Seeds of S. europaea were collected from an alkaline soil area of Xinjiang, China. Plants were germinated in tap water for 14 days and then transferred to half-strength Hoagland’s nutrient solution and allowed to grow for another 14 days before stress treatments. The temperatures of the growth chamber were maintained at 28°C during the day (16 h) and 22°C at night (8 h) under a photoperiod (350-400 µmol m^-2 s^-1). For the salt-stress treatment, NaCl was added to half-strength Hoagland’s nutrient solution with 0 mM, 10 mM, 200 mM, 500 mM, and 800 mM separately. For the K⁺-starvation treatment, KCl was removed from the standard solution. Under 200 mM NaCl treatment, plants tissue samples were harvested at 0, 6, 24, 48, 72, and 120 h after stress treatments and preserved at –80°C for further analyses.

Total RNA was extracted from the shoots of S. europaea using RNeasy plant extraction Mini Kit (Qiagen, Germany) according to the manufacturer’s instructions. Total RNA was reverse transcribed by using an oligo dT primer and MMLV-reverse transcriptase (TaKaRa, China). Full-length cDNA was obtained using 5’- and 3’-RACE (Rapid Amplification of cDNA Ends) techniques. The partial cDNA fragment was amplified by PCR, using a degenerate primer ( Table S1 ) deduced from the conserved regions of HKT and a reverse transcription product as a template. The full-length cDNA of SeHKT1;2 was amplified using primers ( Table S1 ). The PCR product was sub-cloned into the pMD-19 T vector and sequenced.

Transmembrane domains were predicted by DeepTMHMM (https://biolib.com/damgaard/DeepTMHMM-testing). A maximum-likelihood phylogenetic tree of plant HKT proteins was constructed using the MEGA program (version 7.0, Auckland, NewZealand). The accession numbers of genes utilized in the present study are listed in Table S2 . The protein secondary structure was predicted using NetSurfP - 3.0 online program(Høie et al., 2022). The 3D model of SeHKT1;2 protein was build using SWISS-MODEL online program (Waterhouse et al., 2018).

For real-time RT-PCR, total RNA was extracted from the shoots and roots of S. europaea subjected to 0 mM Na⁺, 10 mM Na⁺, 200 mM Na⁺, 500 mM Na⁺, and 800 mM Na⁺, K⁺-starvation, time-course using the RNeasy plant extraction Mini Kit (Qiagen, Germany) according to the manufacturer’s instructions. Two micrograms of DNase-treated RNA were transcribed to cDNA using the Reversal Transcription Reagent Kit (TaKaRa) following the manufacturer’s instructions. The cDNA was diluted 10 times and 1µl of the diluted cDNA was used as the template in each well for quantitative real-time PCR analysis. The cDNA was amplified using Power SYBR Green PCR Master Mix (Applied Biosystems, USA) on CFX96 Real-Time PCR Detection System (Bio-Rad, USA). A α-tubulin gene from S. europaea has been used as internal reference (Lv et al., 2012; Ma et al., 2013; Xiao et al., 2015) and served as an internal standard to normalize the expression data for the SeHKT1;2 gene. The qRT-PCR was performed using sequence specific primers ( Table S2 ). The thermal profile for qRT-PCR was as follows: 2 min at 95°C, 40 cycles of 15 s at 95°C and 30 s at 60°C, and a melting curve protocol (plates read when increased 0.5°C every 5 s from 65°C to 95°C). The melting curve verified the amplification specific and confirmed that there were no primer dimers. All of the samples were run with replicates. The threshold cycle (Ct) values were measured according to the setting of an auto-calculated baseline threshold in Bio-Rad CFX Manager software (Bio-Rad, USA).

The ORF of SeHKT1;2 was inserted into the yeast protein expression vector pYES2. To test the Na⁺ Absorption and efflux function of SeHKT1;2, the Saccharomyces cerevisiae yeast mutant strain AXT3K (enal::Hls3::ena4,nhal::LEuZ,nhxl::KanMx4), which lacks the main plasma membrance Na⁺ transporters, and G19 (MATa ade2ura3leu2his3trp1 ena1Δ::HIS3Δ::ena4Δ) disrupted in the ENA1-4 genes encoding Na ⁺ export pumps were used. The plasmids were introduced by PEG/LiAc method. Positive transformants were selected on Ura-selective medium [0.67% (w/v) yeast nitrogen base without amino acids, 0.077% (w/v) DO supplement-Ura, 2% (w/v) galactose, and 1.5% (w/v) agar]. Growth at variable Na⁺ concentrations with 0, 25, 30, 60, 100 or 150 mM for AXT3K and 0, 60, 100, 150, 300 mM for G19 were tested in arginine phosphate (AP) medium [8 mM phosphoric acid, 10 mM L-Arginine, 2 mM MgSO₄, 0.2 mM CaCl₂, 2% glucose, plus vitamins and trace elements, and 1.5% (w/v) agar, pH 6.5].

Meanwhile, Yeast (Saccharomyces cerevisiae) strain CY162 (MATa, △trk1, trk2:: pCK64, his3, leu2, ura3, trp1, ade2), which is a K⁺-uptake-defective mutant and cannot grow without supplement K⁺ was used to test K⁺-uptake function of SeHKT1;2 gene. Positive transformants were selected on Ura-selective medium with 100 mM KCl. Yeast growth experiments were performed on arginine-phosphate (AP) medium with added K⁺ (1 mM, 10 mM) and supplemented with 100 mM, 150 mM and 300 mM Na⁺ concentrations. Control experiments were performed with the yeast modified with vector pYES2 and pYES2-SeHKT1;2 growing under 100 mM KCl.

For the yeast growth test experiment, all transformed yeasts were cultured overnight at 30°C in AP medium until the OD₆₀₀ reached 0.8, and 10-fold serial diluted cultures were incubated on AP plates containing the indicated concentrations of K⁺ and Na⁺. The plates were incubated at 30°C for 5 days. Control experiments were performed with the yeast wild type modified with vector pYES2.

The coding regions of SeHKT1;2 was sub-cloned into the plant transformation binary vector pBI121 (Clontech, Japan) under the control of CaMV35S promoters. The constructs were introduced into Agrobacterium tumefaciens strain GV3101 and Arabidopsis salt-hypersensitive mutant sos1 was transformed by the floral dip method (Clough and Bent, 1998). Transgenic plants were selected on half-strength Murashige & Skoog medium containing 25 mg/L kanamycin. qRT-PCR showed the introduced gene transcript in 10 out of 15 lines of SeHKT1;2 with varied expression levels among the lines. Two lines transgenic plants with higher expression level were used for phenotype assays of salt stresses. All of the lines used in these experiments were homozygous. Wild type and transgenic Arabidopsis seeds were surface-sterilized with 70% (v/v) ethanol for 2 min and 1% (v/v) NaCl solution for 15 min. The seeds were rinsed three times with water and sown on an MS medium containing 1.5% (w/v) sucrose, 0.8% (w/v) agar, and 50 mg/L kanamycin. Seven-day-old seedlings were transferred from germination medium to soil, supplemented with the indicated amounts of 300 mM NaCl after ten days.

By using degenerate primers deduced from several HKT sequences and standard reverse transcription (RT)-PCR methods, a cDNA homolog of an HKT high-affinity K⁺ transporter from S. europaea was cloned. The length of the amplified cDNA fragment was about 1500-bp long. The full-length cDNA was obtained by 5′ and 3′ RACE (Rapid Amplification of cDNA Ends) which contains an ORF of 1611-bp long ( Figure S1A ). The translated amino acid sequence of this gene exhibited 44.31% amino acid sequence identity with SeHKT1;1 (AKS12114.1) ( Figure S1B ) therefore it was designated as SeHKT1;2 (AKS2645929). Subsequently the evolutionary relationships of SeHKT1;2 with HKT proteins from other plant species. An unrooted phylogenetic tree was generated based on full-length amino acid sequences. Results showed that SeHKT1;2 was grouped into subfamily I of HKT transporters ( Figure 1 ).

The SeHKT1;2 encoding a 536 amino acid polypeptide with 8 predicted transmembrane domains ( Figure 2 ). SMART domain architecture analysis showed that SeHKT1;2 protein has single TrkH conservative domain (pfam ID: PF02386) between amino acids 70-524 ( Figure S2A ). The secondary structure of the SeHKT1;2 protein was predicted, and it was found to have a complex helical folding structure ( Figure S2B ). The 3D models of SeHKT1;2 showed the presence of three glycine residues (Gly244, Gly368 and Gly480) and one serine residue (Ser119) forming an ion selectivity filter, which belong to subfamily I ( Figures 2 , S3 ).

The expression pattern of SeHKT1;2 gene under the salt stress conditions was determined. The SeHKT1;2 gene expression patterns in S. europaea seedlings were completely different between shoots and roots. In a time-course experiment, the SeHKT1;2 was induced by NaCI (200mM) and reached the peak at 48h in shoot, but it was gradually reduced in root ( Figure 3 ). Subsequently, we quantified expression level of SeHKT1;2 gene under the presence of NaCI. As a result, the 10 mM NaCI significantly increased the expression of SeHKT1;2 in shoot, but it remained same in the presence of high concentration (200mM, 500mM, 800 mM) of NaCI ( Figure 3 ). In addition, we also examined the SeHKT1;2 gene expression under K⁺ starved condition. In our experiments, SeHKT1;2 showed a very low expression level in roots while showed 10 folds higher in shoots ( Figure 3 ). The expression of SeHKT1;2 in root was not affected by NaCI.

Previous functional characterization in yeasts showed that SaHKT1 from the Suaeda salsa functions as a Na⁺ selective transporter (Wang et al., 2020). To determine whether SeHKT1;2 is involved in Na⁺ uptake, we expressed SeHKT1;2 proteins in a Na⁺-sensitive mutant yeast strain, G19 (Quintero et al., 1996). Results showed that under the low concentration of K⁺ (1Mm), the increasing of Na⁺ concentration in the growth medium caused severe growth defects in both SeHKT1;2-expressing cells and empty vector-harboring control cells ( Figure 4 left panel). Interestingly, higher K⁺ concentration (10mM) improved the growth of both control and SeHKT1;2-expressing cells ( Figure 4 right panel). Moreover, the higher concentration (300mM) of Na⁺ restricted cell growth of both control and SeHKT1;2-expressing cells ( Figure S4 ). These data indicate that Na⁺ toxicity was reduced by K⁺ uptake, and that cells expressing SeHKT1;2 has a selective advantage under K⁺ limitation suggesting that SeHKT1;2 have a higher affinity for K⁺.

To investigate the transporter activity of SeHKT1;2, we performed complementation assay by ectopic expression of SeHKT1;2 in the K⁺ transporter-deficient yeast strain CY162 (Anderson et al., 1992). We observed no significant differences between SeHKT1;2 and empty vector control in their growth supplied with various concentration of KCl, and both grew well on the medium containing 100 mM KCI ( Figure 5A ). This indicates that SeHKT1;2 failed to complement K⁺ uptake. Additionally, the transformants growth was completely impacted as the non-transformed strain when K⁺ and Na⁺ were applied in combined form, in the CY162 with increased K⁺ (10Mm KCI) in combination with Na⁺, the transformants growth was somehow recovered ( Figure 5B ). These results suggested that a trend toward K⁺ selectivity by SeHKT1;2.

In an attempt to validate the function of SeHKT1;2 in salt tolerance, we over expressed the SeHKT1;2 in Arabidopsis sos1 mutants. As expected, the whole phenotypes of SeHKT1;2-expressing in sos1 and wild type Col-0 plants showed the same growth vigor and had no obvious difference under the normal condition. The growth of sos1 plant, however, was restricted by imposing of salt stresses (300 mM NaCl) and attenuated by overexpression of SeHKT1;2 ( Figure 6 ). Interestingly, the overexpression of SeHKT1;2 cannot restore the salt tolerance of sos1 mutant plant.