The K+ transporter NPF7.3/NRT1.5 and the proton pump AHA2 contribute to K+ transport in Arabidopsis thaliana under K+ and NO3 - deficiency

Nitrate (NO3 -) and potassium (K+) are distributed in plants via short and long-distance transport. These two pathways jointly regulate NO3 - and K+ levels in all higher plants. The Arabidopsis thaliana transporter NPF7.3/NRT1.5 is responsible for loading NO3 - and K+ from root pericycle cells into the xylem vessels, facilitating the long-distance transport of NO3 - and K+ to shoots. In this study, we demonstrate a protein-protein interaction of NPF7.3/NRT1.5 with the proton pump AHA2 in the plasma membrane by split ubiquitin and bimolecular complementation assays, and we show that a conserved glycine residue in a transmembrane domain of NPF7.3/NRT1.5 is crucial for the interaction. We demonstrate that AHA2 together with NRT1.5 affects the K+ level in shoots, modulates the root architecture, and alters extracellular pH and the plasma membrane potential. We hypothesize that NRT1.5 and AHA2 interaction plays a role in maintaining the pH gradient and membrane potential across the root pericycle cell plasma membrane during K+ and/or NO3 - transport.


Introduction
Potassium (K) and nitrogen (N) are essential macronutrients of higher plants.Shortage or excess of these nutrients in soils limit growth and agricultural yield.While potassium is available to plants only as the cation K + , N occurs in soils as inorganic and organic molecules, where nitrate (NO 3 -) is the prevalent form (Xu et al., 2012).For nutrient uptake by the roots and distribution throughout the sporophyte via xylem and phloem, plants have evolved a large number of membrane transporters and channels with different specificities and affinity ranges.In Arabidopsis thaliana NO 3 -transporters and channels are encoded by four gene families (reviewed by Krapp et al., 2014): (1) the NRT1/ PTR (nitrate transporter 1/peptide transporter) family, renamed as NPF family, (2) the NRT2/NNP (nitrate transporter 2/nitrate-nitrite porter) family, (3) the CLC (chloride channel) family, and (4) the SLAC1/SLAHs (slow-type anion channel-associated 1/homologues) family.K + transport is accomplished by the three channel families Shaker, TPK (tandem-pore K + ) and Kir-like (K + inward rectifierlike) (Very and Sentenac, 2003), and the three transporter families KUP/HAK/KT, HKT (high-affinity K + transporters) and CPA (cation-proton antiporters) (Gierth and Maser, 2007;Chanroj et al., 2012).
The intersection between K and N physiology in plant cells is not yet fully understood.K + plays an essential role as a counter ion of NO 3 -, facilitating the uptake, translocation, and distribution of these ions between roots and shoots (reviewed in Rodenas et al., 2017 ).In contrast, as plant K + transporters and channels can also transport ammonium ions (NH 4 + ), K + and NH 4 + compete for uptake and translocation (Ten Hoopen et al., 2010).In addition, the expression of several NRT1 members (i.e.NRT1.1, NRT1.4,NRT1.5, NRT1.8) is influenced directly or indirectly by K + nutrition (Drechsler et al., 2015;Li et al., 2017;Fang et al., 2020;Morales de Los Rıós et al., 2021), indicating regulatory interactions between the K + and NO 3 -transport pathways.
The nitrate transporter NPF7.3/NRT1.5 (subsequently termed NRT1.5) was initially characterized as a proton-coupled, lowaffinity nitrate transporter involved in long-distance NO 3 transport in A. thaliana (Lin et al., 2008).It is expressed in root pericycle cells close to the xylem vessels and localizes to the plasma membrane (PM), suggesting its involvement in xylem loading of nitrate.Subsequently, the expression of NRT1.5 was found to be dependent both on nitrate and potassium availability.Under low nitrate nutrition, in nrt1.5 mutants the root-to-shoot potassium transport is reduced, suggesting a regulatory loop at the level of xylem transport that maintains the balance between nitrate and potassium (Drechsler et al., 2015).A direct link of NRT1.5 to NO 3 and K + homeostasis is supported by several observations.(1) Both NRT1.5 and the K + transporter SKOR contribute to K + translocation from root to shoot, with SKOR activity dominating under high NO 3 -and low K + supply, and NRT1.5 under low NO 3 availability (Drechsler et al., 2015).
Besides transporting nitrate and potassium, NRT1.5 was shown to function as an indole-3-butyric acid (IBA) transporter in A. thaliana root tips, where IBA is converted to the auxin indole-3acetic acid that regulates root gravitropism (Watanabe et al., 2020).NRT1.5 was also reported to be an important component in the regulation and modulation of phosphate deficiency responses (Cui et al., 2019).
NRT1.5 is energized by the H + -gradient across the PM, which is established by H + -ATPases that extrude protons from the cytoplasm by ATP hydrolysis.In A. thaliana 11 different members of PM H + -ATPases are encoded by the AHA (Arabidopsis H + -ATPase) multigene family, with AHA1 and AHA2 being the two major isoforms (Falhof et al., 2016).Proton pumps are known to contribute to nutrient assimilation during root uptake and short and long-distance transport (Sze and Chanroj, 2018;Feng et al., 2020).
The complex regulation of NRT1.5, its dependence on an electrochemical H + -gradient across the PM, and its ability to transport diverse substrates prompted us to investigate whether NRT1.5 interacts with other PM proteins that might directly or indirectly influence its activity and regulation.In a split-ubiquitinscreen, we discovered the proton pump AHA2 to interact with NRT1.5 and confirmed the interaction by BiFC analysis in tobacco plants.The physiological responses of Arabidopsis wild-type plants and nrt1.5 and aha2 mutant plants under different NO 3 -and K + regimes corroborated an interplay between NRT1.5 and AHA2 to control plant K + and H + homeostasis.
2 Materials and methods

Plant material and growth conditions
Arabidopsis thaliana T-DNA insertion lines were obtained from the Arabidopsis Biological Resource Center.The knockout nrt1.5 (GABI_347B03) and aha2 (GABI_219D04) mutant lines are in the Col-0 background.The double mutant nrt1.5/aha2was selected from the F2 progeny of the nrt1.5 × aha2 cross.The absence of fulllength NRT1.5 and AHA2 transcripts in the homozygous mutants was verified by RT-PCR (all oligonucleotides used in this work are listed in Supplementary Table 1).
For growth assays on soil, plants were pre-germinated on a 1:1 mixture of commercial type P soil and unfertilized type 0 soil (Einheitserdewerk Uetersen) under long-day growth conditions (16 h/ 8 h light/dark, 21°C/18°C day/night, light intensity 120 µmol m -2 s -1 , relative humidity 55% to 70%).After 5 days, the seedlings were transferred to either the same soil as control or to unfertilized soil supplemented with different nutrient solutions (Supplementary Table 2).
At the end of each fertilization experiment the frequency of chlorotic leaves (chlorophyll degradation/cell death) was determined for each genotype in 20 individual plants.

DNA isolation
Frozen Arabidopsis thaliana tissue was ground in liquid N 2 , mixed, and homogenized together with 2 metal balls by a mixer mill (Retsch MM 400).Genomic DNA from A. thaliana plants was extracted from frozen plant material according to Chen and Ronald (1999) with modification.600 µl of 2% CTAB buffer (100 mM Tris-Cl pH 8, 1.4 M NaCl, 20 mM EDTA, and 2% CTAB w/v) was added to the ground material and mixed by vortexing, followed by incubation at 65°C for 20 min.The tubes were inverted every 5 min and were centrifuged at room temperature for 10 min at 13000 g.The supernatant was transferred to a fresh tube with one volume chloroform:isoamyl alcohol (24:1) and mixed well by inverting, followed by centrifuging at 13000 g for 5 min.The upper phase was transferred to a fresh tube and mixed with one volume isopropanol, followed by incubating at 4°C for 10 min.
Precipitated DNA was obtained by centrifuging at 13000 g for 10 min.The pellet was washed with 70% EtOH, centrifuged at 13000 g for 3 min at room temperature, air-dried, re-suspended in 200 µl of H 2 O with 20 µg/ml RNaseA and incubated at 50°C for 30 min.

Split ubiquitin screen
Total RNA was extracted from Arabidopsis thaliana Col-0 seedlings 15 days after sowing and from senescing rosette leaves 43 days after sowing as described (Downing et al., 1992), followed by purification of polyadenylated mRNA using oligo (dT)-cellulose Type 7 (Amersham Biosciences).cDNA synthesis, entry cDNA library construction, and cloning into pDONR222 were performed with the CloneMiner ™ Library Construction Kit (ThermoFisher Scientific).The prey library for expression of NubG-cDNA fusion proteins in yeast was generated in pNubG-X (TAIR stock CD3-1737), a modified pNXgate32-3HA vector (Obrdlik et al., 2004).The entry cDNA library was transferred into the attR-sites of pNubG-X using the LR Clonase ™ enzyme mix (ThermoFisher Scientific).Entry cDNA library and prey library plasmid DNA was isolated using the QIAfilter Plasmid Mega Kit (Qiagen).

Complementation of nrt1.5 plants with NRT1.5 and NRT1.5 G209E
The eGFP coding sequence lacking the initial ATG codon was amplified with overhangs to be cloned by Gibson assembly into pTkan+-PHO1NRT1.5vector (Drechsler et al., 2015).The stop codon TAA of NRT1.5 was removed by PCR.The pTkan+-PHO1NRT1.5 vector was opened by restriction enzyme digestion with PstI and ligated with the eGFP fragment using the NEBuilder HiFi DNA assembly kit (New England Biolabs).The pTkan+-PHO1NRT1.5 G209E eGFP vector was constructed from pTkan+-PHO1NRT1.5eGFP using the Q5 Site-Directed Mutagenesis Kit (New England Biolabs).
The constructs were transformed in nrt1.5 plants by the floral dip method (Clough and Bent, 1998).Two independently transformed PHO1p::NRT1.5eGFPand PHO1p::NRT1.5G209E eGFP lines in nrt1.5 background that were homozygous for a single insertion of the transgene were used for this study.

Transient expression of NRT1.5eGFP and NRT1.5 G209E eGFP in Nicotiana benthamiana
The coding sequences of eGFP, NRT1.5eGFP and NRT1.5 G209E eGFP were amplified by PCR with the restriction enzyme attachment sites of BamHI and HindIII using pTkan+-PHO1NRT1.5eGFPand pTkan+-PHO1NRT1.5 G209E eGFP as templates.The amplified products were ligated into Agrobacterium binary vector pCAMBIA containing the minimal CaMV 35S promoter sequence (Hey and Grimm, 2018).Agrobacterium tumefaciens and Nicotiana benthamiana transformations were performed as described for the BiFC assay.

RNA isolation, cDNA synthesis and qPCR
Total RNA from frozen Arabidopsis thaliana seedlings was extracted using the citric acid extraction method (Oñate-Sańchez and Vicente-Carbajosa, 2008) and DNase I treated according to the manufacturer's instructions (New England Biolabs).First-strand cDNA was synthesized from 2 µg total RNA using RevertAid H Minus Reverse Transcriptase (ThermoFisher Scientific).Transcription analyses were performed with tissue from separately grown plants in three independent replicates.qPCR reactions were conducted on a CFX Connect real-time PCR detection system (Bio-Rad Laboratories) using the SYBR Green PCR Master Mix (ThermoFisher Scientific) and the thermal profile: 10 min 95°C and 40 times (15 sec 95°C, 1 min 60°C).For the subsequent melting curve analysis, the temperature was increased from 65 °C to 95 °C within 30 min.Expression values for each gene were normalized to the reference gene UBQ10 (At4g05320), and relative expression levels were determined according to (Czechowski et al., 2005).

Extracellular acidification assay
Extracellular acidification was assayed as described by Haruta et al. (2010) with modifications.Seedlings were pre-germinated in 0.5 × MS medium for five days and then transferred to unbuffered 0.25 × MS-base containing either HK (10 mM K + , 10 mM NO 3 -) or 0K (0 mM K + , 1 mM NO 3 -) fertilizer for another five days.The tenday-old seedlings were incubated in 200 µl modified 0.25 × MS media lacking MES (pH 5.5, unbuffered) with 30 µg/ml fluorescein isothiocyanate-dextran (FITC-dextran) (M r 10,000, Sigma-Aldrich) for 16 h in long-day conditions (16 h/8 h light/dark, 22°C, light intensity 120 µmol m -2 s -1 , relative humidity 40%).Excitation was at 485 nm and fluorescence was detected at 535 nm in a Power-Wave HT microplate spectrophotometer (BioTek).For pH determination, a calibration curve was generated with commercial MS media adjusted to pH values ranging from 4.5 to 7.5.

Live-cell imaging using confocal laser scanning microscopy
Confocal laser scanning microscopy (CLSM) of abaxial tobacco leaf epidermis cells was performed with a Leica TCS SP5 AOBS Confocal Microscope equipped with 20 x and 63 x water immersion objectives (Leica Microsystems).Simultaneous excitation of YFP (l ex = 488 nm, l em = 520-553 nm), mRFP (l ex = 561 nm, l em = 608-642 nm) and chlorophyll (l ex = 561 nm, l em = 642-702 nm) was achieved.The 514 nm argon laser was used to detect YFP and the 561 nm diode-pumped solid-state laser for mRFP and chlorophyll.Post-acquisition image processing and fluorescence quantification was performed using the Leica LAS AF software.

Results
3.1 The K + /NO 3 -transporter NPF7.3/NRT1.5 interacts with the proton pump AHA2 To investigate whether NRT1.5 interacts with other PM proteins, we carried out a split-ubiquitin membrane yeast two-hybrid screen (Stagljar et al., 1998;Thaminy et al., 2003).As AtNRT1.5 is highly expressed in seedling roots and senescing leaves (van der Graaff et al., 2006;Lin et al., 2008;Zheng et al., 2016), we used as prey two A. thaliana cDNA libraries from these tissues and screened them with two bait constructs carrying the Cub-LexA-VP16 reporter module at the Nor C-terminus of NRT1.5, respectively.From the primary NRT1.5-interactorcandidates we selected proteins that are known to be localized in the PM, to be expressed in pericycle cells of Arabidopsis roots, and to be associated with ion-or protein-transport.To confirm the interaction with NRT1.5, the full-length coding sequences of the candidates were amplified from A. thaliana Col-0 cDNA, fused to NubG, and used as prey for the NRT1.5 bait in targeted split-ubiquitin assays.Five putative NRT1.5-interactors were confirmed in this assay (Supplementary Table 3).
Among them was the PM proton ATPase 2 (AHA2) whose interaction with NRT1.5 we investigated in more detail in this study.Co-expression of NRT1.5-Cub and NubG-AHA2 in the auxotrophic yeast strain THY.AP4 led to yeast growth in a medium lacking adenine and histidine, indicating that NRT1.5 directly interacts with AHA2 (Figure 1A).
To evaluate the protein-protein interaction in plant cells, a bimolecular fluorescence complementation assay was carried out in Nicotiana benthamiana cells.Co-expression of the cYFP-NRT1.5 and AHA2-nYFP fusion proteins reconstituted YFP fluorescence, suggesting the interaction of NRT1.5 and AHA2 at the plant PM (Figure 1B).

3.2
The Gly 209 of NTR1.5 is critical for AHA2 interaction and affects NRT1.5 function but not its cellular localization in plants Li and colleagues (2017) had discovered that the low-K +sensitive A. thaliana mutant lks2 carries a Gly 209 to Glu (G209E) Protein-protein interaction of NRT1.Frontiers in Plant Science frontiersin.orgsubstitution in NRT1.5 and that the NRT1.5 G209E mutant protein is compromised in K + efflux in Xenopus laevis oocytes and in yeast cells.We observed that the mutant protein does neither interact with AHA2 in the yeast assay (Figure 1A) or in tobacco cells (Figure 1B).This corroborates the specificity of the NRT1.5-AHA2interaction and demonstrates that Gly 209 of NRT1.5 is critical for the interaction with AHA2.
Previously it was shown that, as a consequence of an impaired K + root-to-shoot transport, the nrt1.5 mutant displays impaired root architecture and develops an early senescence phenotype when plants grow in media lacking potassium and nitrate (Drechsler et al., 2015;Zheng et al., 2016).Additionally, nrt1.5 seedlings exhibit a reduced Hygromycin B (HygB) sensitivity due to a reduction in the PM potential (Drechsler et al., 2015).
The NRT1.5 Gly 209 residue is highly conserved in NRT1.5 homologs in other plant species as well as in the 16 NRT1 family members in A. thaliana (Supplementary Figure 3).To evaluate if NRT1.5 G209E expression in root cells affects NRT1.5 function, we carried out a complementation assay of the nrt1.5 mutant with NRT1.5 G209E .nrt1.5 mutant plants were transformed with a vector containing the promoter of the PHOSPHATE1 (PHO1) gene, and the NRT1.5 or NRT1.5 G209E coding region lacking the stop codon fused to a C-terminal eGFP fluorescent tag.
The 1.6 kb PHO1 promoter (PHO1p) served as substitute for the NRT1.5 promoter in nrt1.5 complementation experiments, because expressing the NRT1.5 coding region under the control of the 2.4 kb upstream region of NRT1.5 or the ubiquitous 35S CaMV promoter does not achieve complementation of the nrt1.5 mutant (Drechsler et al., 2015).PHO1 (At3g23430) was identified as a gene whose expression is like that of NRT1.5 primarily directed to the root vasculature in Arabidopsis plants (Hamburger et al., 2002).Transgenic plants with the NRT1.5 coding region fused behind the PHO1 promoter expressed almost wild-type levels of NRT1.5 transcripts in the roots (Drechsler et al., 2015).Homozygous PHO1p::NRT1.5eGFPand PHO1p::NRT1.5eGFPG209E plants were evaluated phenotypically under K + /NO 3 -deficiency and exposure to HygB.The root growth assay indicated that PHO1p::NRT1.5eGFPplants can rescue the reduction in root density (Figures 2A, B) and the early senescence phenotype of nrt1.5 plants under K + and NO 3 -deficiency (Figure 2C), corroborating previous studies of this complementation line without the fluorescent tag (Drechsler et al., 2015).In contrast, the transgenic nrt1.5 PHO1p::NRT1.5 G209E eGFP plants retained the nrt1.5 root and shoot phenotypes, demonstrating that the mutation G209E prevents complementation of the nrt1.5 mutant.Expression of the NRT1.5 and NRT1.5 G209E proteins was confirmed by the detection of GFP signals in the transgenic plant roots and at the PM of tobacco epidermal cells (Supplementary Figure 1).Hygromycin B (HygB) is used as an indicator of the PM potential status in plant and yeast cells (Haruta et al., 2010;Barreto et al., 2011;Haruta and Sussman, 2012).When the PM potential is reduced, the toxic drug is not able to enter the cells and the plants develop an enhanced HygB tolerance in comparison to wild-type, as was observed in nrt1.5 mutant plants (Drechsler et al., 2015).The two independent PHO1p::NRT1.5eGFPlines were sensitive to 5 µg/ml HygB (reduction of root growth and yellow cotyledons), whereas the PHO1p::NRT1.5G209E eGFP lines were more resistant, indicating that the PM potential was restored upon expression of NRT1.5 in the nrt1.5 mutant background (Figure 2A).In contrast, the NRT1.5 G209E mutant protein is not able to complement the nrt1.5 mutant, most likely because of a disturbed ion transport function of mutant protein, consistent with heterologous expression results in Xenopus laevis oocytes (Li et al., 2017).

The proton pump AHA2 contributes to K + translocation under limiting K + and NO 3 -supply
To investigate if the NRT1.5 interaction with AHA2 is required for nutrient root-to-shoot translocation in Arabidopsis, the nrt1.5/aha2 double mutant was generated.No residual NRT1.5 or AHA2 transcripts were detectable by semi-quantitative RT-PCR in the double mutant (Figure 3A).When grown on unfertilized soil and watered with low-nutrient fertilizer (LK: 1 mM K + , 1 mM NO 3 -), wild-type, single, and double mutant plants showed a similar degree of early shoot chlorosis compared to plants grown under ample nutrient supply (HK: 10 mM K + , 10 mM NO 3 -) (Figure 3B), and no reduction in fresh weight (Figure 3C).Under ample nutrient supply, all plant lines developed chlorosis in less than 10% of leaves, whereas with low nutrient fertilization approximately 40% of the leaves showed leaves chlorosis (Figure 3D).
In an earlier study it was shown that nitrate as well as total N concentrations in rosettes of nrt1.5 mutant plants are similar as in Col-0 wild-type plants under both low and high N fertilization regimes, whereas the K concentration in the mutants remained below 30% and 50% of the wild-type levels (Drechsler et al., 2015).We therefore determined here the concentration of mineral nutrients in 30 days old leaves of Col-0, nrt1.5, aha2, and nrt1.5/aha2 plants by inductively coupled plasma optical emission spectrometry (ICP-OES).When fertilized with 10 mM potassium and nitrate (HK), the concentrations of none of the elements K, Ca, Mg, B, Mn, Zn, Al, P, and S differed between wild-type, nrt1.5, aha2, or nrt1.5/aha2leaves (Figure 4).In contrast, under limiting nutrient supply (LK) Ca, Mg, and S concentrations were increased in nrt1.5 and nrt1.5/aha2leaves, indicating that the lack of NRT1.5 disturbs the distribution of these elements between roots and shoots, whereas K levels were significantly decreased in nrt1.5, nrt1.5/aha2 and also aha2 mutant plants.These results suggest that in low nutrient conditions both NRT1.5 and AHA2 are required for efficient K + translocation to shoots, whereas the distribution of Ca, Mg, and S is not dependent on AHA2.
The contribution of NRT1.5 to K + translocation from roots to shoots under low K + supply was reported by Li and colleagues before (Li et al., 2017).In addition, under low K + in combination with low nitrate conditions, the nrt1.5 mutant exhibits leaf chlorosis and a reduced lateral root density (Zheng et al., 2016;Li et al., 2017).To investigate whether AHA2 is also involved in nutrient translocation, a seedling growth test was carried out.In the presence of HK root development did not differ in Col-0 wildtype and the three mutant plants, where all plants showed the same lateral root (Figures 5A, B) and root hair density (Figures 5C, D).In 0K medium, nrt1.5 mutant seedlings show a significant reduction in lateral root density.In the nrt1.5/aha2double mutant, the lateral root density is similarly reduced by trend, whereas in aha2 seedlings it is similar to Col-0 (Figure 5B).However, in all three mutant lines and most prominently in the nrt1.5/aha2double mutant plants the density of root hairs is significantly reduced in 0K medium compared to wild-type (Figure 5D) and the cotyledons develop an early chlorosis phenotype (Figure 5E).These observations suggest that in addition to NRT1.5 also AHA2 is involved in nutrient translocation to shoots in young seedlings.
3.4 NRT1.5 and AHA2 expression influence the plasma membrane potential and pH homeostasis under low nutrient supply NRT1.5 exports K + from root parenchyma cells into the xylem vessels at weakly acidic extracellular pH conditions like in the xylem sap (Li et al., 2017).AHA2 is the major PM proton efflux pump in root cells (Harper et al., 1990;Haruta et al., 2010).Both transporters effectuate ion movement across the PM of root cells and thus affect the PM potential.In plants, PM H + -ATPases are responsible for the establishment of the cellular membrane potential (Sondergaard  2) for 40 days.et al., 2004).To investigate if NRT1.5 and AHA2 cooperatively contribute to the maintenance of the PM potential, we indirectly measured this parameter in yeast and plant cells overexpressing or lacking NRT1.5 and AHA2.
In yeast cells and plant roots, the uptake of the toxic cationic drug Hygromycin B (HygB) depends on the PM potential.Sensitivity to HygB is often linked to changes in the PM potential, which can be provoked by alterations in K + homeostasis in S. cerevisiae (Barreto et al., 2011) or alterations in H + homeostasis in A. thaliana (Haruta and Sussman, 2012).We observed that the growth of Arabidopsis Col-0 seedling roots was highly sensitive to 5µg/ml HygB.In contrast, root growth of the nrt1.5, aha2, and nrt1.5/aha2mutants is much less sensitive to HygB (Figure 6A), implying that the lack of NRT1.5 and AHA2 leads to membrane depolarization.
In a HygB sensitivity assay with the yeast mutant BYT12, which lacks the two main K + -uptake transporters Trk1 and Trk2, the expression of NRT1.5 causes hyperpolarization of the yeast PM, indicating that NRT1.5 is active in yeast cells (Drechsler et al., 2015).To investigate whether NRT1.5 and AHA2 affect the PM potential, both genes were individually or co-expressed in wild-type BY4741 and mutant BYT12 cells.Subsequently, yeast growth was monitored in plates containing high K + concentration (0.1 M), which supports growth of BYT12 cells, and a HygB concentration gradient.Overexpression of the NRT1.5 G209E mutant protein was included in the assay to evaluate the impact of the mutation in yeast.The expression of NRT1.5 in BYT12 cells caused increased sensitivity to high HygB concentrations (Figure 6C), as previously reported by Drechsler et al. (2015).Expression of the mutant NRT1.5 G209E led to partial growth recovery of BYT12 cells, suggesting the mutation G209E affects the NRT1.5 ion translocation function in yeast.The expression of AHA2 in BYT12 cells results in increased HygB sensitivity compared to BYT12 cells expressing NRT1.5, suggesting that the PM potential is strongly influenced by the expression of Phenotyping of wild-type and mutant plants under low (LK) and high nutrient (HK) conditions.(A) RT-PCR analysis of NTR1.5 and AHA2 transcripts in Col-0 and the nrt1.5/aha2double mutant.GAPC1 (At3g04120) was used as reference housekeeping gene (B) Rosette phenotypes, (C) Leaf fresh weight, and (D) Fraction of chlorotic leaves (%) of Col-0, nrt1.5, aha2, and nrt1.5/aha2plants that were pre-germinated in fertilized soil for seven days and transferred to unfertilized soil type for another 30 days while watered with HK or LK medium, respectively.Shown are the means ± SD of the fraction of chlorotic leaves in 20 individual plants.Different letters above columns indicate statistically significant differences (Tukey's test; P <0.05).
AHA2.BYT12 cells expressing NRT1.5 and AHA2 exhibited the same sensitivity to HygB as cells expressing AHA2 alone (Figure 6C).Thus, this assay does not indicate a biological interaction between NRT1.5 and AHA2 in yeast cells.
Plant PM H + -ATPases use energy from ATP hydrolysis to pump protons from the cytoplasm to the extracellular space, thus creating and maintaining a negative membrane potential at the cytosolic side and a transmembrane pH gradient that acidifies the extracellular space (reviewed by Sondergaard et al., 2004;Falhof et al., 2016) and controls the transport of nutrients across the PM (Cosse and Seidel, 2021).To investigate if NRT1.5 and AHA2 cooperatively influence the extracellular pH, we incubated Col-0, nrt1.5, aha2, and nrt1.5/aha2seedlings in media with different K + and NO 3 -concentrations and determined the pH using the fluorescent dye FITC-dextran.The assay detects pH differences at the Arabidopsis root surface with a sensitivity that is comparable to that of pH-sensitive microelectrodes (Hoffmann and Kosegarten, 1995;Pitann et al., 2009;Haruta et al., 2010;Haruta et al., 2018).At high nutrient supply (HK), the extracellular pH of Col-0 and the nrt1.5 mutant roots were similar (pH ≈5.3), whereas aha2 roots led to an increase to almost pH 5.5 (Figure 6B).An increase in extracellular pH indicates an impaired PM proton-secretion activity and has previously also been observed in two other aha2 mutants (Haruta et al., 2010).The nrt1.5/aha2 double mutants showed an intermediate increase to pH ≈5.4.Wild-type roots showed at 0K supply a reduction of the extracellular pH to ~4.9, supporting that low nutrient availability induces acidification of the extracellular space or rhizosphere (Houmani et al., 2015), presumably due to increased activity of PM H + -ATPases.In contrast, extracellular acidification by nrt1.5, aha2, and nrt1.5/aha2 mutants was reduced under these conditions, where the double mutant showed the highest reduction of acidification to pH ≈5.3.These results are consistent with the hypothesis that in wild-type plants under altered K + and NO 3 -supply NRT1.5 and AHA2 interact and mutually activate each other.As AHA2 exports more protons than NRT1.5 imports, the acidification of the extracellular space is relatively strong.In the mutants the mutual induction of both proteins is absent, resulting in a slightly higher extracellular pH (Figure 6B).The only subtle phenotypic differences of the aha2 and nrt1.5/aha2 double mutants relative to the nrt1.5 mutant prompted us to investigate whether other H + -ATPase family members are upregulated when AHA2 is lacking, we examined the gene expression of all H + -ATPase genes (AHA1-AHA11) in Col-0, nrt1.5, aha2, and nrt1.5/aha2seedlings relative to AHA2 and their response to low nutrient supply.Except for AHA1, which exhibits a similar absolute expression level as AHA2, all other gene family members are expressed at 10-to 10,000-fold lower levels than AHA2 (Supplementary Figure 2A).These results are confirmed by transcriptome data sets in the Arabidopsis eFP Browser (https://bar.utoronto.ca/efp_arabidopsis/cgi-bin/efpWeb.cgi;Winter et al., 2007) which show that AHA1 and AHA2 are the only gene family members that are significantly expressed in seedlings or seedling roots.Under low nutrient supply AHA2 is ~2fold and AHA1 is ~4fold upregulated in the nrt1.5 mutant.Surprisingly though, no upregulation of AHA1 is observed in the nrt1.5/aha2double mutant plants.In the aha2 mutant we observe some upregulation of AHA6 and AHA9 (Supplementary pH and plasma membrane potential are influenced by NRT1.5 and AHA2 expression.(A) Root growth phenotypes of Col-0, nrt1.5, aha2, and nrt1.5/aha2 in response to 5 µg/ml HygB.5-day-old seedlings were transferred to 0.5 × MS medium pH 5.5 supplemented with or without 5 µg/ml HygB for fourteen more days.The white bar represents 1 cm.(B) Extracellular pH measurements in Col-0, nrt1.5, aha2, nrt1.5/aha2roots under HK and 0K nutrition supply.Ten days-old seedlings were incubated with 0.25 x MS liquid medium supplemented with 30 µg/ml of FITC-dextran for 16 h.Media pH was determined by a fluorescence standard curve with a pH range from 4.5 to 7.5.Different letters indicate statistically significant differences (Tukey's test) between mutants and Col-0 with P<0.05, (means ± SD, n =12).(C) BY4741 wild-type and BYT12 mutant yeast strains were transformed with empty expression vectors (p425-TEF and p426-TEF) or expression constructs for the indicated proteins.Twelve 20 µl drops of untransformed (negative control) or transformed yeast cell suspension (OD 600 nm =1) were spotted along the HygB gradient (from 0 to 0.5 g/l) on SD-leu-ura pH 6.0, 0.1 M KCl plates and incubated at 30°C for two days.p425-TEF and p426-TEF are the empty expression vectors.
Figure 2B), however, due to the extremely low basic expression of these genes in seedlings it is very unlikely that they could compensate for the lack of AHA2 in the aha2 mutant.

Discussion
4.1 NRT1.5 and AHA2 coordinately control the H + and K + homeostasis in plants In this study, we demonstrate that the H + /K + -antiporter NRT1.5 interacts in vivo with the plasma membrane H + -ATPase AHA2 and that both proteins contribute to K + transport under K + and NO 3 -deficiency.Our results support the assumption that proton-motive force generated by AHA2 energizes the K + transport across the PM by NRT1.5.AHA2 is expressed in many tissues, most strongly in different root cells, stomatal guard cells, and mesophyll cells (Alexandersson et al., 2004;Sondergaard et al., 2004;Młodzińska et al., 2015;Falhof et al., 2016).In roots it is the major proton pump and, like NRT1.5, it is found in the PM of pericycle cells, which is a prerequisite for a direct interaction between the two proteins (Nicolson, 2014).A close link between proton transport and K + fluxes at the PM has already been concluded by the induction of the K + transporters HAK5 and CHX17 under K + starvation in wild-type (Cellier et al., 2004;Gierth et al., 2005) and aha2 mutant seedlings (Haruta and Sussman, 2012).Moreover, K + uptake under low K + conditions is promoted by the direct interaction between AHA2 and the receptor-like protein kinase BAK1 (Wang et al., 2022).Since the electrical membrane potential and chemical pH gradient generated by the proton-pumping ATPase AHA2 are essential for the function of the majority of plant PM transporters, it is not surprising that AHA2 is regulated by and interacting with many physiological parameters and protein factors (reviewed by Falhof et al., 2016;Michalak et al., 2022).AHA2 expression is affected by changing nitrate supply (Młodzińska et al., 2015) and phosphorus deficiency (Yuan et al., 2017).An additional layer of complexity of AHA2 regulation is added by post-translational control of the AHA2 protein activity.AHA2 can be phosphorylated at its C-terminal end by at least two kinases, which either enables or prevents binding of activating 14-3-3 proteins (Fuglsang et al., 1999;Fuglsang et al., 2007;Haruta et al., 2015).Recently, AHA2 was identified to be a central node in a protein-protein interaction network that undergoes highly dynamic protein-protein interaction assembly and disassembly processes upon nitrogen starvation (Gilbert et al., 2021).
We found that under low potassium and nitrate supply the lack of NRT1.5 and/or AHA2 in Arabidopsis seedlings triggered a decrease of K + in leaves (Figure 4A), and a strong reduction in root hair density relative to Col-0 wild-type plants (Figure 5).Modulation of root architecture by nutrient availability is a wellknown and diverse response in plants (Forde and Lorenzo, 2001;Gruber et al., 2013).It is likely that AHA2 indirectly contributes to these root morphology alterations as it facilitates an acidic cell wall environment by extrusion of protons to the apoplast, which is essential for cell wall expansion and cell growth (Pacifici et al., 2018;Sze and Chanroj, 2018).
Eliminating or overexpressing NRT1.5 and AHA2 in Arabidopsis and yeast, respectively, caused an alteration of the PM potential (Figures 6A, C).In AHA2-lacking seedlings grown under normal N and K + supply the extracellular pH in the medium is slightly elevated (Figure 6B), as was also observed in other studies (Haruta et al., 2010).When wild-type plants face K + deprivation, the extracellular pH drops sharply, indicating an increased H + export or reduced H + import or both.A reduced H + import could results from downregulation of K + /H + antiporters like NRT1.5.In aha2 and nrt1.5 mutant seedlings the drop in extracellular pH is lower under K + deprivation compared to wild type plants, and in the nrt1.5/aha2double mutant no more pH alteration is observed.This indicates that under these conditions NRT1.5 and AHA2 cooperatively modulate the extracellular pH, which is supposedly accomplished by the direct interaction of the two proteins.
In a screen for low-K + -sensitive Arabidopsis mutants Li et al. ( 2017) identified the lks2 mutant, in which a single-nucleotide mutation causes the substitution of Gly 209 to Glu (G209E) in NRT1.5.The NRT1.5 G209E protein is not able to complement the nrt1.5 phenotype in roots or shoots of Arabidopsis (Figure 2), and its expression in yeast does not phenocopy the PM potential status of cells expressing NRT1.5 (Figure 6).As all Arabidopsis NRT1 proteins contain either Gly, Ala or Val at the NRT1.5 G209 homologous position (Supplementary Figure 3B), which is also conserved in NRT1.5 homologs from other plant species (Supplementary Figure 3A), it seems likely that an uncharged amino acid at this position is functionally crucial for NRT1 family transporters.Gly 209 is not exposed at a cytosolic or extracellular loop of the transporter, but is located in the center of transmembrane span 5.It therefore less likely that Gly 209 is directly involved in the interaction with AHA2, but it rather is critical for the structural integrity of NRT1.5.
4.2 Different proton pumps from the H + -ATPases family may contribute to the cellular pH control under low K + and NO 3 -supply In the Arabidopsis thaliana genome 11 PM H + -ATPase (AHA) encoding genes and one pseudogene (AHA12) were identified (reviewed in Gaxiola et al., 2007).AHA1 and AHA2 are the predominantly expressed isoforms in Arabidopsis seedlings, leaves and roots, but also low levels of AHA3, AHA4, and AHA11 were detected by transcriptomics and/or mass spectrometry (Harper et al., 1990;Alexandersson et al., 2004;Haruta et al., 2010).In the root endodermis AHA4 was shown to contribute to ion homeostasis and nutrient transport (Vitart et al., 2001).AHA6 and AHA9 are essential for pollen tube growth and fertility (Hoffmann et al., 2020).Some PM H + -ATPases have been reported to be transcriptionally induced under abiotic stresses including nutrition, light, and salinity (Gevaudant et al., 2007;Caesar et al., 2011;Yuan et al., 2017;Ando and Kinoshita, 2018).However, reports on specific physiological functions of Arabidopsis AHA family members are sparse, presumably because of functional redundancy of the gene family members (Gaxiola et al., 2007;Haruta et al., 2010).For example, AHA1 can functionally complement for AHA2 (Rodrigues et al., 2014).
In aha2 and nrt1.5/aha2mutants we observed under K + deprivation conditions up-regulation of AHA4, AHA6 and AHA9 transcription (Supplementary Figure 2).However, as the regulation of these genes under LK-versus HK-conditions was inconsistent in aha2 and nrt1.5/aha2plants and even the induced expression level of these genes remained more than 100 fold lower than that of AHA2, we have no evidence for complementation of the lack of AHA2 in the mutants by other AHA family members.
Arabidopsis NRT1.5 is not only a nitrate transporter involved in long-distance NO 3 -transport from root to shoot, but under NO 3 deficiency also a potassium transporter loading K + into the xylem.
Here, the identification of the proton pump AHA2 as an interactor partner of NRT1.5 reveals a novel regulatory level in the control of K + and H + homeostasis in Arabidopsis.