Critical role of divalent cations and Na+ efflux in Arabidopsis thaliana salt tolerance

Authors


Correspondence: Frans Maathius. Fax: + 44 1904 434 317; e-mail: fjm3@york.ac.uk

Abstract

Detrimental effects of salinity on plants are known to be partially alleviated by external Ca2+. Previous work demonstrated that the Arabidopsis SOS3 locus encodes a Ca2+-binding protein with similarities to CnB, the regulatory subunit of protein phosphatase 2B (calcineurin). In this study, we further characterized the role of SOS3 in salt tolerance. We found that reduced root elongation of sos3 mutants in the presence of high concentrations of either NaCl or LiCl is specifically rescued by Ca2+ and not Mg2+, whereas root growth is rescued by both Ca2+ and Mg2+ in the presence of high concentrations of KCl. Phenocopies of sos3 mutants were obtained in wild-type plants by the application of calmodulin and calcineurin inhibitors. These data provide further evidence that SOS3 is a calcineurin-like protein and that calmodulin plays an important role in the signalling pathways involved in plant salt tolerance. The origin of the elevated Na : K ratio in sos3 mutants was investigated by comparing Na+ efflux and influx in both mutant and wild type. No difference in Na+ influx was recorded between wild type and sos3; however, sos3 plants showed a markedly lower Na+ efflux, a property that would contribute to the salt-oversensitive phenotype of sos3 plants.

Introduction

Estimates of the total area of saline land world-wide vary between 340 and 950 × 106 Ha, which represents between 2·3 and 6·4% of the total land surface (Flowers et al. 1986). NaCl is the predominant salt in saline environments, but other salts, such as Na2SO4, MgSO4, CaSO4, MgCl2, KCl and Na2CO3, may be present (Flowers et al. 1977). Most crop species are not tolerant of high concentrations of salt and this impacts negatively on agricultural production in regions with saline soils.

In addition to inducing general osmotic stress, high concentrations of Na+ are toxic. Not only does Na+ compete with K+ for uptake into the root, but competition between Na+ and K+ for K+ binding sites in the cytoplasm can also inhibit K+-dependent processes (Hall & Flowers 1973; Wyn Jones & Pollard 1983).

The detrimental effects of salinity can be partially alleviated by external Ca2+. For example, LaHaye & Epstein (1969) reported enhanced salt tolerance in Phaseolis vulgaris L. grown on 50 mm NaCl with more than 3 mm CaSO4. High CaSO4 concentration reduced Na+ content in both roots and shoots, with the most dramatic effect being that reported for leaves.

A greater understanding of the relationship between K+, Na+ and Ca2+ in plants has been gained from the study of Arabidopsis thaliana mutants. Zhu and colleagues identified a number of NaCl-hypersensitive mutants from chemically or fast-neutron mutaganized A. thaliana seeds. Several loci were identified that are responsible for the salt-oversensitive (SOS) phenotype (Wu et al. 1996; Liu & Zhu 1997; Zhu et al. 1998). One of these, sos1, is characterized by a defective high-affinity K+ uptake system and a lower level of Na+ accumulation compared to wild-type plants (Ding & Zhu 1997). In contrast, sos2 mutants, when challenged with high Na+, accumulate more Na+ and retain less K+ than wild-type plants (Zhu et al. 1998) whereas sos3 also shows defective high-affinity K+ uptake but a higher level of Na+ accumulation compared with wild-type plants (Liu & Zhu 1997). The hypersensitivity to NaCl of sos3 mutants could be partially reversed in the presence of high external Ca2+ concentrations.

The SOS1 to SOS3 genes have been identified and subsequent work has shown that they may be part of a linear pathway that is involved in Na+ and K+ homeostasis. SOS1 expression is upregulated after salt stress and encodes a putative Na+/H+ antiport that is predominantly expressed around the xylem tissue. Its expression pattern suggests that SOS1 plays a role in loading Na+ into the xylem (Shi et al. 2000). SOS3 was found to share sequence similarity with CnB, the regulatory B-subunit of protein phosphatase 2B (calcineurin), and with neuronal calcium sensors from animals, suggesting a potential role of Ca2+ binding proteins and Ca2+ signalling in the salinity tolerance mechanism of higher plants (Ding & Zhu 1997). The SOS2 gene is expressed in both root and shoot material (Liu et al. 2000) and predicted to encode a serine/threonine protein kinase that is required for salt tolerance in A. thaliana. Furthermore, in the presence of Ca2+, SOS3 activates the kinase activity of SOS2 by promoting an interaction between the C-terminal regulatory domain of SOS2 and SOS3 (Halfter et al. 2000). Interestingly, although SOS3 is CnB-like, the critical element in the signal transduction pathway is a protein kinase (SOS2), rather than a protein phosphatase (CnA), as is the case in yeast (Serrano et al. 1999).

A range of intriguing questions remains concerning SOS3 function in salt tolerance. The extent to which the sos3 phenotype can be rescued by Ca2+ specifically is of interest in relation to previous reports about the alleviation of NaCl toxicity. The extent to which putative inhibitors of SOS3 action can generate phenocopies of the sos3 mutant in wild-type plants remains to be investigated. The basis for the overaccumulation of Na+ by sos3 mutants was addressed by measuring unidirectional Na+ influx and efflux in both mutant and wild-type plants.

Materials and methods

Ca2+/Mg2+ rescue of seedlings under saline stress

Seeds of wild-type A. thaliana L. Columbia and sos3 (kindly donated by Dr J.-K. Zhu, University of Arizona, USA) were surface-sterilized and sown in rows onto agar plates for germination. The plating medium contained Murashige and Skoog (MS) salts: 20·6 mm NH4NO3, 2·3 mm CaCl2, 0·7 mm MgSO4, 18·9 mm KNO3 and 1·3 mm KHPO4 to provide macronutrients; 0·1 mm H3BO3, 0·1 µm CoCl2, 0·1 µm CuSO4, 0·1  mm disodium ethylenediaminetetraacetic acid (EDTA), 0·1 mm FeSO4, 0·1 mm MnSO4, 5 µm KI, 1 µm Na2MO4 and 30 µm ZnSO4 to provide micronutrients; plus 3% (w/v) sucrose and 1·2% (w/v) agar, pH 5·7 (KOH). Plants were grown on vertically positioned plates at a temperature of 26 °C/19 °C (day/night) and a 16 h per day light period (50–70 µE m−2 s−1).

For studies of sos3 phenotype rescue by divalent cations, plants were transferred after 5 d to plates containing MS salts without Ca2+ and Mg2+, supplemented with either 0 or 3 mm Mg2+ or Ca2+. Thus, ‘0’ Ca2+ or Mg2+ treatments should be interpreted as ‘no added’ Ca2+ or Mg2+, with residual Ca2+ levels of < 0·1 mm. Monovalent cations were added as KCl, NaCl or LiCl for the various treatments. Five seedlings were transferred onto each plate, with four replicate plates per treatment. On completion of transfer, plants were inverted with respect to their previous growth axis and on the fifth day after transfer, new root growth was measured. Root elongation was used previously as a screen to isolate SOS mutants and was shown to correlate well with the effects of NaCl on Arabidopsis seedling growth (Wu et al. 1996). The mean increase in root length and SE represent the values from 20 seedlings. Statistical analysis was performed using anova and multiple range (least significant difference, LSD) tests with P < 0·05 taken as statistically significant.

For pharmacological assays, wild-type plants were transferred to plates containing 50 mm NaCl and 0·15 or 3 mm CaCl2. The following modulators of Ca2+ signalling were applied: the calmodulin inhibitor trifluoperazine dimaleate (TFP) (Calbiochem, Nottingham, UK); cyclosporin A (CsA) (Calbiochem) and fenvalerate (Calbiochem), which are inhibitors of calcineurin. Inhibitors were added to cooling agarose solutions after autoclaving. Inhibitor stocks were made in ethanol. For each treatment, five seedlings were transferred onto each inverted plate. On the fifth day after transfer, new root growth was measured as length.

22Na+ uptake in seedlings

Plants were grown for 7 d as described above. For Na+ uptake, approximately 50 seedlings were preincubated for 1 h in MS salts containing 50 mm NaCl (MS/50NaCl). Four different conditions were tested: (i) MS/50NaCl (i.e. high Ca2+ and high K+) (ii) MS/50NaCl/minus Ca2+ (i.e. high K+ and no Ca2+ added); (iii) MS/50NaCl/0·1K+ (high Ca2+ and 0·1 mm K+), and (iv) MS/50NaCl/0·1K+/minus Ca2+ (no Ca2+ added and 0·1 mm K+). After preincubation, uptake was performed in the same medium supplemented with 0·1 µCi mL−1 22Na+. Uptake was performed for a period of 5 min and followed by two 10 min washes in ice-cold buffer minus 22Na+. Whole seedlings were blotted dry, weighed and radioactivity was determined by scintillation counting.

22Na+ efflux from seedlings

Approximately 300 seedlings were grown as described above and placed in Petri dishes containing 15 mL loading buffer, composed of Ca2+- and K+-free MS salts supplemented with 50 mm NaCl, 0·15 mm CaCl2 and 22Na+ (0·1 µCi/mL). Seedlings were loaded with 22Na+ for 24 h. After loading, seedlings were briefly washed in ice-cold loading buffer without 22Na+ and at t= 0, 15 mL efflux solution was collected rapidly for scintillation counting and replaced with 15 mL of fresh buffer. This procedure, carried out at room temperature, was repeated at time intervals of 5, 10, 15, 20, 30, 45, 60, 90, 120, 180 and 240 min. After 240 min, whole seedlings were blotted dry, weighed and residual radioactivity was determined by scintillation counting. Untransformed data were fitted with exponential functions to derive time constants.

22Na+ efflux from mature plants

Seeds of wild-type and sos3 A. thaliana were sown on compost and grown for 3 weeks. Plants were then removed from the soil and grown hydroponically for a further 6 d on 25% MS medium. Plants were transferred to vials containing 15 mL loading buffer (25% MS, 50 mm NaCl, 0·15 mm CaCl2 with 0·1 µCi/mL 22Na+) and loading continued for 24 h. After loading, roots were washed twice in ice-cold buffer minus 22Na+ and radioactivity in plant roots and external medium was determined as described above.

Results

A. thaliana sos3 mutants are hypersensitive to NaCl and LiCl stress, and sensitivity to Na+ can be partially reversed in the presence of high (>1.5 mm) external Ca2+ concentrations (Liu & Zhu 1997). The specificity of monovalent cation-induced growth inhibition was determined by measuring root elongation of wild-type and sos3 seedlings on agar plates containing 50 mm KCl, 50 mm NaCl or 10 mm LiCl. Treatments were performed in the presence or absence of CaCl2 and/or MgCl2 to examine the divalent cation specificity of the rescue of sos3 root elongation.

Effect of KCl stress on root length

Figure 1 shows that the elongation of wild-type and sos3 roots was not affected by growth on MS with 50 mm KCl compared with the control. In contrast, a large and significant reduction in wild-type root elongation was recorded on Ca2+- and Mg2+-free MS with 50 mm KCl (similar to that recorded for sos3 seedlings). The reduction in root elongation induced by growth on Ca2+- and Mg2+-free MS with 50  mm KCl could be completely reversed in the presence of 3 mm of either Ca2+ or Mg2+ in both wild-type and sos3 seedlings. Thus, this concentration of external KCl appeared to be toxic only in the absence of Ca2+ and Mg2+ and the rescue of root elongation was neither Ca2+- nor sos3-specific.

Figure 1.

Divalent cation-induced rescue of root elongation in wild-type (filled bars) and sos3 (open bars) seedlings grown in the presence of 50 mm KCl. Five-day-old seedlings were transferred to modified MS media, containing 50 mm KCl, in the presence or absence of 3 mm Ca2+ and/or 3 mm Mg2+. Increase in root length was measured after 5 d growth. Bars = SE; n= 20, 19, 20, 19, 10, 10, 20, 20, 20 and 15 seedlings, respectively.

Ca2+-specific rescue of root elongation in the presence of NaCl

To determine whether the partial rescue of sos3 seedlings by Ca2+ on NaCl (Liu & Zhu 1997) was Ca2+-specific, root elongation was measured on plants exposed to 50 mm NaCl in the presence or absence of Ca2+ and/or Mg2+. Figure 2 shows that the presence of 50 mm NaCl resulted in a small but significant reduction in wild-type root elongation to 84 ± 3% of that on control MS. In contrast, the presence of NaCl had a more dramatic effect on the root elongation of sos3 seedlings: the increase in root length was reduced to 31 ± 6% of the control.

Figure 2.

Ca2+-specific rescue of the sos3 phenotype. Five-day-old wild-type (filled bars) and sos3 (open bars) seedlings were transferred to modified high K+ (20 mm) (a) or low K+ (0·1 mm) (b) MS media, containing 50 mm NaCl, in the presence or absence of 3 mm Ca2+ and/or 3 mm Mg2+. Increase in root length was measured after 5 d growth. Bars = SE; n= 20, 20, 20, 18, 20, 20, 20, 18, 19 and 19 seedlings, respectively.

Wild-type seedlings displayed a greater sensitivity to NaCl when grown in the absence of Ca2+ and Mg2+. Thus, in the presence of 50 mm NaCl, root elongation on Ca2+- and Mg2+-free MS was only 50 ± 2% of the control. Notably, sos3 seedlings failed to grow at all on Ca2+- and Mg2+-free MS with 50 mm NaCl. A rise in external Ca2+ to 3 mm restored the growth of wild-type seedlings treated with NaCl to around 80% of that under control conditions. Similarly, growth of sos3 was partially restored at this level of external Ca2+ concentration, to approximately the level observed in standard MS (which contains 2·3 mm Ca2+). In contrast to Ca2+, Mg2+ was considerably less effective in restoring root growth of NaCl-treated roots.

The sos3 mutation is not Na+-specific: it also affects high-affinity K+ uptake. We therefore tested whether the Na+ phenotype was rescued by external Ca2+ in low K+ (0·1 mm) conditions. At low K+, the observed growth pattern as a function of Na+ and Ca2+ is not qualitatively different from that observed at high K+(Fig. 2). In low-K+ control conditions, overall root growth is around 30% and 55% less than that in high-K+ control conditions for wild-type and sos3 plants, respectively. However, in the presence of NaCl, the addition of 3 mm Ca2+ relieves salinity-stress-induced growth reduction in both wild-type and sos3 plants to approximately the same degree as that observed at high-K+ concentrations.

Ca2+-specific rescue of root elongation in the presence of LiCl

The effect of Li+ on root elongation was investigated for the following reasons: because it is presumed that it is transported via the same pathways as Na+, and because Li+ shares with Na+ the same sites of toxic action (Serrano et al. 1999). Figure 3 shows that the presence of 10 mm LiCl had no significant effect on root elongation in wild-type seedlings. In contrast, the elongation of sos3 roots on MS with 10 mm LiCl was significantly reduced, to 28 ± 2% of the control. LiCl was toxic to wild-type seedlings when grown in the absence of Ca2+ and Mg2+, with root elongation reduced to 10 ± 1% of the control. As with 50 mm NaCl, sos3 seedlings completely failed to grow in 10 mm LiCl in the absence of Ca2+ and Mg2+. Wild-type root elongation on Li+ was restored to the control level by 3 mm Ca2+, whereas this concentration of external Ca2+ restored sos3 growth to around 50% of the control. The latter observation demonstrates that, as with Na+, 3 mm external Ca2+ does not fully rescue the sos3 phenotype. Mg2+, in contrast with Ca2+, failed to restore the growth of either wild-type or sos3 seedlings in the presence of 10 mm LiCl.

Figure 3.

Rescue from lithium toxicity by Ca2+ in wild-type (filled bars) and sos3 (open bars) seedlings. Five-day-old seedlings were transferred to modified MS media containing 10 mm LiCl in the presence or absence of 3 mm Ca2+ and/or 3 mm Mg2+. Increase in root length was measured after 5 d growth. Bars = SE; n= 20, 20, 20, 20, 20, 20, 20, 20, 14 and 25 seedlings, respectively.

Effect of interactions between TFP and NaCl on root growth

Calcineurin is activated by calmodulin. We therefore tested the role of calmodulin in salinity tolerance by studying the interactive effects of the calmodulin inhibitor TFP and 50 mm NaCl on root growth of A. thaliana seedlings. Figure 4 shows that no significant reduction in wild-type root growth occurred in the presence of either 50 mm NaCl or 10 µm TFP. In contrast, a large and significant (P < 0·05) decrease in root growth was observed in the presence of both 50 mm NaCl and 10 µm TFP: approximately 20% of that observed in plants from control plates. These data suggest that calmodulin may play an important role in plant response to NaCl stress.

Figure 4.

Effect of interaction between TFP and NaCl on root growth of wild-type (a) and sos3 (b) A. thaliana seedlings. TFP was applied at 10 µm and NaCl at 50 mm. Root growth was recorded after 5 d growth on plates containing MS medium with 0·15 mm Ca2+ and amounted to 22 and 17 mm for wild-type and sos3 plants, respectively, under control conditions. Bars = SE; n = 20 seedlings per treatment.

In sos3 plants, the addition of TFP on its own had a significant effect on growth, indicating that calmodulin-based signalling is of importance in sos3 plants – even in the absence of NaCl. In the presence of both TFP and NaCl, root growth ceased completely.

Effect of calcineurin inhibitors on seedling root length

The effect of the calcineurin inhibitors cyclosporinA (CsA) and fenvalerate on root growth of A. thaliana seedlings was examined in the presence and absence of 50 mm NaCl on MS plates with 0·15 mm Ca2+. Figure 5 demonstrates that in the absence of NaCl, CsA at a concentration of 5 µm did not affect the root elongation of wild-type plants. However, in the presence of CsA plus 50 mm NaCl, the root growth dropped to 38% of that recorded for the control conditions (P < 0·05). The induced sensitivity of wild-type A. thaliana seedlings to CsA was, however, reversed in the presence of high external concentrations of Ca2+. Root elongation for 50 mm NaCl plus 5 µm CsA plus 3 mm Ca2+ was 1·7-fold higher (P < 0·05) than that recorded in 50 mm NaCl with 5 µm CsA. Similar effects on root growth were observed when 5 µm fenvalerate was applied as a calcineurin inhibitor.

Figure 5.

Effect of interaction between calcineurin inhibitors and NaCl on root growth of wild-type (a) and sos3 (b) A. thaliana seedlings. CsA and fenvalerate were applied at 5 µm, and NaCl at 50 mm. Root growth was recorded after 5 d growth on plates containing MS medium with 0·15 mm Ca2+ and amounted to 22 and 17 mm for wild-type and sos3 plants, respectively, under control conditions. For the ‘CsA + NaCl + Ca’ condition, 3 mm Ca2+ was added. Bars = SE; n = 20 seedlings per treatment.

No significant effect of CsA or fenvalerate was apparent during growth of sos3 plants under control conditions. However, in contrast to the results obtained with wild-type plants, the application of CsA or fenvalerate to sos3 plants did not exacerbate salinity stress. This is in agreement with the notion that SOS3 forms a CsA- and fenvalerate-sensitive target.

Wild-type and sos3 seedlings show similar Na+ influx

The higher Na+ content recorded in sos3 seedlings compared with wild type (Liu & Zhu 1997) could result from a higher influx and/or a lower efflux of Na+ across the plasma membrane. To distinguish between these possibilities, we measured unidirectional Na+ fluxes with radio-isotopes. 22Na+ influx was measured under four different conditions (Table 1). Under most conditions, no difference was observed in unidirectional influx between wild-type and sos3 plants. Only when MS minus Ca2+ was used during preincubation and uptake was there a significant difference between wild-type and sos3 plants. However, for the latter condition, 22Na+ uptake rates in sos3 plants were actually lower than those measured in wild-type plants. Similar results were obtained using mature plants with influx of 4·98 ± 0·44 and 5·23 ± 0·32 µmol g−1 FW h−1 for wild-type and sos3 plants, respectively.

Table 1.  Unidirectional Na+ uptake in A. thaliana wild-type and sos3 seedlings. Preincubation for 1 h and uptake of tracer over 5 min were carried out in MS salts supplemented with 50 mm NaCl at different K+ and Ca2+ concentrations. Data are averages (± SE) of four experiments for each condition, using approximately 50 seedlings per experiment
MediumNa+ uptake (µmol g FW−1 h−1)
Wild-typesos3
MS5·32 (± 0·68)5·08 (± 0·46)
MS minus Ca2+7·77 (± 1·06)6·02 (± 0·40)
MS (0·1 mm K+)5·05 (± 0·41)4·72 (± 0·23)
MS minus Ca2+
(0·1 mm K+)
6·79 (± 0·52)7·21 (± 0·22)

Wild-type seedlings show higher Na+ efflux than sos3 seedlings

Since there was no observable difference in Na+ uptake between wild-type and sos3 seedlings, we tested whether the higher Na+ content of sos3 seedlings results from lower Na+ efflux in the mutant. Seedlings were loaded with 22Na+ for 24 h in a medium containing Ca2+- and K+-free MS plus 50 mm NaCl and 0·15 mm CaCl2. The results in Fig. 6 represent the mean efflux from approximately 300 seedlings. Both wild-type and mutant seedlings displayed biphasic Na+ efflux kinetics but, significantly, rates were higher for wild-type plants. Data were effectively described as double exponential functions with fast time constants of 20 ± 5 min and 24 ± 4 min for wild type and mutant, respectively. The slow phase time constants were 8·6 ± 1·2 and 19 ± 2 h for wild-type and mutant seedlings, respectively. Over the entire 4 h period, Na+ content of the wild-type plants reduced by 28%, whereas the Na+ content of sos3 mutants decreased by 16%.

Figure 6.

Comparison of 22Na+ efflux between sos3 (□) and wild-type (▪) seedlings. Seedlings were loaded with 22Na+ for 24 h, and 22Na+ efflux was measured over a 4 h period. Loading and efflux were performed in a buffer consisting of Ca2+- and K+-free MS plus 50 mm NaCl and 0·15 mm CaCl2. The data represent the total efflux from approximately 300 seedlings. Both wild-type and sos3 data were fitted with a bi-exponential decay of the form Q1*exp (–k1*t) +Q2*exp (–k2*t), where (Q1+Q2) is the relative 22Na+ content at t= 0, k1 and k2 are the rate constants and t is time. (Q1+Q2) was 8·8 and 7·0 µmol g FW−1 for the wild-type and mutant, respectively.

Na+ efflux from mature plant roots

To determine whether reduced Na+ efflux in sos3 plants is maintained throughout development, we measured Na+ efflux from mature wild-type and sos3 plant roots after 24 h loading with 22Na+. Loading and efflux were performed in a buffer composed of K+- and Ca2+-free 25% MS supplemented with 50 mm NaCl and 0·15 mm CaCl2. A low Ca2+ concentration was used to ensure that sos3 plants were expressing the sos3 phenotype in terms of hypersensitivity to NaCl. In both wild-type and sos3 roots, the total amount of accumulated Na+ per g fresh weight (FW) was considerably less than that observed in small seedlings. In agreement with results obtained with seedlings, Na+ efflux from wild-type plants was higher than that from sos3 plants, with a 19% reduction of total Na+ over the entire 4 h period in wild-type plants compared with a 3% reduction for sos3 mutants.

Discussion

Salinity stress is alleviated by Ca2+ but not by Mg2+

In agreement with the findings of Liu & Zhu (1997), sos3 seedlings displayed enhanced sensitivity to NaCl and LiCl compared with wild-type seedlings. This sensitivity was reduced in the presence of high external Ca+ concentrations. The divalent specificity of the rescue of sos3 root elongation in the presence of salinity stress was investigated by substituting 3 mm Ca2+ with 3 mm Mg2+. Inhibition of wild-type root growth on Ca2+- and Mg2+-free MS medium in the presence of 50 mm NaCl was almost completely reversed by 3 mm Ca2+, whereas elongation of sos3 roots was only partially restored (Fig. 2). In contrast, NaCl-induced inhibition of root growth of wild-type and sos3 seedlings was only slightly improved by 3 mm Mg2+. Similarly, 3 mm Mg2+ failed to rescue either wild-type or sos3 root elongation on Ca2+- and Mg2+-free medium in the presence of 10 mm LiCl (Fig. 4). Interestingly, KCl stress symptoms could be suppressed by both Ca2+ and Mg2+. Thus, it appears that while Ca2+ has a specific beneficial effect on the presence of NaCl and LiCl, it has no such effect on KCl. These results point to intrinsically different pathways for uptake and/or signalling with respect to K+ and Na+ toxicity at high concentrations.

Effects of calmodulin and calcineurin antagonists in the absence of NaCl

Neither the calmodulin inhibitor TFP nor the calcineurin inhibitors CsA and fenvalerate had much effect on the root growth of wild-type seedlings under control conditions (Figs 4 and 5). Application of the same antagonists to sos3 plants showed a different picture. In the absence of NaCl, root growth of sos3 was significantly reduced by TFP but not by calcineurin inhibitors. The lack of significant effects on wild-type growth of calmodulin and calcineurin inhibitors in the absence of NaCl is somewhat surprising: calmodulin especially, as it is involved in cellular metabolism, ion balance, maintenance of cytoskeleton and protein modification (reviewed in Snedden & Fromm 1998). However, these processes do not appear to be essential for A. thaliana root elongation in non-stress conditions. Alternatively, incomplete inhibition of calmodulin in combination with redundancy in signalling pathways may ensure that normal growth is maintained in the presence of either calmodulin or calcineurin inhibitors. Indeed, the negative impact of TFP (but not of CsA and fenvalerate) on sos3 growth does suggest that for wild-type growth in the presence of TFP or CsA/fenvalerate, residual SOS3 activity may be sufficient to maintain growth. In sos3 plants, the SOS3-based signalling pathway is largely defunct and growth may be maintained through alternative pathways that are more sensitive to inhibition by TFP.

Effects of calmodulin and calcineurin antagonists in the presence of NaCl

The application of either calcineurin or calmodulin inhibitors in the presence of NaCl resulted in a drastic reduction of wild-type root elongation, whereas there was little effect in sos3 plants (Figs 4 and 5). These results point to a role for Ca2+-binding proteins such as calmodulin and SOS3 in the continued root elongation of wild-type seedlings in the presence of 50 mm NaCl. The fact that 3 mm CaCl2 rescued the growth of wild-type A. thaliana seedlings in the presence of 5 µm CsA and 50 mm NaCl further suggests a role for Ca2+ and Ca2+-binding proteins in salinity tolerance. This might occur through increased activation of residual SOS3 via higher cytoplasmic Ca2+ levels, although blockage of Na+ uptake is also a possibility.

In yeast, calcineurin is required for salinity tolerance where it ensures an adequate transcription of ENA1 via interaction with the transcription factors TCN1/CRZ1 (Mendoza et al. 1994; Stathopoulos & Cyert 1997). ENA1 encodes a plasma membrane ATPase induced by Li+, Na+ or alkaline pH and functions in Na+ extrusion from the cell. A similar pathway, mediated by the SOS2 kinase, might exist in higher plants. However, since there is no evidence to support the existence of a Na+ ATPase in higher plants, the end target of such a signalling pathway might be upregulation of Na+/H+ antiport activity. Na+/H+ antiport activity has been described for both vacuolar and plasma membranes of plant cells in several species (for a review see Barkla & Pantoja 1996) and might be modulated either transcriptionally or post-transcriptionally by SOS2. Furthermore, a vacuolar Na+/H+ antiporter that confers enhanced tolerance to salt after overexpression has recently been identified at a molecular level (Apse et al. 1999).

Wild-type plants show higher efflux rates than sos3 plants

The effect of the sos3 mutation on Na+ transport was investigated by comparing 22Na+ fluxes between sos3 and wild-type A. thaliana. A higher tissue Na+ content in sos3 mutants than in wild-type A. thaliana could result from either an enhanced influx or a decreased efflux of Na+ at the plasma membrane. Rescue of the sos3 phenotype by elevated external Ca2+ concentrations might then be hypothesized to modulate either Na+ influx or efflux. Accordingly, both 22Na+ uptake and efflux were examined.

In contrast to Na+ uptake across the plasma membrane, which is thought to occur passively via ion channels (Amtmann & Sanders 1999; Tyerman & Skerrett 1999), Na+ efflux is an energized process in higher plants – at least in saline conditions. We did not observe any difference in Na+ uptake between wild-type and sos3 seedlings except in the ‘high K+/low Ca2+’ condition (Table 1), where the sos3 Na+ influx rate was actually lower than that of the wild type. However, Na+ efflux was found to be higher in wild-type plants than in sos3 plants. The decreased unidirectional Na+ efflux in sos3 mutants was reflected in a larger time constant for washout kinetics. From these data alone, it is not possible to establish whether Na+ transport across the vacuolar or plasma membrane is affected in the mutant, since in a multicompartment system such as an intact root, the experimental time constant will subsume contributions from fluxes that occur in each intracellular compartment (Walker & Pitman 1976). However, Na+ efflux during the slow phase depicted in Fig. 6 is likely to be dominated by vacuolar Na+ release, and thus SOS3 may directly affect Na+ transport activity at the tonoplast. For example, it is possible that SOS3 functions as a moderator of tonoplast H+/Na+ antiport activity (Apse et al. 1999). In sos3 mutants, over-stimulation of tonoplast antiport activity would promote Na+ retention in the vacuole, perhaps to levels that damage cell homeostasis.

SOS3 may also affect Na+/H+ antiport activity at the plasma membrane. Even with transport across the tonoplast dominating during the slow phase, such modulation would still affect the overall efflux. Interestingly, the SOS1 gene has recently been shown to encode a protein with high similarity to the Na+/H+ antiport, located in the plasma membrane, from fungi and bacteria (Shi et al. 2000). SOS1 expression is mainly located in tissues surrounding the xylem but also in epidermal cells (Dr Zhu, University of Arizona, personal communication) and hence SOS1 may participate in Na+ extrusion. SOS1 antiport activity is upregulated via SOS3 (Shi et al. 2000), thus Na+ extrusion via SOS1 would be diminished in sos3 mutants. It is noteworthy that for yeast undergoing salt stress, Na+ efflux, which requires energization, is similarly enhanced via a Ca2+- and calcineurin-mediated signalling pathway (Mendoza et al. 1994).

Conclusions

The SOS3 gene encodes a Ca2+-binding protein that is required to maintain Na+ and K+ homeostasis (Liu & Zhu 1998) in A. thaliana seedlings. We have shown that the growth of sos3 mutants in the presence of 50 mm NaCl can be specifically enhanced by high external concentrations of Ca2+ but not Mg2+. This correlation between Ca2+ and salinity tolerance is suggestive of a role of Ca2+-binding proteins in salinity tolerance, a notion that is substantiated by our observation that both calmodulin and calcineurin inhibitors increase salt sensitivity in wild-type plants but are otherwise without effect on growth. The higher levels of Na+ accumulation in sos3 mutants can be ascribed to a diminished capacity for Na+ extrusion in the mutant. A model can be proposed (Sanders 2000) whereby an increase in external NaCl results directly or indirectly in a rise in cytosolic Ca2+. Ca2+ binds with, and activates, calmodulin, which subsequently binds with SOS3, resulting in the activation of SOS2. The SOS2/SOS3 complex, either pre- or post-transcriptionally, upregulates Na+ export at the tonoplast and/or at the soil–symplast boundary.

Acknowledgments

We thank Dr Zhu (University of Arizona) for kindly donating sos3 seeds, and the BBSRC for funding a studentship to C.H.E.

Received 16 November 2000;received inrevised form 7 March 2001;accepted for publication 13 March 2001

Ancillary