A Single-Pore Residue Renders the Arabidopsis RootAnion Channel SLAH2 Highly Nitrate SelectiveC W

Tobias Maierhofer,a,1 Christof Lind,a,1 Stefanie Hüttl,a,1 Sönke Scherzer,a Melanie Papenfuß,a Judy Simon,b

Khaled A.S. Al-Rasheid,c Peter Ache,a Heinz Rennenberg,b Rainer Hedrich,a Thomas D. Müller,a

and Dietmar Geigera,2

a University of Würzburg, Institute for Molecular Plant Physiology and Biophysics, D-97082 Würzburg, Germanyb Institute of Forest Sciences, Chair of Tree Physiology, University of Freiburg, 79110 Freiburg, Germanyc Zoology Department, College of Science, King Saud University, Riyadh 11451, Saudi Arabia

ORCID ID: 0000-0003-0715-5710 (D.G.)

In contrast to animal cells, plants use nitrate as a major source of nitrogen. Following the uptake of nitrate, this majormacronutrient is fed into the vasculature for long-distance transport. The Arabidopsis thaliana shoot expresses the anionchannel SLOW ANION CHANNEL1 (SLAC1) and its homolog SLAC1 HOMOLOGOUS3 (SLAH3), which prefer nitrate assubstrate but cannot exclude chloride ions. By contrast, we identified SLAH2 as a nitrate-specific channel that is impermeablefor chloride. To understand the molecular basis for nitrate selection in the SLAH2 channel, SLAC1 and SLAH2 were modeled to thestructure of HiTehA, a distantly related bacterial member. Structure-guided site-directed mutations converted SLAC1 intoa SLAH2-like nitrate-specific anion channel and vice versa. Our findings indicate that two pore-occluding phenylalaninesconstrict the pore. The selectivity filter of SLAC/SLAH anion channels is determined by the polarity of pore-lining residueslocated on alpha helix 3. Changing the polar character of a single amino acid side chain (Ser-228) to a nonpolar residue turnedthe nitrate-selective SLAH2 into a chloride/nitrate-permeable anion channel. Thus, the molecular basis of the anion specificityof SLAC/SLAH anion channels seems to be determined by the presence and constellation of polar side chains that act inconcert with the two pore-occluding phenylalanines.

INTRODUCTION

In the model plant Arabidopsis thaliana, SLOW ANION CHAN-NEL1 (SLAC1) represents the founder anion channel of a smallgene family associated with CO2- and abscisic acid–dependentstomatal closure (Negi et al., 2008; Vahisalu et al., 2008). SLAC1,together with SLAC1 HOMOLOGOUS3 (SLAH3), represent themolecular determinants for S-type anion channel activity inguard cells (Geiger et al., 2009b, 2010, 2011), while in mesophyllcells, S-type anion channel activity appears to be driven by justSLAH3 (Geiger et al., 2009b, 2010, 2011; Demir et al., 2013).Both the SLAC1 and SLAH3 channels interact with distinct kinase-phosphatase pairs shown to be associated with water stresssignaling (Geiger et al., 2009b, 2010, 2011; Brandt et al., 2012;Scherzer et al., 2012). Although anion channels of the SLAC/SLAHfamily show homology to tellurite resistance/C(4)-dicarboxylatetransporters expressed in bacteria, archaea, and fungi, SLAC1and SLAH3 exhibit a preference for nitrate and chloride but notfor malate (Schmidt and Schroeder, 1994; Geiger et al., 2009b;Chen et al., 2010). These channels are further characterized by

a NO32/Cl2 permeability ratio of ;10, in the case of SLAC1, and

of 20 for SLAH3 (Geiger et al., 2009b, 2011).Distant relatives of plant SLAC/SLAH anion channels from

Escherichia coli have been annotated as tellurite resistanceproteins (TehA) due to their association to the tehA/tehB operon(Walter et al., 1991; Taylor et al., 1994). While these early studiesdid not elucidate the function of the bacterial SLAC1 homolog,recently, Chen et al. (2010) determined the crystal structure ofTehA from the bacteria Haemophilus influenzae (Hi-TehA). Thestructure of Hi-TehA (and the SLAC1 model derived from it)shows a trimeric assembly composed of quasisymmetrical sub-units. Within the trimer, each monomer subunit consists of10 transmembrane helices. The inner five-helix transmembranering contains a pore with a central, highly conserved, phenylala-nine residue (in Arabidopsis SLAC1, F450) potentially representingthe anion gate. Mutation of this pore phenylalanine turned Hi-TehA and its plant homolog, At-SLAC1, into an open anionchannel (Chen et al., 2010). However, the molecular determinantsof anion selectivity remain unclear.Besides SLAC1, which is exclusively expressed in guard cells,

four SLAC1 homologs (SLAH1 to 4) have been identified inArabidopsis (Negi et al., 2008). Apart from SLAC1, SLAH3 is theonly S-type anion channel that has been functionally characterizedto date (Geiger et al., 2011; Demir et al., 2013). When expressedunder the control of the SLAC1 promoter, SLAH1 and SLAH3 areable to rescue the slac1-2 stomatal phenotype (Negi et al., 2008).By contrast, SLAH2, which is the closest homolog of SLAH3(Dreyer et al., 2012), did not complement the phenotype caused bythe loss of the chloride- and nitrate-permeable SLAC1 (Negi et al.,2008). This raises the question of whether SLAH2 exhibits different

1 These authors contributed equally to this work.2 Address correspondence to geiger@botanik.uni-wuerzburg.de.The author responsible for distribution of materials integral to the findingspresented in this article in accordance with the policy described in theInstructions for Authors (www.plantcell.org) is: Dietmar Geiger (geiger@botanik.uni-wuerzburg.de).C Some figures in this article are displayed in color online but in black andwhite in the print edition.W Online version contains Web-only data.www.plantcell.org/cgi/doi/10.1105/tpc.114.125849

This article is a Plant Cell Advance Online Publication. The date of its first appearance online is the official date of publication. The article has been

edited and the authors have corrected proofs, but minor changes could be made before the final version is published. Posting this version online

reduces the time to publication by several weeks.

The Plant Cell Preview, www.aspb.org ã 2014 American Society of Plant Biologists. All rights reserved. 1 of 14

biophysical properties or interacting regulators than those identi-fied in SLAH3 and SLAC1.

SLAH2 is expressed in the stele of the root, the cells thatimmediately surround the vasculature. These cells determine theanion composition of the sap flow between root and shoot(Köhler and Raschke, 2000; Köhler et al., 2002; Lin et al., 2008).Here, we used Xenopus laevis oocytes to study the anion se-lectivity of SLAH2 and compared it to that of SLAC1. In this way,we were able to establish SLAH2, in contrast to the NO3

2 andCl2 permeable SLAC1, and homologs from other kingdoms, asan anion channel that transports nitrate exclusively (Chen andHwang, 2008; Jentsch, 2008; Picollo et al., 2009; Accardi andPicollo, 2010; Hedrich, 2012; Stauber et al., 2012). Based onhomology models for SLAC1 and SLAH2, structure-guided site-directed mutations were used to explore the molecular basis ofthe extraordinarily high nitrate selectivity of SLAH2. Exchangingthe extracellular half of transmembrane helix (TM) 3 was sufficientto convert the chloride/nitrate permeable SLAC1 into a nitrate-selective channel similar to SLAH2. Within the pore region ofSLAH2, the polarity of a single amino acid determined the distinctivenitrate selectivity of SLAH2, suggesting that the hydrophobicitydifference between chloride and nitrate enables SLAH2 todifferentiate between these two anions.

RESULTS

Stele-Specific Expression of SLAH2

To localize the site of SLAH2 expression, we isolated mRNA fromvarious tissues of 6-week-old Arabidopsis Columbia-0 plants andfound its transcripts restricted to the roots (Supplemental Figure1A). Consistent with this, in plants expressing the SLAH2-promoter:GUS (b-glucuronidase) constructs, the root stele was identifiedas the SLAH2 expression site (Supplemental Figures 1B to 1E).

SLAH2 Physically Interacts with CPK21 and CIPK23Protein Kinases

Based on this stele-specific expression, we tested whetherSLAH2 exhibits stele-specific features that might conflict withthose required for normal guard cell function. As a first step, weinjected SLAH2 complementary RNA (cRNA) into Xenopus oocytesand used the two-electrode voltage clamp technique to analyze thebiophysical properties of the channel. When expressed in Xenopusoocytes alone, SLAH2 was electrically silent. Since the guard cellanion channels SLAC1 and SLAH3 are activated by calcium-dependent protein kinases (CPKs) (Geiger et al., 2010, 2011;Brandt et al., 2012; Scherzer et al., 2012) and CBL-interactingprotein kinase 23 (CIPK23) is required to activate plant K+

channels and nitrate transporters (Geiger et al., 2009a; Ho et al.,2009; Ho and Tsay, 2010), we performed BiFC (bimolecularfluorescence complementation)–based interaction assays inoocytes using these candidate protein kinases and SLAH2.SLAH2 was fused to the C-terminal half of yellow fluorescentprotein (YFP), while the protein kinases of the CPK and CIPKtypes were fused with the complementary (N-terminal) half of theYFP molecule. When the fusion proteins were expressed alone,e.g., SLAH2:YFPCT only, no YFP-specific fluorescence was

observed in the oocytes (Supplemental Figure 2A). Upon co-injection of SLAH2:YFPCT together with CPK21:YFPNT, how-ever, YFP fluorescence (BiFC) could be detected. Similar resultswere obtained with CIPK23:YFPNT/CBL1 (Figures 1A and 1B;Supplemental Figures 2B and 2D). To rule out the possibility thatthe SLAH2-kinase interactions were overexpression artifacts, weperformed experiments using CPK31, a close homolog of thecalcium-dependent protein kinase 21, and CIPK24, a member ofthe CIPK family that activates the sodium/proton antiporter SOS1(Qiu et al., 2002), as negative controls. Coexpression of CPK31:YFPNT or CIPK24:YFPNT with SLAH2:YFPCT in oocytes resulted inno or only faint YFP fluorescence (Supplemental Figures 2C and2E), indicating that the SLAH2-kinase interaction is specific.

Protein Kinase–Activated SLAH2 Mediates Nitrate Fluxes

To study the action of CPK21 on SLAH2 in an environment thatwas not limited by the oocyte’s resting cytosolic free Ca2+ con-centration, we used truncated CPK variants (e.g., CPK21DEF) thatlacked the junction and calcium binding domains. This modifi-cation renders CPK21 Ca2+ independent and, thus, constitutivelyactive (Geiger et al., 2010, 2011). When SLAH2 was coexpressedwith CPK21DEF, steady state currents of up to 10 µA appearedonly in the presence of 30 mM NO3

2 in the external solution(Figure 1C). A similar result was obtained when coexpressingSLAH2 and CIPK23 together with the Ca2+ sensor protein CBL1(Figure 1D).To obtain a quantitative readout of the channel activation ca-

pacity of both protein kinases, we measured SLAH2-mediatedcurrents at a membrane potential of 2100 mV. When 10 ngSLAH2 cRNA was coinjected with 10 ng CPK21 or CIPK23cRNA and 5 ng CBL1 cRNA, similar current amplitudes could bemonitored for each kinase independently (Figure 1E). Coex-pression of SLAH2 with CIPK23 in the absence of CBL1 did notresult in SLAH2 activation. To test whether the SLAH2 channelproperties depend on the nature of the activating kinase, weanalyzed the voltage-dependent gating response of SLAH2 inthe presence of nitrate. Following the expression of SLAH2with either CPK21 or CIPK23/CBL, we analyzed the voltage-dependent open probability Po. The relative Po curves of thetwo kinases superimposed and thus support our hypothesisthat both kinases activate SLAH2 in a similar way, as opposedto modifying the properties of the anion channel in a kinase-specific manner (Figure 1F).

SLAH2 Activation Requires Phosphorylation

The fact that SLAH2-derived anion currents were observed onlywhen protein kinases were coexpressed suggested that phos-phorylation might be necessary for channel activation. We thereforedisrupted CPK21 and CIPK23 kinase activity by exchanging Asp-204 with Ala (CPK21DEFD204A) and Lys-60 with Asn (Harmonet al., 1994; Harper et al., 1994; Franz et al., 2011) and coexpressedSLAH2 alone or with either CPK21DEF, CPK21DEFD204A,CIPK23/CBL1, or CIPK23 K60N/CBL1. In this experiment,SLAH2-mediated anion currents were detected only in oocytescoexpressing SLAH2 with CPK21DEF or CIPK23/CBL1, but notin the absence of a kinase or with the kinase-inactive mutants(Figure 1E). To monitor phosphorylation of SLAH2 by CPK21,

2 of 14 The Plant Cell

we performed in vitro assays using recombinant CPK21 and theN-terminal cytoplasmic domain of SLAH2 (amino acids 1 to 142)as a substrate. Using radiolabeled [g32P]ATP in the presence of2 µM free Ca2+, we observed autophosphorylation of CPK21 aswell as transphosphorylation of the N terminus of SLAH2(Supplemental Figure 3).

SLAH2 Operates as Nitrate Channel

Interestingly, in purely chloride-based media, the SLAH2-kinasepairs remained electrically silent (Figures 1C and 1D). In thepresence of 10 mM NO3

2 at the external face, SLAH2 is acti-vated (Figure 2A). Increasing the NO3

2 concentration in the bath10-fold shifted the half-maximal activation voltage (V1/2) ofSLAH2 from 280 to 2180 mV (Figure 2B). This indicates thatnitrate is operating as a permeating anion and channel opener.During 20-s voltage pulses to negative membrane potentials,SLAH2 currents activated rapidly and deactivated with slow ki-netics (Figures 1C and 1D), a feature reminiscent of S-type anionchannels (Schroeder and Keller, 1992; Marten et al., 2007).Consistent with the hypothesis of a nitrate-gated and -perme-able conductance, SLAH2-mediated anion currents could bedetected only in the presence of nitrate at the extracellular side.Upon reduction of the nitrate concentration, the reversal po-tential shifted to positive membrane voltages (Figure 2C). A 10-fold change in the external nitrate concentration resulted in a58-mV shift of the reversal potential. To compare the permeabilityof SLAH2 to a range of anions, NO3

2 was replaced by HCO32,

malate, sulfate, iodide, bromide, and chloride (Figure 2D). Thepermeability sequence obtained was: NO3

2 (1) > iodide (0.68860.029) >> bromide (0.090 6 0.011) >> chloride (0.012 6 0.007) >malate (0.011 6 0.001) > HCO3

2 (0.009 6 0.002) > sulfate(0.006 6 0.002). The mean nitrate to chloride permeability ratio[P(NO3

2)/P(Cl2)] was 82.4. However, dependent on the quality ofthe respective oocyte batch P(NO3

2)/P(Cl2) values of up to 200were calculated. Thus, the root-expressed SLAH2 representsa nitrate channel that, considering the availability of physiolog-ically relevant anions such as nitrate and chloride in the soil, isfunctionally permeable to nitrate only.To characterize this extraordinarily strong preference for ni-

trate in more detail, we exposed SLAH2-expressing oocytes todifferent chloride/nitrate ratios (Figure 2E). Anion currents in thepresence of 10 mM nitrate and 3 mM chloride in the bath me-dium were rather low. Increasing the chloride concentration to90 mM affected neither the current amplitude nor the reversalpotential or the relative PO; this further underpins our findingsthat chloride is not a SLAH2 substrate or channel modulator(Figures 2E and 2F). This characteristic clearly separates SLAH2from the previously described S-type anion channels homologs,SLAC1 and SLAH3. Here, an increase in the chloride concen-tration in the presence of external nitrate shifted the reversal

Figure 1. Activation of SLAH2 via Interaction with Members of DistinctProtein Kinase Families.

(A) and (B) BiFC experiments revealed that SLAH2 physically interactswith CPK21 and CIPK23. SLAH2:YFPCT was either coexpressed withCPK21DEF:YFPNT (A) (a Ca2+-independent and constitutively activeCPK21 mutant lacking the EF-hands and the junction domain) or withCIPK23:YFPNT/CBL1 (B) in oocytes. Representative cells are shown.(C) and (D) Whole-oocyte currents of SLAH2-expressing Xenopusoocytes recorded in the presence or absence of 30 mM NaNO3

2.Voltage pulses lasting 20 s were applied ranging from +40 to 2180 mVin 20-mV decrements followed by a 3-s voltage pulse to 2120 mV. Theholding potential was clamped to 0 mV. Coexpression of SLAH2 andCPK21DEF (C) or CIPK23/CBL1 (D) in oocytes resulted in macroscopicanion currents in the presence of external NO3

2 only. The currentsslowly deactivated at negative membrane potentials. Representativecells are shown.(E) SLAH2-mediated currents induced by the coexpression with eitherCPK21DEF or CIPK23/CBL1 did not differ significantly in current am-plitudes. The kinase-inactive mutant CPK21DEF D204A was not able toelicit SLAH2-mediated currents comparable to those attained withCPK31DEF. Activation of SLAH2 by CIPK23 appeared only in the pres-ence of CBL1. However, the coexpression of SLAH2 with the kinaseinactive mutant CIPK23 K60N/CBL1 led to no channel activation. Thesame result was obtained after coexpression of SLAH2 with CIPK24/CBL4 (n $ 3, mean 6 SE).

(F) The relative open probability (rel. PO) of SLAH2 coexpressed withCPK21DEF or CIPK23/CBL1 was plotted against the membrane poten-tial. Data points were fitted with a Boltzmann equation (solid lines, n $ 4,mean 6 SD).[See online article for color version of this figure.]

S-Type Anion Channel Selectivity 3 of 14

Figure 2. NO32 Dependence of SLAH2 Currents.

(A) Steady state currents (ISS) of SLAH2- and CPK21DEF-coexpressing oocytes were normalized to the values at +40 mV in 100 mM NO32 and plotted

against the membrane potential. Currents were measured at the indicated voltages in the presence of the indicated concentrations of external NO32

concentrations and 3 mM Cl2 (n $ 5 experiments, mean 6 SD).(B) The relative open probability (rel. PO) of SLAH2 at various NO3

2 concentrations plotted against membrane potential. Data points were fitted witha single Boltzmann equation (solid lines, n $ 5 experiments, mean 6 SD).(C) Reversal potentials Vrev of SLAH2- and CPK21DEF-expressing oocytes are shown as a function of the logarithmic external nitrate concentration. Asexpected for an anion-selective channel, the reversal potential shifted with a slope of 58 mV per decade to more negative values with increasing nitrateconcentrations. To avoid loading of the oocytes with nitrate during the measurement, the reversal potentials were recorded in the current-clamp modus(n = 3 experiments, mean 6 SD).(D) Relative permeability (rel. permeability) of SLAH2 coexpressed with CPK21DEF in Xenopus oocytes (permeability for NO3

2 was set to 1). Standardbath solution contained 50 mM of the indicated anion (pH 7.5). The reversal potentials used for the calculation of the relative permeability were recordedin the current-clamp modus (n = 4 experiments, mean 6 SD).(E) NO3

2 and voltage dependence of steady state currents (ISS) of SLAH2- and CPK21DEF-coexpressing oocytes. Currents were normalized to +40 mVin 100 mM NO3

2 + 3 mM Cl2 and plotted against the membrane potential. With 10 mM NO32 in the bath medium SLAH2-mediated currents could be

observed, which decreased at negative membrane potentials. The addition of 100 mM Cl2 to the bath medium in the presence of 10 mM NO32 had no

4 of 14 The Plant Cell

potential and the relative open probability to negative membranepotentials in both cases, indicating that SLAC1 as well asSLAH3 are permeable for chloride (Geiger et al., 2011). In thepresence of 3 mM chloride in combination with 97 mM nitrate,the SLAH2 current amplitude was substantially higher than inthe presence of elevated chloride concentrations (Figure 2E).Besides affecting the peak current, the 10-fold increase in nitrateconcentration shifted the reversal potential to negative mem-brane potentials (Figure 2E, compared with Figure 2C). Thisfeature characterized SLAH2 as an anion channel permeable toand gated by nitrate only.

Due to the strongly hyperpolarized membrane potential andthe outward-directed anion gradients of plant cells, channel-mediated anion transport appears outwardly directed. Xenopusoocytes contain high concentrations of chloride as the majorcytosolic anion. By contrast, nitrate does not represent a phys-iological anion in animal cells (Park et al., 2013). This raises thequestion about the nature of the charge carrier responsible forthe inward currents (negative currents/anion efflux) observed inSLAH2-expressing oocytes. To substantiate that SLAH2 medi-ates the release of nitrate but not chloride from oocytes, weused a nitrate-preloading approach. At a holding potential of0 mV, which is positive to the reversal potential of nitrate (Vrev),SLAH2-expressing oocytes were loaded with nitrate, while at2100 mV (VH < Vrev) nitrate loading is thermodynamically notsupported by the membrane potential. Following preloading at0 mV, positive as well as negative currents could be recordeddepending on the voltage applied (Figures 3A and 3C, comparedwith Figure 1C and 1D). When challenged with nitrate at 2100mV, however, inward currents (anion efflux) were not observed(Figures 3B and 3C). Note that inward currents at the tail voltageof2120 mV (third segment of the voltage protocol) were presentonly when membrane potentials positive of the Vrev were appliedwithin the preceding test voltage segment (second segment inthe voltage protocol). This indicates that nitrate loading is re-quired to evoke anion efflux. Thus, experiments employing oo-cytes preloaded with nitrate further support the notion thatSLAH2 represents a nitrate-selective anion channel.

To simulate plant cell–like nitrate levels in a heterologousoocyte model, we injected 5 nmol 15N-nitrate into the oocytes(resulting in a final nitrate concentration of ;10 mM). After a5-min incubation in bathing solution containing 10 mM externalnitrate, the amount of 15N-nitrate in the oocytes was determinedby isotope ratio mass spectrometry. Noninjected control oo-cytes as well as oocytes expressing SLAH2, SLAC1, or CPK21alone showed no release of the internal 15N-nitrate into the ex-tracellular space within 5 min (Supplemental Figure 4A). Coex-pression of SLAH2 or SLAC1 together with CPK21DEF resultedin a rapid efflux of 15N-nitrate, confirming that activation by

protein kinases is essential for the transport activity of bothanion channels. In the voltage clamp experiments describedabove (Figures 1C and 1D), we have shown that SLAH2-mediated currents depend on nitrate at the extracellular mem-brane face of oocytes. To test whether external NO3

2 affects effluxof 15N-nitrate in a similar manner, the 15N-nitrate tracer pool inoocytes was followed in the absence or presence of 10 mMexternal nitrate in the bath solution (Supplemental Figure 4B).Whereas SLAH2-expressing oocytes maintained 50% of the15N-nitrate load for 15 min in the absence of external NO3

2, inSLAC1-expressing oocytes (which do not require external NO3

2

for activity) release of 15N-nitrate under identical conditions wasobserved, with a 50% loss after just 5 min and >90% loss after15 min. By contrast, in the presence of external nitrate, the effluxrate of SLAH2-expressing oocytes was similarly high as foroocytes expressing SLAC1. Although the outward-directedchemical gradient was at its maximum in the absence of externalNO3

2, SLAH2-mediated nitrate efflux was accelerated in thepresence of external nitrate (Supplemental Figure 4B, comparedwith Figure 2A). The fact that 15N-nitrate injected, SLAH2-expressing oocytes did not transport nitrate, but required thepresence of external nitrate for channel opening, indicates thatthe anion acts on the extracellular face of SLAH2 only.

Homology Modeling of SLAC1 and SLAH2 to RevealPotential Sites for Nitrate/Chloride Selection

The crystal structure of Hi-TehA, a bacterial homolog of SLAC1,suggests that an inner five-helix transmembrane ring forms thepore with a central phenylalanine residue, which is invariant inthis superfamily and seemingly blocks the pore (Chen et al.,2010; Dreyer et al., 2012). Whereas wild-type Hi-TehA showedno ion transport when expressed in oocytes, mutation of theblocking phenylalanine residue (Phe-262) to alanine renderedthis channel constitutively active (Chen et al., 2010). Similarly,replacement of the pore phenylalanine (SLAC1 F450A) alsogated SLAC1 (constitutively) open even in the absence of anactivating kinase (Chen et al., 2010). However, the mutation didnot markedly alter the anion selectivity profile of SLAC1. Thus,the molecular determinants of anion selectivity within the pore ofSLAC/SLAH anion channels remained unknown.Using two S-type anion channels (SLAC1 and SLAH2) that

strongly differ in anion selectivity, we attempted to identify thepore forming residues involved in shaping the selectivity of theSLAC/SLAH-type channels. To first understand the potentialconstituents of the selectivity filter, 3D homology models ofSLAC1 and SLAH2 were generated based on the crystalstructure of Hi-TehA (Figure 4; PDB accession code 3M71; Chenet al., 2010).

Figure 2. (continued).

influence on the reversal potential (Vrev). By contrast, an increase in the external NO32 concentration shifted the reversal potential to more negative

voltages (n = 4 experiments, mean 6 SD).(F) Relative voltage-dependent open probabilities (rel. PO) of SLAH2- and CPK21-mediated currents under conditions used in Figure 2E. Rel. PO wascalculated from a 2120-mV voltage pulse following the test pulses in the voltage range +60 to 2200 mV in 20-mV decrements. Data points were fittedby a Boltzmann equation (continuous line; n = 4 experiments, mean 6 SD).

S-Type Anion Channel Selectivity 5 of 14

As both 3D models were built from the same template andbecause of the high sequence similarity between SLAC1 andSLAH2, with no insertion or deletions inside the pore region,both homology models of SLAC1 and SLAH2 (SupplementalFigure 5; Figures 4A and 4B) share highly similar pore geome-tries with a diameter of ;5 to 6 Å. The conserved side chain ofthe gating phenylalanines, Phe-450 or Phe-402, are located inthe middle of the pore, as described by Chen et al. (2010) for Hi-TehA and SLAC1. A detailed comparison of the pore regions,however, revealed differences between SLAC1 and both SLAH2and Hi-TehA. Besides the conserved gating phenylalanine, tworesidues above the phenylalanine in the outward facing vestibuleare conserved between SLAC1 (Phe-276 and Trp-404) andSLAH2 (Phe-231 Trp-356) and may contribute to pore blockageas well as rendering the SLAC1/SLAH2 vestibule on the extra-cellular face shallower than that of Hi-TehA (Supplemental

Figure 6). By contrast, the geometry and size of the intracellularfacing cavities of SLAC1 and SLAH2 and the bacterial homologHi-TehA are quite similar. Comparing the three channel proteins,most of the 29 pore-lining residues are either aliphatic or aro-matic, giving the pore a hydrophobic character (SupplementalFigure 5). The pore lining residues provide another clear dis-tinction between the plant anion channels SLAC1 and SLAH2and the bacterial homolog Hi-TehA, with 55% of the pore-liningresidues being identical between SLAC1 and SLAH2 (72%similarity) and only 14% identical between SLAC1 and Hi-TehA.In spite of the high similarity between the pore of SLAC1 andSLAH2, a clearly identifiable difference in anion binding site ascan be observed with the CLC channels that could not be foundin SLAH2 (Supplemental Figure 6). Furthermore, in the porecenter, only a few differences might account for the different ionselectivity observed. However, these changes are additionally

Figure 3. Anion Selectivity of SLAH2.

Whole-oocyte currents of oocytes coexpressing SLAH2, CIPK23, and CBL1 in standard medium containing 30 mM NO32. The experiment was

performed at two different holding potentials using the standard voltage protocol.(A) When the oocytes were clamped to 0 mV, single voltage pulses were applied starting from +60 to 2200 mV in 20-mV decrements. Representativecells are shown.(B) When the oocytes were clamped to 2100 mV, the voltage pulses increased from 2200 to +60 mV in 20-mV steps. Representative cells are shown.(C) Steady state currents (ISS) of SLAH2 and CIPK23/CBL1 coexpressing oocytes measured as in (A) (Vhold = 0 mV pos > neg) or (B) (Vhold = 2100 mVneg > pos) (n $ 5 experiments, mean 6 SD).

Figure 4. Structural Model of SLAC1 and SLAH2.

(A) and (B) Ribbon plot of the homology models of SLAC1 (cyan) and SLAH2 (magenta).(A) Channel seen from the extracellular side, with the side chains of the conserved Phe-450/402 and Phe-276/231 occluding the pore.(B) Same as in (A) but shown as a side view.

6 of 14 The Plant Cell

restrained by the replacement of amino acids with a similar sizeor only slightly different biochemical character.

Pore-Lining Phenylalanine Residues Are Involved in NitrateSelectivity of SLAH2

To examine whether Phe-402 in SLAH2 exhibits similar gatingproperties as Phe-450 in SLAC1 (Supplemental Figure 7Bcompared with Chen et al. [2010]) and Phe-262 in Hi-TehA, wereplaced this residue with alanine and expressed the mutantprotein in Xenopus oocytes. The mutant SLAH2 F402A indeedhad an open channel that no longer required a protein kinase foranion channel activity (Supplemental Figure 7A compared withChen et al. [2010]). Expression of SLAH2 F402A alone displayedmacroscopic anion currents similar to SLAH2 wild-type channelactivated by CIPK23/CBL1 (Figure 5A), but nitrate was never-theless required in the extracellular bath solution for ion con-duction (Supplemental Figure 7A).

In the homology models, another phenylalanine located ontransmembrane helix (TM) 3 in the SLAH2 and SLAC1 pore isin close proximity to the gating Phe (Figures 4A and 4B;Supplemental Figures 6A and 6B). Together with the gating Pheresidue, Phe-231 of SLAH2 and Phe-276 of SLAC1 might also

serve to constrict the pore and may have to also be displaced inorder to establish an anion conducting state (Figures 4A and 4B;Supplemental Figures 6A and 6B). Mutation of Phe-231 alone toAla in SLAH2 did not result in a per se F402A-like open channel.When SLAH2 F231A was coexpressed with CIPK23/CBL1,however, macroscopic chloride currents similar to wild-typeSLAC1 could be monitored in the absence of external nitrate(Supplemental Figure 7C compared with Figure 5B). To bettervalidate the permeability of wild-type SLAH2 and mutantsthereof for different anions, the chord conductance at +40 mVand 2120 mV was calculated. This measure is directly pro-portional to the permeability of a channel for a specific anion.When we compared SLAH2 F231A to wild-type SLAH2, thechloride chord conductance was markedly increased andreached 50% (at 2120 and +40 mV) of the SLAC1 wild-typechord conductance for chloride (Supplemental Figure 7E). Thisindicates that replacement of Phe-231 with alanine converts thenitrate-specific SLAH2 into a chloride-permeable SLAC1-likeanion channel (Supplemental Figure 7E). Mutation of the re-spective Phe in SLAC1 to Ala (F276A) also resulted in enhancedchloride permeability relative to nitrate (Supplemental Figures7D and 7E). These findings show that the replacement of Phe-231 with Ala in the SLAH2 pore results in a SLAC1-like anion

Figure 5. Mutations in TM3 Influence Ion Selectivity of SLAC1 and SLAH2.

Instantaneous currents (Iinst) of Xenopus oocytes in standard buffers containing different anions (gluconate, chloride, or nitrate) are plotted against theapplied voltage.(A) and (C) SLAH2 wild type and the mutant SLAH2C1(271-281).(B) and (D) SLAC1 wild type and the corresponding mutant. The channel mutants and kinases expressed in the oocytes are indicated in the figure (n$ 3experiments, mean 6 SE).

S-Type Anion Channel Selectivity 7 of 14

permeability with broader anion specificity. The latter SLAH2mutation was able to affect channel opening independently fromthe presence of extracellular nitrate. Thus, the molecular basisfor SLAH2 nitrate gating and selectivity appears to be associ-ated with the pore of SLAH2.

The C-Terminal Half of TM3 Determines S-TypeChannel Selectivity

Since both phenylalanine residues are conserved betweenSLAC1 and SLAH2, it is unlikely that these amino acids residuescould directly exert the different anion selectivity. Analysis of the3D models suggests that the side chains of the critical phenyl-alanines, Phe-450 and Phe-276 in SLAC1, and Phe-402 andPhe-231 in SLAH2, have to be displaced to allow anion per-meation, as has also been proposed for Phe-450 in SLAC1 andPhe-262 in HiTehA (Chen et al., 2010). Due to packing, sidechain displacement is restricted by neighboring pore-liningresidues and the helical backbone of the five inner TMs (Figures6A to 6C; Supplemental Figures 6A and 6B). When comparingthe environment of the critical phenylalanines in SLAC1 andSLAH2, the most prominent differences were found in the ex-tracellular half of TM3 (residues 226 to 236 in SLAH2 and resi-dues 271 to 281 in SLAC1; Figures 6A to 6C; SupplementalFigure 5). To study the effect of these residues, we exchangedcorresponding sites between SLAH2 and SLAC1 and expressedthe TM3 chimeras in Xenopus oocytes. The SLAC1 chimeracontaining amino acids 226 to 236 from SLAH2 was namedSLAC1 H2(226-236) and the respective SLAH2 version SLAH2C1(271-281). In the presence of chloride, two-electrode voltageclamp measurements revealed that the chord conductance ofthe SLAC1 chimera was decreased by 80% at 2120 mV and by60% at +40 mV compared with wild-type SLAC1 (Figures 5Dand 6F). This result can be explained by a change in the se-lectivity of SLAC1 or by a transfer of the nitrate activationproperties of SLAH2 to the SLAC1 mutant.

To discriminate between these two possibilities, we per-formed experiments in the presence of SLAH2 activating nitrateconcentrations and varied the chloride concentration at thesame time. In contrast to SLAC1 wild type (Geiger et al., 2011),the increase in the external chloride concentration from 3 to 90mM had no influence on the current amplitude (SupplementalFigure 8A); this result was comparable to the results obtained forSLAH2 wild type (Figure 2E). The reversal potential was also notaffected by the change in the chloride concentration (DVrev = 262.28 mV, n = 5 experiments, mean 6 SE) similarly to SLAH2 wildtype (DVrev = 0.75 6 2.65 mV, n = 4 experiments, mean 6 SE)(compared with Figure 2E). In the experiment performed withSLAC1 wild type under the same conditions, the reversal po-tential shifted by DVrev = 19.25 6 1.56 mV (n = 4 experiments,mean 6 SE) (compared with Geiger et al., 2011). Additionally,chloride had no influence on the open probability of SLAC1 H2(226-236) (Supplemental Figure 8B), again similar to SLAH2 wildtype (Figure 2F). By contrast, the 10-fold increase in the externalnitrate concentration enhanced the current amplitude of SLAC1H2(226-236) (Supplemental Figure 8A) and shifted the reversalpotential to more negative values (DVrev = 54.25 6 1.04 mV) (n =5 experiments, mean 6 SE) (compared with Figure 2E). Thus,

SLAC1 H2(226-236) was characterized by an increased nitrateselectivity. The chimera SLAH2 C1(271-281) resulted in SLAH2with SLAC1-like permeability, since macroscopic anion currentswere observed even without external nitrate (Figure 5C) and thechord conductance in chloride-containing media reached 30%of its nitrate conductance (Figure 6F). However, the conversionof the SLAH2 chimera toward SLAC1-like electrical propertieswas less pronounced than for the SLAC1 chimera. Remarkably,the exchange of the C-terminal half of TM3, which faces theextracellular side, was sufficient to convert the guard cellchannel SLAC1 into a SLAH2-like, strict nitrate-selective anionchannel. Although Phe-231 and Phe-402 in SLAH2 and Phe-276and Phe-450 in SLAC1 are conserved in both channels, ourfindings concerning the molecular nature of the selectivity filterof SLAC/SLAH anion channels strongly implicate the in-volvement of these pore-occluding phenylalanines. One mayhypothesize that the conformational freedom of the side chainsof SLAH2 is restricted by neighboring amino acids on TM3. Ourhomology models suggest that both of the phenylalanine resi-dues forming the gate, Phe-231 and Phe-402 in SLAH2 andPhe-276 and Phe-450 in SLAC1, are in close proximity in the 3Dmodels, with the shortest distance between the two aromaticrings measures ;3.4 Å, which would be too small to let anionspass. A possible mechanism for allowing anion conductionwould require a rotation or flip of the aromatic rings along thex2-torsion angle (Figures 6A to 6C). When focusing on the C-terminal half of TM3, which altered the properties of the chan-nels, three residues were identified from the models that mightinfluence the flexibility of Phe-231 in SLAH2 and Phe-276 inSLAC1. In SLAC1, a bulky methionine residue, Met-275, pressesagainst Phe-276, which together with residues on the TM3 lo-cated one turn up or down, Val-272 and Ile-279, could poten-tially restrict conformational rearrangement of the phenylalanineside chain. In the model of SLAH2, a similar smaller leucineresidue, Leu-230, packs against Phe-231 (Figures 6A to 6C).Thus, based on the homology modeling, the different packing ofthe Phe-231 side chain in SLAH2 compared with its counterpartPhe-276 in SLAC1 could possibly present the molecular causefor the distinct selectivity characteristics of both anion channels.To test this hypothesis, we exchanged Met-275 to Leu in SLAC1(M275L) and Leu-230 of SLAH2 with Met (L230M). Additionally,we created a triple mutant in which we exchanged all neigh-boring residues in SLAC1 and SLAH2 mentioned above thatmight have an influence on the conformational freedom of Phe-231 in SLAH2 and Phe-276 in SLAC1 (SLAH2 I227V L230ML234I and SLAC1 V272I M275L I279L). However, neither thesingle nor the triple mutants showed any differences in theirselectivity compared with the respective wild-type channels(Supplemental Figure 9).In search of further differences in the C-terminal half of TM3

that could alter the chemistry at this site, we found that SLAC1harbors a serine at position 280 in close proximity to the porewhile SLAH2 has a glycine at the respective position (Gly-235).The exchange of the polar Ser-280 to a nonpolar glycine inSLAC1, however, did not change the selectivity of the guard cellanion channel (Figure 6F). Similarly, the respective mutation inSLAH2 (G235S) did not change the nitrate-selective character ofSLAH2 either (Figure 6F). Another difference was observed at

8 of 14 The Plant Cell

Figure 6. Structural Differences of Transmembrane Helix 3 Determine the Selectivity Differences between SLAC1 and SLAH2.

(A) to (C) Structural differences in the pore region of SLAC1 and SLAH2. Close-up view of TM3 of SLAC1 (A), an overlay of SLAC1 and SLAH2 (B), andSLAH2 (C) with the side chains and TM4 of SLAC1 (carbon atoms in cyan) and SLAH2 (carbon atoms in magenta). Pore lining residues are facing left.The position of pore-blocking Phe-450 and Phe-402 on TM9 is shown.(D) and (E) Instantaneous currents (Iinst) of Xenopus oocytes coexpressing the mutants SLAH2 S228V (C) or SLAC1 V273S (D) with CIPK23/CBL1 instandard buffers containing different anions (gluconate, chloride, or nitrate) are plotted against the applied voltage (n $ 4 experiments, mean 6 SE).(F) Chord conductance at +40 and 2120 mV in standard solutions containing 100 mM Cl2. The chord conductance was calculated from the in-stantaneous currents in Cl2- as well as in NO3

2-containing buffers from oocytes coexpressing SLAH2 or SLAC1 wild type, or the different channelmutants, with CIPK23 and CBL1. The chord conductance was normalized to 100 mM NO3

2 (dashed line) (n $ 3 experiments, mean 6 SE).

position 228 in SLAH2 corresponding to residue 273 in SLAC1.While SLAH2 is equipped with a polar serine (Ser-228), SLAC1carries a nonpolar valine (Val-273) at the respective position.Interestingly, this single substitution of Ser-228 by valine wassufficient to convert SLAH2 into a SLAC1-like, chloride- andnitrate-permeable anion channel indicated by macroscopic an-ion currents in nitrate and chloride solutions (Figure 6D). Thechord conductance for chloride of SLAH2 S228V increased to60% of the observed NO3

2 conductance (Figure 6F). Thus, atleast in SLAH2, anion selectivity seems to be encoded locally bya single residue, which is in close proximity to the gating phe-nylalanines. However, performing the reverse amino acid ex-change by substituting Val-273 in SLAC1 to the respectiveserine residue as in SLAH2 did not convert SLAC1 into a NO3

2-selective channel (Figures 6E and 6F). Thus, although a singlemutation in SLAH2 was sufficient to yield an anion channel withSLAC1-like properties, in SLAC1 additional elements besidesthe nature of the amino acid at position 273 are involved inexerting anion selection, possibly indicating that the pore ge-ometry (despite the high sequence similarity between SLAC1and SLAH2) does slightly differ.

DISCUSSION

How does one build and operate an anion-selective channel?Our knowledge concerning the selectivity of animal anionchannels of the CLC type is well advanced (Dutzler et al., 2003).X-ray structures of two prokaryotic CLC Cl2 channel homologsfrom Salmonella enterica serovar typhimurium and Escherichiacoli at 3.0 and 3.5 Å, respectively (Dutzler et al., 2002), revealeda homodimeric membrane protein. Each subunit forms its ownindependent anion selectivity filter. Each anion binding sitestabilizes a single chloride ion at three sites within the highlyconstricted permeation pathway (Dutzler et al., 2003). Togetherwith specific neighboring residues, these binding sites de-termine the selectivity of bacterial CLCs. CLC channels aregated by the side chain carboxylate group of a central highlyconserved glutamate that closes the pore by mimicking a chlo-ride ion. The anion selectivity with the order chloride > bromide >nitrate > iodide found for animal CLC channels confirmed theexistence of a high-field-strength anion binding site (Wright andDiamond, 1977; Picollo et al., 2009).

Arabidopsis contains seven homologs, named CLC-a to CLC-g,which all potentially reside in intracellular membranes (Lv et al.,2009). The most thoroughly investigated plant CLC homologis CLC-a, which, like the bacterial counterpart, utilizes thepH gradient across the vacuolar membrane to transport nitrateinto the organelle (De Angeli et al., 2006; Bergsdorf et al., 2009).The high nitrate selectivity at its cytosolic side is consistent withthe nitrate-specific phenotype of clc-a knockout mutants (Wegeet al., 2010). Using sequence comparisons of different CLCs,a Ser residue was identified that is conserved among animal andbacterial CLC channels and proton-coupled cotransporters(Ser-168 in hCLC-5) but is replaced by a proline in At-CLC-a(Pro-160) (Bergsdorf et al., 2009; Wege et al., 2010). Thisserine contributes to the central binding site, and a substitutionof Ser-168 with proline in hCLC-5 increases the selectivity forNO3

2, such that the anion selectivity of the mutant was similar to

At-CLC-a (Bergsdorf et al., 2009; Wege et al., 2010). Addition-ally, this mutant showed NO3

2/H+-coupled transport reminis-cent to At-CLCa and not uncoupled transport as seen forhCLC-5 wild type. The corresponding “reverse” mutation in At-CLC-a (P160S) rendered the NO3

2/H+ transporter an NO32 and

Cl2 nonspecific H+-coupled transporter (Bergsdorf et al., 2009;Wege et al., 2010).In contrast to the distinct chloride binding site in the highly

constricted pore of animal/bacterial CLC channels, a similarlyidentifiable binding site for halide ions could, however, not bedetected in either SLAC1 or SLAH2. This is supported by therelative anion permeability sequence of nitrate $ iodide > bro-mide > chloride for SLAC1 and SLAH2 (Figure 2D) that corre-sponds to selectivity sequence 1 for anion-selective proteinsthat lack high-field-strength anion binding sites (Wright andDiamond, 1977).The structure of the bacterial Hi-TehA reported by Chen et al.

(2010) provided a molecular basis to analyze ion transport andselectivity for members of this class of anion channel proteins,which exhibit architecture clearly distinct from those of otherknown anion channels such as the CLCs, CFTR, or TMEM16A(Dutzler et al., 2002; Gadsby et al., 2006; Caputo et al., 2008;Schroeder et al., 2008; Yang et al., 2008; Zifarelli and Pusch,2010). This structure guided our functional analysis of the mo-lecular nature of the SLAC/SLAH selectivity filter.Replacing the gating phenylalanine (Phe-402 in SLAH2 and

Phe-450 in SLAC1) in the pore of SLAC1 and SLAH2 by alaninerendered both anion channels open in a kinase-independentfashion (Supplemental Figures 7A and 7B; compared with Chenet al., 2010). However, these mutants did not display markedlyaltered anion selectivity compared with the corresponding wildtype (Figures 5A and 5B). The homology models of SLAH2 andSLAC1 highlighted a second phenylalanine (Phe-231 in SLAH2,Phe-276 in SLAC1) located on TM3 and in close proximity toPhe-402 or Phe-450 on TM9 of SLAH2 and SLAC1, respectively.Both phenylalanine residues on TM3 and TM9 together seem toconstrict the pore. According to Chen et al. (2010), the sidechain of the gating phenylalanine (Phe-450 in SLAC1 and Phe-402 in SLAH2) is present in a high-energy conformation and hasto be displaced to unblock the pore. Because Phe-231 po-tentially restricts the mobility of Phe-450 (and vice versa), itis tempting to speculate that the side chain conformations ofboth phenylalanines rearrange in a concerted manner upon iontransport. Despite the fact that mutation of the Phe-402 inSLAH2 rendered the channel constitutively open, the SLAH2mutant F231A still required coexpression of an activating kinasefor ion translocation. However, the SLAH2 mutant F231A dis-played a SLAC1-like relative permeability and could not furtherdiscriminate between chloride and nitrate. This finding suggeststhat Phe-231 is likely be involved in ion selectivity (SupplementalFigure 7C). Moreover, the SLAH2 mutant F231A did not requirefurther extracellular nitrate as ligand when coexpressed with anactivating kinase in oocytes (Supplemental Figure 7C). This mayindicate that nitrate recognition originates from the pore itselfand is not based on an allosteric effect from a binding site on theextracellular surface of the SLAH2 protein. Thus, ligand bindingand anion conduction and selectivity seem to affect each other.Certain CLC channels operate on a similar basis as SLAH2 by

10 of 14 The Plant Cell

facilitating Cl–-activated Cl– transport (Pusch et al., 1995; Chenand Miller, 1996). Upon chloride conductance, chloride ionspositively modulate the open probability of the CLC channelgate. Here, a glutamate side chain gates the pore open uponbinding to a chloride ion within the selectivity filter of CLCchannels (Dutzler et al., 2002). Thus, selective conduction andgating is coupled within the pore.

In the vicinity of Phe-402/450 of SLAH2 and SLAC1, we un-covered prominent differences in the extracellular half of TM3(Figures 6A to 6C; Supplemental Figure 5). Exchanging thisstretch of amino acids between SLAH2 and SLAC1 convertedSLAC1 into a nitrate-selective anion channel with similar ionselectivity and transport capabilities as wild-type SLAH2. Viceversa, SLAH2 equipped with amino acid residues 271 to 281from SLAC1 provided SLAH2 with SLAC1-like anion channelproperties; however, a complete conversion of SLAH2 intoa chloride- and nitrate-permeable channel like SLAC1 could notbe obtained. This indicates that although the C-terminal half ofTM3 plays a major role as selectivity filter, it is not fully re-sponsible for determining the ion selection. Further inspection ofthe 3D homology models of SLAH2 and SLAC1 highlighted thepolar serine residue at position 228 in SLAH2 within the C-terminal half of TM3. When we exchanged Ser-228 for valine,which is the respective amino acid in SLAC1, SLAH2 gainedSLAC1-like chloride permeability and channel activation be-came independent of extracellular nitrate (Figures 6D and 6F). Inaddition, the movement of the side chain Phe-231 in SLAH2might be restricted by residues present in the C-terminal half ofTM3 in such a way that nitrate, but not chloride, can passthrough the anion channel pore. Thus, the peculiar selectivity ofSLAH2 might originate from the facilitated coordination of nitratebut not chloride at the polar serine residue 228 and may in-fluence the concerted mobility or conformational rearrange-ments of the gating Phe-402 and Phe-231.

SLAH2 Probably Evolved to Facilitate Root-ShootNitrate Transport

Our findings suggest that similar to SLAH3 and SLAC1, the in-teraction of SLAH2 with distinct protein kinases (compared withScherzer et al., 2012) renders the stele-expressed anion trans-porter active. We established that SLAH2 represents a nitrate-selective channel, demonstrating that such channels exist inliving systems. Its strict nitrate specificity distinguishes SLAH2from SLAC1 as well as from SLAH3; therefore, SLAH2 ex-pressed in the SLAC1 knockout mutant might not be capable ofrecovering the chloride release required for stomatal closure(Negi et al., 2008). Given this peculiar selectivity among theSLAC/SLAH family members, the strict nitrate selectivity is of noadvantage to guard cells located at the end of the transpirationstream, but would be highly advantageous in the root, wherewater and nutrients enter the transpiration stream and where theplant obtains soil nitrate and at times is exposed to high NaClload under salt stress. The SLAH2-type anion channels in roots,with a strict NO3

2 selectivity, would facilitate the preferentialtransport of nitrate, which is a major source of nitrogen forplants. Future physiological and electrophysiological studieswith SLAH2 loss-of-function mutants are needed to elucidate

whether these hypothesized physiological functions indeed ex-plain the evolution of this NO3

2-specific channel, SLAH2, inplants.

METHODS

Detection of SLAH2 Transcripts by Quantitative RT-PCR

Quantification of ACTIN2 and 8 as well as SLAH2 transcripts was per-formed by real-time PCR as described elsewhere (Ivashikina et al., 2005).The different tissues of 6- to 8-week-old Arabidopsis thaliana ecotypeColumbia (Col-0) plants were harvested for RT-PCR analysis. FollowingRNA extraction and cDNA synthesis, quantitative PCR experiments wereperformed as described by Ivashikina et al. (2005). Transcripts were eachnormalized to 10,000 molecules of ACTIN2/8. Primer sequences areprovided in Supplemental Table 1.

Cloning and cRNA Generation

The cDNAs of SLAH2, SLAC1, CPK21/31, CIPK23/24, and CBL1/4 werecloned into oocyte (BiFC) expression vectors (based on pGEM vectors,see Supplemental Table 2) by an advanced uracil-excision-based cloningtechnique as described (Nour-Eldin et al., 2006). Site-directed mutationswere introduced by means of a modified USER fusion method as de-scribed by Dadacz-Narloch et al. (2011). Primer sequences are provided inSupplemental Table 1. For functional analysis, cRNA was prepared withthe mMESSAGE mMACHINE T7 transcription kit (Ambion). Oocytepreparation and cRNA injection were performed as described (Beckeret al., 1996). For oocyte BiFC and electrophysiological experiments, 10 ngof SLAH2, SLAC1, or CPK cRNA and 2 ng CIPK or CBL cRNA wereinjected.

Oocyte Recordings

In double-electrode voltage clamp studies, oocytes were perfused withTris/MES buffers. The standard solution contained 10 mM Tris/MES (pH5.6), 1 mM Ca(gluconate)2, 1 mM Mg(gluconate)2, 100 mM NaNO3, and 1mMLaCl3. To balance the ionic strength, we compensated for the NO3

2 orCl2 variations with gluconate. Solutions for anion selectivity measure-ments were composed of 50 mM Cl2, HCO3

2, SO422, NO3

2, gluconate2,or malate2sodium salts; 1 mMCa(gluconate)2; 1 mMMg(gluconate)2; and10mMTris/MES (pH 7.5). Osmolality was adjusted to 220mosmol/kg withD-sorbitol. Standard voltage protocol was as follows: Starting froma holding potential (VH) of 0 mV, single-voltage pulses were applied in 20-mV decrements from +60 to 2200 mV. Steady state currents (ISS) wereextracted at the end of the 20-s test pulses. The relative open probabilityPO was determined from current responses to a constant voltage pulse to2120 mV subsequent to the test pulses. These currents were normalizedto the saturation value of the calculated Boltzmann distribution. The half-maximal activation potential (V1/2) and the apparent gating charge (z) weredetermined by fitting the experimental data points with a single Boltzmannequation. To avoid nitrate loading of the oocytes (Figure 3B), a holdingpotential of 2100 mV was used in combination with a modified standardvoltage protocol (20-mV steps from 2200 to +60 mV).

Instantaneous currents (Iinst) were extracted right after the voltage jumpfrom the holding potential of 0 mV to 50-ms test pulses ranging from +70to 2150 mV. To calculate the chord conductance, the reversal potential(Vrev) was determined by fitting the instantaneous currents surroundingVrev in either chloride- or nitrate-containing standard buffers with a linearfit. Using the instantaneous currents at –120 mV or +40 mV, the chordconductance at –120 mV and +40 mV, respectively, could be calculatedwith the equation ganion = Ianion/(V2 Vrev) as described previously (Hodgkin

S-Type Anion Channel Selectivity 11 of 14

and Huxley, 1952; Hodgkin et al., 1952). To avoid loading of oocytes withnitrate and to keep the cytosolic anion composition constant for correctreversal potential determination in response to nitrate concentrationchanges (Figure 2C) in the external medium or during anion selectivitymeasurements (Figure 2D), the reversal potentials were recorded in thecurrent-clamp modus.

BiFC Experiments

For documentation of the oocyte BiFC results, images were taken witha confocal laser scanning microscope (LSM 5 Pascal; Carl Zeiss Jena)equipped with a Zeiss Plan-Neofluar 203/0.5 objective. Images wereidentically processed (low-pass filtered and sharpened) with the image-acquisition software LSM 5 Pascal (Carl Zeiss).

Protein Purification and in Vitro Kinase Assays

The coding sequence for CPK21, as well as those for the SLAH2 N- andC-terminal domains, were cloned into the USER-compatible expressionvector pGEX 6P1-USER and transformed into Escherichia coli (DE3)pLysS strain (Novagen). Bacteria were grown to an OD of 0.5 to 0.8 at 600nm, and production of glutathione S-transferase fusion proteins wasinduced by 0.4 mM isopropylthio-b-galactoside for 16 h at 18°C. Cellswere harvested by centrifugation at 10,000g and then lysed in 50 mL PBSby five 15-s sonication steps. To remove insoluble bacterial material, thelysate was centrifuged for 30 min at 10,000g. Purification using Gluta-thione Sepharose 4B beads (Amersham Biosciences) was performed incolumns according to the manufacturer’s instructions. Tris-HCl (50 mM)with 10 mM reduced glutathione was used for protein elution. Lysis andwash buffers were composed of 140 mM NaCl, 2.7 mM KCl, 10 mMNa2HPO4, and 1.8 mM KH2PO4 (pH 7.3). Proteins were dialyzed andstored in 150 mM NaCl and 50 mM HEPES. In vitro kinase buffer wascomposed of 50mMHEPES (pH 7.5), 10mMMgCl2, 13 protease inhibitorcocktail (Roche), 2 mM DTT, 5 mM EGTA, and 5 mCi [g-32P]ATP (3000 Ci/mmol). In vitro kinase assays were performed in the presence of 1 µM freeCa2+. Reactions were performed for 10 min at room temperature and thenstopped by addition of 63 SDS loading buffer and subsequent heating to90°C for 5 min. Proteins were separated by SDS-gel electrophoresis withan 8 to 16% gradient acrylamide gel (Precise protein gel; Thermo Sci-entific). Gels were stained with Coomassie blue, and phosphorylatedproteins were determined by autoradiography with a Fujifilm BAS 2000scanner.

Quantification of 15N Abundance and Total Nitrogen Concentration

Xenopus laevis oocytes were injected with 5 nmol 15N-labeled NO32 per

oocyte in the absence of external nitrate. Injected oocytes were incubatedin standard buffers containing 0 or 10 mM nitrate. At different time points,single oocytes were transferred to tin capsules and oven-dried overnightat 80°C. Total nitrogen concentrations and 15N abundances were de-termined using an elemental analyzer (NC2500; CE Instruments), coupledvia a Conflo II interface to an isotope ratio mass spectrometer (Delta Plus;Thermo Finnigan MAT). Working standards (glutamic acid) were analyzedafter every twelfth sample to detect a potential instrument drift over time.Standards were calibrated against the primary standard USGS 41(d15NAir = 47.600) for d13N.

Detection of SLAH2 in Plants with the SLAH2-Promoter:GUS Construct

SLAH2-promoter:GUS plants (Negi et al., 2008) were cultivated on sterileSussman plates (1.2% agar and 1% sucrose, pH 5.7) under the followingconditions for 10 to 12 d: 12-h photoperiod with 250 mmol photons per

square meter per second and 22°C/16°C day/night temperature. GUShistochemical staining was performed according to Jefferson et al. (1987).Ten-micrometer cross sections of GUS-stained roots were prepared witha microtome (Leica RM2165) from roots embedded in Technovit 7100(Heraeus Kulzer). Staining of the endodermis was performed by in-cubation of the sections in Sudan Red 7B (0.1% [w/v] in PEG400/glycerol)for 1 h. Pictures were taken with a fluorescence microscope (BZ 8100;Keyence).

Homology Modeling of the Plant Anion Channel SLAC1 and SLAH2

The 3D homology models for the membrane region of Arabidopsis SLAC1(residues Leu-185 to Gln-501) and SLAH2 (Leu-140 to Arg-452) were builtusing the structure of wild-type TehA (PDB entry 3M71) fromHaemophilusinfluenzae (Hi-TehA consisting of residues Pro-6 to Leu-313) as startingtemplate. Residues in the crystal structure of Hi-TehA (PDB 3M71) withalternative side chain conformations were replaced by a single conformerby either using the conformer withmore than 50%occupancy or using theconformer A. Selenomethionine residues were replaced by regular me-thionines, water, and detergent molecules were removed from the tem-plate coordinates. Amino acid sequences of SLAC1, SLAH2, and SLAH3from different species and the sequence of Hi-TehA were aligned usingthe software ClustalW and the similarity matrices BLOSUM62 or Gonnet.The alignment was iteratively (and manually) optimized using informationfrom transmembrane helix prediction for the SLAC and SLAH sequencesand removing N- and C-terminal cytoplasmic sequences of the SLAC/SLAH members, for which no equivalent data are present in the Hi-TehAtemplate structure.

Using this alignment, residues of the template structure Hi-TehA (PDBentry 3M71) differing in the target anion channel proteins SLAC1 andSLAH2 were manually replaced using the module ProteinDesign of thesoftware package Quanta2008 (MSI Accelrys). These first models werethen refined by applying rotamer searches (Quanta2008 module Xbuild)for residues, which were exchanged. Whenever possible, the side chainconformer most abundant in the PDB databank was used for modelbuilding. Deletion and insertions into a transmembrane helix, which wouldlead to a shift in helix amphipacity, were moved to the beginning or end ofthe affected transmembrane segment. The initial models were then op-timized by a stepwise energyminimization using the Charmm27 force fieldand employing only geometrical energy terms. First, side chain atomswere restrained with a small harmonic potential (5 kcal mol21 Å22),backbone atoms were kept fixed, and hydrogen atoms were allowed tomove freely, then restraints on side chain atoms were lowered to 1 kcalmol21 Å22. In a final step, backbone atoms at positions where prolineresidues were introduced or replaced were similarly minimized bygradually lowering a harmonic potential applied to the proline residue aswell as to the N- and C-terminal neighbor residue. The final models exhibitgood geometry for the side chain as well as backbone atoms; no bad vander Waals contacts were observed.

Accession Numbers

Sequence data from this article can be found in the Arabidopsis GenomeInitiative or GenBank/EMBL databases under the following accessionnumbers: SLAC1 (At1g12480), SLAH2 (At4g27970), CPK21 (At4g04720),CPK31 (At4g04695), CIPK23 (At1g30270), CIPK24 (At5g35410), CBL1(At4g17615), and CBL4 (At5g24270).

Supplemental Data

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

Supplemental Figure 1. Localization of SLAH2 in Arabidopsisthaliana.

12 of 14 The Plant Cell

Supplemental Figure 2. BIFC Interaction Studies in Xenopus Oo-cytes.

Supplemental Figure 3. Specific Phosphorylation of SLAH2 NTerminus (NT) by CPK21 in the Presence of Radiolabeled [g-32P]ATP.

Supplemental Figure 4. Release of Cytosolic15N-Labeled Nitrate.

Supplemental Figure 5. Structure-Based Sequence Alignment ofSLAC1, SLAH2, and Hi-TehA.

Supplemental Figure 6. Stereo View of the Pore of SLAC1 andSLAH2.

Supplemental Figure 7. Selectivity Determination of SLAC1 andSLAH2 Mutants Expressed in Xenopus Oocytes.

Supplemental Figure 8. The Chimera SLAC1 H2(226-236) DisplayedSLAH2-Like Electrical Properties.

Supplemental Figure 9. Chord Conductance Measurements ofSLAC1 and SLAH2 Mutants.

Supplemental Table 1. Primers.

Supplemental Table 2. Constructs.

ACKNOWLEDGMENTS

This work was funded by Deutsche Forschungsgemeinschaft (GermanScience Foundation) grants within GK1342 “lipid signaling” to R.H., D.G.,and T.D.M., within the research group FOR964 to R.H. and withinHE1640/28-1 to R.H. and T.D.M. R.H., D.G., and K.A.S.A. were supportedby grants of the Strategic Technologies Program–National Science,Technology, and Innovation Plan, Saudi Arabia (http://nstip.kacst.edu.sa/NSTIPWCPortal/faces/NSTIPHomePage) Project 10-ENV1181-02. J.S.was financially supported by the European Social Fund and by the Ministryof Science, Research, and Arts Baden-Württemberg (http://cordis.europa.eu/baden-wuerttemberg/int-coop-structuralfunds_en.html). We thank KohIba (Kyushu University) for providing us SLAH2-promoter:GUS plants andJörg Kudla (University of Münster) for providing us CIPK24 and CBL4constructs. We thank Scott McAdam for critically reading the article.

AUTHOR CONTRIBUTIONS

T.M., P.A., K.A.S.A., T.D.M., H.R., R.H., and D.G. designed the research.T.M., C.L., S.H., S.S., M.P., and J.S. performed research. T.M., C.L.,S.H., P.A., T.D.M., H.R., and D.G. analyzed data. T.M., K.A.R., H.R., R.H.,T.D.M., and D.G. wrote the article.

Received March 30, 2014; revised May 10, 2014; accepted May 23,2014; published June 17, 2014.

REFERENCES

Accardi, A., and Picollo, A. (2010). CLC channels and transporters:proteins with borderline personalities. Biochim. Biophys. Acta 1798:1457–1464.

Becker, D., Dreyer, I., Hoth, S., Reid, J.D., Busch, H., Lehnen, M.,Palme, K., and Hedrich, R. (1996). Changes in voltage activation,Cs+ sensitivity, and ion permeability in H5 mutants of the plant K+

channel KAT1. Proc. Natl. Acad. Sci. USA 93: 8123–8128.Bergsdorf, E.Y., Zdebik, A.A., and Jentsch, T.J. (2009). Residues

important for nitrate/proton coupling in plant and mammalian CLCtransporters. J. Biol. Chem. 284: 11184–11193.

Brandt, B., Brodsky, D.E., Xue, S., Negi, J., Iba, K., Kangasjärvi, J.,Ghassemian, M., Stephan, A.B., Hu, H., and Schroeder, J.I.(2012). Reconstitution of abscisic acid activation of SLAC1 anionchannel by CPK6 and OST1 kinases and branched ABI1 PP2Cphosphatase action. Proc. Natl. Acad. Sci. USA 109: 10593–10598.

Caputo, A., Caci, E., Ferrera, L., Pedemonte, N., Barsanti, C.,Sondo, E., Pfeffer, U., Ravazzolo, R., Zegarra-Moran, O., andGalietta, L.J. (2008). TMEM16A, a membrane protein associatedwith calcium-dependent chloride channel activity. Science 322:590–594.

Chen, T.Y., and Hwang, T.C. (2008). CLC-0 and CFTR: chloridechannels evolved from transporters. Physiol. Rev. 88: 351–387.

Chen, T.Y., and Miller, C. (1996). Nonequilibrium gating and voltagedependence of the ClC-0 Cl- channel. J. Gen. Physiol. 108: 237–250.

Chen, Y.H., Hu, L., Punta, M., Bruni, R., Hillerich, B., Kloss, B.,Rost, B., Love, J., Siegelbaum, S.A., and Hendrickson, W.A.(2010). Homologue structure of the SLAC1 anion channel for closingstomata in leaves. Nature 467: 1074–1080.

Dadacz-Narloch, B., Beyhl, D., Larisch, C., López-Sanjurjo, E.J.,Reski, R., Kuchitsu, K., Müller, T.D., Becker, D., Schönknecht,G., and Hedrich, R. (2011). A novel calcium binding site in the slowvacuolar cation channel TPC1 senses luminal calcium levels. PlantCell 23: 2696–2707.

De Angeli, A., Monachello, D., Ephritikhine, G., Frachisse, J.M.,Thomine, S., Gambale, F., and Barbier-Brygoo, H. (2006). Thenitrate/proton antiporter AtCLCa mediates nitrate accumulation inplant vacuoles. Nature 442: 939–942.

Demir, F., Horntrich, C., Blachutzik, J.O., Scherzer, S., Reinders, Y.,Kierszniowska, S., Schulze, W.X., Harms, G.S., Hedrich, R., Geiger, D.,and Kreuzer, I. (2013). Arabidopsis nanodomain-delimited ABA signalingpathway regulates the anion channel SLAH3. Proc. Natl. Acad. Sci. USA110: 8296–8301.

Dreyer, I., Gomez-Porras, J.L., Riaño-Pachón, D.M., Hedrich, R.,and Geiger, D. (2012). Molecular evolution of slow and quick anionchannels (SLACs and QUACs/ALMTs). Front. Plant Sci. 3: 263.

Dutzler, R., Campbell, E.B., and MacKinnon, R. (2003). Gating theselectivity filter in ClC chloride channels. Science 300: 108–112.

Dutzler, R., Campbell, E.B., Cadene, M., Chait, B.T., and MacKinnon,R. (2002). X-ray structure of a ClC chloride channel at 3.0 A revealsthe molecular basis of anion selectivity. Nature 415: 287–294.

Franz, S., Ehlert, B., Liese, A., Kurth, J., Cazalé, A.C., and Romeis, T.(2011). Calcium-dependent protein kinase CPK21 functions in abioticstress response in Arabidopsis thaliana. Mol. Plant 4: 83–96.

Gadsby, D.C., Vergani, P., and Csanády, L. (2006). The ABC proteinturned chloride channel whose failure causes cystic fibrosis. Nature440: 477–483.

Geiger, D., Becker, D., Vosloh, D., Gambale, F., Palme, K., Rehers,M., Anschuetz, U., Dreyer, I., Kudla, J., and Hedrich, R. (2009a).Heteromeric AtKC1/AKT1 channels in Arabidopsis roots facilitategrowth under K+-limiting conditions. J. Biol. Chem. 284: 21288–21295.

Geiger, D., Maierhofer, T., Al-Rasheid, K.A., Scherzer, S., Mumm, P.,Liese, A., Ache, P., Wellmann, C., Marten, I., Grill, E., Romeis, T., andHedrich, R. (2011). Stomatal closure by fast abscisic acid signaling ismediated by the guard cell anion channel SLAH3 and the receptorRCAR1. Sci. Signal. 4: ra32.

Geiger, D., Scherzer, S., Mumm, P., Marten, I., Ache, P., Matschi,S., Liese, A., Wellmann, C., Al-Rasheid, K.A., Grill, E., Romeis, T.,and Hedrich, R. (2010). Guard cell anion channel SLAC1 isregulated by CDPK protein kinases with distinct Ca2+ affinities.Proc. Natl. Acad. Sci. USA 107: 8023–8028.

Geiger, D., Scherzer, S., Mumm, P., Stange, A., Marten, I., Bauer, H.,Ache, P., Matschi, S., Liese, A., Al-Rasheid, K.A., Romeis, T., and

S-Type Anion Channel Selectivity 13 of 14

Hedrich, R. (2009b). Activity of guard cell anion channel SLAC1 iscontrolled by drought-stress signaling kinase-phosphatase pair. Proc.Natl. Acad. Sci. USA 106: 21425–21430.

Harmon, A.C., Yoo, B.C., and McCaffery, C. (1994). Pseudosubstrateinhibition of CDPK, a protein kinase with a calmodulin-like domain.Biochemistry 33: 7278–7287.

Harper, J.F., Huang, J.F., and Lloyd, S.J. (1994). Genetic identificationof an autoinhibitor in CDPK, a protein kinase with a calmodulin-likedomain. Biochemistry 33: 7267–7277.

Hedrich, R. (2012). Ion channels in plants. Physiol. Rev. 92: 1777–1811.

Ho, C.H., and Tsay, Y.F. (2010). Nitrate, ammonium, and potassiumsensing and signaling. Curr. Opin. Plant Biol. 13: 604–610.

Ho, C.H., Lin, S.H., Hu, H.C., and Tsay, Y.F. (2009). CHL1 functionsas a nitrate sensor in plants. Cell 138: 1184–1194.

Hodgkin, A.L., and Huxley, A.F. (1952). The components of membraneconductance in the giant axon of Loligo. J. Physiol. 116: 473–496.

Hodgkin, A.L., Huxley, A.F., and Katz, B. (1952). Measurement ofcurrent-voltage relations in the membrane of the giant axon ofLoligo. J. Physiol. 116: 424–448.

Ivashikina, N., Deeken, R., Fischer, S., Ache, P., and Hedrich, R.(2005). AKT2/3 subunits render guard cell K+ channels Ca2+ sensitive.J. Gen. Physiol. 125: 483–492.

Jefferson, R.A., Kavanagh, T.A., and Bevan, M.W. (1987). GUSfusions: beta-glucuronidase as a sensitive and versatile gene fusionmarker in higher plants. EMBO J. 6: 3901–3907.

Jentsch, T.J. (2008). CLC chloride channels and transporters: fromgenes to protein structure, pathology and physiology. Crit. Rev.Biochem. Mol. Biol. 43: 3–36.

Köhler, B., and Raschke, K. (2000). The delivery of salts to the xylem.Three types of anion conductance in the plasmalemma of the xylemparenchyma of roots of barley. Plant Physiol. 122: 243–254.

Köhler, B., Wegner, L.H., Osipov, V., and Raschke, K. (2002).Loading of nitrate into the xylem: apoplastic nitrate controls thevoltage dependence of X-QUAC, the main anion conductance inxylem-parenchyma cells of barley roots. Plant J. 30: 133–142.

Lin, S.H., Kuo, H.F., Canivenc, G., Lin, C.S., Lepetit, M., Hsu, P.K.,Tillard, P., Lin, H.L., Wang, Y.Y., Tsai, C.B., Gojon, A., and Tsay, Y.F.(2008). Mutation of the Arabidopsis NRT1.5 nitrate transporter causesdefective root-to-shoot nitrate transport. Plant Cell 20: 2514–2528.

Lv, Q.D., Tang, R.J., Liu, H., Gao, X.S., Li, Y.Z., Zheng, H.Q., andZhang, H.X. (2009). Cloning and molecular analyses of the Arabidopsisthaliana chloride channel gene family. Plant Sci. 176: 650–661.

Marten, H., Hedrich, R., and Roelfsema, M.R. (2007). Blue lightinhibits guard cell plasma membrane anion channels in a phototropin-dependent manner. Plant J. 50: 29–39.

Negi, J., Matsuda, O., Nagasawa, T., Oba, Y., Takahashi, H.,Kawai-Yamada, M., Uchimiya, H., Hashimoto, M., and Iba, K.(2008). CO2 regulator SLAC1 and its homologues are essential foranion homeostasis in plant cells. Nature 452: 483–486.

Nour-Eldin, H.H., Hansen, B.G., Nørholm, M.H., Jensen, J.K., andHalkier, B.A. (2006). Advancing uracil-excision based cloning towards

an ideal technique for cloning PCR fragments. Nucleic Acids Res.34: e122.

Park, J.W., Piknova, B., Huang, P.L., Noguchi, C.T., and Schechter, A.N.(2013). Effect of blood nitrite and nitrate levels on murine platelet function.PLoS ONE 8: e55699.

Picollo, A., Malvezzi, M., Houtman, J.C., and Accardi, A. (2009).Basis of substrate binding and conservation of selectivity in theCLC family of channels and transporters. Nat. Struct. Mol. Biol. 16:1294–1301.

Pusch, M., Ludewig, U., Rehfeldt, A., and Jentsch, T.J. (1995).Gating of the voltage-dependent chloride channel CIC-0 by thepermeant anion. Nature 373: 527–531.

Qiu, Q.S., Guo, Y., Dietrich, M.A., Schumaker, K.S., and Zhu, J.K.(2002). Regulation of SOS1, a plasma membrane Na+/H+ exchangerin Arabidopsis thaliana, by SOS2 and SOS3. Proc. Natl. Acad. Sci.USA 99: 8436–8441.

Scherzer, S., Maierhofer, T., Al-Rasheid, K.A., Geiger, D., andHedrich, R. (2012). Multiple calcium-dependent kinases modulateABA-activated guard cell anion channels. Mol. Plant 5: 1409–1412.

Schmidt, C., and Schroeder, J.I. (1994). Anion selectivity of slowanion channels in the plasma membrane of guard cells (large nitratepermeability). Plant Physiol. 106: 383–391.

Schroeder, B.C., Cheng, T., Jan, Y.N., and Jan, L.Y. (2008).Expression cloning of TMEM16A as a calcium-activated chloridechannel subunit. Cell 134: 1019–1029.

Schroeder, J.I., and Keller, B.U. (1992). Two types of anion channelcurrents in guard cells with distinct voltage regulation. Proc. Natl.Acad. Sci. USA 89: 5025–5029.

Stauber, T., Weinert, S., and Jentsch, T.J. (2012). Cell biology andphysiology of CLC chloride channels and transporters. Compr.Physiol. 2: 1701–1744.

Taylor, D.E., Hou, Y., Turner, R.J., and Weiner, J.H. (1994). Locationof a potassium tellurite resistance operon (tehA tehB) within theterminus of Escherichia coli K-12. J. Bacteriol. 176: 2740–2742.

Vahisalu, T., Kollist, H., Wang, Y.F., Nishimura, N., Chan, W.Y.,Valerio, G., Lamminmäki, A., Brosché, M., Moldau, H., Desikan, R.,Schroeder, J.I., and Kangasjärvi, J. (2008). SLAC1 is required forplant guard cell S-type anion channel function in stomatal signalling.Nature 452: 487–491.

Walter, E.G., Weiner, J.H., and Taylor, D.E. (1991). Nucleotidesequence and overexpression of the tellurite-resistance determinantfrom the IncHII plasmid pHH1508a. Gene 101: 1–7.

Wege, S., Jossier, M., Filleur, S., Thomine, S., Barbier-Brygoo, H.,Gambale, F., and De Angeli, A. (2010). The proline 160 in theselectivity filter of the Arabidopsis NO(3)(-)/H(+) exchanger AtCLCais essential for nitrate accumulation in planta. Plant J. 63: 861–869.

Wright, E.M., and Diamond, J.M. (1977). Anion selectivity in biologicalsystems. Physiol. Rev. 57: 109–156.

Yang, Y.D., et al. (2008). TMEM16A confers receptor-activatedcalcium-dependent chloride conductance. Nature 455: 1210–1215.

Zifarelli, G., and Pusch, M. (2010). CLC transport proteins in plants.FEBS Lett. 584: 2122–2127.

14 of 14 The Plant Cell

DOI 10.1105/tpc.114.125849; originally published online June 17, 2014;Plant Cell

Dietmar GeigerKhaled A.S. Al-Rasheid, Peter Ache, Heinz Rennenberg, Rainer Hedrich, Thomas D. Müller and

Tobias Maierhofer, Christof Lind, Stefanie Hüttl, Sönke Scherzer, Melanie Papenfuß, Judy Simon,Selective

Root Anion Channel SLAH2 Highly NitrateArabidopsisA Single-Pore Residue Renders the

 This information is current as of June 19, 2014

 

Supplemental Data http://www.plantcell.org/content/suppl/2014/06/09/tpc.114.125849.DC1.html

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists