Calcium- and potassium-permeable plasma membrane transporters are activated by copper in Arabidopsis root tips: linking copper transport with cytosolic hydroxyl radical production


S. Shabala. E-mail:


Transition metals such as copper can interact with ascorbate or hydrogen peroxide to form highly reactive hydroxyl radicals (OH), with numerous implications to membrane transport activity and cell metabolism. So far, such interaction was described for extracellular (apoplastic) space but not cytosol. Here, a range of advanced electrophysiological and imaging techniques were applied to Arabidopsis thaliana plants differing in their copper-transport activity: Col-0, high-affinity copper transporter COPT1-overexpressing (C1OE) seedlings, and T-DNA COPT1 insertion mutant (copt1). Low Cu concentrations (10 µm) stimulated a dose-dependent Gd3+ and verapamil sensitive net Ca2+ influx in the root apex but not in mature zone. C1OE also showed a fivefold higher Cu-induced K+ efflux at the root tip level compared with Col-0, and a reduction in basal peroxide accumulation at the root tip after copper exposure. Copper caused membrane disruptions of the root apex in C1OE seedlings but not in copt1 plants; this damage was prevented by pretreatment with Gd3+. Our results suggest that copper transport into cytosol in root apex results in hydroxyl radical generation at the cytosolic side, with a consequent regulation of plasma membrane OH-sensitive Ca2+ and K+ transport systems.


Copper (Cu) is an essential micronutrient in higher plants. Cu participates as a catalytic cofactor in many fundamental processes such as electron transport, hormone signalling, cell wall metabolism, pollen formation and fertilization, nodulation and nitrogen fixation, and carbohydrate metabolism (Marschner 2012). In addition, Cu is present as a metal component of the prosthetic group in superoxide dismutases (CuZnSOD) and, thus, plays a central role in detoxification of the superoxide anion free radicals (Shabala 2009a). On the other hand, being a transition (and, hence, highly reactive) metal, Cu ions can lead to the generation of harmful reactive oxygen species (ROS) via Fenton and Haber–Weiss reactions, thus causing oxidative damage to cells (Halliwell & Gutteridge 1984). Copper toxicity is well reported and, despite some Cu-tolerant species that can accumulate as much as 1000 µg g−1 dry weight (DW) of copper, leaf concentrations above 30 µg g−1 DW are considered to be toxic for most crop species (Marschner 2012). This duality of copper has resulted in the development of a complex homeostatic network for copper acquisition and use in aerobic organisms.

Copper transport in plants has been recently reviewed (Burkhead et al. 2009; Pilon 2011). Some members of the ZIP (Zrt, Irt-like protein) family (such as ZIP2 or ZIP4) might be able to mediate copper transport in plant cells, as they complement yeast Cu and Zn transport mutants (Wintz et al. 2003; Burkhead et al. 2009). In addition, ZIP2 transcript levels are up-regulated in response to copper deficiency (Yamasaki et al. 2009). However, ZIP function in Cu transport has not been characterized yet in planta.

The CTR family of trimeric transporters, called COPT in plants, mediates copper transport towards the cytosol in eukaryotic cells (reviewed by Kim, Nevitt & Thiele 2008). It has been suggested that copper transport may require the reduction of Cu2+ to Cu+ by a reductase. Cu+ is then transferred towards the cytosol in a process that requires conserved sequence motifs (reviewed by Peñarrubia et al. 2010). Six members of the family COPT of high-affinity copper transporters have been identified in Arabidopsis (COPT1-6) (Sancenón et al. 2003; Peñarrubia et al. 2010). Some members of the COPT family are induced under copper deficiency via interaction with the SPL7 transcription factor (Yamasaki et al. 2009), and the expression of COPT1 and COPT2 is down-regulated by Cu (Sancenón et al. 2003). Four of the members of the COPT family can complement yeast crt1/crt3D defective growth (COPT 1–3 and COPT5), but only two of them have been characterized in detail. COPT5 it is located at the tonoplast (Garcia-Molina et al. 2011; Klaumann et al. 2011) and it is required for root growth under severe copper deficiency (Garcia-Molina et al. 2011) and for copper ion reallocation from roots to reproductive organs (Klaumann et al. 2011).

The first member of the COPT family to be functionally characterized was COPT1 (Kampfenkel et al. 1995; Sancenón et al. 2004). COPT1 is located at the plasma membrane (PM) in root tips, is expressed under copper deficiency, and is responsible for high-affinity copper uptake in Arabidopsis (Sancenón et al. 2004; Andrés-Colás et al. 2010). COPT1-antisense plants (C1AS) have reduced copper uptake and showed defective root elongation (Sancenón et al. 2004). The effect of overexpressing COPT1 (C1OE) results not only in higher copper accumulation by the seedlings, but also in deregulated development (Andrés-Colás et al. 2010). Some aspects of COPT1 expression, copper accumulation, and the associated phenotypes grown on Murashige & Skoog (MS) medium and MS supplemented with Cu are detailed in Table 1.

Table 1. COPT1 expression, copper uptake and associate phenotype of Arabidopsis seedlings differing in copper-transport activity
  PhenotypeCOPT1 expressionCooper uptake/accumulation
  1. a Sancenón et al. 2004. bAndrés-Colás et al. 2010. cGarcia-Molina et al. 2011. dWu et al. 2009.

  2. BCS, bathocuproinedisulfonic acid; WT, wild type.

Seedlings grown on MS mediumCol-0Root reduction after 3 weeks grown with 100 µm BCScHeart stage embryos, root tips, pollen grains, trichomes, stomata guard cells (COPT1::GUS transgenic plants)a 
COPT1 overexpression (C1oe)Root length similar to the WTb2.5-fold increase in COPT1 expressionbModerate increase in seedling Cu content (determined by graphite furnace atomic absorption spectrometry)b
COPT1-antisense (C1AS)Roots three to six times longer than WT when grown on the sucrose-depleted mediumaTranscript levels of CCH and COPT2 (up-regulated in response to decreasing copper) with fourfold increasea 
Seedlings grown on MS medium supplemented with copperCol-0Root growth unaffected with concentrations up to 20 µmb, and reduced with 50 µmb Cu uptake is prevented when intracellular Cu levels rised
COPT1 overexpression (C1oe)Root length dramatically reduced with 10 µm. Lateral roots.b COPT1 mRNA levels with 3.5-fold increasebModerate increase in seedling Cu contentb. Increase and modification of Cu uptake pattern.b
COPT1-antisense (C1AS)Roots length reverts to WT phenotype with 30 µm copper additiona Both copper uptake and accumulation reduced 40–60%a

As mentioned before, Cu is a transition metal and, as such, can mediate ROS production. In the presence of either ascorbate or hydrogen peroxide (H2O2), copper can produce highly toxic hydroxyl radicals (OH) via Fenton and Haber–Weiss reactions (Fry 1998). OH can cause oxidative damage to proteins and nucleic acids, as well as lipid peroxidation during stress. In addition, it is involved in oxidative stress signalling and programmed cell death (PCD) (Demidchik et al. 2010). The in vivo half-life of OH is approximately 1 ns, which allows OH diffusion over only very short distances (<1 nm) (Halliwell & Gutteridge 1999). Because of this, OH generation was normally associated with two major sites: cell walls and chloroplasts. In the former, OH production was shown to be dependent on the activity of NADPH oxidase, RbohC, that forms H2O2, which is later catalysed by transition metals in the cell wall (Foreman et al. 2003). Here, ascorbic acid is likely to serve as a reductant, because its concentration in the cell wall is very high (1–20 mm) (Fry, Miller & Dumville 2002). In chloroplasts, OH is generated by a Fenton-like reaction of H2O2 with free transition metals present in the stroma (Pospíšil et al. 2004; Šnyrychová, Pospíšil & Nauš 2006; Pospíšil 2009). Copper is a constituent of plastocyanin, and thus is also crucial for the electron transport between photosystem II (PSII) and photosystem I (PSI), and the majority of copper present in plants is located in chloroplasts (Marschner 2012).

Generation of OH has significant implications to cell metabolism. In addition to causing lipid peroxidation (Apel & Hirt 2004), free oxygen radicals have been proven to regulate Ca2+ and K+ PM channels in Arabidopsis root cells (Demidchik et al. 2003, 2010; Demidchik, Shabala & Davies 2007). In this context, earlier works have shown a direct correlation between copper sensitivity and K+ leakage in Arabidopsis roots (Murphy & Taiz 1997; Murphy et al. 1999). Hydroxyl-induced activation of K+ efflux channels results in massive K+ loss from the cell, leading to PCD via stimulation of K+-dependent cell death proteases and endonucleases (Demidchik et al. 2010). Interestingly, ROS sensitivity of root PM channels showed significant tissue specificity (Demidchik et al. 2007), with elongation zone tissues being much more sensitive to exogenously applied H2O2 and Cu2+/ascorbate (OH-generating mixture; Halliwell & Gutteridge 1999). This difference was explained by the existence of different populations of ROS-sensitive channels in different root zones although other causes could contribute to justify the observed differences.

An alternative explanation for the differential sensitivity of PM K+ and Ca2+ permeable channels to ROS could be that: (1) in plant roots OH may be formed not only in the cell wall but also in the cytosol; and that (2) cytosolic Cu concentration may vary between root apex and mature zone tissues. This hypothesis was tested in this study by electrophysiological and imaging assessment of Arabidopsis seedlings differing in their ability to transport copper across the PM into the cytosol. Wild type (Col-0), COPT1 T-DNA insertion mutant (copt1), and COPT1 constitutively overexpressing plants (C1OE) were compared in order to see whether copper can stimulate OH production and in vivo OH activation of ion channels in Arabidopsis root epidermis. Our results suggest that copper transport into cytosol is a key factor contributing to intracellular OH generation and a consequent regulation of PM OH-sensitive Ca2+ and K+ channels in Arabidopsis roots.


Plant material and growing conditions

Arabidopsis (Arabidopsis thaliana L.) seeds ecotype Columbia-0 wild type (Col-0), COPT1 T-DNA insertion mutant (copt1), and COPT1-overexpressing seeds (C1OE) (with a constitutive CaMV35S promoter; Andrés-Colás et al. 2010) were surface sterilized using a commercial bleach and placed on Petri dishes with half-strength Murashige & Skoog (½ MS) medium (M5524; Sigma-Aldrich, Buchs, Switzerland), 1% (w/v) sucrose (Panreac, Barcelona, Spain), 1.2% (w/v) phytagel (Duchefa, Haarlem, The Netherlands), and pH 5.7 adjusted to KOH. Petri dishes were placed for 48 h in darkness at 4 °C to synchronize germination. Seeds were grown placing the plates vertically in a growing chamber (16 h light-8 h darkness, 25 °C, 78 µmol m−2 s−1 irradiance).

Identification of the copt1 knockout mutant

A collection comprising genomic DNAs from 80000 Arabidopsis lines (ecotype Columbia) (J. M. Alonso, W. Crosby and J. R. Ecker; SALK collection) (, containing independent random insertions of a transposable element (T-DNA), was screened by a PCR-derived method (Krysan et al. 1996) to isolate the copt1 mutant (line 39786). A specific primer for the left border of the T-DNA (L, 5′ GGCAATCAGCTGTTGCCCGTCTCACTGGTG) was used in combination with gene-specific primers flanking the COPT1 ORF (F, 5′ CCTCTCCTCCCAACCAAAACACAAGAGCC and R, 5′ TTAATTTGAATCACAAGCATGGATCAATCC) to identify mutant lines. Insertion was verified by DNA gel blot hybridization and sequencing. Combination of both gene-specific primers was used to identify homozygous lines.

Gene expression analysis by quantitative RT-PCR

Total Arabidopsis RNA was extracted with Trizol Reagent (Invitrogen, Carlsbad, CA, USA). RNA was quantified by ultraviolet (UV) spectrophotometry and its integrity was visually assessed on ethidium bromide-stained agarose gels. Total RNA (1.5 µg) isolated from the indicated organs was first converted to cDNA by reverse transcription using SuperScript II reverse transcriptase (Invitrogen) and anchored oligo(dT)15 (Roche, Mannheim, Germany). Real-time quantitative PCR was carried out with SYBR-Green qPCR Super-Mix-UDG with ROX (Invitrogen) and the specific primers for the UBQ10 and COPT1 genes (AtUBQ10 F, TAATCCCTGATGAATAAGTGTTCTAC; AtUBQ10 R, AAAACGAAGCGATGATAAAGAAG; COPT1 F, TTGCAATTTTCCTCTCCTCCCAA; COPT1 R, ATGATGGTCGAGGCATT) in a StepOnePlus Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) under 1 cycle of 95 °C for 2 min and 40 cycles of 95 °C for 30 s and 60 °C for 30 s. Values were normalized to the UBQ10 mRNA levels.

Microelectrode ion flux measurements

A total of 7- to 11-day-old intact seedlings were removed from the dish by cutting a piece of agar without affecting the roots and placing it in a small Petri dish, and immobilized by gently placing a small microscope slide on the top of the agar block, allowing the apical 5–6 mm of the root to be uncovered. The Petri dish with the immobilized seedling was filled with 20 mL of solution containing 1 mm CaCl2, 0.2 mm KCl, pH 5.7, unbuffered. The seedling was placed on a fluorescence inverted microscope (NIKON Eclipse TE2000-E; NIKON Instruments Europe, Amsterdam, The Netherlands). Electrode tips (with external diameter around 5 µm) were positioned at 50 µm from the root surface at 100 × magnification and measured 45 min after immobilization; the process was visualized and recorded by a digital camera (NIKON digital sight DS-U2 controller; NIKON Instruments Europe) and the software NIS-Elements F 2.30 (National Instruments Spain, Madrid). Electrodes were fabricated using a vertical puller PULL-100 (WPI Europe, Hertfordshire, UK) and silanized with dichlorodimethylsilane (Catalogue No. 40136; Fluka, Sigma-Aldrich, Buchs, Switzerland). Electrode tips were filled with commercial ionophore cocktails (60031 for K+ and 21048 for Ca2+; both from Fluka) and calibrated in a range of standard solutions (see Shabala & Shabala 2002 for details). Electrodes with a low response (less than 50 mV/decade for K+ and 25 mV/decade for Ca2+) or a correlation under 0.999 were discarded. Net ion fluxes were measured by the system developed by Gunsé, Rodrigo & Poschenrieder (2010) at the Autonomous University of Barcelona (see for details). In brief, the movement of the electrodes was controlled by a hydraulic manipulator connected to a stepping motor, with a high precision of displacement (1 µm). The electrode movement was fully computerized using the software NI Motion (National Instruments Spain) and a 2-axis card (NI PCI 7330; National Instruments Spain); this allowed a full control of the movement in both axis X/Y. Electric signal was amplified by an electrometer (FD223a; WPI Europe) and sent to a 16 bit data acquisition/conversion card (A/D NI PCI 6220). This acquisition system allows changes in movement, time, frequency and number of samples prior to the beginning of the experiments. The frequency used for ion flux measurements was 0.2 Hz.

Cell viability assays

Seedlings were harvested after 2 h of control and copper treatments, washed with a phosphate buffer, and double stained using fluorescein diacetate (FDA, 31545; Fluka) and propidium iodide (PI, P4170; Sigma, Sigma-Aldrich, Buchs, Switzerland), according to Koyama et al. (1995) as modified from Jones & Senft (1985). Intact cells exhibit green fluorescence due to FDA, and PI provides red fluorescence of nuclei in damaged cells, as PI has very low penetrability across intact membranes (Hamilton, Habbersewtt & Herman 1980). For the Evans blue assay, seedlings were pre-incubated for 1 h with the different channel blockers and stained with an aqueous Evans blue solution (46160; Fluka) for 15 min, and then washed three times with distilled water for 10 min each, according to Baker & Mock (1994). Cell viability was assayed using a fluorescence inverted microscope (NIKON Eclipse TE2000-E). Fluorescence visualization was assayed using a high pressure mercury lamp power supply (NIKON HB-10101AF, NIKON Instruments Europe) and a filter NIKON B-2A (EX 450–590, DM 505, BA 520).

Peroxide detection

Two hours after treatments, roots were washed in 10 mm Tris-HCl buffer and incubated for 30 min at 37° with 25 µm 2′,7′dichlorofluorescein diacetate (DCF-DA, D6883; Sigma), according to the protocol by Sandalio et al. (2008). As a negative control, roots were incubated 1 h prior to the staining with 1 mm ascorbate (peroxide scavenger), as shown in Rodríguez-Serrano et al. (2006). After incubation, roots were washed and stained as described above. Increasing H2O2 concentrations (0.1, 1 and 10 mm) were stained with the probe as a positive control. Peroxide fluorescence was assayed as described for FDA–PI double staining, and peroxide levels (in arbitrary units) were measured with the software Image-Pro Plus 6.0. (Media Cybernetics Inc., Rockville, MD, USA).

Statistical analysis

Statistical significance between data was estimated using analysis of variance (anova) test with the statistic package Statistica7.0 (StatSoft, Inc., Tulsa, OK, USA).


Isolation of the copt1 T-DNA insertion mutant

To investigate the function of COPT1 in Arabidopsis, we have isolated a T-DNA insertion mutant (copt1) in the Columbia (Col-0) ecotype by using a reverse-genetic screen (Krysan et al. 1996). A library encompassing genomic DNA from 80000 randomly mutagenized T-DNA lines (SALK collection) was screened by PCR techniques (see Materials and Methods). Co-segregation of copt1 with the T-DNA was established by PCR and homozygous plants were obtained. Sequence analysis revealed that the T-DNA is inserted 157 bp downstream of the translation start codon of the COPT1 gene (Supporting Information Fig. S1). COPT1 mRNA levels were quantified in wild type and copt1 mutant lines by quantitative RT-PCR. The specific product corresponding to COPT1 was detected only in wild-type plants, whereas no amplification product was obtained in copt1 mutant (Supporting Information Fig. S1). These results demonstrate that copt1 mutant is defective in COPT1 expression.

Short-term copper exposure causes membrane disruptions in C1OE seedlings

The effect of short-term exposure (2 h) to 10 µm copper was analysed by double staining with FDA and PI (Fig. 1) Col-0. No membrane disruption was observed in copt1 (green in the figure), while in C1OE seedlings damage was localized at the root apex (red). The damage at the root apex of C1OE seedlings was prevented by addition of 10 µm of bathocuproinedisulfonic acid (BCS), a known copper chelator (Sandmann & Böger 1983; Marschner 2012).

Figure 1.

Copper-induced damage to the root apex in C1OE seedlings. Intact seedlings were double stained with fluorescein diacetate (intact cells, green) and propidium iodide (damaged cells, red) after 2 h with the different treatments. A representative image for each treatment is shown (n = 6).

Copper stimulates dose-dependent net Ca2+ influx in the root apex but not at mature root level

Calcium fluxes were measured at the root apex (50–100 µm from the root tip) and mature (∼1 mm) zone regions of Arabidopsis roots using ion-selective microelectrodes. At 15–20 min after 10 µm CuSO4 addition, a transient peak of Ca2+ influx was detected at the root apex of Col-0 (76.5% of tested roots, n = 17) and C1OE (64.3% of tested roots, n = 14) seedlings, but none in 10 copt1 seedlings (Fig. 2a). The amplitude of net calcium influx peak in C1OE seedlings was 7.5-fold higher compared with wild type (49.6 ± 14 and 6.6 ± 1.4 nmol m−2 s−1, respectively, significant at P < 0.01; Fig. 2b). Net calcium influx increased proportionally with increasing copper concentrations applied to roots (Fig. 2c).

Figure 2.

Copper induces Ca2+ influx at the root apex of Col-0 and C1OE seedlings. (a) Transient flux kinetics upon addition of 10 µm CuSO4 (indicated by an arrow). Error bars are SEM). (b) Peak calcium influx measured in mature and elongation zones in response to 10 µm CuSO4 addition (n = 6–13). Asterisks indicate statistically significant differences with respect to the wild type, analysis of variance (anova) test (**P < 0.01). (c) Dose-dependency of Ca2+ flux response.

Only slight Ca2+ influx was detected in the mature zone of C1OE seedlings (Fig. 2b), but this was not significant at P < 0.05 (n = 9). Furthermore, no apparent changes in net Ca2+ flux kinetics were seen around the mature region of either Col-0 or copt1 roots (n = 10 for Col-0, and n = 11 for copt1; neither is significant at P < 0.05).

Ca2+ net flux is deregulated in C1OE seedlings grown under neutral photoperiod

Spatiotemporal acquisition of copper is fundamental for A. thaliana development, and C1OE seedlings have been shown to have increased Cu accumulation and deregulated Cu transport. In addition, these transgenics display phenotypes reminiscent of circadian clock mutants (Andrés-Colás et al. 2010). The mechanisms of this process remain elusive. One possibility is that calcium could be an intermediate in the putative pathway between cytosolic copper and circadian clock components. Importantly, basal net Ca2+ flux before Cu addition was lower in C1OE seedlings when compared with the wild type (Fig. 2a). Accordingly, basal Ca2+ fluxes were analysed at the root apex of 7-day-old seedlings grown under neutral photoperiod at Zeitgeber times 0 (T0) and 12 (T12) (e.g. 0 and 12 h after lights on) (Fig. 3). At T0, differences in basal Ca2+ flux between Col-0 and C1OE seedlings had statistical significance at P < 0.01, anova test. No statistically significant difference in Ca2+ flux levels was detected for T12 grown plants, consistent with the above hypothesis of Ca2+ involvement in deregulated copper transport under neutral photoperiod.

Figure 3.

Calcium basal fluxes are deregulated in C1OE seedlings. Basal net calcium flux was measured in seedlings grown under neutral photoperiod at Zeitgeber times 0 and 12. Mean ± SE (n = 5–8 seedlings). Asterisks indicate statistically significant differences with respect to the wild type [**P < 0.01, analysis of variance (anova) test].

Pharmacology of copper-induced Ca2+ influx

The effect of different blockers on calcium net flux was investigated. Seedlings were pre-incubated with either 10 mm tetraethylammonium chloride (TEA+), a potassium channel blocker, 50 µm gadolinium (Gd3+), a known blocker of non-selective cation channels (NSCCs) (Demidchik et al. 2003), or 30 µm verapamil (capable of blocking both KOR and NSCC; Terry, Findlay & Tyerman 1992). In Col-0 plants, TEA+ had no effect on calcium flux, while both gadolinium and verapamil caused significant blockage of the calcium flux response (Fig. 4a). Gadolinium completely suppressed calcium flux (from 6.6 ± 1.4 to 0.6 ± 0.4 nmol m−2 s−1, n = 4; significant at P < 0.01), while verapamil caused approximately 50% reduction in copper-induced Ca2+ influx (n = 4, significant at P < 0.05). Application of the above concentrations of channel blockers to C1OE seedlings resulted in reduction in net Ca2+ influx similar to that observed for Col-0 seedlings (n = 4, Fig. 4b).

Figure 4.

Pharmacology of copper-induced Ca2+ flux responses. (a) Gd3+ and verapamil block net calcium influx at the root apex of wild-type seedlings. Seedlings where pretreated with 0.05 mm Gd3+, 0.02 mm verapamil or 10 mm TEA+ 1 h prior 10 µm CuSO4 addition. Mean ± SE (n = 4–8). (b,c) Effect of Gd3+, TEA+ and verapamil on net Ca2+ influx at the root apex of C1OE (b) and copt1 (c) seedlings. Seedlings were pretreated with 0.05 mm Gd3+, 10 mm TEA+ or 0.02 mm verapamil 1 h prior to 10 µm (for C1OE) or 30 µm (for copt1) CuSO4 addition. Mean ± SE (n = 4). Asterisks indicate statistically significant differences with respect to the wild type [*P < 0.05; **P < 0.01, analysis of variance (anova) test].

As COPT-silenced seedlings show a phenotype similar to wild type when treated with 30 µm copper (Sancenón et al. 2004), the effect of this concentration was tested at the root apex of copt1 seedlings. Under these conditions, the Ca2+ influx response, absent when copt1 roots were treated by 10 µm of copper, becomes evident. In this case, the Ca2+ influx peak had a magnitude comparable to that for Col-0 treated with 30 µm copper (as reported in Fig. 3). Pharmacology experiments revealed that the calcium influx was also blocked in a similar way to that observed in wild-type seedlings (Fig. 4c).

Copper induces K+ efflux in C1OE seedlings

Addition of 10 µm of copper to Arabidopsis root apex also caused net K+ efflux. This efflux was relatively small in Col-0 seedlings (–13.1 ± 9 nmol m−2 s−1, n = 4), but more than fivefold higher in C1OE seedlings (–84.10 ± 27.92 nmol m−2 s−1, n = 5; difference significant at P < 0.05) (Fig. 5). The K+ efflux at the root apex of C1OE was suppressed by pre-incubation in Gd3+ (n = 4), but not in verapamil (n = 3), or TEA+ (n = 5) (see inset in Fig. 5).

Figure 5.

Copper induces K+ efflux at the root apex of C1OE seedlings. Intact seedlings were placed in a chamber with 0.1 mm CaCl2 and 0.2 mm KCl, pH 5.7. A concentration of 10 µm CuSO4 was added at around 10 min (as indicated by an arrow). Mean ± SE (n = 5–8). Inset: Gd3+ blocks potassium net efflux at the root apex of C1OE seedlings. Seedlings were pretreated with 0.05 mm Gd3+, 0.02 mm verapamil or 10 mm TEA+ 1 h prior 10 µm CuSO4 addition. Mean ± SE (n = 3–5). Asterisks indicate statistically significant differences with respect to the control [*P < 0.05, analysis of variance (anova) test].

Gd3+ prevents root tip damage caused by toxic copper exposure

The effect of different ion channel blockers in preventing root damage under toxic concentrations of copper was tested in Col-0 seedlings by Evans blue assay (Baker & Mock 1994). Two hours of treatment with 50 µm CuSO4 caused substantial damage to root apex (dark blue staining in Fig. 6). Verapamil prevented root tip damage in ∼30% of the roots, while Gd3+ completely prevented the toxic effect of copper. No beneficial effect of TEA+ pretreatment was found. No root damage was observed in the presence of 10 µm of BCS.

Figure 6.

Gd3+ prevents root tip damage caused by copper toxicity in Col-0 seedlings. Seedlings were pretreated with inhibitors (0.02 mm verapamil, 0.05 mm Gd3+ or 10 mm TEA+) 1 h prior to root exposure to 50 µm CuSO4 for 2 h. A representative image for each treatment is shown (n = 8–12).

Basal peroxide levels decrease in copper-treated C1OE root apex

In order to assess whether copper is able to react with peroxide and produce OH, roots were washed and incubated with DCF-DA after 2 h of the different treatments (Fig. 7a). Average basal peroxide accumulation was analysed in the section 0–100 µm from the root apex in plants with copper treatments (Fig. 7b). Copper treatment caused a basal peroxide reduction at the root tip of Col-0 and C1OE but not in copt1 seedlings (Fig. 7c). Peroxide accumulation in this section was statistically higher at the root apex of copt1 seedlings with respect to the control (P < 0.01, n = 6). In addition, C1OE had lower peroxide accumulation than Col-0 (significance at P < 0.05, n = 6).

Figure 7.

Copper treatment causes decrease in basal peroxide accumulation at the root apex of C1OE. (a) DFC-DA stained roots. A representative image for each treatment is shown (n = 6). (b) Fluorescence intensity profile at the root tip of the seedlings under copper treatments shown in (a); the inset shows the green colour of increasing H2O2-containing wells and stained with DFC-DA, without plant material. (c) Average fluorescence intensity at the root tip (0–100 µm) of control and copper-treated seedlings. Mean ± SE (n = 6). Red asterisks mark significant differences with respect to Col-0 [**P < 0.01; *P < 0.05, analysis of variance (anova) test].


PM transport systems that are sensitive to hydroxyl radical are activated from the cytosolic side in planta

To the best of our knowledge, all previous studies reporting in planta OH effects of PM ion channel activity assumed that OH generation occurs in the apoplastic space as a result of interaction between transition metals (such as Cu) and H2O2 in the presence of ascorbate located in the cell walls (Demidchik et al. 2003, 2007). In this study, we provide strong evidence that a similar process occurs also in the cell cytosol, and that OH produced in the latter compartment may have significant impact activities of ion transporters at the PM and intracellular homeostasis in intact plants. Several lines of evidence support this claim.

Firstly, the damage caused by application of a low (10 µm) concentration of copper was observed only in C1OE seedlings, with 2.5-fold expression of COPT1 and an increase in Cu content (Andrés-Colás et al. 2010, see Table 1), and is absent in Col-0 and copt1 seedlings (Fig. 1). Localization of the damage at the root tip in Col-0 seedlings (Fig. 6) is consistent with previously reported COPT1 expression (Sancenón et al. 2004), and prevention of damage with the copper chelator BCS confirms that cell membrane disruptions are copper induced.

Secondly, significant differences between Cu-induced net Ca2+ and K+ fluxes at the root apex were found between most C1OE and Col-0 roots (Figs 2 & 5). These differences are consistent with the expression at the root apex of the COPT1 transporter (Sancenón et al. 2004), but cannot be explained only by COPT1 expression, as the expression in C1OE is driven by a constitutive promoter. Different explanations can be put forward. Previous studies have reported differential sensitivity along the root of Ca2+ and K+ permeable PM ion channels to ROS (Demidchik et al. 2002, 2003, 2007), but these differences could also be explained by the fact that internal cell H2O2/OH ratio or some signalling component responsible for channel activation after copper uptake is root apex specific.

Thirdly, basal peroxide (H2O2) detection reveals a decrease at the root apex of C1OE under copper treatment (Fig. 7a–c), being consistent with this cytosolic OH generation from copper reaction with H2O2. The decrease in peroxide at the mature zone of C1OE and copper-treated Col-0 seedlings could be due to a lignification process, as has been reported previously in Arabidopsis roots under copper treatment (Lequeux et al. 2010).

Fourthly, all previous studies on hydroxyl-activated ion channels have used non-physiologically high concentration of OH-generating Cu/ascorbate mix (Demidchik et al. 2002, 2010; Zepeda-Jazo et al. 2011). In this study, fluxes were measured in response to 100 times lower copper addition, 10 µm, and in the absence of ascorbate. Both Ca2+ and K+ transient peaks occurred 15–25 min after copper addition, further suggesting that the effect of copper on transporters mediating fluxes of these ions was indirect and occurred from the cytosolic but not the apoplastic side.

Physiological implications and underlying mechanisms

During root growth, the root apex is the zone that first makes contact with a new soil environment. A higher sensitivity to OH in this root zone may provide an efficient mechanism for signalling changes in the chemical properties in the rhizosphere. OH-induced Ca2+ signalling appears to be essential in this process. Ca2+ flux responses were absent in copt1 mutant treated with 10 µm Cu (Fig. 2), but appeared when plants were treated with higher (30 µm) copper (Fig. 4c). This is consistent with reports that COPT-silenced seedlings show a phenotype similar to wild-type plants when treated with higher Cu concentrations (Sancenón et al. 2004; see Table 1), suggesting a causal relationship between Cu transport and Ca2+ flux changes. The most likely explanation may be that, at lower Cu concentrations, COPT1 is the only transporter mediating Cu uptake into the cell, while at higher concentrations Cu uptake may be carried by some other transporter such a ZIP2. Regardless of whether copper is transported through COPT1 or through another transporter, the effect on activating calcium influx is similar. A moderate OH generation would generate a transient Ca2+ influx necessary for root tip growth. Previous work on COPT1 has shown that copper has a direct effect on the circadian clock components (Andrés-Colás et al. 2010). In addition, Cu fluxes measured in Olea europaea roots have spatial and temporal oscillations (Papeschi, Mancuso & Marras 2000). C1OE seedlings, with modified Cu uptake pattern, have deregulated growth (Andrés-Colás et al. 2010) that could be explained by these differences in basal Ca2+ influx (Fig. 3). Neither Ca2+ influx nor K+ efflux was affected by TEA+ (Figs 4 & 5), suggesting that K+-selective channels were not part of responses to Cu treatment. At the same time, both gadolinium and verapamil, two known NSCC blockers, were highly efficient (Figs 4–6), implicating NSCC as a downstream target of Cu-signalling mechanism. Earlier, Demidchik et al. (2003) reported OH-activated Cain currents mediated by NSCC that had a pharmacological profile similar to our data (insensitivity to TEA+ and sensitivity to lanthanides and verapamil). The other supporting evidence for NSCC involvement may be found in the fact that verapamil has affected fluxes of both Ca2+ and K+ (Fig. 5), consistent with the idea that they may be mediated by the same NSCC channel, as reported elsewhere (White & Davenport 2002). It has been shown in direct patch-clamp experiments that outward-rectifying depolarization-activated K+ permeable GORK channels are also activated by hydroxyl radicals (Demidchik et al. 2010). Both OH-activated Kout currents and net K+ fluxes in the mature root zone of Arabidopsis were sensitive to TEA+ (Demidchik et al. 2003, 2010; Cuin & Shabala 2007). In all these reports, high (1 mm) concentration of a Cu/ascorbate mix was used. In our study, TEA+ did not affect K+ efflux when plants were treated with much lower (10 µm) Cu concentrations. In addition, the K+ flux kinetics was rather different. While in all the above referred papers the peak K+ efflux was achieved within 5 and 7 min of 1 mm Cu/ascorbate application, maximum K+ efflux in our work was observed 25–30 min after 10 µm Cu treatment. This difference may be explained assuming that GORK channels have OH binding sites on the external side (and, hence, activated from the apoplast) while NSCC channels have their OH binding sites on the cytosolic side. If this is the case, it will take longer to activate NSCC channels because Cu must cross the PM to generate OH in the cytosol. It should be noted that such activation of NSCC by hydroxyl radical generation by the Fenton reaction at the cytosolic side has been shown previously for animal cells (Simon et al. 2004).

Copper homeostasis model under deprivation and toxic conditions

Based on the above data, the following model describing copper homeostasis and hydroxyl radical generation/signalling can be put forward (Fig. 8). Under limiting copper conditions, (1) the expression of COPT1 is induced. This high-affinity transporter is located at the PM of root tips and (2) mediates copper transport into the cytosol (Sancenón et al. 2004; Andrés-Colás et al. 2010). Once in the cytosol, copper will be transported by metal chaperones to major intracellular compartments for essential functions in the cells. Part of the copper would generate (3) a moderate OH increase, which activates (4) a moderate Ca2+ influx necessary for root growth. Once this signalling is over, elevated cytosolic calcium would be reverted by the orchestrated action of Ca2+ exchangers and Ca2+-ATPases at the PM and tonoplast. At the same time, the increase in [Cu]cyt (6) regulates COPT1 expression via negative feedback (Sancenón et al. 2004; Peñarrubia et al. 2010) (illustrated in Fig. 8 as black lines).

Figure 8.

The tentative model depicting hydroxyl radical production and activation of plasma membrane Ca2+ and K+ permeable channels in Arabidopsis copper-transport mutants. See text for explanations. Black lines relate to signalling under copper sufficient condition; grey lines refer to copper toxicity. Abbreviations used: COPT1, high-affinity copper transporter 1; ZIP2, Zrt-, Irt-like protein 2; NSCC, non-selective cation channel; NADP(H) oxidase, nicotinamide adenine dinucleotide phosphate-oxidase; SPL7, SQUAMOSA promoter binding protein-like7; PCD, programmed cell death.

If excessive copper supply follows copper deprivation, COPT1 and perhaps other transporters such as ZIP2 (2) would mediate copper entry towards the cytosol (grey lines in Fig. 8). The resultant increase in cytosolic Cu pool would result in (3) a massive generation of OH due to Cu interaction with cytosolic H2O2. This will result in (4) activation of OH-sensitive Ca2+ and K+-permeable NSCC. The effect will be twofold. Firstly, such activation may further increase net Ca2+ uptake into cytosol via (5) positive feedback regulation of NADP(H)oxidase (Lecourieux et al. 2002). Secondly, OH-induced K+ efflux will result (6) in the rapid decrease in [K+]cyt pool, (7) activating caspase-like proteases and leading to PCD (Yu et al. 1997; Yu 2003; Shabala 2009b; Demidchik et al. 2010). This will explain the rapid loss of viability in C1OE plants as indicated by viability staining experiments (Fig. 1).


This work was supported by the Spanish MICINN (Projects BFU2007-60332 and BFU2010-14873). A.R-M. acknowledges her PhD fellowship from Ministerio de Ciencia e Innovación (BES-2008–005096). L.P. and S.S. would like to acknowledge the financial support from BIO2011-24848 and ARC (Discovery Project DP1094663), respectively. The authors would like to thank Joseph R. Ecker (The Salk Institute for Biological Studies, La Jolla, CA) and Vicente Sancenón Galarza who obtained the copt1 mutant and localized T-DNA insertion.