INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Elevated aluminum (Al) levels in acid mineral soils are considered one of the major constraints to crop yields worldwide (Haug, 1984). After 80 years of research on the effects of Al on plant growth and development, the primary mechanisms of Al toxicity and Al resistance, as well as the nature of the target cells, remain largely unknown. However, the root apex could be shown to play an essential role for the perception of the Al signal (Ryan et al., 1993). Within the root apex the distal part of the transition zone (1-2 mm from the root tip; DTZ) is most susceptible to Al (Sivaguru and Horst, 1998; Horst et al., 1999; Sivaguru et al., 1999; Kollmeier et al., 2000). The Al response in this root zone is manifold, including severe changes in the organization of the cytoskeleton (Blancaflor et al., 1998; Horst et al., 1999; Sivaguru et al., 1999) and rhizoplane pH profiles (Kollmeier et al., 2000). Recent results suggest a transduction of the Al signal from the site of perception (DTZ) to its site of action (elongation zone, EZ) through Al-induced alterations in indole-3-acetic acid fluxes in the root apex, finally leading to inhibited root elongation (Kollmeier et al., 2000).

Al toxicity is strongly dependent on the availability of the most rhizotoxic monomeric Al species (Kinraide, 1991). The exudation of organic compounds capable of chelating Al into non-rhizotoxic complexes was predicted to increase Al resistance (Delhaize et al., 1993; Horst, 1995; Kochian, 1995; Pellet et al., 1995, 1996; Ryan et al., 1995a, 1995b; Jones, 1998; Koyama et al., 1999). Depending on the plant species, malate, oxalate, and citrate have been identified as the major Al-induced organic acid anions in root exudates (Miyasaka et al., 1991; Delhaize et al., 1993; Basu et al., 1994; Pellet et al., 1995, 1996; Ryan et al., 1995a, 1995b; Jones, 1998; Zheng et al., 1998). Due to the negative membrane potential and the steep concentration gradient between cytosol and apoplast or rhizosphere, respectively, organic acid anions have been postulated to reach the sites of Al binding in the apoplast through anion channels (Delhaize and Ryan, 1995; Kochian, 1995; Schroeder, 1995; Jones, 1998). In this context, Ryan et al. (1997) could demonstrate the Al-dependent activation of an anion channel in root tip protoplasts isolated from an Al-resistant wheat line. Following the application of Al, channel activation was delayed by 10 to 90 min. Piñeros and Kochian (2001) recently described an Al-activated anion channel in protoplasts derived from the maize (Zea mays) root tip. However, Ryan et al. (1997) and Piñeros and Kochian (2001) did not compare cultivars differing in Al resistance or the defined electrical response of protoplasts isolated from distinct apical root zones.

In search for the molecular mechanisms involved in Al-induced release of organic acid anions from maize roots, we applied the patch-clamp technique to protoplasts derived from the cortex of the root apex of two maize cultivars differing in Al resistance. Since recent observations revealed considerable differences in Al-sensitivity between different apical root zones (Ryan et al., 1993; Ryan et al., 1995a; Sivaguru and Horst, 1998; Horst et al., 1999; Sivaguru et al., 1999; Kollmeier et al., 2000), we furthermore compared protoplasts derived from the DTZ and EZ. The results gained from patch-clamp experiments were compared with exudation experiments on vertically grown root apices of intact seedlings. To our knowledge, we document here for the first time that an Al-activated anion channel in the DTZ is capable of releasing the relevant Al-chelating organic acids malate and citrate.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Effect of Al on the Exudation of Organic Acid Anions

In both cultivars of maize, Al treatment (100 µM, 2 h) enhanced the total exudation rate of organic acid anions (oxalate, cis-aconitate, citrate, -keto-glutarate, malate, trans-aconitate, and fumarate) from the 5-mm root apex (Fig. 1a). Increasing the Al concentration from 100 to 200 µM did not further increase the exudation rate. Among the organic acid anions with stronger Al complexation capacity, citrate and malate were exuded at comparative rates, and the exudation was equally enhanced by Al (Fig. 1b). Al-induced exudation rates of all organic acid anions detected, particularly citrate and malate, were higher in the Al-resistant cv ATP-Y than the Al-sensitive cv Lixis.

View larger version (23K): [in this window] [in a new window]   Figure 1.   Effect of Al on the exudation of organic acid anions from 5-mm root apices of intact maize seedlings. Root apices were incubated in a solution containing 200 µM CaCl2 and 0, 100, or 200 µM AlCl3 (pH 4.3) for 2 h. a, Total exudation rate of organic acid anions. b, Exudation rate of citrate (black bars) and malate (gray bars). Values are means of three independent replicates ± SD. Results shown are representative of two independent experiments. Different letters indicate significant differences at P < 0.05 (Tukey test). Small letters display treatment effects, and capital letters depict genotypical differences.

Al Induces a Large Conductance Anion Channel in Cortical DTZ Protoplasts

To characterize the cell type releasing organic acid anions upon Al application we isolated cortex protoplasts from the DTZ (1-2 mm) and the main EZ (3-5 mm) as indicated in Figure 2. Root protoplasts with average diameters in the DTZ of 25.14 ± 1.76 µm (n = 148) and in the EZ of 52.74 ± 5.46 µm (n = 114) were selected for the patch-clamp experiments.

View larger version (86K): [in this window] [in a new window]   Figure 2.   Origin and size of maize root protoplasts. Protoplasts were isolated from cortical cells of 3- to 4-d-old primary roots by enzymatic digestion. Protoplasts were derived from the DTZ (1-2 mm) or from the main EZ (3-5 mm). The bar represents 10 µm.

Following the application of the patch-clamp technique we studied changes in the electrical properties of the plasma membrane in response to Al. To better resolve anion channels we used electrolyte solutions basically containing tetraethylammonium (TEA⁺) and Cl. The K⁺ channel blocker TEA eliminated inward and outward K⁺ currents (Blatt and Gradmann, 1997). From a holding potential of 52 mV, consecutive 2-s voltage pulses were applied to the plasma membrane in 20-mV steps from 132 to 88 mV. Under conditions of basically 20 mM TEACl in the bath solution and 100 mM TEACl in the pipette, and the absence of Al in the bath and the preincubation medium, only background currents were detected in protoplasts isolated from both root zones of both cultivars as demonstrated for the DTZ in Figure 3, a and b. In protoplasts released from the cortical DTZ of roots pretreated with Al, inward currents were observed in both cultivars (Fig. 3, c-e). In response to hyperpolarization, the macroscopic currents were elicited instantaneously. On top of the macroscopic currents, single channel fluctuations could be resolved. In response to depolarization, slowly activating macroscopic outward currents appeared. The current-voltage curve obtained from steady-state current amplitudes reversed direction close to E_Cl (Nernst). These macroscopic currents were detected in eight out of 10 protoplasts derived from the Al-resistant cv ATP-Y and three out of nine from the Al-sensitive cv Lixis. The average current amplitude was 282 ± 77 pA (cv ATP-Y, n = 8) and 338 ± 71 pA (cv Lixis, n = 3). Removal of Al from the bath solution did not decrease or eliminate the current amplitudes. In accordance with this, in whole-plant experiments, removal of Al from the incubation medium had no significant effect on the Al-induced citrate exudation within 2 h (Fig. 4). Direct Al application (50 µM) to DTZ protoplasts that had not been in contact with Al before elicited anion currents during three out of nine whole-cell measurements on protoplasts derived from the Al-resistant, but not in the Al-sensitive cultivar (zero out of five). When the results obtained from the two treatment types (pretreatment plus direct application and direct application only) were pooled, genotypical differences were statistically significant (one-sided P = 0.036, according to Fisher's exact test). The amplitude of the macroscopic currents, as well as single channel conductance, was identical with those described above. In contrast to protoplasts derived from Al-pretreated roots, inward currents appeared only after a delay of 7 to 10 min when Al was added. To examine whether this delay represents the time required for Al to enter the cytosol we applied Al (10 µM) to the cytosolic face of the plasma membrane through the patch pipette in the whole-cell configuration: Under these conditions anion currents appeared without noticeable delay (n = 3, not shown). However, disintegration of the plasma membrane occurred quickly after application of the first voltage pulses, allowing no further characterization of these currents.

View larger version (23K): [in this window] [in a new window]   Figure 3.   Al induction of anion channels in the plasma membrane of protoplasts isolated from the DTZ of the primary root of an Al-resistant and an Al-sensitive maize cultivar. From a holding potential of 52 mV, consecutive 2-s voltage pulses were applied to the plasma membrane in 20-mV steps from 132 to 88 mV. Whole-cell currents were measured in protoplasts isolated from cortical DTZ cells of intact root apices incubated for 1 h in an agarose gel (0.6%, w/v) containing NS and 0 µM (a and b) or 90 µM (c and d) Al. a and c, Al-resistant cv ATP-Y; b and d, Al-sensitive cv Lixis. The pipette solution contained 100 mM TEACl, 2 mM MgCl2, 2 mM MgATP, 2 mM EGTA, and 10 mM HEPES/Tris (pH 7.2), and the bath solution contained 20 mM TEACl, 1 mM CaCl2, 0.05 mM AlCl3, and 5 mM MES/Tris (pH 4.3). e, Corresponding current-voltage relation of anion channels induced by Al treatment of maize roots in the steady state. Black circles, Al-resistant cv ATP-Y; white circles, Al-sensitive cv Lixis.

View larger version (19K): [in this window] [in a new window]   Figure 4.   Effect of Al removal from the incubation medium on the exudation of citrate from root apices of intact maize seedlings (Al-resistant cv ATP-Y). Root apices (10 mm) were incubated in a solution containing 200 µM CaCl2 ± 100 µM AlCl3 (pH 4.3) for 2 h. Al-treated roots were then rinsed in ultrapure water and were transferred to incubation medium without Al. The roots were incubated for another 2 h in this Al-free solution. Values are means of four independent replicates ± SD. Results shown are representative of two independent experiments. Different letters indicate significant differences at P < 0.05 (Tukey test).

When Al³⁺ was replaced by La³⁺, no inward currents were elicited, whether it was supplied directly (50 µM) to the protoplasts (n = 6) or to root apices of intact seedlings prior to protoplast isolation (n = 5 for ATP-Y; not shown).

The results gained by voltage step protocols, as well as by continuous voltage ramps (15-30 consecutive 200 ms ramps from 132 to 108 mV; not shown) through single channels revealed a depolarization-activated Al-induced anion channel. The single-channel amplitude at depolarized potentials was, however, much smaller. Hence, a much larger number of channels account for the outward currents.

In contrast to DTZ-derived protoplasts, anion-channel activity was not induced upon Al preincubation (n = 5 for both cultivars) or direct Al application (n = 4 for both cultivars) in EZ-derived protoplasts.

Conductance, Selectivity, and Pharmacology

Amplitudes of the single-channel fluctuations observed on top of the macroscopic currents (Fig. 3, c and d) were analyzed in the steady state (continuous voltage application) in the whole-cell configuration (Fig. 5a). Figure 5b depicts the single-channel current/voltage curves in the presence of standard solutions (n = 7). The respective current-voltage curves reversed their direction close to the Nernst potential for Cl. The single-channel conductance was 144 ± 10 pS in the linear range between 132 and 52 mV (n = 7, as depicted by the line in Fig. 5b).

View larger version (18K): [in this window] [in a new window]   Figure 5.   Voltage-dependent single-channel anion currents across the plasma membrane of DTZ-derived protoplasts as measured in the whole-cell configuration. Anion channels were activated by Al treatment of roots from the Al-resistant cv ATP-Y prior to protoplast preparation. The pipette solution contained 100 mM TEACl, 2 mM MgCl2, 2 mM MgATP, 2 mM EGTA, and 10 mM HEPES/Tris (pH 7.2), and the bath solution contained 20 mM TEACl, 1 mM CaCl2, 0.05 mM AlCl3, and 5 mM MES/Tris (pH 4.3). a, Steady-state channel fluctuations at voltages indicated. oi indicates open states, and c the closed state or open state for additional active channels. b, Corresponding current-voltage curves in the presence of standard solutions (n = 7). The straight line represents a linear fit to inward current amplitudes between 132 and 52 mV. The voltage-dependent rectification of the current-voltage curve reverses direction near the Nernst potential for chloride (arrow).

The selectivity of the anion channel was analyzed in the presence of different Cl gradients across the plasma membrane (Fig. 6). From a holding potential of 52 mV the membrane potential was stepped to 140 mV, and during a 2-s ramp to 120 mV, the current response was monitored continuously (Fig. 6a). Although the amplitude of the inward currents was marginally affected, the outward current increased progressively with the external rise in Cl concentration. Changing the external Cl concentration between 4 and 302 mM resulted in a shift of E_rev by 58 mV per 10-fold concentration change, indicating that the halide is the major charge carrier of the current (Fig. 6b). Upon lowering the internal Cl concentration from 104 to 64 mM, the channel conductance decreased by 20% ± 5% (n = 3, not shown).

View larger version (17K): [in this window] [in a new window]   Figure 6.   Chloride dependence of the reversal potential. Protoplasts were isolated from the cortical DTZ of Al-pretreated roots of the Al-resistant cv ATP-Y. The pipette solution consisted of 100 mM TEACl, 2 mM MgCl2, 2 mM MgATP, 2 mM EGTA, and 10 mM HEPES/Tris (pH 7.2). The external solution contained 1 mM CaCl2, 0.05 mM AlCl3, and 5 mM MES/Tris (pH 4.3) in addition to 2, 20, 102, 150, or 300 mM TEACl. The calculated Nernst potentials for the respective ratios of internal and external Cl concentrations are indicated by arrows. a, Currents in response to continuous 2-s voltage ramps in the presence of external [Cl] as indicated ranging from 132 to 88 mV. b, Reversal potentials as a function of the extracellular Cl concentration were obtained through application of step voltage-protocols or 2,000 and 200 ms voltage ramps. The number of replicates for the respective [Cl] is given in brackets. The line represents the predicted Nernst behavior. A decrease in external Cl concentration from 22 to 4 mM shifted the reversal potential (Erev) from 28 ± 7 mV (n = 6) to 67 ± 11 mV (n = 6), whereas an increase to 104 mM shifted the reversal potential to 1 ± 3 mV (n = 5). Increasing the external Cl concentration to 152 and 302 mM shifted the reversal potential to 10 ± 2 mV (n = 3) and 27 ± 2 mV (n = 3), respectively.

Addition of the anion channel inhibitors niflumic acid or 4, 4'-dinitrostilbene-2, 2'disulfonic acid (DIDS; 100 µM) to the incubation medium of 5-mm root apices of intact seedlings of the Al-resistant cv ATP-Y prevented the Al-induced exudation of organic acid anions generally (Fig. 7a) and citrate and malate specifically (Fig. 7b). The non-Al-dependent exudation of organic acid anions as detected in the control treatment was hardly affected, indicating a different mechanism responsible for this efflux.

View larger version (18K): [in this window] [in a new window]   Figure 7.   Effect of anion channel inhibitors on the Al-induced exudation of organic acid anions from 5-mm root apices of intact maize seedlings. The root tips were incubated in a solution containing 200 µM CaCl2 and 0 or 50 µM AlCl3 (pH 4.3) for 2 h with or without niflumic acid or DIDS (100 µM). a, Total exudation rate of organic acid anions. b, Exudation rate of citrate (black bars) and malate (gray bars). Values are means of three independent replicates ± SD. Different letters indicate significant differences at P < 0.05 (Tukey test).

Treatment with anion channel inhibitors affected the anion channel accordingly. The macroscopic currents were blocked to 90% ± 5% (n = 12) by 100 µM niflumic acid within 30 s after application (Fig. 8a). The single-channel conductance was completely blocked instantaneously (n = 8, Fig. 8b). Addition of 100 µM of the stilbene derivative DIDS led to a 60% ± 5% (n = 6, Fig. 8a) inhibition of the currents at 132 mV. The single-channel conductance decreased accordingly (n = 2, Fig. 8b). Addition of 10 µM ZnCl₂, which blocked Cl channels in giant algae (Hille, 1992) and the vacuolar SV channel (Hedrich and Kurkdijan, 1988), however, did not affect the single channels or macroscopic anion currents (n = 8, not shown). Based on the conductance, selectivity, and pharmacology of the macroscopic currents and single anion channels, one might conclude that about 15 of the 144 pS anion channels per cell generate the whole-cell inward currents.

View larger version (14K): [in this window] [in a new window]   Figure 8.   Effect of anion channel inhibitors on Al-induced macroscopic anion currents and single-channel amplitudes. From a holding potential of 52 mV consecutive 2-s voltage pulses were applied to the plasma membrane incrementing in 20-mV steps from 132 to 88 mV. The pipette solution contained 100 mM TEACl, 2 mM MgCl2, 2 mM MgATP, 2 mM EGTA, and 10 mM HEPES/Tris (pH 7.2), and the bath solution contained 20 mM TEACl, 1 mM CaCl2, 0.05 mM AlCl3, and 5 mM MES/Tris (pH 4.3). a, Steady-state current-voltage relations of the Al-induced anion channel before () and after application of 100 µM niflumic acid () or DIDS (). b, Steady-state single-channel fluctuations derived from the whole-cell configuration at 132 mV in the absence (control) or presence of 100 µM niflumic acid or DIDS applied to two different protoplasts.

The Al-Induced Large Conductance Anion Channel Is Permeable to Malate and Citrate

To examine the permeability of the anion channel to Al complexing organic acid anions, 60 mM Cl in the pipette solution was replaced by malate² or citrate³ (Fig. 9). In malate- and citrate-containing pipette solutions, the voltage dependence and kinetics of inward and outward currents were similar to chloride-containing patch solutions, as shown in Figure 9a. The macroscopic currents at 132 mV were reduced from 243 ± 24 pA for chloride (n = 3) to 68 ± 15 pA for malate (n = 3) and 52 ± 10 pA for citrate (n = 3). Similar to the behavior of macroscopic currents, the single-channel amplitudes were reduced by exchange of Cl against organic acid anions (Fig. 9b). From the corresponding current-voltage relation, single-channel conductances of 118 ± 4 pS for chloride (n = 3), 34 ± 4 pS for malate (n = 3), and 21 pS for citrate (n = 2) were determined in the linear range of the current-voltage curve between 132 and 32 mV (Fig. 9c). The relative permeability of these organic anions calculated from macroscopic reversal potentials according to Fatt and Ginsborg (1958) were P_malate/P_Cl = 0.25 ± 0.03 (n = 3) and P_citrate/P_Cl = 0.18 ± 0.07 (n = 3), indicating that the physiologically relevant halide and the two carboxylic acids are able to permeate the anion channel.

View larger version (26K): [in this window] [in a new window]   Figure 9.   Malate and citrate permeability of the Al-activated anion channel. Protoplasts were loaded with 2 mM MgCl2, 2 mM MgATP, 2 mM EGTA, and 10 mM HEPES/Tris (pH 7.2) in addition to 60 mM TEACl, 60 mM TEA2-malate, or 60 mM TEA3-citrate. The external solution contained 60 mM TEACl, 1 mM CaCl2, and 5 mM MES/Tris (pH 4.3). a, Anion currents were measured in response to 2-s voltage pulses ranging from 132 to 88 mV in 20-mV steps starting from a holding potential of 52 mV. b, Steady-state single channel activities from the whole-cell configuration at 132 mV upon loading the protoplasts with 60 mM chloride, malate, or citrate. c, Current-voltage relationship for single-channel conductances as measured at different voltages (132 to 32 mV) under conditions of 60 mM TEACl in the bath and 60 mM TEACl (), 60 mM TEA2-malate (), or 60 mM TEA3-citrate () in the pipette. Lines represent linear fits to the data from n = 3 (chloride and malate) or 2 (citrate).

Al Activates the Large Anion Channel in Cycloheximide-Independent Manner

To examine whether the presence of the channel observed or its activation in response to Al requires protein synthesis, roots of intact seedlings (Al-resistant cv ATP-Y) were incubated with 100 µM of the protein synthesis inhibitor cycloheximide [3-(2-{3, 5-dimethyl-2-oxocyclohexyl}-2-hydroxyethyl) glutarimide] for 45 min before and during Al treatment (1 h). Following this treatment, the number of viable protoplasts severely decreased, whereas frequency of anion channel activation (four out of five for cv ATP-Y, not shown) and conductance remained unchanged. This indicates that the biochemical machinery for transducing the Al signal into channel activation was already established before addition of Al.

Effect of Al on K⁺ Channels

In addition to the increase in organic acid exudation into the culture medium, elevated K⁺ efflux from the root has been observed (Ryan et al., 1995a, 1997). In accordance with this we studied K⁺ channels in the plasma membrane of protoplasts derived from the DTZ of non-Al-pretreated roots of the Al-resistant cv ATP-Y and analyzed their Al sensitivity. These experiments were performed with solutions containing 150 mM K-gluconate in the pipette and 50 mM K-gluconate in the bath. From a holding potential of 63 mV, consecutive 2-s voltage pulses were applied to the plasma membrane from 143 to 177 mV. In response to these hyperpolarizing and depolarizing voltage pulses, time-dependent outward- and inward-rectifying K⁺ currents were recorded (Fig. 10, a and b). Upon application of up to 300 µM AlCl₃, the K⁺ outward rectifier remained unaffected within 10 min (n = 3), whereas the inward rectifier was already inhibited by 70% ± 12% at 143 mV within 5 min (n = 3; Fig. 10c).

View larger version (18K): [in this window] [in a new window]   Figure 10.   Effect of Al on K+ channels located in the plasma membrane of maize root protoplasts. The pipette solution consisted of 150 mM K-gluconate, 2 mM MgCl2, 5 mM MgATP, 1 mM EGTA, and 10 mM HEPES/Tris (pH 7.2). The bath solution was composed of 50 mM K-gluconate, 1 mM CaCl2, and 5 mM MES/Tris (pH 4.3). Protoplasts were isolated from cortical DTZ cells of the Al-resistant cv ATP-Y pretreated without Al. From a holding potential of 63 mV 2-s voltage pulses were applied between 143 and 177 mV in 20-mV steps. a, Slowly activating outward K+ currents upon depolarizing voltage steps. The curve shown is representative for three independent measurements. b, Representative activation of inward K+ channels upon membrane hyperpolarization (n = 3). c, Superposition of current-voltage curves of inward and outward K+ channels. Data points were obtained from protoplasts perfused with solutions containing 0 µM AlCl3 ( or ) or 300 µM AlCl3 (192 µM Almono;  or ). Values are means out of three replicates ± SD.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Al has been reported to inhibit ion channels such as an inward-rectifying K⁺-channel in wheat root-hair protoplasts (Gassmann and Schroeder, 1994) or Ca²⁺ channels in wheat roots (Huang et al., 1994, 1996; Piñeros and Tester, 1995). Al-induced channel activation, however, has been rarely observed. In their pioneering study on protoplasts derived from enzymatically treated wheat root apices (3 mm) of the Al-resistant wheat line ET 8, Ryan et al. (1997) described an Al-activated anion channel. Al-induced anion currents were elicited instantly in 20% of protoplasts, whereas in another 30% of 73 protoplasts, an inward current was activated after a delay of up to 90 min. Piñeros and Kochian (2001) recently reported on Al-sensitive and -insensitive chloride and cation channels in the plasma membrane of protoplasts derived from the maize root apex (5 mm).

In our patch-clamp study we compared the impact of Al on the electrical properties of the plasma membrane of protoplasts derived from roots of two maize cultivars differing in Al resistance. Furthermore, we focused on cortex protoplasts originating from two defined apical root zones of outstanding importance for Al toxicity and resistance: the DTZ (1-2 mm) and the main EZ (3-5 mm). The DTZ of the primary maize root is most Al-sensitive (Sivaguru and Horst, 1998; Horst et al., 1999; Sivaguru et al., 1999), even though it does not considerably contribute to root elongation. In contrast, the EZ is the site of inhibited cell elongation upon Al exposure to the DTZ or the entire root apex, whereas it is far less Al-sensitive upon direct Al treatment (Kollmeier et al., 2000).

There is substantial evidence for a positive correlation between Al-induced exudation of organic acid anions and Al resistance (Delhaize et al., 1993; Basu et al., 1994; Pellet et al., 1995, 1996; Ryan et al., 1995a, 1995b; Schroeder, 1995; Cocker et al., 1998; Jones, 1998). Although most of these studies focused on wheat, substantial data on maize are still rare. In the study presented here, short-term (2 h) application of Al induced the exudation of organic acid anions, primarily citrate and malate, from the apical 5 mm of the root apex (Fig. 1). In contrast to former results presented by Delhaize et al. (1993), Ryan et al. (1995a, 1995b), and Pellet et al. (1995, 1996), which demonstrated the dominant exudation of malate in wheat and citrate in maize, our study showed that Al induced the release of citrate and malate at similar rates in 3-d-old maize seedlings of both maize cultivars. The rates of malate exuded by both cultivars in the controls do not seem to convey Al resistance, since cv Lixis proved to be Al-sensitive in a number of studies (Llugany et al., 1994; Horst et al., 1997; Sivaguru and Horst, 1998; Kollmeier et al., 2000). The Al-induced increase in exudation rates of citrate and malate as well as the cultivar differences observed in this study were less pronounced than in the studies by Delhaize et al. (1993) and Ryan et al. (1995a) on wheat. However, a direct comparison proves difficult since a whole range of experimental factors, i.e. species, seedling age, root zone, Al concentration, as well as treatment duration, may have considerable influence on the results. Furthermore, citrate is much more effective in rendering Al non-phytotoxic than is malate (Ownby and Popham, 1989; Delhaize et al., 1993; Pellet et al., 1995; Jones, 1998). Thus, much lower exudation rates will be sufficient to confer Al resistance in species exuding citrate such as maize.

In protoplasts derived from cortical cells of the DTZ we found an Al-induced large conductance anion channel in both cultivars (Fig. 3), although less frequently in the Al-sensitive cv Lixis. With its large conductance (144 pS under conditions of 104 mM Cl in the pipette and 22 mM Cl in the bath; Fig. 5) this channel is larger than the large conductance anion channel described by Piñeros and Kochian (2001). This might be due to a different protoplast population patch-clamped in that study since size and origin of the protoplasts were more variable than in our study in which cortical protoplasts from defined apical root zones were isolated. The observation of genotypical differences in frequency of channel activation is in agreement with the in vivo exudation of organic acid anions (Fig. 1). Thus, the rate of organic acid anion release from root apices very likely determines the degree of Al resistance (Pellet et al., 1995; Ryan et al., 1995a, 1995b). Nevertheless, the question arises whether the difference in Al resistance between the two maize cultivars investigated in this study is exclusively related to the capacity of the DTZ to release organic acid anions through anion channels. The absence of Al-induced anion channel activity in the EZ indicates that this zone is unlikely to contribute to the Al-induced exudation of organic acid anions. As a consequence, the lower uptake/binding of Al and phytotoxicity (root elongation and callose synthesis) of Al when supplied to the EZ (Sivaguru and Horst, 1998; Kollmeier et al., 2000) can hardly be explained by stimulated release of organic acid anions from this apical root zone.

Permeability of guard cell anion channels to organic acid anions such as malate and citrate has been demonstrated by Hedrich and Marten (1993), Schmidt and Schroeder (1994), and Dietrich and Hedrich (1994, 1998). Here we present the first evidence that the Al-induced root anion channel in maize is permeable for the organic acids malate and citrate. Thereby, we observed relative permeabilities of malate and citrate compared with chloride of 0.25 and 0.18, respectively (Fig. 9). Due to its large conductance and since the channel did not deactivate even after removal of Al, it will mediate considerable secretion of malate and citrate into the apoplast after being triggered by Al. The observation that once activated, the channel remained active even after removal of Al is different from results reported by Ryan et al. (1995a), who demonstrated that exudation of malate from wheat roots ceased within 15 min after removal of Al from the incubation solution, and from Piñeros and Kochian (2001), who, in maize protoplasts, showed immediate deactivation of an Al-induced anion channel after removal of Al. However, our results are in agreement with our whole-plant experiment shown in Figure 4.

The anion channel could not be activated by La³⁺, confirming the specificity for Al. Delhaize et al. (1993) and Ryan et al. (1995a, 1997) also demonstrated that the release of malate from wheat-root apices and anion-channel activation were insensitive to La³⁺. This is in agreement with reports by Kinraide et al. (1992), indicating that genotypical differences in Al sensitivity are not in accordance with La sensitivity.

Application of 100 µM of the anion channel blocker niflumic acid inhibited the macroscopic anion currents and single-channel fluctuations within 2 min after application of the anion channel blocker (Fig. 8). A similar effect of the anion channel blocker on the exudation rate of organic acid anions was observed when root apices of intact seedlings were treated (Fig. 7). These results are in agreement with those reported by Ryan et al. (1995a, 1997) showing that the exudation of organic acid anions from wheat roots decreased upon incubation with niflumic acid, and that an Al-induced anion channel in wheat was blocked by this substance.

Because application of cycloheximide did not affect channel activation and single channel conductance, we conclude that exocytosis or post-translational modifications of the channel protein are responsible for Al-induced channel activation of the large anion channel in maize described here. Ryan et al. (1995a), on the other hand, demonstrated that cycloheximide impaired malate exudation in wheat within 15 min. However, this result is difficult to reconcile with the missing lag phase for Al-induced malate exudation, which does not support the involvement of de novo protein synthesis in the activation process.

The activation of the anion channel was observed after preincubation of the roots with Al and application of the metal ion directly to protoplasts in the whole-cell configuration of the patch-clamp technique. In the latter experiments the activation was delayed by 7 to 10 min. This might be indicative of a complex Al-signaling pathway. It should be noted that secondary messengers such as cytosolic Ca²⁺ (Lindberg and Strid, 1997; Jones et al., 1998a, 1998b; Zhang and Rengel, 1999) or IP₃ (Haug et al., 1994; Jones and Kochian, 1995) have been proposed as cellular mediators of the Al signal. The induction of callose formation by Al, a sensitive indicator of Al injury frequently observed in roots, suspension cells, and protoplasts (Stass and Horst, 1995; Horst et al., 1997), is an indicator for modified plasma membrane characteristics and altered cell wall configuration, as well as an increase in cytosolic [Ca²⁺] (Kauss, 1996). Ryan et al. (1997) did not find any effect on anion channel activation upon changing the concentration of Ca, ATP, or IP₃ in the patch-pipette. This indicates that even though a number of studies clearly demonstrated effects of Al on ATP (Collier et al., 1993), IP₃ (Jones and Kochian, 1995), and cytosolic calcium concentrations (Rengel, 1992a, 1992b; Haug et al., 1994; Jones et al., 1998a, 1998b; Zhang and Rengel, 1999), in roots they might not be involved in this mechanism of Al resistance. These apparent contradictions between studies conducted on the whole plant, tissue, or cellular level and protoplast experiments further stress the importance of studying whole-tissue reactions toward Al and they also support the idea of a role of the apoplast in Al toxicity and resistance as proposed by Horst (1995).

The release of organic acid anions from the cytosol will result in a depolarization of the plasma membrane. Electroneutrality will have to be maintained by an equivalent uptake of anions or efflux of cations (Delhaize and Ryan, 1995; Jones, 1998). Ryan et al. (1995a) observed the concomitant exudation of K⁺ with malate, whereas Murphy et al. (1999) demonstrated that copper-induced, ion channel-mediated efflux of K⁺ in Arabidopsis seedlings was coupled with citrate exudation. These results are in agreement with the insensitivity of the K⁺ outward rectifier toward Al observed in this study (Fig. 10), thus providing the potential for K⁺ release to charge balance Al-induced anion exudation. K⁺ release rather than H⁺ release is also supported by recent studies demonstrating the maintenance of a higher root surface pH in the DTZ of an Al-resistant compared with an Al-sensitive maize cultivar upon Al treatment (Kollmeier et al., 2000), which might be due to protonation of citrate and malate in the apoplast.

Our results with protoplasts isolated from the cortical DTZ of the primary maize root support and allow for the extension of the model proposed by Delhaize and Ryan (1995) on Al-induced effects on the electrophysiological properties of the plasma membrane of root cells. Al induces a large conductance anion channel after a short delay if applied externally. The K⁺ outward rectifier is not affected by Al, thus allowing the concomitant release of K⁺ and organic acid anions stabilizing the membrane potential after initial depolarization (Papernik and Kochian, 1997; Takabatake and Shimmen, 1997; Sivaguru et al., 1999).

The short-term involvement of transcriptional and translational regulation of the Al-activated anion channel appears rather unlikely. Fusion of vesicles containing channel proteins with the plasma membrane might be involved in enhanced channel synthesis. In this context it is intriguing to speculate that the reported maintenance of Al accumulation of tobacco cells by the vesicle transport-inhibitor brefeldin A (Vitorello and Haug, 1999) might be due to reduced plasma membrane permeability for organic acid anions.

The induction of anion currents without any lag phase after Al supply to the internal face of the plasma membrane might suggest an activation mechanism through a signal transduction cascade triggered by cytosolic Al. This would be in line with conclusions drawn by Jones and Kochian (1998) concerning the involvement of an intracellular target site for Al triggering inhibition of root growth.

We suggest that Al binds to the plasma membrane, thus triggering a membrane delimited signal transduction pathway, finally leading to the activation of the already assembled anion channel. The resulting efflux of organic anions provides the potential of reducing the activity of toxic Al in the apoplast, thus increasing Al resistance.

Based on the data gained from whole-plant and patch-clamp experiments presented here, we propose that the Al-activated large conductance anion channel described here contributes to enhanced genotypical Al resistance by facilitating the exudation of organic acid anions from the DTZ of the maize root apex. The transduction pathway involved in channel activation, as well as the elucidation of additional physiological mechanisms conferring Al resistance, are subjects of our ongoing research.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Material and Experimental Conditions

Selected seeds of the maize (Zea mays) cv ATP-Y (Al-resistant, Dr. Thé, Institut de la Recherche Agronomique du Cameroon) and cv Lixis (Al-sensitive, Force Limagrain, Montpellier, France) were germinated in filter paper rolls moistened with nutrient solution (NS). Plants were grown at 30°C/26°C, with a photon flux density of 300 µmol m² s¹ and a day light cycle of 16/8 h. The relative humidity was approximately 70%. Within 24 h before the start of the experiments the seedlings were gradually adapted to low pH (4.3). Three- to 4-d-old seedlings were selected for similar primary root lengths between 9 and 13 cm. For patch-clamp experiments the root apices were treated for 1 h in agarose gel (0.6% [w/v], pH 4.3) containing NS and 0 or 90 µM AlCl₃. The NS (pH 4.3) was composed of (in micromoles): CaSO₄, 250; KNO₃, 400; MgSO₄, 100; FeEDDHA, 20; MnSO₄, 1; ZnSO₄, 0.2; CuSO₄, 0.2; KH₂PO₄, 10; H₃BO₃, 8; (NH₄)₆Mo₇O₂₄, 0.1; and NH₄NO₃, 200.

Exudation of Organic Acid Anions from Root Tips

For each replicate, roots of 10 intact 3-d-old seedlings were bundled. The tips (5 mm) were incubated for 2 h in 4 mL of a solution containing 200 µM CaCl₂ (pH 4.3) and 0, 50, 100, or 200 µM AlCl₃, respectively. Concentrations of monomeric Al as determined according to Kerven et al. (1989) were 0, 45, 89, and 169 µM. The free Al activities as calculated with GEOCHEM were 0, 31, 59, and 113 µM. Experiments were conducted under the conditions described above. The remainder of the roots was kept moist by wrapping them in filter paper soaked with CaCl₂ solution (200 µM). The incubation was performed in filtration columns (Bakerbond SPE, J.T. Baker, Phillipsburg, NJ) loaded with 1 g of an AG 1-X8 anion-exchange resin (100-200 mesh; Bio-Rad Laboratories, Richmond, CA). After removing the roots, the incubation medium was passed through the exchange resin at a rate of 1 mL min¹. The anion-exchange resin was then rinsed with 5 mL of formic acid (8 M) twice at a rate of 1 mL min¹. The formic acid was evaporated in a centrifugal evaporator (RCT 10-22T, Jouan, Saint-Herblain, France), the residue was dissolved in 1 mL of perchloric acid (10 mM), and was then filtered through 0.45-µm filtration inserts (Ultrafree-MC, Millipore, Eching, Germany). Samples were analyzed by isocratic HPLC (Kroma System 2000, Kontron Instruments, Munich, Germany) separated on an Aminex HPX-87H column (Bio-Rad Laboratories) supplemented with a cation H⁺ microguard cartridge, using 10 mM perchloric acid as eluent at a flow rate of 0.5 mL min¹ at 35°C (Oven 480, Kontron Instruments).

Effect of Anion Channel Inhibitors on Exudation of Organic Acid Anions

To examine the involvement of anion channels in the exudation process, the anion channel inhibitors niflumic acid or DIDS (100 µM) were added to the root-tip incubation medium containing 200 µM CaCl₂ and 0 or 50 µM AlCl₃ (incubation for 2 h as described above). None of these substances had an effect on the Al_mono concentrations.

Protoplast Isolation

Protoplasts were enzymatically isolated from two different root zones, the 1- to 2-mm DTZ and the 3- to 5-mm main EZ from the root tip (Fig. 2). After pretreatment the root material was rinsed with distilled water to remove excess solution containing Al from the surface. The material of 40 roots was chopped and pooled for the digestion process yielding 40 × 1 mm (DTZ) and 40 × 2 mm (EZ) sections, respectively. The digestion process was based on the method described by Bregante et al. (1997) adapted for the two root zones. The root fractions were first incubated in 2 mL of a solution containing 1 mM CaCl₂, 0.5% (w/v) polyvinylpyrrolidone, 0.5% (w/v) bovine serum albumin, 0.8% (w/v) cellulase (Onozuka RS, Yakult Honsha Co., Tokyo), 0.1% (w/v) pectolyase (Sigma), 8 mM MES [ 2-(N-morpholino)-ethanesulfonic acid]-KOH to pH 5.5, and sorbitol to 550 mosmol kg¹ and were then agitated at 65 rpm for 50 min at 30°C. Two milliliters of the same solution without pectolyase at pH 5.8 was added to the medium for another 20 min of agitating. To separate the protoplasts from undigested tissue and cell debris, the suspension was then filtered through 75-µm gaze and centrifuged at 60g for 5 min. The supernatant was discarded and the protoplast pellet was resuspended in 5 mL of washing solution (1 mM CaCl₂, 5 mM MES/Tris, pH 5.5, and sorbitol to 570 mosmol kg¹). After two additional washing steps the protoplasts were stored in a solution containing 1 mM CaCl₂, 5 mM MES/Tris (pH 5.5), and sorbitol to 570 mosmol kg¹. Patch-clamp experiments were performed during the following 5 h. Cortex protoplasts were separated from stele parenchyma cells as described by Bregante et al. (1997) and Roberts and Tester (1995). Due to the higher resistance of the rhizodermis cell wall toward cellulases and pectolyases (see Kochian and Lucas, 1983), contamination by epidermal protoplasts could be largely prevented.

Patch-Clamp Recordings

All ion fluxes were studied in the whole-cell configuration of the patch-clamp technique (Hamill et al., 1981). Current measurements were performed using a patch-clamp amplifier (EPC-7, List Electronic, Darmstadt, Germany) and were low-pass-filtered with an eight-pole Bessel filter at 2 kHz. Data were sampled at 5 kHz, digitized (ITC 16, Instrutech Corp., Elmont, NY), stored on hard disc, and analyzed with Wavemetrics software on a MacIntosh Quadra 650 (Apple Computer, Cupertino, CA). Membrane potentials were corrected offline for liquid-junction potentials (Neher, 1992). Since voltage errors due to series resistances were below 5 mV, they were not corrected. The capacitance was corrected before measurements. Patch pipettes were prepared from Kimax-51 glass (Kimble products, Vineland, NY), coated with silicone (Sylgard 184 silicone elastomer kit, Dow Corning, Corning, NY), and heat polished.

From a holding potential of 52 mV, consecutive voltage pulses of 2 s in duration were applied in 20-mV steps from 132 to 88 mV. For deviations see figure legends.

Relative Permeability

Permeability ratios for malate² and citrate³ with respect to Cl (P_{org. anion}/P_Cl) were calculated according to the Goldman-Hodgkin-Katz equation modified after Fatt and Ginsborg (1958) as described earlier by Dietrich and Hedrich (1998).

Solutions

The standard pipette solution replacing the cytoplasm consisted of (in millimoles) 100 TEACl, 2 MgCl₂, 2 MgATP, 2 EGTA, and 10 HEPES [ 4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid]/Tris (pH 7.2). The standard bath solution (control) contained (in millimoles) 20 TEACl, 1 CaCl₂, and 5 MES/Tris (pH 4.3). For Al-containing solutions AlCl₃ was added at a concentration of 50 µM (33 µM Al_mono). All solutions were adjusted to an osmolality of 600 mosmol kg¹ (pipette: 620 mosmol kg¹) with d-sorbitol, and were verified by a water-vapor osmometer (5100C, Wescor, Logan, UT). Changes in the composition of solutions are indicated in the text. The reference electrode was filled with 3 M KCl and a plug containing 3 M KCl in 2% (w/v) agar preventing salt leakage into the bathing solution. Bath solutions were kept on ice, and were continuously perfused at a rate of about 0.1 mL min¹.