for imaging. Exposure times were typically 20 mseconds, and images were collected every 3-20 seconds. In each experiment, the ratio in a region of interest of 6 × 6 μm² was measured using the MetaMorph software, and is shown as sequential line graphs. The mean and maximum values were calculated using Microsoft Excel.

To confirm that the full-length cameleon was expressed in the pollen tube, elongated pollen tubes were monitored with excitation at 442 nm using a spectral imaging fluorescence microscope system (LSM710 META, Carl Zeiss, Jena, Germany). This system is capable of resolving the spectra of various fluorescence images; therefore, we were able to obtain images with no interference from the overlapping fluorescence emissions (Iwano et al., 2009). These experiments were carried out more than 30 times.

Calibration of the [Ca²⁺]_cyt was carried out as described previously (Allen et al., 1999). Serial dilutions of purified YC3.60 were made in Ca²⁺ calibration buffer (Molecular Probes), in which the free [Ca²⁺] ranged from 0 mM to 1 mM. Dilutions of YC3.60 that resulted in similar signal intensities to those seen in YC3.60-expressing pollen tubes were used to determine the minimum (R_min) and maximum (R_max) ratio values. For YC3.60, R_min and R_max values were 1.75 and 5.35 (supplementary material Fig. S1). These values were used to convert the YC3.60 fluorescence ratios into a [Ca²⁺]_cyt by fitting them to YC3.60 calibration curves obtained in vitro.

Statistical analyses were performed using Student's t-test where necessary.

[Ca²⁺]_cyt in pollen tubes growing under in vitro and semi-in vivo conditions has been examined in previous studies (Iwano et al., 2004; Iwano et al., 2009). However, it remained unclear whether ovules affect [Ca²⁺]_cyt dynamics in the pollen tube. First, we examined whether growth and [Ca²⁺]_cyt in pollen tubes changes in the presence of ovules. In these experiments, wild-type ovules were placed about 450 μm away from the pistil excision site. After incubation of the pollinated pistil for ~2 hours at 23°C, pollen tubes emerged from the pistil excision site. The [Ca²⁺]_cyt in pollen tubes were monitored at intervals of 10 seconds using a 20× objective lens. As a result, 75% of ovules successfully attracted pollen tubes, which entered the micropyle and stopped growth at the synergid cells (18/24 samples).

For comparison, excised pollinated pistils were placed on the pollen growth medium without ovules and monitored. In this case, pollen tubes grew straight, even when their lengths were over 400 μm out of excised pistils (Fig. 1A). However, in the presence of ovules, pollen tubes grew straight until they were ~150 μm from a micropyle, then grew in a wave-like pattern towards an ovule (Fig. 1B; supplementary material Movie 1). To elucidate whether the wave-like turning of the pollen tub was related to synergid function, ovules from the myb98 mutant, in which the formation of the filiform apparatus and pollen tube guidance are defective, were placed on the growth medium. In this case, pollen tubes grew straight and only rarely grew into a micropyle (1/29) (Fig. 1C; supplementary material Movie 2).

Next, we monitored the [Ca²⁺]_cyt in pollen tubes growing towards the micropyles of wild-type and myb98 ovules, or without ovules. When the distance from the micropyles of wild-type ovules was about 30-70 μm, the average Venus/ECFP ratio in the tip was 3.18±0.48 (n=18) (Fig. 2A1-A3). Especially, when the pollen tube was turning towards a micropyle, a broader and stronger [Ca²⁺]_cyt signal at the pollen tip was observed (Fig. 2B). By contrast, in pollen tubes growing within 70 μm of the micropyles of myb98 ovules, the [Ca²⁺]_cyt did not change when the pollen tubes approached the micropyles more closely by chance and the average ratio was 2.10±0.16 (n=10), which was significantly different from that in the presence of wild-type ovules (P<0.001) (Fig. 2C,D). Furthermore, in the pollen tubes growing out of a cut style without wild-type ovules, the average ratio was 1.98±0.15 (n=10), which was significantly different from that in the presence of wild-type ovules (P<0.001), but not in the presence of myb98 mutant ovules (Fig. 2C-E). These results were reproducibly obtained in the experiments using another pAct1::YC3.60-expressing line (wild-type ovule, 3.30±0.32; myb98 ovule, 2.21±0.31; without ovule, 2.07±0.21).

To convert the YC3.60 ratios into approximate [Ca²⁺]_cyt values, calibration of [Ca²⁺]_cyt was carried out as described previously (Allen et al., 1999; Iwano et al., 2009). We derived a titration curve (supplementary material Fig. S1, see Materials and methods), from which we estimated that the [Ca²⁺]_cyt in the tip region shown in Fig. 2 ranged from 0.2 to 0.7 μM in the presence of wild-type ovules and from 0.2 to 0.3 μM in the presence of myb98 ovules or in the absence of ovules.

Furthermore, in order to examine whether [Ca²⁺]_cyt in the pollen tip increases just after emergence from the pistil or gradually during its approach to the ovules, the ratio in the pollen tubes emerging from the pistils was monitored every 3-5 minutes over a range of 150 to 400 μm from the pistil excision site, the ovules being 450 to 500 μm away. Without ovules and in the presence of myb98 mutant ovules, there was no obvious change in the ratio (Fig. 2F). In the presence of wild-type ovules, the ratio was no different from that without ovules or with the myb98 ovules in the range of 150-200 μm, but increased between ~250-300 μm, where the pollen tubes started to turn; the increase continued at ~350-400 μm (Fig. 2F). These observations indicate that [Ca²⁺]_cyt in the pollen tip increases as the pollen tubes approach the ovules, before it decreases as the tubes grow closer to the micropyle. Collectively, these results suggest that [Ca²⁺]_cyt dynamics in the pollen tube changes in response to ovules, especially synergid cells, supporting our hypothesis that Ca²⁺ signaling in pollen tubes is involved in micropylar guidance. Moreover, as [Ca²⁺]_cyt in pollen tubes increases at some distance in the presence of wild-type but not myb98 ovules, functional synergids are also involved in producing long-range effects on [Ca²⁺]_cyt in the pollen tube.

Previously, eight independent YC3.60-expressing plants were obtained by transformation with the pAct1::YC3.60 plasmid (Iwano et al., 2009). Surprisingly, one of these transgenic lines expressed YC3.60 not only in the pollen grain and pollen tube, but also in the synergid cells, and to a lesser degree the central cell of the embryo sac and the sporophytic tissues of the ovule (supplementary material Fig. S2A,B). Seed set in this transgenic line (51.8±5.2 seeds, n=30) was no different from that of wild-type plants (50.9±6.7 seeds, n=30). In particular, the intensity of YC3.60 in the synergid cells was the same as that in the pollen tube. To confirm that the synergid cells expressed the ECFP and Venus components of YC3.60, we obtained fluorescence spectra from these cells using a spectral-imaging microscope system with excitation at 442 nm. The YC3.60 spectrum was a combination of the spectra typically observed for recombinant ECFP and Venus proteins. Photobleaching the synergid cells with a 514 nm light induced ~90% increase in ECFP fluorescence (supplementary material Fig. S2C). However, as the expression of the ACTIN1 gene has not previously been reported in synergid cells, several lines expressing YC3.60 in synergids were made using promoters of the DD2 gene expressing specifically in synergid cells (Steffen et al., 2007) and of a paralogous gene (DD2-like). We obtained two lines for pDD2::YC3.60 and four lines for pDD2-like::YC3.60. The YC3.60 protein was specifically expressed in synergid cells in both pDD2-like::YC3.60 lines, and two synergid cells could be distinguished under confocal microscopy (supplementary material Fig. S2D-F); expression in the pDD2::YC3.60 lines was similar but low. Seed sets in these transgenic lines (48.5±6.5 seeds, 49.1±7.0 seeds, n=30) were not different from that of wild-type plants (50.9±6.7 seeds, n=30). Therefore, in this study, two distinct YC3.60 lines, pAct1::YC3.60 and pDD2-like::YC3.60, were used.

In order to examine Ca²⁺ dynamics in the pollen tube and synergid cells during the interaction between these two gametophytic cells, we simultaneously monitored their [Ca²⁺]_cyt dynamics using ovules with pAct1::YC3.60-expressing synergid cells in the semi-in vivo fertilization system described above. In these experiments, ovules were placed 300 μm from the excised edge of the pistil.

First, [Ca²⁺]_cyt in the synergid cells was monitored under a 20× objective before and after pollen tube arrival (Fig. 3). At intervals of 3 to 10 seconds, pollen tubes terminated at 90% of the synergid cells (10/11 samples), and 60% of the pollen tubes that reached the synergid cells burst (6/10 samples).

Next, both the pollen tube and the synergid cells inside the ovule were observed under a 40× objective (data not shown). [Ca²⁺]_cyt dynamics were monitored in these cells as the pollen tube approached at intervals of 3 to 20 seconds. Pollen tubes terminated at about 75% of the synergid cells (9/12 samples), and 55% of the pollen tubes that reached the synergid cells burst (5/9 samples). Before the approach of the pollen tube, the Venus/ECFP ratios of synergid cells were 1.96±0.06 (n=5), and [Ca²⁺]_cyt was estimated at about 0.1 μM (supplementary material Fig. S3); there was no significant [Ca²⁺]_cyt change over time. When a pollen tube arrived at the synergid cells, however, [Ca²⁺]_cyt in the synergid cell increased in the area next to the point of the pollen tube contact (supplementary material Fig. S3B, image number 2). After the pollen tube grew past the synergid cell, [Ca²⁺]_cyt in the entire synergid cell area further increased (supplementary material Fig. S3B, image number 3). Finally, [Ca²⁺]_cyt of the synergid cells increased strongly at the time of pollen tube rupture; however, because both the synergid cells and pollen tubes express the cameleon, it is difficult to determine the exact origin of the signal. On the pollen tube side, the Venus/ECFP ratio of the pollen tube tip growing within the ovule was 2.54±0.26 (n=5, converted to a [Ca²⁺]_cyt of 0.2-0.3 μM), and there was no significant change over time. However, pollen tube tip [Ca²⁺]_cyt increased substantially at the time of pollen tube burst (supplementary material Fig. S3B).

In all experiments under 20× and 40× objective, although we could observe the [Ca²⁺]_cyt dynamics of the pollen tube and the synergid cells during micropylar guidance, it was difficult to distinguish the change of [Ca²⁺]_cyt in the synergid cells from that of the pollen tube after the pollen tube had arrived at the synergid cells because they spatially overlapped. Therefore, to resolve the [Ca²⁺]_cyt dynamics in these cells, we monitored [Ca²⁺]_cyt separately in either only the pollen tube or only the synergid cells (see below).

We observed [Ca²⁺]_cyt changes in pAct1::YC3.60-expressing pollen tubes growing towards and inside wild-type ovules (Fig. 3B,C; supplementary material Movie 3). After entering the micropyle and approaching the synergid cells, pollen tube growth ceased temporarily and pollen tube tips swelled (Fig. 3B1,B2). Pollen tubes later resumed growth (Fig. 3B3,B4) and ruptured shortly afterward (Fig. 3B5,B6). Interestingly, between the time of growth cessation and rupture, [Ca²⁺]_cyt at the pollen tube tip did not significantly change and the tip-focused [Ca²⁺]_cyt gradient was prominent, except for a slight elevation of [Ca²⁺]_cyt in the shank region of the pollen tube (Fig. 3B3,B4). However, [Ca²⁺]_cyt increased from a ratio of 2.5 to 5 (0.15 μM to >1 μM) at the time of pollen tube rupture. In bursting pollen tubes, an area of high [Ca²⁺]_cyt rapidly spread from the tip to the shank. These [Ca²⁺]_cyt changes were observed in all experiments (3/3).

In order to observe the precise [Ca²⁺]_cyt change in the synergid cells alone, we monitored [Ca²⁺]_cyt dynamics in the pAct1::YC3.60-expressing synergid cells that were approached by RFP-expressing pollen tubes (Fig. 4; supplementary material Movie 4). When the pollen tubes made contact with the synergid cells, they temporarily ceased growth, and a Ca²⁺ oscillation was initiated at the contact point near the micropylar end of the synergid cells (Fig. 4A1,A2; 4B1,B2) and subsequently spread throughout the cell (Fig. 4A3; 4B3). When the pollen tubes grew further, the amplitude of the oscillation eventually decreased (Fig. 4A4,B4), but [Ca²⁺]_cyt increased again at about 5 minutes before pollen tube rupture (Fig. 4A5,B5). Finally, the [Ca²⁺]_cyt in the synergid cells reached a maximum (ratio=3.99±1.06; concentration=0.7 μM, n=4) when the pollen tube ruptured and decreased afterwards (Fig. 4B5). These [Ca²⁺]_cyt changes were also observed in all experiments using plants expressing pDD2-like::YC3.60 (3/3) (maximum ratio=4.08±1.05; concentration=0.75 μM, n=3) and indicate that physiological changes in the synergid cells during pollen tube guidance and reception involve [Ca²⁺]_cyt dynamics.

Upon closer examination, we detected some variation in the timing of the appearance of Ca²⁺ oscillation in the synergids after the pollen tubes enter the micropyle. Of the 10 pAct1::YC3.60 synergids and 30 pDD2-like::YC3.60 synergids monitored at intervals of 5 seconds, [Ca²⁺]_cyt in the synergid cells increased after a pollen tube contacted them (8/10, 26/30) (Fig. 5A,B1). But in some synergids (2/10, 4/30), a Ca²⁺ elevation was observed before pollen tube contact when it was still 10 μm to 30 μm away from the synergid cell (Fig. 5A,B2). These results suggest the possible existence of diffusible signals from the pollen tube. However, the level of Ca²⁺ elevation before contact was lower and the interval of oscillation was longer than that after contact. In addition, [Ca²⁺]_cyt increased strongly only after pollen tube contact. These results suggested that pollen tube contact with the synergid is important for Ca²⁺ oscillation.

Using recently developed imaging techniques, micropylar guidance in Arabidopsis has been visualized in a semi-in vivo fertilization system (Palanivelu and Preuss, 2006; Ingouff et al., 2007; Sandaklie-Nikolova et al., 2007; Hamamura et al., 2011). In this study, using Arabidopsis plants expressing the GFP-based Ca²⁺-sensing probe YC3.60 in pollen tubes and synergid cells, we visualized Ca²⁺ dynamics in these cells during micropylar guidance and pollen tube reception.

First, we examined whether ovules on the germination medium affected pollen tube growth and [Ca²⁺]_cyt dynamics. We found that pollen tubes repeatedly turned within 150 μm of the micropyles of wild-type ovules, whereas such turning was never observed near myb98 mutant ovules or in the absence of ovules. In addition, [Ca²⁺]_cyt in pollen tube tips close to wild-type ovules was significantly higher than in those near myb98 ovules or in the absence of ovules. By contrast, [Ca²⁺]_cyt near the pistil in the presence of wild-type ovule was not different from that in the presence of myb98 ovules. From the above observations, we can associate the turning of the pollen tubes with an increase in [Ca²⁺]_cyt. Thus, we hypothesize that pollen tubes sense some substance produced by wild-type ovules that affects the growth direction and [Ca²⁺]_cyt of the pollen tube. Attractants similar to the Torenia LURE-like proteins may also be produced by Arabidopsis synergid cells and may affect Ca²⁺ signaling in the pollen tube. Whether this also occurs in Torenia is currently unknown. However, [Ca²⁺]_cyt decreased when the pollen tube approached the micropyle or funiculus. The attractants from synergid cells are thought to form a concentration gradient around the micropyle in the germination medium. Our results suggest that the attractants elevate [Ca²⁺]_cyt in the pollen tube at a specific concentration, but at higher concentrations attractants reduce [Ca²⁺]_cyt in the pollen tube. In addition, we observed that there was no significant change in [Ca²⁺]_cyt oscillation and concentration at the tip region before the pollen tubes ruptured, although we detected Ca²⁺ oscillation in the synergid cells. These results indicate that a precise sequence of Ca²⁺ dynamics in the pollen tube may be a prerequisite for pollen tube rupture.

Regarding the interaction between pollen tubes and synergid cells, an increase in pollen tube [Ca²⁺]_cyt was observed at the time of pollen tube rupture. Monitoring revealed an absence of the [Ca²⁺]_cyt increase at the tip region before pollen tube burst and a spreading of the high [Ca²⁺]_cyt area from the tip region to the shank after pollen tube rupture. Recent evidence suggests that the pollen tube receptor-like kinases ANXUR1 and ANXUR2, which are closely related to the receptor-like kinase FERONIA, prevent premature rupture during pollen tube growth (Miyazaki et al., 2009; Boisson-Dernier et al., 2009; Kessler and Grossniklaus, 2011). However, it is currently not known whether Ca²⁺ signaling is involved in transducing a signal after the interaction of ANXUR with a putative ligand from the synergid cells. In addition, in aca9 mutants, where a predicted Ca²⁺/calmodulin-activated Ca²⁺ pump is disrupted, pollen tubes normally reach the embryo sac and cease growth, but fail to rupture and release the sperm cells (Schiøtt et al., 2004). Monitoring [Ca²⁺]_cyt dynamics in these aca9 mutants and other mutants, in which the pollen tube fails to rupture within the ovule, will elucidate the precise role of Ca²⁺ signaling in controlling pollen tube rupture.

There was no change in the [Ca²⁺]_cyt of synergid cells until they were approached by the pollen tube after it had entered the ovule. In very few cases, however, [Ca²⁺]_cyt oscillations initiated in the synergid cells before pollen tube reached them. Finally, [Ca²⁺]_cyt in synergid cells reached the maximum level at the time that the pollen tube burst. These [Ca²⁺]_cyt monitoring data suggest that the pollen tube induces [Ca²⁺]_cyt change in the synergid cells, and that Ca²⁺ signaling is involved in pollen tube-synergid cell interactions. In legumes, Rhizobium bacteria cause the development of a nodule, which is the site for bacteria and plant to cooperate in the fixation and assimilation of nitrogen (Mylona et al., 1995). Rhizobium Nod factors induce [Ca²⁺]_cyt oscillations in root hair cells. It has been reported that the oscillation with a mean period of 60 seconds starts within 10 minutes of Nod factor addition (Ehrhardt et al., 1996). In this study, we observed rare cases where [Ca²⁺]_cyt increase occurred in the synergid cells when the pollen tube was more than 20 μm away. This might suggest the existence of a diffusible signal from pollen tube, which affects [Ca²⁺]_cyt in synergid cells, although contact of the pollen tube with the synergid cells would be necessary for the strong Ca²⁺ oscillation observed later on.

The synergid cell that receives the pollen tube undergoes cell death. It is thought that degeneration of the synergid cell is required for pollen tube entry into the synergid cell and to provide access of the sperm cells to the egg and central cell for fertilization (Punwani and Drews, 2008). Sandaklie-Nikolova et al. proposed a model in which a signaling cascade triggered by the contact of the pollen tube with the synergid cell induces synergid cell death in Arabidopsis (Sandaklie-Nikolova et al., 2007). In pathogen-plant interactions, [Ca²⁺]_cyt increase is involved as an early signal in the hypersensitive response (HR) (Levine et al., 1996). The HR involves rapid programmed cell death in the local region surrounding an infection site (Dangl et al., 1996). Therefore, the finding that [Ca²⁺]_cyt increase and oscillation were induced by the interaction of the pollen tube with the synergid cells suggests that synergid cell death occurs downstream of the Ca²⁺ signaling cascade. In this study, when the pollen tube reaches the micropylar pole of the synergid cells, Ca²⁺ oscillations can be observed in all synergid cells after contact with the pollen tube. Furthermore, before pollen tube rupture and cell death of synergid cells occurs, Ca²⁺ oscillations can also be observed in all synergid cells, even in those where subsequently the pollen tube did not rupture. Therefore, it is probable that Ca²⁺ oscillation in the synergid cell is a necessary but not a sufficient condition for pollen tube rupture. During HR, H₂O₂ induces Ca²⁺-mediated cell death. The pollen tube generates reactive oxygen species (ROS) such as H₂O₂ (Dresselhaus and Márton, 2009), which may also function as signaling molecules in this interaction, inducing [Ca²⁺]_cyt changes upstream of programmed cell death.

In this study, a [Ca²⁺]_cyt increase was observed at the time of pollen tube rupture. Changes in [Ca²⁺]_cyt are thought to be regulated by transporters localized in the membrane of Ca²⁺ storage organelles and the plasma membrane (Sze et al., 2000). Previous electron microscopic studies in pearl millet ovaries reported that Ca²⁺ precipitations with antimonate were observed at the cell wall in the filiform apparatus and the developed ER structure under the cell membrane in synergid cells (Chaubal and Reger, 1992a; Punwani and Drews, 2008). We could not clarify the source of Ca²⁺ for the observed [Ca²⁺]_cyt increase in the synergid cells in this study, so it is still unknown whether the Ca²⁺ is delivered from the extracellular region, the ER, or both. Our [Ca²⁺]_cyt imaging results suggest that signaling molecules from the pollen tube, or physical stimulation by the pollen tube, affect Ca²⁺ dynamics in the synergid cells, and vice versa.

We thank Mrs Onishi, Mrs Okamura and Mrs Yamamoto for their technical assistance, and Prof. Higashiyama at Nagoya University and Dr Kaothien Nakayama for critical reading of the manuscript. We also thank Prof. Okada at National Institute for Basic Biology (Okazaki, Japan) for providing myb98 seeds.