The fou2 mutation in the major vacuolar cation channel TPC1 confers tolerance to inhibitory luminal calcium


*(fax +49 931 888 6158; e-mail


The SV channel encoded by the TPC1 gene represents a Ca2+- and voltage-dependent vacuolar cation channel. Point mutation D454N within TPC1, named fou2 for fatty acid oxygenation upregulated 2, results in increased synthesis of the stress hormone jasmonate. As wounding causes Ca2+ signals and cytosolic Ca2+ is required for SV channel function, we here studied the Ca2+-dependent properties of this major vacuolar cation channel with Arabidopsis thaliana mesophyll vacuoles. In patch clamp measurements, wild-type and fou2 SV channels did not exhibit differences in cytosolic Ca2+ sensitivity and Ca2+ impermeability. K+ fluxes through wild-type TPC1 were reduced or even completely faded away when vacuolar Ca2+ reached the 0.1-mm level. The fou2 protein under these conditions, however, remained active. Thus, D454N seems to be part of a luminal Ca2+ recognition site. Thereby the SV channel mutant gains tolerance towards elevated luminal Ca2+. A three-fold higher vacuolar Ca/K ratio in the fou2 mutant relative to wild-type plants seems to indicate that fou2 can accumulate higher levels of vacuolar Ca2+ before SV channel activity vanishes and K+ homeostasis is impaired. In response to wounding fou2 plants might thus elicit strong vacuole-derived cytosolic Ca2+ signals resulting in overproduction of jasmonate.


Following the application of the patch clamp technique to higher plant plasma membranes and vacuoles, ion transport properties of both membranes could be studied individually (Schroeder et al., 1984; Hedrich et al., 1986; Hedrich and Schroeder, 1989 for review). Initial studies on isolated barley vacuoles identified a voltage-dependent cation channel and a proton pump as major conductances of the vacuolar membrane (Hedrich et al., 1986). Using the same electrophysiological approach both vacuolar transporters were identified in almost all plants and cell types (Hedrich et al., 1988; Pottosin and Schönknecht, 2007 for review). The vacuolar cation channel, characterized by its slow activation in response to voltage changes by Hedrich and Neher (1987), was named SV for Slow Vacuolar channel. Furthermore, this study recognized the Ca2+-dependent activation of the SV channel. In line with a Ca2+-modulated channel, voltage activation of SV channel is facilitated in the presence of elevated cytosolic Ca2+ concentrations. A rise of the cytosolic Ca2+ level has a dual effect on SV channel activity, an increase in the maximal number of open SV channels at high positive potentials and a shift of the voltage dependence to less positive potentials (Schulz-Lessdorf and Hedrich, 1995; Pottosin et al., 1997; Pottosin and Schönknecht, 2007 for review).

The SV channel has a similar permeability for K+ and Na+ (Coyaud et al., 1987; Amodeo et al., 1994; Pottosin et al., 2001; Ranf et al., 2008). When these cations are present on the cytosolic side of the vacuolar membrane at about 10 mm and higher concentrations, K+ and Na+ ions enter the vacuole (Ivashikina and Hedrich, 2005). While this situation reflects a physiological condition, Ca2+ transport into the lumen of this organelle was shown when the vacuole was challenged with Ca2+ concentrations exceeding 50 000 times that occurring in vivo in the cytosol (Ward and Schroeder, 1994). When present in the vacuolar lumen, the SV channel of Arabidopsis suspension-cultured cells mediates the uptake and release of the major monovalent cation K+. This process is driven by the voltage and concentration gradient (Ivashikina and Hedrich, 2005). In contrast Na+ and Ca2+ are not released from the vacuole into the cytosol via this channel. At millimolar concentrations Na+ and Ca2+ even block the release of K+.

In search for plant Ca2+ channels, a single copy gene with partial homology to a voltage-dependent Ca2+ channel, called two pore channel (TPC) 1, was identified by database mining (Ishibashi et al., 2000). The Ca2+ channel signature conserved among species (Zagotta, 2006), however, is missing in TPC1 (Furuichi et al., 2001). The channel protein, fused to GFP, was localized in the central vacuole of Arabidopsis (Peiter et al., 2005). Elegant patch clamp studies on a TPC1 loss-of-function mutant, tpc1-2, enabled the authors to unequivocally demonstrate that TPC1 encodes the SV channel. In line with previous findings that showed that the SV channel probably does not catalyze vacuolar Ca2+ release into the cytosol (Pottosin et al., 1997; Ivashikina and Hedrich, 2005), recently Ranf et al. (2008) could demonstrate that Ca2+ signalling of tpc1-2 is indistinguishable from wild type. In contrast to tpc1-2, a point mutation in TPC1 named fou2 showed a pronounced phenotype (Bonaventure et al., 2007a,b). fou2, for fatty acid oxygenation upregulated 2, was identified in a screen for mutants in wounding-induced jasmonate production. fou2 channel activation appeared to be shifted in its voltage dependence to more negative membrane potentials, which resulted in faster activation kinetics compared with the wild-type SV channel. The role of this gain-of-function SV channel activity in wounding/jasmonate signalling is obscure. Also, it is not known how an amino acid residue in the SV channel that is predicted to face the vacuolar lumen is able to shift the voltage gate of the SV channel. Here we show comparative patch clamp studies on Arabidopsis mesophyll vacuoles isolated from wild-type and fou2 mutant plants. While both the responses to changes in the cytosolic Ca2+ level and the Ca2+ impermeability of the fou2 SV channels remained wild-type-like, the SV channel mutation altered sensitivity to luminal Ca2+ levels.


fou2 represents a hyperactive version of the SV channel

Previous studies have shown that the SV channel from barley mesophyll vacuoles under symmetrical 100-mm potassium functions as a Ca2+-dependent outward-rectifying cation channel (Hedrich et al., 1986). Similar results were obtained when the whole-vacuole patch clamp configuration was established on mesophyll vacuoles of Arabidopsis thaliana. Upon voltage stimulation (+70 mV pulses) of wild-type SV channels, outward currents appeared about 5–6 min directly after gaining access to the vacuole lumen during equilibration of the vacuole sap with patch pipette solution, which contained 0.1 mm Ca2+. Steady-state activation was reached after about 12 min in the whole-vacuole mode (Figure 1). Under the same condition with fou2 vacuoles, depolarizing voltage steps elicited outward SV channel currents already in the range of 0.5–2 min (Figure 1). These results indicate that a vacuolar SV channel inhibitor present in the lumen seems to be ‘washed out’ upon equilibration with the patch pipette solution (Maathuis and Prins, 1991). Two possibilities might explain why a potential inhibitor has only a marginal effect in the fou2 mutant: (i) the TPC1 inhibitor concentration is lower in fou2 vacuoles; and/or (ii) the sensitivity of the SV channels towards the inhibitor is reduced in fou2. The latter situation could be gained by a point mutation on the luminal side of the channel protein. To distinguish between the two possibilities, we examined the quantities of various elements in mesophyll vacuoles by energy dispersive X-ray (EDX) analysis (Figure 2). In wild-type and mutant plants the vacuolar Mg, P, S and Cl content was similar while K/Ca ratios appeared quite different. fou2 mesophyll vacuoles contained significantly less K but more Ca than wild-type vacuoles. As wild-type Arabidopsis plants and the fou2 mutant express TPC1 RNA to a similar extent (Bonaventure et al., 2007a), the reduced K level in fou2 might be related to the properties of the mutant form of the SV channel.

Figure 1.

 Transient increase in the macroscopic steady-state currents of wild-type and fou2 SV channels after establishing the whole vacuolar configuration on mesophyll vacuoles of Arabidopsis thaliana.
The steady-state currents Iss were normalized to the maximal value and plotted against the time in the whole vacuolar mode. Closed and open symbols represent wild type and fou2, respectively. Iss were recorded at the following clamped voltages: +70 mV (open downward triangles, open circles and closed symbols) and +50 mV (open upward triangles, open squares). For analysis vacuoles of similar sizes reflected by the measured membrane capacitance Cm (20–30 pF) were chosen. The experiments were performed under symmetrical KCl (100 mm) and pH 7.5 conditions. The luminal and cytosolic Ca2+ concentrations were 0.1 and 1 mm, respectively.

Figure 2.

 Semi-quantitative EDX analysis of ion concentrations in mesophyll vacuoles.
Data were obtained from leaves of four wild-type as well as four fou2 plants. Columns (open = wild type, closed = fou2) show mean values with standard deviations of at least 15 recorded spectra. The scale on the left gives the atomic% of recorded X-ray signals. Note that differences in the relative K+ and Ca2+ concentration were statistically significant (t-test, P < 0.001).

With symmetrical K+ concentrations on both sides of the vacuolar membrane, fou2 SV channel activation was approximately 30 mV more negative than wild-type channels in whole vacuolar studies (Figure 3a, cf. Bonaventure et al., 2007a). Thus, fou2 SV channels conduct inward K+ currents (in other words K+ release from the vacuole into the cytosol) already at low negative voltages. When, in the presence of 1 mm cytoplasmic Ca2+, the potassium gradient is directed out of the vacuolar lumen (high [K+]lumen/low [K+]cytosol), wild-type channels conduct both inward and outward potassium currents (Ivashikina and Hedrich, 2005). Under these asymmetrical conditions we found fou2 to activate at about a 40 mV less depolarized voltage than wild-type SV channels (Figure 3b). As a consequence, in the presence of 30 mm K+ in the cytosol and 150 mm in the vacuole, fou2 SV channels at −5 mV mediate about five times higher inward currents when compared with the wild-type SV channels (Figure 3b). When cytosolic K+ was replaced by 15 mm Ca2+, no current was elicited by voltage stimulation in wild-type and fou2 mesophyll vacuoles (Figure 3c). This result shows that, in the presence of Ca2+ as the sole cytosolic cation, wild-type and fou2 SV channels of A. thaliana mesophyll cell vacuoles neither conduct K+ into the cytosol nor Ca2+ into the vacuolar lumen. Based on these observations, it is unlikely that elevated vacuolar Ca2+ levels in non-stressed fou2 mutants (Figure 2) as well as the predicted wounding-induced rise in cytosolic Ca2+ concentration (Chico et al., 2002; Fisahn et al., 2004; Dombrowski and Bergey, 2007), are related to a change in Ca2+ permeability of the SV channel. This finding, however, does not exclude the possibility that the fou2 mutation had additionally gained altered Ca2+-dependent properties of the SV channel.

Figure 3.

 Dependency of the voltage-dependent activation of fou2 and wild-type SV channels on the K+ and Ca2+ gradient.
Macroscopic currents were recorded in the whole-vacuolar configuration either with 150 mm K+ and 1 mm Ca2+ (a), with 30 mm K+ and 1 mm Ca2+ (b), or with 0 mm K+ and 15 mm Ca2+ (c) in the bath medium (cytosol) at symmetrical pH 7.5. The luminal solution (pipette) contained in (a–c) 150 mm K+ and was nominal Ca2+-free. Representative traces of the current densities (I/Cm) evoked upon 15-mV steps in the voltage range from −50 mV to +70 mV are shown in the left (wild type) and middle lane (fou2). Corresponding steady-state current densities were plotted against the voltages and are shown in the right lane. Please note the different scaling of the current densities for wild type and fou2. Closed symbols = wild type; open symbols = fou2. Arrows indicate the respective voltage thresholds of channel activation. Data points represent the mean, and the error bars give the standard error. The respective number of experiments for wild type was n = 3 in (a–c), for fou2 n = 9 in (a), n = 7 in (b) and n = 3 in (c).

fou2 is activated by cytosolic Ca2+ and tolerates elevated vacuolar Ca2+ levels

SV channels are characterized by pronounced Ca2+ sensitivity (Hedrich and Neher, 1987; Allen and Sanders, 1996; see for review Pottosin and Schönknecht, 2007). Our studies on fou2, shown in Figure 3, were performed with 1 mm cytosolic Ca2+, a level at which this channel is maximally stimulated. To test whether the fou2 mutation alters the sensitivity of the SV channel towards cytosolic Ca2+ concentrations, fou2 and wild-type SV channels were challenged with 200, 300, 500 and 1000 μm cytosolic Ca2+ (Figure 4). Under these conditions, the voltage dependence of the relative channel open probability (Figure 4b,c) was determined by tail-current experiments (Figure 4a). With decreasing cytosolic Ca2+ concentrations the half-activation voltage V1/2 of fou2 and wild-type channels linearly shifted towards more positive voltages (Figure 4b,c). For instance a voltage difference of ΔV1/2 = 46.7 mV was determined for the activation curves with 0.3 and 0.5 mm cytosolic Ca2+. Hence, fou2 and wild-type SV channels exhibit a similar sensitivity towards regulatory cytosolic Ca2+. Given the fact that the luminal calcium concentration feeds back on the transport capacity and the direction of cation fluxes through the wild-type SV channel (see for review Pottosin and Schönknecht, 2007), we examined at the single channel level whether the vacuolar Ca2+ concentration affects the channel activity of fou2 as well. For this test we exposed excised patches with the vacuolar membrane side facing 0, 100 and 1000 μm Ca2+ under symmetrical K+ and H+ concentrations and measured single channel activities in the voltage range from −50 to −30 mV. In line with the macroscopic current recordings the single channel activity in wild-type and fou2 plants declined with negative-going membrane voltages. At voltages positive from −30 mV the channel activities were too high to resolve single channel events. Following an increase in luminal Ca2+ from 0 to 100 or 1000 μm, the wild-type SV channel activity almost completely disappeared at −50 to −30 mV, a range at which single channel analysis was feasible. However, fou2 responded much less dramatically to an increase in luminal Ca2+ levels (Figure 5). In contrast to wild-type channels, fou2 still maintained a pronounced channel activity in the presence of 100 μm, and even 1000 μm, luminal Ca2+. The fact that the wild-type SV channels exhibited a quasi-steady-state inhibition at both Ca2+ levels (100, 1000 μm) points to an increased tolerance of the SV channel mutant to elevated luminal Ca2+ concentrations. Thus, the fou2 point mutation seems to alter the luminal Ca2+ affinity of the SV channel, underlying the Ca2+-mediated inhibition of the SV channel.

Figure 4.

 Effect of cytosolic Ca2+ on the voltage-dependent activation of fou2 and wild-type SV channels.
(a) Representative tail currents shown for wild type and fou2 (framed and enlarged on the right side) were recorded at −60 mV subsequent to +70 mV as a pre-activating voltage pulse. The holding voltage was −60 mV. Traces at cytosolic Ca2+ of 1 and 0.3 mm are shown.
(b) The relative voltage-dependent open-probabilities [G(V) curves] were determined upon tail currents after SV channels were activated at different pre-pulse voltages in the range of −100 mV to +100 mV. The experiments were performed in the presence of 100 mm KCl at pH 7.5 in the bath and pipette medium. The pipette medium (vacuole lumen) contained 100 μm Ca2+. Open and closed symbols represent wild type and fou2, respectively, at the following Ca2+ concentrations: squares = 1 mm; triangles = 0.5 mm and circles = 0.3 mm.
(c) The half-activation voltages (V1/2) determined from the Boltzmann fit of the G(V) curves were plotted against the respective cytosolic Ca2+ concentration.
Number of experiments in (b, c) were nwild type = 7 and nfou2 = 6 for 1 mm Ca2+, nwild type, fou2 = 4 for 0.5 mm Ca2+, nwild type, fou2 = 3 for 0.3 mm Ca2+ and nwild type, fou2 = 4 for 0.2 mm Ca2+. Error bars show standard error of the mean.

Figure 5.

 Effect of different luminal Ca2+ concentrations on the wild-type and fou2 single channel activity under symmetrical KCl and pH conditions.
(a) Single channel fluctuations were recorded from excised membrane patches with the cytoplasmic side of the vacuolar membrane facing the bath medium. The membrane potential was clamped to −30 mV. C indicates the current baseline where all channels are closed. O1,2,3 give the current levels at which 1, 2 or 3 channels were simultaneously open.
(b) Single-channel open probabilities (Po) were determined at −50, −40 and −30 mV from single channel recordings as illustrated in (a). As the resolution limit for analysis of wild-type single-channel fluctuations was reached at elevated luminal Ca2+, the respective single-channel open probabilities were set to 0.
The experiments in (a, b) were performed in the presence of 100 mm KCl and pH 7.5 on both membrane sides. The cytosolic medium (bath) contained 1 mm CaCl2 while the CaCl2 concentration of the luminal medium (pipette) was altered as indicated. The number of experiments was n = 3–5, and the error bars show standard deviation.

Luminal protons and Ca2+ ions affect differentially the activity and K+ transport capacity of fou2 and wild-type
SV channel

Vacuolar proton pumps can modulate the acidity of the vacuolar lumen (see review Martinoia et al., 2007) and luminal pH changes were shown to alter SV channel properties (Schulz-Lessdorf and Hedrich, 1995; Pottosin et al., 1997). Thereby, Ca2+ apparently competes with H+ for the same binding sites. Removal of vacuolar Ca2+ at neutral pH results in a dramatic negative shift of the voltage-dependent SV channel activation that is opposite to the effect of reduced cytosolic Ca2+ levels (Pottosin et al., 1997, 2004). To test whether protons affect the luminal Ca2+ sensitivity of the SV channel too, and thereby rendering this transporter less sensitive to blocking Ca2+ ions, we studied the vacuolar Ca2+ response of the SV channel to vacuolar acidification. When a 100-fold proton gradient (pH 7.5cyt and 5.5vac) was generated across the vacuolar membrane in the absence of luminal Ca2 the wild-type single-channel activity was much lower (Figure 6) than under symmetrical pH 7.5 conditions (Figure 5). An increase in the luminal Ca2+ concentration from 0 to 100 μm at acidic pH (Figure 6b) affects the wild-type SV channel activity less than at symmetrical pH conditions (Figure 5). A further rise to 1000 μm luminal Ca2+ almost completely blocked the wild-type channel. In comparison with neutral pH (Figure 5b), fou2 single-channel activity appeared to be almost similar in the absence of luminal Ca2+ but showed altered sensitivity towards vacuolar Ca2+ loads at acid vacuolar pH values (Figure 6b). A rise from 0 up to 1000 μm luminal Ca2+ only slightly reduced the channel activity in fou2 (Figure 6b). In the absence of luminal Ca2+, acidification of the vacuole decreased SV channel activity in wild type but only marginally in fou2. This observation points to a mutation-related pK change of a critical luminal site of the TPC1 protein. To further characterize this finding, we decreased the luminal pH stepwise from pH 7.5 to 4.5 in the absence of luminal Ca2+ and presence of 1 mm cytosolic Ca2+ (Figure 7). The single channel conductance was in the order of 35–40 pS at symmetrical pH 7.5 for both wild-type and mutant SV channels (Figure 7) and increased upon pH decrease. At pH 5.5 and 4.5 the unitary conductance reached about 80 pS, a transport capacity that is twice as high as at neutral pH. Similar single-channel open probabilities in fou2 plants, in the absence of luminal Ca2+ at luminal pH 5.5 and 7.5 (cf. Figures 5b and 6b), is indicative of an effect of protons on the unitary conductance of SV channels in fou2 only (Figure 7). However, under same experimental conditions, both the open probability (Figures 5b and 6b) and unitary conductance (Figure 7) of SV channels in wild type responded to pH changes in an opposite manner. While the open probability of wild-type SV channels decreased with increasing proton concentration, the unitary conductance increased. Thus, the mutation in fou2 does not seem to affect protonation of the ‘conductance site’ but alters the pK of the site that is associated with the open probability.

Figure 6.

 Effect of an acidic vacuolar pH on the luminal Ca2+ dependency of the wild-type and fou2 single channel activity.
Single channel recordings were performed at varied luminal Ca2+ concentrations under asymmetrical pH conditions (cytosol/bath medium: pH 7.5; vacuolar lumen/pipette medium: pH 5.5).
(a) Current traces were recorded at different Ca2+ levels from excised membrane patches with the cytoplasmic side of the vacuolar membrane facing the bath medium. The membrane potential was clamped to −30 mV.
(b) Single-channel open probabilities (Po) were calculated at −50, −40 and −0 mV from single channel recordings as illustrated in (a). As the resolution limit for analysis of the wild-type single-channel fluctuations was reached at 1 mm Ca2+, the respective single-channel open probability was set to 0. With the exception of the luminal pH, the solutions were composed as in Figure 5. The number of experiments was n = 3–4, and the error bars show standard deviation.

Figure 7.

 Luminal pH-induced changes in the single channel conductance γ of fou2 and wild-type SV channels.
(a) Single channel amplitudes were determined and plotted against the respective voltages. The single channel conductances γ were derived from the slope of a linear regression describing the data points. Open and closed symbols represent fou2 and wild type, respectively. Inset: representative channel fluctuations for fou2 at pH 7.5 (circles) and 4.5 (squares) recorded at −30 mV are shown. C and O1,2,3 have the same meaning as in Figure 5.
(b) Single channel conductances were determined at different vacuolar pH values as indicated. Open and closed bars represent fou2 and wild type, respectively.
(a, b) Channel fluctuations were recorded from excised membrane patches with the cytoplasmic side of the vacuolar membrane facing the bath medium. The experiments were carried out in the presence of 100 mm KCl on both membrane sides. The cytosolic medium (bath) also contained 1 mm CaCl2 and was adjusted to pH 7.5, while the luminal medium (pipette) was nominal Ca2+-free. The data points represent the mean of three to four individual experiments. Error bars show standard deviation.


fou2 mimics Ca2+ activation of wild-type SV channels

As shown by Bonaventure et al. (2007a), under symmetrical K+ conditions on both vacuolar membrane sides fou2 SV channels activate at more negative potentials and shuttle more K+ into the cytosol than wild-type SV channels (Figure 3a). The presence of inward-directed K+ gradients across the vacuolar membrane caused a negative shift in the voltage threshold for activation of the wild-type channel [Figure 3a,b; see also Ivashikina and Hedrich, 2005]. In fou2 the latter process was comparable but associated with higher inward K+ fluxes compared with wild-type TPC1 (Figure 3a,b). Under related physiological conditions, the fou2 mutant seems to gain its phenotype most probably from the increased capacity for K+ release from the vacuole into the cytosol, which results in a reduced K level in fou2 vacuoles (Figure 2). The impaired potassium homeostasis in the mutant is probably related to the voltage-dependent activation of fou2 SV channels at less negative voltages (Figure 3a, b), as well as to its reduced sensitivity towards increased luminal Ca2+ levels (Figures 5, 6) maintaining more SV channels open in fou2 even at elevated luminal Ca2+ levels. Interestingly, the transcript profile of fou2 plants resembles the K+ starvation transcriptome (Armengaud et al., 2004; Bonaventure et al., 2007b). While the expression of the K+ transporters did not change under K+ deficiency, the transcript level of the vacuolar Ca2+/H+ antiporter CAX3 was up-regulated (Armengaud et al., 2004). The latter could contribute to an increase in the vacuolar Ca2+ content (as observed with fou2, Figure 2) because CAX3 promotes Ca2+ entry into the vacuole. In tomato, wound-induced cytosolic Ca2+ signals have been observed (Moyen et al., 1998), which leads to wound-response gene activation (Dombrowski and Bergey, 2007). Considering the elevated vacuolar Ca2+ content in fou2, a more pronounced signal-induced rise in the cytosolic Ca2+ level might occur upon wounding. As one possible consequence, increased lipoxygenase (LOX) gene or enzyme activity could explain the rise in transcript numbers (e.g. 25-fold for LOX2) and LOX activity observed with the fou2 mutant (Bonaventure et al., 2007a).

The SV channel seems to present a primary voltage-dependent cation channel with its voltage gate modulated by Ca2+, luminal pH and transmembrane K+ gradient. Recent studies on red beet vacuoles showed that changes in luminal pH at different luminal Ca2+ levels did not affect the activation potential of SV channels (Pérez et al., 2008). Accordingly, the observed decrease in the open probability (Po) of wild-type SV channels upon luminal acid loads in the absence of Ca2+ (Figures 5, 6) might be related to a change in voltage sensitivity, rather than to voltage dependence. This pH effect on Po is gone in fou2 (Figures 5, 6), which suggests that aspartate at residue 454 of the SV channel protein might be involved in pH control of the single channel activity upon its voltage sensitivity. In line with a Ca2+-activated cation channel, voltage changes triggered SV channel-mediated K+ currents in the presence of elevated cytosolic Ca2+ only (Ward and Schroeder, 1994; Peiter et al., 2005). Thereby, cytosolic Ca2+ seems to shift the voltage threshold for SV channel activation into the range of physiological vacuolar potentials (cf. Hedrich and Neher, 1987). A similar effect of Ca2+ on the voltage range of activity has been described for maxi K channels in neurons (Marty, 1981). In the absence of cytosolic Ca2+, SV channels seem to activate at extreme positive voltages, a condition that is not tolerated by isolated vacuoles under patch clamp conditions. Elevation of the cytosolic Ca2+ level resulted in a similar shift in the voltage-dependent activation of wild-type and fou2 SV channels (Figure 4b,c), which indicates that the SV channel mutation does not affect the cytosolic Ca2+ dependence of the channel. When challenged with increasing Ca2+ concentrations on the luminal side, SV channel activity ceased in wild type, but not in fou2 (Figures 5, 6). Inhibition of K+ currents through the SV channel occurred at luminal Ca2+ concentrations that were similar in range to activating cytosolic Ca2+ concentrations. In contrast to the action of the divalent cation at the cytosolic side, an increase in the vacuolar Ca2+ level, therefore, albeit increasing the driving force for Ca2+ release, results in SV channel closure.

SV channel/TPC1 in mesophyll cells is neither a Ca2+ channel nor capable of mediating CICR

Replacing physiological 100 mm K+ by 5 mm Ca2+ on the cytoplasmic side of the vacuolar membrane in the presence of 50 mm luminal Ca2+, the Vicia faba guard cell SV channel was shown to mediate cation currents into the vacuole lumen (Ward and Schroeder, 1994). From the K+ and Ca2+ gradients (Nernst potentials) and reversal of tail currents, a Ca2+/K+ permeability of about 3:1 was calculated for the V. faba guard cell SV channel (Ward and Schroeder, 1994). Related experiments with non-physiologically high cytosolic Ca2+ loads (15 mm Ca2+) with isolated vacuoles from cultured Arabidopsis cells showed SV channels that mediated Ca2+ currents into the vacuole (Ivashikina and Hedrich, 2005). However, under similar conditions, neither the SV channels of mesophyll vacuoles from Arabidopsis thaliana wild-type nor from fou2 mutant plants (Bonaventure et al., 2007a,b) conducted Ca2+ influx into the vacuole (Figure 3c). Under physiological Ca2+ concentrations (10–1000 nM cytosol/200–2000 μm vacuole) (Felle, 1988; Bethke and Jones, 1994; Pérez et al., 2008) together with K+ (symmetrical 100 mm; cf. Pérez et al., 2008) SV channel-mediated Ca2+ fluxes across the vacuolar membrane have not been observed. Thus the hypothesized capability of the SV channel to mediate CICR (Ca2+ -Induced Ca2+ Release; Ward and Schroeder, 1994), however, seems rather unlikely, because, at luminal Ca2+ concentrations under which this channel type is active, Ca2+ release into the cytosol through SV channels would not result in a pronounced Ca2+ signal anymore. This prediction is further supported by: (i) recent non-invasive Ca2+ flux measurements (Pérez et al., 2008); and (ii) the fact that, in the SV-channel loss-of-function mutant tpc1-2 and in TPC1-overexpressing plants, Ca2+ signals in response to the entire set of stimuli known to trigger a rise in cytosolic Ca2+ concentration are wild-type-like (Ranf et al., 2008).


The gene expression profile of fou2 when compared with wild type is reminiscent of potassium starvation (Bonaventure et al., 2007b). Together with the above-mentioned vacuolar K+ and Ca2+ content and properties of fou2 and wild-type SV channel, this channel type, at least in mesophyll cells, seems to play a role in the control of vacuolar membrane potential and potassium homeostasis rather than Ca2+ release from this major plant organelle. Our data showed that fou2 plants could accumulate higher levels of vacuolar Ca2+ than wild type before SV channels are blocked and potassium homeostasis is impaired. SV channels in fou2 activate in response to small changes in membrane potential. It is thus very likely that, compared with wild-type, fou2 gains more pronounced jasmonate biosynthesis from wounding-induced membrane polarization and in turn shows elevated vacuolar Ca2+ release – mediated by transporters other than TPC1 – that is driven by steeper gradients of this regulatory cation across the vacuolar membrane.

Experimental procedures

Isolation of mesophyll vacuoles

Arabidopsis thaliana ecotype Columbia (Col-0) and fou2 mutant (Bonaventure et al., 2007a) were grown on soil in a growth chamber at a 8/16 h day/night regime, 22/16°C day/night temperature and a light intensity of 800 lx. Fully developed young rosette leaves of 3–5-week-old plants were used for daily enzymatical protoplast isolation. After removing the lower epidermis, the leaves were incubated in 0.5% (w/v) cellulase Onozuka R-10 (Serva, Heidelberg, Germany), 0.05% (w/v) pectolyase Y23 (Seishin Corp., Tokyo, Japan), 0.5% (w/v) macerozyme R10 (Serva, Heidelberg, Germany), 1% (w/v) bovine serum albumin (Sigma-Aldrich), 1 mm CaCl2, 10 mm HEPES/Tris (pH 7.4) for 45 min at 23°C and 80 rpm on a rotary shaker. The enzyme solution was adjusted to an osmolality of 400 mOsmol kg−1 with sorbitol. Released protoplasts were filtered through a 50-μm nylon mesh and washed with 400 mm sorbitol and 1 mm CaCl2. After centrifugation at 60 g and 4°C for 6 min, the enriched protoplasts were stored on ice until aliquots were used for vacuole isolation and patch clamp experiments. Upon exposure to hypotonic medium (10 mm EGTA, 10 mm HEPES/Tris pH 7.4 adjusted to 200 mOsmol kg−1 with d-sorbitol) protoplasts burst and spontaneously released vacuoles.


Patch clamp experiments on mesophyll vacuoles were performed in the whole-vacuole and excised patch configuration essentially as described by Schulz-Lessdorf and Hedrich (1995) and Ivashikina and Hedrich (2005). Patch pipettes were prepared from Kimax-51 glass capillaries (Kimble products, Vineland, NY, USA) and were covered from the inside with Sigmacoat (Sigma-Aldrich Chemie GmbH, Steinheim, Germany). Close to the tip the outside of the patch pipette was coated with silicone (Sylgard 184 silicone elastomer kit; Dow Corning GmbH, Wiesbaden, Germany). Pipette resistance was about 5 to 6 MΩ for single channel recordings, whereas the pipette resistance for whole vacuole experiments was about 3 MΩ in solutions with 100 mm cytosolic potassium and 2 MΩ in solutions with 150 mm cytosolic potassium. Macroscopic and single channel recordings were performed with an EPC-7 patch clamp amplifier (HEKA, Lambrecht, Germany) at a data acquisition rate of 500 μs and 50 μs, respectively. The macroscopic currents were low-pass filtered at 5 kHz and the single channel currents at 1 kHz. Data were digitized by an ITC-16 interface (Instrutech Corp., Elmont, NY, USA), stored on a Maxdata computer and analysed using different software programs such as Pulse and PulseFit (HEKA Elektronik, Lambrecht, Germany), IGORPro (Wave Metrics Inc., Lake Oswego, OR, USA) and TAC V3.0 (Bruxton Coporation, Seattle, WA, USA). The current recordings were performed according to the convention for electrical measurements on endomembranes (Bertl et al., 1992). The vacuolar membrane was clamped to voltages (V) as indicated in the respective figure legends. To allow comparison of macroscopic current magnitudes among different vacuoles, the current densities (Iss/Cm) were determined upon dividing the macroscopic ionic current through the whole-vacuolar membrane capacitance Cm of the individual vacuole as shown in the figures. The relative voltage-dependent open probability, shown as G(V) curves in Figure 4, was determined from tail current experiments in the whole vacuolar mode. The derived conductance values (G) were plotted against the voltages (V), fitted with a Boltzmann distribution and using a fixed number of apparent gating charges (z = 1.6) and normalized with respect to the maximal conductance (G/Gmax). The half-maximal activation voltage V1/2 was derived from the Boltzmann fit and represents the voltage at which 50% of the maximal conductance is reached. Single channel amplitudes (i) were determined from single channel recordings by using the software program TAC V3.0, and the single channel activity expressed as the single-channel open probability (Po) was estimated as described by Bertl and Slayman (1990).

Patch clamp solutions

Bath and pipette solutions were composed of 2 mm DTT, varied KCl/CaCl2 concentrations and set to an osmolality of 400 mOsmol kg−1 with d-sorbitol. Vacuolar side medium additionally contained 2 mm MgCl2. pH values were adjusted to pH 7.5 or pH 6.5 with 10 mm HEPES/Tris, to pH 5.5 with 10 mm MES/Tris, and to pH 4.5 with 2 mm citrate/Tris. The free Ca2+ concentrations for the pipette and bath mediums were calculated with WEBMAXC standard ( Details about the composition of the solutions are given in the figure legends.

EDX analysis

Leaves of 7–8-week-old wild-type and fou2 plants were quickly frozen to about −175°C in liquid nitrogen and freeze-dried (Alpha1-2, Christ, Osterode, Germany). Cross sections were examined by scanning electron microscopy (SEM, S-520 Hitachi) equipped with an energy dispersive X-ray device (EDX eumex Si(Li)-detector, EUMEX GV, Mainz, Germany).


The work was funded by the Deutsche Forschungsgemeinschaft to RH (SFB 487 and FOR 964). We thank Petra Dietrich for helpful discussion.