The role of PQL genes in response to salinity tolerance in Arabidopsis and barley

Abstract While soil salinity is a global problem, how salt enters plant root cells from the soil solution remains underexplored. Non‐selective cation channels (NSCCs) are suggested to be the major pathway for the entry of sodium ions (Na+), yet their genetic constituents remain unknown. Yeast PQ loop (PQL) proteins were previously proposed to encode NSCCs, but the role of PQLs in plants is unknown. The hypothesis tested in this research is that PQL proteins constitute NSCCs mediating some of the Na+ influx into the root, contributing to ion accumulation and the inhibition of growth in saline conditions. We identified plant PQL homologues, and studied the role of one clade of PQL genes in Arabidopsis and barley. Using heterologous expression of AtPQL1a and HvPQL1 in HEK293 cells allowed us to resolve sizable inwardly directed currents permeable to monovalent cations such as Na+, K+, or Li+ upon membrane hyperpolarization. We observed that GFP‐tagged PQL proteins localized to intracellular membrane structures, both when transiently over‐expressed in tobacco leaf epidermis and in stable Arabidopsis transformants. Expression of AtPQL1a, AtPQL1b, and AtPQL1c was increased by salt stress in the shoot tissue compared to non‐stressed plants. Mutant lines with altered expression of AtPQL1a, AtPQL1b, and AtPQL1c developed larger rosettes in saline conditions, while altered levels of AtPQL1a severely reduced development of lateral roots in all conditions. This study provides the first step toward understanding the function of PQL proteins in plants and the role of NSCC in salinity tolerance.

(i.e., yield) must increase to meet future food demands. Increasing the yield of crops grown on saline soils, and the development of salt tolerant crops that can grow on marginal land that is currently considered unsuitable for agriculture, will help to meet these future requirements.
Responses of plants to salinity are classified into two phases: (a) early stress responses which are independent of shoot ion accumulation, the so-called osmotic phase, and (b) the ionic phase, where plants experience the toxic accumulation of ions in the shoot (reviewed by Munns & Tester, 2008). Plants can reduce the detrimental effects experienced during the ionic phase by excluding sodium ions (Na + ) from photosynthetically active tissues, compartmentalizing the ions into the non-photosynthetically active tissues, and at the sub-cellular level, to the vacuoles. Several families of ion transporters (e.g., HKT, SOS) have been demonstrated to be important for controlling Na + and K + movement within the plant (reviewed by Isayenkov & Maathuis, 2019;van Zelm et al., 2020). However, little is known about Na + influx into the root cells from the soil solution (Roy et al., 2014). As this initial Na + influx into plant roots is correlated with final Na + accumulation in the shoot (Horie et al., 2012;Tester & Davenport, 2003), it is important to characterize the genetic components regulating Na + influx into the roots.
The pathways proposed for Na + influx into root cells comprise protein-mediated Ca 2+ -sensitive and insensitive pathways as well as the bypass flow (Tester & Davenport, 2003). K + uptake transporters in the KUP family and high-affinity K + transporters in the HAK family are Ca 2+ insensitive and have been demonstrated to facilitate transport of Na + (Alemán et al., 2014;Bañuelos et al., 2002;Fulgenzi et al., 2008;Santa-María et al., 1997;Takahashi et al., 2007). However, these transporters were shown to transport Na + only in heterologous systems and appear to account for a small fraction of Na + transport under conditions of low K + and high Na + . The addition of extracellular Ca 2+ alleviates the toxic effects of salt stress on plants by reducing Na + influx (reviewed by Tester & Davenport, 2003), suggesting that Ca 2+ -sensitive pathways play an important role in Na + influx. Low Affinity Cation Transporter 1 (LCT1) was previously suggested to mediate the Ca 2+ dependent influx of Na + , When LCT1 was expressed in yeast, it increased Na + influx, and the Na + influx was blocked after addition of Ca 2+ to the external solution (Amtmann et al., 2001).
In yeast, electrophysiological activities of a non-specific cation channel 1 (NSC1) have been reported and this activity is similar to the Ca 2+ -sensitive vi-NSCC activity in plant roots (Bihler et al., 1998(Bihler et al., , 2002. Given the identification of Ca 2+ -sensitive vi-NSCC-like activity in yeast, previous research (Carter, 2009;Tester et al., 2013) sought to identify candidate genes in yeast, as these could inform the search for plant Ca 2+ -sensitive vi-NSCC. Carter (2009) used an in silico screen to find possible candidate genes for vi-NSCC activity in Saccharomyces cerevisiae and identified two candidates, YOL092w (ScPQL1) and YDR352w (ScPQL2). Bioinformatics analysis was performed and revealed that ScPQL1 and ScPQL2 are putative members of the PQ-loop (PQL) protein family (Pfam PF04193;Carter, 2009;Tester et al., 2013). PQL protein family members are predicted to be membrane bound with five or seven transmembrane helices.
Members of this family contain one or two conserved pairs of proline (P) and glutamine (Q) amino acids within a broader, weakly conserved region of 40-60 amino acids (Saudek, 2012). The physiological characteristics of ScPQL1 and ScPQL2 using heterologous expression in the Xenopus laevis oocyte system revealed that both proteins facilitate monovalent cation influx and were inhibited by low external pH and the presence of divalent cations (Carter, 2009;Tester et al., 2013). The transport activities of ScPQL1 and ScPQL2 are similar to NSC1 activity (Bihler et al., 1998(Bihler et al., , 2002, making them a putative candidate for NSCC.
In this manuscript we identified plant homologues of ScPQL1 and
The second strain, LL178 that lacks the native PQL1, PQL2, and PQL3 genes and is derived from the Σ1278b WT strain. The genotype of LL178 is ura3 ypq1Δ ypq2Δ ypq3Δ (Jézégou et al., 2012). LL178 was obtained from Dr. Bruno Gasnier and Dr. Bruno André. The yeast transformations were performed by the lithium acetate method according to Gietz and Woods (2002). The untransformed yeast strains were inoculated into 10 ml of yeast extract peptone dextrose YPD  (Salazar, 2017). A range of salt stress was tested (30 mM NaCl, 35 mM NaCl, 50 mM KCl, and 1 mM LiCl) and the plates were incubated at 30°C for 3 days. The digital images were taken at the end of the experiment using a Canon digital camera.

| Heterologous protein expression and electrophysiological characterization using HEK293 cells
The AtPQL1a and HvPQL1 cDNA were cloned into mammalian ex-  acquisitions, and analysis were performed using pClamp software (ver. 10 package, Molecular Devices). The signals were low-pass filtered at 2 kHz before analog-to-digital conversion and were uncorrected for leakage current or capacitive transients. The data were expressed as mean ± standard error of the mean.
The co-localization was visualized 72 hr after infiltration. Image was captured using confocal laser scanning microscope (LSM-880, Zeiss). Images were analyzed with Fiji image analysis software (Schindelin et al., 2012). The eGFP excitation was at 488 nm and the emission was between 505 and 530 nm, while the excitation for mCherry was at 515 nm and the emission was at 734 nm.

| Stable transformation of AtPQL1b and subcellular localization assay
The UBQ10p::AtPQL1b::eGFP was used for floral dip transformation (Zhang et al., 2006) in order to generate transgenic Arabidopsis overexpression stable line in Col-0 background. Transgenic plants were selected on 0.5 MS medium containing 10 μg/mL BASTA.
Eight-day-old plantlet roots of stable transgenic T3 homozygous lines of Arabidopsis plants overexpressing UBQ10p::AtPQL1b::GFP were stained by Propidium Iodide following (Helariutta et al., 2000) protocol. In order to visualize the GFP and Propidium Iodide fluorescence signals, images were captured with Zen 2.3 image software (Zeiss, Germany) using confocal laser scanning microscope (LSM-880, Zeiss, Germany). Images were processed with Fiji image analysis software (Schindelin et al., 2012). The eGFP excitation was at 488 nm and the emission was at 505-530 nm, while the excitation for Propidium Iodide dye was 515 nm and the emission was between 615-734 nm.   Table S2.

| High-throughput phenotyping of salt stress responses
To study natural variation in the Arabidopsis transgenic materials in response to salt, Photon Systems Instrument (PSI) was used to perform high-throughput phenotyping experiment. The plant materials used were Col-0 as a control, constitutive overexpression lines (35S), and knockdown lines (amiRNA; Shearer, 2013). T-DNA mutant AtPQL1b (Atpql1b-1 SALK_001485 and Atpql1b-2 SALK_129118), and AtPQL1c (Atpql1c-1 -Salk_036418, Atpql1c-2 -Salk_044346, and Atpql1c-3 -Salk_060084) were ordered from NASC. All lines were examined for homozygous T-DNA insertions using the primers designed with T-DNA express. The seeds were germinated and grown as described in Awlia et al. (2016). Seeds were stratified in water and stored in 4ºC for 3 days in the dark. Subsequently, seeds were sown into pots (70 mm x 70 mm x 65 mm) filled with 85 g of fresh soil that was then watered to full water-holding capacity. Fifteen replicates for each genotype were grown in controlled growth chamber (12 hr/ 12 hr light/dark cycle at 22°C and a relative humidity of 60%). The plants were weighed and watered automatically every day. Salt application occurred at 10-leaf stage that was 17 days after germination. The salt applied by soaking the trays in 200 mM NaCl or water for 1 hr to achieve fully saturation stage for the soil. Subsequently, drained the pots for 10 min before placing it back to the PSI chamber. The actual concentration of the salt stress that plant exposed to was approximately 80 mM NaCl. The plants were automatically photographed from above every 12 hr during the day and night (2 p.m. and 2 a.m.) for 7 days. Each measuring round consisted of 15 min dark adaptation period inside the acclimation chamber then kinetic chlorophyll fluorescence (ChlF) and RGB imaging, weighing, and watering. Pixel count, color, and chlorophyll intensity were evaluated from images. The ChlF imaging was optimized using light curve protocol (Henley, 1993). The harvesting occurred after 7 days of salt applications which was the end of the experiment.
The harvested whole shoot used to measure fresh and dry mass and Na + and K + accumulation using inductively coupled plasma-optical emission spectrometry (ICP-OES). Water content was calculated from the differences between the fresh and dry mass. Also, Na + /K + ratio was calculated from the Na + content over the K + content per dry mass of the plant . From the RGB images, and treatments for the RGB traits were done using R studio software (Rstudio team, 2015) and statistically significant differences between the genotypes were determined using student's t-test.

| Identification of plant PQL proteins through phylogenetic relationships
We identified plant homologues of ScPQLs by performing a BLAST search on the NCBI database, using the full-length cDNA sequences 3.2 | AtPQL1a, AtPQL1b, AtPQL1c, and HvPQL1 transport Na + , K + , and/or Li + in yeast The transport properties of AtPQL1a, AtPQL1b, AtPQL1c, and HvPQL1 were initially examined using yeast growth assays. The

| AtPQL1a and HvPQL1 showed permeability to monovalent ions such as Na + , K + and/or Li +
In order to investigate further the transport properties of AtPQL1a and HvPQL1, the patch clamp assay was performed using HEK293 cells. When measuring currents in whole-cell configuration from a non-transfected HEK293 cell, the only detectable current in Na +containing bath solutions was an I A -like current that has an ini-   (Kumar et al., 2016) utilizing the Neighbor-Joining tree building method (Saitou & Nei, 1987). Numbers on the nodes indicate bootstrap values as a percentage. Red box indicates clade 1, green box indicates clade 2, and blue box indicates clade 3 of PQL proteins. Scale bar indicates the number of amino acid substitutions per site. (b) Predicted protein topologies of PQLs belonging to clade 1. The protein transmembrane structures are predicted using the HMMTOP (Tusnády & Simon, 1998 and visualized using the TMRPres2D (Spyropoulos et al., 2004) with default settings. The topology of PQLs in clade 1 is predicted to have seven transmembrane domains and two PQ motifs, indicated by the red arrows. The part above and below the lipid bilayer indicates the extracellular and intracellular loop, respectively HvPQL1 discriminated poorly between monovalent cations such as K + , Na + , and Li + , with Li + showing slightly higher permeability relative to both Na + and K + (Figure 3c). The inward currents were inhibited by addition of high external calcium ( Figure 3c) and marginally inhibited by external acidification (Figure 3d), two notable properties of Na + influx currents in plants. The electrical currents facilitated by AtPQL1a (Figure 3e) are similar to currents observed for HvPQL1, and the currents facilitated by AtPQL1a is more sensitive to inhibition by external calcium compared to HvPQL1.

| PQLs localize into the internal membrane structures in planta
In order to examine the subcellular localization of PQL proteins in planta, the subcellular localization of AtPQL1a, AtPQL1b, AtPQL1c, and HvPQL1 fused to GFP at the C-terminal end was performed Summarizing, we observed that clade 1 PQLs localize to internal membrane structures. Based on the morphological characteristics typical to the vacuoles, and co-localization with the tonoplast marker, the most probable localization of the plant PQLs is the vacuolar membrane.

| Expression of clade 1 PQLs in plants is responsive to salinity stress and nutrient deprivation
As NSCCs are suggested to play an important role in ion transport, in particular of Na + (Munns & Tester, 2008), it is reasonable to F I G U R E 3 Electrophysiological characterization of HvPQL1 and AtPQL1a (a) Currents in whole-cell configuration recorded from HEK293 cells transfected with pcDNA6.2-HvPQL1-EmGFP vector (TC, red) and non-transfected (NTC, black) cells in 10 mM Na + external solution (left panels, voltage protocol shown below the traces) and the I-V plot (right panel) showing the average currents from NTC (5 cells) and TC (8 cells) in 10 mM Na + external solution. (b) The sketch of the setup used, with the internal solution containing 2.5 mM Na + while the external solutions contained three different Na + concentrations 50 mM Na + (•), 20 mM Na + (▲), and 10 mM Na + (■). Lower panels represent currents in whole cell configuration recorded from HEK293 cells transfected with HvPQL1 in 10 mM (■), 20 mM (▲), and 50 mM (•) external Na + . Middle upper panel represents I-V plot indicating the shift in the reversal potential for Na + (see red arrow), while upper right panel is E Na / E rev relationship indicating that the voltage shift in E rev is consistent with increasing Na + concentrations. hypothesize that they could play a role in salinity and nutrient stress.
Therefore, we examined tissue-specific expression of the clade 1 PQLs in Arabidopsis thaliana and barley (v. Morex; Figure 6). We observed that 1 and 7 days after application of salt stress treatment, the expression of AtPQL1a was upregulated compared to that seen in control conditions. This upregulation was specific to shoot tissue ( Figure 6a). The expression of AtPQL1b was also upregulated in response to salinity stress in shoot tissue, but the difference was found to be significant only 4 days after stress application (Figure 6a).  Figure S2). This observation could be caused by developmental effects of PQL1, although we did not observe any differences between PQL1 mutant lines and Col-0 grown under control conditions (Figure 8, Figure S2).
The effect of salt stress on the photosynthetic performance was assessed daily at different photon irradiances using chlorophyll fluorescence images, using the light curve protocol (Henley, 1993).
The significant differences between Col-0 and some mutant lines were observed for maximal fluorescence for dark-adapted state (F m ), variable fluorescence for dark-adapted state (F v ), instantaneous fluorescence level (F t ), and photochemical quenching (F q ) under control and salt stress conditions at the last day of experiment ( Figure S4).
In summary, the above results suggest that modifications in the clade 1 PQLs levels, either through overexpression or knock-down of the PQL genes, result in larger plant size and improved photosynthetic performance at the later stages of salt stress exposure compared to the wild-type plants.

| AtPQL1a and AtPQL1b affect development of lateral roots under control and ionic stress conditions
Although the expression of the clade 1 PQLs is predominant in the shoot tissue, the PQLs could potentially play an important role in root system architecture, contributing to the ion compartmentalization into the vacuole in concert with other proteins. Therefore, we studied the Root System Architecture phenotypes of PQL mutants and compared them to Col-0. Four major Root System Architecture traits were analyzed: main root length, total lateral root length, lateral root number, and total root length.

F I G U R E 5
AtPQL1b localizes in the internal membrane compartments in stably transformed Arabidopsis root tips. The root tips of stably transformed Arabidopsis plants were imaged using confocal laser scanning microscopy. The left panel shows the root epidermal cells stably transformed with UBQ10p::AtPQL1b-GFP. The middle panel shows the propidium iodide staining of the same cells, while the right panel is the merge between the two channels. Similar results were observed in 12 replicates, visualized during 3 independent microscopy sessions F I G U R E 6 Transcripts of Arabidopsis and barley PQLs are responsive to environmental stress. (a) Relative expression level of AtPQL1a, AtPQL1b, and AtPQL1c in shoot and root tissues of Arabidopsis grown under control (blue), salt stressed (100 mM NaCl; red), and nutrient starved (0.1 mM CaCl2; yellow) conditions. Three-week-old plants were used to apply the treatment and the samples for transcript level estimation were collected 1, 4, or 7 days after exposure to stress conditions. (b) The expression level of HvPQL1 was examined in leaf-blade, leaf-sheath, and root tissues of plants grown under control (blue), salt stressed (100 mM NaCl, red), and nutrient starved (0.1 mM CaCl2, yellow) conditions. Two-week-old plants were exposed to stress and the samples for transcript level estimation were collected 1 or 10 days after the exposure to stress conditions. The box plots represent the median expression value based on six biological replicates. The boxes represent the 1.5*Interquartile Range. Statistically significant differences between control and other stress conditions are indicated as * (p < .05) and ** (p < .01) as determined using ANOVA with pairwise Tukey HSD test F I G U R E 7 The Arabidopsis PQLs clade 1 does not significantly affect the ion accumulation. Two overexpression lines of AtPQL1a (35S::AtPQL1a-1 and 35S::AtPQL1a-2) and knock-down lines of AtPQL1a, AtPQL1b, AtPQL1c (amiRNA-AtPQL1a, Atpql1a, Atpql1b1, Atpql1b2, Atpql1c3, Atpql1c4, Atpql1c6), and Col-0 were germinated under control conditions. Salt stress (~80 mM NaCl) was applied at 17 days after germination, and the ion content (Na + , K + ), fresh and dry mass were determined 7 days after stress application for 5 replicates per condition per genotype.
The bars represent the median of 5 biological replicates. The boxes represent the 1.5*Interquartile Range. The differences between Col-0 and individual mutant lines were tested using a ttest. No significant differences were observed The lines with increased or decreased expression of AtPQL1a (35S::AtPQL1a-1, 35S::AtPQL1a-2, amiRNA-AtPQL1a, Atpql1a) were severely impaired in root development, mainly due to reduced lateral root development (Figure 9a

| D ISCUSS I ON
Salinity tolerance is determined, in part, by ion transport, which encompasses the initial flux of ions into the root and compartmentalization of ions into non-photosynthetic tissues and/or vacuoles (Volkov & Beilby, 2017). Non-selective cation channels were previously shown to be important for Na + influx into root cells (Amtmann & Sanders, 1998;Davenport & Tester, 2000;Demidchik & Maathuis, 2007;White, 1999), yet their genetic F I G U R E 8 Disruption of clade 1 PQL function results in larger plants under salt stress conditions. The Projected Rosette Area estimated 7 days after application of treatment for Col-0, 35S::AtPQL1a-1, 35S::AtPQL1a-2, amiRNA-AtPQL1a, Atpql1a, Atpql1b1, Atpql1b2, Atpql1c3, Atpql1c4, and Atpql1c6. Seventeen-day-old plants were exposed to treatment (control or ~ 80 mM NaCl) and their Projected Rosette Area was scored using automated phenotyping system (PSI, Czech Republic) from the RGB images. Box plots represent the median of 15 replicates per genotype and treatment. The boxes represent 1.5*Interquartile Range. The significant differences between Col-0 and the other genotypes per condition are indicated with *, **, or *** for p-values below .05, .01, and .001, respectively, as calculated using t-test F I G U R E 9 AtPQL1a and AtPQL1b affect the development of lateral roots under control conditions. Four-day-old seedlings were transferred from 0 mM NaCl to 75, 125 mM NaCl, 125 mM KCl, and 20 mM LiCl. RSA was quantified in 12 replicates after 6 days of treatment. The box plots represent the distribution of the samples for (a) the total root size (cm), (b) Lateral Root Number (#LR / MR), (c) Main Root Length (cm), (d) average Lateral Root Length, as well as (e) the relative Total Root Length (fraction of average Total Root Length at 0 mM NaCl) and (f) the relative Main Root Length (fraction of average Main Root Length at 0 mM NaCl). The significant differences between the Col-0 and studied mutant lines are indicated by *, **, and *** for pvalue below .05, .01, and .001 per treatment and calculated using t-test constituents are yet to be identified. In this study, we identified plant homologues of yeast PQLs (Figure 1), and identified clade 1 to contain the closest homologues of yeast PQLs, previously described to mediate the flux of several cations, including K + and Na + , in yeast (Carter, 2009). We performed functional characterization of clade 1 PQL from Arabidopsis and barley, including ion transport characteristics and their role in plant development, growth, and salt tolerance.
We observed that AtPQL1a and HvPQL1 are involved in the transport of monovalent cations, such as Na + , K + , and Li + (Figures 2-3). The PQL1 permeability was inhibited by high external Ca 2+ and pH acidification (Figure 3c,d), which is also typical for previously described Ca 2+ -sensitive voltage-insensitive non-selective cation channels (vi-NSCCs) reported for root-derived protoplasts from several plant species (Buschmann et al., 2000;Spalding et al., 1992;Tyerman et al., 1997). Only one difference was noticed, which is vi-NSCC favored K + over Na + , while in clade 1 PQLs showed higher permeability to Li + than K + and Na + (Figure 3c). The characteristics of clade 1 PQLs were previously reported to be similar in their transport properties to vi-NSCC (Carter, 2009;Shearer, 2013). As vi-NSCC is considered to be the major pathway for Na + influx into the plant root (Amtmann & Sanders, 1998;Davenport & Tester, 2000;Demidchik & Maathuis, 2007;White, 1999), we initially hypothesized that clade 1 PQLs were located to the plasma membrane of plant cells.
However, results presented in this study suggest that the clade 1 PQLs in Arabidopsis and barley are localized to internal membrane compartments, most likely the tonoplast (Figures 4-5). The NSCCs that have been reported to be in the tonoplast were either slow (SV) or fast (FV) vacuolar channels Pottosin & Dobrovinskaya, 2014). SV cation channels are activated by Ca 2+ and voltage Pottosin & Dobrovinskaya, 2014), which is different from the transport properties of clade 1 PQLs ( Figure 3). Clade 1 PQLs are also unlikely to belong to non-selective voltage-dependent FV cation channels, as the activity of PQLs is not voltage dependent. While both FV channels and clade 1 PQLs channels are inhibited by increased non-cytosolic Ca 2+ , the FVs are also inhibited by cytosolic Ca 2+ and Mg 2+ , which were not tested in this study. Additionally, clade 1 PQLs showed slight inhibition in response to low pH, while the same range of pH was not tested for FV .
In summary, AtPQL1a and HvPQL1 are Ca 2+ -sensitive, voltage-insensitive NSCC located to the vacuolar membrane. Further experiments will reveal the role of PQLs in ion transport and their relationship to other known NSCCs such as the FV channels.
We did not observe any changes in ion accumulation (Figure 7), unlike for the mutants of other ion transporters, including sos1 and hkt1, where increased accumulation of Na + ions under salt stress conditions was observed (Ji et al., 2013;Møller et al., 2009;Wu et al., 1996;Yang et al., 2009). However, SOS1 and HKT1 have presumably high specificity for Na + over other transported ions (Møller et al., 2009), which is unlike the PQLs, which may provide some explanation for no significant differences observed in our study.
Interestingly, we found that clade 1 PQLs play an important role in lateral root development (Figure 9). Some of the ion transporters were earlier described to affect plant development, including NHX (Bassil et al., 2011) and HKT1 (Julkowska et al., 2017). The effect of the clade 1 PQLs on root development is clearly visible in the nonstress conditions (Figure 9). Whether the single mutations in the clade 1 PQLs affect lateral root development by affecting vacuolar function remains to be studied.
Interestingly, we did not observe any developmental phenotypes This study is a first step toward the understanding of the function of the plant homologues of yeast ScPQL proteins. We present evidence that plant clade 1 PQLs transport monovalent cations in a non-specific fashion are localized to internal membranes and affect development of lateral roots in Arabidopsis seedlings.