Spatial variation in H2O2 response of Arabidopsis thaliana root epidermal Ca2+ flux and plasma membrane Ca2+ channels

Authors


(fax +44 1223 333 953; e-mail jmd32@cam.ac.uk).

Summary

Hydrogen peroxide is an important regulatory agent in plants. This study demonstrates that exogenous H2O2 application to Arabidopsis thaliana root epidermis results in dose-dependent transient increases in net Ca2+ influx. The magnitude and duration of the transients were greater in the elongation zone than in the mature epidermis. In both regions, treatment with the cation channel blocker Gd3+ prevented H2O2-induced net Ca2+ influx, consistent with application of exogenous H2O2 resulting in the activation of plasma membrane Gd3+-sensitive Ca2+-influx pathways. Application of 10 mm H2O2 to the external plasma membrane face of elongation zone epidermal protoplasts resulted in the appearance of a hyperpolarization-activated Ca2+-permeable conductance. This conductance differed from that previously characterized as being responsive to extracellular hydroxyl radicals. In contrast, in mature epidermal protoplasts a plasma membrane hyperpolarization-activated Ca2+-permeable channel was activated only when H2O2 was present at the intracellular membrane face. Channel open probability increased with intracellular [H2O2] and at hyperpolarized voltages. Unitary conductance decreased thus: Ba2+ > Ca2+ (14.5 pS) > Mg2+ > Zn2+ (20 mm external cation, 1 mm H2O2). Lanthanides and Zn2+ (but not TEA+) suppressed the open probability without affecting current amplitude. The results suggest spatial heterogeneity and differential sensitivity of Ca2+ channel activation by reactive oxygen species in the root that could underpin signalling.

Introduction

Reactive oxygen species (ROS) are now acknowledged as signalling and regulatory agents, rather than simply the potentially hazardous products of metabolic imbalance (Apel and Hirt, 2004; Mittler et al., 2004; Pastori and Foyer, 2002). ROS are involved in the regulation of defence responses and cell death (Alvarez et al., 1998; Zhang et al., 2003), gravitropism (Joo et al., 2001), stomatal aperture (Kwak et al., 2003; McAinsh et al., 1996; Murata et al., 2001; Pei et al., 2000), cell expansion and polar growth (Coelho et al., 2002; Foreman et al., 2003; Liszkay et al., 2004; Rodríguez et al., 2002, 2004; Schopfer et al., 2002) and leaf and flower development (Sagi et al., 2004). Additionally, ROS produced during abiotic stresses (such as hypoxia and nutritional restriction) act to signal change and regulate gene expression (Baxter-Burrell et al., 2002; Desikan et al., 2001; Mittler et al., 2004; Pastori and Foyer, 2002; Shin and Schachtman, 2004; Shin et al., 2005).

As studies advance, it is becoming clear that ROS: (a) accumulate in different cells during pathogen challenge (Trujilo et al., 2004); (b) can have opposite effects in a single cell type (for example, hydrogen peroxide can inhibit root hair growth, Jones et al., 1998; while hydroxyl radicals can stimulate it, Foreman et al., 2003); and (c) are capable of differential gene activation in a single cell type or organelle (op den Camp et al., 2003; Wisniewski et al., 1999). The mechanisms mediating such distinct responses rely in part on the complement of enzymes for production and scavenging of ROS in a given cell or organelle (Mittler et al., 2004) plus the proteins and lipids lying upstream or downstream of the ROS, such as phospholipase D and phosphatidic acid (Zhang et al., 2003), ROP GTPases (Baxter-Burrell et al., 2002) and MAP kinases (Kovtun et al., 2000; Rentel et al., 2004). In the control of stomatal aperture, cell expansion and polar growth, plasma membrane (PM) Ca2+ channels appear to be downstream of ROS production (Coelho et al., 2002; Foreman et al., 2003; Kwak et al., 2003; Murata et al., 2001; Pei et al., 2000). The resultant elevation of cytosolic Ca2+ could act as a second messenger or regulator of exocytosis and the cytoskeleton. ROS activation of Ca2+ channels probably forms the basis of a regulatory network in which specificity of ‘output’ is determined by the input combination of an individual ROS (superoxide anion, hydroxyl radical or H2O2) and a target Ca2+ channel in any given cell type. This would permit cell specificity and spatio-temporal heterogeneity in ROS/Ca2+-mediated signalling reactions.

To date, very little is known about regulation of a given cell's Ca2+ channel complement by different ROS. Sensitivity of root PM Ca2+ channel activity to hydroxyl radicals (OH·) decreases from the epidermis to the pericycle, and from the elongation zone epidermis to the mature epidermis (Demidchik et al., 2003). These results are broadly consistent with a role for Ca2+ channel activation by OH· in elongation growth (Demidchik et al., 2003; Foreman et al., 2003), gravitropism (Joo et al., 2001), and initial sensing of stress conditions at the plant–soil interface (Demidchik et al., 2003). Another key result emerging from root studies is a marked insensitivity of root epidermal PM Ca2+ channels to H2O2 (Demidchik et al., 2003; Foreman et al., 2003), tending to support the concept of differential ROS effects. Unlike guard cells, root-cell PM Ca2+ influx conductances have proved to be refractory to H2O2 applied to the extracellular membrane face (Demidchik et al., 2003; Foreman et al., 2003; Köhler et al., 2003; Kwak et al., 2003; Murata et al., 2001; Pei et al., 2000). While guard cell PM Ca2+ channels respond to micromolar extracellular H2O2 (Köhler et al., 2003), it was found previously that even prolonged exposure to 1–20 mm H2O2 at the extracellular membrane face had no significant effect on the PM Ca2+-conductance of Arabidopsis root mature epidermis (measured in the whole-cell recording mode; Demidchik et al., 2003).

It is now clear that H2O2 is involved in nutritional, hypoxic, and possibly low-temperature stress in roots (Baxter-Burrell et al., 2002; Lee et al., 2004; Shin and Schachtman, 2004; Shin et al., 2005). The relative abundance of apoplastic H2O2, OH· and superoxide anion is now also considered to be a significant factor in integrating growth with stress responses. While OH· could co-ordinate wall extension and channel-mediated Ca2+ influx, notable apoplastic H2O2 accumulation also occurs in the elongation zone of unstressed roots (del Carmen Córdoba-Pedregosa et al., 2003; Liszkay et al., 2004). This peroxide may simply be the precursor for OH· production, but could also modulate Ca2+ influx. Here the effect of H2O2 on Arabidopsis root Ca2+ transport and PM Ca2+ influx conductances has been re-examined. First, spatial variation in the effect of H2O2 application on net Ca2+ influx at the root epidermis of intact roots has been explored using a slowly vibrating Ca2+-selective microelectrode. Here we show dose-dependent, transient elevation in net Ca2+ influx in response to exogenous [H2O2]. This elevation was greater in the elongation zone than in mature regions. The results suggest that extracellular H2O2 could, directly or indirectly, activate Ca2+-influx pathways in intact root cells. High extracellular [H2O2] (representative of apoplastic [H2O2] under stress conditions) stimulated a PM hyperpolarization-activated Ca2+ conductance in elongation zone epidermal protoplasts. Critically, this conductance differed from that stimulated by extracellular OH·, and may operate in stress signalling. A single-channel examination using excised PM patches from the mature epidermis revealed that a hyperpolarization-activated Ca2+ channel in this cell type differs from that found in the root hair, and responds only to intracellular H2O2. It is proposed that, in the mature epidermis, OH· may be the only extracellular ROS capable of regulating Ca2+ influx. Thus a spatially distinct pattern emerges of differential ROS effects and differential channel activation that may be involved in growth and environmental sensing.

Results

Effect of exogenous H2O2 on net epidermal Ca2+ flux

Net epidermal Ca2+ flux was recorded under control conditions at the elongation zone (100–150 μm from the apex) and in a fully expanded, mature region (1–1.5 mm from the apex). Under control conditions, mean ± SE net basal Ca2+ influx at the elongation zone was 7.10 ± 0.32 nmol m−2 sec−1 (n = 77), and at the mature epidermis was 3.30 ± 0.19 nmol m−2 sec−1 (n = 39). The greater net influx to the elongation zone confirms this previously reported pattern (Demidchik et al., 2002), and is consistent with a greater requirement for Ca2+ influx to this zone for growth and nutrition. Pre-treatment with the cation channel blocker Gd3+ (30 μm) resulted in statistically significant (P < 0.001; unpaired t-test) reductions of basal net Ca2+ influx (elongation zone 1.10 ± 0.15 nmol m−2 sec−1, n = 25; mature epidermis 0.30 ± 0.01 nmol m−2 sec−1, n = 24). These results indicate that, in both regions, basal net Ca2+ influx was likely to be mediated by Ca2+-permeable cation channels.

In control experiments, addition of bathing solution alone did not evoke net Ca2+ influx (Figure 1a). A typical response of the mature epidermis to the application of 10 mm H2O2 is shown in Figure 1(b). Transient increases in net Ca2+ influx were observed immediately after the resumption of recording, and the magnitude of the increase in net Ca2+ influx was dose-dependent (Figure 1c). At 10 mm H2O2, the mean ± SE duration of transient influx was 134 ± 20 sec, and the mean ± SE maximum increase in net Ca2+ influx was 8.4 ± 1.7 nmol m−2 sec−1 (n = 13). Pre-treatment with 30 μm Gd3+ prevented the response to even the highest (10 mm) H2O2 treatment; mean net Ca2+ influx was 0.15 ± 0.02 nmol m−2 sec−1 (n = 8) when measured during a period similar to the control experiment.

Figure 1.

 Effect of exogenous H2O2 on net Ca2+ flux at the Arabidopsis root epidermis.
(a) Addition of control buffer (marked by arrow) did not evoke net Ca2+ influx (n = 2).
(b) Time course of net Ca2+ flux measured at the mature epidermis using a Ca2+-selective vibrating electrode. Arrow marks addition of 10 mm H2O2 to the bathing medium (0.1 mm KCl, 0.1 mm CaCl2, 5 mm MES, 2 mm Tris base, pH 6.0). Upward deflection corresponds to an increase in net Ca2+ influx.
(c) Effect of increasing exogenous H2O2 on mean ± SE (n = 4–21) maximum value of the transient Ca2+ influx at the elongation zone (control,•; 30 μm Gd3+ pre-treatment, bsl00072) and at the mature epidermis (control, bsl00043). For clarity, the effect of 30 μm Gd3+ pre-treatment on the mature epidermis, which reduced the signal to that of the equivalent test on the elongation zone, has been omitted.

Application of H2O2 to the elongation zone epidermis also caused dose-dependent transient increases in net Ca2+ influx (Figure 1c). Duration of the transient response increased with increasing [H2O2]: 10 μm, 121 ± 16 sec, n = 4; 10 mm, 312 ± 36 sec, n = 5. In addition to greater duration of response, the magnitude of mean maximum increases of net Ca2+ influx was statistically significantly greater (P < 0.001; unpaired t-test) than those of the mature epidermis at all concentrations tested (10 μm to 10 mm). At 10 mm H2O2 the mean ± SE maximum increase in net Ca2+ influx was 29.5 ± 2.4 nmol m−2 sec−1 (n = 7). As with the mature epidermis, pre-treatment with 30 μm Gd3+ suppressed the response to H2O2 (Figure 1c: At 10 mm H2O2, 1.1 ± 0.25 nmol m−2 sec−1, n = 8). Overall, the data are consistent with application of exogenous H2O2 resulting (directly or indirectly) in the activation of PM Gd3+-sensitive Ca2+-influx pathways in both elongation zone and mature regions of the root.

Electrophysiological response of elongation zone protoplasts to exogenous H2O2

Protoplasts were isolated from elongation zone epidermis and exposed to 10 mm extracellular H2O2 shortly after establishing the whole-cell patch-clamp configuration. A hyperpolarization-activated, inwardly directed conductance reached its maximum after approximately 20 min exposure to 10 mm extracellular H2O2 [Figure 2a,b,e: Bathing solution (BS) 20 mm CaCl2; n = 7]. Under control conditions (no H2O2) in this period, the mean ± SE reversal potential (Erev) was 43 ± 7 mV, and current at a holding potential (HP) of −160 mV was −22 ± 3 pA (n = 7). After H2O2 exposure, the mean ± SE Erev was 59 ± 9 mV, and current at HP = −160 mV was −111 ± 8 pA (n = 7). The positive Erev of the H2O2-responsive conductance is consistent with Ca2+ permeability.

Figure 2.

 Effect of exogenous H2O2 on whole-cell currents of elongation zone epidermal protoplasts.
(a) Control currents in response to voltage steps from −160 to 80 mV in 20 mm CaCl2 bathing solution (BS).
(b) Currents approximately 20 min after application of 10 mm H2O2 to BS.
(c, d) Sequential substitution of Ca2+ with Ba2+ and Zn2+ in continued presence of 10 mm H2O2.
(e) Mean ± SE current–voltage relationships:bsl00043, 20 mm CaCl2 control;•, 20 mm CaCl2 + 10 mm H2O2; bsl00072, 20 mm BaCl2 + 10 mm H2O2; bsl00083, 20 mm ZnCl2 + 10 mm H2O2 (n = 5–7).

The H2O2-responsive conductance was also permeable to Ba2+, indicating that it was not generated by K+ channels. The current generated with Ba2+ as the charge carrier was greater than that with Ca2+ (Figure 2c,e: Mean ± SE current at HP = −160 mV, −215 ± 22 pA; n = 5). This is typical of animal voltage-regulated calcium channels (Wang et al., 2005) and plant hyperpolarization-activated calcium channels (reviewed by White, 2000). The H2O2-responsive conductance was poorly permeable to Zn2+ (Figure 2d,e: mean ± SE current at HP = −160 mV, −65 ± 7 pA; n = 5) which can act as a cation channel blocker (Kiss and Osipenko, 1994). The temporal development of the H2O2-responsive hyperpolarization-activated Ca2+ channel (HACC) conductance was far more rapid than the constitutive HACC conductance normally observed in protoplasts from the elongation zone (40–60 min; Demidchik et al., 2002). Critically, the H2O2-responsive HACC conductance comprised a time-dependent current component (most clearly seen with Ba2+ permeation; Figure 2c). This distinguishes it from an OH·-activated Ca2+-permeable conductance reported in this cell type, and also in protoplasts from the mature epidermis (Demidchik et al., 2003). To assess whether extracellular H2O2 was permeating the PM and acting at the intracellular face, catalase was incorporated into the pipette solution (PS) (up to 100 units ml−1), but this compromised seal formation and stability.

H2O2 sensitivity of protoplasts from the mature epidermis and cortex

In whole-cell recordings on protoplasts derived from the mature epidermis, an inwardly directed HACC conductance develops in approximately half the protoplasts sampled, 40–60 min after establishing the configuration (Demidchik et al., 2002). Here, in whole-cell recordings on protoplasts from the mature epidermis, treatment (up to 30 min) with 0.1, 0.3, 1 or 3 mm H2O2 added to the BS or PS did not induce inward currents either more quickly or more frequently (BS, 20 mm CaCl2; n = 8 for each H2O2 concentration either in BS or PS). Thus the apparent insensitivity of this protoplast type to exogenous H2O2 (1–20 mm; Demidchik et al., 2003) was confirmed at lower [H2O2] and persisted even when H2O2 was present at the intracellular face.

It is feasible that the underlying cortical cells could have contributed to the net Ca2+ influx in response to H2O2 applied to intact roots. As a direct test of the H2O2-responsiveness of the cortex, cortical protoplasts were isolated from the mature zone and patch-clamped. The response to 20–30 min exposure in whole-cell configuration to 10 mm H2O2 is shown in Figure 3. Under control conditions, the mean ± SE Erev was 51.3 ± 2 mV, and current at a holding potential (HP) of −170 mV was −27.1 ± 4.1 pA (n = 4). After H2O2 exposure, the mean ± SE Erev was 45.5 ± 2.5 mV, and current at HP = −170 mV was −22.6 ± 5 pA (n = 4). The positive Erev values are consistent with Ca2+ permeability. As exogenous H2O2 did not increase inwardly directed Ca2+ transport into cortical or epidermal protoplasts, it is feasible that in the whole-root studies the H2O2 was converted into a more reactive ROS, which activated net Ca2+ influx. However, to explore the possibility that the null response in whole-cell recordings on epidermal protoplasts was the consequence of low functional channel abundance or poor permeability to extracellular H2O2, the response of excised membrane patches was examined.

Figure 3.

 Effect of exogenous H2O2 on whole-cell currents of root cortical protoplasts from the mature zone.
(a) Control currents in response to voltage steps from −170 to 100 mV in 20 mm CaCl2 bathing solution (BS).
(b) Currents approximately 20 min after application of 10 mm H2O2 to BS.
(c) Mean ± SE current–voltage relationships: bsl00084, 20 mm CaCl2 control; bsl00043, 20 mm CaCl2 + 10 mm H2O2 (n = 4).

Constitutive channel activity in plasma membrane from the mature epidermis

Excised membrane patches from mature epidermis were most readily obtained for the outside-out configuration, and formed the basis of this study. Under control conditions (BS; 20 mm CaCl2) with no exogenous H2O2 present, a constitutive single channel with a conductance of approximately 4–5 pS was observed in approximately half of the patches, but activity was lost within 40–60 min. This magnitude of single-channel conductance corresponds to a Ca2+-permeable non-selective cation channel that was characterized in our previous study (Demidchik et al., 2002). A larger conductance channel was also observed under control conditions in six patches out of 18 examined (Figure 4a). The open probability (Popen) of this channel (determined at HP = −100 mV) increased with time after patch excision (Figure 4b). Single-channel activity increased with approximately the same time course as the HACC activity previously measured as whole membrane conductances (Demidchik et al., 2002).

Figure 4.

 The constitutive hyperpolarization-activated Ca2+ channel of Arabidopsis root plasma membrane (mature epidermis).
(a) Single-channel activity in an outside-out patch observed at given time points after patch excision (holding potential, HP = −100 mV).
(b) Increase over time of mean ± SE Popen of the first open state; HP = −100 mV, n = 4–6.
(c) Effect of HP on mean ± SE Popen of the first open state (O1), n = 4–6.
(d) Current–voltage relationship of the channel first open state (n = 4).
In all assays (a–d), bathing solution (BS) contained 20 mm CaCl2; pipette solution (PS) contained 25 mm K-gluconate, 5 mm KCl and 100 nm Ca2+.

Single-channel activity was further characterized approximately 50 min after excision: Popen increased as the membrane voltage was hyperpolarized (Figure 4c) and the mean ± SE single-channel conductance of the first open state was estimated to be 15.3 ± 0.9 pS (n = 4; Figure 4d). The mean extrapolated Erev of this inward current was 55 ± 4 mV (n = 4). This was away from the estimated equilibrium potential for Cl (ECl = −52.4 mV) and towards that of Ca2+ (ECa = 153.7 mV), indicating Ca2+ permeation of this hyperpolarization-activated channel. In light of its voltage sensitivity and Ca2+ permeability, the channel was termed the constitutive hyperpolarization-activated Ca2+ channel (HACC). In vivo, the membrane voltage of mature epidermal cells can be hyperpolarized sufficiently (Maathuis and Sanders, 1993) to support Ca2+ influx through this HACC. On the basis of unitary conductance, the constitutive HACC is unlikely to be the same protein as the root hair HACC (15 pS in 20 mm CaCl2 compared with the root-hair channel at 22 pS in 10 mm external Ca2+; Véry and Davies, 2000).

Channel activity in mature epidermal PM in the presence of H2O2

Extracellular application of H2O2 (3 mm in BS) had no appreciable effect on single-channel activities in outside-out membrane patches from protoplasts of mature epidermis (BS: 20 mm CaCl2; n = 6), even though this concentration effected net Ca2+ influx in vivo (Figure 1c). However, in 29 of 67 outside-out patches, addition of 0.1–3 mm H2O2 to the PS resulted in single-channel activity (Figure 5a), which was apparent immediately after the formation of the outside-out configuration. Decrease of Ca2+ (from 20 to 0.3 mm) in the BS abolished H2O2-responsive currents (1 mm H2O2 in PS; n = 4), demonstrating that these currents were due to Ca2+ influx from the outer to the cytosolic membrane face.

Figure 5.

 Hydrogen peroxide-responsive unitary conductances in outside-out patches (mature epidermis).
(a) Currents recorded at different [H2O2] in pipette solution at holding potential (HP) = −50 mV. Note the four open states (O1−4) at 3 mm H2O2.
(b) Mean ± SE Popen of the first open state at different [H2O2] (HP = −50 mV), n = 4.
(c) Typical time course of Popen change showing ‘run-down’ of H2O2-activated channel (1 mm H2O2; HP = −50 mV).
In all assays (a–c), bathing solution (BS) contained 20 mm CaCl2; PS contained 25 mm K-gluconate, 5 mm KCl and 100 nm Ca2+.

The Popen of this channel's first open state (estimated at −50 mV) was dependent on the H2O2 concentration (Figure 5b); the greatest activation was found at the highest H2O2 concentration (3 mm). However, as can be seen in Figure 5(c), channel activity declined with time. The H2O2-responsive channel was voltage-dependent, revealing higher Popen upon hyperpolarization of membrane voltage (Figure 6a,b; Popen was calculated from the first open state of the channel in the presence of 1 mm H2O2). The single-channel conductance was determined at 20 mm external CaCl2 and with 1 mm cytosolic H2O2; Figure 6(c). The mean conductance was 14.4 ± 0.6 pS (n = 4). Under these conditions, the extrapolated Erev of the unitary inward currents ranged from +50 to +100 mV (mean ± SE Erev = 72 ± 9 mV; n = 4). This was more positive than ECl (ECl = −52.4 mV) and indicative of Ca2+ permeability (ECa = 153.7 mV). This channel was termed the H2O2-responsive HACC. The unitary conductances of the constitutive and H2O2-responsive HACCs were not statistically significantly different (Student's t-test), suggesting that they may be the same channel protein.

Figure 6.

 Properties of the H2O2-responsive channel (mature epidermis, outside-out patches).
(a) H2O2-responsive currents recorded at different holding potentials (HP) as indicated.
(b) Mean ± SE voltage-dependence of Popen (n = 4).
(c) Mean ± SE current-voltage relationship of the H2O2-responsive channel (n = 4).
(d) Mean ± SE current of H2O2-responsive channel recorded at HP = −100mV (n = 3) measured in bathing solution (BS) containing 20 mm Cl salt of Ba2+, Ca2+, Mg2+, Co2+ or Zn2+. Mean ± SE unitary conductances are shown. For seal stability, 0.3 mm CaCl2 was added in the cases of Mg2+ and Co2+.
(e) Pharmacology of the H2O2-responsive channel: The effect of blockers (applied to BS) on Popen, HP = −150 mV (n = 3–4).
(a–c, e) BS contained 20 mm CaCl2. In all assays (a–e), PS contained 25 mm K-gluconate, 5 mm KCl, 100 nm free Ca2+ and 1 mm H2O2.

Mean unitary conductance measured with 20 mm divalent cation in the BS (Cl salt) was: Ba2+ (21.3 pS) > Ca2+ (14.5 pS) > Mg2+ (9.2 pS) ≈ Co2+ (8.9 pS) > Zn2+ (3.9 pS) (n = 3; Figure 5d). To maintain seal stability, 0.3 mm Ca2+ was added to the bath in the case of Mg2+ and Co2+. This permeability sequence was based on unitary conductance values, estimated from the current recorded at HP = −100 mV with 1 mm H2O2 in the PS. The Popen of the H2O2-responsive HACC was not affected by external application (to BS) of the K+ channel blocker TEA+ (5 mm); however, it was reduced significantly by Zn2+(5 mm), La3+(20 μm) or Gd3+(20 μm) (Figure 6e), which are known to act as Ca2+ channel antagonists in many plant preparations. Single-channel conductance was not affected by the blockers tested (data not shown).

Discussion

The elongation zone epidermis contains an H2O2-responsive plasma membrane HACC conductance

Patch-clamping protoplasts from the elongation zone epidermis revealed previously that exogenously applied 2 mm H2O2 had no effect on PM Ca2+ channel activity. However, when the same concentration of H2O2 was used to generate exogenous OH· by further addition of 0.2 mm ascorbate/0.2 mm Cu2+, a hyperpolarization-activated and inwardly directed Ca2+ conductance was revealed (Foreman et al., 2003). Here, exogenous application of far higher [H2O2] (10 mm) as the sole ROS caused the appearance of a PM HACC conductance. This H2O2-responsive HACC conductance appeared more rapidly (20 min) than either the constitutive HACC in this cell type (40–60 min; Demidchik et al., 2002, 2003) or the mature epidermis and root hair (40–60 min; Demidchik et al., 2002; Véry and Davies, 2000). The kinetics of both the elongation zone epidermal constitutive HACC and the H2O2-responsive HACC differ markedly from the PM OH·-activated Ca2+ conductance in this cell type (Demidchik et al., 2002, 2003; Foreman et al., 2003). This supports the premise that, in the elongation zone epidermal PM, different apoplastic ROS (H2O2 and OH·) regulate distinct Ca2+-permeable pathways to modulate Ca2+ influx according to cellular need.

Dose-dependent effects of apoplastic H2O2 in elongation zone epidermis could be mediated by two Ca2+ conductances

Apoplastic H2O2 is generated in vivo by cell wall peroxidases, polyamine oxidases, amine oxidases, oxalate oxidases and by spontaneous or superoxide dismutase-mediated dismutation of extracellular superoxide anions (produced by PM NADPH oxidase) (Apel and Hirt, 2004; Cona et al., 2003; Mittler et al., 2004). Zonal quantification of apoplastic H2O2 during root growth has yet to be achieved, but overall apoplastic [H2O2] in Arabidopsis has been estimated to be between 0.6 and 7 mm (Karpinski et al., 1999; Veljovic-Jovanovic et al., 2001). In Zea seedlings, apoplastic [H2O2] is around 0.1–0.2 mm, rising to 10 mm after chilling stress (Prasad et al., 1994). Histochemical detection of H2O2 in growing, unstressed onion roots revealed greatest accumulation in apical and elongation zone epidermal cell walls with close association with the PM; cytosolic H2O2 was rarely detected (del Carmen Córdoba-Pedregosa et al., 2003). In growing, unstressed maize roots, apoplastic H2O2 accumulated specifically in elongation zone epidermis rather than the mature zone (Liszkay et al., 2004).

Elongation zone apoplastic H2O2 is thought to be converted (by apoplastic Fenton reagents Cu+ or Fe+) to OH·, which remodel wall carbohydrates to permit extension and also activate Ca2+ influx for growth (Demidchik et al., 2003; Foreman et al., 2003; Liszkay et al., 2004; Renew et al., 2005; Rodríguez et al., 2002; Schopfer et al., 2002). However, exogenously applied (mm) H2O2 signals oxidative stress in the elongation zone of Arabidopsis root, resulting in Ca2+ influx-dependent expression of Glutathione-S-Transferase1 (GST1; Rentel and Knight, 2004). As cellular responses to H2O2 are dose-dependent (Apel and Hirt, 2004; Mittler et al., 2004; Zhang et al., 2003), we propose that, in the present study, addition of micromolar H2O2 to the intact elongation zone epidermis resulted in its conversion to OH·, which then stimulated Ca2+ influx through the previously characterized growth-related OH·-activated Ca2+ conductance (Demidchik et al., 2003; Foreman et al., 2003). As exogenous [H2O2] was increased to mm levels (consistent with H2O2 accumulation at the PM as the result of stress-induced NADPH oxidase activity; Lee et al., 2004; Zhang et al., 2003), an increasingly greater part of the Ca2+ influx would be mediated by the H2O2-responsive HACC conductance observed in patch-clamp experiments (Figure 7).

Figure 7.

 Summary of reactive oxygen species effects on plasma membrane calcium channels in the elongation zone and mature epidermis of Arabidopsis roots.
In the elongation zone, high concentrations of apoplastic H2O2 would activate a HACC (red), however it is unknown at which membrane face H2O2 exerts its effects. Hydroxyl radicals act at the extracellular membrane face and increase a Ca2+-permeable, non-selective cation channel conductance (green; Demidchik et al., 2003). In the mature epidermis, cytosolic H2O2 activates the constitutive HACC (this study) while extracellular OH· positively regulate a non-selective cation channel conductance (Demidchik et al., 2003).

A more specific role in stress signalling can be proposed for the elongation zone H2O2-responsive HACC conductance. Rentel and Knight (2004) invoked a PM H2O2-activated HACC in mediating the Ca2+ influx necessary for induction of GST1 expression in the elongation zone in response to 10 mm exogenous H2O2. However, at the time only the null response to 2 mm exogenous H2O2 in patch-clamp experiments was known (Foreman et al., 2003), compromising the model. The finding here that 10 mm exogenous H2O2 activates an elongation-zone PM HACC now strongly supports Rentel and Knight's original proposal.

Peroxide acts at the intracellular membrane face of the mature epidermal PM

In studies on protoplasts from the mature epidermis, H2O2 was ineffective when applied to the extracellular face of outside-out patches or whole membrane. Only H2O2 at the intracellular face evoked HACC activity. This is in stark contrast with the guard cell PM H2O2-responsive Ca2+ channel (Köhler et al., 2003) or the H2O2-responsive cation-permeable channel of Fucus rhizoid apical PM (Coelho et al., 2002), where extracellular H2O2 was effective. However, it is notable that in all three systems similar [H2O2] caused an increase in channel number without altering single-channel conductance. On the basis of unitary conductance, the H2O2-responsive HACC of the mature epidermis is most probably the constitutive HACC, and its relative Ca2+/Mg2+ permeability again suggests that this channel is a different protein from the root hair HACC (Véry and Davies, 2000). No single-channel measurements are available for the OH· response of root epidermis and root hairs (Demidchik et al., 2003; Foreman et al., 2003), for comparison. It is feasible that H2O2 acts directly on the HACC, given that activation was independent of intracellular ATP and that H2O2 activation of channels in artificial bilayers can result in a similar increase in open probability without affecting single-channel conductance (Kourie, 1998).

There may not always be a strict correlation between Ca2+ influx and current in root protoplasts, which may be due to charge compensation (Gilliham et al., 2006). The discrepancy between the positive in planta response of the mature epidermis and cortex to extracellular H2O2 and the null response of protoplasts could mean that, in whole roots, H2O2 was converted apoplastically to a more reactive ROS such as OH·. Certainly, the cell wall is critical to peroxide conversion and breakdown. The half-life of H2O2 applied exogenously to Arabidopsis suspension cell protoplasts was estimated to be 1 h, but only 2–5 min for the original cells (Neill et al., 2002). The discrepancy could also reflect a ‘change of state’ in protoplasts, for example the loss or inactivity of an H2O2 sensor, or a lowered PM permeability to H2O2 (Branco et al., 2004). Physiological changes induced by protoplasting may explain why (even when protoplasts were dialysed with H2O2 from the PS reservoir) there was no response to intracellular H2O2 during whole-cell experiments. Protoplasting involves wounding and biotic challenge with wall-degrading enzymes – processes that evoke upregulation of cellular antioxidant capacity, such as increased catalase activity (Desikan et al., 2001; Guan and Scandalios, 2000; Neill et al., 2002). It is envisaged that far less native antioxidant capacity would be retained in the outside-out patch configuration. Therefore incorporation of H2O2 in the PS may bring sufficient H2O2 into proximity with the channel to cause activation.

Physiological significance of the vectorial response to H2O2 in the mature epidermis

The mature epidermis is a key site for NADPH oxidase-mediated nutrient stress signalling, with an increase in [H2O2] occurring just behind the elongation zone upon K+ deprivation (Shin and Schachtman, 2004; Shin et al., 2005). The site of H2O2 generation (extracellular or intracellular) can have a profound effect on stress-responsive gene expression (Avsian-Kretchmer et al., 2004). Apoplastic [H2O2] is not significant in the mature epidermis of unstressed, growing roots (Liszkay et al., 2004), but would increase upon stress due to PM NADPH oxidase activity (Lee et al., 2004; Shin and Schachtman, 2004; Shin et al., 2005). Intracellularly, H2O2 could be produced by cytoplasmic polyamine oxidase (Cona et al., 2003) or diffuse into the cytoplasm from organelles. Taken together with our previous findings, a clearer picture is now emerging of ROS-mediated Ca2+ influx at the mature epidermis (Figure 7). We propose that, in this cell type, stress results in elevated cytoplasmic H2O2, which activates Ca2+ influx through the H2O2-responsive HACC. This Ca2+ signal could then be amplified by Ca2+ activation of the PM NADPH oxidase (through its EF hands), stimulating the eventual production of apoplastic OH·, which then (acting at the extracellular face) causes further Ca2+ influx through the PM OH·-activated Ca2+-permeable conductance (Demidchik et al., 2003).

Experimental procedures

Plant material and growth conditions

Arabidopsis thaliana (ecotype Columbia and line J0481 from Dr J. Haseloff, University of Cambridge) was grown aseptically and vertically at 22°C for 5–15 days (16 h daylength; 100 μmol m−2 sec−1 irradiance) on medium comprising 0.3% (w/v) Phytagel (Sigma, Poole, UK), full-strength Murashige–Skoog medium (Duchefa, Haarlem, Netherlands) and 1% (w/v) sucrose.

Measurement of net Ca2+ flux

Net Ca2+ fluxes were measured using the non-invasive, slowly vibrating microelectrode ion flux estimate (MIFE) technique (Shabala et al., 1997). Plants were sampled after 5–15 days’ growth. Excised roots were mounted in a Perspex chamber by an agar drop and immersed in 0.5 ml assay solution comprising (in mm) 0.1 KCl, 0.1 CaCl2, 5 MES, 2 Tris base (pH 6.0). The root was viewed under an Olympus IX50 microscope using ×200 magnification. Ca2+-selective microelectrodes with an external tip diameter of approximately 3 μm were manufactured and calibrated as described previously (Shabala, 2000; Shabala et al., 1997). Electrodes were calibrated before and after use, using the standard procedure (Shabala, 2000). Peroxide (up to 10 mm) did not affect calibration values or electrode response time. The microelectrode was placed 20 μm above the root surface. During measurements, the electrode was moved between two positions, 20 and 50 μm above the root surface, in a square-wave manner with a 5-sec half cycle. Measurements of mature epidermal flux were taken at 1–1.5 mm from the root apex; those of elongation zone epidermis were taken 100–150 μm from the root apex. Net Ca2+ flux was measured for 5 min prior to the addition of 0.5 ml assay solution containing H2O2 (Shabala, 2000; Shabala et al., 1997). Measurements were resumed after 20–30 sec when unstirred layer conditions were reached.

Patch-clamp electrophysiology

Plants were harvested after 10–15 days’ growth. The procedure for protoplast isolation from root mature and elongation zone epidermis was described in full by Demidchik et al. (2003). Cortical protoplasts were isolated using the same protocol as for the mature epidermis and identified by the absence of GFP fluorescence in a line expressing GFP in the epidermis (line J0481; Demidchik et al., 2002; Kiegle et al., 2000). They were then selected for in the wild type by their uniform darker colour and slightly larger size than epidermal protoplasts. Protoplasts used for patch clamping were 22–23 μm in diameter (mature cortex), 20 ± 1.5 μm (mature epidermis) or 15–25 μm (elongation zone). Seal formation was carried out in basic bathing solution (BS) comprising (in mm): 20 CaCl2, 2 MES, pH 5.7 with Tris. A BS with different divalent cation content was then introduced as required. The pipette solution (PS) was optimized for each type of protoplast to effect high sealing rates and good seal stability. In studies on protoplasts from the mature epidermis, the PS contained (in mm): 25 K-gluconate, 5 KCl, 5 1,2-bis(2-aminophenoxy)ethane-N,N,N′, N′-tetra-acetic acid [BAPTA; free Ca2+ adjusted to 100 nm with Ca(OH)2] pH 7.2 adjusted with 5 mm HEPES/KOH. For the mature zone cortical protoplasts, PS comprised (in mm): 20 Na-gluconate, 10 NaCl, 2 BAPTA [free Ca2+ adjusted to 100 nm with Ca(OH)2] pH 7.2 with 5 mm HEPES/NaOH. In studies on protoplasts from the elongation zone epidermis, the PS comprised (mm): 25 Na-gluconate, 5 NaCl, 4 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid pH 7.2 with 5 HEPES/NaOH. BS and PS were adjusted to 290–300 mOsM with d-sorbitol (ultrapure grade; Sigma, Poole, UK). Solutions containing H2O2 (Sigma) were freshly prepared directly before application. Ion activities were calculated using geochem (Parker et al., 1995). Liquid junction potentials were measured and corrected as described previously (Demidchik et al., 2003).

Acknowledgements

We thank Professor Enid MacRobbie FRS for useful discussion and Dr Jim Haseloff for line J0481. Financial support from the BBSRC (8/D/13399), the Leverhulme Trust (Project Grant F/09 741/C) and ARC (to S.N.S.) is appreciated.

Ancillary