A Calcium-Selective Channel from Root-Tip Endomembranes of Garden Cress

Abstract

A Ca²⁺ channel from root-tip endomembranes of garden cress (Lepidium sativum L.) (LCC1) was characterized using the planar lipid-bilayer technique. Investigation of single-channel recordings revealed that LCC1 is voltage gated and strongly rectifying. In symmetrical 50 mm CaCl₂ solutions, the single-channel conductance was 24 picosiemens. LCC1 showed a moderate selectivity for Ca²⁺ over K⁺ (9.4:1) and was permeable for a range of divalent cations (Ca²⁺, Ba²⁺, and Sr²⁺). In contrast to Bryonia dioica Ca²⁺ channel 1, a Ca²⁺-selective channel from the endoplasmic reticulum of touch-sensitive tendrils, LCC1 showed no bursting channel activity and had a low open probability and mean open time (2.83 ms at 50 mV). Inhibitor studies demonstrated that LCC1 is blocked by micromolar concentrations of erythrosin B (inhibitor concentration for 50% inhibition [IC₅₀] = 1.8 μm) and the trivalent cations La³⁺ (IC₅₀ = 5 μm) and Gd³⁺ (IC₅₀ = 10 μm), whereas verapamil showed no blocking effect. LCC1 may play an important role in the regulation of the cytoplasmic free Ca²⁺ concentration in root-tip and/or root-cap cells. The question of whether this ion channel is part of the gravitropic signal transduction pathway deserves further investigation.

Changes in cytosolic free Ca²⁺ concentration play an important role in plant signal transduction pathways in response to stimuli such as gravity and mechanical forces. To generate a Ca²⁺ signal, the cell may use intracellular and/or extracellular pools of Ca²⁺. For example, it has been demonstrated that plant cells respond to touch with transient increases in cytoplasmic Ca²⁺ concentration (Knight et al., 1991, 1993; Haley et al., 1995) and that this Ca²⁺ originates from some intracellular store rather than from the apoplast (Haley et al., 1995). In contrast, for the graviresponse, most data suggest that apoplastic Ca²⁺ may be critical (for review, see Konings, 1995; White, 1998). Thus, plant cells may use different stores of Ca²⁺ and generate locally confined, transient fluctuations in cellular Ca²⁺ levels to encode different signals. It is less clear how such differential Ca²⁺ signaling is brought about and how Ca²⁺ signaling is regulated by the cell. Thus, there is an urgent need to identify the transport components of the Ca²⁺-signaling machinery in plant tissues at the molecular level before proceeding to functional studies.

A detailed knowledge of these processes will require the molecular identification and analysis of Ca²⁺-transport systems located in different cell membranes; this analysis has just begun. In the plasma membrane of root cells from rye and wheat, two types of Ca²⁺-conductive channels have been detected using the planar lipid-bilayer technique: a maxi-cation channel, which is rather unselective and shows highest conductance and permeability for K⁺, and voltage-dependent cation channel 2, which has lower unitary conductance and also transports a range of cations, including K⁺ and Ca²⁺ (White, 1993, 1994; Pineros and Tester, 1995; for review, see White, 1998). Patch-clamp experiments demonstrated that vacuolar membranes from sugar beet tap roots contain voltage-dependent Ca²⁺ channels (Johannes et al., 1992; Pantoja et al., 1992; Gelli and Blumwald, 1993). The ion channel described by Johannes et al. (1992) activates at negative membrane potentials and is insensitive to inositol-1,4,5-trisphosphate and the cytoplasmic Ca²⁺ concentration. A very similar channel has been shown to reside in the tonoplast of fava bean guard cells (Allen and Sanders, 1994). Stomatal guard cells also harbor the so-called slow vacuolar ion channel, which is thought to be involved in Ca²⁺-induced Ca²⁺ release (Ward and Schroeder, 1994).

It was shown by Klüsener et al. (1995) that the planar lipid-bilayer technique (Mueller et al., 1962) is suitable to identify and characterize Ca²⁺-selective ER channels from plants. Their study revealed the first such channel from the ER of mechanosensitive Bryonia dioica tendrils, BCC1, and led to a hypothesis on how this channel might be integrated in the mechanotransduction process (Klüsener et al., 1995, 1997). The identification of BCC1 was facilitated greatly by the fact that tendrils have a profuse ER that can be isolated readily (Liss and Weiler, 1994; Klüsener et al., 1995) and the fact that the channel density at the ER of this organ, geared to perceive and transduce a mechanical stimulus, apparently was quite high, allowing a high-yield reconstitution of BCC1 in planar lipid bilayers.

To study Ca²⁺ channels in tissue specialized to perceive gravistimulation and mechanical force, root tips seem most appropriate. A technique has been developed (Buckhout et al., 1982) that allows the large-scale preparation of root tips (including root caps) of garden cress (Lepidium sativum L.), which is a close relative of Arabidopsis, a species that is excluded from preparative studies because of the small amounts of root tip tissue that can be prepared. Ion channels identified in garden cress should have close relatives in Arabidopsis, thus permitting molecular genetic studies.

Using garden cress root-tip tissue, we isolated enriched ER membranes, developed a technique to incorporate ion channels from this membrane preparation into planar lipid bilayers, and succeeded in identifying electrophysiologically a strongly rectifying, Ca²⁺-selective channel, LCC1, which could be a component of Ca²⁺ homeostasis and/or Ca²⁺ signaling in root tips.

MATERIALS AND METHODS

Plant Material

Seeds of garden cress (Lepidium sativum L. cv Krause) were soaked for 30 min in water and then spread over a wire-mesh rack (Buckhout et al., 1982). After growing for 48 h in the dark, the roots were harvested and stored in an ice-cold 50 mm Suc solution (50 mm Suc and 25 mm Hepes-KOH, pH 7.1) for further preparation.

Preparation Techniques

Isolation of Root Tips

Roots were cut in a blender (Waring) for 5 s twice and mixed with a 20% Percoll solution. The solution was then centrifuged for 10 min at 5000g at 4°C. The resulting pellet consisted of the root tips, which, because of their high amount of amyloplasts, have a higher specific density than the other root tissues. To remove the Percoll from the root tips, the pellet was washed two times with transport buffer (250 mm Suc, 6 mm MgSO₄, and 25 mm Hepes-KOH, pH 7.2).

Preparation of Enriched ER

ER vesicles from root tips were prepared using the density-shift technique, which was described in detail by Liss and Weiler (1994). The root tips were ground under liquid N₂ with a mortar and pestle and then with 5 mL g⁻¹ fresh mass of homogenization medium (50 mm Hepes-KOH, 3 mm DTT, 10% [w/w] Suc, 2 mm EGTA, 0.6% [w/v] insoluble PVP, and 1 mm PMSF, pH 7.5). The homogenate was passed through gauze and the tissue was reextracted once with the same buffer. Afterward, the extract was centrifuged for 10 min at 10,000g at 4°C in a rotor (model SS-34, Sorvall). The supernatant was diluted with one-half of its volume of transport buffer (250 mm Suc, 6 mm MgSO₄, and 25 mm Hepes-KOH, pH 7.2) and centrifuged for 55 min at 100,000g at 4°C in a rotor (model T8.65, Kontron Instruments, Eching, Germany). The resulting pellet consisted of microsomal membranes and was resuspended in 5 mm Hepes-KOH (pH 7.1) containing 6% (w/w) Suc, 1 mm DTT, 3 mm MgSO₄, 0.5 mm PMSF, and 50 μg mL⁻¹ chymostatin. Two to three milliliters of this suspension was layered on top of a Suc step gradient consisting of 6 mL of 50%, 9 mL of 40%, 10 mL of 30%, and 5 mL of 20% (w/w) Suc layers in 5 mm Hepes-KOH, 1 mm DTT, and 3 mm MgSO₄, pH 7.1, and was then centrifuged in a swinging-bucket rotor (model TST 28.38, Kontron) for 2.5 h at 110,000g at 4°C. The fractions from 31% to 40% Suc contained the rER contaminated with plasma membrane, tonoplast, and broken mitochondria. From this material, the rER was further enriched by an EDTA-dependent density-shift technique, as follows. The pooled fractions from a 31% to 40% Suc interface from two gradients were diluted with 1 volume of 5 mm Hepes-KOH (pH 7.1) containing 6% Suc, 1 mm DTT, and 3 mm EDTA and centrifuged. The sediments were resuspended in the same buffer and layered as a 2-mL aliquot on top of a Suc gradient prepared as described above, but with 3 mm MgSO₄ replaced by 3 mm EDTA. The sample was centrifuged (110,000g, 2.5 h, 4°C, TST 28.38 rotor), the ER shifted to a density range of 1.085 to 1.127 g cm⁻³ (21%–30% Suc) because of the loss of ribosomes and thus separated from contaminating membranes. The shifted ER fraction was collected, diluted with 1 volume of transport buffer, and repelleted at 100,000g.

Enzyme Assays

The activity of the NADH-Cyt c reductase (antimycin A insensitive) was determined according to the methods of Moore and Proudlove (1983) and Lord (1983). The vanadate-sensitive, K⁺-stimulated, and Mg²⁺-dependent ATPase was determined according to the method of Hodges and Leonard (1974) with phosphate determinations, as described by Lanzetta et al. (1979). The NO₃⁻-sensitive ATPase was determined as described by Gräf and Weiler (1989), and succinate dehydrogenase was determined according to the method of Singer et al. (1973).

Miscellaneous Assays

Protein was assayed according to the method of Bradford (1976). The density of the fractions collected from density gradients was measured by refractometry.

Electrophysiological Techniques

Electrophysiological experiments were carried out exactly as described previously (Klüsener et al., 1995, 1997). Planar lipid bilayers were formed from a solution of 80 parts (w/w) 1-palmitoyl-2-oleoyl-glycero-3-phosphatidylcholine and 20 parts (w/w) 1,2-dioleoyl-glycero-3-phosphatidylethanolamine (Avanti Polar Lipids, Inc., Alabaster, AL) dissolved in n-decane (15 mg mL⁻¹). We used self-made Perspex cuvettes, and the hole on which the bilayer was painted was 0.1 mm in diameter. The sign of the membrane voltage refers to the cis compartment with respect to the grounded trans compartment. A positive current (upward deflections) therefore corresponds to a cation transfer from the cis to the trans compartment. ER vesicles were always added to the cis compartment. All experiments were carried out under voltage-clamp conditions using a current amplifier (model BLM-120, Biologic, Echirolles, France). The amplifier signal was filtered with a low-pass, linearized, five-pole Tchebicheff filter (Biologic), at a corner frequency of 1 kHz and recorded continuously on a digital audiotape recorder (model DTR-1204, Biologic). For data evaluation on a Power Macintosh 4400/200 computer, the recorded signals were digitized with a sample rate of 10 kHz using an ITC-16 computer interface (Instrutech Co., Great Neck, NY) and Pulse software (HEKA Electronic, Lambrecht/Pfalz, Germany). Single-channel current amplitudes and kinetic properties were analyzed with TAC software (Instrutech Co.), which uses the 50% threshold method for the detection of signals (Colquhoun and Sigworth, 1983).

RESULTS

Suc-Density Step-Gradient Analysis

ER-enriched membrane fractions of root tips from garden cress were prepared by a two-step process described in Methods. In this process, microsomes were first separated on a Suc gradient containing Mg²⁺ in the absence of EDTA. The distribution of marker enzyme activities is summarized in Table I. Fraction C/Mg²⁺ (corresponding to a density range of 1.132–1.176 g cm⁻³), which contained the rER and contaminations with plasma membrane, tonoplast, and broken mitochondria, was collected, washed, and rerun on a second step gradient containing EDTA in the absence of Mg²⁺. Because of the loss of ribosomes, the ER was shifted to a density range of 1.085 to 1.127 g cm⁻³ (21%–30% Suc, corresponding to fraction B/EDTA in Table I), where it could be collected in an enriched form with clearly reduced contaminations of plasmalemma and mitochondrial membranes. However, it was not possible to reduce tonoplast contaminations, most likely because membrane fusions took place. Although the ER is enriched in fraction B/EDTA, we cannot exclude that the ion channel we characterized (LCC1; see below) originates from some other cellular membrane. To analyze this possibility further, we performed reconstitution experiments with fraction C/EDTA, which consisted mainly of plasmalemma and mitochondrial membranes. LCC1 activity was not observed in this fraction. Instead of LCC1, we found a voltage-dependent ion channel with a very fast time-dependent inactivation kinetic and a single-channel conductance of 124 pS in a 100 mm KCl solution (data not shown). This channel was absent from the enriched ER used to characterize LCC1. Therefore, one can assume that LCC1 originates from neither plasmalemma nor mitochondrial membranes.

Table I.

Distribution of marker enzyme activities in a Mg²⁺ and a subsequent EDTA gradient

Enzyme	Specific Activitya								Ratio of B/EDTA:C/Mg2+
	A/Mg2+	B/Mg2+	C/Mg2+	D/Mg2+	A/EDTA	B/EDTA	C/EDTA	D/EDTA
	nkat mg−1 protein
NADH-Cyt c-reductaseb	0.69	0.12	0.46	1.19	0.13	0.68	0.48	0.45	1.48
Succinate dehydrogenasec	16.14	7.22	15.99	160.60	0.80	5.93	31.10	51.91	0.37
VO43−-inhibited ATPased	0.17	0.43	2.22	2.19	0.08	0.76	1.05	2.19	0.34
NO3−-inhibited ATPasee	0.22	0.22	0.48	0.27	0.14	0.47	0.34	0.27	0.98

Shown are the mean values of three gradients.

^aDensity zones: A, 1.059 to 1.077 g cm⁻³ (15%–19% Suc); B, 1.085 to 1.127 g cm⁻³ (21%–30% Suc); C, 1.132 to 1.176 g cm⁻³ (31%–40% Suc); and D, 1.182 to 1.224 g cm⁻³ (41%–49% Suc). 

^bAntimycin A-insensitive ER marker enzyme. 

^cMitochondrial marker enzyme. 

^dPlasma membrane marker enzyme. 

^eTonoplast marker enzyme. 

Electrophysiology

Enriched ER membrane vesicles were incorporated into artifical planar lipid bilayers. In the presence of divalent-cation chloride solutions, single-channel fluctuations could be observed if a sufficiently high voltage was applied to the membrane. Control pure lipid bilayers did not show any current fluctuations within the range of experimental parameters used in this study. Figure 1A shows typical current fluctuation traces of the root-tip LCC1 in a symmetrical 50 mm CaCl₂ solution. The open channel exhibited an ohmic current/voltage relationship with a single-channel conductance of 24 pS (Fig. 1B). All current/voltage curves shown in this paper are derived from single-channel recordings at positive membrane potentials. The ion channel is permeable for a range of divalent cations, with single-channel conductances declining in the order Ca²⁺ (24 pS) ≈ Ba²⁺ (21 pS) > Sr²⁺ (16 pS).

Figure 1.

A, Typical current-fluctuation traces of LCC1 at various applied voltages. Electrolyte solution (cis and trans): 50 mm CaCl₂ and 10 mm Hepes-KOH, pH 7.0. B, Current/voltage relationship of the ER Ca²⁺ channel in a symmetrical 50 mm CaCl₂ solution. Single-channel conductance = 24 pS, n = 5.

The Ca²⁺:K⁺ permeability ratio of LCC1 was determined by biionic potential measurements (Hille, 1992). Replacing the 50 mm CaCl₂ solution in the trans compartment with a 100 mm KCl solution shifted the reversal potential to −30.20 mV (Fig. 2A). From the general assumptions of constant field theory and assuming no difference between internal and external surface potential, the selectivity coefficient is described by the following equation (Lee and Tsien, 1984):

where V_rev. = reversal potential, F = Faraday's constant, R = gas constant, and T = absolute temperature (T = 298 K). If we take the molar ion activities for 50 mm CaCl₂ (a_Ca²⁺ = 29.5 mm) and 100 mm KCl (a_K⁺ = 77.1 mm) into account, the permeability ratio P_Ca²⁺:P_K⁺ is 9.4:1. Figure 2B shows typical single-channel recordings of the ER Ca²⁺ channel under biionic conditions. Even at zero applied voltage, current fluctuations are visible.

Figure 2.

A, Reversal potential of LCC1 under bi-ionic conditions (n = 6). Reversal potential was determined by extrapolation of the single-channel current/voltage curve to the zero-current axis. B, Typical single-channel current-fluctuation traces at various applied voltages. Salt concentrations: cis, 50 mm CaCl₂ and 10 mm Hepes-KOH, pH 7.0; trans, 100 mm KCl and 10 mm Hepes-KOH, pH 7.0.

The fact that LCC1 conducts Ca²⁺ and not Cl⁻ is emphasized by the data shown in Figure 3. Reducing the Cl⁻ concentration of the electrolyte solution by replacing CaCl₂ with Ca(NO₃)₂ or hemicalcium gluconate changed neither the conductance value (Fig. 3) nor the kinetic behavior of LCC1 (data not shown).

Figure 3.

Current/voltage relationships of the ER Ca²⁺ channel in Ca²⁺ solutions with reduced Cl⁻ concentrations. Electrolyte solutions: ○, 50 mm CaCl₂ and 10 mm Hepes-KOH, pH 7.0; ▵, 80 mm hemicalcium gluconate, 10 mm CaCl₂, and 10 mm Hepes-KOH, pH 7.0; and □, 40 mm Ca(NO₃)₂, 10 mm CaCl₂, and 10 mm Hepes-KOH, pH 7.0.

The open-channel conductance of LCC1 was dependent on the Ca²⁺ concentration of the electrolyte solution. In Figure 4 the open-channel conductance is plotted against the Ca²⁺ concentration. Data points can be fitted to a simple Michaelis-Menten-type hyperbola. At a Ca²⁺ concentration of K_m = 7.3 mm, the half-maximum conductance was achieved.

Figure 4.

Ca²⁺ concentration dependence of LCC1 open-channel conductance. Data from at least three bilayer recordings were averaged and fitted to a simple Michaelis-Menten-type hyperbola, with maximum conductance (Λ) of 26.79 pS and K_m of 7.27 mm.

LCC1 had strong rectifying properties (Fig. 5). Depending on the orientation of the channel protein in the bilayer, channel fluctuations were observed only at positive or negative membrane potentials. Switching the membrane polarity completely and instantaneously rendered the ion channel silent. After a return to the original membrane polarity, a rapid recovery to the previous channel activity took place. The voltage dependence of LCC1 is also demonstrated in Figures 6 and 7. With increasing membrane voltages, the mean closed time of the ion channel was reduced from 248 ms at 20 mV to 16 ms at 90 mV (Fig. 6). This dependence could be fitted by an exponential function. In contrast to the mean closed time, the mean open time exhibited no voltage dependence. The dependence of open-state probability (P_o) on membrane potential is shown in Figure 7. For the fitting of the data points, the following equation was used:

where a = α × F/R × T (F = Faraday's constant, R = gas constant, and T = absolute temperature [298 K]), V = applied voltage, V_½ = +48.59 mV, and P_omax = 0.27. We set P_omin at 0. The fit gives, for the constant factor a in the exponential function, a value of 66.24 V⁻¹, which means a formal gating charge of α = 1.7.

Figure 5.

Rectifying properties of LCC1. At the times shown (arrows), the membrane voltage was switched between +50 and −50 mV. Channel activity in this case was seen only when positive voltage was applied to the membrane.

Figure 6.

Kinetic properties of the ER Ca²⁺ channel. The mean open (t_o) and closed times (t_c) of LCC1 are plotted against the membrane voltage. Data from at least five different experiments are shown with their se values. Electrolyte solution (cis and trans): 50 mm CaCl₂ and 10 mm Hepes-KOH, pH 7.0.

Figure 7.

Voltage dependence of the open-state probability (P_o). Experimental conditions are the same as those described for Figure 6.

Inhibitor studies revealed that the root tip Ca²⁺ channel was blocked by micromolar concentrations of the Ca²⁺-ATPase inhibitor erythrosin B. In Figure 8A, the dose-response relationship for the channel blockade by erythrosin B is depicted. Data points were fitted to a Hill function of the following equation:

where block_max = 100%, K_m = IC₅₀, and ν = exponent of the Hill function. For the channel blockade by erythrosin B, an IC₅₀ value of 1.8 μm was determined. The exponent of the Hill function was 1.13, indicating a 1:1 stoichiometry for the binding of erythrosin B.

Figure 8.

Inhibitors of LCC1 activity. A, Dose-response relationship (n = 3) for the channel blockade by erythrosin B. Block (%) is defined as 100 × (1 − P_oa/P_ob), where P_oa and P_ob are the single-channel open-state probabilities after and before, respectively, the addition of erythrosin B to the trans compartment. B, Effect of Cu²⁺ and La³⁺ on the single-channel properties of LCC1. Shown are current recordings of the ER Ca²⁺ channel before and after the addition of Cu²⁺ and La³⁺ to the cis compartment. Electrolyte solution: 50 mm CaCl₂ and 10 mm Hepes, pH 7.0. The membrane voltage was 50 mV.

Some divalent and trivalent cations also blocked LCC1 in low concentrations: Cu²⁺ (IC₅₀ = 5 μm), La³⁺ (IC₅₀ = 5 μm), and Gd³⁺ (IC₅₀ = 10 μm). Figure 8B shows the inhibitory effect of Cu²⁺ and La³⁺ on the ER Ca²⁺ channel. With 10 μm Cu²⁺ or La³⁺ at the Ca²⁺-entry side of the channel protein, LCC1 was nearly completely blocked. The phenylalkylamine verapamil, a typical blocker of L-type Ca²⁺ channels, did not inhibit LCC1 activity (data not shown).

DISCUSSION

An understanding of Ca²⁺ homeostasis and Ca²⁺ signaling requires the molecular identification and analysis of plasmalemma and endomembrane ion channels. Whereas the plasmalemma and the tonoplast can be studied by patch-clamp analysis, this is impossible for endomembranes such as the ER. In these cases, the lipid-bilayer technique (Mueller et al., 1962) seems to be the method of choice (Klüsener et al., 1995, 1997). The ER acts as a Ca²⁺ store (for review, see Meldolesi and Pozzan, 1998) and is loaded through the action of primary active Ca²⁺ ATPases (Buckhout, 1983; Chen et al., 1993; Liss and Weiler, 1994). Free Ca²⁺ concentrations in the ER may reach 5 to 50 mm; however, to our knowledge, there are no data available for plant cells (Meldolesi and Pozzan, 1998). Release of Ca²⁺ from the ER requires the presence of Ca²⁺-release channels, one of which, BCC1, has recently been described and characterized electrophysiologically (Klüsener et al., 1995, 1997). BCC1 is a strongly rectifying, Ca²⁺-selective channel, and its activity is regulated through the transmembrane electrical potential, the transmembrane chemical potential of Ca²⁺ (Klüsener et al., 1995), and the cytosolic pH (Klüsener et al., 1997). BCC1 was found in mechanosensitive tendril tissue of B. dioica.

The involvement of Ca²⁺ in plant mechanotransduction is well documented (Braam and Davis, 1990; Knight et al., 1991, 1993; Haley et al., 1995), and release of Ca²⁺ induced by mechanical force takes place from intracellular stores (Haley et al., 1995). Root tips are subject to mechanical forces in the soil, and they respond to mass acceleration to direct gravitropic orientation of the organ. Both processes are dependent on Ca²⁺ (for review, see Konings, 1995) but in a different, incompletely understood manner. Intracellular Ca²⁺ seems more important for root mechanotransduction (Legué et al., 1997), whereas root gravitropism seems more critically dependent on extracellular Ca²⁺ (Lee et al., 1983a, 1983b; Ishikawa and Evans, 1992). Whereas root plasmalemma Ca²⁺ channels have been characterized (White, 1993, 1994; Pineros and Tester, 1995; for review, see White, 1998), this is the first report, to our knowledge, to describe the identification and characterization of a Ca²⁺-selective channel from endomembranes of the root tip of a higher plant. LCC1 displays a moderate Ca²⁺ selectivity, is voltage dependent, and is strongly rectifying (Figs. 2 and 5).

The exact intracellular location of LCC1 needs to be shown. LCC1 was not observed in fractions from the Suc step gradient enriched for mitochondrial membranes or plasmalemma. Likewise, a prominent channel in these fractions was absent from the enriched ER fraction. Thus, LCC1 resides either in the rER (the dominant membrane in our enriched ER preparation) or in the tonoplast, which is present to some extent in this preparation.

Voltage-dependent vacuolar Ca²⁺ channels have been identified in the root storage tissue of sugar beet and the guard cells of broad bean. Based on the single-channel properties of these ion channels (for review, see Pineros and Tester, 1997), it cannot be excluded that LCC1 originates from the tonoplast membrane. For example, LCC1 is comparable to the voltage-dependent Ca²⁺ channel from sugar beet described by Johannes et al. (1992) with regard to its single-channel conductance (11 pS at 5 mm CaCl₂) and its formal gating charge (α = 1.7 and 1.4, respectively), but it differs from it by its lower Ca²⁺:K⁺ permeability ratio. Finally, to determine the membrane origin of LCC1, it will be necessary to analyze tonoplast ion channels from cress root tips in patch-clamp experiments.

Although the bilayer technique does not allow us to determine the native channel orientation, it is reasonable to assume that LCC1 releases Ca²⁺, which is stored in the endomembrane compartment that harbors LCC1, into the cytosol. The dependence of unitary conductance on the Ca²⁺ concentration (K_m = 7.3 mm) covers the range of Ca²⁺ concentrations reported for the ER lumen (for review, see Meldolesi and Pozzan, 1998) and the vacuole (Gilroy et al., 1993). This further supports the proposed orientation of LCC1.

Although LCC1 shows some similarities to BCC1, a voltage-dependent Ca²⁺ channel from the ER of mechanosensitive tendrils (Klüsener et al., 1995, 1997), it clearly differs from it in its kinetic and some of its pharmacological properties. Whereas BCC1 shows a bursting channel activity, no such patterned current fluctuations could be observed for LCC1. Whereas the ER Ca²⁺ channel from tendrils exhibits at least two open states (Klüsener et al., 1995), only one open state, with a mean open time of 2.83 ms, was determined for LCC1. Therefore, the gating mechanism of the ER Ca²⁺ channel from mechanosensitive tendrils seems to be much more complex and complicated than that of the root-tip Ca²⁺ channel. Like BCC1, LCC1 is blocked by micromolar concentrations of the divalent cation Cu²⁺ and the lanthanide ions La³⁺ and Gd³⁺. On the other hand, the phenylalkylamine verapamil, which blocks a range of plant Ca²⁺ channels such as BCC1 from tendrils and voltage-dependent cation channel 2 from the plasma membrane of wheat and rye roots (White, 1998), did not show any effect on LCC1 in the tested concentrations (10–100 μm). The Ca²⁺-ATPase inhibitor erythrosin B has a very strong blocking effect on LCC1. It has been shown that the Ca²⁺-ATPases from animal sarcoplasmic reticulum can operate as Ca²⁺ channels if the catalytic and transmembrane domains of the ATPase are uncoupled (De Meis et al., 1996). Such an uncoupling of the two domains could easily occur during the membrane-preparation procedure. Therefore, it will be important to determine whether the protein we characterized in planar lipid-bilayer experiments is an ion channel in the narrow sense or is formed by the channel-like transmembrane domain of a Ca²⁺-ATPase.

LCC1 could be involved in Ca²⁺ homeostasis and/or Ca²⁺ signaling. Its precise role(s) remains to be determined. Thus, further experiments with the aim of cloning root-tip Ca²⁺ channels of higher plants are required.

Abbreviations:

BCC1

Bryonia dioica Ca²⁺ channel 1

IC₅₀

inhibitor concentration for 50% inhibition

LCC1

Lepidium sativum Ca²⁺ channel 1

pS

picosiemens

rER

rough ER