Ca2+ transient induced by extracellular changes in osmotic pressure in Arabidopsis leaves: differential involvement of cell wall–plasma membrane adhesion

Authors


  • *

    Present addresses: Department of Applied Biological Science, Tokyo University of Science, 2641 Yamazaki, Noda, Chiba 278-8510, Japan, and Department of Biology, Graduate School of Science, Kyushu University, 4-2-1 Ropponmatsu, Chu-o-ku, Fukuoka 810-8560, Japan.

Teruyuki Hayashi. Fax: +81 4 7123 9767; e-mail: teruh@rs.noda.tus.ac.jp

ABSTRACT

We investigated the mechanism underlying the perception of extracellular changes in osmotic pressure in Vallisneria gigantea Graebner and transgenic Arabidopsis thaliana (L.) Heynh. expressing cytoplasmic aequorin. Hypertonic and hypotonic treatments of A. thaliana leaves each rapidly induced a Ca2+ transient. Both responses were essentially dependent on the presence of extracellular Ca2+ and were sensitive to Gd3+, a potential blocker of stretch-activated Ca2+ channels. Immediately after plasmolysis caused by hypertonic treatment and subsequent deplasmolysis caused by hypotonic treatment, the cells did not respond to a second hypertonic treatment and exhibited an impaired adhesion of the plasma membrane (PM) to the cell wall (CW). Recovery of the responsiveness required about 6 h. By contrast, no refractory phenomenon was observed in response to hypotonic treatment. Pretreatment with cellulase completely inhibited the Ca2+ transient induced by hypertonic treatment, but it did not affect the response to hypotonic treatment. V. gigantea mesophyll cells pretreated with cellulase exhibited an impaired adhesion of the PM to the CW. The leaf cells of multicellular plants can respond to both hypertonic and hypotonic treatments through the stretch-activated Ca2+ channels, whereas cellulase-sensitive adhesion of the PM to the CW is involved only in the response to hypertonic treatment.

INTRODUCTION

Plant cells respond to extracellular changes in osmotic pressure, as well as to various kinds of mechanical stimuli, including touch, wind and gravity (Wayne, Staves & Leopold 1992; Staves & Wayne 1993; Braun 1996; Shimmen 1996; Sato, Kadota & Wada 1999). An increase in extracellular osmotic pressure caused by hypertonic treatment has been reported to induce cellular responses such as stomatal closure (Liu & Luan 1998; Asai et al. 1999), accumulation of compatible solutes (Bartels & Nelson 1994; Bohnert, Nelson & Jensen 1995) and reduction of cytoplasmic motility (Okazaki et al. 1996; Hayashi & Takagi 2003). On the other hand, a decrease in extracellular osmotic pressure caused by hypotonic treatment also induces cellular responses such as activation of K+ and/or Cl efflux channels for turgor regulation (Brownlee et al. 1999; Shepherd, Beilby & Shimmen 2002; and references therein), incorporation of exocytotic vesicles into the plasma membrane (PM) (Gilkey & Staehelin 1989) and reduction of cytoplasmic motility (Tazawa, Shimada & Kikuyama 1995; Okazaki, Ishigami & Iwasaki 2002).

Recently, cytoplasmic Ca2+ has been found to play the central role of second messenger in the perception of environmental abiotic stimuli and the regulation of many responses induced by those stimuli (Gilroy & Trewavas 1994; Bush 1995; Brownlee et al. 1999; Knight 2000). A number of studies have demonstrated that the cytoplasmic concentration of Ca2+ ([Ca2+]cyt) increases rapidly and transiently after perception of mechanical stimuli (Knight et al. 1991; Knight, Smith & Trewavas 1992; Haley et al. 1995; Plieth 2001), and upon hypertonic (Lynch, Polito & Läuchli 1989; Okazaki et al. 1996; Knight, Trewavas & Knight 1997) or hypotonic (Okazaki et al. 1987, 2002; Tazawa et al. 1995; Taylor et al. 1996; Takahashi et al. 1997) treatment. Mechanical stimuli and extracellular changes in osmotic pressure similarly produce deformation in the PM, and such deformation is thought to activate mechano-sensitive ion channels in both animal (Yang & Sachs 1989) and plant (Pickard & Ding 1993) cells. The presence of mechano-sensitive Ca2+ channels activated by stretching of the PM has been suggested in several types of plant cell (Falke et al. 1988; Cosgrove & Hedrich 1991; Ding & Pickard 1993). Recently, a gene encoding a stretch-activated Ca2+-permeable ion channel (Mid1) has been identified in yeast (Kanzaki et al. 1999). Although such Ca2+-channel activities may play important roles in the Ca2+ transient induced by hypertonic and hypotonic treatments, the activating mechanisms for the Ca2+ channels after the perception of extracellular changes in osmotic pressure have been poorly investigated, especially in multicellular plants.

In yeasts, molecular genetic approaches have elegantly revealed that the cell wall (CW) is essential for the two-component histidine kinase (Sln1) signalling pathways in the response to high osmotic stress (Reiser, Raitt & Saito 2003). In animal cells, focal adhesions play an important role in sensing physical forces and in transducing mechano-sensory signals into the cells (Ingber 1991). At focal adhesions, extracellular matrix proteins that contain an Arg-Gly-Asp (RGD) motif bind to specific PM receptors – integrins – and are thus associated indirectly with the intracellular actin cytoskeleton (Hynes 1987; Burridge et al. 1988). On the other hand, because the CW of plant cells, which is composed of cellulose microfibrils, is often constructed as a multilayered network and has a dynamic nature similar to that of the extracellular matrix in animal cells, the CW can be considered as the extracellular matrix of plant cells (Roberts 1989). Wyatt & Carpita (1993) reported that plant cells might also have a CW–PM–cytoskeleton continuum that is sensitive to exogenously applied synthetic RGD peptide (Wyatt & Carpita 1993; Canut et al. 1998; Mellersh & Heath 2001; Takagi et al. 2001; references therein). Therefore, such a continuum may be involved in the mechanisms of mechano-perception in plant cells (Baluška et al. 2003).

In our previous study, we demonstrated that cytoplasmic streaming in mesophyll cells of the aquatic angiosperm Vallisneria gigantea Graebner was transiently inhibited by hypertonic (Hayashi & Takagi 2003) or hypotonic (unpublished results) treatment. The response to hypertonic treatment depends strictly on the presence of extracellular Ca2+ and is inhibited by blockers of stretch-activated Ca2+ channels. Because cytoplasmic streaming in V. gigantea mesophyll cells is inhibited by high concentrations of Ca2+ (Takagi & Nagai 1986), we assume that the cessation of cytoplasmic streaming is mediated by an increase in [Ca2+]cyt caused by extracellular changes in osmotic pressure. Moreover, we have indicated that the response to hypertonic treatment is disrupted by plasmolysis, and that recovery of responsiveness requires trypsin-sensitive protein factor(s) (Hayashi & Takagi 2003). On the basis of observations of the types of plasmolyses (Oparka 1994), we have also confirmed that the intactness of the CW–PM adhesion is restored concomitantly with the recovery of responsiveness. Therefore, we assume that functional adhesion of the PM to the CW is necessary for the Ca2+ transient, which is mediated by stretch-activated Ca2+ channels, in the response of mesophyll cells of V. gigantea to hypertonic treatment. However, we had difficulty measuring the changes in [Ca2+]cyt in materials from this species. The most difficult problem was that we could not effectively introduce any type of Ca2+ indicators into the cytoplasm (Cyt), which formed a very thin layer.

In the present study, using mature Arabidopsis thaliana (L.) Heynh. leaves transformed genetically with apoaequorin (Harada, Sakai & Okada 2003), we demonstrated that both hypertonic and hypotonic treatments induce a transient and rapid increase in [Ca2+]cyt and that stretch-activated Ca2+ channels are involved in both responses. We further examined the possible involvement of cellulose microfibrils and other factors in the response to extracellular changes in osmotic pressure. In the light of our results, we discuss the different natures of the responses induced by hypertonic and hypotonic treatments.

MATERIALS AND METHODS

Plant materials

V. gigantea was grown under a cycle of 12 h of darkness and 12 h of light (irradiance of 0.5 W m−2 from white fluorescent lamps) at 20–25 °C, as described previously (Izutani, Takagi & Nagai 1990). T3 plants of transgenic A. thaliana[ecotype Wassilewskija (Ws)] expressing cytosolic apoaequorin (Harada et al. 2003) were grown at 22 °C under continuous light (irradiance of 8.5 W m−2 from white fluorescent lamps) on agar plates containing half-strength Murashige and Skoog medium (Wako Pure Chemical, Osaka, Japan), 0.05% 2-morpholinoethanesulfonic acid monohydrate, 300 µg mL−1 pyridoxine–HCl, 500 µg mL−1 nicotinic acid, 30 µg mL−1 kanamycin and 0.8% agar; the pH was adjusted to 5.7 with 1 m KOH.

Aequorin reconstitution and estimation of [Ca2+]cyt

Reconstitution of aequorin from expressed apoaequorin and exogenously applied coelenterazine was performed in vivo, as described by Harada et al. (2003). Briefly, one of the first four leaves was cut from a 2- to 3-week-old plant and incubated on freshly prepared coelenterazine cp (Molecular Probes, Eugene, OR, USA) at 2.5 µm in artificial pond water (APW) [0.05 mm KCl, 0.2 mm NaCl, 0.1 mm Ca(NO3)2, 0.1 mm Mg(NO3)2 and 2 mm piperazine-1,4-bis(2-ethanesulfonic acid) (PIPES) buffer at pH 7.0] in the dark for 14–20 h at 22 °C. The incubated leaves were then floated for varying periods of time on APW supplemented with the test chemical(s), as described below, before hypertonic treatment. Luminescence emitted from the leaves was measured with a luminometer (Lumicounter 2500; Microtech Nichi-on, Chiba, Japan). The data were stored on a computer running luminescence curve-analysing software (Microtech Nichi-on). All of the aequorin remaining in the leaves after each experiment was discharged by a series of rapid injections of ice-cold water. The in vivo Ca2+ concentrations were then estimated according to Baum et al. (1999).

Osmotic treatment and luminescence measurements

Each leaf in which aequorin had been reconstituted by the addition of coelenterazine was placed in a transparent plastic cuvette with 300 µL of APW supplemented with the test chemical(s). The cuvette was set on the holder of the Lumicounter in total darkness. Measurements were started after the intensity of luminescence had stabilized. Sixty seconds after the start of measurements, 700 µL of hypertonic APW containing the test concentration of sorbitol, to give a final concentration described in the text, was injected into the cuvette by inserting a 1 mL syringe into a light-tight port in the Lumicounter. For hypotonic treatment, APW lacking sorbitol was injected into the cuvette to give a final volume of 2 mL after hypertonic treatment. Occasionally, a slight change in aequorin luminescence was detected in response to APW treatment without sorbitol within 20 s after starting the treatment (e.g. in Fig. 2a, trace I). This might have been the result of a touch-induced increase in [Ca2+]cyt in response to application of any solutions (Knight et al. 1991). Consequently, we excluded each value of [Ca2+]cyt for the first 0–20 s of hypertonic treatment from our calculations of Δ[Ca2+]cyt. To examine the responsiveness of [Ca2+]cyt after plasmolysis and subsequent deplasmolysis, we used a pair of leaves derived from the same plant. The first hypertonic and hypotonic treatments were applied to one leaf, as mentioned above. During each measurement of aequorin luminescence, another leaf was treated for between 6 and 7 min with APW containing 0.7 m sorbitol. The concentration of sorbitol was then decreased gradually by addition of normal APW at a rate of about 10 mm s−1 to avoid any possible damage that a sudden increase in turgor pressure might cause, and finally the solution was replaced with APW lacking sorbitol. The second hypertonic and hypotonic treatments were applied to this leaf, as mentioned above. In a parallel experiment, we confirmed by video microscopy that almost all epidermal and mesophyll cells in the leaves bathed in hypertonic APW exhibited plasmolysis. All measurements were performed in a dark room at 22 °C.

Figure 2.

Ca2+ transient induced in Arabidopsis thaliana leaves by hypertonic treatment. (a & b) Typical examples of changes in cytoplasmic concentration of Ca2+ ([Ca2+]cyt) induced by treatment with various concentrations of sorbitol. In (a), a leaf was sequentially treated with normal artificial pond water (APW) (trace I) and then with APW containing 0.7 m sorbitol (trace II). In (b), leaves were treated with APW supplemented with 0.1–0.6 m sorbitol. Each arrowhead indicates the start of treatment. (c & d) Dependence of response on the concentration of sorbitol. The difference between [Ca2+]cyt before treatment and that at the peak (Δ[Ca2+]cyt) (c) and the duration of time required for the [Ca2+]cyt to reach the peak level (d) are each plotted against the concentration of sorbitol. Vertical bars indicate SE from three to nine independent experiments.

Microscopic observation

The procedures for the pretreatment of mesophyll cells of V. gigantea (Ryu, Takagi & Nagai 1995) and for microscopic observation (Hayashi & Takagi 2003) have been described in detail. Briefly, a leaf segment was cut into small pieces about 0.8 mm long to expose the end walls of the cells. The leaf pieces were incubated in APW under the original light regimen for 36–48 h and then treated with APW supplemented with the test chemical(s) before observation. Each leaf of A. thaliana in APW was degassed by a vacuum pump for 10 min and then cut into small pieces about 1 mm long. The leaf pieces were incubated in APW under the original light condition for 30 min, then treated with APW supplemented with the test chemical(s) before observation.

Each specimen was fastened onto a glass slide with a small amount of vaseline, in such a way that the end walls of the mesophyll cells of V. gigantea and the periclinal walls of the epidermal cells of A. thaliana were exposed; then a drop of normal APW was applied to each specimen. For hypertonic treatment, each specimen was perfused with APW supplemented with 0.4 or 0.5 m sorbitol under video microscopy. On the recorded optical images, the types of plasmolyses induced in the V. gigantea mesophyll cells and A. thaliana epidermal cells were examined. In addition, the five most rapidly moving cytoplasmic vesicles in the V. gigantea mesophyll cells were chosen, and the velocity of cytoplasmic streaming was calculated from the distance that each cytoplasmic vesicle moved in 1 s.

Staining of actin filaments

Actin filaments of epidermal cells of A. thaliana were visualized by staining with fluorescent phalloidin (Alexa Fluor 488 phalloidin; Molecular Probes, Eugene, Oregon, USA) as described by Ryu et al. (1995). After incubation in a staining solution for 20 min at 22 °C in darkness, the stained cells were examined with an epifluorescence microscope (BX50; Olympus, Tokyo, Japan). The optical images were captured digitally with a charge-coupled device camera (CoolSNAP; RS Photometrics, Tucson, AZ, USA).

Chemicals

1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA) was dissolved in 200 mm PIPES–NaOH (pH 7.0) as a 500 mm stock solution. The stock solution was diluted into APW that did not contain CaCl2. GdCl3 was dissolved in APW as a 10 mm stock solution. Latrunculin B (LatB) (Alexis Biochemicals, San Diego, CA, USA) was dissolved in dimethyl sulfoxide (DMSO) as a 2 mm stock solution. Each stock solution, except BAPTA, was diluted into normal APW. After pretreatment with APW supplemented with the test chemical(s) at the concentrations and for the periods of time indicated later in the text, the leaves were treated with the hypertonic solution containing 0.7 m sorbitol. APW supplemented with 0.05% DMSO was used as a control for LatB.

Cellulase (Onozuka RS; Yakult; Tokyo, Japan) was dissolved directly in APW. The synthetic hexapeptides GRGDSP (RGD peptide) and GRGESP [Arg-Gly-Glu (RGE) peptide] were purchased from Asahi Techno Glass (Chiba, Japan) and then dissolved in a buffered solution that contained 0.1 mm KCl, 0.1 mm NaCl, 1 mm ethyleneglycoltetraacetic acid (EGTA), 10 µm free Ca2+ ions (added as CaCl2) and 10 mmN-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid (HEPES), with the pH adjusted to 7.0 using NaOH, as described by Ryu et al. (1997). After pretreatment with either cellulase, RGD peptide or RGE peptide at the concentrations and for the periods of time indicated later in the text, the leaves were incubated in normal APW for 20 min before hypertonic treatment. All the experiments were carried out in a dark room at 22 °C.

RESULTS

Transient cessation of cytoplasmic streaming and type of plasmolysis induced by hypertonic treatments

As shown in our previous study (Hayashi & Takagi 2003), hypertonic treatment with APW containing 0.5 m sorbitol induced a transient cessation of cytoplasmic streaming in about 61% of V. gigantea mesophyll cells (n = 41). A typical example is demonstrated in Fig. 1a. After transient cessation of cytoplasmic streaming, responsive cells exhibited a concave-type plasmolysis in the hypertonic solution (Fig. 1c). The concave-type plasmolysis is assumed to reflect the strong adhesion of the PM to the CW (Oparka 1994). After pretreatment with APW supplemented with 1% cellulase for 3 h at 22 °C, the response to hypertonic treatment was reduced to about 28% of the observed mesophyll cells (n = 75). A typical example is demonstrated in Fig. 1b. In mesophyll cells that did not respond to hypertonic treatment after pretreatment with cellulase, the PM was detached smoothly and completely from the CW to instantaneously produce a convex-type plasmolysis (Fig. 1d). This convex-type plasmolysis is assumed to reflect the weak adhesion of the PM to the CW (Oparka 1994). In addition, after pretreatment with 1% cellulase for 3 h followed by incubation in normal APW for 18 h, the response to hypertonic treatment was substantially recovered (data not shown) and almost all the responsive cells exhibited a concave-type plasmolysis, as shown in Fig. 1c. Consequently, we concluded tentatively that cellulose microfibrils are involved in the adhesion of the PM to the CW; this adhesion may exert a stretching force on the PM in a hypertonic solution and bring about the opening of stretch-activated Ca2+ channels.

Figure 1.

Effects of cellulase on the transient cessation of cytoplasmic streaming and on the type of plasmolysis induced by hypertonic treatment in mesophyll cells of Vallisneria gigantea. (a & b) Typical examples of changes in the velocity of cytoplasmic streaming induced by treatment with artificial pond water (APW) containing 0.5 m sorbitol after pretreatment with normal APW for 3 h (a) or with APW that contained 1% cellulase (b). Hypertonic treatment was started at 0 s (arrowhead). Vertical bars indicate SE from five cytoplasmic vesicles. (c) Concave-type plasmolysis induced by hypertonic treatment in the cell analysed in (a). (d) Convex-type plasmolysis induced by hypertonic treatment in the cell analysed in (b). The time after the start of hypertonic treatment is given in seconds in the upper-right corner of each photograph (c & d). Outlines of the cell wall (CW) and plasma membrane (PM) were traced and are illustrated in the lower panels of (c) and (d). Cyt, cytoplasm; P, plastid. Scale bars = 10 µm.

Ca2+ transient induced by hypertonic and hypotonic treatments

To measure the changes in [Ca2+]cyt induced by extracellular changes in osmotic pressure, we used A. thaliana leaves that expressed cytoplasmic aequorin. In rosette leaves from 3-week-old plants of A. thaliana, the values estimated for the resting level of [Ca2+]cyt from [Ca2+]cyt-dependent aequorin luminescence in APW lay between 40 and 100 nm in most experiments. The [Ca2+]cyt was not altered when the leaf was treated gently with normal APW (Fig. 2a, trace I). By contrast, the treatment of the same leaf with APW that contained 0.7 m sorbitol induced a rapid, transient increase in [Ca2+]cyt (Fig. 2a, trace II). It took about 35 s after the start of hypertonic treatment for the [Ca2+]cyt to reach the peak level, which was about 360 nm. The [Ca2+]cyt decreased smoothly for about 180 s and then returned gradually to the original level, taking a few minutes. Figure 2b shows typical examples of the changes in [Ca2+]cyt induced by treatment with various concentrations of sorbitol. A Ca2+ transient was rarely observed after treatment with sorbitol at concentrations lower than 0.1 m, but was observed clearly at concentrations of 0.2 m or higher. The increase in [Ca2+]cyt induced by each treatment (Δ[Ca2+]cyt) was calculated as the difference between the [Ca2+]cyt before treatment and that at the peak. As the concentration of sorbitol increased from 0.2 to 0.7 m, the average Δ[Ca2+]cyt appeared to increase in a concentration-dependent manner (Fig. 2c), and the [Ca2+]cyt peaked more rapidly (Fig. 2d). In a parallel experiment, we observed that mesophyll and epidermal cells exhibiting signs of plasmolysis first appeared with 0.2 m sorbitol (data not shown).

Immediately after the [Ca2+]cyt had returned to the resting level following the transient increase induced by treatment with 0.7 m sorbitol (Fig. 3a, trace I), we applied hypotonic treatment to the same leaf (Fig. 3a, arrow). Upon hypotonic treatment, the leaf also exhibited a Ca2+ transient (Fig. 3a, trace I′). Although the pattern of response induced by hypotonic treatment was similar to that induced by hypertonic treatment, the [Ca2+]cyt increase by hypotonic treatment took a longer time – about 10 min – to return to the resting level (data not shown). Similarly, after overnight incubation in APW that contained 0.1 m sorbitol, other leaves of the same plant were treated with normal APW. The milder hypotonic treatment also induced a Ca2+ transient (data not shown). Using transgenic A. thaliana leaves, we thus revealed that both hypertonic and hypotonic treatments can induce a rapid, transient increase in [Ca2+]cyt.

Figure 3.

Effects of 1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA) on the Ca2+ transient induced by hypertonic and hypotonic treatments in Arabidopsis thaliana leaves. (a) Typical examples of the changes in cytoplasmic concentration of Ca2+ ([Ca2+]cyt) induced by sequential hypertonic and hypotonic treatments. A leaf was treated with artificial pond water (APW) containing 0.7 m sorbitol (trace I), and then with an equivalent volume of normal APW (trace I′). Another leaf derived from the same plant was pretreated for 30 min with Ca2+-free APW that contained 5 mm BAPTA; the leaf was then treated with Ca2+-free APW supplemented with 0.7 m sorbitol (trace II), and then with Ca2+-free APW (trace II′). Hypertonic and hypotonic treatments were started at 0 s (arrowhead) and at 360 s (arrow), respectively. (b & c) Δ[Ca2+]cyt induced by hypertonic (b) and hypotonic (c) treatments, respectively, in the absence (open column) or presence (closed column) of BAPTA. Vertical bars indicate SE from eight independent experiments.

Involvement of Ca2+ influx in the Ca2+ transient induced by hypertonic and hypotonic treatments

After pretreatment of leaves with the Ca2+-chelating agent BAPTA at 5 mm for 30 min, the Ca2+ transient induced by hypertonic treatment was almost completely inhibited (Fig. 3a, trace II, b). On the other hand, the response to hypotonic treatment was incompletely but substantially suppressed to about 54% of the control response in the presence of 5 mm BAPTA (Fig. 3c). We further examined the effects of Gd3+, a potential blocker of stretch-activated Ca2+ channels (Yang & Sachs 1989). Gd3+ inhibited the response to hypertonic treatment to about 35% of the control response at 0.1 mm, and almost completely at 1 mm(Fig. 4a, trace II, b). Although the inhibitory effect of Gd3+ on the response to hypotonic treatment was very small at 0.1 mm, Gd3+ at 1 mm suppressed the response to about 40% of the control level (Fig. 4c). These results strongly suggest that a Ca2+ influx across the PM, probably through stretch-activated Ca2+ channels, was induced not only by hypertonic treatment but also by hypotonic treatment.

Figure 4.

Effects of Gd3+ on the Ca2+ transient induced by hypertonic and hypotonic treatments in Arabidopsis thaliana leaves. (a) Typical examples of the changes in cytoplasmic concentration of Ca2+ ([Ca2+]cyt) induced by sequential hypertonic and hypotonic treatments. A leaf was treated with artificial pond water (APW) containing 0.7 m sorbitol (trace I), and then with an equivalent volume of normal APW (trace I′). Another leaf derived from the same plant was pretreated for 30 min with APW that contained 1 mm Gd3+ the leaf was then treated with the Gd3+-containing APW supplemented with 0.7 m sorbitol (trace II), and then with Gd3+; containing APW (trace II′). Hypertonic and hypotonic treatments were started at 0 s (arrowhead) and at 360 s (arrow), respectively. (b & c) Dependence of the Δ[Ca2+]cyt induced by hypertonic (b) and hypotonic (c) treatments, respectively, on the concentration of Gd3+. Vertical bars indicate SE from six to nine independent experiments.

Effects of plasmolysis and deplasmolysis on the Ca2+ transient induced by hypertonic and hypotonic treatments

We demonstrated previously that the responsiveness of V. gigantea mesophyll cells to hypertonic treatment was drastically impaired by plasmolysis and deplasmolysis (Hayashi & Takagi 2003). In the present study, we asked whether the same was true of the responsiveness of [Ca2+]cyt in A. thaliana leaves. Figure 5a shows typical examples of the response to hypertonic and hypotonic treatments in a pair of leaves derived from the same plant. The first hypertonic and hypotonic treatments each induced a Ca2+ transient in one leaf (Fig. 5a, traces I & I′). Another leaf was treated with 0.7 m sorbitol to plasmolyse the cells, and then was treated gently with APW lacking sorbitol to de-plasmolyse the cells (see Materials and Methods). In such leaves, the resting level of [Ca2+]cyt was estimated to be slightly higher than in the first leaf (< 150 nm), although the reasons for this are unclear. Immediately after deplasmolysis, the Ca2+ transient induced by the second hypertonic treatment was markedly reduced (Fig. 5a, trace II, b). By contrast, the response to the second hypotonic treatment was clearly observed (Fig. 5a, trace II′), although it was reduced to about 40% of the first hypotonic treatment (Fig. 5c).

Figure 5.

Effects of plasmolysis and deplasmolysis on the Ca2+ transient induced by hypertonic and hypotonic treatments in Arabidopsis thaliana leaves. (a) Typical examples of the changes in cytoplasmic concentration of Ca2+ ([Ca2+]cyt) induced by sequential hypertonic and hypotonic treatments. A leaf was treated with artificial pond water (APW) containing 0.7 m sorbitol (trace I), and then with an equivalent volume of normal APW (trace I′). Another leaf derived from the same plant was subjected to the same plasmolysis and subsequent deplasmolysis, and immediately afterwards it was treated again with APW supplemented with 0.7 m sorbitol (trace II), and then with normal APW (trace II′). Hypertonic and hypotonic treatments were started at 0 s (arrowhead) and at 300 s (arrow), respectively. (b & c) Δ[Ca2+]cyt induced by hypertonic (b) and hypotonic (c) treatments, respectively, applied for the first (open column) or second (closed column) time to the leaves. (d) Recovery of response to the second hypertonic treatment. The ratio of Δ[Ca2+]cyt induced by the second hypertonic treatment to that by the first hypertonic treatment (recovery ratio) was plotted as a percentage against the time of incubation after deplasmolysis. Vertical bars indicate SE from three to six independent experiments.

We examined the types of plasmolysis induced in A. thaliana epidermal cells bathed in a hypertonic APW containing 0.4 m sorbitol (Fig. 6). The cells exhibited a typical concave-type plasmolysis, in which Hechtian strand-like structures (Reuzean & Pont-Lezica 1995) connected the protoplasm to the CW (Fig. 6c′), in response to the first hypertonic treatment (Fig. 6a–c). Those cells were then mildly de-plasmolysed in APW lacking sorbitol. When the cells, immediately after deplasmolysis, were subjected to the second hypertonic treatment (Fig. 6d–f), they exhibited a typical convex-type plasmolysis, in which the PM was detached smoothly and completely from the CW (Fig. 6f′).

Figure 6.

Concave- and convex-type plasmolysis induced in epidermal cells of Arabidopsis thaliana by hypertonic treatment. Optical images of the cytoplasmic layer along the outer periclinal wall of the same cell were captured successively before treatment (a), in the process of plasmolysis induced by a first hypertonic treatment with 0.4 m sorbitol (b), at the time of concave-type plasmolysis (c), immediately after deplasmolysis in normal artificial pond water (APW) (d), in the process of plasmolysis induced by a second hypertonic treatment with 0.4 m sorbitol (e) and at the time of convex-type plasmolysis (f). The time after the start of hypertonic treatment is given in seconds in the upper-right corner of each photograph. Outlines of the cell wall (CW) and plasma membrane (PM) in (c) and (f) were traced and are illustrated in (c′) and (f′), respectively. Arrowheads in (c′) indicate Hechtian strand-like structures. Cyt, cytoplasm. Scale bars = 10 µm.

We further tested the recovery of the increase in [Ca2+]cyt induced by the second hypertonic treatment. Although the leaves had hardly responded to the second hypertonic treatment after about 15 min of deplasmolysis, after incubation in APW for 1, 3 and 6 h their responsiveness recovered gradually to about 37, 64 and 78%, respectively, of that after the first hypertonic treatment (Fig. 5d). These results may support the idea that some mechanism, which is impaired by plasmolysis and deplasmolysis, functions in the response to hypertonic treatment.

Effects of cellulase on the Ca2+ transient induced by hypertonic and hypotonic treatments

As described above, cellulose microfibrils may be involved in the transient cessation of cytoplasmic streaming induced by hypertonic treatment and in the CW–PM adhesion of V. gigantea mesophyll cells (Fig. 1). We examined the effects of cellulase on the Ca2+ transient induced by extracellular changes in osmotic pressure in A. thaliana leaves. After pretreatment with 1% cellulase for 3 h at 22 °C, the Ca2+ transient induced by hypertonic treatment was significantly inhibited (Fig. 7a, trace II) in comparison with that induced without cellulase pretreatment (Fig. 7a, trace I). The Δ[Ca2+]cyt was reduced to approximately 43% of the control value (Fig. 7b). After pretreatment with 1% cellulase for 3 h followed by incubation in normal APW for 18 h, the Δ[Ca2+]cyt induced by hypertonic treatment recovered to close to the control value (Fig. 7b). By contrast, the response to hypotonic treatment was not inhibited by pretreatment with cellulase at all (Fig. 7a, traces I′ & II′), but rather seemed to be promoted (Fig. 7c). Moreover, we confirmed that the response to hypotonic treatment after pretreatment with cellulase was considerably reduced in the presence of 1 mm Gd3+ (data not shown). These results suggest that (a) cellulase-sensitive factor(s) contribute(s) to the response only to hypertonic treatment, although stretch-activated Ca2+ channels are involved in the responses induced by both hypertonic and hypotonic treatments.

Figure 7.

Effects of cellulase on the Ca2+ transient induced by hypertonic and hypotonic treatments in Arabidopsis thaliana leaves. (a) Typical examples of the changes in cytoplasmic concentration of Ca2+ ([Ca2+]cyt) induced by sequential hypertonic and hypotonic treatments. A leaf was treated with artificial pond water (APW) containing 0.7 m sorbitol (trace I), and then with an equivalent volume of normal APW (trace I′). Another leaf derived from the same plant was pretreated for 3 h with APW that contained 1% cellulase; the leaf was then treated with APW supplemented with 0.7 m sorbitol (trace II), and then with normal APW (trace II′). Hypertonic and hypotonic treatments were started at 0 s (arrowhead) and at 360 s (arrow), respectively. (b & c) Δ[Ca2+]cyt induced by hypertonic (b) and hypotonic (c) treatments, respectively, after pretreatment of the leaves either for 3 h with normal APW (control), for 3 h with APW that contained 1% cellulase (cellulase → APW 0 h) or for 3 h with APW that contained 1% cellulase, followed by normal APW for another 18 h (cellulase → APW 18 h). Vertical bars indicate SE from five or six independent experiments.

Effects of LatB on the Ca2+ transient induced by hypertonic and hypotonic treatments

Several studies have demonstrated that structural changes in actin filaments modulate the activities of ion channels in animal (Cantiello et al. 1993) and plant (Hwang et al. 1997) cells. By staining A. thaliana leaves with fluorescent phalloidin, we confirmed that numerous linear bundles of actin filaments appeared to be stretched across the whole epidermal cells (Fig. 8a). By contrast, after pretreatment for 20 min with 1 µm LatB, which is known to interfere with normal polymerization of actin filaments, such bundles of actin filaments were almost completely disrupted, except for some faint, cloudy fluorescence around the periphery of the cells, in both the epidermal (Fig. 8b) and mesophyll (data not shown) cells. However, in the presence of 1 µm LatB, the Ca2+ transient was induced by hypertonic and hypotonic treatments in similar manners to those in its absence (Fig. 9). In addition, we confirmed that simple treatment of leaves with APW that contained LatB did not affect the [Ca2+]cyt (data not shown).

Figure 8.

Effects of latrunculin B (LatB) on the configuration of actin filaments in epidermal cells of Arabidopsis thaliana. Actin filaments on the outer periclinal walls were visualized by staining with fluorescent phalloidin after pretreatment with artificial pond water (APW) that contained 0.05% dimethyl sulfoxide (DMSO) (a) or 1 µm LatB (b) for 20 min. (a′ & b′) Bright-field images of (a) and (b), respectively. Scale bars = 10 µm.

Figure 9.

Effects of latrunculin B (LatB) on the Ca2+ transient induced by hypertonic and hypotonic treatments in Arabidopsis thaliana leaves. (a) Typical examples of the changes in cytoplasmic concentration of Ca2+ ([Ca2+]cyt) induced by sequential hypertonic and hypotonic treatments. A leaf was treated with artificial pond water (APW) containing 0.7 m sorbitol (trace I), and then with an equivalent volume of normal APW (trace I′). Another leaf derived from the same plant was pretreated for 20 min with APW that contained 1 µm LatB; the leaf was then treated with the LatB-containing APW supplemented with 0.7 m sorbitol (trace II), and then with the LatB-containing APW (trace II′). Hypertonic and hypotonic treatments were started at 0 s (arrowhead) and at 360 s (arrow), respectively. The solutions used in each experiment contained 0.05% dimethyl sulfoxide (DMSO). (b & c) Δ[Ca2+]cyt induced by hypertonic (b) and hypotonic (c) treatments, respectively, in the absence (open column) or presence (closed column) of LatB. Vertical bars indicate SE from six independent experiments.

We further examined the effects of an RGD peptide. Neither the RGD peptide nor the control RGE peptide had a significant effect on the responses, in terms of the Δ[Ca2+]cyt and the duration of time required for [Ca2+]cyt to reach the peak level, induced by hypertonic (Fig. 10) and hypotonic (data not shown) treatments. In A. thaliana, neither actin-dependent nor RGD-sensitive adhesion of the PM to the CW appears to function in the sensing of extracellular changes in osmotic pressure.

Figure 10.

Effects of Arg-Gly-Asp (RGD) peptide on the Ca2+ transient induced by hypertonic treatment in Arabidopsis thaliana leaves. After pretreatment of the leaves for 18 h with either a buffer solution (control), a buffer solution that contained 1 mm RGD peptide (RGD) or a buffer solution that contained 1 mm Arg-Gly-Glu (RGE) peptide [a control peptide (RGE)], the leaves were treated with artificial pond water (APW) supplemented with 0.7 m sorbitol. Δ[Ca2+]cyt (a) and the duration of time required for the cytoplasmic concentration of Ca2+ ([Ca2+]cyt) to reach the peak level (b) detected in each experiment are shown. Vertical bars indicate SE from five to seven independent experiments.

DISCUSSION

Ca2+ transient induced by hypertonic and hypotonic treatments in A. thaliana leaves

We clearly showed that the [Ca2+]cyt in A. thaliana leaves that expressed aequorin was transiently increased within several tens of seconds after hypertonic and hypotonic treatments (Figs 2a & 3a). We can infer that the transient cessation of cytoplasmic streaming induced by hypertonic treatment in V. gigantea mesophyll cells (Hayashi & Takagi 2003; Fig. 1) is mediated by a transient increase in [Ca2+]cyt. The Ca2+ transients induced by hypertonic and hypotonic treatments were sensitive to Gd3+ (Fig. 4), and the response to hypertonic treatment was induced by sorbitol in a concentration-dependent manner (Fig. 2). Because Gd3+ is known to inhibit mechano-sensitive ion channels in the PM (Yang & Sachs 1989), we concluded that hypertonic and hypotonic treatments are perceived as mechanical stimuli in A. thaliana leaves. On the other hand, a putative osmosensor protein that is orthologous to Sln1 in yeast has been identified in plants (Urao et al. 1999). Such an osmosensor might also be implicated in upstream of the calcium-signalling pathway for the response to extracellular changes in osmotic pressure.

The time required for the [Ca2+]cyt to return to the original resting level was longer in the hypotonic response than in the hypertonic response (Fig. 3a, traces I & I′). This might be due to the second and prolonged Ca2+ transient, regulated by protein phosphorylation, which is induced by hypotonic treatment (Takahashi et al. 1997). Unique characteristics in the manner of changes in [Ca2+]cyt have also been reported in responses to various kinds of stimuli, such as heat shock, cold shock and wind (Knight et al. 1992; Gong et al. 1998; Trewavas & Malhó 1998; Plieth 2001). However, there is also a possibility that different responses induced by hypertonic and hypotonic treatments were caused by different populations of responsive cells (Kiegle et al. 2000). This problem might be solved by producing transgenic A. thaliana plants that express aequorin only in specific types of cells or tissues.

The Ca2+ transient induced by hypertonic treatment was inhibited completely in the presence of BAPTA or Gd3+ (Figs 3 & 4), and that induced by hypotonic treatment was inhibited considerably by each of these chemicals (Figs 3 & 4). These results strongly suggest that an influx of extracellular Ca2+ through the stretch-activated Ca2+ channels plays an essential role in the Ca2+ transient induced by extracellular changes in osmotic pressure. The inhibitory effects of BAPTA on the response (Fig. 3) might be attributable not only to the depletion of extracellular Ca2+ but also to the loss of responsiveness because of changes in the nature of adhesion between the CW and the PM (Decreux & Messiaen 2005). Nevertheless, the crucial requirement of extracellular Ca2+ to the Ca2+ transient is evident because Ca2+ channel blockers did not affect the type of plasmolysis in the mesophyll cells of V. gigantea (data not shown), but completely inhibited the response even in the presence of extracellular Ca2+ in both V. gigantea mesophyll cells (Hayashi & Takagi 2003) and A. thaliana leaves (Fig. 4). On the other hand, many studies have demonstrated that the Ca2+ source for the response to hypotonic treatment is not only extracellular (Okazaki & Tazawa 1990; Taylor et al. 1996; Takahashi et al. 1997; Stento et al. 2000) but also intracellular (Tazawa et al. 1995; Gao et al. 2004). In the internodal cells of a freshwater charophyte, cytosolic hydration caused by hypotonic treatment induced a release of Ca2+ from internal Ca2+ stores (Tazawa et al. 1995; Shimada, Kikuyama & Tazawa 1996). Because the inhibitory effects of BAPTA and Gd3+ on the response to hypotonic treatment were not complete (Figs 3c & 4c), mechanisms other than stretch-activated Ca2+ channels located at the PM might also participate in the response to hypotonic treatment in A. thaliana leaves.

Role of CW–PM adhesion in the Ca2+ transient induced by hypertonic treatment

In the present study, we showed that both the Ca2+ transient induced by a second hypertonic treatment (Fig. 5b) and the adhesion of the PM to the CW (Fig. 6f) were impaired immediately after plasmolysis and deplasmolysis in A. thaliana. These results suggest that in A. thaliana leaves, a similar mechanism to that operating in V. gigantea (Hayashi & Takagi 2003) might function in the Ca2+ transient induced by hypertonic treatment and might be disrupted by physical detachment of the PM from the CW. We confirmed that the first and second hypertonic treatments of A. thaliana epidermal cells with sorbitol at the same concentration induced plasmolysis with similar time courses (Fig. 6). Therefore, the activation of some mechanisms for osmoregulation, including accumulation of compatible solutes (Bartels & Nelson 1994; Bohnert et al. 1995), does not likely affect the response to the second hypertonic treatment under the present condition. On the other hand, responsiveness gradually recovered over about 6 h of incubation after deplasmolysis (Fig. 5d). This time course was similar to that of the recovery of responsiveness of [Ca2+]cyt after a second heat shock in tobacco seedlings (Gong et al. 1998). Because the response of [Ca2+]cyt to cold treatment was retained in A. thaliana leaves after plasmolysis and deplasmolysis (data not shown), this refractory period was not caused by some modification of aequorin. Although the recovery of responsiveness to hypertonic treatment after 6 h of deplasmolysis was inhibited by 10 µm cycloheximide, a general inhibitor of protein synthesis, we could not conclude that protein synthesis is genuinely involved in the recovery process, because the total aequorin luminescence was considerably reduced in the presence of cycloheximide (data not shown).

By contrast, the Ca2+ transient induced by the second hypotonic treatment was clearly detected immediately after plasmolysis and deplasmolysis (Fig. 5a, trace II′). The result indicates that even when the mechanism for the response to hypertonic treatment is abolished by plasmolysis and deplasmolysis, because the protoplasm can easily increase its surface area upon hypotonic treatment, both stretching of the PM and activation of Ca2+ channels may still occur normally. The result also excludes the possibility that inhibition of the response to hypertonic treatment by plasmolysis was due to physical impairment of the Ca2+ channels.

Involvement of cellulose microfibrils in the Ca2+ transient induced by hypertonic treatment

Cellulose microfibrils are synthesized at the outer surface of the PM (Delmer & Amor 1995; references therein) and are thought to play a role as direct and/or indirect scaffolding for the adhesion of the PM to the CW. After pretreatment with cellulase, both the transient cessation of cytoplasmic streaming in V. gigantea mesophyll cells (Fig. 1b) and the Ca2+ transient in A. thaliana leaves (Fig. 7b) induced by hypertonic treatment were inhibited. We confirmed that the unresponsive mesophyll cells of V. gigantea exhibited a convex-type plasmolysis, in which the protoplasm was retracted smoothly from the CW (Fig. 1d). Moreover, after 18 h of removal of cellulase, the V. gigantea mesophyll cells again exhibited the concave-type plasmolysis, concomitantly with the recovery of responsiveness to hypertonic treatment (data not shown). Although we could not obtain quantitative data on the pattern of plasmolysis in the A. thaliana epidermal cells after pretreatment with cellulase, mainly because of the shape and size of the cells, these results strongly suggest that cellulose microfibrils are involved not only in the adhesion of the PM to the CW, but also in the Ca2+ transient (which is probably mediated by stretch-activated Ca2+ channels) induced by hypertonic treatment. A possible involvement of the CW in jasmonate- and ethylene-dependent signalling pathways has also been suggested (Ellis et al. 2002). Therefore, the effects of cellulase on the Ca2+ response to hypertonic treatment might be attributed to impairments of some common signalling complex(es) other than the mechano-sensory pathway. However, even when the adhesion of the PM to the CW was abolished by cellulase, the Ca2+ transient was normally induced by hypotonic treatment (Fig. 7c). This may be compatible with the results that plasmolysis and deplasmolysis did not totally abolish the Ca2+ transient induced by hypotonic treatment (Fig. 5c).

On the other hand, it has been reported that a Ca2+ transient is induced by injection of an isotonic solution into tobacco protoplasts, and that external Ca2+ is not essential, but instead Ca2+ release from internal stores is responsible for the response (Haley et al. 1995). These results indicate that mechanisms that do not depend on the adhesion of the PM to the CW and on stretch-activated Ca2+ channels in the PM might also work in the perception of mechanical stimuli in plant cells.

Role of actin filaments in the Ca2+ transient induced by hypertonic and hypotonic treatments

In our previous study (Hayashi & Takagi 2003), we could not discuss the role of actin filaments in the response to hypertonic treatment, because actin-dependent cytoplasmic streaming was used as an indicator of responsiveness. In this study, we demonstrated that although LatB almost completely disrupted the actin filaments in A. thaliana leaf cells (Fig. 8), the Ca2+ transient was normally induced by hypertonic and hypotonic treatments in those samples (Fig. 9). While there is a possibility that the resistant actin filaments to LatB still survived in the vicinity of the PM (Fig. 8b) under the present experimental conditions, our results strongly suggest that actin filaments do not work upstream of the activation of Ca2+ channels in the signalling pathways evoked by extracellular changes in osmotic pressure.

In addition, the RGD peptide had no effects on the Ca2+ transient induced by hypertonic (Fig. 10) and hypotonic (data not shown) treatments of A. thaliana. The Ca2+-dependent cessation of cytoplasmic streaming induced by hypertonic treatment of mesophyll cells of V. gigantea is also insensitive to RGD peptide (Hayashi & Takagi 2003), although the patterns of streaming were sensitive (Ryu et al. 1997). These results indicate that RGD-sensitive adhesion of the PM to the CW may function in the maintenance of organization of the actin cytoskeleton, but is not involved in the perception of extracellular changes in osmotic pressure. Recently, several attractive candidates for putative plant-specific linker molecules, including cell-wall-associated kinases (He, Fujiki & Kohorn 1996), arabinogalactan proteins (Schultz et al. 1998; Shi et al. 2003), pectins and cellulose synthases, have been demonstrated (Baluška et al. 2003; references therein). These factors might also be implicated in the perception of mechanical stimuli, including extracellular changes in osmotic pressure.

ACKNOWLEDGMENTS

We acknowledge Professor/Dr I. Terashima of Osaka University for critical discussions. We thank Drs S. Inada and T. Mayama of RIKEN Plant Science Center for critical reading of the manuscript. This work was partly supported by Grant-in-Aid for Scientific Research no. 16570033 from the Japanese Society for the Promotion of Science and by the 21st Century Center of Excellence (COE) Program.

Ancillary