Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
The high affinity potassium transporter, HKT1 from wheat was introduced into Florida wheat in sense and antisense orientation under control of a ubiquitin promoter. Ten transgenic lines expressing the transgene were identified and two of these showed strong down-regulation of the native HKT1 transcript. One line (271) was expressing the antisense construct and the other (223) was expressing a truncated sense construct. The two lines were examined further for phenotype relating to cation transport. Membrane depolarisations were measured in low (0.1 mm) K+ and high (100 mm) NaCl. Under these conditions there was no difference between line 271 and the control at low K+, but at high Na+ there was a rapid depolarisation that was significantly larger in control plants. 22Na uptake was measured in this line and there was a significant decrease in uptake at 100 mm NaCl in the transgenic line when compared with the control. The two transgenic lines were grown at high NaCl (200 mm) and analysed for growth and root sodium content. Lines 271 and 223 showed enhanced growth under salinity when compared with the control and had lower sodium in the root. Secondary ion mass spectrometry (SIMS) analysis of transverse sections of the root showed that Na+ and K+ were strongly localised to stelar regions when compared with other ions, and that the Na+ : K+ ratios were reduced in salt-stressed transgenic tissue when compared with the control.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Potassium (K+) is a major macronutrient in plant cells and is essential for many plant processes, including balancing the large excess of negative charge on proteins, providing the correct ionic environment for protein synthesis, and activation of K+-dependent enzymes in the cytosol (Maathuis and Sanders, 1996). In addition, K+ is the predominant cationic contributor to cell turgor and therefore involved in cell growth, regulation of stomatal aperture and in maintaining osmotic homeostasis in the cell. In plants grown with no limitations on K+ supply, this ion can contribute up to 10% of the DW of tissues (Leigh and Wyn Jones, 1984).
The HKT1 gene was isolated from wheat roots and initial heterologous expression studies in yeast suggested that it mediates H+-coupled high affinity K+ uptake (Schachtman and Schroeder, 1994). However, subsequent studies showed that HKT1-mediated K+ transport was energised through coupling to Na+ rather than H+ (Gassmann et al., 1996; Rubio et al., 1995). This finding suggested that (part of) high affinity K+ uptake in plants is driven by an electrochemical Na+ gradient rather than a H+ gradient. Further support for a role of HKT1 in K+ uptake came from observations that, in both barley and wheat, HKT1 transcription was induced by K+ starvation (Wang et al., 1998), although the induction is transient and appears to be specific to young roots. Previous studies also showed the occurrence of Na+-coupled K+ uptake in various aquatic plant species (Smith and Walker, 1989). However, during a study using barley, wheat and Arabidopsis, no evidence was found for a significant contribution of Na+-coupled K+ transport to K+ uptake in terrestrial plants (Maathuis et al., 1996).
Functional analysis of HKT1 in heterologous expression systems showed that at high Na+ : K+ ratios, K+ transport ceases and both binding sites are involved in low affinity Na+ transport (Rubio et al., 1995). A highly charged loop region in the protein is thought to be involved in Na+ uptake (Diatloff et al., 1998; Rubio et al., 1999). Deletion of this loop gives the transporter a greater selectivity for K+ over Na+ and increases salt tolerance of yeast cells expressing the modified HKT1 (Liu et al., 2000). A potential role of HKT1-like transporters in Na+ rather than K+ transport was further substantiated by work on the Arabidopsis HKT1 orthologue expressed in oocytes (Uozumi et al., 2000), which showed that AtHKT1 selectively transported Na+ but not K+. In addition, hkt1 insertional mutants of Arabidopsis, which lack functional HKT1, show decreased Na+ uptake when grown under saline conditions, supporting a role for HKT1 in Na+ influx in planta (Rus et al., 2001). In addition to the wheat and Arabidopsis HKT1 genes, there are homologues in Eucalyptus (Fairbairn et al., 2000), database entries for rice homologues (Horie et al., 2001; Su et al., 2001), and we, as well as others, have sequenced homologous partial clones from barley. Recent characterisation in oocytes of two Eucalyptus homologues showed Na+-enhanced K+ currents but also transport of K+ in the absence of Na+ (Liu et al., 2000).
Thus, it appears that the HKT superfamily contains members which may display a variety of functional and mechanistic properties and, in particular, their Na+ and K+ selectivities may vary. This may reflect different physiological roles for family members in different plants. Characterisation of ‘loss of function’ mutants (Rus et al., 2001) and antisense plants with decreased expression of HKT1 can potentially reveal invaluable information about the function of HKT1 in planta. In this paper we describe the production of transgenic wheat plants in which native HKT1 expression is significantly reduced through the introduction of antisense and sense transgenes. Phenotypic characterisation of these lines shows that these plants are compromised in their Na+ uptake and exhibit enhanced salt tolerance indicating that, despite the greater K+ selectivity of wheat HKT1, it mediates Na+ uptake, as in Arabidopsis.
Generation of transgenic wheat lines
Using biolistic bombardment of immature scutellae, we produced 10 transgenic wheat lines containing the HKT1 gene in either sense or antisense orientation. A segregation analysis, carried out to study gene integration and inheritance, showed that inheritance patterns in the T1 generation were variable with only two of the lines showing 3 : 1 inheritance (positive to null) of the introduced HKT1. Some lines failed to inherit the transgene in the T1 generation. Lines homozygous for the introduced HKT1 were produced by crossing or by producing dihaploid lines by anther culture (Massiah et al., 2001).
Southern analysis of native, non-transformed tissue showed a background pattern of five bands hybridising to the dioxigenin probe (Figure 1, control). All co-transformed lines were shown to contain the five native bands and full transcription cassette (Figure 1) and additional bands ranging from two to six in number were recorded. Generally less than four additional bands were seen in each transgenic plant. There was no obvious relationship between numbers of copies, orientation of transgene in the construct, or inheritance patterns (not shown).
RT-PCR of all the lines was carried out, using the native and polylinker primers for transcription of the introduced HKT1 gene, and two native primers for native gene transcription. The native gene could be distinguished from the transgene by primers against part of a 400-bp sequence present in the 5′ region of the native gene but deleted from the transgene. In all root material tested (eight lines), there was evidence of transcription of the introduced construct (data not shown). However, the results were less clear when looking for evidence of down-regulation of native HKT1. Initial experiments were designed following the previous work of Wang et al. (1998), and material was K+ starved to induce HKT1 transcription. Homozygous transgenic T2 or anther culture derived material was tested for changes in native HKT1 transcription. As controls, plants transformed with the marker gene constructs alone were compared with co-transformed material. Under K+ starvation conditions with 4–6-d-old roots, transcription of the native gene was strong but there was no evidence of reduced native HKT1 transcript in the plants expressing the antisense or sense orientation constructs. A more prolonged K+ starvation did not alter this observation. Wheat HKT1 was previously shown to be capable of Na+ transport, either on its own or in symport with K+ (Rubio et al., 1995). We therefore included 100 mm NaCl in the growth medium to test whether these conditions would alter HKT1 transcription. Exposure to NaCl during the initial 14-d growth period followed by K+-starvation in the absence of NaCl induced a strong reduction of native HKT1 transcription. In line 271, expressing the antisense construct, the native gene transcript could not be detected after 24 h of the K+ starvation treatment (Figure 2). There was also a down-regulation of native gene transcription in line 223, expressing the sense construct, although in this case it took longer and was not apparent until after 96 h. On the basis of these results we looked in closer detail at the two transgenic lines (271 and 223) showing a clear reduction in native HKT1 transcript. For these experiments, down-regulation of HKT1 was induced by growing plants in FNSN followed by FNS-K for 24 or 96 h (depending on line being used; Figure 2).
Transgenic wheat with down regulation of HKT1 shows enhanced NaCl tolerance
Lines 271 and 223 were tested for K+ and Na+-dependent growth phenotypes. No difference was found between these lines and control lines when plants were grown in K+ deficient conditions (results not shown). However, when down regulation of HKT1 was induced in line 271 and plants were then exposed to 200 mm NaCl in the continued absence of K, significantly better growth was observed compared with the control line with relative growth rates of 2.3 ± 0.3% and 1.6 ± 0.3% d−1 for line 271 and the control, respectively. RT-PCR was again used to demonstrate that down-regulation of the native gene was occurring, and lower levels of gene activity were still observed in the trangenic plants than in the control under these growth conditions (data not presented). When exposed to NaCl, FW increase was maintained at least at the same level for plants from line 271, whether they were grown with or without 200 mm NaCl and there was a slight reduction in FW increase in plants from line 223 exposed to salinity. High NaCl had a considerable affect on growth of plants from the control line – salt-stressed plants showed a reduction of 50% in FW increase when compared with plants from the same line returned to non-stress conditions (Figure 3). A similar reduction was seen in line 293, which was transformed but showed no down-regulation of the native gene.
To investigate the basis of the improved salt tolerance, the effect of Na+ on membrane potential of root cortical cells was investigated in plants of line 271 and control plants. After growth under conditions that down-regulated HKT1 expression (Figure 2c), the membrane potential was measured in a simple buffer solution and then either 100 mm NaCl or 0.1 mm KCl was added to the solution. Addition of NaCl resulted in a rapid depolarisation of the membrane but this depolarisation was significantly smaller in line 271 compared with control plants (Table 1). In contrast, the depolarisation induced by addition of 0.1 mm K+ was comparable in both lines. These experiments suggest that reduced expression of HKT1 in line 271 decreases the membrane Na+ conductance in cortical root cells whereas the K+ conductance appears unaffected.
Table 1. Effects of K+ and Na+ on membrane potentials in root cortical cells of transgenic line 271 and a control line. Plants were grown under conditions that down-regulated HKT1 (see text) and the effects of 100 mm NaCl or 0.1 mm KCl on the membrane potential was measured. The resting potential (Vrest) was measured for roots perfused with 1 mm CaCl2, 2 mm Mes/Tris pH 6.0. Subsequently roots were exposed to the same perfusion solution supplemented with NaCl or KCl to determine membrane depolarisations (ρV). Data are expressed as the mean ± SE in mV for 6–8 independent impalements
100 mm NaCl
138 ± 8
42 ± 3.0
144 ± 12
30 ± 3.4
0.1 mm KCl
138 ± 8
17 ± 1.5
144 ± 12
20 ± 5.6
Membrane depolarisations can, at best, provide a qualitative comparison of membrane conductance toward Na+. We therefore measured short-term unidirectional Na+ influxes in roots from line 271 and control line. After growth under conditions to induce HKT1 down-regulation, there was no significant difference in Na+ uptake between the transgenic and control lines at moderate (20 and 50 mm) NaCl concentrations but there was a much smaller Na+ uptake in the transgenic line at 100 mm NaCl (Table 2). Reduced unidirectional Na+ influx was reflected in tissue Na+ accumulation. Roots from transgenic lines 223 and 271 contained less Na+ than did control lines (Figure 4). Statistical analysis of the differences between the samples was carried out. The difference between the control line and line 223 was small (t = 0.96, P= 0.4), with a more significant difference observed between control and 271 (t = 2.14, P= 0.09). Comparison of the two lines showing no down-regulation of HKT1 (control and 293) with the down-regulating lines (223 and 271) gave more significant differences still (t = 3.17, P = 0.01). We were interested to see whether the reduced uptake of sodium resulted in reduced translocation of ions to the shoot. After 5 days growth in 200 mm NaCl the shoots were excised and root exudates were collected using a pressure bomb. It was apparent that the water potentials of control plants were much higher than the transgenic plants, and, while it was possible to collect approximately 20 µl sap from transgenic plants over a 5-min period, only a few µl could be collected from the controls. The sodium content of the root exudate from the transgenic plants showed almost a four-fold decrease when compared with the control (Table 3). Overall, the reduced Na+ uptake and accumulation in lines 271 and 223 is likely to contribute to the improved salt tolerance of these transgenic lines.
Table 2. Uptake of 22Na sodium into roots of transgenic line 271 and a control line. Plants were grown under conditions that down-regulated HKT1 in the transgenic line (14 d in FNSN followed by 24 h in FNS-K) and then 5 cm root segments from 5 to 7 plants were pre-incubated for 30 min at room temperature in the Na+ uptake solution (1 mm CaCl2, 2 mm Mes/Tris pH 6.0, and 20, 50 or 100 mm NaCl). Uptake of 22Na was initiated by adding 0.2 µCi ml-1 of 22Na and was allowed to proceed for 30 min at room temperature. After the uptake period, two 8 min washes were carried out in ice cold solution of the same composition but without radiotracer. Results are the mean ± SE of 3–4 independent experiments for each treatment
Na uptake (µmol/g FW/h)
4.1 ± 0.9
4.9 ± 1.1
9.3 ± 2.0
8.2 ± 0.7
17.0 ± 2.2
9.8 ± 2.4
Table 3. Sodium content of root exudate. Seedlings were grown for 14 d in FNSN and then transferred to high stress conditions (FNS – K + 200 mm NaCl). After 5 d growth at high salt, shoots were excised and root exudate was collected for 5 min using a pressure bomb. Measurements are means of three samples
Na content of root sap (mM) Mean ± SE
14.6 ± 3.0
20.8 ± 4.52
67.86 ± 7.38
One of the major strategies that plants employ to counter salinity stress is sequestration of harmful ions such as Na+ in specific cellular and/or tissue compartments (Blumwald, 2000). Tissue-specific expression of transporters capable of Na+ translocation such as HKT1 can thus affect Na+ contents in a tissue-specific matter. We therefore investigated relative abundance of Na+ and the Na+ : K+ ratio in root tissues by using secondary ion mass spectrometry (SIMS). SIMS uses an energetic destructive primary ion beam to ionise the surface of the sample. Ionised sample fragments are released and characterised by mass spectrometry to reveal ion composition and location at the sample surface. Plants were harvested from time points comparable with the molecular studies, that is either grown on FNS (control) or FNSN followed by K+ starvation. Analysis of control samples showed an even distribution of ions such as calcium (Figure 5a) and phosphorus (not shown) across the transverse section of the root, whereas Na+ and K+ were preferentially localised to the stelar regions. A similar pattern was seen in line 223 (Figure 5b) and 271 (data not shown) for K+, although it appeared that Na+ was less concentrated around the stele and showed a more diffuse pattern across the cortex. With calcium coloured red, sodium green and potassium blue (UTHSCSA ImageTool), the difference in sodium distribution between the transgenic line 223 and the control at lower magnification is very clear (Figure 5c). A semiquantitative estimate of the Na+ : K+ ratio across the central stelar region showed a considerable reduction in Na+ : K+ for the two HKT1 down-regulating lines compared with a control line after salt treatment (Figure 6).
The results present the first evidence of down-regulation of the putative K+ transporter HKT1 in hexaploid wheat. In one of the two independent lines this was through the introduction of an antisense construct and in the other through co-suppression with a sense construct.
The conditions inducing down-regulation of HKT1 were not simple and required growth in saline conditions followed by a period in K+-free conditions in the absence of salt. This led to disappearance of all detectable HKT1 transcript after 24 h in the antisense line and after 96 h in the sense line. The reason why suppression of expression required such a manipulation remains obscure. With the evidence from Arabidopsis of multiple families of known transporters (Maser et al., 2001) it is clear that the regulation of cation transport is a complex affair, as would be expected from the great variety of environmental and developmental conditions under which the plant must be able to maintain ion homeostasis. Previous work has shown that in younger cereal roots (6–9 d), HKT1 is up-regulated by K+ deficiency (Wang et al., 1998) but we found that HKT1 is constitutively expressed in 2-week-old wheat roots and it may be that transcriptional control is different in older plants.
In both lines that showed strong HKT1 down-regulation (271 and 223) no apparent K+-dependent phenotype was detected. In contrast, both lines exhibited a clear Na+-dependent phenotype, consistent with HKT1 being involved in Na+ influx. Thus, after induction of HKT1 down-regulation, both transgenic lines showed better growth than controls under saline conditions (Figure 3), smaller Na+-induced membrane depolarisations in root cortical cells than control lines (Table 1), and lower 22Na influx into roots (Table 2). The results are consistent with the conclusion that, as in Arabidopsis (Rus et al., 2001), HKT1 in wheat mediates Na+ influx, and that decreased expression leads to an inhibition of this pathway and hence improved ability to cope with saline conditions through decreased tissue Na+ concentrations. Despite the evidence from heterologous expression studies that wheat HKT1 mediates K+ transport, we found no evidence that K+-dependent membrane depolarisations were affected (Table 2) suggesting that other transport pathways dominate uptake of this ion in wheat roots at this stage of growth.
SIMS analysis of the root transverse sections suggested that decreased HKT1 expression has its major impact in the stele (Figures 5 and 6). SIMS is a highly sensitive surface analytical technique that is capable of detecting and mapping elements at the parts per million level with submicron resolution (Benninghoven et al., 1987). The method is particularly sensitive for K and Na. As we did not have access to freeze-substitution facilities and the SIMS did not have a cold transfer stage, our method of sample preparation resulted in some image artefacts. In particular, information about the sub-cellular localization of K+ and Na+ could not be gained, as a characteristic redistribution of these ions to the margins of the cells as was observed (Grignon et al., 1997) when comparing different preparation methods for leaf material in soybean. In spite of this there were interesting differences in the distribution of ions across the transverse section, with marked differences obvious in the way Na+ and K+ were localised compared with Ca2+. Although quantifying SIMS is difficult, it is possible to compare multiple samples prepared by the same protocol to obtain the average ratios of the elements present in the tissue sample. This confirmed an overall reduction in Na+ content in HKT1 down-regulating lines and a lower Na+ : K+ ratio in the stele. This was reflected in a more diffuse spread of Na+ across the cortex in the transgenic plants with reduced HKT1 activity. Again, with the redistribution of ions during fixation, there must be cautious interpretation of this result. The evidence that transport of sodium to the shoot is restricted in the transgenics when compared with the controls supports the evidence from SIMS, however. A reduction in the stelar Na+ : K+ ratio is likely to decrease translocation of Na+ to the shoot and thus limit Na+ toxicity in photosynthesising tissues.
The original in situ studies carried out to localise HKT1 showed that it is primarily expressed in the root cortex (Schachtman and Schroeder, 1994). In combination with the high affinity K+ transport capability that was established in heterologous expression systems, it was concluded that HKT1 was likely to function as a constituent of high affinity K+ uptake in wheat. However, subsequent studies showed that HKT1 does not significantly participate in high affinity K+ uptake in wheat (Maathuis et al., 1996). All HKT1 orthologues that have been characterised so far show a distinct Na+ transport capacity (Liu et al., 2000; Rubio et al., 1995; Uozumi et al., 2000). Our results and those of Rus et al. (2001) suggest that the probable role of HKT1–like transporters in planta is in Na+ rather than K+ transport.
Donor wheat (Triticum aestivum L. cv. Florida) plants for transformation were grown in a glasshouse with supplementary lighting provided by 400 W sodium lamps. There was a 16/8 h light/dark photoperiod and day/night temperature regime of 16–18°C/14°C. Caryopses were harvested at approximately 10–15 days post anthesis for embryo isolation. Plant material generated through tissue culture was transferred to soil, acclimatized in a propagator for 2 weeks, vernalised at 4°C for 8 weeks and then grown to maturity under greenhouse conditions.
Plasmid constructs pAHC25 contained the marker genes bar (encoding phosphinothricin acetyltransferase) and uidA (encoding β-glucuronidase), each under the transcriptional control of a separate maize ubiquitin (ubi-1) promoter and first intron (Christensen and Quail, 1996; Christensen et al., 1992), and terminated by the nos. 3′ untranslated region (Bevan et al., 1983). The plasmid containing the gene of interest was constructed as follows: the 5′-truncated HKT1 gene, missing the first 400 bp, was cut out of a yeast expression construct, pYES2 (donated by J.I. Schroeder) at the HindIII site. The HKT1 insert was introduced to the vector pUPLN. pUPLN was produced by modification of the vector pRPLN (Massiah et al., 2001) which contains the rubisco small subunit promoter followed by a polylinker cassette and terminated by nos as in pAHC25. Modification involved the removal of the Rubisco small subunit promoter, excising at thePstI site, with replacement by the ubiquitin promoter from pAHC25. The HKT1 HindIII fragment was cloned into pUPLN at the Spe1 site in the polylinker, with a 2-bp fill-in of vector and insert. The vector containing the maize ubiquitin promoter, the 5′-truncated fragment of HKT1 and the nos terminator was termed pUHKT1N. Constructs containing the gene in forward and reverse orientation were used in transformation.
Production of transgenic wheat T0 lines
Wheat transformation was essentially as described by Rasco-Gaunt and Barcelo (1999) using immature caryopses. The plasmids pAHC25 and pUHKT1N were used in equimolar concentrations. Coated gold particles were bombarded into isolated scutella at acceleration pressures of between 650 and 1100 psi, using a PDS1000/He particle gun (Biorad Laboratories, Hemel Hempstead, UK). Control bombardments, of gold without precipitated DNA and gold with only pAHC25, were carried out for each experiment to generate control, tissue culture-derived plants for each transgenic line for use in the subsequent phenotypic analysis. All plantlets that successfully survived the selection and regeneration cycles were transferred to soil, acclimatized in a propagator and tested for GUS activity where appropriate. GUS-expressing plantlets were vernalised and grown to maturity in the glasshouse.
Generation of homozygous lines
Homozygous lines were generated by normal crossing. T1 lines were analyzed for gene presence by PCR on genomic DNA and positive lines grown on to T2. Between 10 and 20 T2 seeds were germinated and homozygous lines were identified where all progeny tested positive for transgenes. Dihaploid plants were also produced by an anther culture technique (Massiah et al., 2001).
For uida (GUS)-expressing lines it was also possible to identify homozygous progeny by analysis of dry half-seeds from the T1 generation. Endosperm was incubated in GUS assay mix containing 5-bromo-4-chloro-3-indolyl-β-D-glucuronide using the method of McCabe et al. (1988). Blue colouration developed in the tissue after incubation for 24 h at 37°C. Where all endosperm stained blue, co-transformation was confirmed by growing up the embryo half of the seed and analyzing shoot material for HKT1 presence by PCR. This screening method had the advantage of characterizing the T2 status with minimum growth of tissue and DNA extractions, as only the putative homozygotes were grown further.
Molecular analysis of plants
Total genomic DNA was isolated from leaf tissue harvested from primary transformants (T0), first generation (T1) and second generation (T2) seed-derived or T1 and T2 anther culture-derived progenies using a cetyltrimethylammonium bromide (CTAB) method (Stacey and Isaac, 1994). The polymerase chain reaction (PCR) was used to confirm the presence of transgenes. Approximately 250 ng of genomic DNA was used as template in 50 µl reaction volumes. Primers, reaction components and cycling conditions for detection of the marker genes were as previously described in Barro et al. (1997). Annealing temperatures for the reactions were 57°C and 60°C for bar and uidA, respectively. The forward primers for the HKT1 constructs were specific to the gene and the reverse primer was specific to the polylinker sequence immediately upstream of the nos terminator in the plasmid. Gene specific primers for the native HKT1 gene were against a sequence in the 400 bp region missing from the 5′ end of the truncated version used for transformation.
Primers and annealing temperatures were as follows;
PCR reactions were routinely carried out using the Hybaid Omnigene system except for reverse transcriptase PCR (RT-PCR), where optimum results were gained using a Stratagene Robocycler. The products of PCR amplification were analysed by electrophoresis on 1% (w/w) agarose 1 × TAE (40 mm Tris-acetate, 10 mm EDTA, pH 8.0) gels.
Patterns of HKT1 transgene integration were analysed by Southern analysis using a double digest with the restriction enzymes EcoRV and SphI, which excises the complete transcription cassette. Digested genomic DNA (8 µg) was electrophoresed on a 0.6% (w/v) agarose gel and blotted onto positively charged nylon filters (Roche Diagnostics Ltd., Lewes, UK) using standard procedures (Sambrook et al., 1989). An HKT1 digoxigenin-labeled probe was generated by PCR using plasmid pUHKT1N as template and the primers designed for detecting HKT1 described above. Probe synthesis, hybridisation of filters with probe and detection using chemiluminescence were carried out following the manufacturer's instructions (Boehringer Mannheim). Post hybridisation washes were at high stringency (2 × SSC, 0.1% SDS at room temperature followed by 0.1 X SSC, 0.1% SDS at 68°C).
RNA extraction and RT-PCR
RNA was extracted from 100 mg of root tissue using the Qiagen RNeasy protocol. The Omniscript RT kit (Qiagen, Crawley, UK) was used for RT-PCR. To allow semiquantitative comparison of transcript levels, RNA starting concentrations were equal for each sample, and two control reactions were carried out using primers for the constitutively expressed bar and uidA constructs. Band intensities of the bar controls were used as indications of cDNA loading.
Plant growth and ion content analysis
Seeds were germinated on 10 mm CaCl2 and transferred to full strength Hoagland's solution (FNS) once roots were approximately 3 cm long. The seedlings were grown for 2 weeks in FNS plus 100 mm NaCl (FNSN), and then K-starved in the absence of 100 mm NaCl for 12–96 h (FNS-K) and samples were taken for RT-PCR. For growth studies, the plants were grown under these same conditions (FNSN for 2 weeks followed by FNS-K for 24 h) to induce down-regulation of HKT1. Half of the plants from each line were then exposed to high salt stress at 200 mm NaCl in FNS-K (FNShighN) for 4–5 d before harvesting. The other half of the plants were returned to unstressed non-inducing conditions of normal FNS. Fresh weights of shoots and roots were recorded and the percentage difference in FW between salt-stressed and unstressed material was calculated. For ion content analysis, 4–6 plants were sampled from each line. Plant material was dried at 80°C for 48 h, ground in a mortar and pestle and placed into 5 ml of 2.5% (v/v) HCl at 90°C for 30 min The extract was cooled and filtered through Whatman no. 1 filter paper and made up to 50 ml with water. Ion concentration was measured using a flame photometer (Jencons, Leighton Buzzard, Bedfordshire, UK).
Analysis of root exudate from salt-stressed plants
Shoots were excised from plants that had been exposed to high salt conditions as described above (14 d on FNS + 100 mm NaCl followed by 24 h on FNS-K and 5 d on FNShighN) and the cut end was inserted into flexible polythene tubing to facilitate collection of exudate. Shoots were sampled randomly from all pots between 13.00 and 15.00 h on the same day. The root material was placed in the chamber of a pressure bomb (ELE model EL540-300, ELE International Ltd., Leighton Buzzard, Bedfordshire, UK) and pressure applied until root exudate appeared at the cut surface. Exudate was collected for a period of 5 min and ion concentration was measured using a flame photometer as described above. Three plants were sampled from each line and results presented as means ± SE.
Sodium flux experiments and membrane depolarisation analysis
Plants were grown in inducing conditions (FNSN for 14 d followed by FNS-K for 24 h), and 5 cm root segments of 5–7 plants were acclimatised for 30 min to the Na+ uptake solution which comprised 2 mm Mes/Tris pH 6.0, 1 mm CaCl2 and 20, 50 or 100 mm NaCl. Uptake of 22Na+ was initiated by adding 0.2 µCi ml-1 of 22Na+ and allowed to proceed for 30 min at room temperature. After the uptake period, two 8 min washes were carried out in ice cold solution of the same composition but without radiotracer. Subsequently, roots were blotted dry and weighed into scintillation vials and tissue radioactivity was determined by scintillation counting. Results are the average (±SE) of 3–4 independent experiments for each treatment. Membrane potential recordings were carried out as described previously (Maathuis et al., 1996) in buffer containing 2 mm Mes/Tris pH 6.0, 1 mm CaCl2, supplemented with either 0.1 mm KCl or 100 mm NaCl.
Secondary ion mass spectrometry (SIMS)
Root material was cut into transverse sections using a razor blade directly into chemical fixative solution. The samples were chemically fixed using a solution of 3.7% paraformaldehyde and 2.5% glutaraldehyde in 25 mm PIPES buffer pH 6.9 (Chaffey et al., 1998). Ultrapure water was used in the preparation of all reagents. The samples were vacuum infiltrated with the fixation solution for 30 min followed by 4 h incubation with continuous rotation at atmospheric pressure and then washed in 25 mm PIPES buffer for 3 × 15-min periods. The fixed samples were dehydrated in an increasing ethanol:water series (30%, 50%, 70%, 80%, 90%, 100% v/v ethanol), incubating for 30 min in each solution with continuous rotation. A PolyscienceJB-4TM. kit was used to embed the samples in resin. The samples were incubated with the catalysed infiltration resin for 2 × 30 min, 1 × overnight and 1 × 30 min after which the samples were placed in the embedding solution (as per manufacturer's protocols) in aerobically sealed containers for 60 min The embedded samples were removed from their containers and a flat surface for analysis was provided by a glass knife mounted on an ultramicrotome (Ultracut, Reichert-Jung). Prior to analysis, the samples were attached to steel microscope stubs and coated with gold in an Edwards ‘Scancoat six’ plasma-coating unit for 90 s to produce a film of approximately 100 nm thickness. Flat aluminium plates were then attached to the tops of the samples to maintain a flat surface and to facilitate a uniform electric field for accelerating secondary ions into the mass spectrometer.
The SIMS instrument was constructed at the University of Bristol and comprises a focused gallium ion gun (FEI electronically variable aperture type) fitted to a Vacuum Generators model 7035 double-focusing magnetic sector mass analyser. A Thornley-Everhard electron detector allows the acquisition of secondary electron images from the sample at spatial resolutions determined by the diameter of the ion beam. The sample potential is 4 kV and secondary ions are accelerated into the mass spectrometer through an ion extraction lens system. They then travel through an electrostatic energy filter with adjustable pass energy, a magnetic mass fileter and a slit, into a channeltron detector. The instrument is similar to that described by Schuetzle et al., 1989) except that the electrostatic energy analysier is situated ahead of the magnetic mass analyser. To select an area for analysis, ion-induced secondary electron images were first obtained and then the gallium ion beam was used to acquire secondary ion images. Images were obtained in 640 × 480 pixel format, each image taking 90 s to acquire with an ion beam current of 1 nA. The images presented here have been adjusted for brightess for presentation purposes, but within each image, the brightness of each pixel is proportional to the signal received at that point. Sodium : potassium ratios were obtained from the SIMS images, taking the brightness of the area of interest in each image, the instrument sensitivity settings used to obtain the images and the SIMS relative sensitivity factors for sodium and potassium as 2000 and 1800, respectively (Sparrow, 1977), as given below;
Because of the sample matrix effects that afflict SIMS, the absolute accuracy of the results thus obtained is acknowledged to be questionable. However, comparisons between the samples here are believed to be valid because the matrix material is substantially the same in each case.
IACR receives grant-aided support from the Biotechnology and Biological Sciences Research Council (BBSRC) of the United Kingdom and SL and SJB were supported by a ROPA award from the BBSRC. We are grateful to Dr J.I. Schroeder (University of California at San Diego) for the gift of the HKT1 gene, and to the University of the West of England for loan of equipment.