Identification and functional characterization of cation–chloride cotransporters in plants

Authors

  • José M. Colmenero-Flores,

    Corresponding author
    1. Centro de Genómica, Instituto Valenciano de Investigaciones Agrarias, Ctra. Moncada-Náquera Km. 5, 46113 Moncada, Valencia, Spain, and
    Search for more papers by this author
  • Germán Martínez,

    1. Centro de Genómica, Instituto Valenciano de Investigaciones Agrarias, Ctra. Moncada-Náquera Km. 5, 46113 Moncada, Valencia, Spain, and
    Search for more papers by this author
  • Gerardo Gamba,

    1. Molecular Physiology Unit, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán and Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Tlalpan 14000, Mexico City, México
    Search for more papers by this author
  • Norma Vázquez,

    1. Molecular Physiology Unit, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán and Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Tlalpan 14000, Mexico City, México
    Search for more papers by this author
  • Domingo J. Iglesias,

    1. Centro de Genómica, Instituto Valenciano de Investigaciones Agrarias, Ctra. Moncada-Náquera Km. 5, 46113 Moncada, Valencia, Spain, and
    Search for more papers by this author
  • Javier Brumós,

    1. Centro de Genómica, Instituto Valenciano de Investigaciones Agrarias, Ctra. Moncada-Náquera Km. 5, 46113 Moncada, Valencia, Spain, and
    Search for more papers by this author
  • Manuel Talón

    1. Centro de Genómica, Instituto Valenciano de Investigaciones Agrarias, Ctra. Moncada-Náquera Km. 5, 46113 Moncada, Valencia, Spain, and
    Search for more papers by this author

(fax +34 963424001; e-mail chema@ivia.es).

Summary

Chloride (Cl) is an essential nutrient and one of the most abundant inorganic anions in plant tissues. We have cloned an Arabidopsis thaliana cDNA encoding for a member of the cation–Cl cotransporter (CCC) family. Deduced plant CCC proteins are highly conserved, and phylogenetic analyses revealed their relationships to the sub-family of animal K+:Cl cotransporters. In Xenopus laevis oocytes, the A. thaliana CCC protein (At CCC) catalysed the co-ordinated symport of K+, Na+ and Cl, and this transport activity was inhibited by the ‘loop’ diuretic bumetanide, a specific inhibitor of vertebrate Na+:K+:Cl cotransporters, indicating that At CCC encodes for a bona fide Na+:K+:Cl cotransporter. Analysis of At CCC promoter-β-glucuronidase transgenic Arabidopsis plants revealed preferential expression in the root and shoot vasculature at the xylem/symplast boundary, root tips, trichomes, leaf hydathodes, leaf stipules and anthers. Plants homozygous for two independent T-DNA insertions in the CCC gene exhibited shorter organs such as inflorescence stems, roots, leaves and siliques. The elongation zone of the inflorescence stem of ccc plants often necrosed during bolt emergence, while seed production was strongly impaired. In addition, ccc plants exhibited defective Cl homeostasis under high salinity, as they accumulated higher and lower Cl amounts in shoots and roots, respectively, than the treated wild type, suggesting At CCC involvement in long-distance Cl transport. Compelling evidence is provided on the occurrence of cation–chloride cotransporters in the plant kingdom and their significant role in major plant developmental processes and Cl homeostasis.

Introduction

Cation transport and homeostasis are issues of special significance in plant biology. In recent years, potassium (K+) and sodium (Na+) transporters have been studied extensively because of their roles in both plant nutrition and salt tolerance (Amtmann et al., 2004; Horie and Schroeder, 2004; Pauline et al., 2003; Rodriguez-Navarro and Rubio, 2006). To understand the molecular mechanisms that plants have developed to cope with salinity, most approaches have been directed to studying cation transporters and their regulation. These studies have reported the characterization of transport pathways involved in Na+ uptake, organellar Na+ compartmentalization, Na+ efflux across the plasma membrane, and the recovery of K+ homeostasis (Essah et al., 2003; Pardo et al., 2006; Rus et al., 2004; Tester and Davenport, 2003). However, it is known that acclimatization of plants to salt stress also requires appropriate regulation of chloride (Cl) homeostasis, not only because of its toxicity when over-accumulated on the plant symplast (Marschner, 1995; White and Broadley, 2001), but also because Cl is the most important counter-anion in most saline soils, altering Na+ and K+ availability and distribution within plant tissues and cell compartments (Gaxiola et al., 1998). Cl toxicity is a common phenomenon, and high Cl accumulation (≥10 mm in leaf water) is toxic to sensitive species such as wheat, soybean, bean, cotton and most fruit trees, resulting in a major constraint to horticultural production on arid, semi-arid, irrigated and saline soils (Marschner, 1995; Storey and Walker, 1999; White and Broadley, 2001; Xu et al., 2000).

Cl, like K+, is an essential nutrient for higher plants and a major osmotically active solute in the vacuole, implicated in osmoregulation and cell elongation. It is also involved in photosynthetic O2 evolution, enzyme activity regulation, stabilization of membrane potential and intracellular pH regulation (Marschner, 1995; White and Broadley, 2001). Despite these facts, the involvement of Cl transport in salt stress or plant nutrition has not been properly investigated at the molecular level. Electrophysiological studies, however, have provided valuable information (reviewed by Barbier-Brygoo et al., 2000; Roberts, 2006). Under non-saline conditions ([Cl]ext < [Cl]cyt), Cl is actively taken up by a ΔpH-driven Cl/H+ symport (Babourina et al., 1998a;Beilby and Walker, 1981; Felle, 1994; Sanders, 1980). Under saline conditions, an outward rectifying anion channel permeable to NO3 and Cl has been characterized in isolated wheat protoplasts (Skerrett and Tyerman, 1994). Furthermore, three distinct anion conductances, X-QUAC, X-SLAC and X-IRAC, have been identified in xylem parenchyma cells of barley and maize (Gilliham and Tester, 2005; Köhler and Raschke, 2000). While Cl transport mechanisms in plants have been scarcely characterized, striking information is available in the animal field.

In animals, the cation–Cl cotransporter (CCC) family is essential for adequate homeostasis of the most abundant electrolytes, K+, Na+ and Cl, playing key roles in cell ionic and osmotic regulation (reviewed by Delpire and Mount, 2002; Gamba, 2005; Hebert et al., 2004; Russel, 2000). According to the Human Gene Nomenclature Database, this family has recently been called SLC12 (Hebert et al., 2004). CCC proteins are secondary active transporters that mediate the movement of Cl tightly coupled to that of K+ and/or Na+ across the plasmalemma (Haas, 1994). Regarding the ions involved in the symport mechanism, CCCs are divided into three groups: K+:Cl cotransporters, known as the KCC group; Na+:Cl cotransporters, NCC group; and Na+:K+:Cl cotransporters, NKCC group. Ion transport studies have demonstrated that the members of all three groups shared an absolute requirement for both Cl and at least one cation (Na+ and/or K+), and that the three cotransport processes are electrically silent or electroneutral (Russel, 2000). In secretory and absorptive epithelia, CCCs promoted net transepithelial salt and water movement (Delpire et al., 1994; Haas and Forbush, 1998; Isenring and Forbush, 1997; Xu et al., 1994). In non-epithelial tissues, CCCs participate actively in the regulation of cell volume (Haas and Forbush, 1998; Isenring and Forbush, 1997; Lauf et al., 1992). They also controlled Cl gradients in neurons, allowing modulation of the excitability state (Jang et al., 2001; Rivera et al., 1999). A role for CCCs in cell proliferation, survival and differentiation has also been proposed (Panet et al., 1999; Shen et al., 2001).

A putative plant CCC gene from tobacco (AXI 4), reported by Harling et al. (1997), was proposed to be involved in stimulation of protoplast division in an auxin-independent manner. However, results supporting growth stimulations by AXI 4 were subsequently retracted (Schell et al., 1999) and plant CCC genes remained functionally uncharacterized to date. In this work, we show that members of the CCC gene family in mono- and dicotyledonous plants belong to a phylogenetic compact group of highly conserved genes. In addition, it is shown that the Arabidopsis thaliana CCC gene (At CCC) encodes for a cation–Cl cotransporter involved in plant development and ion homeostasis.

Results

Identification of plant CCC genes, sequence analyses and phylogenetic relationships

Given the participation of CCC proteins in the regulation of Cl homeostasis in animal systems, we decided to deal with the identification and functional characterization of plant CCC genes. The predicted amino acid sequence of the AXI 4 cDNA, which shared 36–38% amino acid sequence identity with KCC proteins (Harling et al., 1997), was used as a query sequence in blastp similarity searches against the A. thaliana genome, giving rise to the identification of a single CCC gene (AGI reference At1g30450), referred to here as At CCC. Using At CCC gene-specific primers situated in the predicted 5′- and 3′-untranslated (UTR) regions, the corresponding cDNA was cloned by RT–PCR and sequenced. A complete open reading frame (ORF) was found in the At CCC cDNA sequence, and its comparison with the genomic sequence allowed us to determine the intron/exon structure of the At CCC gene. tblastn searches identified orthologue CCC genes in other plant genomes. Two CCC genes were identified in rice (Oryza sativa, japonica), Os CCC1 and Os CCC2, and one more in Medicago truncatula, Mt CCC. The gene structure of CCC orthologue genes was deduced from the integrated results of the blastp, tblastn and genscan gene-prediction programs. Interestingly, these plant CCC genes contained 12 introns located in the same position, indicating that the exon/intron structure was perfectly conserved (Figure S1 in Supplementary Material). Description of plant CCC genes is summarized in Table S1. Polypeptides encoded by the CCC genes reported here are 975–994 aa long, with predicted molecular masses between 106.6 and 108.7 kDa. A multiple alignment with the ClustalX program showed a very high degree of homology among plant CCC proteins, with 83–89% similarity through the whole sequences (Figure S2). Animal CCCs are integral membrane glycoproteins, consisting of a central hydrophobic region of 10–12 transmembrane (Tm) segments flanked by large hydrophilic N- and C-termini that reside within the cytoplasm (Gerelsaikhan and Turner, 2000). The same structure was predicted for all the plant CCC proteins according to different hydropathy- and secondary structure-prediction programs (see Experimental procedures). The position of the putative Tm domains, as well as a predicted Asn-glycosylation site, are given in Figure S2. A tentative model of plant CCC topology based on these considerations is presented in Figure 1(a). In the proposed model, a single glycosylation site is situated in the extracellular hydrophilic segment linking the Tm3 and Tm4 domains. A multiple alignment, including functionally characterized proteins from animal CCCs, was also performed (Figure S3), revealing a significantly higher conservation of plant CCCs with animal KCCs (approximately 60% similarity) rather than with sodium-dependent NCCs and NKCCs (approximately 46% similarity). Plant and animal CCCs were highly conserved in the C-termini and the central hydrophobic core, particularly within the putative Tm domains and the predicted intracellular loops. In general, extracellular loops showed lower sequence identity while the N-termini were only barely conserved (Figure S3). A schematic representation obtained from the ProDom program (Corpet et al., 2000; http://www.toulouse.inra.fr/prodom.html), displaying conserved structural domains of plant and animal CCCs, is presented in Figure 1(b). The major apparent structural difference between CCCs was related to the position of the potential glycosylation sites.

Figure 1.

 Structural and phylogenetic relationships between plant and animal cotransporters.
(a) Putative topology model of plant CCCs. The grey box represents the membrane containing the predicted 12 transmembrane spanning domains. Open circles, amino acid residues; branched lines in the predicted extracellular domain between transmembrane domains 3 and 4, potential N-glycosylation site.
(b) Analysis of structural domain arrangements and homology relationships among CCC sub-families obtained from the ProDom database. Mean values of amino acid identities (Id.) and similarities (Sim.) were determined with the NCBI blastp program using the Arabidopsis thaliana CCC protein as query sequence. Boxes represent protein domains that correspond to ProDom entries. Box length is not proportional to the length of the represented protein domain. Filled boxes indicate high conservation with the corresponding plant CCC domain, and their inclusion in the same ProDom domain family (ProDom accession numbers are indicated under the boxes). Grey boxes indicate the position of putative transmembrane regions, which are numbered above. Branched lines correspond to potential N-linked glycosylation sites in a predicted extracellular domain.
(c) Phylogenetic tree of representative CCC proteins from animals, fungi and plants constructed with the UPGMA method. Genetic distances were calculated with the Poisson correction for proteins. The accession numbers and sources of the CCC proteins are as follows. Human cotransporters: HsKCC4, NM_006598; HsKCC2, AF208159; HsKCC1, NM_005072; HsKCC3, AF314956; HsCCC6 (CIP1), NM_020246; HsNCC, NM_000339; HsNKCC2, NM_000338. Drosophila melanogaster cotransporters: DmCCC3, NP_726377.1; DmCCC1, NP_726379.1; DmCCC2, NP_571976.2; DmCCC6, NP_648572.1; DmCCC4, NP_730938.1; DmCCC5, NP_730939.1; DmCCC7, NP_609887.1. Pseudopleuronectes americanus PaNCCa, P55019. Manduca sexta MsNKCC, Q25479; fungal cotransporters: Schizosaccharomyces pombe SpCCC, NP_595740.1; Saccharomyces cerevisiae ScCCC, YBR235 W; Neurospora crassa NcCCC, XP_323085.1. Plant cotransporter accession numbers are given in Table S1.

A multiple alignment, including CCC-homologous proteins from fungal organisms, was also obtained and used to calculate genetic distances among CCC proteins, allowing the construction of a phylogenetic tree with the neighbour-joining method (Figure 1c). In this tree, three main family branches were observed. Two branches distinguished Na+-dependent from Na+-independent cotransporters, and a third grouped the human cotransporter interacting protein (CIP1) with fungal CCC-homologous proteins. Plant CCCs grouped into a cluster more closely related to animal KCCs.

At CCC encodes for an Na+:K+:Cl cotransporter

Most common assays employed to demonstrate CCC activity consist of the stimulation of 22Na+ and/or 86Rb+ (as K+ analogue) uptake in a Cl-dependent manner in Xenopus laevis oocytes or mammalian cells transfected with the corresponding CCC gene (Gamba, 2005). To determine the functional activity and properties of At CCC, we used the heterologous expression system of X. laevis oocytes, which has been extensively used by us and others to express all members of the human CCC gene family (Gamba, 2005; Hebert et al., 2004). As plant CCCs are more closely related to KCC proteins (Figure 1), we first assayed Cl-dependent 86Rb+ uptake in oocytes injected with either water, At CCC cRNA or KCC4 cRNA (a well characterized mammalian K+:Cl cotransporter). Because K+:Cl cotransporters exhibit little to no activity in isotonic conditions and are activated by cell swelling (Mount and Gamba, 2001), 86Rb+ uptake was performed 4 days after injections in both isotonicity and hypotonicity. As shown in Figure 2(a), 86Rb+ uptake was similar in water and At CCC cRNA-injected oocytes in both isotonicity and hypotonicity. The small Cl-dependent 86Rb+ uptake in water and At CCC cRNA-injected oocytes was due to the activity of the endogenous K+:Cl cotransporter present in oocytes, as described previously (Mercado et al., 2001). In contrast to At CCC cRNA, KCC4 cRNA-injected oocytes showed high 86Rb+ uptake (approximately 75-fold increase) in hypotonic conditions. Thus no K+:Cl cotransport-like activity was induced in oocytes by At CCC cRNA injection. However, when Na+ was added to the extracellular medium we were able to define the functional properties of At CCC. As shown in Figure 2(b), when uptake assays were performed in isotonic conditions in the presence of Na+, K+ and Cl, microinjection of At CCC cRNA resulted in a threefold increase in 86Rb+ uptake (At CCC, 3.068 ± 123 versus H2O-injected oocytes, 878 ± 100 pmol per oocyte h−1, P < 0.01). This uptake was Cl-dependent (156 ± 33 pmol per oocyte h−1). The uptake observed in water-injected oocytes in these conditions was due to the endogenous Na+:K+:2Cl cotransporter that we and others have described before (Gamba et al., 1994; Plata et al., 2002; Shetlar et al., 1990; Suvitayavat et al., 1994). In addition to ion selectivity, CCCs can be grouped according to their affinities for specific inhibitors (Gamba, 2005). We show here that At CCC-dependent ion uptake was sensitive to the NKCC-specific inhibitor bumetanide (45 ± 6.6 pmol per oocyte h−1) (Figure 2b). The observation that At CCC cRNA injection in Xenopus oocytes induced a Cl-dependent and bumetanide-sensitive 86Rb+ uptake, active in isotonic conditions, suggested that At CCC encodes for an NKCC cotransporter. Supporting this conclusion, we demonstrated that At CCC cRNA microinjection of X. laevis oocytes also increased 22Na+ and 36Cl uptake. Figure 2(c) showed that At CCC induced 22Na+ uptake, and that this was reduced in the presence of 100 µm bumetanide. Similarly, At CCC cRNA microinjection resulted in a significant increase in 36Cl uptake when compared with the control water-injected oocytes (Figure 2d). 36Cl uptake was reduced in the absence of Na+ or K+ in the extracellular medium and, as above, in the presence of bumetanide. Results shown in Figure 2 indicate that At CCC encodes for a transport pathway that simultaneously increases 22Na+, 86Rb+ and 36Cl uptake. In all cases the tracer uptake was reduced by bumetanide. In addition, we have shown that increased 86Rb+ uptake was Cl- and Na+-dependent, while increased 36Cl uptake was both Na+- and K+-dependent. Therefore At CCC encodes for the plant homologue of the animal bumetanide-sensitive Na+:K+:2Cl cotransporter.

Figure 2.

 Functional expression of At CCC in Xenopus laevis oocytes.
(a) 86Rb+ uptake in oocytes injected with either water, At CCC cRNA or KCC4 cRNA. Uptake was performed in the absence of extracellular Na+, in hypotonic (110 mOsm kg−1 H2O) or isotonic (220 mOsm kg−1 H2O, adjusted with sucrose) solutions in the presence (grey bars) or absence (open bars) of extracellular Cl. Each bar represents the mean ± SE of 20 oocytes extracted from two different toads.
(b) 86Rb+ uptake in oocytes injected with either water or At CCC cRNA in isotonic conditions in the presence of extracellular Na+, K+ and Cl (grey bars); in the presence of the three ions plus10−4m bumetanide (closed bars); and in the presence of the two cations without extracellular Cl (open bars). Each bar represents the mean ± SE of 20 oocytes extracted from two different toads.
(c) 22Na+ uptake in oocytes injected with either water or At CCC cRNA in isotonic conditions, in the presence of extracellular Na+, K+ and Cl, and in the absence (grey bars) or presence (black bars) of 10−4m bumetanide. Each bar represents the mean ± SE of 10 oocytes extracted from one toad.
(d) Cl uptake in oocytes injected with either water or At CCC cRNA in isotonic conditions in the presence of extracellular Na+, K+ and Cl (open bars); in the presence of the three ions plus10−4m bumetanide (closed bar); without extracellular Na+ (grey bar); and without extracellular K+ (hatched bar). Results were tested using anova, resulting in statistically significant differences (< 0.005); comparisons of means were determined through the least significant differences (LSD) method (< 0.05). Columns with different letters indicate statistically significant differences.

While Na+-coupled mechanisms dominate secondary transport in animal systems, active transport in terrestrial plants relies on H+-dependent cotransporters energized by H+-ATPases. Earlier studies had identified H+/Cl symport systems as responsible for active Cl transport in plants (Babourina et al., 1998a; Felle, 1994). To ascertain whether plant CCCs may also function as H+-driven symporters, the effect of pH on At CCC activity was studied. H+-driven transport mechanisms are stimulated at low external pH (4–5) and inhibited at high external pH (7–9) as shown previously (Felle, 1994; Sanders et al., 1985). In contrast, a progressive pH increase from 5 to 9 stimulated At CCC activity, which displayed the highest transport activity around pH 8 (Figure 3). This indicated that At CCC did not function as an H+-dependent cotransporter.

Figure 3.

 pH-dependence of At CCC activity.
(a) 86Rb+ uptake in oocytes injected with water or At CCC cRNA in the presence of extracellular Na+, K+ and Cl. Each bar represents the mean ± SE of 10 oocytes extracted from two different toads.
(b) Net 86Rb+ transport activity after subtracting uptake values of water-injected oocytes shown in (a).
anova analyses showed statistically significant differences for (a) (= 0) and (b) (< 0.001). Comparisons of means were determined through the least significant differences (LSD) method. Columns with different letters indicate statistically significant differences (< 0.05).

At CCC expression and tissue specificity

To determine tissue and cell specificity, At CCC expression patterns were examined in transgenic plants carrying the Escherichia coliβ-glucuronidase (GUS) gene under the control of the At CCC promoter region (see the entire sequence of the At CCC promoter–GUS construct in Figure S4). In 5-day-old seedlings, At CCC promoter was active in cotyledon tips, plant vasculature, root tips and axillary buds (Figure 4a,b). In vegetative tissues of adult plants, the same expression pattern was observed in roots, with GUS activity localized to the vascular strand and the root tip (Figure 4c,d). Cross- and longitudinal section analyses of paraffin-embedded roots from 5-week-old plants revealed that GUS activity was located in the pericycle and other parenchyma cells bordering xylem vessels (Figure 4e,f). In emerging secondary roots, GUS activity was also associated with epidermal cells (Figure 4f). At CCC gene expression related to the xylem/symplast boundary was also observed in aerial organs of mature plants such as the rosette stem (Figure 4g), rosette leaves (Figure 4h–j) and cauline leaves (Figure 4k). Other leaf organs showing At CCC gene expression were trichomes (Figure 4l) and hydathodes (Figure 4m,n).

Figure 4.

 At CCC promoter–GUS expression pattern in transgenic Arabidopsis plants.
All panels show GUS expression in mature plants except for (a, b) corresponding to 5-day-old Arabidopsis seedlings.
(a, b) GUS expression in 5-day-old Arabidopsis seedlings: cotyledon tips and hypocotyl vasculature (a); and axillary buds, indicated by arrows (b).
(c, d) GUS expression in the root vasculature and root tip of mature plants.
(e) Longitudinal section of a secondary root in a CCC–GUS plant showing GUS activity in pericycle and inner tissues surrounding xylem vessels.
(f) Cross-sections of the proximal zone of the primary root in a CCC–GUS plant showing GUS activity in the cells surrounding xylem vessels and also in the epidermis from emerging secondary roots.
(g) Oblique sections of the rosette stem in a CCC–GUS plant showing GUS activity in xylem parenchyma cells.
(h–j) Cross-section of a rosette leaf in a CCC–GUS plant showing GUS activity in xylem parenchyma cells: a leaf section stained with toluidine blue is first shown in (h); an adjacent section showing GUS activity in (i); and GUS activity associated to xylem parenchyma cells is shown in detail in (j).
(k) Cross-section of a cauline leaf in a CCC–GUS plant showing GUS activity in xylem parenchyma cells.
(l) GUS expression in trichome cells.
(m, n) GUS expression associated to leaf hydathodes: in a cotyledon (m) and a rosette leaf (n).
(o–r) GUS expression in leaf stipules: longitudinal section stained with toluidine blue showing an inflorescence leaf primordia (o); an adjacent section showing GUS activity in a stipule (p); cross-section of very young cauline leaves showing GUS activity in stipules (q); GUS activity in stipules of an expanded cauline leaf (r).
(s,t) GUS activity in cauline leaves surrounding the shoot apex of an emerging inflorescence stem.
(u, v) GUS activity in stamens: stamens of an open flower (u) and detail of stamens showing stronger GUS activity in pollen grains (v).
Co, cortex; c, cotyledon; Ep, epidermis; Is, inflorescence stem; Lp, leaf primordia; M, shoot meristem; p, pericycle; Ph, phloem; Sp, stipules; X, xylem; xP, xylem parenchyma. Scale bars are in µm.

In the shoot meristems of inflorescence stems, microscope analyses of paraffin-embedded tissues revealed that GUS expression was associated to leaf stipules (Figure 4o–q). During emergence of the inflorescence stem, GUS activity was intense in stipules, veins, trichomes and hydathodes of young cauline leaves surrounding the apex (Figure 4r–t). Finally, high GUS activity was found in flower stamens, mainly associated to pollen grains (Figure 4u,v).

Supporting data for experiments on tissue localization of At CCC transcript were obtained through quantitative RT–PCR with Syber-Green chemistry and gene-specific primers. In agreement with GUS observations, higher At CCC transcript accumulation was detected in seedlings, leaf hydathodes, root tips and stamens (Figure S5). No evidence was found on At CCC regulation in response to Cl starvation, Cl resupply, salinity or water deficit (data not shown).

Identification of homozygous T-DNA insertion alleles of At CCC

To ascertain the role of plant CCCs, two independent T-DNA insertion lines from the SALK collection, mapping into the At1g30450 locus, were used: SALK-048175 (referred to here as ccc-1) and SALK-145300 (ccc-2). The position of the T-DNA left-border insertion sites in these two lines is shown in Figure 5(a). The genotype of the plants was determined by PCR using both At CCC gene-specific primers flanking the T-DNA insertion points, Pccc-F and I1-R, and the gene-specific primer Pccc-F in combination with a T-DNA-specific primer placed in the left border of the insertion element, LBb1 (Figure 5a). Plants homozygous for the T-DNA-insertion in the At CCC gene were not identified in a first screening for any insertion lines, indicating possible participation of the At CCC gene in plant viability and/or fertility processes. However, heterozygous plants for both T-DNA insertions were found. Progenies obtained from self-fertilization of heterozygous plants allowed the identification of plants homozygous for both T-DNA insertions. To verify the absence of integral At CCC transcript in the ccc knockout lines, RT–PCR on total RNA from plants homozygous and azygous for the T-DNA insertions were performed with At CCC gene-specific primers that flanked the T-DNA insertion points, Atkcc-F and RT-R. As shown in Figure 5(b), At CCC gene-specific RT–PCR fragments were absent in the mutant lines, whereas a product was obtained from the azygous lines and an At CCC-overexpressing transgenic line.

Figure 5.

 Mutant alleles of CCC.
(a) Schematic illustration indicating the position of the T-DNA insertions in exon E1 and intron I1 in ccc-1 and ccc-2 mutant alleles, respectively. Arrowheads in white boxes indicate position of T-DNA left border. Arrows indicate position and orientation of primers used for T-DNA genotyping and for RT–PCR amplification of the CCC mRNA.
(b) RT–PCR products obtained with At CCC gene-specific primers and total RNA isolated from the following lines: ccc-1 (1); ccc-2 (2); azygous, wild type, line resulting from the segregation of a heterozygous CCC/ccc-1 plant (3); azygous, wild type, line resulting from the segregation of a heterozygous CCC/ccc-2 plant (4); CCC overexpressing transgenic Col line 142 (5). Expression of the housekeeping α-tubulin gene TUA6 was used to control RNA loading.
(c) Shorter inflorescences in the knockout lines ccc-1 and ccc-2.

The most significant phenotype of the ccc knockout plants was the scarce development of the inflorescences (Figure 5c). Twenty-three out of 96 individuals that segregated from a CCC/ccc-1 heterozygous plant exhibited short inflorescences and were genotyped as homozygous for the T-DNA insertion. This 1:4 ratio corresponded to the segregation of a single recessive mutant allele (χ2 = 0.055; > 0.9 for a 1:4 segregation ratio). Thirty-eight individuals with wild-type phenotype were also genotyped for the presence of the T-DNA insertion in the At CCC gene, and the analyses indicated that 27 were heterozygous and 11 were azygous. A CCC/ccc-2 heterozygous plant corresponding to the second insertion line exhibited similar segregation ratios, and homozygous ccc-2/ccc-2 segregants also developed short inflorescences (Figure 5c).

Reliable evidence for the relationship of the observed phenotype with the loss of At CCC function was obtained through genetic complementation, rather than with direct transformation, because of the partial sterility of the mutants. To obtain At CCC expression in a homozygous ccc background, line ccc-1 was crossed with a Col transgenic line overexpressing the At CCC gene (see the entire sequence of the complementation construct in Figure S6). This cross provided an F1 heterozygous progeny, as confirmed with PCR (data not shown). Plants homozygous for the T-DNA insertion and overexpressing the At CCC gene were subsequently identified in the segregating F2 progeny. Six out of 40 F2 plants analysed with PCR (data not shown) showed the desired genotype, exhibiting a wild-type phenotype with inflorescences of normal appearance (Figure S7). Taken together, these data demonstrated that both ccc-1 and ccc-2 lines were homozygous for the T-DNA insertions in the At CCC gene, and also that the short inflorescence phenotype was associated with the lack of At CCC function.

ccc mutants exhibit developmental alterations

Arabidopsis thaliana CCC-knockout lines exhibited a bushy phenotype, with short inflorescences containing a higher number of stems (Figure 6a). In general, potted juvenile ccc plants exhibited a wild-type appearance, although their rosette diameter was slightly smaller compared with wild-type Col-0 (Col) plants (Figure 6b). Roots from hydroponically growing plants were shorter in ccc plants (Figure 6c). Most visible developmental alterations of the knockout mutants affected reproductive organs. Although inflorescence emergence and even flowering initiation were similar in Col and ccc plants (data not shown), mutant lines exhibited shorter inflorescences with higher number of stems per plant (Figure 6d,e). The dwarf phenotype might be apparently linked to a frequently observed collapse of the inflorescence stem-elongation zone, which eventually become necrotic (Figure 6 g,h). Cauline leaves also exhibited a lower size in ccc plants (Figure 6f). In addition, ccc plants were impaired in silique development and seed production. During the early to mid-flowering stages, silique production rate was strongly reduced in ccc plants (Figure 6i), which developed many siliques with no seeds (Figure 6j,k). Later on, ccc plants gave rise to viable siliques, although they were always smaller and contained fewer seeds than those produced by Col plants (Figure 6l). The production of flowers and siliques also lasted longer in ccc plants than in Col plants (data not shown). Developmental alterations associated with the lack of function of At CCC could not be overcome with the application of higher Cl, K+ and Na+ concentrations to the growing media (data not shown).

Figure 6.

 Phenotype of ccc plants.
(b, d–f, i) bsl00001, Col plants; □, ccc-1 plants; ⋄, ccc-2 plants. Days are given relative to date of sowing, including a 3-day stratification period at 4°C to synchronize seed germination. Plant measurements were obtained in soil-based assays, except for values in (c) that were generated in hydroponic culture.
(a) Dwarf and bushy phenotype of 55-day-old ccc plants.
(b) Comparison over time of rosette diameter between ccc-1 and Col plants. Data are given as averages ± SD for 20 individual plants.
(c) Comparison of primary root length between ccc-1 and Col plants grown in hydroponic culture. Data are averages ± SD of 24 and 17 plants for ccc-1 and Col-0, respectively.
(d) Main bolt length comparison over time between ccc lines and Col plants. Data are averages ± SD of 21, 34 and 18 plants for Col, ccc-1 and ccc-2, respectively.
(e) Comparison over time of the number of side bolts between ccc lines and Col plants. Only emerging stems >0.5 mm were considered. Data are averages ± SD of 21, 34 and 18 plants for Col-0, ccc-1 and ccc-2, respectively.
(f) Comparison over time of cauline leaf length between ccc lines and Col plants. Only the first and second cauline leaf of the main stem were measured. Data are averages ± SD of 21, 34 and 18 plants for Col, ccc-1 and ccc-2, respectively.
(g) Typical necrosis in the elongation zone of emerging inflorescence stems from ccc-1 plants.
(h) Percentage of necrosed inflorescence stems from 55-day-old ccc-1 plants. Data are averages ± SD of 20 individual plants.
(i) Comparison over time of silique number between ccc lines and Col plants. Data are given as averages ± SD for 10 individual plants.
(j) Siliques from 40-day-old Col-0 and ccc-1 plants.
(k) Percentage of aborted siliques in 45-day-old Col and ccc-1 plants. Data are averages ± SD of 10 individual plants.
(l) Number of seeds per viable silique in 54-day-old Col-0 and ccc-1 plants. Data are averages ± SD of 10 individual plants.

Plant Cl distribution is altered in ccc knockout mutants

The preferential expression of At CCC at the xylem/symplast boundary, as indicated by the GUS reporter gene, may suggest a role of plant CCCs regulating long-distance ion transport through controlling xylem loading or unloading. To investigate this possibility, Cl accumulation in the shoot of ccc-1 and wild-type potted plants treated with different concentrations of chloride salts, applied as equimolar amounts of NaCl and KCl, was compared. The results indicated that mutant plants subjected to salinity treatments accumulated higher Cl concentration in the shoot compared with Col plants (Figure 7a), suggesting a regulatory role of AtCCC in Cl accumulation in aerial tissues. To obtain supporting data for this hypothesis, we studied Cl distribution among roots and shoots in hydroponically growing Col and ccc plants treated with 50 mm Cl salts (Figure 7b). In comparison with Col plants, both ccc-1 and ccc-2 mutant lines accumulated higher Cl concentration in shoots and lower in roots. This unbalanced Cl distribution in the ccc lines may indicate a main involvement of At CCC at the xylem parenchyma cells in Cl unloading in the root and/or Cl withdrawal from aerial tissues.

Figure 7.

 Cl accumulation in shoots and roots of Col and ccc mutant plants under saline conditions.
(a) Shoot Cl contents in soil-grown plants subjected to 0, 30, 60 and 90 mm Cl treatments for 15 days. Two independent experiments were performed, and similar patterns of Cl accumulation were obtained.
(b) Cl content and distribution between shoot and root over time in hydroponically grown plants subjected to 50 mm Cl salts. Two independent experiments were performed, and similar patterns of Cl accumulation were obtained.
Closed bars, wild type; open bars, ccc-1; grey bars, ccc-2. Error bars, ± SD (= 4–8). Parameters were tested statistically by anova; comparisons of means were determined using the least significant differences (LSD) method. Columns with different letters indicate statistically significant differences (< 0.05)

Discussion

We describe in this work the characterization of a novel group of inorganic ion transporters in plants belonging to the CCC family. CCCs have been largely characterized in animals, where they are involved in the regulation of intracellular Cl content, vectorial salt transport in secretory and absorptive epithelia and cell volume regulation. In plants, therefore, CCCs are probably involved in the homeostasis of Cl. We have identified CCC genes from mono- and dicotyledonous plant species (Figure S1; Table S1) that conform a compact phylogenetic cluster with highly conserved genes (Figures 1 and S2). Gene conservation was evident for both gene structure (identical number and position of introns; Figure S1) and sequence homology (87–89% similarity in predicted amino acid sequences; Table S1). The hydrophobic central domain displayed the highest sequence homology among animal and plant CCCs (Figures 1 and S3), in agreement with the involvement of the transmembrane segments in ion translocation. Although CCCs are plasma membrane transporters in animal cells, and At CCC activity in oocytes suggests that this may also be true in plant cells, the subcellular localization of plant CCCs is still an open question.

Functional expression of At CCC in Xenopus oocytes (Figure 2) showed the following functional properties of the plant cotransporter: (i) At CCC mediated Na+, K+ (Rb+) and Cl transport; (ii) At CCC activity required the simultaneous presence of all three co-ions (Cl, K+ and Na+); (iii) At CCC activity was specifically sensitive to the loop diuretic bumetanide. These are functionally defining and unique characteristics of NKCC cotransporters (Gamba, 2005; Russel, 2000). As there was a contribution of endogenous oocyte CCCs to the measured values of At CCC activity, a minor possibility that the At CCC protein modulates oocyte transporters cannot be ruled out completely. Thus the dominant structural unit of certain animal CCCs in the plasma membrane is a homodimer (De Jong et al., 2003;Moore-Hoon and Turner, 2000; Starremans et al., 2003). Furthermore, there is evidence indicating that CCC-related proteins, such as CIP1 and an NKCC2-splicing variant, specifically inhibit the activity of NKCC1 and NKCC2, respectively (Caron et al., 2000; Plata et al., 1999). However, no data showing stimulation of CCC transporters through co-expression with another member of the family have been reported to date. Thus the possibility that At CCC may enhance endogenous NKCC1 activity is remote. Despite the Na+-dependent co-transport activity of At CCC, plant CCCs are more closely related to the animal KCC group of K+-coupled cotransporters (Figure 1), indicating that they do not fulfil the correlation between phylogenetic distribution and ion specificity observed in animal KCCs and NKCCs. This provides an excellent tool to delimit more precisely those protein domains involved in ion- and inhibitor-specificity of CCC proteins. For instance, it was proposed that the position of the extracellular glycosylated loop, which distinguishes Na+-dependent from Na+-independent animal cotransporters, is a major structural property of cation discrimination (Gillen et al., 1996; Payne et al., 1996). However, the presence of the conserved glycosylation domain of plant CCCs in an extracellular loop different from those of animal cotransporters (Figure 1) suggests that this structural characteristic is not involved in cation distinction.

Direct evidence for the participation of plant CCCs in Cl homeostasis in plants came from the analysis of ccc lines, which showed the involvement of At CCC in long-distance Cl transport (Figure 7). Consistent with this role, At CCC–GUS expression was found in xylem parenchyma cells (Figure 4). This allows us to hypothesize a role of plant CCCs in controlling Cl loading/unloading at the xylem/symplast boundary. Transport of inorganic ions between the plant symplast and the xylem is thought to be an electroneutral process, requiring the presence of anion conductances that match the cation permeabilities (Köhler and Raschke, 2000). Furthermore, Cl flux studies in transgenic Arabidopsis plants expressing a fluorescent anion probe supported the notion of a coupling between Cl transport and its accompanying permeable cations (Lorenzen et al., 2004). As CCC-mediated ion transport does not require the proton motif gradient, as suggested in Figure 3, ion translocation mediated by CCC proteins could be considered an efficient and energetically inexpensive mechanism to facilitate electroneutral salt distribution within the plant . The direction of net cotransport should be either inwards or outwards, depending on the sum of the electrochemical gradient of the three ions Cl, K+ and Na+.

Glycophyte plants cope with salinity through several strategies, including the control of ion accumulation in the shoot. This mechanism, also known as ion exclusion, is thought to be partially controlled by transport mechanisms regulating ion loading/unloading along the sap-conducting tissues of the vascular plants. The above results suggest that At CCC may participate, under salt stress conditions, in the Cl exclusion mechanism. Under salinity, progressive ion accumulation in leaf tissues generates a strong concentration gradient that eventually may facilitate At CCC-mediated net efflux of Cl from xylem parenchyma cells to xylem vessels. Similarly, the electrochemical gradient at the root xylem–symplast boundary may favour ion uptake into xylem parenchyma cells, promoting net ion retrieval from shoot to root. Other transporters, like the plasma membrane Na+/H+ antiporter SOS1, have been implicated in long-distance ion transport in plants (Shi et al., 2002). Under salt stress, sos1 mutant plants accumulated higher Na+ levels in the shoot than wild type, as ccc plants did with Cl. In addition, other parallels between SOS1 and CCC include expression in root tips, the plant vasculature at the xylem–symplast boundary and leaf hydathodes (Shi et al., 2002).

The impaired growth of different organs in ccc plants (Figure 6) may indicate the participation of CCC in cell elongation, a developmental process regulated primarily by the phytohormone auxin. Interestingly, there was an obvious parallel between sites of elevated free-auxin production and CCC expression. Stipules and hydathodes, leaf organs with strong CCC expression (Figure 4), have been identified as primary sites of high free-auxin production (Aloni et al., 2003). Other sites of auxin accumulation that correlates with CCC expression are the stamens of young flower buds (Aloni, 2004), trichomes and parenchyma cells adjacent to xylem vessels (Aloni et al., 2003). A large body of evidence indicates that auxin controls ion transport through the plasma membrane. While auxin-stimulated H+ efflux has been related to cell-wall loosening, a prerequisite for cell elongation, the induction of Cl and K+ uptake mechanisms is apparently also required for the sustenance of turgor pressure, another condition of cell growth (Babourina et al., 1998b; Long and Iino, 2001; Yamagami et al., 2004). We may think that the necessary presence of Na+ in the cotransport process is difficult to reconcile with the proposed turgor-regulation mechanism. However, Shabala and Lew (2002) demonstrated that fast turgor regulation in Arabidopsis epidermal root cells required the uptake of K+, Cl and Na+. The emergence and growth of inflorescence stems is one of the fastest organ-elongation events during Arabidopsis development, and therefore requires an efficient mechanism of nutrient and water acquisition. As ccc plants develop shorter stems, electrolyte mobilization necessary for efficient stem elongation may be mediated by CCC. Moreover, in ccc plants the frequent collapse of the stem-elongation zone, a powerful water sink due to a higher osmolite accumulation (Colmenero-Flores et al., 1999; Nonami and Boyer, 1993), appears to be compatible with the idea that a failure in the synchronization of physical growth and osmolite (or water) supply occurred during bolting.

The At CCC gene was also highly expressed in hydathodes, passive pores on the leaf margin involved in guttation (Figure 4m,n). The plant cotransporter may be implicated in this secretion phenomenon in two different ways: (i) mediating ion reabsorption from the guttation solution, similar to the function of animal CCCs in the kidney; (ii) as the water-driving force of guttation, resembling the involvement of some members of the animal CCC family in controlling water movement during cell-volume regulation.

In conclusion, evidence is presented in this work for the occurrence of cation–Cl cotransporters in the plant kingdom. The functional characterization of the At CCC gene indicates that the protein operates as an Na+:K+:Cl co-transporter, and analyses of ccc knockout lines strongly suggest its involvement in pivotal developmental processes and Cl homeostasis.

Experimental procedures

Plant materials and growth conditions

Arabidopsis thaliana ecotype Columbia (Col-0) and T-DNA insertion mutant lines obtained from this ecotype were used in the study. For soil-based analysis, seeds were sown on artificial soil (2:1:1 moss:vermiculite:perlite, pH adjusted to 5.75 with CaCO3), kept at 4°C for 3 days to synchronize germination, then placed in a plant growing chamber at 24°C under a 12-h light/12-h dark regime, with cool-fluorescent illumination providing a light intensity between 140 and 160 µE m−2 sec−1, relative humidity 72–75%.

For hydroponic culture, seeds were surface-sterilized, placed on half-strength MS medium (Murashige and Skoog, 1962) salts (Duchefa, http://www.duchefa.com) containing 2% sucrose solidified with 0.5% Phytoagar (Duchefa) and kept for 3 days at 4°C. Agar plates with cold-stratified seeds were then transferred to the above plant growth chamber. After emergence of two to three rosette leaves (8–10 days after stratification), plants were transferred to hydroponic containers with low-Cl medium (LC, approximately 20 µm residual Cl) and regular amounts of standard macro-elements (1.25 mm KNO3, 0.625 mm KH2PO4, 0.5 mm MgSO4, 0.5 mm Ca(NO3)2). Microelement content was as described by Arteca and Arteca (2000), pH was adjusted to 5.7 and medium was replaced every week. When Cl was required, the medium was supplemented with equimolar amounts of KCl and NaCl.

Isolation of At CCC cDNA

The cDNA of the A. thaliana CCC gene was cloned by reverse-transcriptase and PCR amplification using information from The Arabidopsis Information Resource database (TAIR, http://www.arabidopsis.org, AGI locus At1g30450). Total RNA from young leaves was extracted using the RNeasy plant mini kit (Qiagen, http://www1.qiagen.com). Total RNA (2 µg) was reverse transcribed with the ThermoScript kit for first-strand cDNA synthesis (Invitrogen, http://www.invitrogen.com) using an oligo dT primer. CCC cDNA was obtained by PCR amplification with gene-specific primers (forward primer: 5′-GATTCTCTGAAATCTGAAAAATCATCTC-3′; reverse primer: 5′-AAGGGCAGGCTTTTTGATGCTCATG-3′). Gene-specific primers were designed based on both EST (GenBank accession numbers AV825750 and AV793105) and genomic sequences.

Sequence analysis

Database searches looking for sequence similarities were performed using the blast program (Altschul et al., 1997) at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/BLAST). Multiple sequence alignments were made using the ClustalX program (Thompson et al., 1997). Phylogenetic and molecular evolutionary analyses were conducted using mega ver. 2.1 (Kumar et al., 2001). Genetic distances for amino acid sequences were calculated with the Poisson correction method (Nei and Chakraborty, 1976). The phylogenetic tree was constructed with the neighbour-joining (Saitou and Nei, 1987), unweighted pair group method with arithmetic mean (UPGMA; Swofford and Selander, 1981) and maximum parsimony (Fitch and Farris, 1974) methods, with the bootstrap test carried out with 5000 iterations. Transmembrane-spanning domains were predicted with the combined information obtained from the ARAMEMNON database (http://aramemnon.botanik.uni-koeln.de) and the PHD secondary structure prediction method from the Columbia University Bioinformatics Center, New York (http://cubic.bioc.columbia.edu/predictprotein). To scan amino acid sequences for the occurrence of patterns stored in the PROSITE database, ScanProsite from the ExPASy proteomics server of the Swiss Institute of Bioinformatics (http://www.expasy.org/tools/scanprosite) was used. Analysis of protein domains automatically generated from the SWISS-PROT and TrEMBL sequence databases was performed with the ProDom tool (Corpet et al., 2000) available at http://www.toulouse.inra.fr/prodom.html. The GenScan program from the GENSCAN web server at Pasteur Institute, Paris (http://bioweb.pasteur.fr/seqanal/interfaces/genscan.html) was used to predict the exon/intron structure of the plant CCC genes.

Functional expression of CCC in X. laevis oocytes

The A. thaliana CCC cDNA was cloned in the pGEMHE vector to induce enhanced expression in X. laevis oocytes (Liman et al., 1992). The KCC4 clone was as described previously (Mercado et al., 2000). cRNA was transcribed in vitro using the T7 RNA polymerase mMESSAGE mMACHINETM (Ambion, http://www.ambion.com) transcription system. cRNA product integrity was confirmed on agarose gels, and concentration was determined by absorbance reading at 260 nm (DU 640, Beckman, http://www.beckmancoulter.com). cRNA was stored frozen in aliquots at –80°C until use. Mature X. laevis oocytes were defoliculated and injected with either water or in vitro-transcribed cRNA (0.5 µg μL−1), following our previously described methods (Gamba et al., 1994; Plata et al., 1999, 2002). Oocytes were incubated during 4 days in frog Ringer ND96 containing sodium pyruvate and gentamicin. The night before uptake experiments, oocytes were switched to a Cl-free Ringer (Plata et al., 2002).

The function of the CCC protein was assessed by measuring tracer 22Na+, 86Rb+ or 36Cl uptake (NEN, http://las.perkinelmer.com/About+Us/PKI+Heritage/NEN.htm) according to described incubation protocols (Plata et al., 1999, 2002): 30 min incubation in K+- and Cl-free ND96 medium with 1 mm ouabain, followed by a 60-min uptake period in ND96, supplemented with 1 mm ouabain and 2.0 µCi 22Na+, 86Rb+ or 36Cl uptake in the absence of extracellular Na+, K+ or Cl in hypotonic (110 mOsm kg−1 H2O) or isotonic (220 mOsm kg−1 H2O, adjusted with sucrose) conditions for oocytes following our standard assay (Mercado et al., 2001). Statistical analyses were performed with statgraphics plus for windows, ver. 2.1 (Statistical Graphics, http://www.statgraphics.com).

At CCC promoter–GUS fusion construct lines

The A. thaliana CCC promoter region, from −739 to −16 (relative to the A of the predicted start codon) was amplified by PCR with the forward primer 5′-TAAGAAAGCTTCGAGCATTAGTTCTG-3′ and the reverse primer 5′-CCGTTCTAGAACAGATCTCTTTCGACAATG-3′, and cloned into pBI101 (Clontech, http://www.clontech.com), giving rise to a transcriptional fusion with the GUS–NOSt cassette. The construct was transformed into A. thaliana Col plants via the Agrobacterium tumefaciens-mediated floral dip method (Clough and Bent, 1998). Transgenic plants were selected on MS (0.5×) media (Murashige and Skoog, 1962) containing 50 mg L−1 kanamycin sulfate. T2 individuals homozygous for the t-DNA insertion were identified according to the percentage of kanamycin-resistant plants in the segregating T3 progeny. Five independent homozygous transgenic lines, tested for GUS activity, showed a consistent staining pattern. Three lines were selected for further analysis. For GUS histochemical staining, plant material was vacuum-infiltrated with GUS staining solution (50 mm sodium phosphate buffer pH 7.0, 0.5 mg ml−1 5-bromo-4-chloro-3indolyl-β-glucuronic acid) for 5 min and incubated overnight at 37°C. Tissue was cleared with 70% ethanol washings and analysed under a stereomicroscope (Olympus, http://www.olympus.com). For sectioning, stained organs were fixed in 100 mm 4% formaldehyde/sodium phosphate buffer pH 7.2, dehydrated in ethanol, briefly equilibrated in Histoclear (National Diagnostics, http://www.nationaldiagnostics.com) and embedded in Paraplast medium (Sigma-Aldrich, http://www.sigmaaldrich.com). Embedded tissue was cut into 10-µm sections using a microtome (Leica RM 2135) and studied under light microscopy (Leica ASLMD, http://www.leica.com).

Generation of transgenic Arabidopsis lines overexpressing the At CCC gene

The At CCC cDNA was amplified with the primers 5′-CTTGGATCCTCGAGAAATGGATAGCGGCGAC-3′ and 5′-TCGGGATCCTCTAGATTTAAACTATGTAAACAAAGTAAC-3′, and cloned in the binary vector pBI321 (Martínez-Atienza et al., 2007), leaving the At CCC gene under the regulation of the cauliflower mosaic virus 35S promoter and the NOS terminator. This construction was transformed into A. thaliana Col plants as explained above. Quantitative real-time RT–PCR analysis was used to measure At CCC transcript abundance in leaves of homozygous transgenic individuals. The transgenic line 142, with approximately 11-fold higher mRNA levels, was utilized for genetic complementation of the ccc-1 line.

Isolation of homozygous T-DNA insertion mutants

Information about the T-DNA lines was obtained from the SIGnAL web site at http://signal.salk.edu. Two T-DNA insertion lines that mapped into the At CCC gene, SALK-048175 (ccc-1) and SALK-145300 (ccc-2), were obtained from the Nottingham Arabidopsis Stock Centre at the University of Nottingham. The genotype of plants was determined by PCR using two pairs of primers: first, two gene-specific primers flanking the insertion point of the wild-type allele (Pccc-F: 5′-TAAGAAAGCTTCGAGCATTAGTTCTG-3′ and I1-R: 5′-CAACGAGCCCCATTTCTACTG-3′); second, the gene-specific primer Pccc-F and a T-DNA left border-specific primer 5′-GCGTGGACCGCTTGCTGCAACT-3′ (LBb1). Integrity of the CCC mRNA in the mutant lines was assayed through retro-transcription (RT) and PCR analysis with gene-specific primers flanking the position of the T-DNA inserts (Atkcc: 5′-ATGGATAGCGGCGACATTGAAGAA-3′ and RT-R: 5′-CGCGACCAATGAGGTAATATGG-3′) and designed to span five introns in order to distinguish genomic DNA contamination. The expression of the housekeeping α-tubulin gene TUA6 was used to control RNA loading. Amplification of cDNAs was arrested at the exponential phase of the PCR reaction (18 cycles for α-tubulin and 23 cycles for CCC). Fifteen microlitres of the CCC reactions and 5 µl of the α-tubulin reactions were separated on an 8-mm 1.5% agarose gel stained with ethidium bromide 1.2%.

Quantitative real-time RT–PCR analysis of CCC gene expression

Total RNA was extracted using the RNeasy plant mini kit (Qiagen) and treated with DNaseI for 1 h to ensure the absence of genomic DNA. When a very low amount of plant material was available from root tips, hydathodes and anthers, total RNA was extracted with the RNeasy Micro Kit (Qiagen). Reverse transcription and real-time PCR were performed in a single 20-μl capillary with a LightCycler 2.0 instrument (Roche Diagnostics, http://www.roche-applied-science.com/index.jsp) using double-stranded DNA-binding dye SYBR with the LightCycler FastStart DNA Master Sybr Green I kit (Roche Diagnostics). PCR was performed with gene-specific primers (5′-ACATGGATTGTTGGCATGGC-3′ and 5′-CGCGACCAATGAGGTAATATGG-3′) designed to span intron I5 in order to distinguish genomic DNA contaminants. Relative mRNA levels were normalized to total RNA amounts as described previously (Hashimoto et al., 2004). Cycle conditions were as follows: one reverse transcription cycle (48°C for 30 min); one cycle for activation of the Hot Start DNA polymerase (95°C for 10 min); and 45 cycles of denaturing (95°C for 10 sec), annealing (60°C for 15 sec) and elongation (72°C for 5 sec). After PCR, a melting cycle analysis was performed to ensure a single amplification product.

Ion content assays

Fresh and dry weight of samples was determined before and after drying for 2 days at 60°C. Samples were then ground with a Polytron (PT 2100, Kinematica, http://www.kinematica.ch), extracted with deionized water and filtrated (0.45 µm). Filtrates were analysed with an ion chromatograph (ICS-2000, Dionex, http://www1.dionex.com) equipped with an IonPac AS11-HC (250 × 4-mm) column. Cl was detected with an ASRS-4 mm suppressor column operating at 241 mA in the recycle mode. A 0- to 60-mm potassium hydroxide concentration gradient was used as eluent. Column and detector cell were adjusted at 40°C, and chromeleon 6.6 chromatography management software was used for system control, data processing and quantification. Measurements were replicated three times with similar trends. Statistical analyses were performed as described above.

Acknowledgements

We thank Joseph R. Ecker and the Salk Institute Genomic Analysis Laboratory for providing the sequence-indexed T-DNA insertion mutants, and the Nottingham Arabidopsis Stock Centre for distributing the seeds. We acknowledge Victoria San Andrés for technical support in plasmid constructions; Ana Conesa for assistance in statistical analyses; Javier Terol for assistance in phylogenetic analyses and manuscript reading; and F. J. Quintero for providing the pBI321 binary vector. This work was supported by the Spanish Ministerio de Ciencia y Tecnología through grant A6L2003-08502-C04-01.

Accession numbers: Sequence data from this article can be found in the EMBL/GenBank data libraries under accession numbers AM113986, BN000834 and BN000835.

Ancillary