We have identified a plasma membrane Na+/H+ antiporter gene from tomato (Solanum lycopersicum), SlSOS1, and used heterologous expression in yeast to confirm that SlSOS1 was the functional homolog of AtSOS1. Using post-transcriptional gene silencing, we evaluated the role played by SlSOS1 in long-distance Na+ transport and salt tolerance of tomato. Tomato was used because of its anatomical structure, more complex than that of Arabidopsis, and its agricultural significance. Transgenic tomato plants with reduced expression of SlSOS1 exhibited reduced growth rate compared to wild-type (WT) plants in saline conditions. This sensitivity correlated with higher accumulation of Na+ in leaves and roots, but lower contents in stems of silenced plants under salt stress. Differential distribution of Na+ and lower net Na+ flux were observed in the xylem sap in the suppressed plants. In addition, K+ concentration was lower in roots of silenced plants than in WT. Our results demonstrate that SlSOS1 antiporter is not only essential in maintaining ion homeostasis under salinity, but also critical for the partitioning of Na+ between plant organs. The ability of tomato plants to retain Na+ in the stems, thus preventing Na+ from reaching the photosynthetic tissues, is largely dependent on the function of SlSOS1.
Soil salinity disturbs ion homeostasis in plants, leading to membrane dysfunction, attenuation of metabolic activity and secondary effects that cause growth inhibition and, ultimately, cell death (Hasegawa et al. 2000). Salinity may reduce K+ uptake because of competition between Na+ and K+ for plasma membrane uptake sites because of physical–chemical similarities between both ions. This causes a nutritional deficiency in K+ with fatal consequences for plant cells because this ion is an indispensable macronutrient essential for many metabolic processes and a major contributor to cell turgor (Zhu, Liu & Xiong 1998; Chérel 2004). Therefore, it is crucial for the plant to maintain an adequate K+ concentration to prevent the negative effects of Na+ toxicity and osmotic stress associated.
Plants have cell transport mechanisms that contribute to salt tolerance and prevent excessive Na+ accumulation in the cytosol through Na+ sequestration into vacuoles or its extrusion out of the cell across plasma membrane (Apse & Blumwald 2007). In plants, cation/H+ antiporters are the main transport systems involved in Na+ and K+ homeostasis through vacuolar- and plasma membrane-associated transport processes. The process of exclusion of Na+ from the cytosol in exchange of H+ is an energy-costing process for the cell where primary proton pumps provide the necessary force to transport Na+ against electrochemical gradient (Blumwald, Aharon & Apse 2000). Only two classes of Na+/H+ antiporters have been well characterized in plants so far, belonging to either the NHE/NHX or NhaP (SOS1) subfamilies of the CPA1 family (Tester & Davenport 2003; Pardo et al. 2006). The first NHX protein described in plants was AtNHX1, a member of the vacuolar clade of NHE/NHX proteins (Gaxiola et al. 1999; Quintero, Blatt & Pardo 2000). NHX1 is involved in vacuolar Na+ compartmentalization minimizing the toxic accumulation of the ion in the cytosol (Blumwald et al. 2000 and references therein). Over-expression of NHX1 has been used to enhance salt tolerance in different plant species including tomato (Apse et al. 1999; Zhang & Blumwald 2001; Zhang et al. 2001).
The discovery and pioneer studies on sos mutants in Arabidopsis were crucial to understand the mechanism of salt stress tolerance and ion homeostasis in plants. The sos mutants are specifically hypersensitive to high external Na+ or Li+, and unable to grow under low external K+ concentrations (Zhu et al. 1998). While other Na+ efflux systems may exist, in Arabidopsis the defective phenotype of Atsos1 plants suggested that Na+ efflux is dominated by the plasma membrane-localized Na+/H+ antiporter AtSOS1 (Wu, Ding & Zhu 1996; Shi et al. 2000). SOS1 is part of a functional module, the SOS signal transduction pathway, that has been described as crucial for ion homeostasis and salt tolerance in plants (Zhu 2002, 2003). In this pathway, a calcium-binding protein, SOS3, senses cytosolic calcium changes elicited by salt stress (Liu & Zhu 1998; Ishitani et al. 2000). SOS3 physically interacts with and activates the serine/threonine protein kinase, SOS2 (Halfter, Ishitani & Zhu 2000; Liu et al. 2000). The SOS3/SOS2 kinase complex phosphorylates and activates the plasma membrane Na+/H+ exchanger encoded by the SOS1 gene (Qiu et al. 2002; Quintero et al. 2002; Shi et al. 2002). Over-expression of AtSOS1 improves salt tolerance in Arabidopsis by limiting Na+ accumulation in plant cells (Shi et al. 2003). SOS1 expression was observed at the epidermal cells of the root tip, implying a role of this transporter in extruding Na+ to the soil (Shi et al. 2002). The preferential expression of SOS1 in the cells surrounding the vasculature also suggested a role of this transporter in the control of long-distance Na+ transport in plants, because this ion is transported from the root to the shoot via the xylem.
The use of Arabidopsis thaliana mutants has been crucial to discover and study many genes involved in salt tolerance in plants. However, this plant might not be the best choice to understand how ion transporters involved in long-distance transport work in salt tolerance in crop plants (Essah, Davenport & Tester 2003; Tester & Bacic 2005). Tomato is a good alternative to study Na+ long-distance transport because of its physiological and anatomical structure, and its agricultural significance. This crop presents a high genotypic diversity related to ion homeostasis. In this sense, the more tolerant species accumulate higher amounts of salts in stems and leaves, while the more sensitive species accumulate salt mainly in roots (Cuartero & Fernandez-Muñoz 1999). Physiological evidences indicate that tomato roots could determine the Na+ concentration reaching aerial parts depending on the intensity of the stress (Estañet al. 2005), probably involving the SOS pathway.
In this work, we have isolated SlSOS1, a putative Na+/H+ antiporter gene from tomato highly homolog to AtSOS1. SlSOS1 has been functionally characterized in planta in order to gain insights of the role played by this transporter in tomato halotolerance. We have confirmed that SlSOS1 plays a crucial role in the survival of tomato plants under saline conditions, thus extending the observations from the model system Arabidopsis to a crop plant. We have gone a step further to show the critical importance of SlSOS1 in the stem to control Na+ and K+ distribution along the plant.
MATERIALS AND METHODS
Isolation of tomato SOS1
A degenerate pair of primers, corresponding to sequences of the conserved amino acidic regions of the plant SOS1 genes, was used (Supporting Information Fig. S1). Forward primer was 5′-GC(A/T)TGCAT(G/C)A(G/C)TT(C/T)TGGGAAATG-3′ and reverse 5′-GTCTGGACAGCAT(A/C)(A/G)TGAAGATG-3′. A 1.2 kb cDNA fragment from the middle region of the putative gene was amplified by RT-PCR (Enhanced Avian HS RT-PCR kit; Sigma-Aldrich, Madrid, Spain) using 5 µg total RNA from tomato (Solanum lycopersicum cv. Pera) stem (RNeasy plant mini kit; Qiagen, Hilden, Germany) and primers described earlier. Based on this 1.2 kb sequence, two additional primers were designed to amplify the full-length cDNA using the Smart 5′-3′ RACE cDNA Amplification Kit (Clontech; BD Biosciences, San Jose, CA, USA); forward primer, GSP2 (5′-GGTGCTGTAGCGTTGTCCCTTTC ATTG-3′) and reverse primer, GSP1 (5′-TCCAGATG AGGTATGAGGATGCACTGG-3′). First-strand cDNA synthesis was carried out using 1 µg total RNA and primers provided by the kit, following manufacturer's guidelines. For 3′RACE, 5 µL of 3′RACE-Ready cDNA was used as template together with a combination of GSP2 primer and the UPM primer provided in the kit. The 2.5 kb PCR fragment obtained was subcloned in pGEM-T Easy Vector (Promega Biotech Iberica, Madrid, Spain), and fully sequenced. For 5′RACE, 5 µL of 5′RACE-Ready cDNA was used as template and together with a combination of GSP1 primer and the UPM primer provided in the kit. PCR product was purified and sequenced. Contiguous sequences of cloned fragments and PCR products were assembled to get the full-length SlSOS1 (formerly LeSOS1) (deposited in GenBank/EMBL data libraries under accession number AJ717346). The open reading frame of SlSOS1 was obtained by PCR using 5′RACE-Ready cDNA as template and a pair of primers with engineered restriction sites to be subcloned as an XbaI–SalI fragment into the yeast shuttle vector pYPGE15 under PGK1 promoter (Brunelli & Pall 1993).
Yeast strains and media
The Saccharomyces cerevisiae strain AXT3K, lacking several major endogenous Na+ transporters (Δena1::HIS3::ena4,Δnha1::LEU2,Δnhx1::KanMX4), pSOS1-1 and pFL32T containing AtSOS1-1 and AtSOS2/AtSOS3 full-insert sequences, respectively, has been described elsewhere (Quintero et al. 2000, 2002), as well as pSOS2T/DΔ308 containing a truncated SOS3-independent hyperactive kinase version of AtSOS2 (Guo et al. 2001). Yeast transformation with the different plasmid constructs was performed using a standard lithium–polyethylene glycol method. Cells were grown at 30 °C in YPD medium (1% yeast extract, 2% peptone, 2% glucose) or APD (10 mm arginine, 8 mm phosphoric acid, 2% glucose, 2 mm MgSO4, 1 mm KCl, 0.2 mm CaCl2, trace minerals and vitamins) (Rodriguez-Navarro & Ramos 1984), which is essentially free of alkali cations. Na+ tolerance in drop tests was performed in APD medium supplemented with 1 mm KCl and different concentrations of NaCl, as indicated, and grown for 3–5 d at 30 °C (Venema et al. 2003).
RNA blot analysis
Tomato seeds of S. lycopersicum cv. Moneymaker were surface-sterilized, and then germinated and grown for 5 weeks in sterile perlite under a light irradiance of 180 µmol m−2 s−1 (16/8 h photoperiod) at 25/18 °C day/night temperatures, and 65–70% relative humidity. One-tenth-strength Hoagland's nutrient solution (Hoagland & Arnon 1950) was applied from the emergence of the first leaf, and raised to quarter-strength (2 weeks after sowing) thereafter. After 2 weeks, the plants were treated for 3 d with 0 or 225 mm NaCl. Tissues were harvested at 0, 1 and 3 d of salt treatment, immediately frozen in liquid nitrogen and stored at −70 °C.
DNA probe was obtained by PCR amplification of an 1148 bp fragment of the N-terminal coding region for SlSOS1, using gene-specific primers. Total RNA from salt-treated and non-treated tissues was isolated and purified according to Logemann, Shell & Willmitzerm (1987). Total RNA (15 µg) was run on 1.2% (w/v) denaturing formaldehyde agarose gels, and blotted onto nylon membranes (Hybond N+; Amersham Pharmacia Biotech, Buckinghamshire, UK) (Sambrook, Fritsch & Maniatis 1989). DNA probe was radioactively labelled with [α-32P]-dCTP by random priming using the Rediprime II kit (Amersham Pharmacia Biotech). Nylon filters were prehybridized and hybridized at 42 °C in ULTRAhyb Ultrasensitive Hybridization Buffer (Ambion-Europe, Ltd, Austin, TX, USA). Blots were washed twice in 2× SSC, 0.1 (w/v) sodium dodecyl sulphate (SDS), at 42 °C for 5 min each; twice in 0.1× SSC, 0.1 (w/v) SDS, at 42 °C for 15 min each; and once in 0.1× SSC, 0.1 (w/v) SDS, at 60 °C for 20 min. Nylon filters were exposed to Phosphorimager screen (BAS MS, Fugi Photofilm, Kanagawa, Japan), and hybridization signals were recorded with a Phosphorimager analyser (BioRad Molecular Imager System).
RNAi silencing of SlSOS1
Stable gene silencing via agrotransformation was done using a pKANNIBAL vector (Wesley et al. 2001) designed for producing a hairpin RNA construct of SlSOS1. A 459 bp PCR fragment encoding 156 amino acids of the central region of the long hydrophilic tail in the C-terminal end (amino acids from 763 to 918; Supporting Information Fig. S1) was obtained using primers 5′-GCGGATCCTCGAGCAAAGGCAACTAGAGTTTGGC-3′ (BamHI and XhoI restriction sites underlined) 5′-CGACAAGCTTGAATTCGAGCTGAAAGAAGCCCTTC-3′ (HindIII and EcoRI restriction sites underlined). The whole Not I cassette bearing the RNAi construct was subcloned into the corresponding site of the binary vector pART27, under the control of the CAMV35S promoter, which was introduced into Agrobacterium tumefaciens strain LBA4404 cells and used for plant transformation of S. lycopersicum cv. Moneymaker as described in Ellul et al. (2003).
Analysis of transgenic plants
The tomato plants proceeding from each independent experiment of transformation (F0) were screened by a PCR assay using pKANNIBAL-specific primers and DNA obtained from tomato leaves (Gen Elute Plant Genomic DNA miniprep kit; Sigma-Aldrich) in order to detect the presence of the RNAi construct. Positive plants were selected to study SlSOS1 expression pattern by semiquantitative RT-PCR, and those with a reduced expression were chosen for phenotypic analysis. Total RNA was isolated from roots, stem and leaves of tomato plants (RNeasy plant mini kit; Qiagen) and treated with RNAse-free DNase, according to the manufacturer's protocol. First-strand cDNAs were synthesized from 4 µg of total RNA and oligo-dT primer provided (Enhanced Avian HS RT-PCR kit; Sigma-Aldrich). Tomato elongation factor, ef1-α (acc. AB061263), was used as internal positive control for equal cDNA amounts; forward primer 5′-ATTGGAAACGGATATGCTCCA-3′ and reverse primer 5′-TCCTTACCTGAACGCCTGTCA-3′. cDNA for SlSOS1 was amplified by PCR using the reverse primer for RNAi construct and a forward primer sequence 5′-TATCTGTGCTGGATTTCTCCG-3′, not present in the gene-specific SlSOS1 cDNA fragment in the RNAi construct. Each sample was run in triplicate and repeated twice from pooled samples of independently silenced and non-silenced plants.
Several transformant lines F0 (L1, L2 and L3) were obtained, and subsequent F1 seeds were collected from each line for further experiments where different physiological parameters were studied under different saline conditions. F1 seeds from transformed plant lines were germinated on plates containing Murashige and Skoog (MS) (Murashige & Skoog 1962) media and kanamycin 300 mg L−1. Plants bearing three to four leaves were transferred to hydroponic culture system containing 2.5 L of aerated one-fourth-strength Hoagland's nutrient solution and placed in a glasshouse since September to June under a natural light irradiance (16/8 h photoperiod) supplemented with artificial light of 122 µmol m−2 s−1, at 24/18 °C day/night temperatures, and 40–50% relative humidity. After 2 weeks of growth in hydroponic culture, the plants were treated with or without 100 mm NaCl in fresh one-fourth-strength Hoagland's nutrient solution. Tissue samples of leaves, stems and roots corresponding to 6 plants per pot were washed four times with deionized water, weighed for fresh weight (FW) determination, oven dried for 48 h at 80 °C, weighed (dry weight) and prepared for mineral analysis by digestion in a HNO3 : HClO4 (2:1, v/v) solution. Na+ and K+ contents were determined by inductively coupled plasma spectrometry (ICP) (Ionomic Service; CEBAS–CSIC, Murcia, Spain). Relative growth rate (RGR) measures the quantity of biomass deposited per gram of biomass per unit of time, and was calculated using the following equation (Hunt 1982):
where W1 and W2 are total FW at harvest time 1 and 2, respectively. Net uptake rate (NUR) of Na+ and K+ in whole plants was calculated as the accumulation rate of each cation in the whole plant per dry weight unit of root and time unit according to the following equation (Guerrier 1996):
where M1 and M2 are the whole plant contents of either Na+ or K+ at harvest time 1 and 2, respectively; t2 – t1 is the time interval between harvests; and Rw is the root dry weights at each harvest. For the calculation of net translocation rates (NTRs) from roots to shoots, Na+ and K+ contents in the whole plants (M1 and M2 in Eqn 2) were replaced by their contents in stems or leaves. Fluxes to individual organs were calculated by Eqn 2, with M denoting the Na+ or K+ content in the organ, and Rw the dry weight of the organ under consideration.
Xylem sap was collected as root exudates after decapitation of shoots of mature hydroponically cultivated plants as described by Navarro et al. (2003). Briefly, the plants were decapitated above the roots, leaving the base of the stem, which was sealed with silicone grease inserted into a tapered plastic tube. The first exuded xylem sap, 10 min after decapitation, was discarded in order to prevent contamination of xylem sap with content from damaged cells or phloem sap. Xylem sap volume that emerged afterwards was collected as a function of time between 10 and 60 min. Sap flowed freely in plants from the control and the 25 mm NaCl 3 d treatments, and a few microlitres could be collected only after 1 d treatment in 100 mm NaCl, and nothing thereafter. The roots were removed and weighted. Sap flow was expressed in mg h−1 g−1 root FW. Na+ and K+ concentrations in xylem sap were determined by ICP spectrometry (Ionomic Service, CEBAS–CSIC). Na+ and K+ fluxes into the xylem were estimated as the product of ion concentration by sap flow, and expressed as µmol h−1 g−1 root FW.
Purification of plasma membrane vesicles
Microsomal membrane fractions of 100 mm NaCl-treated silenced and non-silenced tomato root tissues were isolated as previously described (Ballesteros, Donaire & Belver 1998). Plasma membrane vesicles were purified from microsomal fraction by phase partitioning following the procedures described in Buckhout et al. (1988) for tomato roots. Protein content was determined in the presence of Triton X-100 (Gogstad & Krutsnes 1982).
Na+/H+ exchange assay
Na+/H+ exchange activity was measured at 26 °C by the initial rate of dissipation of inside-acidic pH gradients across plasma membrane vesicles, previously generated by the plasma membrane H+-ATPase, using the fluorescence quenching of ACMA as described by (Ballesteros et al. 1997). H+ gradient formation by the H+-ATPase was assayed by incubating plasma membrane vesicles (50 µg protein) in a reaction medium (1 mL) containing 10 mm BTP–MES, pH 7.0; 25 mm KNO3; 1 µm ACMA; 1.5 mm ATP–BTP, pH 7.0; and 250 mm mannitol, and then started by the addition of 1.5 mm MgSO4. When fluorescence reached a steady-state level, then Na+ (as Na2SO4) was added to determine the initial rate of Na+-dependent fluorescence recovery per minute, the total fluorescence recovery (100%) being the difference between the maximal level of quenching at steady-state level and the maximal level of fluorescence obtained upon addition of 10 mm NH4SO4 at the end of the assays, which abolish pH gradient. Fluorescence was recorded with a spectrofluorimeter containing a thermostated and stirred cell (model QM2000; Photon Technology International, Sussex, UK) at excitation and emission wavelengths of 415 and 485 nm, respectively.
Tomato SOS1 isolation and functional analysis in yeast
By using RT-PCR, degenerated primers corresponding to sequences conserved in SOS1-like protein from other plant species, and subsequent 5′RACE, a full-length SOS1 cDNA from tomato was obtained. The nucleotide sequence is predicted to encode a protein highly homologous to the plasma membrane Na+/H+ antiporter SOS1 from Arabidopsis and other plant species, and sharing common features such as 12 transmembrane regions and a large cytoplasmic tail, a Na+/H+ exchange domain and a putative cyclic nucleotide-binding domain centrally located in the long C-terminal tail (Supporting Information Fig. S1; Pardo et al. 2006; Oh et al. 2007; Aramemnon plant membrane database, http://crombec.botanik.uni-koeln.de/).
SlSOS1 was expressed in the AXT3K yeast strain which is unable to grow in arginine phosphate (AP) medium with Na+ concentrations higher than 70 mm (Quintero et al. 2000, 2002). Expression of SlSOS1 alone allowed cells to grow in AP medium containing up to 100 mm NaCl (Fig. 1). Co-expression of SlSOS1 together with the Arabidopsis SOS2/SOS3 kinase complex or the hyperactive kinase version of SOS2 (SOS2T/DΔ308), that is independent of SOS3 (Guo et al. 2001, 2004), improved the halotolerance of yeast cells, allowing them to grow in a medium containing up to 300 mm NaCl. These results clearly indicate that the SOS1-like protein from tomato is orthologous of AtSOS1.
Silencing of SlSOS1
In order to study the function of SlSOS1 gene in planta, the gene was silenced in transgenic tomato. Twenty-eight transgenic lines F0 harbouring the RNAi construct were identified. Out of the 28 positive lines, three were randomly selected for further analysis, showing reduced SlSOS1 gene expression. Silencing level in leaves was found to be 70% for line 1 (L1), 40% for line 2 (L2) and 60% for line 3 (L3), as determined by semiquantitative RT-PCR (Supporting Information Fig. S2). The effect of SlSOS1 silencing on the salt tolerance of tomato was studied in heterozygous F1 plant lines. The salt treatment, 100 mm NaCl for 5 d, strongly affected plant growth in all silenced lines (Fig. 2a). Indeed, the observed growth reduction was 67% for L1, 57% for L2 and 75% for L3 compared to plants grown under normal conditions, whereas in the case of wild type (WT) plants it was only 16%. These results were indicative of an increased sensitivity to salt stress in the silenced lines. SlSOS1 transcript levels, in WT and silenced plants, in different tissues under saline conditions were examined (Fig. 2b). Transcript levels of SlSOS1 were reduced in leaves and stems of silenced plants. In roots, although there was a reduction in the transcript level of SlSOS1 in silenced lines, it was not as high as the observed in the other tissues. In tomato WT, there was a great increase in gene expression in roots after salt treatment. These results are in agreement with the increase in transcript level observed by Northern blot in WT tomato treated for 3 d with 225 mm NaCl (Fig. 2b). Taking into account the mRNA levels measured in all the experiments carried out and all the tissues analysed, we established that the silencing level in these heterozygous F1 lines was similar to the values obtained in F0.
Transgenic line L1 showed the highest growth reduction in 100 mm NaCl media and the strongest gene silencing (∼70%). Therefore, F2 seeds from L1 selected in kanamycin were used for further experiments in which different physiological parameters were studied. To evaluate the difference in plant size, RGR was measured after 5 d of salt treatment (Fig. 3). The accumulation of new plant biomass upon stress treatment depends on the initial plant size, duration of treatment and growth rate. RGR eliminates differences in biomass production related to treatment duration and/or initial plant size (at the beginning of salt treatment). Thus, RGR gives a relative basis for comparison of the effect of salt on plant growth among species and genotypes (Hunt 1982). The silenced plants showed a reduced growth after 5 d under control conditions and, above all, under salt treatment compared to WT plants. Salinity had no effect on the RGR of WT plants, which grew the same regardless of the treatment. The difference in RGR between silenced and WT was even higher under salt stress, being approximately five times lower for SlSOS1-silenced plants at 100 mm NaCl.
In order to check whether SlSOS1 gene silencing correlated with a proportional reduction in plasma membrane Na+/H+ exchange activity, plasma membrane vesicles were purified by two-phase partitioning from microsomal fractions from L1 and WT tomato roots treated with 100 mm NaCl for 3 d. Salt treatment was applied in order to get a maximal SlSOS1 expression (Fig. 2b and Qiu et al. 2002). Na+/H+ exchange activity was tested in isolated plasma membrane vesicles from both WT and SlSOS1-silenced roots. Purity of plasma membrane-enriched fractions was checked using inhibitors specific for different subcellular ATP-dependent H+ transport activities (Supporting Information Table S1). Firstly, an ATP-dependent H+ gradient across the vesicles was generated by activation of the plasma membrane H+-ATPases (Fig. 4a). The H+-pumping activity, measured as initial rate of ACMA quenching, was significantly lower in SlSOS1-silenced than in WT roots, suggesting that SlSOS1 silencing affected the ATP-dependent H+ transport activity of the plasma membrane H+-ATPase (Supporting Information Table S1). At the steady state of ΔpH, addition of sodium salts (as Na2SO4) resulted in fluorescence recovery, or dissipation of the pH gradient, which indicated Na+-dependent H+ efflux from the vesicles. This fluorescence recovery was significantly lower in membranes from SlSOS1-silenced roots than in WT roots (Fig. 4a). The concentration dependence of the Na+-dependent H+ efflux was tested by adding different amounts of Na2SO4 (Fig. 4b). The Na+/H+ exchange activity was clearly lower for vesicles from SlSOS1-silenced roots under all tested concentrations. Na+/H+ exchange activity appeared to be specific and not because of passive, electrochemically coupled exchange, because the addition of K+ and valinomycin did not affect the rate of Na+-dependent H+ movements (data not shown).
SlSOS1 silencing alters tissue Na+ and K+ contents
The distribution of Na+ and K+ contents in plant organs was analysed. When SlSOS1-silenced and WT plants were exposed to mild and severe salt stress (25 or 100 mm NaCl), their Na+ contents increased following different patterns (Fig. 5). At 25 mm NaCl, a mild saline stress for tomato (see Fig. 3) suppressed lines showed greater Na+ content than control plants in leaves and roots, but not in stems, 3 and 5 d after the onset of treatment (Fig. 5). This is to be expected from the known role of SOS1 in Na+ efflux in roots (Shi et al. 2002). By contrast, at 100 mm NaCl, Na+ contents were also higher in leaves and roots of the suppressed lines, but significantly lower in stems compared to control plants, even at the first sampling 1 d after stress imposition. Notably, the Na+ content in the stems of the suppressed plants remained fairly constant (3–4% at 3–5 d) regardless of the intensity of stress, whereas it doubled in stems of control plants subjected to 100 mm NaCl relative to 25 mm NaCl (5–6% Na+ content versus 2.5–3%; Fig. 5b). Also worth noting is that at 25 mm NaCl, the greatest Na+ accumulation in control plants was in stems at any time point, whereas the suppressed plants showed preferential deposition in roots, particularly at longer times of treatment (Fig. 5). Together, these data indicate that SlSOS1 is also involved in the root-to-shoot export of Na+, presumably by acting at xylem loading (Shi et al. 2002). Because of the known implications of SOS1 in K+ nutrition (Wu et al. 1996), the amount of this ion was analysed as well. Total K+ accumulation in whole plants was lower in salt-treated plants than in controls after 5 d of both mild (25 mm NaCl) and severe salt treatment (100 mm NaCl) (data not shown). However, while no difference was observed between WT and SlSOS1-silenced plants in the shoot (data not shown), the difference was striking in roots (Fig. 6). In 25 mm NaCl (Fig. 6a), the roots of the suppressed plants, but not WT, showed a significant drop in K+ content after day 3. Under severe stress (100 mm NaCl), the roots of WT plants showed a K+ loss after the third day. At day 5, the content in the suppressed lines was significantly lower than in WT plants (Fig. 6b).
Physiological parameters as the rate of net uptake of K+ and Na+ in whole plant (NUR), NTR of K+ and Na+ from roots to shoots and net fluxes to individual organs were calculated as described by Guerrier (1996), as a measure of the net balance between total cation influx and efflux in whole plant, cation fluxes from roots to shoots, as well as the partitioning of Na+ among different organs, in a time interval between day 1 and day 5 after stress imposition. Plants of the suppressed line L1 showed a greater NUR as expected from their compromised extrusion of Na+ out of the root (Table 1). Translocation (NTR) of Na+ from roots to shoots (i.e. stem, petioles and leaves together) was also higher in the suppressed plants compared to control plants under saline conditions (Table 1). However, net fluxes of Na+ to specific organs indicated that Na+ in the suppressed plants was preferentially accumulated in leaves, while WT plants retained Na+ mostly in stems under severe salt stress (Table 2). Apparently, stems of SlSOS1-silenced plants seemed to have lost significant capacity to retain Na+. Net flux into roots was also higher in roots of L1, but only at 25 mm. Unexpectedly, Na+ net flux into roots was similar for both types of plants under severe salt stress, which may indicate that the imposed Na+ gradient at 100 mm NaCl had greatly exceeded the capacity of roots to maintain homeostasis by extruding Na+ back to the medium (Tester & Davenport 2003). This is supported by the much greater Na+ content in the roots of control plants in 100 mm NaCl compared to 25 mm NaCl, and approaching that of SlSOS1-silenced plants (Fig. 5c). Potassium NUR in whole plants (Table 1) was nearly zero in both control and L1 plants under mild saline conditions, but very negative in 100 mm NaCl, signifying a strong K+ leak in root or K+ efflux. Net K+ flux into leaves, stems and, above all, roots from the suppressed plants was lower, even negative under severe salt stress, indicating a K+ lost (Table 2). Usually, salt stress causes an excessive leakage of K+ from the cell (Shabala et al. 2005). In SlSOS1-silenced plants where the main mechanism for Na+ extrusion out of the root is impaired, this leak seems even greater.
Table 1. Net uptake rates (NURs) in whole plants and translocation to shoots [net translocation rate (NTR)] for K+ (expressed as µmol g−1 root DW d−1) and Na+ (expressed as mmol g−1 root DW d−1) in plants of tomato Moneymaker [wild type (WT)] and SlSOS1-silenced L1 grown for 5 d under control or salinized conditions with 25 or 100 mm NaCl
Values were obtained from data means at harvest time between 1 and 5 d of at least eight plants of a representative experiment.
DW, dry weight.
25 mm NaCl
100 mm NaCl
25 mm NaCl
100 mm NaCl
Table 2. Net K+ and Na+ fluxes (µmol g−1 DW d−1) to leaves, stems and roots in tomato wild type (WT) and SlSOS1-silenced (L1) treated with 0, 25 and 100 mm NaCl for 5 d
Salt treatment (mm)
Results are means of 12 plants. Values were obtained from data means at harvest time between 1 and 5 d of at least eight plants of a representative experiment.
Measurements of Na+ concentration in xylem sap collected as root exudates showed that both SlSOS1-silenced and WT plants accumulated more Na+ in xylem sap because of salt treatments, being Na+ content always higher for SlSOS1-silenced plants (Table 3). However, Na+ fluxes into the xylem were significantly reduced by salt stress especially in the suppressed plants, probably caused by a strong reduction (8 to 15-fold) in xylem sap flux under these conditions (Table 3). Indeed, net Na+ fluxes carried over by xylem sap movement were much lower in L1 plants (Table 2). Values for the suppressed plants grown for 3 d with 100 mm were not determined because of lack of root pressure. The same pattern was observed for K+ xylem concentration and fluxes (Table 3). These results indicate a role of SOS1 in xylem loading for long-distance transport, as previously suggested in Arabidopsis by Shi et al. (2002).
Table 3. Xylem sap flux, Na+, K+ concentrations, and Na+, K+ fluxes in xylem sap from roots of tomato wild type (WT) and SOS1-silenced L1 grown for 3 d with 0, 25 or 100 mm NaCl
Days in salt
Values are means ± SD of three plants. Values followed by different letters in each row differ significantly (Duncan test, P = 0.05).
FW, fresh weight; ND, not determined.
Xylem sap flux
141 ± 39e
217 ± 31f
82.3 ± 5.0d
10.6 ± 4.1b
21.5 ± 8.8c
1.4 ± 0.6a
(mL h−1 g−1 root FW)
48.0 ± 13.2c
101 ± 30e
25.1 ± 11.3bc
6.7 ± 0.3a
Sodium in xylem sap (mm)
0.1 ± 0.1a
0.21 ± 0.0a
14.4 ± 3.4b
25.1 ± 1.6c
73.0 ± 17d
168 ± 20e
0.1 ± 0.1a
0.14 ± 0.1a
27.2 ± 2.2b
34.1 ± 3.6b
Sodium flux into xylem sap
0.01 ± 0.00a
0.04 ± 0.02a
1.00 ± 0.28c
0.27 ± 0.10b
1.50 ± 0.21c
0.31 ± 0.20b
(mmol h−1 g−1 root FW)
0.01 ± 0.00a
0.01 ± 0.00a
0.65 ± 0.22c
0.23 ± 0.01b
Potassium in xylem sap (mm)
13.9 ± 3.5a
9.2 ± 2.2a
23.1 ± 10.7ab
48.7 ± 12.6c
19.4 ± 9.9ab
49.6 ± 7.8c
8.58 ± 1.1a
5.07 ± 1.91a
25.6 ± 8.1b
23.1 ± 5.6b
Potassium flux into xylem sap
1.96 ± 0.30c
2.01 ± 0.24c
1.90 ± 0.35c
0.48 ± 0.10b
0.38 ± 0.05b
0.07 ± 0.04a
(mmol h−1 g−1 root FW)
0.41 ± 0.07b
0.54 ± 0.34b
0.62 ± 0.21b
0.15 ± 0.03a
Since the identification in Arabidopsis of the Na+/H+ antiporter SOS1 and the establishment of its critical role in salt stress tolerance (Shi et al. 2000, 2003), SOS1 genes from different plant sources have been functionally characterized, mainly in heterologous systems (Garciadeblás, Haro & Benito 2007; Martinez-Atienza et al. 2007; Oh et al. 2007; Wu et al. 2007; Xu et al. 2008). In the present study, we have isolated, cloned and characterized in planta a SOS1 gene from tomato. We provide evidence of the importance of SOS1 in the salt tolerance of a crop species and how gene silencing affects ion distribution in the plant.
Heterologous expression of SlSOS1 in yeast was able to restore the halotolerance of AXT3K lacking major Na+ transporter systems (Fig. 1). As with Arabidopsis and rice counterparts (Quintero et al. 2002; Martinez-Atienza et al. 2007), SlSOS1 co-expression with either AtSOS2/SOS3 complex or a hyperactive and SOS3-independent AtSOS2 mutant could completely restore AXT3K halotolerance allowing it to grow on 300 mm NaCl. This salt concentration was considerably higher than that tolerated by AXT3K expressing the Arabidopsis SOS system (Quintero et al. 2002), and similar to that of AXT3K expressing the rice SOS system (Martinez-Atienza et al. 2007).
Similar to results previously obtained with the Arabidopsis sos1 mutant (Qiu et al. 2002, 2003), vesicles from SlSOS1-silenced plants showed a reduced plasma membrane Na+/H+ exchange activity (Fig. 4), which correlated with the reduction in the level of SOS1 transcript (Fig. 2) and their greater salt sensitivity compared to WT plants (Fig. 3). We also found that the ATP-dependent H+ transport in isolated plasma vesicles from tomato roots of suppressed plants was affected (Supporting Information Table S1). These results support the idea that SOS1 or SOS signal transduction pathway somehow controls H+ pumping within the mature zone of the root (Shabala et al. 2005).
Previous studies showed that SOS1 is not essential for normal plant growth and development, but critical for the salt tolerance of Arabidopsis (Wu et al. 1996; Shi et al. 2002). Here, we show that silencing SlSOS1 had strong negative effects on growth under salt stress, thus expanding these observations from the model plant to a crop species. Our data are consistent with the notion that the main action for SOS1 is to extrude Na+ out of the root cells (Qiu et al. 2002; Shi et al. 2002, 2003; Fig. 7). NURs of Na+ in whole plants under both saline conditions used were threefold higher in SlSOS1-silenced plants than in WT plants (Table 1). As discussed by Tester & Davenport (2003), the rates of unidirectional entry of Na+ (influx) in roots are very high, but they do not translate into a commensurate Na+ accumulation (net uptake). The usually much lower rate of net Na+ uptake (less than 10-fold) than unidirectional influx implies a substantial efflux rate. By some estimates, about 70–90% of the Na+ fluxed into the root symplasm is extruded back to the apoplast (Tester & Davenport 2003). Thus, even small changes in the rate of Na+ efflux, to which SOS1 is known to be critical, would have dramatic effects in the root Na+ content. In tomato, salt tolerance is associated with Na+ exclusion and lower Na+ uptake rates (Cano et al. 1991; Taleisnik & Grunberg 1994; Guerrier 1996), and our data indicate that SlSOS1 is a critical component in Na+ extrusion.
Based on the expression pattern of AtSOS1 and the physiological characterization of Arabidopsis sos1 plants, Shi et al. (2002) suggested that AtSOS1 controls long-distance Na+ transport. Arabidopsis sos1 mutant accumulated less Na+ in shoot than WT under mild stress condition, probably because of a reduction in the active loading of Na+ into the xylem. However, Arabidopsis sos1 mutant accumulated more Na+ on shoots than WT under severe salt stress (100 mm NaCl). One limitation of these studies is that the very short stem of Arabidopsis plants did not allow a precise dissection of the relative content of Na+ in stem versus leaf, and this separation would be critical to assess a role of SOS1 in xylem loading and/or unloading, Na+ export by roots and retention in stems. In tomato, SlSOS1-silenced plants under saline conditions accumulated more Na+ in roots and leaves than control plants under both mild and severe types of salt stress (Figs 5 & 6). By contrast, the stems of the suppressed plants accumulated similar amounts of Na+ than controls in 25 mm NaCl, but significantly less under severe stress (100 mm NaCl). In fact, average Na+ contents in stems of the suppressed plants remained mostly unchanged, whereas those in control plants doubled in 100 mm NaCl compared to 25 mm NaCl. The Na+ content in stems of control plants was greater than that in roots and leaves, particularly under mild stress. This ability to retain Na+ in stems was clearly lost in the suppressed plants, in which an acropetal gradient was established. Roots have a remarkable ability to control their own Na+ content, and many plants regulate Na+ delivery to the shoot (Tester & Davenport 2003). As most glycophytes, tomato plants exclude salts from the photosynthetic tissues in leaves by removing Na+ from the root while retaining it in the stem (Taleisnik & Grunberg 1994; Guerrier 1996). By contrast, SOS1-suppressed plants simply showed a root-to-leaf gradient of Na+ distribution together with a reduced ability to accumulate Na+ in the stem under severe salt stress (Fig. 5). At 25 mm NaCl, a mild treatment that permits the comparison under less stringent conditions, particularly for L1 plants, the net fluxes of Na+ into plant organs from day 1 to 5 after stress imposition were almost 10- and 40-fold higher in roots and leaves of the suppressed plants compared to WT (Table 2), whereas the difference in stems was less than twofold.
Increasing amounts of Na+ in roots should translate, in principle, into higher translocation rates through the xylem. Indeed, the concentration of Na+ in the xylem exudates of detached roots was higher in the suppressed plants than in controls (Table 3). However, this finding seems at odds with the proposal that SOS1 loads Na+ into the xylem (Shi et al. 2002). We suggest that the uptake of Na+ into the xylem sap of the suppressed lines is probably caused by passive loading of the ion because higher cytosolic Na+ concentration in the root xylem parenchyma cells and a relatively depolarized plasma membrane would favour Na+ movement into the xylem (Apse & Blumwald 2007). In other words, a greater difference in the concentration of Na+ in the xylem sap of the control and suppressed plants is to be expected, considering the large difference in the Na+ content in roots and the greater accumulation of Na+ in leaves of the suppressed plants. For instance, the Na+ content in roots of the suppressed plants after 3 d in 25 mm NaCl was threefold greater than in control roots (Fig. 5c), whereas the small increment in the Na+ concentration in the xylem sap was below statistical significance (Table 3). It is also intriguing that net Na+ transport within the xylem sap was actually lower in the suppressed plants than in WT, essentially because of a drastic reduction in its xylem sap flux (Table 3), and yet the deposition of Na+ in the suppressed leaves was greater. Again, this points out to a role of SOS1 in Na+ efflux at the xylem parenchyma in leaves restricting the unloading of Na+ (Fig. 7), but we cannot rule out that the sap flow rate determined using detached root system did not reflect the natural transpiration conditions of intact plants (Schurr 1998). On the other hand, proteins belonging to the HKT family are known to control Na+ unloading in the xylem of Arabidopsis, rice and wheat (Ren et al. 2005; Byrt et al. 2007; Davenport et al. 2007), and by doing so they ultimately control the long-distance transport of Na+ to the leaf. Insufficient SOS1 activity is also known to affect the function of the K+ channel AKT1 (Qi & Spalding 2004) and the plasma membrane H+-ATPase (Shabala et al. 2005) in Arabidopsis. A functional link and possible interplay between SOS1 and HKT1;1 have also been suggested (Rus et al. 2004). It appears that the transport functions of the SOS and HKT systems are coordinated to achieve Na+ (and K+) homeostasis (Pardo et al. 2006; Fig. 7). Dysfunction of either system alters long-distance transport and adequate partition of Na+, thereby resulting in salt-sensitive phenotypes. It remains to be seen whether the depletion of SOS1 in suppressed tomato does indeed affect the function of HKT-like proteins in the xylem parenchyma of tomato.
We have observed that SlSOS1-silenced plants are more sensitive to salt stress than WT. Because K+ is an essential element for many metabolic processes, it was important to establish how salt stress altered K+ homeostasis. In this regard, SlSOS1-silenced plants suffered a great loss of K+ in roots after 5 d exposure to 100 mm NaCl. These results are consistent with those reported with Arabidopsis plants. These studies showed that mutation of AtSOS1 caused a defect in K+ uptake and that the functioning of the entire root was affected, causing an excessive leakage of K+ (Wu et al. 1996; Zhu et al. 1998; Shabala et al. 2005). Qi & Spalding (2004) reported the impairment by Na+ of K+ permeability in root of the sos1 mutant.
Our results provide evidence of the importance of the SOS1 antiporter in maintaining ion homeostasis in tomato, a crop plant that is a Na+ excluder with the capacity to tolerate moderate salt concentrations (Taleisnik & Grunberg 1994; Guerrier 1996; Cuartero & Fernandez-Muñoz 1999). We also show that, besides its main action in extruding Na+ out the root, SlSOS1 is critical for the partitioning of Na+ in plant organs and the ability of tomato plants to retain Na+ in the stems, thus preventing Na+ from reaching the photosynthetic tissues.
We thank Dr Francisco Javier Quintero and Dr Susana Rivas for critically reading the manuscript. This work was supported by grants BIO2006-01955 (A.B.) and BFU2006-06968/BMC (J.M.P.) from the Ministerio de Educación y Ciencia, and by CVI 124 (A.B.) and CVI 148 (J.M.P.) from the Consejería de Innovación, Ciencia y Empresa, Junta de Andalucia. R.O. was supported by a grant of Programa Averroes from the Consejeria de Inovación, Ciencia y Empresa, Junta de Andalucia; J.L. by a grant from the Spanish Ministerio de Educacion y Ciencia; and P.A.M. by an I3PCPG fellowship from CSIC.