Diversity in Expression Patterns and Functional...Diversity in Expression Patterns and Functional...

Diversity in Expression Patterns and FunctionalProperties in the Rice HKT Transporter Family1[W]

Mehdi Jabnoune, Sandra Espeout, Delphine Mieulet, Cécile Fizames, Jean-Luc Verdeil,Geneviève Conéjéro, Alonso Rodrı́guez-Navarro, Hervé Sentenac, Emmanuel Guiderdoni,Chedly Abdelly, and Anne-Aliénor Véry*

Biochimie et Physiologie Moléculaire des Plantes, UMR 5004 CNRS/INRA/SupAgro-M/UM2, CampusSupAgro-M/INRA, 34060 Montpellier cedex 1, France (M.J., S.E., D.M., C.F., G.C., H.S., A.-A.V.); Laboratoired’Adaptation des Plantes aux Stress Abiotiques, Centre de Biotechnologie de Borj-Cédria, Hammam-Lif 2050,Tunisia (M.J., C.A.); UMR Développement et Amélioration des Plantes, Centre de Coopération Internationaleen Recherche Agronomique pour le Développement/INRA/SupAgro-M/UM2, 34398 Montpellier cedex 5,France (S.E., D.M., J.-L.V., E.G.); and Departamento de Biotecnologı́a, Universidad Politécnica de Madrid,28040 Madrid, Spain (A.R.-N.)

Plant growth under low K+ availability or salt stress requires tight control of K+ and Na+ uptake, long-distance transport, andaccumulation. The family of membrane transporters named HKT (for High-Affinity K+ Transporters), permeable either to K+

and Na+ or to Na+ only, is thought to play major roles in these functions. Whereas Arabidopsis (Arabidopsis thaliana) possesses asingle HKT transporter, involved in Na+ transport in vascular tissues, a larger number of HKT transporters are present in rice(Oryza sativa) as well as in other monocots. Here, we report on the expression patterns and functional properties of three riceHKT transporters, OsHKT1;1, OsHKT1;3, and OsHKT2;1. In situ hybridization experiments revealed overlapping butdistinctive and complex expression patterns, wider than expected for such a transporter type, including vascular tissues androot periphery but also new locations, such as osmocontractile leaf bulliform cells (involved in leaf folding). Functionalanalyses in Xenopus laevis oocytes revealed striking diversity. OsHKT1;1 and OsHKT1;3, shown to be permeable to Na+ only,are strongly different in terms of affinity for this cation and direction of transport (inward only or reversible). OsHKT2;1displays diverse permeation modes, Na+-K+ symport, Na+ uniport, or inhibited states, depending on external Na+ and K+

concentrations within the physiological concentration range. The whole set of data indicates that HKT transporters fulfilldistinctive roles at the whole plant level in rice, each system playing diverse roles in different cell types. Such a large diversitywithin the HKT transporter family might be central to the regulation of K+ and Na+ accumulation in monocots.

Although it is not clear what levels of Na+ are toxicin the plant cell cytosol and actually unacceptable invivo, the hypothesis that this cation must be excludedfrom the cytoplasm is widely accepted. The mostabundant inorganic cation in the cytosol is K+, in plantas in animal cells. This cation has probably beenselected during evolution because it is less chaotropic

than Na+ (i.e. more compatible with protein structureeven at high concentrations; Clarkson and Hanson,1980). Its selection might also be due to the fact that inprimitive cells, which originated in environmentalconditions (seawater) where Na+ was more abundantthan K+, a straightforward process to energize the cellmembrane was to accumulate the less abundant cationand to exclude the most abundant one.

In the cell, K+ plays a role in basic functions, such asregulation of cell membrane polarization, electricalneutralization of anionic groups, and osmoregulation.Concerning the latter function, K+ uptake or release isthe usual way through which plant cells control theirwater potential and turgor. Although toxic at highconcentrations, Na+ can be used as osmoticum andsubstituted for K+, mainly in the vacuole, when theplant is facing low K+ conditions and Na+ is availablein the soil solution. This use of Na+, however, requiresa tight regulation of K+ and Na+ transport and com-partmentalization that becomes crucial in conditionsof high Na+ concentrations in the soil solution. Controlof Na+ and K+ uptake, long-distance transport in thexylem and phloem vasculatures, accumulation in ae-rial parts, and compartmentalization at the cellular

1 This work was supported by the Agropolis Fondation underthe Rice Functional Genomics platform (Montpellier, France), by anAgence Universitaire de la Francophonie graduate fellowship and anational doctoral fellowship from the Tunisian Ministry of HigherEducation, Scientific Research, and Technology (LR02CB02 toM.J.), by the European Research Area Network Plant GenomicsProgramme (grant no. ERA–PG FP/06.018B to E.G. and H.S.), andby the Biotechnology and Biological Sciences Research Council-Institut National de la Recherche Agronomique (grant to H.S. andA.-A.V.).

* Corresponding author; e-mail [email protected] author responsible for distribution of materials integral to the

findings presented in this article in accordance with the policydescribed in the Instructions for Authors (www.plantphysiol.org) is:Anne-Aliénor Véry ([email protected]).

[W] The online version of this article contains Web-only data.www.plantphysiol.org/cgi/doi/10.1104/pp.109.138008

Plant Physiology�, August 2009, Vol. 150, pp. 1955–1971, www.plantphysiol.org � 2009 American Society of Plant Biologists 1955

https://plantphysiol.orgDownloaded on May 31, 2021. - Published by Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.

https://plantphysiol.org

and tissue levels have actually been shown to beessential in plant adaptation to salt stress (GreenwayandMunns, 1980; Flowers, 1985; Hasegawa et al., 2000;Mühling and Läuchli, 2002). Thus, accumulation of Na+

as osmoticum during K+ shortage or plant adaptationto salt stress requires integration at the whole plantlevel of Na+ and K+ membrane transport systemactivities (Apse et al., 1999; Shi et al., 2002; Qi andSpalding, 2004; Ren et al., 2005; Maathuis, 2006; Pardoet al., 2006; Horie et al., 2007).

This report concerns transport systems named HKTupon first identification (for High-Affinity K+ Trans-porters) that are active at the plasma membrane andpermeable to either K+ and Na+ or to Na+ only(Schachtman and Schroeder, 1994; Rodrı́guez-Navarroand Rubio, 2006). Several members of the HKT familyhave already been shown, by genetic approaches, toplay important roles in plant salt tolerance (Berthomieuet al., 2003; Ren et al., 2005; Huang et al., 2006; Byrtet al., 2007) or growth in conditions of K+ shortage(Horie et al., 2007). In Arabidopsis (Arabidopsis thali-ana), the HKT family comprises a single member,AtHKT1;1, which is permeable to Na+ only (Uozumiet al., 2000) and contributes to Na+ removal from theascending xylem sap and recirculation from the leavesto the roots via the phloem vasculature (Berthomieuet al., 2003; Sunarpi et al., 2005). Interestingly, the HKTfamily comprises a much larger number of members inrice (Oryza sativa), with seven to nine genes dependingon the cultivar (Garciadeblás et al., 2003). In line withprevious reports using rice as a model species todecipher the roles that HKT transporters can play inthe plant, we have analyzed the expression patterns ofthree rice HKT genes, OsHKT2;1, OsHKT1;1, andOsHKT1;3, and investigated the functional propertiesof these transporters after heterologous expression,revealing new patterns of expression for HKT trans-porters and striking functional diversity.

RESULTS

Expression Patterns of OsHKT2;1, OsHKT1;1,and OsHKT1;3

The expression patterns ofOsHKT2;1,OsHKT1;1, andOsHKT1;3 were analyzed using the in situ hybridiza-tion technique in 30-d-old rice plants grown in standardconditions, K+ shortage, or salt stress conditions. Crosssections of roots (Fig. 1) and leaves (Fig. 2) werehybridizedwith antisense or sense RNA probes specificof OsHKT2;1, OsHKT1;1, or OsHKT1;3 transcripts. Pos-itive controls using an 18S ribonucleic RNA antisenseprobe indicated that strong hybridization signals couldbe observed in all root and leaf tissues in our experi-ments (Figs. 1G and 2G). Negative controls withHKTor18S sense probes did not lead to any significant labeling(Figs. 1, B, D, F, and H, and 2, B, D, F, and H).

In roots, labeling of OsHKT2;1 expression with theantisense probe was strongest in peripheral layers(epidermis, exodermis, and cortex differentiated into

an aerenchyma). It was also detected in the stele,mainly in phloem (sieve elements and companioncells; Fig. 1A). The expression pattern of OsHKT1;1 inroots was reminiscent of that of OsHKT2;1 (Fig. 1C);however, the labeling required longer incubation timesin color development buffer, suggesting that the levelofOsHKT1;1 expression in roots was lower than that ofOsHKT2;1. Expression of OsHKT1;3 was also detectedboth in peripheral layers (mainly in the cortex) andstele (vascular tissues), but compared with OsHKT2;1and OsHKT1;1, the labeling was strongest in thephloem (Fig. 1E). Whatever the HKT gene studied,changes in growth conditions (see “Materials andMethods”) did not seem to affect its expression pattern(data not shown).

In mature leaves, transcripts of the three HKT geneswere identified in bulliform cells, large highly vacuo-lated cells of the adaxial epidermis involved in leafrolling (Fig. 2, A, C, and E). Staining of OsHKT1;3transcripts was particularly strong in these cells. Ex-pression of the three HKT genes was also detected invascular tissues, phloem, and xylem (parenchyma cellsand differentiating vessels; Fig. 2, A, C, and E). To alower extent, OsHKT2;1 and OsHKT1;3 expression wasfound inmesophyll (Fig. 2, A and E). As in the roots, theexpression patterns of the threeHKT genes in the leaveswere not modified by the different growth conditions(data not shown). Staining, on the other hand, wasdependent on the developmental stage: young imma-ture leaves still folded in the sheath displayed a broaderexpression pattern of the threeHKT genes. In particular,high labeling of mesophyll was observed in sheath forthe three HKT genes (data not shown).

Rectification Properties of HKT Transporters Expressedin Xenopus laevis Oocytes

Functional diversity among the rice HKT trans-porter family was investigated by characterizingOsHKT2;1, OsHKT1;1, and OsHKT1;3 in parallel ex-periments after heterologous expression in Xenopusoocytes. OsHKT2;1 had already been partly character-ized in yeast (Saccharomyces cerevisiae) and Xenopusoocytes, but conclusions were controversial, the trans-porter being described either as a highly selective Na+

transporter (Horie et al., 2001; Garciadeblás et al.,2003) or a weakly selective alkali cation transporter(Golldack et al., 2002). OsHKT1;1 had been reportedto function as a Na+-selective transporter in yeast(Garciadeblás et al., 2003).

For each of the three transporters, OsHKT2;1,OsHKT1;1, and OsHKT1;3, injection of 50 ng of trans-porter copy RNA (cRNA) per oocyte resulted in a highlevel of expression 1 to 2 d after the oocyte injection.Figure 3 shows typical examples of current traces andmean currents recorded in control oocytes injectedwith water and in oocytes expressing OsHKT2;1,OsHKT1;1, or OsHKT1;3. Currents were recorded inNa+-containing medium, since the three rice HKTtransporters were already reported or could be pre-

Jabnoune et al.

1956 Plant Physiol. Vol. 150, 2009



dicted (based on phylogenic relationships; see “Dis-cussion”) to be permeable to Na+. For each HKTtransporter, the level of inward currents recorded inoocytes after 1 to 2 d of expression was at least 25 timeslarger than that recorded in control oocytes. Thus, eachof the three transporters was by far the main transportsystem active at the oocyte membrane.Currents mediated by OsHKT2;1, OsHKT1;1, or

OsHKT1;3 in Na+-containing solutions activated in-stantaneously and remained stable or slightly de-creased over a 1-s voltage clamp stimulation (Fig. 3,B–D). Surprisingly, the three rice HKT transportersexhibited marked differences in rectification proper-ties. Two of them, OsHKT2;1 and OsHKT1;3, were ableto mediate both inward and outward currents, asalready observed in members of the plant HKT familycharacterized so far (Rubio et al., 1995; Fairbairn et al.,2000; Uozumi et al., 2000; Horie et al., 2001; Ren et al.,2005), but the former almost did not rectify while thelatter displayed weak inward rectification, its outwardcurrents saturating when the driving force increased

(Fig. 3, B, D, F, and H). The third transporter,OsHKT1;1, strongly differed from the two others,since it was strongly inwardly rectifying, mediatingno outward current (Fig. 3, C and G). Measurements ofoocyte membrane resting potentials in bath solutionscontaining a few millimolar external Na+ yieldedvalues consistent with the rectification properties ofthe three transporters. For example, the membranepotential in the presence of 3 mM external Na+ wasmuch more negative in oocytes expressing OsHKT2;1or OsHKT1;3 (272 6 1 mV and 290 6 4 mV) than inoocytes expressing OsHKT1;1 (2196 2 mV) or controlwater-injected oocytes (2176 2 mV). A strong inwardrectification such as that observed in OsHKT1;1 hadnot, to our knowledge, been previously reported in theplant HKT family.

Affinity for Na+

Na+ uptake kinetics with respect to external Na+

concentration was studied in the three rice HKT trans-

Figure 1. Localization of OsHKT2;1,OsHKT1;1, and OsHKT1;3 expression inrice root by in situ hybridization. Twelve-day-old plants were subjected to reducedK+ (100 mM) and mild salt stress (50 mMNaCl) conditions for 18 d before tissuecollection. Hybridizations of root crosssections were revealed with VectorBlueKit III. A and B, Localization of OsHKT2;1transcripts. C and D, Localization ofOsHKT1;1 transcripts. E and F, Localiza-tion of OsHKT1;3 transcripts. G and H,Control: hybridization with 18S ribonu-cleic RNA probe. A, C, E, and G, Antisenseprobes. B, D, F, and H, Sense probes. a,Aerenchyma; c, cortex cells; en, endo-dermis; ep, epidermis; ex, exodermis; pe,pericycle; ph, phloem; sc, sclerenchyma;xy, xylem. Bars = 100 mm.

Diversity among Rice HKT Transporters

Plant Physiol. Vol. 150, 2009 1957



porters (Fig. 4). External Na+ concentration was variedin the range 4 mM to 100 mM, external K+ being kept astraces only (approximately 2 mM). Increasing externalNa+ concentration increased the macroscopic inwardconductance of the three transporters in a saturablemanner (Fig. 4). In the two nonrectifying or weaklyrectifying transporters OsHKT2;1 and OsHKT1;3, in-creasing external Na+ also resulted in positive shifts ofthe zero current potential (Fig. 4, A and E). Fitting theinward conductance versus external Na+ with hyper-bolic functions gave rise to apparent half-saturationconstants (K1/2) that were strikingly different in thethree transporters: 1 order of magnitude separated theK1/2 of OsHKT2;1 (0.15 6 0.07 mM Na

+ [n = 5]) fromthat of OsHKT1;3 (3.5 6 2 mM Na+ [n = 5]), which wasitself separated by more than 1 order of magnitudefrom that of OsHKT1;1 (766 8 mM Na+ [n = 6]). Hence,the rice HKT transporter family comprises membersdisplaying high, low, or very low Na+ uptake affinity.It should also be noted that the inward conductanceversus external Na+ relationships could be fitted wellwith a single hyperbolic function in OsHKT1;1 andOsHKT1;3 but not in OsHKT2;1 (Fig. 4B). Fittingconductance data in OsHKT2;1 with the sum of twohyperboles resulted in high-affinity and low-affinityhalf-saturation constants (9.5 6 0.2 mM Na+ and 2.2 60.2 mM Na+, respectively), suggesting that OsHKT2;1may possess a dual mode of Na+ uptake (see below).

Cation Selectivity

The cation selectivity of the three rice HKT trans-porters was examined by comparing the currents me-diated in the presence of different monovalent cations(Fig. 5). Figure 5, A and B, shows mean OsHKT2;1currents recorded successively in the presence of Na+,K+, Rb+, Cs+, and Li+. The external cation concentrationwas either 10 mM (Fig. 5A) or 1 mM (Fig. 5B). In thepresence of 10 mM external Na+, OsHKT2;1 couldmediate large inward and outward currents reversingclose to250 mV, as shown previously (Fig. 4A). WhenLi+ replaced Na+ at the same concentration, bothinward and outward OsHKT2;1 currents could stillbe observed, but the current-voltage relationship wasshifted 50 mV negatively and the inward conductancewas reduced (Fig. 5A). This indicates that OsHKT2;1 isselective for Na+ against Li+, probably by favoring Na+

at a conducting site. When oocytes were bathed with10 mM K+, Rb+, or Cs+, both inward and outwardcurrents through OsHKT2;1 almost disappeared (Fig.5A) and the reversal potential of residual currentsremained close to that observed in the presence of Na+.OsHKT2;1, therefore, is also highly selective for Na+

against K+, Rb+, and Cs+ in these conditions, but themechanism of selectivity seems different from thatoccurring between Na+ and Li+. The strong reductionof both inward and outward currents and the absence

Figure 2. LocalizationofOsHKT2;1,OsHKT1;1, and OsHKT1;3 expres-sion in rice leaf by in situ hybridiza-tion. Twelve-day-old plants weresubjected to reduced K+ (100 mM)and mild salt stress (50 mM NaCl)conditions for 18 d before tissue col-lection.Hybridizations ofmature leafblade cross sections were revealedwith VectorBlue Kit III. A and B,LocalizationofOsHKT2;1 transcripts.C and D, Localization of OsHKT1;1transcripts. E and F, Localization ofOsHKT1;3 transcripts. G and H,Control: hybridization with 18Sribonucleic RNA probe. A, C, E,and G, Antisense probes. B, D, F,and H, Sense probes. abe, Abaxialepidermis; ade, adaxial epidermis;bc, bulliform cells; bs, bundle sheathcells; m, mesophyll; ph, phloem; xy,xylem. Bars = 50 mm.

Jabnoune et al.




of any negative shift of the reversal potential whenexternal K+, Rb+, or Cs+ at 10 mmol L21 replaced Na+

suggest that the selectivity for Na+ against the lattercations is not due to high discrimination but thatexternal K+, Rb+, or Cs+, present at 10 mmol L21,renders OsHKT2;1 very poorly conductive. Similarcurrent patterns were observed when external cationswere present at 100 mmol L21 (data not shown): (1) thereversal potential of OsHKT2;1 currents in bath con-taining Na+ was then close to zero (5 6 2 mV [n = 16])and it was shifted by 281 6 1 mV when Li+ wassubstituted for Na+; (2) almost no current was ob-served in the presence of K+, Rb+, or Cs+. Interestingly,a different current pattern was observed when theexternal cation concentration was reduced to 1 mmolL21 (Fig. 5B). Indeed, when K+, Rb+, and Cs+ werepresent at this concentration, contrary to what wasobserved when they were present at 10 or 100 mM,noticeable inward (Fig. 5B) and outward (data notshown) currents could be recorded. These currentsreversed at a potential more positive than in thepresence of 1 mM Na+, excluding the possibility thatthey could be due to Na+ contamination. Thus, at highexternal cation concentration, OsHKT2;1 would bealmost only permeable to Na+, but at low externalcation concentration it would be permeable also to K+,Rb+, and Cs+. Concerning Li+, the zero current poten-tialwasmorenegative than that recorded in thepresenceof Na+, as already observed in 10 or 100 mM solution,confirming that Li+ was less permeant than Na+. Fur-thermore, the zero currentpotential poorlydependedonthe concentration of Li+ in the externalmedium (1, 10, or

100 mM), suggesting that OsHKT2;1 is not significantlypermeable to Li+. The inward currentsmeasured in LiClsolutions are thus likely mostly carried by Na+ and K+

present as contaminants (see Fig. 5B legend and Fig. 7Abelow).

In order to further analyze the permeability to K+

in OsHKT2;1, currents were recorded in solutionscontaining both Na+ and K+ at different concentra-tions, 0.3, 3, and 10 mM (Fig. 5, C and D). Whenexternal K+ was present at 0.3 mM, increasing exter-nal Na+ from 0.3 to 3 mM and then to 10 mM led toslight increases in OsHKT2;1 inward and outwardconductance (1.8 6 0.02 times between 0.3 and 3 mMand 1.4 6 0.02 times between 3 and 10 mM, theconductances being determined close to the currentreversal potentials) and positive shifts of the currentreversal potential (shift of 61 6 3 mV between 0.3and 10 mM; Fig. 5C), in line with what had beenobserved when K+ was present at nominal concen-tration (Fig. 4A). In the symmetrical experiment,when external Na+ was present at 0.3 mM, increasingexternal K+ from 0.3 to 3 mM and then to 10 mM ledto quite similar shifts of current reversal potential(54 6 2 mV between 0.3 and 10 mM) but to strongdecreases in OsHKT2;1 inward and outward con-ductances (4.8 6 1 times between 0.3 and 3 mM and8.3 6 1 times between 0.3 and 10 mM; Fig. 5D). Thisindicates that when Na+ and K+ are both present inthe bath, at least in the concentrations used in thisexperiment, OsHKT2;1 carries both cations, but ex-ternal K+ exerts an inhibitory effect on OsHKT2;1macroscopic inward and outward conductances.

Figure 3. Rice HKT transporters differin their rectification properties. A to D,Examples of currents recorded in Xen-opus oocytes. E to H, Mean current-voltage (I-V) curves. A and E, Control(water-injected) oocytes. B and F, Oo-cytes expressing OsHKT2;1. C and G,Oocytes expressing OsHKT1;1. D andH, Oocytes expressing OsHKT1;3.Currents were measured in a bath so-lution containing either 3 mM Na-Glu(3 Na; B, D, F, and H) or 30 mM Na-Glu(30 Na; A, C, E, and G). Applied volt-ages varied in the range 2180 to 215mV with an increment of +15 mV(most positive and negative voltagesapplied are indicated at right of thecorresponding current traces). Thedashed lines in A to D mark the zerocurrent level. Data presented in I-Vrelationships (E–H) correspond to totaloocyte currents (means 6 SE; n $ 4).





The selectivity to movovalent cations in OsHKT1;1and OsHKT1;3 was analyzed in a similar way (Fig.5, E–H). When oocytes were bathed successively in100 mmol L21 Na+, K+, Rb+, Cs+, or Li+, significantOsHKT1;1 inward currents were recorded in the pres-ence of Na+ only (currents in Na+ were about 20 timesgreater than those in the presence of the other cations;Fig. 5E). Absence of permeability to K+ in OsHKT1;1 insolutions containing only traces of Na+ was furtherchecked at lower K+ concentration. OsHKT1;1 currentswere negligible and almost constant when external K+

concentration was varied in the range 3 to 50 mM,while they strongly increased when external Na+ con-centration was raised within the same range (Fig. 5F).This confirmed that OsHKT1;1, at least in the absenceof Na+, is not permeable to K+. The question ofOsHKT1;1’s possible permeability to K+ in the pres-

ence of Na+ was difficult to address, because of thestrong inward rectification and hence the absence ofdetectable variation with external concentration of thereversal potential in this transporter. However, we didnot observe any current increase when 5 to 70 mM K+

was added to 30 mM external Na+ (rather, the currentdecreased; Fig. 6, G and I). Thus, no indication thatOsHKT1;1 might work as a Na+-K+ cotransporter wasobtained.

In OsHKT1;3, as in OsHKT1;1, no indication of asignificant permeability to another cation than Na+

was obtained (Fig. 5, G and H). When Na+, K+, Rb+,Cs+, and Li+ were present separately at 10 mmol L21 inthe bath, OsHKT1;3 currents reversed either close to250 mV (247 6 1 mV) in the case of Na+, as inOsHKT2;1, or about270 mVmore negatively (211662 mV) in the case of K+, Rb+, Cs+, or Li+ (Fig. 5G). The

Figure 4. Rice HKT transporters differ in their affinityfor Na+. A, C, and E, Current-voltage (I-V) curves fromoocytes expressing OsHKT2;1 (A), OsHKT1;1 (C), orOsHKT1;3 (E) in the presence of varying external Na+

concentrations. The background solution contained 4mM Na+ and 2 mM K+ as contaminants. Additional Na+

was present as Glu salt (final concentrations as indi-cated). Currents flowing through HKT transporterswere extracted from total oocyte currents by sub-tracting mean currents obtained in water-injectedoocytes from the same batch in the same ionicconditions (see “Materials and Methods”). B, D, andF, Variation of rice HKT inward conductance withexternal Na+ concentration. Macroscopic conduc-tances were defined as slopes of I-V relationships,either close to reversal potentials in the nonrectifyingOsHKT2;1 and OsHKT1;3 transporters (determina-tion with three I-V points) or between 2150 and2120 mV in the inwardly rectifying OsHKT1;1 trans-porter. Macroscopic inward conductances in B, D,and F were extracted from I-V data shown in A, C,and E, respectively. Inward conductance values ob-tained for each oocyte were plotted versus externalNa+ concentration, then normalized by the maximalconductance determinedwith a hyperbolic fit (sum oftwo Michaelis-Menten hyperboles in B and a singlehyperbole in D and F) to suppress variability causedby differences in oocyte levels of expression. Meannormalized inward conductances were fitted againwith the same kind of hyperbolic equations (solidlines) to determine the apparent K1/2 (half-saturationconcentration regardless of the fitting equation used).Fit parameters of the two hyperboles in B were asfollows: Km1 = 9.5 mM, Km2 = 2.2 mM, Gmax1 = 50%,Gmax2 = 50%. In D, it should be noted that OsHKT1;1conductance in 100 mM Na+ (the highest concentra-tion used) was still far from saturation. Therefore, theapparent K1/2 obtained for this transporter is probablynot very accurate. The inset in B is a magnification ofconductance data and fit at low Na+ concentrations.Data are means6 SE (n = 5 in A, B, E, and F and n = 6in C and D) and are representative of at least twoexperiments performed on different oocyte batches.

Jabnoune et al.




negative shift in reversal potential when K+, Rb+, Cs+,or Li+ replaced Na+ was accompanied by a reductionin OsHKT1;3 inward and outward conductance (5 61-fold reduction; Fig. 5G). In solutions containing bothNa+ and K+, no indication of K+ permeability wasobtained either (Fig. 5H): increasing K+ concentrationfrom 0.3 to 10 mM in the presence of 0.3 mM Na+ waswithout any significant effect on OsHKT1;3 reversalpotential and on the inward and outward conduc-tances, whereas increasing Na+ concentration in thesymmetrical experiment (K+ concentration fixed to 0.3mM) increased OsHKT1;3 conductance and shiftedOsHKT1;3’s reversal potential positively by 706 3mV.K+ uptake in the wheat (Triticum aestivum) TaHKT2;1

transporter was first reported to be energized by H+

(Schachtman and Schroeder, 1994), and several HKT-

related transporters in fungi (Trk family), which havetheir activity regulated by external pH, are thought towork as H+-K+ symporters, at least in certain condi-tions (Lichtenberg-Frate et al., 1996; Bihler et al., 1999;Rodrı́guez-Navarro, 2000). Therefore, we investigatedwhether external pH regulates K+ and/or Na+ trans-port in the three rice HKT transporters (SupplementalFig. S1). Experiments were performed with solutionscontaining either K+ “only” (about 1 mM contaminantNa+) or Na+ “only” (approximately 2 mM contaminantK+) for OsHKT2;1 (Supplemental Fig. S1, A and B),since this transporter was shown to be permeable toboth cations, or Na+ only for OsHKT1;1 and OsHKT1;3(Supplemental Fig. S1, C and D). In every case, to favorobservation of H+-Na+ or H+-K+ cotransport behavior,if any, external Na+ concentrations were chosen far

Figure 5. Rice HKT transporters display differentcation selectivities. A to D, Oocyte currents mediatedby OsHKT2;1. E and F, Oocyte currents mediated byOsHKT1;1. G and H, Oocyte currents mediated byOsHKT1;3. A, B, E, and G, Bath solutions withstandard background successively supplementedwith Na+, K+, Rb+, Cs+, or Li+ (chloride salts) atconcentrations of 10 mM (A and G), 1 mM (B), or 100mM (E). C, D, F, and H, Bath solutions with varyingconcentrations, eventually mixed, of Na+ and K+ (Glusalts). Na+ and K+ concentrations are indicated (inmM). Bath solutions containing 1 mM (B), 10 mM (Aand G), and 100 mM (E) cations contained approxi-mately 10, 10, and 25 mM K+ and 10, 15, and 20 mMNa+, respectively. Data are means6 SE (n = 4 in A, C,and D, n = 7 in B and H, and n = 6 in E and F) and arerepresentative of at least two experiments performedon different oocyte batches.





from values saturating the conductance (Na+ concen-tration lower than K1/2 for each transporter), andexternal K+ concentration for OsHKT2;1 was fixedvery low (12 mM). Decreasing external pH from 7.5 to6.5 and 5 did not affect the conductance of the threetransporters and did not induce any significant shift inthe current-voltage relationships. Thus, OsHKT2;1,OsHKT1;1, and OsHKT1;3 would not cotransport H+

or be regulated by external pH.

Inhibition of Na+ Transport by Monovalent Cations in

Rice HKT Transporters

Inhibition by external K+ of high-affinity Na+ uptakethrough OsHKT2;1 had been evidenced by fluxanalysis after expression of the transporter in yeast(Garciadeblás et al., 2003). In Figure 5D, we show thatwhen external Na+ was 0.3 mM, increasing external K+

concentration to the millimolar range strongly de-

Figure 6. Inhibition of Na+ trans-port by monovalent cations in riceHKT transporters. Effect of externalmonovalent cations on Na+ trans-port through OsHKT2;1 (A–F),OsHKT1;1 (G–I), and OsHKT1;3(J). Na+ external concentration(chloride salt) was 30 mM. Addedconcentrations of K+ (A, G, and J),Cs+ (B, H, and J), Rb+ (C and J), andLi+ (D and J; chloride salts) areindicated (in mM). Chloride con-centration was kept constant inall bath solutions by adding,when necessary, HCl titrated withN-methyl-D-glucamine (NMDG). Ato D, G, H, and J, Current-voltage(I-V) curves. E, F, and I, Kinetics ofinhibition by external cations ofinward (E and I) and outward (F)rice HKT conductances. Inwardand outward conductances weredetermined in the nonrectifyingOsHKT2;1 transporter using I-Vdata at potentials between 2150and 2105 mV and between 0 and+15 mV, respectively. Inward con-ductances were determined forthe inwardly rectifying OsHKT1;1transporter between 2150 and2120 mV. External cation concen-trations leading to half conduc-tance inhibition (Ki, as indicated)were obtained by mass action lawfits (solid lines). Data are means 6SE (n = 6 in A–F and n = 5 in G–J).

Jabnoune et al.




creased both inward and outward OsHKT2;1 macro-scopic conductances. In these conditions, the increasesin external K+ concentration also resulted in positiveshifts of OsHKT2;1 reversal potential, indicative of K+

permeability.Sensitivity to external K+ in OsHKT2;1 was further

examined using a 10 times higher Na+ basis than inFigure 5D: 30 mM (Fig. 6A). External K+ was varied inthe range approximately 10 mM (nominal K+ concen-tration) to 70 mM, the ionic strength of the solutionsbeing kept constant by addition of the nonpermeantlarge organic cation N-methyl-D-glucamine, to whichOsHKT2;1 was insensitive (data not shown). In 30 mMNa+, as in 0.3 mM Na+, external K+ at a concentration ofa few millimolar efficiently inhibited both the inwardand the outward OsHKT2;1 macroscopic conduc-tances (Fig. 6A). Half-inhibition concentration, de-rived from mass action law fit, was about the samefor the inward and the outward conductances: close to1 mM (Fig. 6, E and F). Contrary to what was observedin 0.3 mM Na+, however, additions of K+ to the 30 mMNa+ external solution did not induce positive shifts ofOsHKT2;1 reversal potential (Fig. 6A), indicating thatin the presence of high external Na+ concentration,OsHKT2;1 is not significantly permeable to K+. Thus,in these conditions, external K+ inhibits OsHKT2;1without being transported.OsHKT2;1 sensitivity to external Cs+, Rb+, and Li+

was similarly examined in the presence of the same 30mM Na+ background (Fig. 6, B–D). Cs+ (Fig. 6B) and Rb+

(Fig. 6C), like K+, inhibited OsHKT2;1 inward andoutward conductances with about the same efficiencyin the inward and outward directions but withoutmodifying the current reversal potential, suggesting amechanism of inhibition similar to that occurring withexternal K+. Among these three inhibitors, the mostefficient was K+ (Ki approximately 1 mM in presence of30 mM Na+), then Cs+ (Ki approximately 2 mM) and Rb

+

(Ki approximately 3.5 mM; Fig. 6, E and F). In contrastto K+, Cs+, and Rb+, Li+ at a concentration up to 70 mMin the 30 mM Na+ medium did not have any effect onOsHKT2;1 inward and outward conductances (Fig.6D). Li+ did not show any sign of permeation throughOsHKT2;1 either. Interestingly, these results are rem-iniscent of those obtained with K+, Cs+, Rb+, or Li+

“alone” (in the presence of a nominal concentration ofNa+; Fig. 5, A and B), where Li+ was already noticed tobehave differently from K+, Cs+, and Rb+ in permea-tion of OsHKT2;1. Thus, both blockage and permea-tion experiments suggest the presence of binding site(s)with higher affinity for K+, Cs+, and Rb+ than for Li+ inthe OsHKT2;1 translocation pathway.Sensitivity to external K+, Cs+, Rb+, and Li+ in the

presence of 30 mMNa+ was also analyzed in OsHKT1;1(Fig. 6, G–I) and OsHKT1;3 (Fig. 6J). OsHKT1;1 cur-rents were inhibited by K+ (Fig. 6G) and Cs+ (Fig. 6H),although with a strongly lower sensitivity (Ki approx-imately 85 mM; Fig. 6I) than OsHKT2;1 currents.OsHKT1;3 currents were not sensitive to any of thefour alkali cations (Fig. 6J).

Interactions between Na+ and K+ in OsHKT2;1 at

Micromolar and Millimolar External Concentrations

Figures 5 and 6 show that permeation properties inOsHKT2;1 were highly dependent on external con-centrations of Na+ and K+. OsHKT2;1 was found to bepermeable to both Na+ and K+ in the presence of 0.3mM Na+ and a few millimolar external K+ (Fig. 5D). Itwas permeable to Na+ only when external Na+ con-centration was 30 mM (Fig. 6A). Whatever the Na+

concentration, it displayed a very weak macroscopicconductance when external K+ concentration wasmore than 10 mM (Figs. 5, A and D, and 6A).

OsHKT2;1 permeation properties were further ana-lyzed in the presence of very low external Na+ con-centrations (4 mM Na+), external K+ being varied fromthe micromolar to the millimolar range (Fig. 7, A andB). Increasing K+ concentration induced positive shiftsof OsHKT2;1 zero current potential. The magnitude ofthe shift was about 52 mV upon a 100-fold increase inK+ concentration from 10 mM to 1 mM (Fig. 7A). Thewhole set of results indicated that in the presence oflow external Na+ concentrations, even as low as a fewmicromolar, OsHKT2;1 is permeable to K+ within a K+

concentration range from at least a few micromolar toa few millimolar. In this experiment, an inhibitoryeffect of external K+ on OsHKT2;1 macroscopic con-ductance was also clearly apparent (Fig. 7, A and B), asalready mentioned with 0.3 mM external Na+ back-ground (Fig. 5D). However, a detailed investigation ofthe effect of varying K+ concentration in the 4 mM Na+

background revealed that addition of K+ at very lowconcentration first increased OsHKT2;1 conductancebefore it had, at higher concentration, an inhibitoryeffect, with the maximal conductance being observedat approximately 10 mM external K+ (Fig. 7B). It isworth noting that the initial increase in conductancewhile increasing external K+ (up to 10 mM) providesfurther support to the conclusion drawn from reversalpotential analysis that OsHKT2;1 at low external Na+

is permeable to K+. It should also be noted that themodulation (stimulation and inhibition) of OsHKT2;1macroscopic conductance by external K+ was observedin both the inward and the outward directions.

Compiling experiments performed in the presenceof various external Na+ backgrounds highlights com-plex relationships between OsHKT2;1 conductanceand external K+ or Na+ concentrations (Fig. 7C). Inhi-bition of OsHKT2;1 conductance by external K+ wasdetermined neither by the sole external K+ concentra-tion nor by the Na+/K+ concentration ratio (Fig. 7C).

Diverse Conduction Modes in OsHKT2;1

Putative coupling between Na+ and K+ transportin OsHKT2;1 was examined by analyzing more pre-cisely the effect of external K+ and Na+ activities onthe OsHKT2;1 zero current potential (Fig. 7D). In thepresence of 4 mM external Na+ and K+ activity in the 2 to900 mM range, variation of OsHKT2;1 zero current





potential was linear on a logarithmic scale, with a slopeof 26 mV per decade of K+ activity (Fig. 7D, blackdiamonds). Linear variations of OsHKT2;1 zero currentpotential, with similar slope values, were also observedin symmetrical experiments (i.e. the external activity ofK+ being kept constant at 2 mM and that of Na+ varied inthe submillimolar range, 4–900 mM; Fig. 7D, blacktriangles). The slope values were then of about 22 mVper decade of Na+ activity. In summary, variations withexternal K+ or external Na+ of OsHKT2;1 zero currentpotential at submillimolar K+ and Na+ concentrationswere very similar, linear on a logarithmic scale withslope values (22–26 mV) close to that expected for a

Na+-K+ symport with a stoichiometry of 1:1 (slope of 29mV per decade of Na+ or K+ activity). That such slopevalues were slightly lower than the theoretical ones fora symport with a stoichiometry of 1:1 may indicate aNa+:K+ stoichiometry slightly divergent from 1:1 and/ora slight permeability to another ion (Kuroda et al., 2004).Overall, these results strongly suggest that the transportof Na+ and K+ by OsHKT2;1 at submillimolar K+ andNa+ concentrations is coupled. The fact that the coupledtransport behavior is still observed with [Na+]-to-[K+]or [K+]-to-[Na+] ratios of about 500 (Fig. 7D) suggeststhe presence of high-affinity and selective binding sitesfor Na+ and K+, able to discriminate efficiently between

Figure 7. Interaction between Na+ and K+ in OsHKT2;1. A and B, Effect of increasing external K+ on OsHKT2;1 transportproperties in the presence of low external Na+. The external concentration of Na+ was 4 mM (Na+ present as contaminant).External K+ concentrations (2 mM to 1mM, 2 mM being the nominal concentration) are indicated in A. Additional K+ was present asits Glu salt. A, Current-voltage (I-V) curves. B, Conductances extracted from I-V data shown in A. Inward and outwardconductances were determined close to reversal potentials (three-point and two-point determinations, respectively). C, Absenceof unequivocal relationship between the sensitivity of OsHKT2;1 inward conductance to K+ and the K+-to-Na+ concentrationratio in the bath solution. GNa and GNaK refer to OsHKT2;1 conductance in the absence and presence, respectively, of added K

+

(contaminant K+ concentration of 2–10 mM). The concentration of Na+ in the bath solution was 0.004 (contamination inbackground), 0.05, 0.3, 1, or 30 mM. D, Sensitivity of OsHKT2;1 and OsHKT1;3 zero current potential (Erev) to external Na

+ or K+

activity. In each set of solutions, either the concentration of K+ was fixed and that of Na+ was variable or that of Na+ was fixed andthat of K+ was variable (activities indicated in mM). Solid lines correspond to logarithmic fits of Erev versus Na

+ activity or K+

activity. Black symbols represent OsHKT2;1. Slope of the logarithmic fits in the presence of 4 mM Na+ was 26mV per decade of K+

activity. In the presence of 2 mM K+ and varying Na+ concentrations, regressions were performed on Erev values at activities eitherlower than 1 mM or higher than 1 mM, leading to slope values of 22 or 48 mV per decade of Na+ activity, respectively. Whitesymbols represent OsHKT1;3. Slope of the fits was 46 mV per decade of Na+ activity, independent of the K+ activity (0.002 or0.25 mM). Data are means 6 SE (n = 7 in A and B and n = 5–7 in C and D).

Jabnoune et al.




one ion and the other in the OsHKT2;1 ion translocationpathway.When the external concentration of Na+ was in-

creased in the millimolar range (Na+ activity from 1 to25mM, K+ kept constant at 2 mM), the slope of OsHKT2;1zero current potential variation changed from approx-imately 22 mV to approximately 50 mV per decade ofNa+ activity (Fig. 7D, black triangles, dotted line),suggesting that OsHKT2;1 then behaves like a Na+

uniport (slope expected for a selective uniport: 58 mVper decade). This is in agreement with the conclusionpreviously drawn from Figure 6A that at 30 mM exter-nal Na+, OsHKT2;1 is not permeable to K+. It is worthnoting that OsHKT1;3, which was found to be perme-able to Na+ and not to K+ (Fig. 5, G and H), displayedzero current potential variations with a single slopefrom the submillimolar to the millimolar Na+ con-centration range, close to 50 mV per decade of Na+

activity (Fig. 7D, white squares and triangles).External K+ in OsHKT2;1 symport or uniport modes

inhibits the macroscopic conductance. We proposethat this inhibition, which concerns both the inwardand outward directions (Figs. 5D, 6A, and 7A), is dueto the presence in the OsHKT2;1 population of trans-porters in nonconductive (or very weakly conductive)states. Since external K+ has almost the same effect oninward and outward currents, K+ binding leading tononconductive states should involve relatively exter-nal binding sites. Similarly, the presence of an external

binding site is expected to explain Na+ control of theOsHKT2;1 conduction mode (symport or uniport)from the external side.

Figure 8 summarizes the different conductionmodes proposed to be encountered by OsHKT2;1,with examples of external concentrations of K+ andNa+

at which each mode is preponderant. Schematically,the symport mode is expected to operate mainly atsubmillimolar external Na+ and K+, the Na+ uniportmode at external Na+ in the millimolar range andexternal K+ in the submillimolar range, and noncon-ductive states at external K+ in the millimolar to 10 mMrange. Thus, the different conductionmodes displayedby OsHKT2;1 while expressed in Xenopus oocytes alloperate at physiological concentrations of Na+ and K+.Operation of these different conduction modes inplanta, therefore, is likely.

DISCUSSION

OsHKT2;1 Permeability to Monovalent Cations

The question of OsHKT2;1 ionic selectivity wasdebated (Horie et al., 2001; Golldack et al., 2002;Garciadeblás et al., 2003). Our data show thatOsHKT2;1 is permeable to Na+, K+, Cs+, and Rb+ butnot (or weakly) to Li+ (Fig. 5B) at external alkali cationconcentrations lower than 1 mM, while it is signifi-cantly permeable to Na+ only at higher alkali cation

Figure 8. Scheme of OsHKT2;1 conduction modesin Xenopus oocytes depending on external K+ andNa+

concentrations. Three types of conduction modeproposed to be encountered by OsHKT2;1 (Na+-K+

symport, Na+ uniport, and nonconductive states) arepositioned in the scheme according to external K+

and Na+ concentrations. Sets of external Na+ and K+

concentrations (in mM) at which each mode is ex-pected to be preponderant are indicated with refer-ence to supporting figures below each mode. Notethat the two independent arrows present in thediagrammatic representation of the symport modedo not mean that Na+ and K+ could not move throughthe same pore.





concentrations (Figs. 5A, 6, A–D, and 7D). However, inthe presence of high Na+ concentrations, it displays avery low conductance when K+, Cs+, or Rb+ is alsopresent at a high concentration (.1 mM) in the externalsolution (Figs. 5A, 6, A–C, and 7B).

Thus, OsHKT2;1 displays complex permeationproperties and ionic selectivity, differing deeplywhen the external concentrations of cations (Na+, K+,Rb+, and Cs+) are varied from the micromolar to the 10to 100 mM range. Initial and apparently contradictoryanalyses of its ionic selectivity (at least concerning Na+

and K+ in the Xenopus oocyte system) are thus likely todescribe distinct transport behaviors restricted to par-tial concentration ranges.

Na+-K+ Interactions and Transport Mechanismin OsHKT2;1

As outlined in Figure 8 and further discussed below,our electrophysiological analyses in Xenopus oocytesindicate (see last paragraph of “Results”) thatOsHKT2;1 displays three conductive modes depend-ing on external K+ and/or Na+: a Na+-K+ symport,which is the dominating mode at submillimolar exter-nal Na+ and K+; a Na+ uniport, preponderant when theexternal concentration of Na+ is within or above themillimolar range and external K+ is in the submilli-molar range; and nonconductive states at external K+

in the millimolar to 10 mM range. This behavior isreminiscent of (but probably more complex than) thatof the wheat TaHKT2;1 transporter, which has beenshown to work as a Na+-K+ symport when external Na+

andK+concentrationsarebalancedandasaNa+-selectiveuniport when external Na+ is in excess (Rubio et al., 1995;Gassmann et al., 1996; see below).

Kinetic models were extensively developed in the1980s to account for transport properties in various iontransport systems, including symporters (Hansenet al., 1981; Sanders et al., 1984; Blatt et al., 1987). Inthese models, symporter behavior is classically de-scribed by six-state schemes (when stoichiometry is1:1) comprising steps of successive binding of the twotransported ions, voltage-dependent ion translocationacross the membrane, release of the transported ions,and voltage-independent recycling in the membrane.The attractiveness of these models relies on theirsimplicity, since the six-state model can be operation-ally reduced to a two- to four-state model when theconcentration of only one transported ion is varied atone side of the membrane and on the absence ofpresupposition concerning the rate-limiting steps(McCulloch et al., 1990; Meharg and Blatt, 1995;Maathuis et al., 1997; Haro and Rodrı́guez-Navarro,2002). Such models do not include the eventuality ofcompetitive binding by the cotransported ion at one oreach of the two binding sites. It was not possible,within this framework, to fit OsHKT2;1 current-voltagerelationships determined over a large range of Na+

concentration (up to 30 mM; Fig. 4A) with the switchfrom the symport to the uniport mode or to fit the

complex effect of external K+ (stimulation or inhibition,depending on K+ concentration) on OsHKT2;1 currents(data not shown).

To account for the dual mode of transport inTaHKT2;1 (Na+-K+ symport or Na+ uniport), Rubioet al. (1995) and Gassmann et al. (1996) have taken intoaccount the eventuality of competitive binding by thecotransported ion at the binding sites. They proposedthat two high-affinity binding sites, one for K+ and theother one for Na+, named K+ and Na+ coupling sites,respectively, are present in the transporter and need tobe occupied for uptake to occur. Competitive bindingof K+ and Na+ at the K+ “coupling” site determines thepermeation mode of the transporter, namely symportor Na+ uniport. Our data regarding the effects ofvarying external Na+ (Fig. 7D) are compatible withthis model. Since the symport behavior could beobserved in the presence of only a few micromolar K+

and concentrations of Na+ up to 500 times higher (e.g.2 mM K+ and 1 mM Na+; Fig. 7D), the K+ coupling site inOsHKT2;1 must display very high affinity and selec-tivity for K+ against Na+ (higher than in TaHKT2;1,where symport behavior was only observed in quitebalanced Na+ and K+ concentrations). Furthermore,experiments with other alkali cations suggest that theK+ coupling site in OsHKT2;1 would poorly discrim-inate between K+, Cs+, and Rb+ but would be selectiveagainst Li+ (Fig. 5B). Inhibition of symport activity by“high” external K+ (Figs. 5D and 7A) could also fit thecoupling site model, assuming that the binding of K+

at the Na+ coupling site results in a nonconductive (orweakly conductive) state in OsHKT2;1. Similarly, con-cerning the Na+ coupling site in the framework of thediscussed model, the fact that symport activity wasstill observed in the presence of a few micromolar Na+

and concentrations of K+ up to 200 times higher (e.g. 4mM Na+ and 1 mM K+; Fig. 7D) indicates that the Na+

coupling site must display high affinity and selectivityfor Na+ against K+.

The assumption of only two (Na+ and K+) highlyselective coupling sites controlling the transportermode, however, does not help explain the inhibitionby external K+ of OsHKT2;1 in the Na+ uniport mode(Fig. 6A). Indeed, addition of K+ in the millimolarrange in the presence of 30 mM Na+, assuming thepresence of a highly selective K+ coupling site, shouldrestore symport activity, a phenomenon that can bediscarded owing to the absence of a shift of OsHKT2;1reversal potential upon addition of K+ in these condi-tions. The presence of a supplementary site, therefore,should be proposed. Binding of K+ to this regulationsite would prevent conduction. Although this cou-pling and regulation site model remains speculative,our analysis strongly suggests the presence of at leastthree binding sites able to interact with bothNa+ andK+.The presence of a set of sites displaying distinctiveaffinities for Na+ and K+ is not incompatible with thehypothesis that permeation occurs through a narrowpore where several ions move simultaneously in asingle file, hopping from one binding site to the next

Jabnoune et al.




one. Within the framework of this hypothesis, thethree binding sites might be located within the con-duction pathway itself. It should be noted, however,that our data clearly indicate that the OsHKT2;1 con-duction mode is controlled from the external side ofthe transporter (at a location not accessible from theinternal side), since the three conduction modes (sym-port, uniport, and nonconductive states) are deter-mined by the external concentrations of K+ and Na+,even when polarization conditions result in net ef-fluxes through OsHKT2;1 (Figs. 6A, and 7, A and D).

Functional Diversity in the HKT Family

The HKT family was described as comprising twomajor types of transporters when expressed in yeast orXenopus oocytes (Rodrı́guez-Navarro and Rubio,2006): those permeable to Na+ only, like AtHKT1;1,the sole member of the HKT family present in Arabi-dopsis; and those permeable to both Na+ and K+ andeventually endowed with Na+-K+ symport activity, asevidenced for two of them, TaHKT2;1 (Rubio et al.,1995; Gassmann et al., 1996) and HvHKT2;1 (Haroet al., 2005). The presence in rice of nine HKT genes(Garciadeblás et al., 2003) raised the question of thefunctional diversity within this transporter family.Four rice members were already partly characterized:OsHKT1;1 and OsHKT1;5 (reported to be permeable toNa+ only; Garciadeblás et al., 2003; Ren et al., 2005);OsHKT2;1, the permeability of which was debated(see above); and OsHKT2;2, described as permeable toboth K+ and Na+ (Horie et al., 2001). This reportprovides detailed electrophysiological analyses fortwo of these members, OsHKT2;1 and OsHKT1;1,and for a new member, OsHKT1;3, that was notpreviously characterized. OsHKT2;1 is shown to bepermeable to both Na+ and K+ and to be able tomediate symport activity in some conditions (Figs. 5,A–D, 7D, and 8; see above). OsHKT1;1 and OsHKT1;3are shown to be permeable to Na+ only (Fig. 5, E–H).Thus, the new data are consistent with the view thatHKT transporters can be sorted into two types whentheir ionic permeability to Na+ and K+ is considered,all transporters being permeable to Na+ and some alsoto K+. In HKT transporters of the latter type, thequestion of interactions between the two transportedspecies has not been systematically addressed, butwhen such analyses have been carried out (i.e. forTaHKT2;1, HvHKT2;1, and OsHKT2;1), they havealways led to the conclusion that the transporterswere able to behave as Na+-K+ symporters, at least insome environmental conditions (Rodrı́guez-Navarroand Rubio, 2006; Fig. 7D). This may indicate that everyHKT transporter permeable to K+ is endowed with thecapacity to function as a Na+-K+ symport.The rice K+-permeable OsHKT2;2 transporter has

not been characterized in detail, but the availableinformation suggests that its apparent affinity for Na+

is lower than that of OsHKT2;1 and its sensitivity to(blockage by) external K+ is weaker (Horie et al., 2001;

Figs. 4, A and B, 5D, and 6A). Similarly, as discussedabove, OsHKT2;1 displays a higher affinity for Na+

and a stronger sensitivity to K+ than the wheatTaHKT2;1. Our data on OsHKT1;1 and OsHKT1;3also indicate that rice HKT transporters permeable toNa+ only strongly differ in terms of affinity for Na+,rectification, and sensitivity to external K+ (Ren et al.,2005; Figs. 3, C and D, 4, C–F, and 6, G–J). Thephysiological implication of differences in rectificationamong HKT transporters is so far unknown. Evidencethat HKT transporters are not all able to mediateinward and outward Na+ fluxes raises the question ofthe possible occurrence and role of cation effluxesthrough nonrectifying members in vivo.

Phylogenetic Relationships and Functional Properties in

HKT Transporters

Phylogenetic analyses of publicly available HKTsequences revealed, 2 years ago, that the family splitsinto two major branches, which were named subfam-ilies 1 and 2 (Platten et al., 2006). This led to theproposal of a new nomenclature to sort out membersaccording to their subfamily membership, the namesHKT1;x or HKT2;y referring to transporters belongingto subfamily 1 or 2, respectively (Platten et al., 2006).Based on available data, it was suggested that sub-family 1 would correspond to HKT transporters per-meable to Na+ only and subfamily 2 to transporterspermeable to both Na+ and K+. The analysis alsosuggested differences between dicots and monocotsconcerning the number and functional properties ofHKT transporters. Dicotyledonous species appearedto constitute a low number of HKT genes, 1 (as inArabidopsis or poplar [Populus trichocarpa]) or 2 (as inEucalyptus camaldulensis). Furthermore, HKT trans-porters permeable to both Na+ and K+ (subfamily 2)were present only in monocotyledonous species. Amuch larger number of HKT sequences are availablenow (see the phylogenetic tree in Supplemental Fig.S2). The data indicate that some dicotyledonous spe-cies can display a larger number of HKT members, forexample, at least four in grapevine (Vitis vinifera).However, in agreement with previous hypotheses, allnewly identified HKT transporters from dicotyledon-ous species, including the four grapevine members,belong to subfamily 1, based on sequence analysis. Thephylogenetic tree shown in Supplemental Figure S2and our electrophysiological analyses also suggest thatdistinct subgroups might exist in monocot subfamily1. Indeed, OsHKT1;1 and OsHKT1;3, which are low-affinity slightly or strongly rectifying transporters,appear to belong to the same subbranch, distinctfrom that of OsHKT1;5, which seems to be a high-affinity nonrectifying transporter (Ren et al., 2005).

Expression Patterns and Roles in the Plant

Information on tissue localization was availablefor three members of each HKT subfamily so far:





AtHKT1;1 from Arabidopsis (Berthomieu et al., 2003;Sunarpi et al., 2005), OsHKT1;5 from rice (Ren et al.,2005), and McHKT1;1 from ice plant (Mesembryanthe-mum crystallinum; Su et al., 2003) for subfamily 1;TaHKT2;1 from wheat (Schachtman and Schroeder,1994), OsHKT2;1 (Golldack et al., 2002; Kader et al.,2006; Horie et al., 2007), and OsHKT2;2 (Kader et al.,2006) from rice for subfamily 2. The whole set of dataemphasized that HKT transporters are strongly ex-pressed in tissues (root epidermis and cortex, xylemand phloem, vascular bundle regions) playing crucialroles in ion uptake or long-distance transport andredistribution in the plant. The subfamily 1 trans-porters have been constantly reported to be expressedin the plant vasculature and rarely in other tissues.Expression patterns of the subfamily 2 transportershave always been found to include root periphery cellsand often tissues in or close to the plant vasculature.

The tissue specificity of OsHKT2;1 expression pre-sented in this study on the japonica rice ‘Nipponbare’is consistent with the pattern previously reported inthe same cultivar by Horie et al. (2007) and overlapswith, but also differs slightly from, the expression pat-terns reported in indica rice varieties (Golldack et al.,2002; Kader et al., 2006). Our study also provides thefirst information on two other HKT transporters,OsHKT1;1 and OsHKT1;3. Taken as a whole, the newdata reveal broad expression patterns, not restricted toroot periphery cells and vascular regions. For example,hybridization results shown in Figure 2 indicate thatthe three HKTs under investigation are all expressed inbulliform cells, enlarged epidermal cells present inlongitudinal rows in leaves of grasses and involvedin leaf rolling and unrolling. It is tempting to speculatethat these transporters are involved in ion fluxestriggering turgor changes in bulliform cells, allowingthem to control leaf folding in response to environ-mental conditions. Thus, the roles of HKT transporterswould not be restricted to ion uptake from the soilsolution and long-distance transport and redistribu-tion in the plant. Our data also indicate that rootperiphery cells can express HKT transporters fromboth subfamilies 1 and 2. For example, data shown inFigure 1 indicate that root cortical cells expressHKT2;1, HKT1;1, and at a lower level HKT1;3. Also,the three HKTs are expressed in root phloem. In otherwords, the information available in monocots at pres-ent, where the two subfamilies of HKT transporterscoexist, points to complex expression patterns result-ing in intriguing overlaps of transporters endowedwith clearly distinctive functional properties in termsof permeation and selectivity (uniport and/or sym-port activity) and affinity. It is tempting to speculatethat this complexity and diversity allow HKT trans-porters to fulfill highly integrated roles at the wholeplant level.

Reverse genetics approaches in Arabidopsis andanalysis of quantitative trait loci for salt tolerancein rice have highlighted the roles in planta of twoHKT transporters from subfamily 1, AtHKT1;1 and

OsHKT1;5. In Arabidopsis, AtHKT1;1 is expressed instelar tissues of both roots and shoots and plays a rolein preventing Na+ overaccumulation in shoots byunloading Na+ from the ascending xylem sap andloading it into the descending phloem (Berthomieuet al., 2003; Sunarpi et al., 2005; Davenport et al., 2007).Similarly, in rice, OsHKT1;5 expression was detectedmainly in shoot and root vascular tissues, shown toplay a role in Na+ transport by the xylem vasculatureand control Na+ accumulation in leaves (Ren et al.,2005). At the whole plant level, each of these twosystems contributes to adaptation to salt stress.

The roles of HKT transporters from subfamily 2remain more mysterious. A role for OsHKT2;1 in high-affinity Na+ uptake by rice roots has recently beendemonstrated using a reverse genetics approach(Horie et al., 2007). When grown under low Na+ andnominal K+ (“K+-free”) conditions, knockout mutantrice lines disrupted in OsHKT2;1 displayed stronglyreduced high-affinity Na+ uptake by roots comparedwith wild-type plants. They also displayed reducedsize and biomass, suggesting a beneficial role of Na+

nutrition on rice growth at low Na+ concentration inconditions of severe K+ starvation. In wheat, a role forTaHKT2;1 in root Na+ uptake from external solutionscontaining high Na+ concentrations (100–200 mM) wasdeduced from comparative analyses of wild-type andantisense wheat lines (Laurie et al., 2002). Thus,depending on their affinity for this cation, HKT trans-porters can contribute to Na+ uptake in a large con-centration range. On the other hand, no evidence isavailable so far that these systems contribute to K+

transport in the plant. Furthermore, no indication ofNa+-K+ symport activity has been obtained in higherplant tissues yet (Maathuis et al., 1996; Walker et al.,1996; Haro et al., 2005). The absence of evidence thatsubfamily 2 HKT transporters contribute to K+ fluxesin planta, while these systems are clearly permeable toK+, highly expressed in plant cells, and functional inthe cell membrane (as shown by their strong contri-bution to Na+ transport) might be due to redundancyin K+ transport systems from different families. Forinstance, in root periphery tissues, when plants aregrown in classical environmental conditions and/orwhen experiments are performed in classical bathsolutions, K+ transport activity of HKT transportersmight be masked by that of K+ channels from theShaker family (Sentenac et al., 1992; Hirsch et al., 1998;Santa-Marı́a et al., 2000; Pilot et al., 2003; Véry andSentenac, 2003) and/or H+-K+ symporters from theKUP/HAK family (Santa-Marı́a et al., 2000; Bañueloset al., 2002; Gierth and Mäser, 2007). The question ofthe existence of Na+-K+ symport activity in plants,owing to its importance, clearly deserves to be read-dressed.

Two HKT transporters, OsHKT2;1 and TaHKT2;1,are now clearly demonstrated to be able to mediateNa+-K+ symport activity. Furthermore, they are bothinduced upon K+ starvation and can complementyeast mutants defective in K+ uptake (Schachtman

Jabnoune et al.




and Schroeder, 1994; Wang et al., 1998; Horie et al.,2001; Golldack et al., 2002). The eventuality that HKTtransporters can actually function as Na+-K+ uptakesystems in planta is thus worth being further exam-ined, taking advantage of the important progress thathas been made recently in molecular identification ofK+ transport systems and using rice as a model plant.

MATERIALS AND METHODS

Plant Material

Seeds of rice (Oryza sativa japonica ‘Nipponbare’) were provided by the

French Agricultural Research Centre for International Development. They

were treated with 70% ethanol for 1 min, rinsed with sterile distilled water,

treated with 15% potassium hypochlorite solution for 20 min, and rinsed three

times again with sterile distilled water. They were then dipped in water for 12

h. For germination, five seeds were placed on one layer of Whatman germi-

nation paper in a plastic petri dish into which 10 mL of sterile distilled water

had been introduced. The dishes were then stored in a growth chamber (14 h

of light per day, 500 mE m22 s21, 28�C/25�C day/night). Seedlings weretransferred into aerated hydroponic tanks 12 d after germination. The growth

solutions used were based on that described by Yoshida et al. (1976). The

medium used for control treatment contained 0.7 mM KNO3, 0.8 mM KH2PO4,

1.2 mM Ca(NO3)2, 0.5 mM (NH4)2SO4, 1.6 mM MgSO4, 60 mM NaFeEDTA, 45 mM

H3BO3, 20 mM MnSO4, 1.6 mM CuSO4, 1.4 mM ZnSO4, and 0.3 mM (NH4)6Mo7O24.

In experiments with reduced K+ (“2K+”), NaH2PO4 was substituted forKH2PO4 and 0.35 mM Ca(NO3)2 was substituted for 0.7 mM KNO3. For salt

stress treatment, 50 mM NaCl was added to Yoshida or 2K+ medium. The pH

of the growth solutions was adjusted between 5.3 and 5.5 every 2 d.

Concentrations of contaminant K+ and Na+ were about 100 mM. Plants were

grown in a phytotron with a photoperiod of 12 h of light/12 h of dark at 500

mE m22 s21 at day/night temperatures of 28�C/24�C under 65% hygrometry.Experiments were performed on 1-month-old plants.

In Situ Hybridization

RNA Probe Synthesis

Primers were designed for PCR amplification of specific regions of

OsHKT2;1, OsHKT1;1, or OsHKT1;3 cDNAs, respectively, as follows:

OsHKT2;1-Up and OsHKT2;1-Down (5#-CATACTCGTTGGCTCGTTGC-3#and 5#-GGTTGCTTGGCTTCAGGAAC-3#), OsHKT1;1-Up and OsHKT1;1-Down (5#-GGAAGTCAATTTCTTCTGATCCA-3# and 5#-TCATTTCAGGAT-GAACTCCTTG-3#), and OsHKT1;3-Up and OsHKT1;3-Down (5#-CTCCA-AACTTTCCGCACATT-3# and 5#-TTGACTTGTCTCGTGGCTTG-3#). Primerswere also designed for PCR amplification of an 18S ribosome cDNA region

(18S ribosome probe used as a positive control): Rib-Up (5#-CCGACCCT-GATCTTCTGTGAAGGG-3) and Rib-Down (5#-CCAAGTCAGACGAAC-GATTTGCACG-3#). Primers containing the above sequences but extended attheir 5# ends with the T7 RNA polymerase promoter sequence (5#-GCGAAAT-TAATACGACTCACTATAGGGAGA-3#) were also designed and were namedOsHKT2;1-T7-Up and OsHKT2;1-T7-Down, OsHKT1;1-T7-Up and OsHKT1;1-

T7-Down, OsHKT1;3-T7-Up and OsHKT1;3-T7-Down, and RibT7-Up and

RibT7-Down. Finally, one primer corresponding to the T7 end was also

designed, E-T7 (5#-GCGAAATTAATACGACTCAC-3#). Sense and antisenseprobes were synthesized in two steps: a first PCR was performed with one

primer Up and one primer T7-Down or with one primer Down and one primer

T7-Up. The PCRwas carried out with 2 ng of DNA, 13 Taq polymerase reactionbuffer, 1.5 mM MgCl2, 10 mM dATP, dCTP, dGTP, and dTTP, 0.4 mM forward and

reverse oligonucleotide primers, and 5 units of Taq polymerase in a final volume

of 50mL. Samplesweremaintained for 3min at 95�C, subjected to 35 cycles (95�Cfor 30 s, 55�C for 30 s, and 72�C for 1min), and thenmaintained for 7min at 72�C.A second PCR was performed using 2003 dilutions of the first PCR products,E-T7 primer, and primer Up or Down to obtain the DNA template for the

desired sense or antisense probe. The PCR was carried out with 1 mL of DNA

dilution using the same protocol as in the first PCR. Amplification products

were purified by electrophoresis (on a 1% agarose gel). After purification in

ethanol, the fragments were used to generate sense or antisense digoxigenin-

labeled transcripts by in vitro transcription (T7 MAXIScript Kit; Ambion). Final

purification of the probes was done using ethanol precipitation. Probes were

labeled using the dIG Oligonucleotide Tailing Kit (Roche).

Tissue Preparation

Root and leaf samples were fixed with 4% paraformaldehyde in phosphate

buffer (0.2 M, pH 7.5) for 16 h. They were washed in phosphate buffer

containing 0.1 M Gly and then dehydrated through a series of incubations in

ethanol-butanol solutions: ethanol 50%, twice for 7 min; ethanol 70%, 30 min

and then 1 h; ethanol 95%, twice for 30 min; butanol 100%, twice for 1 h;

butanol 100%, 3 d. They were then progressively embedded in paraffin

(ParaplastPlus; Structure Probes) through several baths: butanol:Safesolv

(Labonord; 2:1), 1 h; butanol:Safesolv (1:2), 1 h; Safesolv:Paraplast (1:1), 1 h;

Paraplast, twice for 1 h and then 48 h. After inclusion, samples were cut into

8-mm sections and mounted on silanized slides (Dako). Sections were

subjected to dewaxing with Safesolv and rehydration through an ethanol

series: ethanol 100%, twice for 10 min; ethanol 70%, twice for 1 min; ethanol

50%, 2 min; and two 2-min baths into diethyl pyrocarbonate (DEPC) sterile

water. Slides were then subjected to protein digestion with proteinase K (0.1

unit mL21) for 30 min at 37�C, the reaction being stopped with an arrest bufferbrought twice for 5 min at room temperature. Slides were washed with

phosphate-buffered saline (PBS) plus 0.2% Gly for 2 min and then twice for 2

min in PBS only, then dehydrated through an ethanol series: ethanol 50%,

1 min; ethanol 70%, 1 min; ethanol 100%, twice for 1 min.

Hybridization

Hybridization was done overnight at 45�C in a humidity chamber with adenatured (5 min at 65�C) hybridization mixture: 50% deionized formamide,10% 103 SSC, 10% dextran sulfate, 2% Denhardt’s reagent, 0.1 mg mL21

tRNA, approximately 200 ng of the probe of interest, and DEPC water such

that the final volume for one hybridization was 100 mL. Slides were washed

(23 SSC twice for 5 min, 0.23 SSC twice for 30 min at 50�C, and 13NaCl-Tris-EDTA buffer [NTE] for 5 min at 37�C), subjected to RNase treatment (13NTE,20 mg mL21 RNase A for 30 min at 37�C), washing (13NTE twice for 5 min at37�C, 0.23 SSC twice for 30 min at 55�C), blocking (13 Blocking Reagent[Roche] for 30 min at 37�C), incubation with anti-digoxigenin antibodyconjugated with alkaline phosphatase (1:500 dilution [Roche]; 1 h at 37�C),washing (PBS three times for 10 min at room temperature), and incubation in

the color development buffer (100 mM Tris-HCl, pH 9.5, 100 mM NaCl) twice

for 10 min at room temperature. Hybridization signals were detected with

VectorBlue Kit III (Vector Laboratories) in color development buffer in the

dark at room temperature (incubation until signal appears). Color develop-

ment was stopped by incubation in water. Sections were then mounted in

Mowiol. Slides were observed with a Leica DM6000 microscope under white

light. Photographs were taken with a Retiga 2000R camera (QImaging), and

images were processed through Volocity 4.0.1 (Improvision). In situ hybrid-

ization experiments have been conducted on the Plate-Forme d’Histocytologie

et Imagerie Cellulaire Végétale (http://phiv.cirad.fr/) using microscopes of

the Montpellier Rio Imaging platform (www.mri.cnrs.fr).

Expression in Xenopus laevis Oocytes

Rice HKT cDNAs were subcloned into the pGEMDG vector (D. Becker)

downstream from the T7 promoter and between the 5# and 3# untranslatedregions of the Xenopus b-globin gene. Capped and polyadenylated cRNAs

were synthesized in vitro from linearized vector using the mMESSAGE

mMACHINE T7 kit (Ambion). Oocytes were isolated as described previously

(Véry et al., 1995), injected with 50 ng of OsHKT cRNA in 50 nL of DEPC-

treated water or with 50 nL of DEPC-treated water (for control oocytes), and

then kept at 18�C in ND96 solution (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2,1 mM MgCl2, 2.5 mM Na-pyruvate, and 5 mM HEPES-NaOH, pH 7.4)

supplemented with 0.5 mg L21 gentamycin until electrophysiological record-

ings.

Electrophysiology

Whole oocyte currents were recorded using the two-electrode voltage-

clamp technique 1 to 3 d after cRNA injection. The voltage-clamp amplifier

was an Axoclamp 2A (Axon Instruments). Voltage-pulse protocols, data





acquisition, and data analyses were performed using pClamp9 (Axon Instru-

ments) and Sigmaplot8 (Jandel Scientific) software. Both membrane potential

and current were recorded. Correction was made for voltage drop through the

series resistance of the bath and the reference electrode using a voltage-

recording microelectrode in the bath close to the oocyte surface (potential

difference between the local electrode and the reference subtracted from the

potential of intracellular electrodes in the amplifier for real-time series

resistance correction). All electrodes were filled with 1 M KCl. The oocyte

was continuously perfused during the voltage-clamp experiment. All bath

solutions contained a background of 6 mM MgCl2, 1.8 mM CaCl2, and 10 mM

MES-1,3-bis[tris(hydroxymethyl)methylamino]propane, pH 5.5. Monovalent

cations were added as Glu or chloride salts (chloride concentration constant in

each set of solutions). D-Mannitol was added when necessary to adjust the

osmolarity (same osmolarity in each set of solutions in the range 220–280

mosmol). In experiments where external pH was changed to 7.5, HEPES

replacedMES in the background solution. K+ and Na+ actual concentrations in

solutions were systematically measured by flame photometry. K+ and Na+

nominal concentrations were in the 2 to 25 mM range (see figure legends for

detailed values). To extract OsHKT-mediated currents from total oocyte

currents, mean currents recorded in water-injected control oocytes (n = 4–7)

from the same batch in the same ionic conditions were subtracted from those

recorded in OsHKT-expressing oocytes. Unless otherwise stated, presented

current-voltage relationships were constructed with OsHKT-extracted cur-

rents.

Sequence data from this article can be found in the GenBank/EMBL data

libraries under accession numbers AB061311 (OsHKT2;1 mRNA), AJ491816

(OsHKT1;1 mRNA), and AJ491818 (OsHKT1;3 mRNA).

Supplemental Data

The following materials are available in the online version of this article.

Supplemental Figure S1. Insensitivity of OsHKT2;1, OsHKT1;1, and

OsHKT1;3 to extracellular pH.

Supplemental Figure S2. Phylogenetic relationships among HKT trans-

porters.

ACKNOWLEDGMENTS

We are grateful to Mike Blatt and Jean-Baptiste Thibaud for helpful

discussions and to Julia Davies for critical reading of the manuscript.

Received March 3, 2009; accepted May 23, 2009; published May 29, 2009.

LITERATURE CITED

Apse MP, Aharon GS, Snedden WA, Blumwald E (1999) Salt tolerance

conferred by overexpression of a vacuolar Na+/H+ antiport in Arabi-

dopsis. Science 285: 1256–1258

Bañuelos MA, Garciadeblás B, Cubero B, Rodrı́guez-Navarro A (2002)

Inventory and functional characterization of the HAK potassium trans-

porters of rice. Plant Physiol 130: 784–795

Berthomieu P, Conéjéro G, Nublat A, Brackenbury WJ, Lambert C, Savio

C, Uozumi N, Oiki S, Yamada K, Cellier F, et al (2003) Functional

analysis of AtHKT1 in Arabidopsis shows that Na+ recirculation by the

phloem is crucial for salt tolerance. EMBO J 22: 2004–2014

Bihler H, Gaber RF, Slayman CL, Bertl A (1999) The presumed potassium

carrier Trk2p in Saccharomyces cerevisiae determines an H+-dependent,

K+-independent current. FEBS Lett 447: 115–120

Blatt MR, Rodrı́guez-Navarro A, Slayman CL (1987) Potassium-proton

symport in Neurospora: kinetic control by pH and membrane potential.

J Membr Biol 98: 169–189

Byrt CS, Platten JD, Spielmeyer W, James RA, Lagudah ES, Dennis ES,

Tester M, Munns R (2007) HKT1;5-like cation transporters linked to Na+

exclusion loci in wheat, Nax2 and Kna1. Plant Physiol 143: 1918–1928

Clarkson DT, Hanson JB (1980) The mineral nutrition of higher plants.

Annu Rev Plant Physiol 31: 239–298

Davenport RJ, Muñoz-Mayor A, Jha D, Essah PA, Rus A, Tester M (2007)

The Na+ transporter AtHKT1 controls xylem retrieval of Na+ in

Arabidopsis. Plant Cell Environ 30: 497–507

Fairbairn DJ, LiuW, Schachtman DP, Gomez-Gallego S, Day SR, Teasdale

RD (2000) Characterisation of two distinct HKT1-like potassium trans-

porters from Eucalyptus camaldulensis. Plant Mol Biol 43: 515–525

Flowers TJ (1985) Physiology of halophytes. Plant Soil 89: 41–56

Garciadeblás B, Senn ME, Bañuelos MA, Rodrı́guez-Navarro A (2003)

Sodium transport and HKT transporters: the rice model. Plant J 34:

788–801

Gassmann W, Rubio F, Schroeder JI (1996) Alkali cation selectivity of

the wheat root high-affinity potassium transporter HKT1. Plant J 10:

852–869

Gierth M, Mäser P (2007) Potassium transporters in plants: involvement

in K+ acquisition, redistribution and homeostasis. FEBS Lett 581:

2348–2356

Golldack D, Su H, Quigley F, Kamasani UR, Muñoz-Garay C, Balderas E,

Popova OV, Bennett J, Bohnert HJ, Pantoja O (2002) Characterization

of a HKT-type transporter in rice as a general alkali cation transporter.

Plant J 31: 529–542

Greenway H, Munns R (1980) Mechanisms of salt tolerance in nonhalo-

phytes. Annu Rev Plant Physiol 31: 149–190

Hansen UP, Gradmann D, Sanders D, Slayman CL (1981) Interpretation

of current-voltage relationships for “active” ion transport systems. I.

Steady-state reaction-kinetic analysis of class-I mechanisms. J Membr

Biol 63: 165–190

Haro R, Bañuelos MA, Senn ME, Barrero-Gil J, Rodrı́guez-Navarro A

(2005) HKT1 mediates sodium uniport in roots: pitfalls in the expression

of HKT1 in yeast. Plant Physiol 139: 1495–1506

Haro R, Rodrı́guez-Navarro A (2002) Molecular analysis of the mechanism

of potassium uptake through the TRK1 transporter of Saccharomyces

cerevisiae. Biochim Biophys Acta 1564: 114–122

Hasegawa PM, Bressan RA, Zhu JK, Bohnert HJ (2000) Plant cellular and

molecular responses to high salinity. Annu Rev Plant Physiol Plant Mol

Biol 51: 463–499

Hirsch RE, Lewis BD, Spalding EP, Sussman MR (1998) A role for the

AKT1 potassium channel in plant nutrition. Science 280: 918–921

Horie T, Costa A, Kim TH, Han MJ, Horie R, Leung HY, Miyao A,

Hirochika H, An G, Schroeder JI (2007) Rice OsHKT2;1 transporter

mediates large Na+ influx component into K+-starved roots for growth.

EMBO J 26: 3003–3014

Horie T, Yoshida K, Nakayama H, Yamada K, Oiki S, Shinmyo A (2001)

Two types of HKT transporters with different properties of Na+ and K+

transport in Oryza sativa. Plant J 27: 129–138

Huang S, Spielmeyer W, Lagudah ES, James RA, Platten JD, Dennies ES,

Munns R (2006) A sodium transporter (HKT7) is a candidate for Nax1, a

gene for salt tolerance in durum wheat. Plant Physiol 142: 1718–1727

Kader MA, Seidel T, Golldack D, Lindberg S (2006) Expressions of

OsHKT1, OsHKT2, and OsVHA are differentially regulated under NaCl

stress in salt-sensitive and salt-tolerant rice (Oryza sativa L.) cultivars.

J Exp Bot 57: 4257–4268

Kuroda T, Bihler H, Bashi E, Slayman CL, Rivetta A (2004) Chloride

channel function in the yeast TRK-potassium transporters. J Membr Biol

198: 177–192

Laurie S, Feeney KA, Maathuis FJ, Heard PJ, Brown SJ, Leigh RA (2002) A

role for HKT1 in sodium uptake by wheat roots. Plant J 32: 139–149

Lichtenberg-Frate H, Reid JD, Heyer M, Höfer M (1996) TKHp in

Schizosaccharomyces pombe. J Membr Biol 152: 169–181

Maathuis FJM (2006) The role of monovalent cation transporters in plant

responses to salinity. J Exp Bot 57: 1137–1147

Maathuis FJM, Sanders D, Gradmann D (1997) Kinetics of high-affinity K+

uptake in plants, derived from K+-induced changes in current-voltage

relationships: a modelling approach to the analysis of carrier-mediated

transport. Planta 203: 229–236

Maathuis FJM, Verlin D, Smith FA, Sanders D, Fernández JA, Walker NA

(1996) The physiological relevance of Na+-coupled K+-transport. Plant

Physiol 112: 1609–1616

McCulloch SR, Beilby MJ, Walker NA (1990) Transport of potassium in

Chara australis. II. Kinetics of a symport with sodium. J Membr Biol 115:

129–143

Meharg AA, Blatt MR (1995) NO32 transport across the plasma membrane

of Arabidopsis thaliana root hairs: kinetic control by pH and membrane

voltage. J Membr Biol 145: 49–66

Mühling KH, Läuchli AE (2002) Effect of salt stress on growth and cation

Jabnoune et al.




compartmentation in leaves of two plant species differing in salt

tolerance. J Plant Physiol 159: 137–146

Pardo JM, Cubero B, Leidi EO, Quintero FJ (2006) Alkali cation exchangers:

roles in cellular homeostasis and stress tolerance. J Exp Bot 57: 1181–1199

Pilot G, Pratelli R, Gaymard F, Meyer Y, Sentenac H (2003) Five-group

distribution of the Shaker-like K+ channel family in higher plants. J Mol

Evol 56: 418–434

Platten JD, Cotsaftis O, Berthomieu P, Bohnert H, Davenport RJ,

Fairbairn DJ, Horie T, Leigh RA, Lin HX, Luan S, et al (2006) Nomen-

clature for HKT transporters, key determinants of plant salinity toler-

ance. Trends Plant Sci 11: 372–374

Qi Z, Spalding EP (2004) Protection of plasma membrane K+ transport by

the Salt Overly Sensitive1 Na+-H+ antiporter during salinity stress. Plant

Physiol 136: 2548–2555

Ren ZH, Gao JP, Li LG, Cai XL, Huang W, Chao DY, Zhu MZ, Wang ZY,

Luan S, Lin HX (2005) A rice quantitative trait locus for salt tolerance

encodes a sodium transporter. Nat Genet 37: 1141–1146

Rodrı́guez-Navarro A (2000) Potassium transport in fungi and plants.

Biochim Biophys Acta 1469: 1–30

Rodrı́guez-Navarro A, Rubio F (2006) High-affinity potassium and sodium

transport systems in plants. J Exp Bot 57: 1149–1160

Rubio F, Gassmann W, Schroeder JI (1995) Sodium-driven potassium

uptake by the plant potassium transporter HKT1 and mutations con-

ferring salt tolerance. Science 270: 1660–1663

Sanders D, Hansen UP, Gradmann D, Slayman CL (1984) Generalized

kinetic analysis of ion-driven cotransport systems: a unified interpretation of

selective ionic effects on Michaelis parameters. J Membr Biol 77: 123–152

Santa-Marı́a GE, Danna CH, Czibener C (2000) High-affinity potassium

transport in barley roots: ammonium-sensitive and -insensitive path-

ways. Plant Physiol 123: 297–306

Schachtman DP, Schroeder JI (1994) Structure and transport mechanism of

a high-affinity potassium uptake transporter from higher plants. Nature

370: 655–658

Sentenac H, Bonneaud N, Minet M, Lacroute F, Salmon JM, Gaymard F,

Grignon C (1992) Cloning and expression in yeast of a plant potassium

ion transport system. Science 256: 663–665

Shi H, Quintero FJ, Pardo JM, Zhu JK (2002) The putative plasma

membrane Na+/H+ antiporter SOS1 controls long-distance Na+ trans-

port in plants. Plant Cell 14: 465–477

Su H, Balderas E, Vera-Estrella R, Golldack D, Quigley F, Zhao C, Pantoja

O, Bohnert HJ (2003) Expression of the cation transporter McHKT1 in a

halophyte. Plant Mol Biol 52: 967–980

Sunarpi, Horie T, Motoda J, Kubo M, Yang H, Yoda K, Horie R, Chan WY,

Leung HY, Hattori K, et al (2005) Enhanced salt tolerance mediated by

AtHKT1 transporter-induced Na+ unloading from xylem vessels to

xylem parenchyma cells. Plant J 44: 928–938

Uozumi N, Kim EJ, Rubio F, Yamaguchi T, Muto S, Tsuboi A, Bakker EP,

Nakamura T, Schroeder JI (2000) The Arabidopsis HKT1 gene homolog

mediates inward Na+ currents in Xenopus laevis oocytes and Na+ uptake

in Saccharomyces cerevisiae. Plant Physiol 122: 1249–1259

Véry AA, Gaymard F, Bosseux C, Sentenac H, Thibaud JB (1995) Expres-

sion of a cloned plant K+ channel in Xenopus oocytes: analysis of

macroscopic currents. Plant J 7: 321–332

Véry AA, Sentenac H (2003) Molecular mechanisms and regulation of K+

transport in higher plants. Annu Rev Plant Biol 54: 575–603

Walker NA, Sanders D, Maathuis FJM (1996) High-affinity potassium

uptake in plants. Science 273: 997–998

Wang TB, Gassmann W, Rubio F, Schroeder JI, Glass ADM (1998) Rapid

up-regulation of HKT1, a high-affinity potassium transporter gene, in

roots of barley and wheat following withdrawal of potassium. Plant

Physiol 118: 651–659

Yoshida S, Forno DA, Cock JH, Gomez KA (1976) Routine procedure for

growing rice plants in culture solution. In S Yoshida, DA Forno, JH

Cock, KA Gomez, eds, Laboratory Manual for Physiological Studies of

Rice, Ed 2. International Rice Research Institute, Los Baños, Philippines,

pp 61–66




Diversity in Expression Patterns and Functional...Diversity in Expression Patterns and Functional...

Documents

Transcript of Diversity in Expression Patterns and Functional...Diversity in Expression Patterns and Functional...