Seeds of G. uralensis were obtained from the Mongolian Autonomous County of Hoboksar, Tarbagatay Prefecture, Xinjiang Uygur Autonomous Region (latitude 46°82′N, longitude 85°75′E). The seeds were planted in plastic pots (5 cm in diameter, 8 cm in height), containing a 2:1 mixture of sand and nursery soil, and placed in a growth chamber at Beijing Forestry University, Beijing, China. The potted G. uralensis were well irrigated, according to evaporation demand, and fertilized with one-quarter-strength Hoagland solution weekly. The temperature and relative humidity were maintained at 25–28°C and 60–70%, respectively. A photoperiod of 14 h (9:00–23:00) was applied, and photosynthetically active radiation varied from 280 to 350 μmol m^-2s^-1. After 2 weeks of culture, rooted liquorice seedlings were transferred to 300-ml pots containing one-quarter-strength Hoagland’s nutrient solution for hydroponic equilibration.

Hydroponic-equilibrated seedlings of G. uralensis were subjected to 0 or 100 mM NaCl for 24 h. Na⁺ concentrations in root and leaf cells were examined after 6, 12, and 24 h of treatment. Na⁺, K⁺, and H⁺ fluxes were measured along the root axes with the NMT technique. The effects of PM transporter/channel inhibitors were examined in NaCl-treated G. uralensis. A blocker of the Na⁺/H⁺ antiporter, amiloride (50 μM), and a specific inhibitor of the H⁺-ATPase, sodium orthovanadate (500 μM), were used to inhibit the Na⁺/H⁺ antiport system in the PM (Sun et al., 2009a). A typical K⁺ channel inhibitor, tetraethylammonium chloride (TEA, 50 μM), was used to reduce NaCl-elicited K⁺ efflux (Lu et al., 2013; Lang et al., 2014). In our study, control and NaCl-treated roots were treated with these inhibitors for 30 min before the flux recordings. In addition, two series of experiments (described below) were carried out to determine the involvement of Ca²⁺, H₂O₂, NO, and eATP in regulating Na⁺ and K⁺ fluxes and gene expression in NaCl-treated G. uralensis roots.

We added exogenous agonists, CaCl₂ (10 mM), H₂O₂ (10 mM), the NO donor, sodium nitroprusside (SNP, 100 μM), and ATP-Na₂ (300 μM), and measured the effects on NaCl-induced Na⁺ and K⁺ fluxes in young roots of G. uralensis. The chemicals were added to one-quarter-strength nutrient solution in the presence and absence of NaCl (100 mM). Control plants treated with or without salt were cultured in nutrient solution without the application of the chemicals mentioned above. The steady-state fluxes of K⁺ and Na⁺ were recorded along the root axis after 24-h NaCl treatments.

We also examined the expression levels of genes involved in salt transport and signaling after salt and agonist (Ca²⁺, H₂O₂, SNP, and ATP) treatments. Specifically, we examined expression of the PM H⁺-ATPase gene, GuAHA; the PM Na⁺/H⁺ antiporter gene, GuSOS1; the salt overly sensitive 3 gene, GuSOS3; the calcineurin B-like protein (CBL)-interacting protein kinase gene, GuCIPK; the respiratory burst oxidase homolog protein D gene, GuRbohD; the nitrate reductase gene, GuNIR; and the mitogen-activated protein kinases 3 and 6 genes, GuMAPK3 and GuMAPK6.

Control and NaCl (100 mM, 24 h)-stressed G. uralensis seedlings were treated with or without pharmacological agents for 30 min. These agents were: LaCl₃, an inhibitor of the PM Ca²⁺ channel (5 mM); DMTU, a ROS scavenger (5 mM); cPTIO, a scavenger of NO (300 μM); and PPADS, an antagonist of animal PM P2 receptors (300 μM) (Sun et al., 2010a,b; Chen et al., 2013; Zhao et al., 2016). Next, young roots with apices of 2.0–3.0 cm were sampled and equilibrated in measuring solution for 30 min. Then, steady-state fluxes of K⁺ and Na⁺ along the root axes were recorded in plants after treating with NaCl and antagonist (LaCl₃, DMTU, cPTIO, and PPADS). We also examined the abundances of GuAHA and GuSOS1 transcripts in these roots.

We used the NMT technique (NMT-YG-100, Younger United States LLC, Amherst, MA, United States) to measure the net Na⁺, K⁺, and H⁺ fluxes in G. uralensis roots. The microelectrodes were prepared and calibrated as previously described (Sun et al., 2009a,b; Lang et al., 2014).

After roots were exposed to NaCl treatment, with either an agonist (Ca²⁺, H₂O₂, SNP, and ATP) or an antagonist (amiloride, sodium orthovanadate, TEA, LaCl₃, DMTU, cPTIO, and PPADS), root segments with 2.0–3.0 cm apices were selected and washed two or three times with redistilled water. When placed in a buffer with a lower Na⁺ concentration, the preloaded Na⁺ would diffuse from the surface of salt-stressed roots. To decrease the effect of this excess salt release on flux recordings, roots were equilibrated prior to flux recordings in a measuring solution (0.1 mM NaCl, 0.1 mM MgCl₂, 0.1 mM CaCl₂, and 0.5 mM KCl) for 30 min. The concentrations of Ca²⁺ and K⁺ in the measuring solution were set to 0.1 and 0.5 mM, respectively (Li et al., 2012), to reduce interference from Ca²⁺ and K⁺ on the Na⁺ electrodes (Cuin et al., 2011). The pH of the measuring solution was adjusted to 5.7 with HCl and KOH.

After equilibration, roots were immobilized on the bottom of a measuring chamber with 10 ml of fresh measuring solution. Flux measurements were started at 200 μm from the root apex and conducted along the root axis, up to 2700 μm from the root apex, at intervals of 200 or 300 μm (vigorous ion fluxes were typically observed at the apical regions; Lu et al., 2013; Lang et al., 2014). A 6–8 min continuous recording was performed at each measuring point in the apical zones. Five or six individual seedlings were measured from each treatment group.

To evaluate the NaCl-induced Na⁺ distribution in G. uralensis roots and leaves, we used a specific fluorescent probe, CoroNa-Green AM (Sun et al., 2012a). Two-week-old seedlings were exposed to 0 or 100 mM NaCl for 6, 12, or 24 h. Then, the roots and leaves were exposed to CoroNa-Green AM (20 μM) for 2 h in a 5 mM Mes-KCl loading buffer (pH 5.7). Cellular Na⁺ was visualized with a Leica SP5 confocal microscope (Leica Microsystems GmbH, Wetzlar, Germany). The confocal settings were as follows: excitation 488 nm, emission 510–530 nm, frame = 512 × 512.

In G. uralensis roots, we used specific fluorescent probes to detect cellular signal contents. We used Rhod-2 AM (Biotium) to detect cytosolic Ca²⁺ (Sun et al., 2012a; Zhang et al., 2015); H₂DCF-DA (Eugene) to detect H₂O₂ (Sun et al., 2010a,b); and DAF-FM DA (Eugene) to detect NO (Sun et al., 2012a). Briefly, young roots were exposed to 0 or 100 mM NaCl for 30 min. Then, the roots were transferred to a 5 mM Mes-KCl loading buffer (pH 5.7) containing 2 μM Rhod-2 AM, 50 μM H₂DCF-DA, or 10 μM DAF-FM DA. The staining was performed in the dark for 1 h at room temperature. Next, the roots were washed 4–5 times with Murashige and Skoog (MS) liquid medium prior to confocal microscope measurements. The confocal settings were as follows: excitation 488 nm, emission 510–530 nm for H₂DCF-DA and DAF-FM DA; and excitation 543 nm, emission 570–590 nm for Rhod-2 AM (frame = 512 × 512).

Extracellular ATP levels were monitored with the Enlighten ATP assay system bioluminescence kit (Promega, Madison, WI, United States; Sun et al., 2012a; Deng et al., 2015). Briefly, G. uralensis roots were exposed to 0 or 100 mM NaCl at room temperature. The liquid culture medium of control and NaCl-treated roots was sampled at 0, 5, 10, 20, 40, 60, 120, and 240 min, then immediately frozen in liquid nitrogen. eATP was measured in an assay with luciferin-luciferase Turner Designs Modulus^TM Microplate Multimode Reader (Promega Corp., Madison, WI, United States). The eATP levels were calculated, based on a standard curve created by measuring a linear range (0.01–100 nM) of standard eATP concentrations (Sun et al., 2012a; Deng et al., 2015).

The transcription levels of genes related to the PM Na⁺/H⁺ transport system and salt signaling were evaluated in salt-stressed plants. Quantitative real-time PCR assays were conducted according to Deng et al. (2015) with some modifications. Briefly, total RNA was isolated from G. uralensis roots with TRIzol reagent (Invitrogen). DNA was eliminated by treating for 0.5 h with DNase I (Promega). An aliquot of purified RNA (1 μg) was used as template for first strand cDNA synthesis with M-MLV reverse transcriptase (Promega) and oligo (dT) primers. Specific primers for GuAHA, GuSOS1, GuSOS3, GuCIPK, GuRbohD, GuNIR, GuMAPK3, and GuMAPK6 were designed, based on homologous sequences found in Populus trichocarpa or Arabidopsis. Forward and reverse primers are listed in Supplementary Table 1. Amplification was performed as described by Ding et al. (2010): 95°C for 5 min, followed by 32 cycles of 94°C for 30 s, 55°C for 30 s, and finally, 72°C for 30 s, with a final step of 72°C for 10 min. The transcripts of target genes were normalized to the expression level of the G. uralensis β-actin 2 gene (GuACT2), and relative expression was calculated with the 2^-ΔΔC_T method (Livak and Schmittgen, 2001). Each experiment was replicated at least three times, and mean values are shown.

Ion fluxes were evaluated with JCal V3.0, which was created by Yue Xu¹. In the present study, positive values denote cation efflux and negative values denote cation influx. All experimental data were processed with SPSS 17.0 for statistical tests. Data were subjected to an Analysis of Variance (ANOVA), and comparisons between means were performed with Duncan’s multiple range test. P-values less than 0.05 were considered statistically significant.

The Na⁺ concentrations in roots and leaves of G. uralensis were detected with a Na⁺-sensitive fluorescent dye, CoroNa-Green AM. The Na⁺ fluorescence in roots and leaves increased with the duration of salt exposure (6, 12, and 24 h) (Figure 1). Intracellular Na⁺ was detected as a bright green fluorescence, which was typically observed in vacuoles (Figure 1). However, Na⁺ levels in roots were 1.69- to 2.40-fold higher than that in leaves over the observation period (Figure 1). This result indicated that G. uralensis roots could take up and accumulate high Na⁺ within a short period of salt treatment. Therefore, the roots were used to evaluate the effects of salt signaling molecules on ion fluxes and gene transcription.

Rhod-2 AM, H₂DCF-DA, and DAF-FM DA, respectively, were used to detect cytosolic Ca²⁺, H₂O₂, and NO elicited by NaCl in G. uralensis roots (Sun et al., 2010a,b, 2012a). Confocal assays (Figure 2) revealed that cytosolic Ca²⁺ (color: pseudo-red), H₂O₂ (color: pseudo-green), and NO (color: pseudo-green) significantly increased by 71–111% after a 30 min salt shock. Similarly, in an ATP-bioluminescence assay, NaCl caused a marked rise in eATP after 20 min of stress, and the peak level occurred at 40 min of stress (Figure 3).

A short period of NaCl exposure caused cellular Na⁺ accumulation, which was more pronounced in roots than in leaves (Figure 1). The buildup of Na⁺ in root cells resulted in remarkable increases in cytosolic Ca²⁺, H₂O₂, NO, and eATP (Figures 2, 3). The rapid increase of these signaling molecules indicated that G. uralensis roots could sense NaCl stress, and they set into motion a wide range of cellular processes required for salt adaptation (Chen and Polle, 2010; Chen et al., 2014; Polle and Chen, 2015). Accordingly, our NMT and RT-qPCR data revealed that eATP, Ca²⁺, H₂O₂, NO, and their interactions played crucial roles in regulating ion fluxes and gene transcription (Figures 4–10). These findings were similar to findings from our previous study in a salt-resistant poplar, P. euphratica (Sun et al., 2010a,b, 2012a).

Clear interactions occurred between these stress signals to accelerate the transcription of salt-adaptive signaling pathway genes in G. uralensis roots. Ca²⁺ increased the GuSOS3 expression (Figure 10), thus leading to enhanced Na⁺ extrusion via the SOS-signaling pathway (Zhu, 2001). In NaCl-treated roots of G. uralensis, H₂O₂, NO, or eATP, promoted the transcription of GuSOS3/GuCIPK (Figure 10), which indicated that these stress signals predominantly activated the Ca²⁺-SOS signaling pathway. A previous study in P. euphratica cells showed that exogenously applied H₂O₂ increased Ca²⁺ influx, which led to elevated cytosolic Ca²⁺ (Sun et al., 2010b). Based on our present results in G. uralensis roots, we suggest that H₂O₂ increased cytosolic Ca²⁺, which then mediated PM Na⁺/H⁺ antiport upregulation via the SOS-signaling pathway (Zhu, 2001, 2016). Furthermore, we found that NO enhanced the transcription of GuSOS3/GuCIPK in NaCl-stressed liquorice roots (Figure 10). Thus, NO-simulated Ca²⁺-SOS signaling would promote Na⁺ efflux and alleviate cellular Na⁺ toxicity in G. uralensis. Similarly, in the secretor mangrove, A. corniculatum, NO enhanced Na⁺ efflux elicited by Ca²⁺ (Lang et al., 2014).

Extracellular ATP signaling is a novel player in salt-stress acclimation. We found that eATP increased the expression of GuSOS3, GuCIPK, GuRbohD, GuNIR, GuMAPK3, and GuMAPK6 (Figure 10). Moreover, GuMAPK3 and GuMAPK6 expression levels were enhanced by H₂O₂ in salinized G. uralensis roots (Figure 10). This indicated that eATP interacted with H₂O₂ and Ca²⁺ signaling to maintain K⁺/Na⁺ homeostasis. Previously, eATP was shown to interact with H₂O₂ and Ca²⁺ to increase Na⁺ extrusion in two mangrove species, Kandelia obovata and A. corniculatum (Lang et al., 2014). In P. euphratica cells, eATP signaling was mediated by H₂O₂ and cytosolic Ca²⁺ in the salt response (Sun et al., 2012a). Accordingly, eATP is thought to bind P2-like receptors in the PM (Choi et al., 2014), which leads to an increase in H₂O₂ and a transient elevation in cytosolic Ca²⁺ (Jeter et al., 2004; Demidchik et al., 2009; Sun et al., 2012a). Thus, eATP could initiate the H₂O₂ and Ca²⁺ signaling cascades and cause an increase in Na⁺/H⁺ exchange across the PM of G. uralensis roots under NaCl stress.

Extracellular ATP also increased the expression of GuNIR (Figure 10). This finding indicated that NO was a downstream component of eATP signaling. Similarly, in P. euphratica cells, NO was triggered by eATP, although NO played a negligible role in eATP-stimulated cell death (Sun et al., 2012b). There are species–specific interactions between eATP and NO in the mediation of K⁺/Na⁺ homeostasis (Lang et al., 2014). In this study, eATP signaling appeared to be mediated by NO in G. uralensis roots (Figure 10). However, in the non-secretor, K. obovata, NO was redundant in the presence of eATP, because eATP alone exerted a pronounced effect on Na⁺/H⁺ antiporters (Lang et al., 2014).