Expression of KT/KUP Genes in Arabidopsis and the Role of Root Hairs in K⁺ Uptake

Intron-Exon Structure of AtKT/KUP Genes

The 13 members of AtKT/KUPs are distributed on all five chromosomes of Arabidopsis (Maser et al., 2001). To determine the structure of the genes encoding the AtKT/KUPs, we compared the predicted cDNA sequences to the gene structure predictions from The Institute for Genomic Research (TIGR) and The Arabidopsis Information Resource (TAIR). For three of the genes (AtKT/KUP8, AtKT/KUP9, and AtKT/KUP12), no full-length cDNA sequences were available; therefore, those gene predictions have not been verified. The AtKT/KUPs have six to 10 exons that are separated by introns of various lengths (Fig. 1). We have placed the AtKT/KUP genes into five groups according to similar intron and exon structures (Fig. 1). One group comprises AtKT/KUP1, AtKT/KUP2, AtKT4/KUP3, and AtKT3/KUP4. Members of this group also appear to be related at the protein level (Maser et al., 2001). A second group is comprised of AtKT/KUP5 and AtKT/KUP7. These genes encode proteins that are very closely related (Maser et al., 2001). The third group encodes the closely related proteins AtKT/KUP6 and AtKT/KUP8 (Maser et al., 2001). Although at the protein level AtKT/KUP6 and AtKT/KUP8 are closed related to AtKT2/KUP2 (Maser et al., 2001), the gene structures are relatively divergent. A fourth group is comprised of AtKT/KUP9, AtKT/KUP10, AtKT/KUP11, and AtKT/KUP12. Although the proteins encoded by AtKT/KUP9, AtKT/KUP10, and AtKT/KUP11 are closely related, the relationship to the AtKT/KUP12 protein sequence is more distant (Maser et al., 2001). A fifth group is comprised of the single gene AtHAK5, which has a long intron (1,484 bp) between the third and fourth exons (Fig. 1). The gene and protein structure of AtHAK5 is distinct from the other AtKT/KUPs (Maser et al., 2001).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Gene structure of the Arabidopsis AtKT/KUPs. Prediction programs from TAIR (http://www.Arabidopsis.org) and TIGR (http://www.tigr.org) were used. Predictions were compared with full-length cDNAs except for the AtKT/KUP8, AtKT/KUP9, and AtKT/KUP12 genes for which full-length cDNA data were not available. Gray boxes, UTRs; black boxes, exons; solid lines, introns. Scale shown in base pairs.

Steady-State Levels of AtKT/KUP Gene mRNA

Plants grown under conditions of sufficient K⁺ (1.75 mm) were harvested at the flowering stage (approximately 45 d after sowing) and separated into roots, older leaves, younger leaves, developing siliques, and flowers. The abundance of the transcripts encoding AtKT/KUP transporters was determined relative to ubiquitin (At4g05320) by real-time reverse transcription RT-PCR (Fig. 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Relative expression levels of individual AtKT/KUP genes as determined by real-time RT-PCR. Plants were grown under control conditions (1.75 mm KCl) and harvested at flowering stage. Transcript abundance was quantified in total RNA from roots (R), older leaves (OL), younger leaves (YL), developing siliques (DS), and flower (F). Expression levels were calculated relative to ubiquitin. The results shown represent real-time RT-PCR analysis of the cDNA synthesized from one experiment. The mean ± se of three reactions are shown as error bars, and the numerical values for SE are shown above the columns where expression was detected. The same results were obtained on two other independent sets of cDNA synthesized for different sets of plants. If relative expression levels were below 0.010, transcripts were considered to be not detectable (ND).

The expression profiles of individual AtKT/KUP genes in different organs under control conditions revealed some differences in AtKT/KUP gene expression. The relative transcript levels of AtKT/KUP11 and AtHAK5 were barely detectable under these conditions. In contrast, many AtKT/KUP genes were expressed in all organs including root, older leaf, younger leaf, developing silique, and flower (Fig. 2). Overall, the expression levels of AtKT/KUPs were higher in the aboveground parts of the plants. For example, AtKT/KUP6, AtKT/KUP8, and AtKT/KUP9 were expressed most highly in young leaves, whereas AtKT/KUP2, AtKT4/KUP3, and AtKT3/KUP4 were highly expressed in developing siliques. The expression of AtKT/KUP10 and AtKT/KUP12 was restricted to older and younger leaves. In flowers, AtKT/KUP2 and AtKT/KUP8 were more highly expressed than other AtKT/KUPs (Fig. 2).

K⁺Content in Tissues

To determine the changes in K⁺ content and AtKT/KUP gene expression in the different Arabidopsis organs, flowering plants were deprived of K⁺ for 1 and 6 d. Dry weight and K⁺ content in roots, older leaves, younger leaves, developing siliques, and flowers were then measured. There were no significant differences in the total dry weight or in the dry weight of the individual tissues 1 or 6 d after deprivation (data not shown). The highest K⁺ content was observed in the roots and stems of plants grown under control conditions (1.75 mm KCl). However, K⁺ content decreased in roots and older leaves 1 and 6 d after deprivation (Table I). Younger leaves and stems had lower K⁺ content after 6 d of K⁺ deprivation than controls. K⁺ content of developing siliques and flowers did not change up to 6 d after deprivation (Table I). K⁺ concentrations in the nutrient solutions 1 and 6 d after K⁺ deprivation were 31 and 12 μm, respectively.

AtHAK5 Is Up-Regulated in Roots and Older Leaves under K⁺ Deprivation

We tested if K⁺ deprivation, K⁺ deprivation then resupply of K⁺, and growth under excess K⁺ and Na⁺ altered AtKT/KUP gene expression. Flowering plants were treated, and gene expression levels were quantified in different tissues and compared with plants grown under control conditions.

In plants at the flowering stage, the expression of most AtKT/KUP genes was insensitive to the various external conditions tested (data not shown). Only the expression of one gene, AtHAK5, was altered by K⁺ deprivation (Fig. 3). The expression level of AtHAK5 was lower than that of other AtKT/KUP genes (except AtKT/KUP11) in all organs including roots under K⁺-sufficient conditions (Fig. 2). In many different experiments, this gene was up-regulated in roots 1 and 6 d after K⁺ deprivation (Fig. 3A). Expression of AtHAK5 in older leaves increased after 6 d of K⁺ deprivation (Fig. 3B). AtHAK5 was down-regulated rapidly in roots when K⁺ was resupplied for 6 and 30 h. AtHAK5 expression was not altered by exposure for 6 d to excess K⁺ (50 and 100 mm) or Na⁺ (50 mm; Fig. 3A). The expression of AtHAK5 in other tissues was relatively insensitive to these treatments.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Temporal expression of AtHAK5 in roots (A) and older leaves (B) of flowering plants grown under hydroponic conditions for 6 weeks. The growth conditions included: K⁺ sufficient (1.75 mm) as a control (lane 1), K⁺ deprivation for 1 (lane 2) and 6 (lane 3) d, K⁺ deprived for 6 d then transferred to K⁺-sufficient conditions for 6 h (lane 4), and 30 h (lane 5). Plants were exposed to excess K⁺ (lanes 6 and 7) and 50 mm Na⁺ (lane 8) for 6 d. Real-time RT-PCR results are shown for roots (A) and older leaves (B). Shown above the real-time PCR data are the end point PCR products from each reaction after agarose gel electrophoresis. Values are mean ± se (n = 3) and representative of at least two independent experiments.

Expression of AtKT/KUP Genes in Root Hairs, Root Tips, and Whole Roots

Under control conditions, all the genes encoding AtKT/KUPs except for AtKT/KUP12 and AtHAK5 were expressed in roots (Figs. 2 and 3). To begin to explore whether the genes encoding the AtKT/KUPs localize to different regions of the root and different root cell types, we used RT-PCR because it provides a very sensitive technique for determining gene expression and does not require large amounts of tissue.

For determination of root specific expression patterns, 6- to 7-d-old seedlings grown on agar plates were harvested. Root hairs were isolated using an enzymatic method (Ivashikina et al., 2001), and root tips were excised using a razor blade under a microscope (Fig. 4A). β-Tubulin (At5g23860) was used for normalization because it could be detected in all root cells. Contamination of root hair protoplasts by other cells was tested using specific primers for AtSKOR, which is not expressed in root hairs, and for AtKC1, which is expressed in root hairs (Ivashikina et al., 2001). AtSKOR expression was not detected in root hair protoplasts, and AtKC1 was present (Fig. 4B). In contrast to our previous results, however, we found that AtHAK5 was expressed in roots under control conditions (Fig. 4C). In root hairs, we found that 10 AtKT/KUP genes were expressed. Only AtKT/KUP5 and AtKT/KUP12 were relatively specific to root hairs. In contrast to previous experiments on more mature plants, we found that AtKT4/KUP3 was not expressed in seedling roots. Although a large number of AtKT/KUPs were found to be expressed in root hair cells (10), only five genes encoding AtKT/KUPs were expressed in root tips (Fig. 4C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Root and leaf expression patterns of AtKT/KUP genes in 1-week-old seedlings grown on vertically oriented agar plates and 6-week-old flowering plants grown in hydroponic solution, both containing 1.75 mm KCl. The diagram shows the regions of the root sampled (A). AtKC1 and SKOR were amplified to determine whether root hair cells were contaminated by other cell types (B). Results of RT-PCR on root hairs (RH), root tip (RT), whole roots (WR), and leaf (L) in 1-week-old and whole roots from 6-week-old plants. Forty cycles of PCR were used (C).

The root expression of certain AtKT/KUP genes that were detected using RT-PCR was regulated by the plant's developmental stage. Comparison of roots from seedlings (1 week) and mature flowering plants (6 weeks) showed that AtKT4/KUP3, AtKT/KUP5, AtKT/KUP7, and AtKT/KUP11 had higher expression levels at the reproductive stage. In contrast, AtHAK5 transcripts were more abundant in plate-grown seedlings (Fig. 4C).

The Importance of Root Hairs in K⁺ Uptake

The expression of 10 of 13 AtKT/KUP genes in root hairs suggested an important role for root hairs in K⁺ uptake. Therefore, we tested two different root hair mutants for K⁺ uptake properties under K⁺-sufficient and -limiting conditions. The trh1-1 (tiny root hairs) mutant (Rigas et al., 2001) had fewer root hairs than the wild type; rhd6 (root hair defective; Masucci and Schiefelbein, 1994) almost completely lacked root hairs (Fig. 5A). Many plant species do not form root hairs in hydroponic conditions, but Arabidopsis roots produce root hairs under these conditions (Fig. 5A). We tested the growth of these lines and found that under K⁺-limiting conditions, the presence of root hairs was important. A decrease in whole-plant biomass and K⁺ content in roots at K⁺-limiting conditions (100 μm KCl) was related to the amount of roots hairs present (Fig. 5, B and C). To confirm the importance of root hairs in K⁺ uptake, we compared the rubidium (Rb⁺) uptake at 20, 100, and 1,000 μm in Ws, trh1-1, and rhd6 (Fig. 6). We found that only rhd6 had significantly lower Rb⁺ uptake rates than Ws (Fig. 6) at 20 μm. At 100 and 1,000 μm, trh1-1 and rhd6 had significantly lower Rb⁺ uptake rates than Ws (Fig. 6). Differences were significant at P < 0.05 (Student's t test).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Light microscope images showing the difference between root hairs of the Wassilewskija (Ws; wild type), trh1-1, and rhd6. Roots were grown in hydroponic culture for 6 weeks (large panels) and on vertically oriented agar plates for 1 week (small panels). A, Whole-plant biomass (B) and K⁺ content in roots (C) of Ws, trh1-1 (tiny root hairs), and rhd6 (root hair defective). Plants were grown with sufficient K⁺ (1.75 mm KCl) and under limiting conditions (100 μm KCl) for 21 d. Values are mean ± se. The numbers above the columns are the calculated percent of biomass or K⁺ content as compared with control plants grown with sufficient K⁺. Scale is same for all photos. Bars = 1 mm.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

⁸⁶Rb⁺ uptake rate at 20, 100, and 1,000 μm RbCl of Ws (wild type), trh1-1 (tiny root hairs), and rhd6 (root hair defective) exposed to K⁺ deprivation for 2 d before uptake was measured (n = 6 ± se).

Complementation and Rb⁺ Uptake of Seven AtKT/KUPs

Only a few AtKT/KUPs have been shown to transport K⁺. These include AtKT/KUP1 (Fu and Luan, 1998), AtKT/KUP2 (AtKT2; Quintero and Blatt, 1997), AtKT3/KUP4 (Rigas et al., 2001), and AtHAK5 (Rubio et al., 2000) in yeast (Saccharomyces cerevisiae) and AtKT/KUP1 and AtKT/KUP2 (Kim et al., 1998; Elumalai et al., 2002) in E. coli. To confirm that AtKT/KUP5, AtKT/KUP6, AtKT/KUP7, AtKT/KUP10, and AtKT/KUP11 function as K⁺ transporters, we expressed these cDNAs in an E. coli mutant deficient in K⁺ uptake. We used AtHAK5 and AtKT3/KUP4 as controls because they have been shown previously to complement a yeast mutant. The E. coli strain TK2420, in which the K⁺ transporters Kdp^-, Kup^-, and Trk^- have been deleted (Epstein et al., 1993), does not grow on K⁺-limited medium when transformed with the empty plasmid pPAB404 (Fig. 7). The plasmid pPAB404 that we used for expression has an IPTG-inducible promoter; therefore, we tested the strains expressing the AtKT/KUP cDNAs with and without IPTG. As shown in Figure 7, AtKT3/KUP4, AtKT/KUP5, AtKT/KUP6, AtKT/KUP7, AtKT/KUP10, AtKT/KUP11, and AtHAK5 all complemented the E. coli mutant TK2420 with IPTG on K⁺-limiting medium. Some of the strains expressing the AtKT/KUP cDNAs grew but more slowly without IPTG on the medium containing 0.5 mm KCl (Fig. 7B). All strains grew on plates containing 30 mm KCl (Fig. 7C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

The complementation of the E. coli TK2420 strain by several AtKT/KUP transporters. The E. coli TK2420 is defective in three K⁺ transporters and was transformed with the empty pPAB404 plasmid or with the plasmid containing AtKT3/KUP4, AtKT/KUP5, AtKT/KUP6, AtKT/KUP7, AtKT/KUP10, AtKT/KUP11, and AtHAK5. The strains were dropped on medium containing: A, 0.5 mm KCl with 0.5 mm isoprophyl-1-thio-β-d-galactropyranoside (IPTG); B) and 0.5 mm KCl without IPTG; and C, 30 mm KCl without IPTG.

The E. coli stains expressing the AtKT/KUP cDNAs were tested for uptake rates first at 2 mm RbCl, and if uptake rates over time did not increase, strains were tested at 100 μm RbCl. The control strain containing the empty vector was tested at both 2 mm and 100 μm RbCl. Rb⁺ uptake over time did not increase in the control strain over a 3-h time course at either RbCl concentration. The strain expressing AtKT/KUP5 showed the highest uptake rate at 2 mm RbCl, whereas the strains expressing AtKT3/KUP4 and AtKT/KUP6 showed increases in uptake over time, but at a lower rate (Table II). AtHAK5-expressing strains showed the highest uptake at 100 μm RbCl, and measurable increases in uptake were found for the strains expressing AtKT/KUP7, AtKT/KUP10, and AtKT/KUP11 (Table II).

Temporal, Spatial, and Conditional Gene Expression

The AtKT/KUP family represents the largest family of K⁺ transport proteins in Arabidopsis (Maser et al., 2001; Very and Sentenac, 2003). A large AtKT/HAK/KUP gene family containing 17 genes also has been identified in rice (Banuelos et al., 2002). To gain insight into where and under what conditions AtKT/KUPs function, we initiated a study to characterize the expression of all 13 members of this gene family. There are many possible reasons why there are a large number of genes encoding these proteins in Arabidopsis. One possible reason is that they are required in different tissues under different conditions. We tested this hypothesis by conducting realtime PCR experiments on flowering plants grown under control conditions. We found that most of the genes encoding AtKT/KUPs were expressed in all the tissues tested under all conditions. Only AtKT/KUP9,10,11,12 group were found to have a more limited range of expression: AtKT/KUP10 and 12 were mainly expressed in leaves, whereas AtKT/KUP11 expression was barely detectable.

To determine whether AtKT/KUPs are involved in plant response to changes in environmental conditions, we tested whether specific genes in the AtKT/KUP family were regulated by K⁺ deprivation, K⁺ excess, or Na⁺ excess. The expression levels of all the AtKT/KUP genes were tested under these conditions, but only a single gene was responsive to changes in external K⁺. AtHAK5 was clearly induced by K⁺ deprivation in flowering plants and was repressed within 6 h of K⁺ resupply. The regulation of AtHAK5 is not the only aspect of the gene that differs from other family members. AtHAK5's intron/exon and protein structure were also different from those of other family members. Rubio et al. (2000) also tested for the conditional expression of AtHAK5 in leaves and roots but found that the expression of this gene was up-regulated by K⁺ deprivation in leaves and down-regulated in roots. The reason for the discrepancy between results is not clear. Therefore, we repeated these experiments numerous times using different biological samples; in all cases, we found that AtHAK5 was up-regulated in roots and older leaves by K⁺ deprivation. The expression of AtHAK5 was also correlated to the changes in K⁺ that we observed in the roots and in older leaves. A tomato homolog of AtHAK5 was found to be up-regulated by K⁺ deprivation and phosphorus and iron deprivation (Wang et al., 2002). In a previous report, the expression of AtKT4/KUP3 was induced by K⁺ deprivation in seedlings (Kim et al., 1998). However, in flowering plants, AtKT4/KUP3 was constitutively expressed.

The developmental changes that occur throughout the lifecycle of a plant may also require different AtKT/KUPs to be expressed or repressed. For example, we found that the background level of AtHAK5 expression was higher in seedlings than in flowering plants, but AtHAK5 expression still increased in response to K⁺ deprivation in both seedlings (data not shown) and flowering plants. We also found that the expression of AtKT4/KUP3 and AtKT/KUP11 was developmentally regulated. These genes were not expressed at the seedling stage but were present in flowering plants. The functional significance of these developmental changes is not known at this time, but the results highlight the varied roles that AtKT/KUPs may play throughout plant development.

Expression of AtKT/KUPs in Root Hairs and Role of Root Hairs in K⁺ Uptake

To further dissect the expression patterns of AtKT/KUPs in roots, we isolated root hairs and root tips and tested for expression using RT-PCR. We found that 10 of the AtKT/KUP genes were expressed in root hairs, and only five AtKT/KUP genes were expressed in the root tip cells. This result was unexpected because although root tips are comprised of many different cell types, root hairs are of a single cell type. The plethora of AtKT/KUPs expressed in the root hairs may provide indirect support for the suggestion put forward by Senn et al. (2001) that the AtKT/KUPs are localized to different membranes, such as the plasma and tonoplast membrane (Senn et al., 2001; Banuelos et al., 2002). Based on functional data showing a difference in the optimal pH for uptake and phylogenetic groupings, AtKT/HAK/KUPs from both barley (Hordeum vulgare) and rice were also suggested to be localized on the vacuolar membrane. At least one rice HAK was tentatively localized to the vacuolar membrane using a green fluorescent protein fusion (Banuelos et al., 2002). In Arabidopsis, the 10 genes expressed in root hairs fall into each of the different clusters of genes (Fig. 1) and proteins (see Maser et al., 2001).

Previous data have shown that the periphery of the root is important in K⁺ uptake (Kochian and Lucas, 1983). Our data show that many AtKT/KUP transporters are expressed in the root hairs, themselves a part of the root periphery. Therefore, it is likely that AtKT/KUP transporters in root hairs are important for K⁺ uptake. Although root hairs have been shown to be important in phosphate uptake in Arabidopsis (Bates and Lynch, 2000) and in other plant species, there are few reports about how root hairs influence the uptake of K⁺ (Jungk, 2001). In contrast with phosphorus acquisition (Schachtman et al., 1998), it is not necessary for roots to explore a large volume of soil in search of K⁺ because of the bulk flow of K⁺ to the root. To test the importance of root hairs on K⁺ uptake, we used the trh1-1 mutant (Rigas et al., 2001), which is the knockout of AtKT3/KUP4 and has a tiny root hairs phenotype, and rhd6, which is nearly devoid of root hairs (Masucci and Schiefelbein, 1994). We compared both the growth and Rb⁺ uptake of these lines and found that when K⁺ was limiting, root hairs played an important role in providing plants with adequate quantities of K⁺ for growth. The quantity of root hairs was qualitatively correlated to growth at low K⁺, so the trh1-1 mutant (which had few root hairs) grew slightly better than the rhd6 mutant, which was devoid of root hairs, but did not grow as much as the wild type. Although our results suggest that root hairs play an important role in K⁺ acquisition, more detailed studies are required to determine which is responsible: the lack of root hairs or the associated loss of transport proteins from the periphery of the root.

Our results support previous work showing that the volume of the root cylinder, which is influenced by the amount of root hairs, is correlated to K⁺ uptake (Jungk, 2001). Although previous studies compared the differences in root hairs between plant species, our studies used mutants of the same species to demonstrate that root hairs play an important role in K⁺ uptake, particularly when K⁺ concentrations are low. Our results are also in accord with previous findings that K⁺ uptake is evenly distributed across the root hair surface (Jones et al., 1995). Based on an even distribution of K⁺ uptake along the root hairs, one would predict that a reduction in root hair surface area would result in a reduction in the absorption surface for K⁺ and, hence, a decrease in uptake at all K⁺ concentrations. In our studies, a decrease in uptake at all concentrations was measured, but only at the lower K⁺ concentrations was a decrease in growth observed.

AtKT/KUPs Function as K⁺ Transporters

In Arabidopsis, AtKT/KUP1, 2, 4 and AtHAK5 have been shown to complement either an E. coli mutant deficient in K⁺ uptake or a yeast strain deficient in K⁺ uptake (Quintero and Blatt, 1997; Santa-Maria et al., 1997; Fu and Luan, 1998; Kim et al., 1998; Rubio et al., 2000; Rigas et al., 2001). Characterization of AtHAK5 showed that this protein functions as a high-affinity transporter. This may be related to the inducibility of this gene when plants are deprived of K⁺, which we demonstrated in this study. Early work in the field of plant ion transport showed that high-affinity K⁺ uptake is induced by K⁺ deprivation (Glass, 1976, 1983). In our study, we show that five more AtKT/KUPs (AtKT/KUP5-7 and 10 and 11) complement an E. coli mutant deficient in K⁺ uptake, which provides additional confirmation that these proteins play a role in K⁺ transport. The Rb⁺ uptake studies support the complementation results.

Based on our findings, it appears that except for AtHAK5, genes encoding AtKT/KUPs are not regulated on the transcriptional level in flowering plants and are expressed in multiple tissues throughout the plant. Although it is clear that root hairs are important for K⁺ uptake by Arabidopsis, the reason for the large number of AtKT/KUP genes expressed in root hairs needs to be resolved. Data presented on AtKT/KUP transporter function support the classification of these proteins as K⁺ transporters.

Exon-intron structure of the AtKT/KUP Gene

Experimentally determined full-length cDNA sequences were located in public gene databases. Predicted gene structures were obtained mainly from TAIR (http://www.Arabidopsis.org) and TIGR (http://www.tigr.org). We compared the predicted cDNA sequences based on the gene structure predictions from TIGR and TAIR with the experimentally determined cDNAs. These sequences were aligned, and pair-wise comparisons were performed with ClustalW (Thompson et al., 1994) using Vector NTI (version 7.0, InforMax, Inc., Frederick, MD). This comparison provided an additional measure of confidence in the gene structures predicted by TIGR and TAIR.

Plant and Growth Conditions

Seedlings of Arabidopsis (Columbia-0, Ws, trh1-1, and rhd6) were grown under short-day conditions (8-/16-h day/night cycle), with 200 μmol m^-2 s^-1 light in aerated hydroponic culture and transferred to long days (16-/8-h day/night cycle) under 200 μmol m^-2 s^-1 light to induce flowering after 5 weeks. The modified nutrient solution contained 1.25 mm KNO₃, 1.5 mm Ca(NO₃)₂, 0.75 mm MgSO₄, 0.5 mm KH₂PO₄, 75 μm FeEDTA, 50 μm H₃BO₃ 10 μm MnCl, 2 μm ZnSO₄, 1.5 μm CuSO₄, and 0.075 μm (NH₄)₆Mo₇O₂₄. Solutions were changed once a week. Plants (Columbia-0) were deprived of K⁺ for either 24 h or 6 d. For the K⁺ deprivation experiment, roots were rinsed and transferred to K⁺-free solution, 1.25 mm KNO₃ and 0.5 mm KH₂PO₄ being replaced by 0.5 mm phosphoric acid adjusted to pH 5.8 with Ca(OH)₂. The solutions for the plants growing in sufficient K⁺ contained 1.75 mm KCl. Plants were also exposed to excess K⁺ (100 mm) and Na⁺ (50 mm) for 6 d after initial K⁺ deprivation. To determine biomass and K⁺ content, Ws (wild type), trh1-1, and rhd6 mutants were grown in 100 μm and 1.75 mm KCl for 21 d. Flowering plants were harvested and separated into roots, older leaves, younger leaves, developing siliques, and flowers. Biomass was measured, K⁺ was extracted from plant tissue in 0.5 m HCl at 37°C, and ion analysis was performed using an Atomic Absorption Spectrophotometer (Aanalyst 300, Perkin-Elmer Applied Biosystems, Foster City, CA).

Vertical Plate Culture and Protoplast Isolation

Seeds were surface sterilized in 70% (v/v) ethanol and 0.1% (v/v) Triton X-100 and then planted on 10-cm-diameter sterile plates on modified nutrient medium (see above) containing 1% (w/v) phytagar (PhytoTechnology Laboratory, Shawnee Mission, KS) and 2% (w/v) Suc. Plates containing seeds were vernalized for 1 d in the dark at 4°C. Plants were grown under a 12-/12-h day/night cycle under 110 μmol m^-2 s^-1 light during the day. After 6 to 7 d, plants were harvested by dividing into root tip, whole root, and leaf. The root tips (0-1 mm) were excised using a razor blade and frozen immediately in liquid N₂. Root hair protoplasts were isolated from separately grown seedlings following the method described by Ivashikina et al. (2001). The enzyme solution contained 0.8% (w/v) cellulase (Onozuka R-10), 0.1% (w/v) pectolyase (Sigma, St. Louis), 0.5% (w/v) bovine serum albumin, 0.5% (w/v) polyvinylpyrrolidone, 1 mm CaCl₂, and 8 mm MES/Tris (pH 5.6). Osmolarity of the enzyme solution and 1 mm CaCl₂ buffer (pH 5.6) were adjusted to 280 mosmol kg^-1 using d-sorbitol. Roots were incubated in enzyme solution at 30°C for 30 min. Protoplasts released from the tip of the root hairs were filtered through a 50-μm screen and washed in 1 mm CaCl₂ buffer by centrifugation twice for 10 min each at 500 rpm and 4°C. The protoplast suspension was stored on ice, and aliquots were used for RT-PCR. To confirm that the isolated protoplasts were from root hairs, specific primers were used for a gene that is not expressed in root hairs, AtSKOR (At3g02850), and for a gene found in root hairs, AtKC1 (At4g32650; Ivashikina et al., 2001). Primers used were: SKORsense (5′-TGA CCC GAA TAA GAC AG-3′), SKORantisense (5′-TGT GTT TCC CCA TCT G-3′), AtKC1sense (5′-ATA TTG CGA TAC ACA AG-3′), AtKC1antisense (5′-GAC CTA ACT TCG CTA AT-3′; Szyroki et al., 2001), β-tubulin (At5g23860) sense (5′-GCC AAT CCG GTG CTG GTA ACA-3′), and β-tubulin antisense (5′-CAT ACC AGA TCC AGT TCC TCC TCC C-3′).

mRNA Isolation and cDNA Preparation for RT-PCR

After K⁺ deprivation, total RNA was isolated in Trizol (Invitrogen, Carlsbad, CA) using a FastPrep (FP120, Qbiogene, Carlsbad, CA) bead beater system for roots, older leaves, and younger leaves. Acid phenol-LiCl was used for isolation of developing siliques and flowers. The quantity of RNA was measured using a spectrophotometer, and its quality was checked by agarose gel electrophoresis. RNA was treated with RNase-free DNase-I (Ambion, Austin, TX) for 45 min followed by DNase-I removal as specified by the manufacturer. DNase-treated RNA was checked for DNA contamination before RT by performing PCR with primers for AtKT/KUP genes with 5× diluted RNA. Two micrograms of DNA free RNA was then reverse transcribed using First-Stand Synthesis System for RT-PCR (Invitrogen) according to the manufacturer's specifications. cDNA concentrations were then normalized using β-tubulin and ubiquitin primers.

Determination of Gene Expression

Gene-specific primers for real time PCR were designed using Vector NTI (version 7) software. Each PCR primer amplified an approximately 150-bp region (for details, see Table III). PCR products were optimized by altering temperature and primer concentration for each primer pair. Products were visually inspected by electrophoresis to ensure that only a single band of the expected size was amplified and that primer dimmer artifacts were absent. PCR products for AtKT/KUP1, AtKT/KUP2, AtKT3/KUP4, AtKT/KUP7, AtKT/KUP8, and AtHAK5 genes were cloned and sequenced to ensure that the PCR primers were specific. After normalization and standardization, the real-time PCR was performed in 50 μL of reaction mixture composed of cDNA, 1 unit of JumpStart TaqDNA Polymerase (Sigma), manufacturer's buffer, gene-specific primers, and 1 μL of a 100,000 dilution of SYBR Green I added to each PCR (Molecular Probes, Eugene, OR). A Bio-Rad iCycler was used, and amplification of PCR products was monitored via intercalation of SYBR Green I. Experiments were conducted using 40 cycles, and data were analyzed as follows. The PCR cycle at which the increase of SYBR green fluorescence becomes statistically significant is called the threshold value (Ct).

The expression of AtKT/KUP genes relative to ubiquitin (At4g05320; primer sense 5′-GAA TCC ACC CTC CAC TTG GTC-3′ and antisense 5′-CGT CTT TCC CGT TAG GGT TTT-3′) was calculated by the equation as follows: where RE is the relative expression and ΔCt: Ct_ubiquitin - Ct_KUPs.

PCR conditions were optimized for amplification efficiency of 100% ± 5% for all primer pairs used. Efficiency was calculated by comparing the experimentally determined and theoretically expected Ct in dilution series of the plasmid DNA using 10, 1, 0.1, 0.01, or 0.001 ng reaction^-1. After completion of each PCR reaction, products were checked by melt-curve analysis and by agarose gel electrophoresis. Two independent experiments were conducted.

Rb⁺ Uptake Assay of Ws, trh1-1, and rhd6

For the Rb⁺ uptake experiments, control solutions (1.75 mm KCl) were replaced 2 d before the experiment with solutions that lacked K⁺, as described above. At the time of transfer, plant roots and rockwool plugs were rinsed with water to remove as much K⁺ as possible. After 2 d of starvation, plants were moved to 250-mL beakers containing Rb⁺ at 20, 100, and 1,000 μm concentrations. After an equilibration time of 5 min, trace amounts of ⁸⁶Rb⁺ were added to start a 10-min uptake period. After the uptake period, plants were moved to beakers containing 0.5 mm CaSO₄, and roots were desorbed for 10 h. Roots were then blotted, weighed, and placed in scintillation vials containing scintillant. Radioactivity in the roots was counted using a Beckman LS6500 scintillation counter (Beckman Instruments, Fullerton, CA).

Complementation of AtKT/KUP Genes in Escherichia coli

The E. coli strain TK2420 deficient in the three K⁺ uptake systems (Kdp^-, Kup^-, and Trk^-) was used for complementation studies (Epstein et al., 1993). The KUP cDNAs AtKT3/KUP4, AtKT/KUP5, AtKT/KUP6, AtKT/KUP7, AtKT/KUP10, AtKT/KUP11, and AtHAK5 were cloned into the plasmid pPAB404 containing an IPTG-inducible promoter (Buurman et al., 1995) and then transformed into TK2420 strain. Each E. coli strain containing a AtKT/KUP cDNA and the strain containing an empty plasmid were grown in medium containing 10 g L^-1 tryptone, 5 g L^-1 yeast extract, and 10 g L^-1 KCl with containing 100 μg mL^-1 ampicillin. Cells were harvested when OD₆₀₀ reached 1.0 and washed four times with deionized water. For complementation tests, 10 μL of cells resuspended in water was placed on minimal media plates (Senn et al., 2001; 5 mm phosphoric acid, 0.4 mm MgSO₄, 6 μm FeSO₄, 1 mm citric acid, 1 mg L^-1 thiamine, 0.2% [v/v] glycerol, 8 mm Asn, 20 μm CaCl₂, and 1.5% [w/v] Bactoagar) with 0.5 and 30 mm KCl. Arg base was used to neutralize the medium to pH 7.5. Cells were grown at 37°C for 16 h on the 30.0 mm KCl plates and for 72 h for the 0.5 mm KCl plates.

Rb⁺ Uptake Assay of E. coli Strains

Eight strains of the E. coli TK2420 were tested for Rb⁺ uptake experiments as follows. The strains contained the empty vector and the vector containing the following cDNAs: AtKT/KUP4 to 7 and 10 and 11 and AtHAK5. Cells were grown to mid-log phase for approximately 5 h in LB containing 30 mm K⁺ and 100 μg mL^-1 ampicillin. During the last 15 min of growth, 0.5 mm IPTG was added. Cells were then washed in K⁺-free medium that contained 2% (w/v) Glc and 10 mm MES buffer (pH 5.5) with CaOH. Cells were then resuspended in K⁺-free medium with 0.5 mm IPTG and placed on a shaker at 37°C for 30 min. Cells were then pelleted and resuspended in fresh K⁺-free medium without IPTG. One milliliter of cells was aliquoted to microfuge tubes. RbCl to a final concentration of either 2 mm or 100 μm with trace amounts of ⁸⁶Rb⁺ were added to tubes and uptake was measured after 0.5, 1, 2, and 3 min. Cells were collected on 0.45-μm-pore membrane filters and washed with 10 mL of 25 mm MgCl₂. Filters were then placed in scintillation vials, and cell-associated ⁸⁶Rb⁺ was quantified with a Beckman LS6500 scintillation counter.