The Arabidopsis vacuolar anion transporter, AtCLCc, is involved in the regulation of stomatal movements and contributes to salt tolerance

Authors

  • Mathieu Jossier,

    1. Institut des Sciences du Végétal, Centre National de la Recherche Scientifique, 1 Avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex, France
    2. Université Paris 7 Denis Diderot, Unité de Formation et de Recherche Sciences du Vivant, 35 rue Hélène Brion, 75205 Paris Cedex 13, France
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Laëtitia Kroniewicz,

    1. Laboratoire des Echanges Membranaires et Signalisation, Unité Mixte de Recherche 6191, Centre National de la Recherche Scientifique/Commissariat à l’Energie Atomique, Université Aix-Marseille II, Commissariat à l’Energie Atomique Cadarache Batiment 156, 13108 St Paul-lez-Durance, France
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Fabien Dalmas,

    1. Laboratoire des Echanges Membranaires et Signalisation, Unité Mixte de Recherche 6191, Centre National de la Recherche Scientifique/Commissariat à l’Energie Atomique, Université Aix-Marseille II, Commissariat à l’Energie Atomique Cadarache Batiment 156, 13108 St Paul-lez-Durance, France
    Search for more papers by this author
  • D. Le Thiec,

    1. Institut National de la Recherche Agronomique, Nancy Université, Unité Mixte de Recherche1137 Ecologie et Ecophysiologie Forestières, Institut Fédératif de Recherche 110, EFABA, F-54280 Champenoux, France
    Search for more papers by this author
  • Geneviève Ephritikhine,

    1. Institut des Sciences du Végétal, Centre National de la Recherche Scientifique, 1 Avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex, France
    2. Université Paris 7 Denis Diderot, Unité de Formation et de Recherche Sciences du Vivant, 35 rue Hélène Brion, 75205 Paris Cedex 13, France
    Search for more papers by this author
  • Sébastien Thomine,

    1. Institut des Sciences du Végétal, Centre National de la Recherche Scientifique, 1 Avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex, France
    Search for more papers by this author
  • Hélène Barbier-Brygoo,

    1. Institut des Sciences du Végétal, Centre National de la Recherche Scientifique, 1 Avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex, France
    Search for more papers by this author
  • Alain Vavasseur,

    1. Laboratoire des Echanges Membranaires et Signalisation, Unité Mixte de Recherche 6191, Centre National de la Recherche Scientifique/Commissariat à l’Energie Atomique, Université Aix-Marseille II, Commissariat à l’Energie Atomique Cadarache Batiment 156, 13108 St Paul-lez-Durance, France
    Search for more papers by this author
  • Sophie Filleur,

    Corresponding author
    1. Institut des Sciences du Végétal, Centre National de la Recherche Scientifique, 1 Avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex, France
    2. Université Paris 7 Denis Diderot, Unité de Formation et de Recherche Sciences du Vivant, 35 rue Hélène Brion, 75205 Paris Cedex 13, France
      (fax +33 1 69 82 37 68; e-mail filleur@isv.cnrs-gif.fr or fax +33 4 42 25 23 64; e-mail nathalie.leonhardt@cea.fr).
    Search for more papers by this author
  • Nahalie Leonhardt

    Corresponding author
    1. Laboratoire des Echanges Membranaires et Signalisation, Unité Mixte de Recherche 6191, Centre National de la Recherche Scientifique/Commissariat à l’Energie Atomique, Université Aix-Marseille II, Commissariat à l’Energie Atomique Cadarache Batiment 156, 13108 St Paul-lez-Durance, France
      (fax +33 1 69 82 37 68; e-mail filleur@isv.cnrs-gif.fr or fax +33 4 42 25 23 64; e-mail nathalie.leonhardt@cea.fr).
    Search for more papers by this author

(fax +33 1 69 82 37 68; e-mail filleur@isv.cnrs-gif.fr or fax +33 4 42 25 23 64; e-mail nathalie.leonhardt@cea.fr).

Summary

In plant cells, anion channels and transporters are essential for key functions such as nutrition, resistance to biotic or abiotic stresses, and ion homeostasis. In Arabidopsis, members of the chloride channel (CLC) family located in intracellular organelles have been shown to be required for nitrate homeostasis or pH adjustment, and previous results indicated that AtCLCc is involved in nitrate accumulation. We investigated new physiological functions of this CLC member in Arabidopsis. Here we report that AtCLCc is strongly expressed in guard cells and pollen and more weakly in roots. Use of an AtCLCc:GFP fusion revealed localization to the tonoplast. Disruption of the AtCLCc gene by a T-DNA insertion in four independent lines affected physiological responses that are directly related to the movement of chloride across the tonoplast membrane. Opening of clcc stomata was reduced in response to light, and ABA treatment failed to induce their closure, whereas application of KNO3 but not KCl restored stomatal opening. clcc mutant plants were hypersensitive to NaCl treatment when grown on soil, and to NaCl and KCl in vitro, confirming the chloride dependence of the phenotype. These phenotypes were associated with modifications of chloride content in both guard cells and roots. These data demonstrate that AtCLCc is essential for stomatal movement and salt tolerance by regulating chloride homeostasis.

Introduction

The chloride channel (CLC) family are ubiquitous proteins that are present in prokaryotes and eukaryotes, and, unusually, comprise both channels and transporters (Mindell and Maduke, 2001; Jentsch, 2008). Seven AtCLC genes (AtCLCa–g) have been identified in the Arabidopsis genome (Hechenberger et al., 1996; Lv et al., 2009). Intracellular localization of some of these proteins shows that AtCLCs are present in various membranes including the vacuolar membrane (AtCLCa and AtCLCb) (De Angeli et al., 2006; von der Fecht-Bartenbach et al., 2010), Golgi vesicles (AtCLCd and AtCLCf) or chloroplast membranes (AtCLCe) (von der Fecht-Bartenbach et al., 2007; Marmagne et al., 2007). Physiological characterization of Arabidopsis mutants suggested the involvement of AtCLCa, AtCLCb and AtCLCc in the regulation of nitrate levels in planta (Geelen et al., 2000; Harada et al., 2004; von der Fecht-Bartenbach et al., 2010). AtCLCa is located at the tonoplast in Arabidopsis mesophyll cells. It has been shown to function as a 2NO3/1H+ antiporter that is able to specifically accumulate nitrate into the vacuole. In agreement, knockout mutant plants accumulate 50% less nitrate than wild-type plants (Geelen et al., 2000; De Angeli et al., 2006). Thus AtCLCa plays a major role in nitrate homeostasis, probably partly in cooperation with AtCLCe, which is localized in the chloroplast compartment (Monachello et al., 2009). It has also been proposed that AtCLC proteins located in intracellular organelles participate in establishment of an acidic intra-organellar pH, and AtCLCd has been shown to be targeted to the trans-Golgi network (TGN), to co-localize with V-type ATPase, and to be involved in adjustment of the luminal pH of this compartment (von der Fecht-Bartenbach et al., 2007).

AtCLCa plays a role in the translocation of nitrate into the vacuole, but all the other characterized CLCs in bacteria and mammals are involved in chloride transport. It was recently shown that this difference in selectivity is due to an amino acid change in the selectivity filter (Bergsdorf et al., 2009; Wege et al., 2010). Three of other AtCLCs, AtCLCc, AtCLCd and AtCLCg, possess a selectivity filter in favour of chloride transport (Zifarelli and Pusch, 2009a). It has been suggested that these proteins could be involved in physiological phenomena that require Cl transport, such as stomatal movement or salt tolerance (Hansch and Mendel, 2009; Teakle and Tyerman, 2010).

Stomata comprise two highly specialized guard cells located on the aerial organs of plants. They play an essential role in controlling gaseous exchange for adaptation to environmental conditions. In the past decade, our knowledge of guard cell signalling has significantly increased. Investigations on the osmotic changes driving guard cell movement have mainly focused on the role of plasma membrane-associated ion channels and transporters and on signalling elements regulating these transport systems. Stomatal opening and closure mainly occur in response to K+ fluxes across the plasma membrane that allow swelling and shrinking of guard cells (for review, see Very and Sentenac, 2003; Fan et al., 2004). However, these K+ fluxes must be accompanied by fluxes of anions such as Cl, NO3 and malate to counterbalance the charges. Anion channels at the guard cell plasma membrane have been studied in detail, but little is yet known about their molecular identity. Several anion transporters have been found to be involved in stomatal movement, such as AtNRT1.1 (CHL1), which is at least partly responsible for NO3 uptake (Guo et al., 2003), SLAC1, which controls guard cell anion homeostasis (Negi et al., 2008; Vahisalu et al., 2008, 2010), and AtABCB14, a malate transporter (Lee et al., 2008).

In contrast to the plasma membrane, knowledge of the changes occurring in intracellular compartments of guard cells during stomatal movements is less extensive, even though their volume can change by more than 40%. Description of the dynamic changes in vacuolar configuration during stomatal movement showed that a great number of small vacuoles are present in guard cells of closed stomata, and only a few large ones are present in guard cells of fully opened stomata (Couot-Gastelier et al., 1984; Gao et al., 2005; Tanaka et al., 2007). These observations suggest the important roles of the vacuole and ion fluxes across the tonoplast during stomatal movements. Until now, only cation channel activities have been identified at the tonoplast, including fast vacuolar (FV), slow vacuolar (SV) and K+-selective vacuolar (VK) cation channels (Ward and Schroeder, 1994; Pei et al., 1996; Bihler et al., 2005; Peiter et al., 2005). Little is known regarding anion channels at the tonoplast, and only MRP5, an inositol hexakisphosphate transporter, has been shown to be involved in stomatal movements (Nagy et al., 2009).

AtCLCs may also be involved in responses to salt stress. The effect on plants of NaCl, the most widespread salt in soil, has been well described. NaCl stress involves two distinct phases, a rapid osmotic phase in response to the osmotic effect of salt, followed by a slow ionic phase resulting from toxic accumulation of sodium in the cytoplasm. One strategy for tolerance requires compartmentalization of sodium and chloride inside the cell. Studies on NaCl tolerance have mainly focused on the cation (Na+) component of this stress at the expense of anion component (Cl), as Na+ reaches a toxic concentration in many species before Cl does (for review, see Munns and Tester, 2008). Although Cl is an essential micronutrient for higher plants (White and Broadley, 2001; Hansch and Mendel, 2009), it can become toxic at high concentrations if accumulated in the cytoplasm. Indeed, for some species such as soybean (Glycine max), citrus (Citrus spp.) and grapevine (Vitus spp.), Cl is more toxic than Na+ (Läuchli, 1984; Storey and Walker, 1999). Nevertheless, whatever the species analysed, Na+ and Cl must both be sequestered in the vacuole to detoxify the cytoplasm. Currently, only vacuolar cation transporters have been identified as being involved in salt tolerance, such as members of the NHX gene family (Yokoi et al., 2002; Apse et al., 2003), and little is known about the vacuolar anion component of salt tolerance.

In this study, we report evidence that a member of the CLC family, AtCLCc, is involved in stomatal movements and salt tolerance in Arabidopsis thaliana. The AtCLCc protein is localized in the tonoplast, and AtCLCc is highly expressed in guard cells and up-regulated by ABA and salt treatment in the whole plant. Four T-DNA mutants in AtCLCc in two accessions (WS and Col-0) showed impaired light-induced stomatal opening and ABA-induced stomatal closure. These alterations were associated with modifications of the chloride content in guard cells. The clcc mutants also exhibited a hypersensitive phenotype to salt stress compared to wild-type. Together, these results demonstrate that AtCLCc plays an important role in regulation of stomatal movements and is involved in salt tolerance through participation in anion homeostasis.

Results

AtCLCc is strongly expressed in guard cells and pollen but weakly in roots

The expression level of AtCLCc mRNAs was determined in leaves, flowers and pollen from soil-grown plants and germinated seeds and roots from in vitro cultures. Quantitative reverse transcription PCR showed transcript accumulation mainly in aerial organs, with maximal amounts in pollen (Figure 1a). To further characterize the AtCLCc expression pattern, its spatial expression was explored. A genomic region comprising 1437 bp of the AtCLCc promoter upstream of the start codon was fused in-frame with the uidA reporter gene, and introduced into Arabidopsis wild-type plants. All T3 plants obtained from eight independent AtCLCc::GUS transgenic lines showed a similar qualitative pattern of uidA expression that corresponded to the quantitative reverse transcription PCR data. In all transgenic lines, GUS staining was predominantly detected in guard cells and pollen (Figure 1b–d), showing that the AtCLCc promoter was strongly active in both cell types, and corroborating the data available at http://www.genevestigator.com. Low levels of GUS staining were also observed in the maturation zone (Figure 1e,f) and root tip (Figure 1e,h) of plants grown vertically in sterile agar plates. No staining was detected between these two zones (Figure 1g).

Figure 1.

 Tissue-specific expression of AtCLCc.
(a) Quantitative reverse transcription-polymerase chain reaction analysis of AtCLCc expression in leaves, flowers, germinated seeds, pollen and roots.
(b–h) Histochemical analyses of representative transgenic Arabidopsis thaliana plants expressing the PAtCLCc:GUS construct. Micrographs of (b) a 5-day-old seedling, (c) part of a mature leaf, showing strong staining in guard cells, (d) the stamen, showing strong staining in pollen grains, and (e–h) 7-day-old seedlings, showing staining in the primary root and the root cap.

AtCLCc is localized in the tonoplast

To determine the subcellular localization of AtCLCc, green fluorescent protein (GFP) was fused to its N-terminus, and the fusion protein was transiently expressed in protoplasts from Arabidopsis cell suspensions under the control of the CaMV 35S promoter. Confocal imaging showed that labelling by the GFP:AtCLCc fusion coincided with the tonoplast (Figure 2a–c). Stable transformation of A. thaliana with the same construct confirmed tonoplast localization in stomata (Figure 2d–f).

Figure 2.

 Subcellular localization of AtCLCc in protoplasts and leaf epidermis of Arabidopsis.
(a) Confocal cross-section through an Arabidopsis protoplast from a cell suspension transiently transformed with an N-terminal GFP translational fusion of AtCLCc.
(d) Confocal cross-section of Arabidopsis leaf epidermis of transgenic lines expressing 35S::GFP-CLCc. (b,e) Merged images; (c,f) transmission images. Scale bars = 10 μm.

Identification of atclcc null mutants

To determine the function of AtCLCc in planta, four homozygous T-DNA insertion lines were isolated. Two, clcc-1 and clcc-2, were in the Col-0 ecotype, and were obtained from the Syngenta Arabidopsis Insertion Library (SAIL). The other two, clcc-3 and clcc-4, were in the WS ecotype, and were obtained from the INRA Versailles collection. To confirm the T-DNA insertion sites and select homozygous lines, PCR-based screening was performed using AtCLCc-specific primers and T-DNA primers. All T-DNA insertions were located within the coding region of AtCLCc (Figure S1a,c). The alleles clcc-1, clcc-3 and clcc-4, possess insertions in the 6th, 3rd and 1st exons, respectively, which encode the ‘voltage CLC’ domain that constitutes the pore of the CLC channel family (Meyer and Dutzler, 2006). The second allele, clcc-2, has an insertion in the 6th intron between exons that correspond to the ‘cystathionine β-synthase’ (CBS ) domain in the C-terminal part of the protein. This regulatory domain, which has been described in numerous proteins (reviewed by Ignoul and Eggermont, 2005), is only present in eukaryotic CLC proteins (Estevez et al., 2004; De Angeli et al., 2009). No full-length transcript was detected for clcc-2 (Figure S1b,d). The resistance to BASTA® of progeny obtained from self-pollinated heterozygous plants segregated at a ratio of 3:1 for all alleles, indicating a single T-DNA insertion in each of the four mutants. Under our growth conditions on soil, no difference in macroscopic phenotype was observed between homozygous clcc lines and the corresponding wild-type plants, indicating that AtCLCc gene disruption does not affect the general development of the plant. As AtCLCc is strongly expressed in pollen, the pollen fertility was investigated. We took advantage of the fact that the disrupting T-DNA carries a copy of the bar gene, which confers resistance to phosphinothricin (PPT). F1 homozygous wild-type plants that do not contain the transforming T-DNA have been identified as sensitive to the antibiotic. In the F1 progeny, hemizygous for the bar transgene, the mean percentage of homozygous wild-type plants, sensitive to PPT, was close to 25%, as expected for Mendelian segregation of a single recessive trait. These results suggest that the pollen fertility is not affected in the atclcc-1 mutant.

AtCLCc plays a role in stomatal function

Studies of AtCLCc promoter activity (Figure 1) and analysis of the guard cell transcriptome (Leonhardt et al., 2004) revealed that AtCLCc is highly expressed in guard cells. To confirm these results, quantitative reverse transcription PCR experiments were performed using RNAs prepared from highly purified guard cells and mesophyll cells from Arabidopsis plants treated or not with 100 μm ABA for 4 h. As shown in Figure 3(a), the expression level of AtCLCc was very high in guard cells compared to mesophyll cells. In addition, ABA treatment led to a significant increase in AtCLCc expression in guard cells. To further examine the role of AtCLCc in stomatal function, stomatal apertures were measured. After 2 h illumination, the stomatal apertures were 25–30% higher in wild-type plants compared to atclcc mutants (Figure 3b). Moreover, all clcc mutant plants were less sensitive to ABA than wild-type plants even at 100 μm ABA (Figure 3b), but the size of the guard cells and the stomatal density were similar between mutants and wild-type plants (data not shown).

Figure 3.

 AtCLCc expression, stomatal movements and leaf temperature of clcc mutants plants and wild-type.
(a) Expression of AtCLCc in wild-type plants (Col-0) was measured by quantitative reverse transcription-polymerase chain reaction using specific primers for RNA isolated from guard cells (GC) or mesophyll cells (MC) in response to a 4 h 100 μm ABA treatment. Asterisks indicate statistically significant differences between treated and control conditions (< 0.05, Student’s t test).
(b) ABA-induced stomatal closure. Leaf epidermis from wild-type or clcc mutants was incubated in stomatal opening solution for 3 h in the light. Then 10 or 100 μm ABA was added, and stomatal aperture was measured 2 h after treatment. Values are means ± SEM of three independent repetitions (= 8). Asterisks indicate a lack of statistically significant differences between mutants for each condition (< 0.05, Student’s t-test).
(c) Leaf temperature is affected in the AtCLCc KO mutants. Infra-red images of leaves from all clcc mutants (clcc-1, clcc-2, clcc-3 and clcc-4), the complemented mutant (clcc-4 35S::CLCc) and control plants (Col and WS) were captured by an infra-red thermography device immediately after excision during daylight.

Further confirmation that AtCLCc is involved in stomatal opening was obtained by measuring leaf temperature. Infra-red thermography allows measurement of the leaf temperature, which is representative of the transpiration flux and consequently the stomatal aperture (Figure 3c). Leaves were excised during the light period in the growth chamber under well-watered conditions. The mean leaf temperatures for clcc-1 (24.79°C ± 0.03) and clcc-2 (24.61°C ± 0.08) mutants were approximately 0.5°C higher than in the Col-0 wild-type (23.80°C ± 0.45). Similar differences were observed for clcc-3 and clcc-4 compared to the WS wild-type (data not shown). However, the leaf temperatures of the AtCLCc over-expressing plants in the clcc-4 mutant background (clcc-4 35S:CLCc) were similar to those of the wild-type (WS). These observations confirm that stomatal apertures of the clcc mutant plants were smaller than those of their corresponding wild-type plants. Then the stomatal closure kinetics of excised leaves were estimated by measuring the leaf temperature every 10 min for 1 h. In wild-type plants, an increase in leaf temperature was observed during the first 30 min of the experiment due to closure of stomata to limit dehydration of the leaf. This response was strongly decreased in knockout plants. Together, these results show that invalidation of AtClCc alters the stomatal responses to light and ABA.

Because CLC genes encode channels or anion/H+ exchangers that are selective for chloride and/or nitrate, the stomatal aperture was examined in the presence or absence of chloride or nitrate in wild-type and clcc mutants. Light-induced stomatal opening was measured in the presence of KNO3 or KCl. All clcc mutant alleles showed significantly impaired stomatal opening in white light when the leaf epidermis was incubated with 30 mm KCl (Figure 4). This deficiency was dependent on chloride, as no significant difference in stomatal opening was observed between the clcc-3 mutant and wild-type (WS) when Cl was replaced with NO3, and the difference between clcc-1 and Col-0 was strongly reduced. Thus, AtCLCc contributes substantially to the light-induced opening of stomata when chloride is provided, but is not necessary in the presence of nitrate.

Figure 4.

 Measurement of stomatal movements in clcc mutants in response to chloride and nitrate.
The stomatal apertures of epidermal peels of wild-type (Col-0 and WS) or clcc mutants (clcc-1 and clcc-3) were measured after 3 h incubation under light in 30 mm KCl or 30 mm KNO3. Values are means ± SEM of three independent repetitions (= 5). Asterisks indicate statistically significant differences between mutants and wild-type for each condition (< 0.05, Student’s t-test).

Further confirmation that AtCLCc contributes to chloride homeostasis in stomata was obtained by examining the Cl and K+ contents in intact guard cells using X-ray microanalyses. The Cl and K+ contents were determined for intact closed stomata in normal air after 16 h darkness (Table 1). Chloride was much less abundant in the guard cells of the clcc-3 and clcc-4 mutants compared to wild-type, but no significant variation was observed for K+. In addition, the nitrate contents were analysed and no significant change was observed (data not shown). These results show that AtCLCc plays a major role in Cl homeostasis in guard cells during stomatal movements.

Table 1.   Measurement of ion contents in guard cells of wild-type and clcc mutants
 ChloridePotassium
  1. Element contents (% dry weight) in guard cells of wild-type (WS) and clcc mutants (clcc-3 and clcc-4) were determined by microanalysis on 5-week-old leaves grown on soil. Data are means ± SEM of 32 individual guard cells of two independent plants. Asterisks indicate significant differences (< 0.05) between the clcc mutant lines and the wild-type (Student’s t-test).

WS0.4 ± 0.033.58 ± 0.24
clcc-30.07 ± 0.02*3.54 ± 0.37
clcc-40.07 ± 0.01*4.40 ± 0.59

AtCLCc is involved in NaCl tolerance

In yeast and mammals, CLC genes encode chloride channels or Cl/H+ exchangers. To determine whether AtCLCc is involved in chloride homeostasis under salt stress, we measured the expression of AtCLCc in plants grown in a range of NaCl concentrations. Quantitative reverse transcription PCR experiments were performed on RNA extracted from leaves and roots of wild-type plants (Col-0 and WS) grown for 3 weeks in vitro on half-strength MS only or supplemented with 25 or 50 mm NaCl. AtCLCc expression was significantly increased in the leaves of both wild-type accessions (Figure 5a) in response to 25 and 50 mm NaCl in WS and 50 mm NaCl in Col-0. Induction was also observed in roots in response to 50 mm NaCl in the WS accession only.

Figure 5.

 Involvement of AtCLCc in NaCl tolerance.
(a) Induction of AtCLCc expression in leaves and roots of wild-type plants (Col-0 and WS) grown in vitro for 3 weeks on half-strength MS only or supplemented with 25 or 50 mm NaCl. Asterisks indicate statistically significant differences between treated and control conditions (< 0.05) (Student’s t-test).
(b) Phenotypes of wild-type (WS) and clcc mutant (clcc-3 and clcc-4) plants grown in soil and watered with deionized water only or water containing 200 mm NaCl. Plants are shown at 7 weeks after germination (5 weeks after the start of salinity treatment).
(c) Phenotypes of wild-type (Col-0 and WS), clcc knockout mutants (clcc-1 and clcc-4) and complemented plants (clcc-4 35S::CLCc) grown for 3 weeks in vitro on half-strength MS only or supplemented with 25 or 50 mm NaCl.
(d) Shoot and root fresh weights for the plants shown in (c). Values are means ± SEM of three independent repetitions (= 12). Asterisks indicate statistically significant differences (< 0.05) between mutants and wild-type (Student’s t-test).

To further characterize the involvement of AtCLCc in the NaCl response, the clcc-3 and clcc-4 mutants and WS wild-type were first grown on soil and watered for 5 weeks with deionized water only or water supplemented with 200 mm NaCl. Under these conditions, both mutants exhibited strong hypersensitivity to NaCl compared to the wild-type (Figure 5b). To analyse this phenotype in more detail, in vitro experiments were performed. clcc-1 and clcc-4 showed shoot growth reduction of 20% compared to wild-type (Col-0 and WS, respectively) when grown on half-strength MS containing 50 mm NaCl (Figure 5c,d). At the root level, clcc-4 showed a reduction of 40% in root fresh weight in response to NaCl stress (Figure 5d). A similar result was obtained for clcc-3 (data not shown). This phenotype was abolished in a clcc-4 mutant over-expressing AtCLCc (clcc-4 35S:CLCc) (Figure 5c,d). Althrough the differences between the wild type accessions and the mutants were small, they were reproducible in three independent experiments in all the mutant lines (data not shown). In parallel to the NaCl hypersensitivity, a slight decrease in chloride content in roots (−10%) and an increase in nitrate content (+12%) in comparison with wild-type were observed in clcc-4 (Table 2) and clcc-3 (data not shown) in response to 50 mm NaCl. No difference was found in the shoot or for the Col-0 mutants clcc-1 (Table 2) and clcc-2 (data not shown). As modifications with respect to anion content may be associated with modifications in cation content, measurements of Na and K using the inductively coupled plasma-atomic emission spectroscopy (ICP-AES) technique were performed on the same samples, but no modifications were observed in clcc mutants either in roots or shoots of plants in media containing 0, 25 or 50 mm NaCl (Table S2).

Table 2.   Anion content in roots in response to NaCl in wild-type and clcc mutant plants
 ChlorideNitrate
0 mm25 mm50 mm0 mm25 mm50 mm
  1. Anion contents in roots of clcc mutants (clcc-1 and clcc-3) and their corresponding wild-type (Col-0 and WS) (μmol g−1 FW) were measured on plants grown in vitro for 3 weeks on half-strength MS only or supplemented with 25 or 50 mm of NaCl, as indicated. Values are means ± SE for six individual samples (1–3 plantlets per sample) in two independent experiments. Asterisks indicate statistically significant differences (< 0.05) between the clcc mutant lines and the wild-type (Student’s t-test).

Col-017.0 ± 0.953.6 ± 2.087.8 ± 1.376.5 ± 2.380.0 ± 2.478.5 ± 2.3
clcc-115.4 ± 0.953.4 ± 1.382.9 ± 3.875.8 ± 1.182.7 ± 1.781.2 ± 2.8
WS19.9 ± 1.556.5 ± 1.796.3 ± 3.974.7 ± 1.667.4 ± 1.362.4 ± 2.8
clcc-423.0 ± 1.763.2 ± 6.086.7 ± 0.9*78.2 ± 1.974.9 ± 1.369.7 ± 1.7*

To test whether clcc mutant hypersensitivity to NaCl was specifically due to sensitivity to chloride rather than sodium, wild-type (Col-0 and WS) and mutant (clcc-1 and clcc-4) plants were grown for 3 weeks on half-strength MS supplemented with 25 or 50 mm KCl. Similar results were obtained in response to NaCl and KCl, with clcc-1 and clcc-4 showing decreases in shoot fresh weight of 20 (clcc-1) and 40% (clcc-4) in response to KCl (Figure 6a). In roots, a 30% decrease in fresh weight was observed in clcc-1 (Figure 6a) and clcc-2 (data not shown) in response to 25 mm KCl. A decrease of shoot (30%) and root (20%) fresh weight was also observed in the mutants clcc-4 (Figure 6a) and clcc-3 (data not shown) in response to 50 mm KCl. To confirm that the clcc mutant hypersensitivity to NaCl was not due to the osmotic component of NaCl stress, experiments were performed on medium containing mannitol as an osmotic control. Mannitol is not efficiently metabolized by plants and causes a constant osmotic stress (Arenas-Huertero et al., 2000). Wild-type (Col-0 and WS) and mutant plants (clcc-1 and clcc-4) were grown for 3 weeks on half-strength MS supplemented with 55 or 110 mm mannitol, generating osmolarities of 100 and 155 mOsm, respectively. These osmolarities due to mannitol are similar to those generated by NaCl or KCl in half-strength MS medium at concentrations of 25 and 50 mm, i.e. 120 and 165 mOsm, respectively. No difference was observed for either mutant when experiments were performed on half-strength MS supplemented with mannitol (Figure 6b). This results that the growth decrease may be due preferentially to chloride rather than osmotic stress (Figure 6).

Figure 6.

 Effect of KCl and mannitol treatments on wild-type and clcc plants.
Shoot and root fresh weights were measured for clcc-1 and clcc-4 mutants and their corresponding wild-type, Col-0 and WS, respectively, grown in vitro for 3 weeks on half-strength MS supplied or not with (a) 25 or 50 mm KCl, or (b) 55 or 110 mm mannitol. Values are means ± SEM of three independent repetitions (= 12). Asterisks indicate statistically significant differences (< 0.05) between mutants and wild-type (Student’s t-test).

The expression of the other tonoplastic CLCs are regulated by salt stress, light and ABA in guard cells but are not affected in clcc mutants

AtCLCc belongs to a multigenic family with possible functional redundancy between the seven members. We therefore determined whether other AtCLC genes are regulated at the transcriptional level during stomatal movement in guard cells or salt stress in roots and shoots using quantitative reverse transcription-polymerase chain reaction. Analysis of gene expression in stomata was performed in response to light and ABA treatment of guard cells (Figure 7a). Exposure of plants to light for 4 h resulted in dramatically decreased expression of AtCLCa and AtCLCc in guard cells. In addition, ABA treatment (10 μm ABA sprayed once at the start of the 4 h treatment) slightly increased the expression of AtCLCa. Expression of the remaining genes, AtCLCb, AtCLCd, AtCLCe, AtCLCf and AtCLCg, was almost undetectable, suggesting that mainly AtCLCa and AtCLCc are expressed in guard cells. To analyse the salt-stress response, the plants were grown in vitro as previously described (Figure 5c,d). AtNHX2 was used as a positive control for NaCl treatment as it is a tonoplast protein that is induced in shoots by NaCl (Yokoi et al., 2002). For the CLC family, salt treatment resulted in an increase in expression of AtCLCc, AtCLCd and AtCLCg in shoots (Figure 7b) and of AtCLCb and AtCLCc in roots (Figure 7c). Together, these results suggest possible redundancy between the AtCLCs, but the clcc mutations did not have any effect on the expression in shoots and roots of the other AtCLCs (encoded by AtCLCa, AtCLCb and AtCLCg, Figure S2) that have been previously shown to be localized at the tonoplast (Lv et al., 2009).

Figure 7.

 Expression of AtCLC genes in response to ABA and NaCl treatments in wild-type plants.
Quantitative reverse transcription-polymerase chain reaction analysis of AtCLC genes was performed using specific primers for each gene and RNA isolated from guard cells from plants sprayed or not with 100 μm ABA (a) or RNA isolated from shoots (b) and roots (c) of wild-type plants (WS) grown in vitro for 3 weeks on half-strength MS only or supplemented with 25 or 50 mm NaCl. Asterisks indicate statistically significant differences between treated and control conditions (< 0.05) (Student’s t-test).

Discussion

Anion channel activities and regulation mechanisms have been well characterized using electrophysiological approaches, but identification of the corresponding genes is still in its infancy. Members of the CLC family are the best candidates for encoding plant anion channels. Characterization of mutant plants has provided insight into the function of some CLC proteins in nitrate accumulation and anion homeostasis. In the present study, we characterize new functions of a member of the CLC family, AtCLCc. Using promoter::GUS fusions and reverse transcription quantitative PCR, we found that AtCLCc is strongly expressed in guard cells and pollen and weakly in roots (Figure 1). This finding is supported by a previous study on the guard cell transcriptome (Leonhardt et al., 2004) and expression pattern analyses of all AtCLCs (Lv et al., 2009). In agreement with two tonoplastic proteome analyses (Jaquinod et al., 2007; Whiteman et al., 2008), we have demonstrated, using GFP tagging, that AtCLCc is located to the tonoplast (Figure 2).

A QTL analysis for shoot nitrate accumulation previously suggested that AtCLCc is involved in nitrate storage (Harada et al., 2004). Although Harada et al. found impaired nitrate accumulation in a clcc mutant, we found no difference in global nitrate content between mutant and wild-type plants under limiting (2 mm) or non-limiting (10 mm) nitrate conditions (data not shown). Moreover, no phenotype was observed in vitro on media containing 0.1, 1, 7, 10 or 20 mm KNO3 (data not shown). Harada et al. (2004) used different culture conditions and a different accession (No-O) than used in this study (Col-0 and WS). We found differences between WS and Col-0 in terms of nitrate and chloride content in response to NaCl treatment (Table 2). Thus, the discrepancy between our study and the study by Harada et al. (2004) could be linked to the use of different accessions. In this study, identification of four T-DNA insertional mutants of AtCLCc in two ecotypes (Col-0 and WS) allowed us to assess the contribution of AtCLCc to ion homeostasis in wild-type plants. Interestingly, the two phenotypes of clcc mutant plants that we identified, reduced stomatal movement and salt sensitivity, are directly related to the movement of chloride across the tonoplast membrane.

The first phenotypic trait identified in clcc mutants was de-regulation of stomatal movement associated with a strong diminution in chloride content in guard cells. The aperture of stomatal pores in leaf epidermis is regulated by turgor and volume changes in guard cells. The large central vacuole plays an important role in guard-cell turgor regulation, as 90% of K+ and anions accumulated and released by guard cells during stomatal movements are shuttled into and out of the guard cell vacuoles (Humble and Raschke, 1971; MacRobbie, 1981, 1983, 1990). Several studies have shown that Cl uptake into guard cells is crucial to balance K+ uptake during stomatal opening (Penny and Bowling, 1974; Penny et al., 1976; Schnabl and Raschke, 1980; Bowling, 1987; Lascève et al., 1987). Interestingly, both light-induced stomatal opening and ABA-induced stomatal closure were significantly impaired in all clcc mutants in the presence of KCl (Figure 3). This deficiency was dependent on chloride, as no significant impairments of these responses were observed when Cl was replaced with NO3 (Figure 4). In addition, chloride was found to be dramatically less abundant in the guard cells of mutant lines compared to wild-type (Table 1). Thus, AtCLCc is able to transport Cl into the vacuole and regulate stomatal movements. This hypothesis suggests that stomatal opening in clcc mutants is be directly impaired due to lack of transport of Cl by AtCLCc, whereas closure is indirectly affected due to the large decrease in Cl content in the vacuole, meaning that Cl cannot flow from the vacuole to promote vacuole depolarization.

The second trait of clcc mutants is reduced tolerance to NaCl or KCl. Growth reduction was observed in clcc mutants grown on soil watered with 200 mm NaCl or in vitro on half-strength MS containing 50 mm NaCl. Transformation of mutant plants with 35S:CLCc transgene restored wild-type growth, confirming that AtCLCc is involved in salt tolerance. Other members of the CLC family have already been shown to be involved in NaCl response. In rice, OsCLC-1, which is located on the tonoplast, is expressed in response to NaCl treatment (Nakamura et al., 2006). In soybean, GmCLC-1, which is also located at the tonoplast and induced by NaCl treatment, allows better tolerance to NaCl by accumulating chloride in the vacuole in BY2 cells (Li et al., 2006). In Physcomitrella patens, a member of the CLC family has been identified by proteomic analyses in response to NaCl treatment (Wang et al., 2008). However, none of these studies have shown a direct link between CLC and salinity tolerance.

The reduction in shoot growth in response to salinity occurs in two phases: a rapid response to the increase of external osmotic pressure, and a slower response due to accumulation of toxic ions (Munns and Tester, 2008). Under our conditions, no difference was detected in clcc mutants in response to mannitol, which generates osmotic stress. However, AtCLCc appears to participate to the second phase of salinity response, the ionic phase. Moreover, long exposure to NaCl treatment (3 weeks in vitro) is necessary to observe shoot growth reduction in clcc mutants (Figure 5). This is consistent with the fact that AtCLCc is required for the response to salt stress during the ionic phase. We show here that, in A. thaliana, the hypersensitivity of clcc mutants to NaCl stress is specific to chloride rather than sodium, as the mutants show a similar phenotype in the presence of KCl (Figure 6). No modification of sodium content was detected in the shoot or the root in response to NaCl treatment, but less chloride was found in roots, and this was associated with a slight increase in nitrate content (Tables 2 and S2). Interestingly, only clcc mutants in the WS accession were less tolerant than the wild type to NaCl stress at the root level, consistent with the expression of AtCLCc in roots in response to NaCl (Figure 5A) and with the reduction in Cl root content (Table 1), which was only observed in WS accession. These data suggest that, in clcc mutant roots, chloride cannot be sequestered in the vacuole, at least for WS accession mutants. The two clcc phenotypes, NaCl sensitivity and altered control of stomatal aperture, may be linked, as salinity affects stomatal conductance. The rapidity of this response excludes the involvement of transport of ABA from the roots to the shoots, and another mechanism may be involved (Fricke et al., 2004; Munns and Tester, 2008). AtCLCc could be involved in this cooperation between roots and stomata by taking part in chloride homeostasis.

Many CLC proteins have been previously described and characterized as Cl channels. The high-resolution crystal structure of the bacterial protein CLC-ec1 reveals that two highly conserved glutamate residues, Glu148 and Glu203, are crucial for ion transport, and that the serine in the signature sequence GSGIPE is essential for gating and selectivity (Dutzler et al., 2002, 2003). Interestingly, Glu203 is conserved in both channels and transporters, while Glu148 is strictly conserved only among transporters. Sequence analysis of AtCLCc revealed the presence of two glutamate residues in positions 173 and 212, suggesting that AtCLCc might function as an anion/proton antiporter like AtCLCa does.

The recent discovery that AtCLCa is a NO3-selective H+-coupled exchanger, in which a proline replaces the serine in the signature sequence, confirms the role of this residue in determining selectivity (Bergsdorf et al., 2009;Picollo et al., 2009; Zifarelli and Pusch, 2009b; Wege et al., 2010). Since AtCLCc has a serine in its signature sequence, AtCLCc is expected to be highly selective for Cl. To further characterize the transport properties of AtCLCc, we used an electrophysiological approach to investigate anion currents across the tonoplast of Arabidopsis guard cells. Unfortunately, due to their very small size, we were not able to successfully use the patch–clamp technique on isolated guard cell vacuoles. However, the two phenotypes previously described for clcc mutant plants, including the hypersensitivity of the mutant plants to NaCl or KCl salts and de-regulation of stomatal movement, both associated with a decrease in chloride content, point in this direction.

Together, the data presented here provide genetic and physiological evidence for a role of a new CLC member in Cl homeostasis in plants. In contrast with AtCLCa, which is involved in NO3 homeostasis, our data support a role for AtCLCc in vacuolar Cl accumulation. The lack of AtCLCc activity in clcc mutants dramatically reduces Cl content in guard cells and affects stomatal movement, but also modifies sensitivity to salt stress. However, expression analyses of the AtCLC family suggest that other CLCs in addition to AtCLCc may be involved in these processes. AtCLCa, whose product is involved in NO3 homeostasis, is also strongly expressed in guard cells and regulated by ABA, suggesting involvement of AtCLCa in stomatal movement. AtCLCd and AtCLCg (in shoots) and AtCLCb (in roots) are up-regulated in response to NaCl (Figure 7). Expression of the tonoplastic AtCLCs does not seem to be affected in the clcc mutants (Figure S2), but we cannot exclude an effect at the protein level. As tissue localization and exact selectivity are not known for all CLC members, further experiments are necessary to clarify their contribution to stomatal movements and salt tolerance.

Finally, our data support a role for AtCLCc in vacuolar Cl accumulation, and highlight the active involvement of anion channels/transporters in osmoregulation and stomatal movement. The results of this study suggest that other AtCLC members might also participate in the same physiological processes, and future work is required to obtain further insight into their interactions.

Experimental procedures

Plant materials and growth conditions

Wild-type and clcc mutant plants of A. thaliana were grown on soil in a plant growth chamber with an 8 h light period (250 μmol m−2 sec−1) at 23°C and a 16 h dark period at 19°C, and relative humidity of 75%. For stomatal measurements, plants were watered daily with nutrient solution as previously described (Verret et al., 2004). For NaCl treatment, plants were grown under the same chamber conditions but with a 16 h light period and a 8 h dark period. They were watered with deionized water for 2 weeks, and then with 0, 150 or 200 mm NaCl for 5 weeks. For in vitro culture, seeds were germinated on diluted Murashige and Skoog medium (half-strength MS) containing 1% w/v sucrose (except the experiment with mannitol), supplemented or not with 25 or 50 mm NaCl or KCl, or with 55 or 110 mm mannitol, at 21°C under a 16 h light/8 h dark regime for 21 days. The osmolarity of the various media was checked using a freezing point osmometer (automatic type 15 osmometer, Löser Messtechnik, http://www.loeser-osmometer.de/).

The T-DNA insertion lines in the Columbia ecotype, clcc-1 and clcc-2, were obtained from Syngenta (Torrey Mesa Research Institute, San Diego, CA, USA) (Garlic Lines 523D08 and 539G05), and the lines in the Wassilewskija ecotype, clcc-3 and clcc-4, were obtained from the Versailles collection (lines 5DWK25 and DYB166, respectively). The resistance to kanamycin of progeny obtained from self-pollinated heterozygous plants with both alleles segregated approximately 3:1, indicating a single insertion in the mutants.

Gene constructs for protoplast and plant transformations

For complementation of clcc mutant lines and protoplast transformation, CLCc cDNA was amplified by PCR from a leaf cDNA library using the oligonucleotides GWCLCc-F and GWCLCc-R (Table S1). A BP Gateway recombination was performed with the pDONR207 vector as described by the manufacturer (Invitrogen, http://www.invitrogen.com). CLCc cDNA was then transferred by a LR reaction into the destination vector pMDC43 to generate an N-terminal GFP fusion (Curtis and Grossniklaus, 2003) and pH2GW7 for protoplast transformation and complementation of clcc lines (Karimi et al., 2002). For the promoter–reporter fusion, a 1.4 kb fragment of the AtCLCc promoter was isolated from genomic DNA using PRCLCc-F and PRCLCc-R (Table S1). This sequence was inserted into pBI101 vector to allow cloning and testing of promoter activities on the basis of β-glucuronidase expression.

To identify the subcellular localization of AtCLCc expression, the GFP–AtCLCc construct was introduced into Arabidopsis cell suspension protoplasts by PEG-mediated transformation (Thomine et al., 2003).

For stable expression, the various vectors were introduced into clcc plants by Agrobacterium tumefaciens-mediated transformation (Clough and Bent, 1998). Transgenic plants were selected on half-strength MS medium containing hygromycin (20 μg L−1). Seeds were obtained by self-fertilization, and homozygous lines from the T3 generation were used.

Microscopy

Confocal microscopy was performed using an inverted Leica TCS-SP2 confocal laser scanning microscope (Leica Microsystems, http://www.leica.com/), with excitation at 488 nm, and the fluorescence emission signal of GFP was recovered between 500 and 525 nm.

Stomatal aperture measurements

Stomatal aperture measurements were performed as described by Merlot et al. (2007) on epidermis from leaves of 3–4-week-old-plants grown as described above in 30 mm KCl or KNO3, 10 mm MES/Tris, pH 6.0. After 30 min in darkness, stomatal apertures were measured. In light-induced stomatal opening experiments, epidermal peels were incubated for 3 h under light (300 μmol m−2 sec−1) at 22°C, and then stomatal apertures were measured. For ABA-induced stomatal closure experiments, isolated epidermal cells were incubated in stomatal opening solution for 3 h under light. Then ABA was added at a concentration of 10 or 100 μm, and stomatal aperture was measured 2 h after treatment. Values are the means for at least 60 apertures from at least three experiments. Error bars represent standard errors of the mean (SEM), with a confidence interval of 95%.

Reverse transcription quantitative-polymerase chain reaction experiments

Total RNA was extracted from roots, flowers, leaves, germinated seeds and pollen using TRIzol® reagent according to the manufacturer’s instructions (Invitrogen). In addition, total RNA from guard cells and mesophyll cells was extracted from protoplasts as described previously by Leonhardt et al. (2004). Reverse transcription was performed on 2 μg of total RNA using a first-strand cDNA synthesis kit (Amersham-Pharmacia Biotech, http://www5.amershambiosciences.com/). PCR reactions were performed using SYBR Green mix (Takara, http://www.takara-bio.com/) in a 96-well plate using an ABI PRISM 7900HT sequence detection system (Applied Biosystems, http://www.appliedbiosystems.com/) or a Light Cycler 480 II (Roche, http://www.roche.com), as previously described (Herbette et al., 2006). Specific primers were designed using the LightCycler probe design software (Roche) for Actin2 and AtCLC genes (Table S1). The thermal profile used was 2 min at 50°C and 10 min at 95°C, followed by 45 cycles of 30 sec at 95°C, 30 sec at 60°C and 30 sec at 72°C. The presence of a single amplicon in each PCR reaction was confirmed by reference to melting curves. Standard curves were derived from reactions with Actin2-specific primers and a series of five dilutions of the cDNA template. The relative amount of transcripts for each RNA sample was determined by normalizing against the standard curve, and was calculated as the arithmetic mean of three independent repeated reactions. Data are representative of two or three independent biological experiments. The Genome Initiative numbers for the AtCLC genes are At5g40890 (AtCLCa), At3g27170 (AtCLCb), At5g49890 (AtCLCc), At5g26240 (AtCLCd), At4g35440 (AtCLCe), At1g55620 (AtCLCf) and At5g33280 (AtCLCg).

Histochemical GUS activity

For GUS staining, plant tissues were infiltrated in 50 mm NaPO4 buffer, pH 7.0, containing 0.01% w/v Triton X-100, 1 mm K3Fe(CN)6, 1 mm K4Fe(CN)6 and 1 mg ml−1 5-bromo-4-chloro-3-indolyl-β-d-glucuronide (X-Gluc), and incubated overnight at 37°C. Pigments were removed by incubation with five increasing concentrations of ethanol for 1 h each. The samples were then rinsed with the above phosphate buffer and examined using a light microscope or a binocular microscope equipped with a digital camera.

Infra-red thermography

Plants were cultivated on soil in a plant growth chamber as described above for up to 3–4 weeks. Leaves were cut after 4 h of the light period and exposed for 1 h to a Thermacam PM250 infra-red camera (Inframetrics, http://www.americaninfrared.com) as described by Belin et al. (2006). Using the flir pm250 software on a PC computer, an area of standard size was drawn on the leaves, and the mean temperature of the pixels within this area was recorded. The results for the leaf temperature recorded by the camera are the mean temperatures in this defined area.

Ion content measurements and X-ray microanalysis

Inorganic anion and cation contents were analysed on 3-week-old plants grown in vitro as described above. Inorganic anions were extracted from 50 mg fresh weight of tissue using nine volumes of water and three series of freezing and thawing at −20°C. The nitrate and chloride contents of shoots and roots were then analysed by HPLC on a DX-120 analyser (Dionex, http://www.dionex.com/). Elution buffer (1 mm NaHCO3, 3.5mm Na2CO3) was applied at a flow rate of 1.25 ml min−1 on a Dionex IonPac® AS14 anion-exchange column (4 × 250 mm). For sodium and potassium contents, roots and shoots were dried for 48 h at 50°C and mineralized. The cation content of plants was determined using ICP-AES (Vista MPX, Varian, http://www.varianinc.com).

To measure chloride and potassium contents in guard cells, whole leaves were sampled on 5-week-old plants. The material to be analysed was prepared as described by Le Thiec et al. (1994). The X-ray microanalytical studies were performed under standardized conditions using a Leo 1450 VP electron microscope (Zeiss, http://www.zeiss.com/) fitted with Oxford INCA ATW (for EDX, http://www.x-raymicroanalysis.com) spectrometer systems (SDD spectrometer). To ensure that the beam of primary electrons did not penetrate other cells, one layer of cell on a stub of aluminium was analysed. The absence of an aluminium signal in the spectra confirmed that the beam electrons did not penetrate the layer.

Acknowledgements

We thank Benoît Guillemardet (ISV, CNRS, Gif-sur-Yvette, France) and Laurie Piette (IBEB, CEA, St-Paul-lez-Durance) for technical support, Eugene Diatloff (ISV, CNRS, Gif-sur-Yvette, France) and Jean-Luc Montillet (IBEB, LEMS, CEA, St-Paul-lez-Durance) for critical reading of the manuscript, Stéphanie Boutet, Sylvie Citerne and Grégory Mouille (IJPB, INRA, Versailles, France) for nitrate content analysis, Paul Soreau (IBEB, CEA, St Paul lez Durance) for ICP-AES measurements,and Marie-Noëlle Soler (Plateforme Imagif, CNRS, Gif-sur-Yvette, France) for her help with confocal microscopy analysis. This reseach was financially supported by the Centre National Recherche Scientifique (CNRS), University of Paris VII, the Commissariat à l’Energie Atomique (CEA), the University of Aix Marseille II (grant number MENRT 26591-2007) and the Agence Nationale à la Recherche (ANR-Nitrapool: grant number ANR-08-BLAN-0008-02). Initial research of N.L. on CLCc in guard cells was supported by U.S. DOE grant number DOE-DE-FG02-03ER15449 to Julian Schroeder.

Ancillary

Advertisement