Calcium is a critical structural and regulatory nutrient in plants. However, mechanisms of its uptake by root cells are poorly understood. We have found that Ca2+ influx in Arabidopsis root epidermal protoplasts is mediated by voltage-independent rapidly activating Ca2+-permeable non-selective cation channels (NSCCs). NSCCs showed the following permeability (P) sequence: PCa (1.00) = PBa (0.93) > PZn (0.51), PCa/PNa = 0.19, PCa/PK = 0.14. They were inhibited by quinine, Gd3+, La3+ and the His modifier diethylpyrocarbonate, but not by the Ca2+ or K+ channel antagonists, verapamil and tetraethylammonium (TEA+). Single channel conductance measured in 20 mm external Ca2+ was 5.9 pS. Calcium-permeable NSCCs co-existed with hyperpolarisation-activated Ca2+ channels (HACCs), which activated 40–60 min after forming the whole-cell configuration. HACCs activated at voltages <−130 to −150 mV, showed slow activation kinetics and were regulated by cytosolic Ca2+ ([Ca2+]cyt). Using aequorin-expressing plants, a linear relationship between membrane potential (Vm) and resting [Ca2+]cyt was observed, indicating the involvement of NSCCs. Intact root 45Ca2+ influx was reduced by Gd3+ (NSCC blocker) but was verapamil and TEA+ insensitive. In the root elongation zone, both root net Ca2+ influx (measured by Ca2+-selective vibrating microelectrode) and NSCC activity were increased compared to the mature epidermis, suggesting the involvement of NSCC in growth. A Ca2+ acquisition system based on NSCC and HACC co-existence is proposed. In mature epidermal cells, NSCC-mediated Ca2+ influx dominates whereas in specialised root cells (root hairs and elongation zone cells) where elevated [Ca2+]cyt activates HACCs, HACC-mediated Ca2+ influx predominates.
Here, it is shown that Arabidopsis root epidermis NSCCs can mediate steady-state Ca2+ influx over the Vm range at which activity of HACCs may be restricted. Critically, activity of NSCCs was found to be greatest in the elongation zone epidermis where net Ca2+ influx was highest. NSCCs co-localise with [Ca2+]cyt-regulated HACCs in the epidermis, providing a dynamic Ca2+ influx pathway for growth.
Whole-cell Ca2+ currents through non-selective cation channels in the plasma membrane of protoplasts from the root mature epidermis
Characteristic properties of Arabidopsis root NSCCs are low selectivity between monovalent cations, rapid activation, weak voltage dependence, insensitivity to K+ and Ca2+ channel blockers, inhibition by lanthanides, quinine and the His modifier diethylpyrocarbonate (DEPC) (Demidchik and Tester, 2002). In all protoplasts derived from mature epidermis, rapidly activating, weakly voltage-dependent Ca2+ influx currents (ICa) were found (Figure 1a,b) (currents were recorded 10–30 min after establishing the whole cell configuration). Decrease of external Ca2+ to 0.1-mm abolished inward currents (in the latter condition, mean ± SE ICa was −2.3 ± 0.1 pA at Vm = −160 mV; n = 24), demonstrating that these currents were due to Ca2+ influx. Increased external Ca2+ shifted the reversal potential to more positive values (Figure 1b, inset) according to the shift in the Ca2+ equilibrium potential (ECa) (from mean ± SE 32 ± 4.6 mV at 2 mm Ca2+ to 78 ± 8.5 mV at 50 mm Ca2+n = 10).
Inward calcium currents slightly rectified at Vm < −120 mV in 55% of protoplasts (recorded at different bath Ca2+ concentrations, total n = 64), and this rectifying component of ICa was sensitive to 20 µm verapamil (Figure 1b). This verapamil sensitivity is indicative of HACC activity (Kiegle et al., 2000). In 45% of protoplasts, ICa was linear and insensitive to verapamil. Further pharmacological analyses demonstrated that in 100% of protoplasts tested (n = 29), ICa was not inhibited by the K+ channel blocker, TEA+ (tetraethylammonium, 1–5 mm; Figure 1c). The currents were inhibited by quinine (1 mm; five protoplasts out of 11), Gd3+ or La3+ (0.1 mm; all protoplasts tested; n = 30) and long-term (1 h) exposure to DEPC (0.1 mm; in eight protoplasts from 13) (see Figure 1c,d). The effects of quinine, Gd3+ and La3+ were reversible on washout, while DEPC changed ICa irreversibly (data not shown). These properties show that the dominant linear component of ICa was mediated by the Arabidopsis root NSCCs that were previously characterised as a Na+ influx pathway (Demidchik and Tester, 2002). Indeed, the substitution of Ca2+ with Na+ in the external solution did not change the main characteristics of the NSCC-mediated currents; they remained rapidly activating and weakly voltage dependent (Figure 1a). The amplitude of Na+ currents was significantly larger (Figure 1b) indicating greater permeability to Na+ than to Ca2+ and confirming the capacity of NSCC to conduct both Ca2+ and Na+.
In multicationic conditions, the analysis of ion channel selectivity based on reversal potential (Erev) measurements can be problematic (Demidchik et al., 2002; Pottosin et al., 2001). Estimates of cation selectivity were therefore based on values of the relative conductance calculated from inward currents at voltages between Erev and −160 mV using protoplasts with no inward-rectifying verapamil-sensitive currents. Measurements were performed in an external cation activity of 50 mm (Cl– salt). This yielded the following permeability sequence for divalent cations: PCa (1.00) = PBa (0.93 ± 0.21) > PZn (0.51 ± 0.15) (n = 5), PCa/PNa = 0.19 ± 0.01 (n = 14), PCa/PK = 0.14 ± 0.01 (n = 14) (note: Na+ and K+ currents were measured at 0.1 mm external CaCl2, since some Ca2+ was needed to maintain seal stability; values are mean ± SE).
It should be noted that in 38% of protoplasts (n = 60), a DACC-like conductance was evident immediately following formation of the whole cell configuration. This conductance significantly varied in individual protoplasts, was unstable, showed fast ‘run down’ and was not evident in the 10–30-min observation period used for NSCC characterisation. The peak amplitude of the DACC-like conductance was observed at −74 ± 4.5 mV (−6.4 ± 0.3 pA, 20 mm external Ca2+ mean ± SE, n = 20). This conductance was not characterised further.
Single-channel Ca2+ currents
Single-channel measurements were performed using outside-out patches obtained from mature epidermis protoplasts in conditions directly comparable with those described in the preceding section (Figure 2). In 20 mm external Ca2+, the estimated single channel conductance was 5.9 pA (Figure 2b). The mean extrapolated Erev (+100 mV) was significantly nearer to ECa than to ECl (+161 and −39 mV, respectively) demonstrating that currents were predominantly driven by Ca2+ influx rather than Cl– efflux (Figure 2b). Moreover, lowering the pipette Cl– to 0.2 mm had no effect on Erev (data not shown). The open probability of single channels only slightly increased at more negative voltages (Figure 2c). Addition of 0.1 mm Gd3+ or La3+ caused a (mean ± SE) 91 ± 4.5% decrease of single NSCC open probability over the voltage range tested (n = 8), while 5 mm TEA+ or 20 µm verapamil had no effect.
NSCCs co-exist with HACCs in the root cell plasma membrane
While NSCCs were dominant early on during recordings, 40–60 min after establishing the whole cell configuration, the HACC-like conductance was clearly evident in 51% (n = 110) of protoplasts (Figure 3a–c). The time course of appearance and current–voltage characteristics agreed with the previously described root hair HACC (Véry and Davies, 2000). At physiological [Ca2+]cyt (100 nm), HACC-mediated Ca2+ influx currents activated at Vm negative of −140 mV, thus adding to NSCC currents at more negative Vm. The steep voltage dependence of HACC activity means that at Vm more negative than −160 mV, HACC-mediated currents quantitatively exceeded those through NSCCs. After addition of 20 µm verapamil, activity of NSCCs remained unchanged while HACCs were significantly blocked, thus distinguishing between these two types of cation channels co-existing in the PM. HACCs were previously observed only at the very apex of the Arabidopsis root epidermis (Kiegle et al., 2000) but an extended observation period clearly demonstrated that this type of channel also exists in mature epidermis.
An increase in pipette Ca2+ ([Ca2+]cyt) from 100 nm to 1 µm shifted activation of HACCs to more positive Vm (by +37 ± 5 mV; mean ± SE, n = 7; Figure 3d), in agreement with the activation voltage of the root hair HACC (Véry and Davies, 2000). In contrast, decreasing pipette Ca2+ from 100 to 10 nm and to 1 nm shifted activation to more negative Vm by −28 ± 8 and −62 ± 16 mV, respectively (mean ± SE, n = 5). In most protoplasts, NSCC activity was not significantly changed by changing pipette Ca2+ but at elevated pipette Ca2+ (1 µm) 40% of protoplasts exhibited quicker run-down of NSCCs.
Resting [Ca2+]cyt of mature epidermal cells shows a linear dependence on membrane potential
The co-existence of two Ca2+-permeable influx routes (NSCCs and HACCs) differing in their voltage sensitivity prompted an examination of the effects of Vm on [Ca2+]cyt. If NSCCs were to provide Ca2+ influx into root cells, their voltage dependence should be reflected in changes of resting [Ca2+]cyt as Vm is varied. A physiologically relevant way to manipulate Vm non-invasively is to change external [K+] (Maathuis and Sanders, 1993; Sokolik and Yurin, 1981). It is known that the cytosolic K+ activity in Arabidopsis root epidermal cells is approximately 80 mm, and above 0.1 mm external K+, the steady-state PM Vm is very close to the equilibrium potential for K+, EK (Leigh, 2001; Maathuis and Sanders, 1993). A range of external [K+] was therefore applied to protoplasts derived from mature epidermis of aequorin-transformed plants. A linear dependence of steady-state [Ca2+]cyt on the estimated EK was observed (Figure 4a) under conditions where EK should set the resting Vm (Maathuis and Sanders, 1993). Application of the cation channel blocker Gd3+ (1-h pre-treatment with 0.1 mm) eliminated the effect of external [K+] on steady-state [Ca2+]cyt, suggesting that under control conditions, Vm-dependent changes in [Ca2+]cyt were mediated by Ca2+ influx through Gd3+-sensitive PM channels. Significantly (as demonstrated earlier), 0.1 mm Gd3+ blocks NSCCs in this cell type over the voltage range predicted to be obtained under the experimental conditions used. The root hair HACC is also blocked by submillimolar Gd3+ (Véry and Davies, 2000). However, given the Vm and [Ca2+]cyt dependence of HACCs in the mature epidermis, their contribution to steady-state [Ca2+]cyt was likely to be substantial only at the most negative EK (and hence Vm) applied. At less negative Vm, the relationship between Vm and steady-state [Ca2+]cyt reflects that of voltage on NSCC-mediated Ca2+ influx currents.
NSCCs mediate 45Ca2+ uptake by intact roots
To corroborate the in vivo significance of the data described above, the NSCC-mediated Ca2+ pathway in intact roots was examined using 45Ca2+. In 20-min accumulation assays, the HACC blocker verapamil (20 µm) and K+ channel blocker TEA+ (1 mm) did not inhibit Ca2+ uptake, while 1 mm Gd3+ (the most potent NSCC blocker) inhibited Ca2+ uptake by 14.4 times (Figure 4b). This TEA+ and verapamil insensitivity, coupled with extreme Gd3+ sensitivity, clearly demonstrate that NSCCs are involved in Ca2+ acquisition by Arabidopsis roots.
NSCC-mediated Ca2+ uptake plays a role in root growth
According to Cramer and Jones (1996), Arabidopsis root growth responds linearly to an increase in [Ca2+]cyt of immature apical cells and significantly accelerates with even slight elevation of [Ca2+]cyt. Raising external Ca2+ increased [Ca2+]cyt (Cramer and Jones, 1996). Therefore, it is likely that elongation zone epidermis requires significantly higher Ca2+ influx than its mature form. Accordingly, Ca2+ influx of elongation zone and mature epidermal cells of intact roots was measured using the vibrating Ca2+-selective microelectrode technique (Figure 5). Mean net Ca2+ influx was larger in intact epidermal cells of the elongation zone than in cells of mature epidermis (n = 4). Notably, 0.1 mm Gd3+ suppressed 95% of net Ca2+ influx both in elongation zone and mature epidermis, suggesting that influx was mediated by lanthanide-sensitive pathways such as the NSCCs.
In real soil conditions, the contribution of NSCCs and HACCs to Ca2+ influx would depend on values of the resting plasma membrane Vm. Thus, the observed elevated Ca2+ uptake in elongation zone epidermal cells could originate in more negative Vm values of this tissue. Therefore, Vm of elongation zone epidermal cells was measured in a range of external [K+] (Table 1) to compare with previously reported Vm values from mature epidermal cells (Maathuis and Sanders, 1993). Resting plasma membrane Vm in elongation zone cells did not differ significantly from values previously obtained from the mature epidermis and approximated a linear dependence on extracellular [K+] at concentrations greater than 0.1 mm.
Table 1. Resting plasma membrane Vm measured in epidermis cells of Arabidopsis root elongation zone
Extracellular K+, mm
Resting Vm, mV (±SE)
Assay solution contained 2 mm CaCl2, 1 mm Mes/Tris pH 6.0 and different [K+] (applied as KCl). Data are from 16 plants.
−164 ± 11.9
−156 ± 3.4
−151.9 ± 2.4
−145.5 ± 4.4
−135 ± 3.2
−125.5 ± 6.4
−118.7 ± 5.3
−112 ± 4.6
−88.3 ± 15.8
−85.5 ± 6.1
−49 ± 9.9
Patch clamp studies on the PM of epidermal protoplasts derived from the root elongation zone showed significantly greater NSCC activity in these protoplasts compared to those from mature epidermal cells (Figure 5). Mean NSCC currents recorded approximately 20 min after establishing whole cell configuration were three times larger (at −160 mV) than in mature epidermis under identical conditions. HACCs were also observed in elongation zone epidermal protoplasts and exhibited similar overall characteristics to those of the mature epidermis (current amplitude, kinetics and voltage dependence). However, HACC activity was observed in all elongation zone protoplasts tested (n = 6). Thus, the increased activity of both NSCCs and HACCs appears to underlie the elevated Ca2+ influx in elongation zone epidermal cells required for root growth.
Calcium uptake and root plasma membrane Ca2+ channels
Despite the physiological importance of Ca2+, relatively little is known about the mechanisms and regulation of its uptake by roots. The two main types of root PM Ca2+-permeable channels characterised to date (HACCs and DACCs) differ in their voltage sensitivity and there appears to be a voltage range over which neither type of channel would be particularly efficient at Ca2+ uptake. This voltage range matches that observed in Arabidopsis mature epidermal cells (−101 to −144 mV; Maathuis and Sanders, 1993) and warrants an examination of other channel types that could mediate Ca2+ influx at such moderately negative Vm. In this study, PM non-selective cation channels (NSCCs) have been found to mediate Ca2+ influx over the entire physiological range of Vm, including the moderately negative Vm that would compromise HACC and DACC activity.
Root plasma membrane non-selective cation channels are Ca2+ permeable
Plant PM NSCCs became a subject of broad interest only in the last few years and although significant advances have been made using artificial phospholipid bilayers (Aleksandrov et al., 1976; Lunevsky et al., 1980; White and Tester, 1992; White, 1993, 1994, 2000), information on NSCC properties (particularly their behaviour in intact membranes) is still limited. Although the rapidly activating and voltage-insensitive NSCC found in Nitella flexilis internodes is known to be Ca2+ permeable (Sokolik, 1999), none of the rapidly activated NSCCs that were found in native PM of higher plants to date (Demidchik et al., 2002) have been examined for Ca2+ permeability. Here, the PM of Arabidopsis root mature epidermal cells was always found to contain rapidly activating, weakly voltage-dependent NSCCs that were Ca2+ permeable in both whole cell and single channel patches (Figures 1 and 2). The kinetics and voltage dependence, coupled with the pharmacological profile (lanthanide, quinine and DEPC sensitive; TEA+ and verapamil insensitive, Figure 1) identify these NSCCs with those found previously to mediate Arabidopsis root Na+ influx (Demidchik and Tester, 2002).
NSCCs mediate Ca2+ uptake in roots and are involved in [Ca2+]cyt homeostasis
The Ca2+ permeability of NSCCs and their co-existence with HACCs (Figures 1 and 3) and DACCs in mature epidermal cells appears to render the mature epidermis competent for Ca2+ uptake over an extensive Vm range. Critically, in patch clamp experiments, the NSCCs of the mature epidermis were found to be active across the physiologically relevant Vm range at which HACC- and DACC-mediated Ca2+ influx would be small (−101 to −144 mV; Maathuis and Sanders, 1993). The competence of NSCCs in Ca2+ uptake at such moderately negative Vm was tested here by non-invasively manipulating the Vm of aequorin-expressing mature epidermal protoplasts (Figure 4). The largely linear relationship between the Vm (set by EK) and resting [Ca2+]cyt, plus the significant reduction of [Ca2+]cyt by Gd3+, clearly implicates the Gd3+-sensitive NSCCs in Ca2+ influx and [Ca2+]cyt homeostasis. The resting level of [Ca2+]cyt is maintained by the concerted action of ion channels that allow Ca2+ to enter the cytoplasm from external and internal stores, Ca2+ binding proteins (Anil and Rao, 2001) that buffer cytoplasmic Ca2+ and active transport mechanisms for the removal of Ca2+ from the cytoplasm (Sze et al., 2000). It appears from this study that PM Ca2+-permeable NSCCs are a novel component of Ca2+ homeostasis in mature root cells. Furthermore, whole root 45Ca2+ influx measured in the presence of 2 mm KCl (which should de-polarise Vm to moderately negative values) was strongly inhibited by a high level of Gd3+ but not by the HACC-blocker verapamil. This again implicates NSCCs in Ca2+ uptake.
NSCCs and HACCs may act in concert to enhance Ca2+ uptake
It is clear from the patch clamp studies on mature epidermal cells that NSCCs and HACCs not only co-exist in the same membrane but also may be functionally coupled by [Ca2+]cyt (Figure 3). In contrast to the ubiquitous NSCCs, HACCs were found in 50% of protoplasts derived from the mature epidermis. We propose that this frequency of occurrence reflects the trichoblast/atrichoblast identity of the protoplasts and is critical to the formation of root hairs. The root hair bulges formed by trichoblasts do not exhibit an apex-high [Ca2+]cyt gradient until they enter into the main phase of elongation growth (Wymer et al., 1997). It is already known that in Arabidopsis root hairs, it is an apical PM HACC that mediates Ca2+ influx to maintain the apex-high [Ca2+]cyt gradient that is associated with growth (Véry and Davies, 2000; Wymer et al., 1997). NSCC activity in trichoblast root hair bulges could perhaps provide an initial increase in [Ca2+]cyt which in turn shifts the activation voltage of HACCs to values more positive than the resting Vm (Figure 6). This would allow HACCs to mediate a greater Ca2+ influx and so initiate the [Ca2+]cyt gradient for entry into the elongation phase of hair growth. This scheme envisages precise spatial channel localisation that would be lost on the production of protoplasts.
The special case of the elongation zone
The coupled activity of NSCCs and HACCs takes on even greater significance in epidermal cells of the elongation zone. Arabidopsis apical root cell [Ca2+]cyt is known to correlate positively with root growth (Cramer and Jones, 1996) which strongly suggests that enhanced Ca2+ influx is required for cells undergoing extension growth. Indeed in this study, net Ca2+ influx (measured using a vibrating Ca2+-selective microelectrode) at the elongation zone epidermis was over 20 times larger than that at the mature epidermis (Figure 5). This enhanced Ca2+ uptake was sensitive to Gd3+, a potent blocker of both NSCCs and HACCs. In contrast to the mature epidermis, all elongation zone epidermal protoplasts contained both NSCCs and HACCs, with NSCC activity significantly larger in the elongation zone than in the mature region. Again, it is feasible that NSCC-mediated elevation of [Ca2+]cyt shifts the HACC activation voltage in a positive direction, thus enabling high-capacity HACC-mediated Ca2+ influx for growth. The resting Vm of the elongation zone epidermis (Table 1) was comparable with that of the mature epidermis (Maathuis and Sanders, 1993) in terms of a near-linear dependence on external [K+] above 0.1 mm (although in contrast to the mature epidermis, the response was not close to Nernstian). In normal soil [K+] of 0.2–2 mm (Bergmann, 1992; Leigh, 2001), NSCCs would be competent to mediate Ca2+ influx and their high activity in the elongation zone would augment that of HACC. Increased levels of soil K+ (e.g. through excessive fertiliser application) or the presence of other cations such as Na+ or NH4+ would lead to de-polarised cells, thus severely limiting HACC activity but having relatively less effect on voltage-independent NSCC-mediated Ca2+ influx in this critical cell type.
Voltage modulation of the NSCC/HACC Ca2+ uptake system would be a potent regulator for Ca2+ entry to the root cell cytoplasm and it can be envisaged that a cell's specific PM H+-ATPase complement would play a pivotal role in regulating voltage and hence Ca2+ influx (Miedema et al., 2001).
Ca2+-permeable NSCCs in other tissues
It is feasible that NSCCs are involved in regulating [Ca2+]cyt in other plant organs and specific cell types. Background Ca2+ influx mediated by NSCCs is likely to underlie the [Ca2+]cyt to Vm relationship (approximately linear) at Vm between −50 and −150 mV observed in guard cells (Grabov and Blatt, 1998). In this case, significant non-linear [Ca2+]cyt bursts started only at more negative Vm (−180 to −200 mV; Grabov and Blatt, 1998) and probably corresponded to activity of guard cell PM HACC (Hamilton et al., 2000). This points to the possible ubiquity of Ca2+ influxes mediated by NSCCs in plant cells with different functions.
NSCCs – multiple physiological roles?
NSCCs, co-existing with HACCs, provide a flexible Ca2+ uptake system, adjusting Ca2+ homeostasis to diverse physiological and developmental states. It is clear now that these channels play a broader physiological role than only toxic Na+ uptake (Davenport and Tester, 2000; Demidchik and Tester, 2002; Tyerman and Skerrett, 1999). In addition to the delineation of a role for NSCCs in Ca2+ uptake by roots, the present study has also demonstrated that root NSCCs are permeable to Zn2+. Transport of such divalent cationic micronutrients is a critical physiological process, directly related to quality and quantity of crop yield (Bergmann, 1992; Marschner, 1995). The mechanistic basis of this process remains poorly understood (Welch, 1995; White et al., 2002). That root PM NSCCs are permeable to Zn2+ suggests that, apart from their role in plant cell Ca2+ homeostasis, root Ca2+-permeable NSCCs can mediate the passive influx of divalent cationic micronutrients to the root cell cytoplasm. In this respect, the importance of high activity of NSCCs in elongation zone cells is confirmed.
Arabidopsis thaliana (Heyn) Columbia and C24 were from our laboratory stock. Plants constitutively expressing cytosolic aequorin (driven by the CaMV 35S promoter; Knight et al., 1996) and plants expressing green fluorescent protein (GFP) in root epidermis (line J0481; Kiegle et al., 2000) were kindly provided by Dr M. Knight (University of Oxford) and Dr J. Haseloff (University of Cambridge), respectively. Seedlings were grown aseptically at 22°C for 10–15 days (16-h day length; 100 µmol m−2 sec−1 irradiance) on standard medium comprising 0.3% (w/v) Phytagel (Sigma), full-strength Murashige–Skoog medium (Duchefa, Haarlem, the Netherlands) and 1% (w/v) sucrose. For vibrating Ca2+-selective microelectrode experiments, Columbia seedlings were grown for 3–6 days on standard medium or on wet paper rolls (Shabala et al., 1997). For 45Ca2+ accumulation studies, Columbia seedlings were mounted on perforated polycarbonate discs in groups of 10 and grown on standard medium solidified with 0.8% (w/v) agar. Growth was at 24°C for 16-h day length (75 µmol m−2 sec−1 irradiance) and 16°C at night. Plants were grown for 21 days and then discs were transferred to hydroponics. Plants were grown for 14 days over an aerated solution containing: 0.5 mm KOH, 0.25 mm KH2PO4, 0.75 mm MgSO4, 0.03 mm CaCl2, 0.10 mm FeNaEDTA, 4.0 mm Ca(NO3)2, 0.03 mm H3BO3, 0.01 mm MnSO4, 0.001 mm ZnSO4, 0.003 mm CuSO4 and 0.0005 mm Na2MoO4 (pH 5.6). For examining resting PM potentials, Columbia seedlings were hydroponically grown as described elsewhere (Maathuis and Sanders, 1996) on 25% strength Murashige–Skoog medium. Growth conditions were: 16-h day length (70 µmol m−2 sec−1 irradiance), 24/20°C (day/night).
Isolation of protoplasts from mature root epidermis
The method was adapted from Demidchik and Tester (2002). Roots from 50 seedlings were cut into approximately 1-mm-long pieces in 4 ml of enzyme solution. This comprised 1.5% (w/v) Cellulase Onozuka RS (Yakult Honsha, Tokyo, Japan), 1% (w/v) cellulysin (CalBiochem, Nottingham, UK), 0.1% (w/v) pectolyase Y-23 (Yakult Honsha, Tokyo, Japan), 0.1% (w/v) bovine serum albumin (Sigma), 10 mm KCl, 10 mm CaCl2, 2 mm MgCl2, 2 mm MES, pH 5.6 with Tris; 290–300 mOsM adjusted with 330 mm sorbitol. After gentle shaking (60 r.p.m.) in the enzyme solution for 30–50 min at 28°C, protoplasts were filtered (50-µm pore mesh) and rinsed with holding solution (HS: 5 mm KCl, 2 mm CaCl2, 1 mm MgCl2, 10 mm sucrose, 10 mm glucose, 2 mm MES, pH 5.7 with Tris; 290–300 mOsM with sorbitol). Protoplasts were collected by 5-min centrifugation at 200 g and approximately 0.5 ml was diluted with 8–10 ml of HS. Most viable protoplasts isolated by this procedure were derived from mature epidermis as judged by direct observation and comparison with protoplast release from roots expressing GFP in the epidermis. All protoplasts used for patch clamping (mean ± SE 20 ± 1.5 µm diameter and uniformly grey) were derived only from mature epidermis.
Isolation of protoplasts from root elongation zone epidermis
Root tips (3–4 mm) were isolated directly in the patch clamp chamber and exposed to the enzyme solution (without shaking) for 1–2 h (22°C, dark). Direct observation confirmed protoplast release solely from elongation zone epidermis. Protoplasts (mean ± SE 20 ± 1.5 µm diameter) were washed with sealing solution (see below) for 5 min and then patch clamped.
Patch clamp electrophysiological protocols were adapted from Demidchik and Tester (2002). Potassium was excluded from experimental solutions in order to delineate Ca2+ currents clearly. Sealing solution comprised: 20 mm CaCl2, 2 mm MES, pH 5.7 with Tris (290–300 mOsM adjusted with sorbitol). After seal formation, assay solutions with different CaCl2 concentrations were applied. The pipette solution contained: 20 mm Na-gluconate, 10 mm NaCl, 2 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (free Ca2+ adjusted to 100 nm with Ca(OH)2), pH 7.2 adjusted with 5 mm HEPES/NaOH. Ion activities were calculated using GEOCHEM (Parker et al., 1995). A standard patch-clamp amplifier (IM/PCA, List, Darmstadt, Germany), Digidata 1200 digitiser and pClamp software, version 6.0.2 (Axon Instruments, Foster City, CA) and 8-pole Bessel filter (Frequency Devices, Haverhill, MA) were used. Liquid junction potentials were measured and corrected as described elsewhere (Ward and Schroeder, 1994). Curve fitting was performed using Statistica (StatSoft, Tulsa, OK) or Sigma Plot (SPSS Science, Chicago).
Membrane potentials were recorded in epidermal cells of the root elongation zone (approximately 0.5–3 mm behind the root apex) as described previously (Maathuis and Sanders, 1996) in assay solution composed of 2 mm CaCl2, 1 mm Mes/Tris pH 6.0; [K+] was varied by addition of KCl.
Computer-operated luminometry techniques were used for recording aequorin luminescence (Knight et al., 1996). Protoplasts from mature epidermis were prepared as described previously. Protoplasts (in HS) were treated with 4 µg ml−1 coelenterazine (free base, NanoLight Technology, Prolume Ltd, Pittsburgh, PA) for 3 h at 28°C (dark, 20 r.p.m.). Protoplasts were washed in recording solution (1 mm CaCl2; osmotic potential and pH were the same as HS; coelenterazine concentration was maintained at 4 µg ml−1) and placed in luminometer cuvettes (0.5-ml aliquots). One hour before recording resting [Ca2+]cyt, various [K+] (added as KCl) were added to each cuvette as required (solution pH and osmotic potential adjusted to that of HS). Luminescence was recorded at 1-Hz sampling for 1 h.
This technique is described in detail elsewhere (Shabala et al., 1997; Shabala, 2000). Each plant was mounted in a Perspex holder by an agar drop and roots were immersed in assay solution comprising 0.1 mm KCl, 0.1 mm NaCl, 0.1 mm MgCl2 and 0.05 mm CaCl2, pH 5.6 with 1 Tris/MES. The microelectrode was placed 20 µm above the root surface. Net Ca2+ flux was measured for 30 min.
Calcium uptake by excised roots
One hour before determining 45Ca2+ uptake, discs holding plants were placed over an aerated solution containing 2 mm KCl and 0.1 mm CaCl2. Immediately prior to experiments, shoots were removed and isotopically labelled Ca (approximately 3 MBq l−1 45Ca2+ NEN Life Science Products, Zavantem, Belgium) was introduced. Calcium (45Ca2+) uptake was assayed over 20 min in the absence or presence of 1 mm GdCl3, 1 mm TEACl or 20 µm verapamil. After 20 min, roots were transferred to a 45Ca2+-free solution containing 2 mm KCl, 0.1 mm CaCl2 and 1 mm LaCl3 for 5 min to remove 45Ca2+ from the cell wall (preliminary experiments demonstrated that this La3+ exposure was sufficient to displace cell wall Ca2+). Roots were separated, blotted, weighed and 45Ca content was determined (Beckman LS 6000TA scintillation counter; Beckman Instruments, Fullerton, CA, USA).
This paper is dedicated to Professor V.M. Yurin and Dr A.I. Sokolik. We thank Dr M. Knight (University of Oxford) for 35S plants and Dr J. Haseloff (University of Cambridge) for line J0481. We also thank Dr Henk Miedema for useful discussions and help with developing the protocol for production of elongation zone protoplasts, John Hammond for growing the plants for calcium influx experiments and John Banfield for technical advice. The work was supported by the Australian Research Council and the BBSRC (8/D13399).