Single-Cell Measurements of the Contributions of Cytosolic Na⁺ and K⁺ to Salt Tolerance

Cortical Cell Electrode Measurements of pH, Membrane Potential Difference (E_m), Na⁺ Activity (a_Na), and K⁺ Activity (a_K)

The Na⁺ and K⁺measurements were separated into two populations by assigning each value to either the cytosol (Fig. 1) or the vacuole (Fig. 2) using measured pH values (Walker et al., 1995). Two normally distributed populations having means of pH 5.6 and 7.4 (see TableI) described the combined pH measurements obtained using both Na⁺- and K⁺-selective microelectrodes. Figures 1 and 2show the mean intracellular pH, E_m, a_Na, and a_K measurements obtained using these microelectrodes in barley root cortical cells of plants growing for either 5 or 8 d in NaCl. After 5 d of salinization (1 d at 200 mm NaCl), the mean cytosolic a_Na in Gerbel was only 1.7 mm, whereas Triumph had a markedly higher value of 19.5 mm, but these varietal differences had gone by 8 d (3 d at 200 mm NaCl; Fig. 1). At 5 d, cytosolic a_K was very similar in Triumph and Gerbel, with both varieties having values of about 60 mm, but by 8 d, only in Triumph, the NaCl-sensitive variety, was there a significant decrease to 39 mm (Fig. 1). The cytosolic pH was not significantly different between varieties on either day.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Histograms showing the results of triple-barreled Na⁺ and K⁺ measurements in the cytosol of Triumph (white) and Gerbel (shaded) cortical root cells of seedlings treated with NaCl for 5 and 8 d. The values are means ± se of E_m, a_Na, a_K, and pH and each value was obtained from between five and 28 samples.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Histograms showing the results of triple-barreled Na⁺ and K⁺ measurements in the vacuoles of Triumph (white) and Gerbel (shaded) cortical root cells of seedlings treated with NaCl for 5 and 8 d. From the top to the bottom, Mean ± se values of E_m, a_Na, a_K, and pH measured in the cytosol. Each value was obtained from between four and 28 samples.

Figure 3 shows a summary of the triple-barreled microelectrode data for the cortical cell compartments. At 5 d, both varieties had higher a_Na in the vacuole than in the cytosol, and this difference was greater in the tolerant Gerbel than in Triumph (Fig. 3). By 8 d, these varietal differences had gone and in the tolerant Gerbel the vacuolar a_Na had decreased when compared with 5 d (Fig. 2), but these changes were not significant (Fig. 3; ANOVAP = 0.12). The measurements of vacuolar a_K showed similar values for both varieties around 18 mm at 5 d, although in Gerbel vacuolar K⁺ accumulation had increased by 8 d; this change was not statistically significant (Fig. 3; ANOVAP = 0.21). For both varieties, a_Kwas larger in the cytosolic than in vacuolar compartment (Fig.3).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Diagram showing a summary of the mean intracellular a_K (mm), a_Na (mm), pH, and E_m (mV) in root cortical cells of Triumph (top) and Gerbel (bottom) during the transition from 5 to 8 d treated with NaCl. For each value when no change could be measured, a single value is shown. For comparison, the a_K, a_Na, and pH values of the nutrient solution bathing the roots are also shown. The significance of changes was tested by ANOVA. An asterisk indicates a probability of less than 5%.

After 5 d growing in NaCl, the E_m values reported from both intracellular compartments differed between the varieties with Triumph having more negative values (comparing Figs. 1and 2). By 8 d, both varieties showed significantly more negative vacuolar E_m values than at 5 d (Fig. 3), but at 8 d the only varietal differences were in cytosolic E_m, which was more negative for Triumph (−94 ± 6 mV) compared with Gerbel (−72 ± 5 mV). After 8 d of salinization, the trans-tonoplast potential was −9 mV for Triumph and +9 mV for Gerbel (using the convention of Bertl et al., 1992). The microelectrode measurements of a_K, a_Na, and E_m (Figs. 1 and 2) were used to investigate the energetic feasibility of likely transport mechanisms for Na⁺ and K⁺ across the plasma membrane and tonoplast.

Thermodynamics of Na⁺ Transport across the Plasma Membrane and Tonoplast

The Na⁺ electrochemical potential differences (Δμ_Na) across both the plasma membrane and the tonoplast were calculated to determine whether active or passive transport was required to maintain these gradients (TableII). Extracellular a_Na was measured by the Na⁺microelectrodes at 150 mm, both in the nutrient solution containing 200 mm NaCl and in the apoplast between epidermal and cortical cells (data not shown). This value of a_Na is similar to the calculated value of 142 mm for the nutrient solution obtained using an activity coefficient of 0.71, determined using the Debeye-Hückel equation (Robinson and Stokes, 1970). The values in Table II show that for the cytosol, regardless of variety or time in NaCl, there was a large inwardly directed driving force for Na⁺across the plasma membrane of between −113 and −182 mV. The vacuolar accumulation of Na⁺ required active transport at the tonoplast, for Gerbel on 5 d, but this requirement had gone by 8 d. These results indicate that energy is needed to maintain this ion gradient, with efflux mechanisms removing Na⁺from the cytosol into either the extracellular solution or the vacuole, or both. The feasibility of several active Na⁺ transport mechanisms was assessed by calculating the associated free energy change (ΔG′/F) for each mechanism (Table III). In most of the conditions for the intracellular electrode measurements, an Na⁺/H⁺ antiport operating at the plasma membrane, transporting Na⁺ out of the cell, would be energetically feasible for maintaining the measured ion gradients. An exception is Gerbel on 5 d; for these conditions, the ΔG′/F at 5 d is +27 mV, showing that operating alone an antiport mechanism is not feasible for the measured ion gradients. The energy provided by the hydrolysis of ATP for Na⁺ efflux via either an Na⁺ or Na⁺-K⁺ ATPase was also assessed, despite a lack of evidence for such a mechanism in higher plant cells. This calculation reveals the efflux of Na⁺ across the plasma membrane via an ATPase to be energetically feasible (Table III).

In most cases for both varieties, a tonoplast H⁺/Na⁺ antiport was energetically feasible to explain the measured ionic gradients; the one exception was for Gerbel on d 5 (Table III).

Thermodynamics of K⁺ Transport across the Plasma Membrane and Tonoplast

Using the data for intracellular compartmentation of K⁺ (Figs. 1 and 2), the feasibility of K⁺ cotransport mechanisms was also assessed. Under these experimental conditions, K⁺ uptake into the cytoplasm across the plasma membrane cannot occur by a channel mechanism. The positive electrochemical potential values for K⁺, with Δμ_K/F between +68 and +106 mV, show that active transport was required for K⁺ uptake from the external solution into the cytosol across the plasma membrane (Table II). Transport of K⁺ across the tonoplast into the cytosol would also require energy input with the Δμ_K/F of +17 to +22 mV (Table II). Testing the thermodynamics of K⁺ uptake at the plasma membrane by both Na⁺- and H⁺-coupled symport mechanisms showed that both are feasible for maintaining the measured cytosolic a_K in both varieties at both 5 and 8 d (data not shown).

Varietal Differences in Cellular Cation Activities

The microelectrode measurements of a_Na and a_K in the root cortex show that there are large cellular differences between the varieties after just 5 d of treatment with NaCl (summarized in Fig. 3). After 5 d, the tolerant variety Gerbel seemed to be more effective at cytosolic Na⁺ exclusion and vacuolar sequestration than the more sensitive variety Triumph (Fig. 3), despite the large cytosol-directed electrochemical potentials for Na⁺ (Table II). However, statistical comparison of the mean values of a_Na for 5 d for the two varieties were not significantly different (ANOVA P= 0.08). After 8 d, Gerbel was also better at maintaining cytosolic a_K in the high background of 200 mm NaCl (ANOVA P < 0.05). Expressing the cytosolic K⁺:Na⁺ ratios, 34.7 for Gerbel and 3.2 for Triumph, illustrates clearly the large differences in the response of the two varieties after 5 d in NaCl. However, after 8 d of growing in 200 mm NaCl, the cytosolic K⁺:Na⁺ ratios of the two varieties were very similar (Gerbel, 2.1; and Triumph, 2.9). These changes in cytosolic cation activities agree with the whole-plant responses of the two varieties showing that NaCl tolerance under these conditions was a matter of differing survival times.

K Deficiency, Compartmental pH, and NaCl Stress

Undoubtedly, the application of NaCl to plants alters the intracellular pools of K⁺, but is this cellular response similar to that brought about by a lack of K⁺ in the nutrient solution? Subcellular compartmentation of K⁺ in barley is known to change in response to changes in external K supply (Memon et al., 1985). Under K⁺-replete conditions, there is cytosolic a_K homeostasis at around 70 mm, whereas vacuolar a_K changes in response to changes in external K⁺ (Walker et al., 1996). Vacuolar activities decreased from 100 mm in replete plants to less than 25 mm for deficient plants, when the external K⁺ supply was decreased (Walker et al., 1996, 1998). During extreme K⁺ starvation, barley root epidermal cells showed a decrease in cytosolic a_K to 40 mm accompanied by an acidification of the cytosol to pH 6.7 (Walker et al., 1996). In the present study, treatment of plants with 200 mm NaCl, combined with low external K⁺ of 0.1 mm, produced cortical cell cytosolic a_K as low as 39 mm in the variety Triumph by 8 d (Figs. 1 and 3). At this cytosolic a_K, decreases in the rates of K-dependent biochemical processes and, hence, growth would result (Leigh and Wyn Jones, 1984; Walker et al., 1998). Cytoplasmic K⁺ concentrations (rather than activities) of 60 to 80 mm, for plants growing at the same sodium concentration but 60 times the potassium concentration for 15 d and calculated from x-ray microanalyses (Flowers and Hajibagheri, 2001), were remarkably similar to our microelectrode measurements.

The pH electrode measurements of both K⁺ and Na⁺ triple-barreled electrodes separated into two distinct populations, and the more alkaline population of measurements was assumed to be from the cytosol. These values for the cytosolic pH are very similar to those measured previously for unstressed barley roots using microelectrodes (7.3 ± 0.1; Walker et al., 1996, 1998). In the salinized cells described here, there was no evidence for an associated acidification with depletion of cytosolic a_K (Figs. 1 and 3). This result suggests that within cells, the depletion of cytosolic a_K resulting from NaCl treatment is physiologically different from that associated with K starvation, although another possible explanation is that epidermal and cortical cells respond differently to depletion of cytosolic a_K. In support of this idea, we know that the expression pattern of a high-K⁺/low-Na⁺ affinity transporter (HKT1) in wheat is stronger in root cortical cells (Schachtman and Schroeder, 1994) and during K⁺ withdrawal (Wang et al., 1998).

The pH values for both the cytosol and the vacuole broadly agree with values published for barley roots obtained using³¹P-NMR (Martinez and Läuchli, 1993; Katsuhara et al., 1997). These NMR measurements for unsalinized barley roots gave mean pH values of 7.5 to 7.7 and 5.5 to 5.7 in the cytosol and vacuole, respectively. In both reports, the addition of NaCl resulted in an alkalization of the vacuole (to over pH 6), but almost no change in the pH of the cytosol (Martinez and Läuchli, 1993; Katsuhara et al., 1997). Although these values obtained using NMR are similar to the results reported here, where the mean cytosolic pH was 7.5 for both varieties, there was no significant alkalization of the vacuolar pH from 5 to 8 d (Fig. 2). In wheat, NaCl stimulation of proton pumping at the plasma membrane has been measured (Ayala et al., 1997), and this is consistent with the measured more negative membrane potential in Triumph compared with Gerbel at 8 d (Figs. 1 and 3), possibly indicating a greater level of NaCl stress in this variety.

Transport Mechanisms That Are Important in NaCl Tolerance

Na⁺ might enter cells through nonselective cation channels (Schachtman et al., 1991;Davenport and Tester, 2000; Maathuis and Sanders, 2001), and HAK1- (Santa-Marı́a et al., 1997) and HKT1-type transporters (Rubio et al., 1995; Rus et al., 2001). The ability of the plant cell to prevent, at least in the short term, a net rise in cytosolic Na⁺ is demonstrated here by the recorded values of between 2 and 33 mm, for 5 to 8 d in the high background of 200 mm Na⁺in the external solution (a_Na = 142 mm). After 15 d at the same salinity, barley root cell cytoplasmic Na⁺ concentrations (rather than activities) were calculated from x-ray microanalyses to be 180 to 245 mm (Flowers and Hajibagheri, 2001). These plants had been grown at 60 times the K⁺concentration used in the experiments described here and although the cytoplasmic concentration of Na⁺ was estimated to be 1.4 times higher for Triumph than Gerbel, the difference was not statistically significant: There were no significant differences in K⁺ concentrations between varieties. Several membrane transport mechanisms have been proposed to explain K⁺/Na⁺ selectivity in higher plants. These include exclusion at the uptake stage, when there can be high selectivity against Na⁺ entry though channels (e.g. Amtmann and Sanders, 1999; Tyerman and Skerrett, 1999), and efflux of Na⁺back into the external solution or the vacuole. The electrochemical gradients measured here for Na⁺ indicate that there was a large cytosol-directed gradient (Table II), and active transport would be required to remove cytosolic Na⁺. Although a Na⁺-pumping ATPase is implicated in the NaCl tolerance of marine algae (Shono et al., 2001), and although energetically feasible (see Table III), a similar direct role for a primary pump removing Na⁺ from the cytosol in higher plants has yet to be shown. The role and molecular identities of H⁺/Na⁺ exchangers, at both the plasma membrane (Qiu et al., 2002) and tonoplast (Apse et al., 1999; Zhang and Blumwald, 2001), in the NaCl tolerance of higher plants is now well established. For most of the conditions tested in this paper, a plasma membrane H⁺/Na⁺ antiport with a stoichiometry of 1:1 could operate to achieve the measured ion gradients; only for Gerbel on d 5 was this mechanism not energetically feasible (Table III). This result for Gerbel at 5 d may indicate that the antiport has a variable H⁺/Na⁺ stoichiometry ratio enabling the maintenance of such a low cytosolic a_Na. The activity of a plasma membrane H⁺/Na⁺ antiport should lead to an acidification of the cytosol between 5 and 8 d, but no such change was measured (Figs. 1 and 3), a result that agrees with measurements obtained using ³¹P-NMR (Martinez and Läuchli, 1993; Katsuhara et al., 1997). An increased activity of the plasma membrane H⁺ pump may function to regulate cytosolic pH, compensating for any acidification caused by an antiporter removing Na⁺ from the cytosol. Our microelectrode measurements show that there was a hyperpolarization of the plasma membrane potential in Triumph root cells between 5 and 8 d, becoming more negative by between 10 and 20 mV (Figs. 1 and 3). However, this change in membrane potential was not statistically significant (ANOVA P = 0.2), but a similar change in the membrane potential was also measured in root cells of Triumph treated directly with 200 mm NaCl (Carden et al., 2001).

At the tonoplast, there was also a cytosol-directed Δμ_Na/F indicating that an active transport process was operating on the tonoplast transporting Na⁺ into the vacuole (Table II). Na⁺/H⁺ antiports have been found on the tonoplast of barley, with energy provided by the proton pumps (Garbarino and DuPont, 1989; Martinez and Läuchli, 1993). Moreover, the tonoplast ATPase of barley is stimulated by the presence of NaCl (Garbarino and DuPont, 1988; Matsumo and Chung, 1988). The operation of a H⁺/Na⁺antiport might be expected to result in an alkalization of the vacuole, but this was not observed between 5 and 8 d (Fig. 2). NMR measurements showed barley vacuolar alkalization upon exposure to NaCl and these measurements used excised root tips measured for up to 6 h after NaCl treatment (Martinez and Läuchli, 1993; Katsuhara et al., 1997). The microelectrode measurements do show significant vacuole alkalization when the values in Figures 2 and 3 are compared with values for root cells not treated with NaCl (pH 5.1–5.3; Walker et al., 1995, 1996). Therefore, there is evidence to support a tonoplast-located Na⁺/H⁺ antiport as the main mechanism for vacuolar sequestration of Na⁺. The microelectrode measurements show that in the longer term response to NaCl stress, the increased activity of tonoplast H⁺ pumps can maintain vacuolar pH. This result agrees with the observation that transgenic plants overexpressing a vacuolar H⁺/Na⁺ antiport have improved NaCl tolerance, accumulating NaCl in the vacuole but also able to maintain normal pH regulation within intracellular compartments (Apse et al., 1999; Zhang and Blumwald, 2001). In addition, the removal of protons from the cytosol by the tonoplast pumps would offset an acidification of the cytosol by the actions of a plasma membrane Na⁺/H⁺ antiport. This may also explain the lack of cytosolic acidification observed between 5 and 8 d as the cytosolic a_K is depleted and that occurs during K⁺ deficiency (Walker et al., 1996, 1998).

Clearly salinity tolerance depends on sustaining the cytosolic environment, limiting Na⁺ accumulation and maintaining K⁺ concentration. Microelectrode measurements of root cell responses to salinity suggest that for the first 5 d, the tolerant variety Gerbel is better able to exclude Na⁺ and then by 8 d better at maintaining K⁺ when compared with Triumph (see Fig. 3). In root cells, these varietal differences only seem to occur early on in the salinity response; by 15 d, no significant differences were measured (Flowers and Hajibagheri, 2001). The differences between Gerbel and Triumph at 5 d cannot be explained by the activity of a plasma membrane H⁺/Na⁺ antiport because this mechanism is not energetically feasible to explain the measured electrochemical gradients in the tolerant plant. Possible differences in the Na⁺ exclusion mechanisms for uptake may be important for explaining these early differences between the two varieties. Both this result and the differences after 8 d between Gerbel and Triumph in the ability to maintain cytosolic K⁺ could be explained by subtle changes in the properties of either a root HKT1-type transporter (Rubio et al., 1995; Rus et al., 2001) or HAK1-type transporter (Santa-Marı́a et al., 1997).

The microelectrode data have provided direct intracellular measurements of three cations during the early onset of NaCl stress well before any visual stress symptoms have appeared in the plants. This nondestructive technique can be used for the analysis of transport mechanisms in single cells of mutant plants with altered expression of transporters; for example, Arabidopsis mutants with altered expression and activity of H⁺/Na⁺ exchangers (Qiu et al., 2002). The subsequent data analysis can show the feasibility of transport mechanisms at both the plasma membrane and tonoplast in single cells.