Aluminum Induces a Decrease in Cytosolic Calcium Concentration in BY-2 Tobacco Cell Cultures

Abstract

Al toxicity is a major problem that limits crop productivity on acid soils. It has been suggested that Al toxicity is linked to changes in cellular Ca homeostasis and the blockage of plasma membrane Ca²⁺-permeable channels. BY-2 suspension-cultured cells of tobacco (Nicotiana tabacum L.) exhibit rapid cell expansion that is sensitive to Al. Therefore, the effect of Al on changes in cytoplasmic free Ca concentration ([Ca²⁺]_cyt) was followed in BY-2 cells to assess whether Al perturbed cellular Ca homeostasis. Al exposure resulted in a prolonged reduction in [Ca²⁺]_cyt and inhibition of growth that was similar to the effect of the Ca²⁺ channel blocker La³⁺ and the Ca²⁺ chelator ethyleneglycol-bis(β-aminoethyl ether)-N,N′-tetraacetic acid. The Ca²⁺ channel blockers verapamil and nifedipine did not induce a decrease in [Ca²⁺]_cyt in these cells and also failed to inhibit growth. Al and La³⁺, but not verapamil or nifedipine, reduced the rate of Mn²⁺ quenching of Indo-1 fluorescence, which is consistent with the blockage of Ca²⁺- and Mn²⁺-permeable channels. These results suggest that Al may act to block Ca²⁺ channels at the plasma membrane of plant cells and this action may play a crucial role in the phytotoxic activity of the Al ion.

Crop production is severely limited in many areas of the world where the low pH of acidic soils solubilizes the rhizotoxic, trivalent metal Al³⁺. Al has been shown to rapidly (<1 h) inhibit both primary root and root hair growth, resulting in poor nutrient acquisition, and consequently leading to shoot nutrient deficiencies and poor crop yields (Taylor, 1990; Kochian, 1995). Although the actively dividing and expanding cells of the root apex have been identified as the principal site of toxicity (Ryan et al., 1993), the causes of Al toxicity have remained elusive. It is known that Al can rapidly enter the cytoplasm (Lazof et al., 1994), but it is still far from clear whether the primary site(s) of toxicity is external (i.e. interactions with the cell wall or external face of plasma membrane) or internal (affecting cytoplasmic functions or activities in internal membranes/compartments). After prolonged exposure (e.g. >12 h), Al can affect many physiological processes either directly or indirectly (Kochian, 1995); however, to date, none of these inhibited processes have been correlated with the growth inhibition event.

It has been postulated by numerous authors that Al may interfere with cellular Ca²⁺ homeostasis, leading to a breakdown of the Ca²⁺-dependent signal transduction cascades that may be necessary for both cell division and cell elongation (Haug, 1984; Taylor, 1990; Rengel, 1992; Delhaize and Ryan, 1995; Kochian, 1995). Tentative evidence for this was provided by the fact that high external Ca²⁺, as well as other ions, can ameliorate Al toxicity (Rengel, 1992; Kinraide et al., 1994; Pineros and Tester, 1995). It was recently shown that, at the toxic concentrations normally found in soils (10–100 μm), Al³⁺ is capable of blocking voltage-gated plasma membrane Ca²⁺ channels and disrupting inositol 1,4,5-trisphosphate-mediated signaling events in wheat roots (Jones and Kochian, 1995; Huang et al., 1996). Other potential intracellular target sites for Al include occupation of Ca²⁺-binding sites in Ca²⁺-requiring enzymes and proteins (e.g. phospholipase C, calmodulin), the complexing of ligands required by Ca²⁺-dependent enzymes (e.g. ATP for Ca²⁺-ATPase), the prevention of Ca²⁺-mediated vesicle fusion, and the alteration of Ca²⁺-mediated cytoskeletal dynamics (Haug, 1984; Taylor, 1990; Rengel, 1992; Delhaize and Ryan, 1995; Kochian, 1995). Thus, Al may affect diverse aspects of Ca²⁺-regulated cellular events that may in turn disrupt cell division and expansion.

The role of Ca²⁺ in plant cell division and expansion is still being defined. Transient changes in Ca²⁺ have been observed to accompany mitotic progression (for review, see Hepler, 1994) and an involvement of Ca²⁺ in the machinery that performs nuclear envelope breakdown, nuclear envelope reformation, cell plate formation, and ana-phase progression have been proposed (Hepler, 1994; Jurgens et al., 1994; Staehelin and Hepler, 1996). The role of cytoplasmic Ca²⁺ in cell expansion remains more elusive. Ca²⁺ promotes elongation in many plant cells (Takahashi et al., 1992; Hyde and Heath, 1995; Levina et al., 1995), Ca²⁺ antagonists can block elongation growth (Muto and Hirosawa, 1987; Jackson and Hall, 1993; Cho and Hong, 1995), and changes in [Ca²⁺]_cyt may accompany growth-altering hormonal treatments such as auxin application (Gehring et al., 1990). Sustained gradients in Ca²⁺ have also been shown to be central regulators of expansion of tip-growing plant cells such as pollen tubes (Herth et al., 1990; Miller et al., 1992; Pierson et al., 1994, 1996; Malh 243 et al., 1995) and root hairs (Clarkson et al., 1988; Schiefelbein et al., 1992; Herrmann and Felle, 1995; Felle and Hepler, 1997; Wymer et al., 1997).

Irrespective of the proposed role of Ca²⁺ changes in regulating cell expansion and division, maintained homeostatic control of [Ca²⁺]_cyt is known to be essential for continued cell viability (Bush, 1995). Disrupting this homeostatic system represents one of the most widely proposed explanations of Al toxicity (Rengel, 1992). Indeed, Al has been reported to induce a rapid, transient increase in [Ca²⁺]_cyt in wheat root protoplasts (Lindberg and Strid, 1997). We have recently shown that Al toxicity in growing root hairs of Arabidopsis thaliana involves a disruption in [Ca²⁺]_cyt (D.L. Jones, S. Gilroy, and L.V. Kochian, unpublished). However, the inhibition of root hair elongation occurred as much as 20 min before a detectable change in the root hair [Ca²⁺], suggesting that this disruption in [Ca²⁺]_cyt was not required to initiate the process of Al toxicity. Similarly, measurements of Al effects on Ca²⁺ fluxes into root hairs using a vibrating Ca²⁺-selective microelectrode system have revealed that the Al levels that inhibited root hair growth failed to block Ca²⁺ fluxes (Jones et al., 1995). These results suggest that Al toxicity is not always preceded by an alteration in Ca²⁺ homeostasis in these cells. However, the disruption of the tip growth of root hairs may represent a unique mechanism of Al toxicity compared with the action of this ion on the dividing or diffuse growing cells of the root apex.

Cultured tobacco (Nicotiana tabacum L.) cells have proved to be a highly useful system to analyze the mechanism(s) of Al toxicity (Yamamoto et al., 1994, 1996; Ono et al., 1995; Ezaki et al., 1996). These cultures undergo rapid cell division and expansion and grow as single cells or small groups of cells that are readily visible using fluorescence microscopy. Actively growing cultured cells exhibit Al toxicity, whereas those in stationary phase are resistant (Yamamoto et al., 1994), which is consistent with the finding that the site of Al toxicity in roots seems to be limited to the actively growing cells of the apex (Ryan et al., 1993). We therefore tested whether Al could affect expansion in these cultured tobacco cells and whether this effect could be mediated through a blockage of Ca channels in the plasma membrane. We present data showing that Al treatment does lower [Ca²⁺]_cyt and that this effect is mimicked by Ca channel antagonists such as La. These data suggest that Al can block Ca²⁺-permeable channels in higher plant cells and that this action may interfere with the normal Ca²⁺ homeostasis required for sustained cell division and expansion.

MATERIALS AND METHODS

Tobacco (Nicotiana tabacum L. cv BY-2) suspension-cultured cells were maintained as described by Kuss-Wymer and Cyr (1992). The basic growth medium (pH 5.0) contained the following macronutrients (mm): KNO₃, 30.0; NH₄NO₃, 10.3; MgSO₄, 1.5; CaCl₂, 3.0; KH₂PO₄, 3.0; Suc, 88.0; and Mes, 2.0; and the following micronutrients (μm): H₃BO₃, 100; CoCl₂·6H₂O, 0.1; CuSO₄·5H₂O, 0.1; Fe-EDTA, 100; MnSO₄·H₂O, 110; KI, 5.0; Na₂MoO₄·2H₂O, 1.0; ZnSO₄, 50; and 2,4-D, 0.90; inositol, 555. Five days after subculturing, the cells were transferred to new growth medium from which EDTA and PO₄ had been removed (to prevent Al chelation and precipitation), centrifuged at 100g, and washed twice with fresh medium. The cells were then placed in new medium supplemented with 25 mm dimethylglutamic acid, pH 4.5, and 50 μm Indo-1 and incubated for 1 h. The loaded cells were washed with fresh growth medium, settled onto a no. 1 coverslip that formed the bottom of a perfusion chamber (1 mL total volume, flow rate 2 mL min⁻¹). Treatments (Al, La³⁺, Mn²⁺, verapamil, nifedipine, and EGTA) were applied by perfusing the chamber with the appropriate solution using a peristaltic pump. The perfusion chamber equilibration time was 30 to 60 s. All chemicals and reagents were supplied by Sigma unless stated otherwise. Al does not interfere with the Indo-1 Ca²⁺ fluorescent signal (D.L. Jones and S. Gilroy, unpublished data).

Measurement of [Ca²⁺]_cyt

For fluorescence ratio-imaging of [Ca²⁺]_cyt, the acid-loaded cells were placed on the stage of an Axiovert inverted microscope attached to a LSM410 laser scanning confocal microscope (Zeiss) and imaged using a ×40, 0.75 numerical aperture, dry objective (Zeiss). Fluorescence from the dye was excited with the 364-nm line of a UV laser (Enterprise, Coherent, Auburn, CA) using an 80/20 beam splitter. Emitted light was simultaneously detected at 400 to 435 nm and 480 ± 20 nm using a 460-nm dichroic mirror and the appropriate Zeiss interference filters on each of the two photomultiplier detectors. Each frame represents a single 8-s scan of the laser. Photobleaching represented <5% per channel per scan for each ratio image. Transmission images were also taken for each ratio image using the transmission detector of the confocal microscope and illumination by the 633-nm He/Ne laser of the confocal attenuated to 10% with neutral density filters. Pseudocolor ratio images of the [Ca²⁺]_cyt distribution were calculated essentially as described by Gilroy et al. (1991). Image processing was carried out on a PowerMac 8100 computer (Apple) using IP Labs Spectrum image-analysis software (Signal Analysis, Vienna, VA). Autofluorescence and dark current represented <5% of the Indo-1 fluorescence signal at each detector.

For Mn²⁺ quench experiments, fluorescence emission was also monitored at the Ca²⁺-insensitive wavelength of Indo-1 (460 ± 20 nm). Lucifer Yellow fluorescence was monitored using 488-nm excitation, 488-nm dichroic mirror, and 515- to 540-nm emission. Each image represents a single 8-s scan of the laser.

Ratio images were calibrated using in vitro Ca²⁺ calibration standards from Molecular Probes (Eugene, OR) as described by Gilroy (1996). Confirmation of the applicability of this in vitro calibration to in vivo data was made by performing an in vivo calibration of the dye. Indo-1-loaded cells were perfused with calibration solutions containing 5 mm EGTA and known free [Ca²⁺] and 25 μm Ca²⁺-ionophore Br-A23187 for 15 min. Ratio images of these cells showed the expected changes in ratio values to within 10% of those predicted from the in vitro calibration.

Ca²⁺ levels in the media were determined using a Ca-selective electrode (Orion, Boston, MA), which showed a linear response to [Ca²⁺] to 100 nm. The electrode was calibrated using Ca²⁺ calibration standards from two sources (World Precision Instruments, New Haven, CT, and Molecular Probes).

Video Imaging and Determination of Growth Rate

For calculation of growth rates, a 1-mL sample of cells was placed in the perfusion chamber on the stage of the Axiovert inverted microscope attached to a LSM410 laser scanning confocal microscope. Cells were imaged using the transmission detector of the confocal microscope and illumination by the 633-nm He/Ne laser of the confocal attenuated to 10% with neutral density filters. Cell size was used as an indication of expansion growth and was monitored as cross-sectional area of each cell using IP Labs image-analysis software. Typical BY-2 cells were found to be 20 to 30 μm in width and 20 to 50 μm in length. Cell areas ranged from 400 to 1000 μm². Repetitive measurements of the same image of a cell revealed a se of ± 3% (n = 17) in calculation of the area using this image-analysis software. Measuring the area of the same cell but varying the focal plane of the image to the point where data would be rejected because the cell was obviously out of focus revealed that the largest change in area introduced by a focal plane effect was ± 7% (se, n = 12).The accuracy of these cell area determinations allowed us to detect significant changes in growth of individual cells during a 10-h period. Critically, measurements made using this approach allowed us to monitor the effects of treatments with Al and Ca²⁺-channel antagonists on expansion growth under conditions identical to those used for imaging [Ca²⁺]_cyt.

Microinjection

BY-2 cells were embedded in growth medium supplemented with 0.5% (w/v) Phytagel and pressure microinjected with Indo-1, Indo-1 linked to 10-kD dextran or Lucifer Yellow (Molecular Probes) as described by Gilroy and Jones (1992). Micropipettes (10–20 MΩ resistance) were pulled from filament electrode glass (World Precision Instruments) using a PC-84 pipette puller (Sutter Instruments, Novato, CA). The micropipettes were loaded with 1 mm Indo-1, Indo-1 conjugated to dextran (10,000 M_r), or Lucifer Yellow. Fluorescent dye was then pressure injected using a PV830 pneumatic picopump (World Precision Instruments) using a series of 0.14-MPa pressure pulses. Injected cells were allowed to recover from the microinjection for 20 min prior to ratio imaging. Cells that failed to maintain a turgid appearance or that showed disruption of cytoplasmic structure (typically a rapid condensation of cytoplasmic contents) were excluded from further analysis. Intracellular dye concentration was calculated as described by Gilroy et al. (1991).

Chemical Equilibria Predictions

Theoretical chemical equilibria predictions of Al activities were made using the computer simulation program Geochem-PC, version 2.0 (Parker et al., 1995).

RESULTS

Indo-1 Loading Does Not Affect the Growth of BY-2 Cells

The phytotoxic action of Al has been proposed to involve disruption of normal cellular Ca²⁺ homeostasis through blockage of Ca²⁺ channels at the plasma membrane. We therefore decided to test this possible mode of action by assessing the effects of Al on [Ca²⁺]_cyt and growth in plant cells. Tobacco BY-2 suspension-cultured cells were chosen as our experimental system since they are a homogeneous cell preparation that is highly amenable to growth analysis and fluorescence imaging and show a rapid inhibition of growth in response to Al (see below). These cells were loaded with the fluorescent [Ca²⁺] indicator Indo-1 by incubation at pH 4.5 (acid loading; Bush and Jones, 1987).

We first ensured that the acid loading of Indo-1 into these cells did not affect the growth kinetics or the effect of Al on these cells. Figure 1 shows the growth kinetics of BY-2 cells monitored as an increase in cell size. Indo-1 loading, at up to 50 μm Indo-1 in the acid-loading medium, did not affect growth of these cells. Indo-1-loaded and control cells were morphologically indistinguishable for as long as we observed their growth (up to 24 h; data not shown). Acid loading under these conditions led to an internal Indo-1 concentration of approximately 10 μm. Figure 2 shows that addition of 100 to 200 μm Al (11.6 and 23.4 μm Al³⁺ activities, respectively) led to a rapid inhibition of growth that was identical in Indo-1-loaded cells and unloaded controls. The relatively high requirement of 100 to 200 μm Al for a highly reproducible inhibition of growth may reflect a degree of Al tolerance of BY-2 cells or Al binding or chelation under the culture conditions used. However, a similar requirement for 100 to 200 μm Al for phytotoxicity has been reported previously for suspension-cultured tobacco cells (Yamamoto et al., 1994, 1996)

Figure 1.

Growth kinetics of BY-2 cells. Cellular expansion was determined after placing 1 mL of a 5-d-old cell-suspension culture into a perfusion chamber mounted on the microscope stage. Time- lapse video images of cell expansion were then taken as described in Methods. The growth rate of individual cells was calculated as the increase in cell area measured from individual frames of the video. Identical experiments were performed with cells acid loaded with Indo-1 (▪, 50 μm, 1 h) and with unloaded control cells (•). Results represent means ± se, n ≥ 35.

Figure 2.

Growth kinetics of BY-2 cells treated with Al. Cells were acid loaded with Indo-1 and treated with 0, 100, or 200 μm AlCl₃. For comparison, non-Indo-1-loaded cells were also treated with 200 μm Al (control). Cellular expansion was determined after placing 1 mL of a 5-d-old cell-suspension culture into a perfusion chamber mounted on the microscope stage. Time-lapse videos of cell expansion were then taken as described in Methods. The growth rate of individual cells was calculated as the increase in cell area measured from individual frames of the video. Results represent means ± se, n > 50.

Indo-1 Reports [Ca²⁺]_cyt in BY-2 Cells

Having established that Indo-1 loading did not disrupt the growth response of BY-2 cells to Al, we next ensured that this indicator was reliably reporting [Ca²⁺]_cyt. Dyes such as Indo-1 may be taken up by organelles in some plant cells (Read et al., 1992). Once localized in an organelle, the dye cannot be used to monitor [Ca²⁺]_cyt. We therefore ensured that in our experiments acid loading of Indo-1 led to cytosolic localization of the indicator and consequently was a valid monitor of [Ca²⁺]_cyt. Several lines of evidence suggested that this was the case. The ratio images of [Ca²⁺] from BY-2 cells acid loaded with Indo-1 were similar to those from cells that had been microinjected with Indo-1 (data not shown) or with Indo-1 linked to a 10-kD dextran (compare A and C in Fig. 3). In both cases the Indo-1 signal was localized to the cytoplasm. Vacuoles excluded the indicator and appear as dark regions in the ratio images. However, dye-loaded cytoplasmic strands were visible crossing these vacuolar regions. Dextran-conjugated dyes are not thought to cross organelle membranes and, thus, once introduced into the cytoplasm, should remain there and reliably report [Ca²⁺]_cyt (Read et al., 1992). Thus, as dextran-conjugated and acid-loaded indicator showed similar distributions, it is unlikely that acid-loaded Indo-1 was reporting vacuolar or cell wall [Ca²⁺]. Also, upon plasmolysis of the acid-loaded cells, the Indo-1 signal remained with the plasmolyzed cytoplasm and was not evident in the wall (Fig. 3, D and E). Addition of the Ca²⁺ ionophore Br-A23187 also led to a rapid increase in [Ca²⁺] monitored by the Indo-1, suggesting that the dye was localized in a compartment showing a low, stable [Ca²⁺] (Fig. 3, A and B), consistent with a cytosolic location. We cannot discount that some of the free Indo-1 or the dextran-bound form of the indicator may be sequestered by organelles. However, the close parallels between the ratio images obtained with acid-loaded Indo-1 and its microinjected, dextran-conjugated form suggest that both are measuring [Ca²⁺]_cyt.

Figure 3.

Ca²⁺ ratio imaging of BY-2 cells after treatment with ionophore, mannitol, or Al. BY-2 cell before (A) and 10 min after (B) treatment with 20 μm Ca²⁺ ionophore Br-A23187. C, BY-2 cell microinjected with Indo-1 linked to a 10-kD dextran. D, BY-2 cell before plasmolysis in 500 mm mannitol solution. E, BY-2 cell after plasmolysis in 500 mm mannitol solution. F, Time course (min) of the effect of 200 μm Al on [Ca²⁺]_cyt in BY-2 cells. G, Time course (min) of the effect of 50 μm Al on [Ca²⁺]_cyt in BY-2 cells. Cells were either acid loaded with Indo-1 (A, B, and D–G) or microinjected with Indo-1-dextran (C) and maintained in a perfusion chamber on the microscope stage. Ca²⁺ distribution was then determined by confocal ratio imaging. Treatments were added by perfusing the cells with medium supplemented with the appropriate addition. The perfusion chamber completely equilibrated in 30 to 60 s. Cytoplasmic Ca²⁺ levels have been pseudocolor coded according to the inset scale. A to G, Corresponding transmission detector images of the cells shown in A′ to G′. Results are typical of n ≥ 10 individual experiments. cs, Cytoplasmic strand; n, nucleus; and v, vacuole. Scale bar = 10 μm.

Al Induces a Decrease in [Ca²⁺]_cyt and Inhibits Growth

Having established that Indo-1 was a viable reporter of [Ca²⁺]_cyt in the BY-2 cell, we next monitored [Ca²⁺]_cyt using confocal ratio imaging as the cells were subjected to Al stress. Figures 3 and 4 show that perfusion of cells with 200 μm Al led to a rapid reduction in [Ca²⁺]_cyt from resting levels of 256 ± 43 to 64 ± 51 nm (n = 37). This result is consistent with the proposed phytotoxic mode of action of Al through blockage of Ca²⁺ channels required to maintain normal cellular [Ca²⁺]_cyt. This decrease in [Ca²⁺]_cyt was not reversed by perfusing the cells with fresh, Al-free medium for up to 70 min (Fig. 4). These Al-treated cells were arrested in growth (Fig. 2). It was possible that this Al-induced decrease in [Ca²⁺]_cyt was an artifact of an Al-induced compartmentalization of the acid-loaded Indo-1 into a cellular site of low [Ca²⁺]. This possibility was tested by monitoring the effect of Al on [Ca²⁺]_cyt using cells microinjected with Indo-1 dextran. This dextran-conjugated form of the indicator is much less likely to undergo compartmentalization than the free, acid-loaded indicator. Dextran-conjugated Indo-1 revealed an equivalent decrease in [Ca²⁺]_cyt in cells treated with 200 μm Al, as did the acid-loaded indicator (data not shown). Similarly, 200 μm Al had no effect on the fluorescence from cells microinjected with the Ca²⁺-insensitive dye Lucifer Yellow-CH (data not shown). These results suggest that the effect of Al was not due to some nonspecific toxic effect on dye fluorescence but was specific to dyes reporting [Ca²⁺] localized to the cytosol.

Figure 4.

Effect of Al on [Ca²⁺]_cyt in BY-2 cells. Cells were acid loaded with Indo-1 and maintained in a perfusion chamber on the microscope stage. Cells were perfused with 0, 50, 100, and 200 μm AlCl₃ as indicated and Ca²⁺ distribution was determined by confocal ratio imaging. After 20 min of Al treatment, the cells were perfused with Al-free medium and the effect on [Ca²⁺]_cyt was monitored. Ca²⁺ level over the entire cell was calculated at each time using IP Labs image-analysis software. Results are means ± se, n ≥ 30.

La and EGTA Cause a Decrease in [Ca²⁺ ]_cyt and Inhibit Growth

To test whether the [Ca²⁺] decrease caused by Al was potentially an effect of blocking Ca²⁺-permeable channels, [Ca²⁺]_cyt was monitored in cells treated with other agents proposed to block plasma membrane Ca²⁺ channels of plant cells. Figures 5A and 6A indicate that under our growth conditions, 100 μm nifedipine and verapamil, Ca²⁺-channel blockers, had little effect on [Ca²⁺]_cyt and also did not inhibit BY-2 cell growth. However, the Ca²⁺-channel blocker La³⁺ at 1 mm induced a rapid, steady-state decrease in [Ca²⁺]_cyt and also inhibited growth (Figs. 5B and 6A). Chelation of external Ca²⁺ with 5 mm of the Ca²⁺ buffer EGTA also led to a rapid decline in [Ca²⁺]_cyt and in the growth rate (Figs. 5B and 6A). These results suggest that a supply of external Ca²⁺ is required for BY-2 cells to sustain normal, resting [Ca²⁺]_cyt and growth. Both the La³⁺- and EGTA-induced Ca²⁺ decrease were reversed after cells were washed free of these inhibitors by perfusion with fresh growth medium (Fig. 5B). Growth inhibition by La³⁺ and EGTA was also found to be reversible. Thus, when BY-2 cells were pretreated for 30 min with 1 mm La³⁺ or 5 mm EGTA (at which time the decrease in [Ca²⁺]_cyt induced by these compounds was complete, Fig. 5B) and then perfused with inhibitor-free medium, growth recovered (Fig. 6B). In contrast, growth inhibition by a 30-min pulse of 200 μm Al was irreversible (Fig. 6B), suggesting that Al may have toxic effects in addition to causing a reduction in [Ca²⁺]_cyt, and that these other effects are not shared by La³⁺ and EGTA.

Figure 5.

Effect of La³⁺, verapamil, nifedipine, and EGTA on [Ca²⁺]_cyt. A, Mean [Ca²⁺]_cyt values in BY-2 cells treated with 100 μm verapamil or nifedipine. B, Mean [Ca²⁺]_cyt values in BY-2 cells treated with 1 mm La³⁺ or 5 mm EGTA. Cells were acid loaded with Indo-1 and maintained in a perfusion chamber on the microscope stage. Ca²⁺ distribution was then determined by confocal ratio imaging and the average [Ca²⁺]_cyt was calculated from the ratio images. LaCl₃ (1 mm), verapamil (100 μm), nifedipine (100 μm), or EGTA (5 mm) were perfused into the chamber and the effect on [Ca²⁺]_cyt was monitored. At the indicated times, the cells were perfused with inhibitor-free medium. The Ca²⁺ level over the entire cell was calculated at each time using image-analysis software. Results are means ± se, n ≥ 20.

Figure 6.

Effect of Al, La³⁺, verapamil, and EGTA on growth of BY-2 cells. A, Growth rate in cells treated with 100 μm verapamil (○), 1 mm La³⁺ (⋄), 200 μm Al (□), or 5 mm EGTA (▪). B, Recovery in growth rate of cells pretreated with 1 mm La³⁺ (⋄), 200 μm Al (□), or 5 mm EGTA (▪) for 30 min and then perfused with inhibitor-free medium. Results are means ± se, n ≥ 30.

Al Reduces the Rate of Mn²⁺ Quenching

Mn quenching of fluorescence has been used as a probe for Ca²⁺ channel activity in plant cells loaded with Ca²⁺-indicating dyes such as Indo-1 (Malhó et al., 1995; McAinsh et al., 1995; Wymer et al., 1997). Mn is thought to enter cells through Ca²⁺-permeable channels, and once in the cytosol, it binds to and quenches the fluorescent indicator Indo-1. We therefore used this Mn²⁺-quench approach to determine whether Al and La³⁺ were blocking Mn²⁺-permeable channels. Such a blockage would reduce the rate of Mn²⁺ entry into the cytoplasm and therefore reduce the rate of quenching of Indo-1. Figure 7 shows the quenching kinetics for Indo-1-loaded BY-2 cells treated with 100 μm Mn²⁺ and with 100 μm Al, 1 mm La³⁺, or 100 μm verapamil. Al and La³⁺ reduced the quenching effect of Mn²⁺ by 50%, measured 5 min after Mn²⁺ addition, whereas verapamil had no detectable effect on the kinetics of dye quenching. As expected, addition of 20 μm Mn²⁺-permeant ionophore Br-A23187 almost entirely quenched the Indo-1 signal. Although Mn²⁺ quenching is an indirect approach to monitoring Ca²⁺ channel activity, these results are consistent with Al blockage of Ca²⁺-permeable channels in these cells.

Figure 7.

Effect of Al, La³⁺, and verapamil on Mn²⁺ quenching of Indo-1 fluorescence. Cells were acid loaded with Indo-1 and maintained in a perfusion chamber on the microscope stage. Indo-1 fluorescence was monitored at its Ca²⁺-insensitive wavelength (460 nm). MnCl₂ (100 μm) supplemented with nothing (○, control), 100 μm AlCl₃ (▪), 1 mm LaCl₃ (□), or 100 μm verapamil (•) was then perfused into the chamber. When Mn²⁺ entered the cell it quenched the Indo-1 fluorescence, resulting in a reduction of signal. Br-A23187, Fluorescence signal monitored 5 min after adding 20 μm divalent cationophore to the perfusion chamber. Results are means ± se, n ≥ 8.

DISCUSSION

It has been postulated by numerous authors that Al may interfere with cellular Ca²⁺ homeostasis, leading to a breakdown of the Ca²⁺-dependent signal transduction cascades that are necessary for both cell division and cell elongation (Haug, 1984; Taylor, 1990; Rengel, 1992; Delhaize and Ryan, 1995; Kochian, 1995). We have observed that cytotoxic levels of Al lead to a rapid (within minutes) reduction in [Ca²⁺]_cyt in BY-2 cells and that this change correlates with the inhibition of growth in these cells. These results suggest that Al may inhibit the Ca²⁺influx across the plasma membrane required to maintain growth. In contrast, Lindberg and Strid (1997) reported an immediate, transient (2-min duration), oscillating increase in [Ca²⁺]_cyt in wheat root protoplasts exposed to 80 μm Al. However, this change was relatively small, from approximately 160 to 225 nm, identical in protoplasts isolated from Al-resistant and Al-sensitive cultivars, and occurred in only 60% of the protoplasts studied. Thus, the relationship of this transient increase in [Ca²⁺]_cyt to the phytotoxic action of Al remains to be determined.

Ryan et al. (1997) showed that the phytotoxic effects of Al are unlikely to result from the displacement of Ca²⁺ from critical sites in the apoplast. However, there are many reports of a requirement for high extracellular Ca²⁺ to sustain plant cell expansion and division (Hepler and Wayne, 1985). We have confirmed that chelating extracellular Ca²⁺ with 5 mm EGTA (leading to a free Ca²⁺ of <300 nm in the growth medium) inhibits growth in BY-2 cells. The role of this extracellular Ca²⁺ requirement is unknown, but stabilization of wall structure, membrane integrity, and as a source for intracellular regulatory events are all possibilities. The reduction in [Ca²⁺]_cyt induced by the Ca²⁺ channel antagonist La³⁺ suggests that transplasma membrane fluxes represent an important role for this extracellular pool.

Although indirect, the inhibition of Mn²⁺ quenching by Al provides further evidence that one mode of action of Al in these cells is to block Ca²⁺-permeable channels. At toxic levels, Al and La³⁺ inhibited Mn²⁺ quenching of intracellular Indo-1, whereas 100 μm nifedipine and verapamil had no effect on quenching or cell growth. Despite successful application of the Mn²⁺ quench technique to plant cells (Malh 243 et al., 1995; McAinsh et al., 1995), the quench data alone provide very tentative evidence for Ca²⁺ channel activity. However, in conjunction with the ratio-imaging data showing an Al-induced reduction in [Ca²⁺]_cyt, the similarities between the toxicity of Al and that of the Ca²⁺-channel antagonist La³⁺, and the extensive literature implicating Al blockage of Ca²⁺ channels as a potential mode of Al toxicity, the Mn²⁺-quench data strongly suggest that Al is blocking Ca²⁺ channels in these BY-2 cells. We await patch-clamping data from isolated BY-2 cell protoplasts to confirm that this is the case.

Recently, it was shown that at the toxic concentrations normally found in soils (10–100 μm), Al³⁺ is capable of blocking voltage-gated plasma membrane Ca²⁺ channels and disrupting inositol 1,4,5-trisphosphate-mediated signaling events in wheat roots (Jones and Kochian, 1995; Huang et al., 1996). Inositol 1,4,5-trisphosphate has been implicated in both cytoskeletal regulation and the progression through cell division (Berridge, 1993), as well as in the control of plant cell tip growth (Franklintong et al., 1996). The cytoskeleton shows well-characterized Ca²⁺-dependent regulation (Lonergan, 1985; Billger et al., 1993; Bokros et al., 1996), providing one mechanism for regulation of growth and division by Ca²⁺. It is interesting that Al has been reported to also affect cytoskeletal dynamics, causing both the actin and microtubule network to become rigidified (MacDonald et al., 1987; Grabski and Schindler, 1995).

Much work remains to be done to determine the sites of phytotoxicity of Al in plants. The data presented herein suggest that Al interaction with Ca²⁺ channels and disruption of cellular Ca²⁺ homeostasis may well represent one mode of phytotoxic action. Changes in [Ca²⁺]_cyt are known to be associated with an enormous range of signal transduction and cellular regulation processes in plant cells (Bush, 1995). Disruption of these events by Al blockage of Ca²⁺ fluxes should then inevitably lead to catastrophic disruption of the regulation and maintenance of cell activities. However, Al toxicity is likely to be more complex than simply blocking Ca²⁺ fluxes by binding to the extracellular face of the Ca²⁺ channel. An intracellular site of Al action is suggested by the irreversibility of the phytotoxic effect of Al on BY-2 cells (Fig. 6B), compared with the reversible nature of growth inhibition by the Ca²⁺ channel antagonist La³⁺. Thus, once within the cell Al may affect a range of activities, such as the complexing of ligands required by Ca²⁺-dependent enzymes (e.g. ATP for Ca²⁺-ATPase), the prevention of Ca²⁺-mediated vesicle fusion, and the inhibition of Ca²⁺-mediated cytoskeletal dynamics (Haug, 1984; Taylor, 1990; Rengel, 1992; Delhaize and Ryan, 1995; Kochian, 1995). Such intracellular sites of Al action should be fruitful areas of future research.

Abbreviations:

Al

Alⁿ⁺

[Ca²⁺]_cyt

cytoplasmic free Ca concentration