Ectopic Expression of Gs5PTase8, a Soybean Inositol Polyphosphate 5-Phosphatase, Enhances Salt Tolerance in Plants

Inositol polyphosphate 5-phosphatases (5PTases) function in inositol signaling by regulating the catabolism of phosphoinositol derivatives. Previous reports showed that 5PTases play a critical role in plant development and stress responses. In this study, we identified a novel 5PTase gene, Gs5PTase8, from the salt-tolerance locus of chromosome 3 in wild soybean (Glycine soja). Gs5PTase8 is highly up-regulated under salt treatment. It is localized in the nucleus and plasma membrane with a strong signal in the apoplast. Ectopic expression of Gs5PTase8 significantly increased salt tolerance in transgenic BY-2 cells, soybean hairy roots and Arabidopsis, suggesting Gs5PTase8 could increase salt tolerance in plants. The overexpression of Gs5PTase8 significantly enhanced the activities of catalase and ascorbate peroxidase under salt stress. The seeds of Gs5PTase8-transgenic Arabidopsis germinated earlier than the wild type under abscisic acid treatment, indicating Gs5PTase8 would alter ABA sensitivity. Besides, transcriptional analyses showed that the stress-responsive genes, AtRD22, AtRD29A and AtRD29B, were induced with a higher level in the Gs5PTase8-transgenic Arabidopsis plants than in the wild type under salt stress. These results reveal that Gs5PTase8 play a positive role in salt tolerance and might be a candidate gene for improving soybean adaptation to salt stress.


Introduction
Cultivated soybean (Glycine max) is an important economic crop as a main source of dietary protein and oil. However, its development and agricultural productivity could be severely restricted by salt stress [1,2]. Unfortunately, around 20% irrigated land is salt affected and the salt-affected regions continue to expand [3,4]. Thus, it is necessary to improve salt tolerance in cultivated soybean. On the other hand, wild soybean (Glycine soja), the close related species of Glycine max, could confer higher tolerance to salinity or other stresses. Wild soybean could supply resistant gene resources for molecular breeding in soybean. It is meaningful to identify new salt-tolerance genes from Glycine soja and characterize their molecular mechanism.
The 5PTase genes are widespread among plants. For example, there are 15 members in Arabidopsis thaliana, 21 in rice and 39 in soybean [13]. Until now, the knowledge on the functions of plant 5PTases is still limited, and most is from the model plant Arabidopsis by examining the plants containing gain or loss of function in a certain 5PTase. At5PTase7 and At5PTase9 have been identified to take a crucial role in salt tolerance [14,15]. Mutations in At5PTase7 or AtPTase9 would result in increased salt sensitivity, whereas overexpression of At5PTase7 or AtPTase9 would enhance salt tolerance in Arabidopsis [14,15]. Both of the At5ptase7 and Atptase9 mutants would disturb reactive oxygen species (ROS) production and the expression of salt-responsive genes. Besides, the Atptase9 mutants also showed reduced Ca 2+ influx and decreased fluid-phase endocytosis, suggesting At5PTase7 and AtPTase9 have non-redundant roles in the regulation of plant salt tolerance [14,15].
On the other hand, GmSAL1, a soybean inositol 1-phosphatases, has been found to be involved in salt and drought response by the hydrolysis of inositol-1,4,5-trisphosphate (InsP 3 ) [16], suggesting the inositol signaling would also play a role in stress responses in soybean. The components including 5PTases, involved in inositol signaling, would be good candidates for soybean breeding of salt tolerance as well. However, the function of most 5PTases remains unknown in soybean.
Recently, a salt-tolerance locus has been identified in chromosome 3 of Glycine soja and the salt-tolerant Glycine max varieties [17][18][19]. In order to dig out more salt-tolerance genes, we performed post-genomonic analyses on this locus. In this study, a novel soybean 5PTase gene, Gs5PTase8, has been identified from this salt tolerance locus due to its high induction by salinity in the Glycine soja accession W05. The analyses of sequences showed that Gs5PTase8 contains the two conserved domain as the other known 5PTases, indicating it is a potential 5PTase. Its effects on plant salt responses were monitored via ectopic expression in BY-2 cells, soybean hairy roots, and Arabidopsis thaliana. Its impacts on the activities of antioxidant enzymes, ABA signaling, and stress-responsive gene expression were also determined here.

Gs5PTase8 Encoding a Putative Inositol Polyphosphate 5-Phosphatas is Induced by Salt Stress
A novel salt tolerance locus has been identified in the chromosome 3 of soybean previously [17][18][19]. From the transcriptome data of soybean under salt stress [20], we identified a salt-induced gene, Gs5PTase8 (XM_028369920), in the salt tolerance locus of the G. soja accession W05. To verify the induction of Gs5PTase8 by salinity, the expression analyses were performed in W05 by qRT-PCR ( Figure 1A). The results showed that the expression of Gs5PTase8 was highly up-regulated, especially at 24 h after salt treatment in both leaves and roots, suggesting Gs5PTase8 would be involved in salt responses.

Ectopic Expression of Gs5PTase8 Increased Salt Tolerance in Plants
In order to investigate its function in plant salt responses, the effects of Gs5PTase8 were assessed by the utilization of the three transformation systems of BY-2 cell lines, soybean hairy roots and Arabidopsis thaliana. The functional analysis in salt responses was performed in BY-2 cells containing the transgenes of Gs5PTase8 (Figure 2). The expression of the transgenes was confirmed by qRT-PCR. The dead cells were visualized by trypan blue staining. Under normal growth conditions, all the cell lines have around 98% survival rates. Under 100 mM NaCl treatments, all the transgenic lines of Gs5PTase8 exhibited significantly higher survival rates than the wild type (WT) and the vector-only control line (V). The soybean cotyledons were infected for hairy root induction by the wild type of A. rhizogenes stain K599 (CK), K599 containing the empty binary vector pTEV8 (V) or K599 containing the recombinant construct V7-Gs5PTase8 (Gs5PTase8), respectively. The transformation of hairy roots were confirmed by PCR genotyping ( Figure 3A). Nine roots were positive transformants in 10 roots, indicating that the transformation was mainly successful. All the hairy roots grew well with no obvious difference under normal growth condition, whereas the growth of Gs5PTase8 transformed hairy roots was inhibited at a less level with more fresh weight than that of the controls under 150 mM NaCl ( Figure 3). The results revealed that the overexpression of Gs5PTase8 alleviated salt stress in soybean hairy roots. Sample means with standard errors (SE) were calculated with three replicates each containing 5-6 soybean cotyledons. The significant tests were performed using ANOVA (Duncan's test, p < 0.05). The significant differences were indicated by different letters above the bars.
Three independent homozygous transgenic Arabidopsis lines of Gs5PTase8 were obtained for functional analyses under salt stress ( Figure 4A). Under normal growth conditions, the seeds of all the lines germinated at a similar rate in 2 days after stratification and there are no significant growth difference between them. Under salt treatment, the germination rates of all the Gs5PTase8-transgenic lines were significantly higher than the rates of the wild type (Col-0), especially at 3.5 d after stratification ( Figure 4B,C). Moreover, when the 3-day seedlings were treated by 175 mM NaCl for one week, the results showed that the Gs5PTase8-transgenic plants had significantly higher survival rates and relative chlorophyll contents comparing to the wild type ( Figure 4D-F). The results demonstrated that the ectopic expression of Gs5PTase8 could increase salt tolerance in Arabidopsis as well, in consisted with its function in BY-2 cells and soybean hairy roots.

Enhanced Activities of Antioxidant Enzymes in Gs5PTase8-Transgenic Arabidopsis
The evidences suggested that salinity would lead to oxidative stress [3]. To assess the contribution of antioxidant enzymes in reduction of oxidative stress, the activities of catalase (CAT), ascorbate peroxidase (APX), peroxidase (POD), and superoxide dismutase (SOD) were measured in the wild type and the three Gs5PTase8-transgenic Arabidopsis lines. As shown in Figure 5, the application of salt stress increased the four ROS scavenging enzymes more or less. Among them, the activities of CAT and APX were significantly higher in the Gs5PTse8-transgenic lines than the wild type. It demonstrated that the overexpression of Gs5PTse8 could enhance the activities of CAT and APX under salt treatment, probably to alleviate the following oxidative stress caused by salinity.

Altered ABA Sensitivity of Gs5PTase8-transgenic Arabidopsis
Since 5PTase enzymes have been observed to be involved in ABA signaling pathway [7], ABA sensitivity were detected with the Arabidopsis plants over-expressing Gs5PTase8 in seed germination ( Figure 6). The germination rates of all the plant lines declined when treated with the increased ABA for around 3.5 d after incubation at 21 • C. In general, the Gs5PTase8-transgenic lines showed less sensitive to ABA-elicited inhibition of seed germination compared to the wild type. With exposure to over 5 µM ABA, all the seeds hardly germinated. To further verify the differences on ABA sensitivity between the transgenic lines and the wild type, we used 1 µM ABA to perform the time course tests. It illustrated that the Gs5PTase8-transgenic plants germinated at a higher speed compared to the wild type. The ectopic expression of Gs5PTase8 could impart ABA insensitivity on seed germination, suggesting it might be involved in ABA signaling.

Expression of the Stress-Responsive Genes in Gs5PTase8-Transgenic Arabidopsis
To investigate whether the ectopic expression of Gs5PTase8 affects the stress-responsive genes, the expression levels of AtABA1, AtRD22, AtRD29A, and AtRD29B were detected in the Gs5PTase8-transgenic Arabidopsis lines and the wild type by qRT-PCR (Figure 7). AtABA1 participates in ABA synthesis, and the other three, AtRD22, AtRD29A, and AtRD29B have been showed to be induced by salt and other abiotic stresses [21,22]. All the tested genes were up-regulated in the first 3 or 6 h post salt treatment, then down-regulated and again up-regulated at 24 h post salt treatment. The expression of AtABA1 seemed not to be altered significantly by ectopic expression of Gs5PTase8 in Arabidopsis, indicating Gs5PTase8 might not affect the ABA biosynthesis. On the other hand, the expression levels of the other three stress-responsive genes, AtRD22, AtRD29A, and AtRD29B, were induced more in the transgenic lines compared to the wild type under salt stress, suggesting Gs5PTase8 might take an important role in the regulation of those stress-responsive genes.  Effects of Gs5PTase8 on the expression levels of the stress-responsive genes. The three-week seedlings were treated by 100 mM NaCl in a hydroponic system for 0 h, 1 h, 3 h, 6 h, 12 h, and 24 h. The expression of AtABA1, AtRD22, AtRD29A, and AtRD29B were analyzed by qRT-PCR in the three independent Gs5PTase8-transgenic lines (Gs5PTase8 #1, #2, #3) and the wild type (Col-0). The data were normalized to the reference gene ACT2. The value of the samples collected from Col-0 at 0h was set at 1. The data were analyzed using ANOVA by the Dunnett's post hoc test. Asterisk indicates a significant difference between the wild type and the transgenic lines at the same time point (p < 0.05).

Subcellular Localization of Gs5PTase8
To examine the subcellular localization of Gs5PTase8, the recombinant construct of Gs5PTase8-GFP under the control of the 35S promoter were transformed into onion epidermal cells by particle bombardment (Figure 8). The 35S::GFP construct was used as the control. Before plasmolysis, the GFP signal appeared throughout the entire cell except the vacuole for 35S::GFP, whereas it was observed in the nucleus and plasma membrane for 35S::Gs5PTase8-GFP ( Figure S1). After plasmolysis, the signals the Gs5PTase8-GFP protein were observed in the nuclear, plasma membrane, cell wall and especially in the apoplast (probably Hechitan strands), compared to the GFP alone protein. It revealed that Gs5PTase8 is expressed in the nucleus and plasma membrane. Besides, Gs5PTase8 targets to apoplast particularly, suggesting it might function in apoplastic pathways.

Discussion
Inositol polyphosphate 5-phosphatases (5PTases) are one of the key components in inositol pathway [8,9]. Previous studies suggested that they could play an important roles in plant development and adaption to environmental stresses [10,11]. They have been identified to be widely distributed in all kingdoms of organisms [13]. However, only a few plant 5PTases have been investigated, and most are from the model plant, Arabidopsis thaliana. In this study, a novel Gs5PTase8 was cloned from Glycine soja and characterized for its function in salt responses. The sequence analysis illustrated that Gs5PTase8 contains the two conserved domain of the reported 5PTases (Figure 1), suggesting its potential function as an enzyme to hydrolyze the phosphate at the 5' site of the inositol ring.
Based on the analyses of gene expression, Gs5PTase8 is highly induced in salt-tolerant accession W05. Moreover, the further functional analyses showed that ectopic expression of Gs5PTase8 could also confer salt tolerance in transgenic BY-2 cells (Figure 2), soybean hairy roots ( Figure 3) and Arabidopsis (Figure 4). The activities of two antioxidant enzymes, CAT and APX, were increased in the Gs5PTase8-transgenic Arabidopsis lines under salt treatment ( Figure 5), suggesting Gs5PTase8 might relieve the oxidative damage by salt stress and thus increase salt tolerance. All these revealed that Gs5PTase8 would be a positive regulator in salt tolerance, which might be a good candidate for soybean breeding.
5PTases have been found to affect ABA signaling pathway, which could play a vital role in plant adaption, especially to abiotic stresses [7,23]. We have determined the effect of Gs5PTase8 on ABA signaling. The results revealed that the overexpression of Gs5PTase8 in Arabidopsis caused ABA insensitivity on seed germination (Figure 6), which is similar as the previous reports on the mammalian type I 5PTase, At5PTase1, At5PTase2, and At5PTase6. The transgenic Arabidopsis plants expressing the mammalian type I 5PTase exhibited less sensitivity to ABA treatment and increased drought tolerance comparing to the wild type [24]. At5PTase1 and At5PTase2 could alter ABA signaling via the hydrolysis of inositol-1,4,5-trisphosphate (InsP 3 ) [25,26]. Their mutants showed a ABA hypersensitivity phenotype with faster seed germination and longer hypocotyls when grown in dark [27]. Similarly, the cvp2 mutant (deficient in At5PTase6), with open vein network in leaf, also showed an increased sensitivity to ABA in seed germination [28]. But mutation in At5PTase13 caused ABA insensitivity [29,30]. Thus, it suggested that Gs5PTase8 might also impact ABA signaling for its function.
Among the reported Arabidopsis 5PTases, At5PTase7, and At5PTase9 have been characterized as a positive regulator in salt tolerance [14,15]. The overexpression of At5PTase7 or At5PTase9 displayed enhanced salt induction of the salt-induced marker genes, AtRD22 and AtRD29A [14,15]. We monitored the expression levels of three known stress-marker genes of AtRD22, AtRD29A, and AtRD29B [31] under salt stress. The stress-marker genes could be up-regulated under abiotic stress or ABA treatment. Our observations showed that the expression of AtRD22, AtRD29A, and AtRD29B exhibited increased salt induction in Gs5PTase8-transgenic lines compared to the wild type as well (Figure 7). The subcellular localization analysis demonstrated that the fluorescent signal of Gs5PTase8-GFP could be detected in the nucleus. Gs5PTase8 might be involved in the regulation of stress-responsive gene expression by itself or via the ABA signaling pathway.
Except for the nuclear localization, Gs5PTase8-GFP was detected in plasma membrane with a relatively strong signal in the apoplast as well (Figure 8). The apoplast is the portion outside the cell membrane, including the cell wall and the intercellular space, with the function as a barrier and a linker between the environment and the protoplast [32]. Due to its apoplastic localization, we supposed that Gs5PTase8 might hydrolyze the membrane-bound phosphoinositide and be involved in the lipid signaling. However, there are no exploration on the substrates of Gs5PTase8 and its closest Arabidopsis homolog, At5PTase8, yet. Previous reports showed that various components exist in the apoplast, including ions, reactive oxygen species (ROS), and secreted proteins [32]. The apoplastic ROS and the scavenging enzymes have been identified to provide the signals for stress responses [33]. At5PTase7 and At5PTase9 have been identified to increase the production of ROS in the initial period of salt treatment, thus probably triggering the scavenging pathway to tolerate salt stress [14,15]. We have detected the activities of four antioxidant enzymes, CAT, APX, POD, and SOD at 24 h after salt treatment. Among them, CAT and APX were activated with a significant higher level in the Gs5PTase8-transgenic Arabidopsis plants than the wild type under salt treatment ( Figure 5), suggesting that Gs5PTase8 would regulate ROS signaling probably in the apoplast under salt stress.
In conclusion, we identified a novel salt-induced Inositol polyphosphate 5-phosphatase gene, Gs5PTase8, in soybean and characterized its function in salt responses. The gain-of-function tests demonstrated that Gs5PTase8 conferred salt tolerance in transgenic BY-2 cells, soybean hairy roots and Arabidopsis. Overexpression of Gs5PTase8 increased the activity of CAT and APX, altered the ABA signaling pathway and up-regulated the expression of salt-responsive genes under salt stress, which might result in enhanced salt tolerance. Gs5PTase8 is localized not only inside the protoplast, but also the apoplast. This study provided a novel gene for improving salt tolerance in soybean and other plants. However, further investigation would be performed to reveal its substrates and its comprehensive roles in the ABA signaling and apoplastic pathways.

Plant Materials and Growth Conditions
The Glycine soja accession W05 (variety name: Mengjin1, salt tolerant) and Glycine max accession C08 (variety name: Union, salt sensitive) supplied by Dr. Honming Lam's lab, Chinese University of HongKong, were used for gene cloning and expression analyses in this work [17,34,35]. After soybean seeds were germinated on vermiculite for one week, the seedlings were transplanted to a hydroponic system of half-strength Hoagland's solution in a greenhouse [36].
Tobacco BY-2 cells [37] and Arabidopsis thaliana (ecotype: Columbia-0) were employed to obtain the transgenic lines for functional analyses. BY-2 cells were cultured in Murashige and Skoog (MS) media (Hopebio, Qingdao, China) supplemented with 0.025% KH 2 PO 4 and 3% sucrose at 25 • C in dark. Arabidopsis plants were grown on 1/2 MS medium or in hydroponic system of 1/10 MS solution under a 16 h light / 8 h dark cycle at 21 • C and 70%-80% humidity. The hydroponic system was used only for the gene expression analyses on Arabidopsis under salt treatment. And the plants were grown on 1/2 MS medium for all the other experiments.

Gene Expression Detected by qRT-PCR
For soybean, when the first trifoliate of the soybean seedling opened, the nutrient solution was replaced by the fresh 1/2 Hoagland's solution containing 0.9% NaCl (w/v) [17,20,38]. To extract the total RNA, the leaves and roots were collected and ground in liquid nitrogen separately at 0 h, 4 h, 24 h, 48 h, and 72 h after the start of salt treatments [20,38,39]. For Arabidopsis, three-week-old seedlings grown in the hydroponic system were transferred to the fresh 1/10 MS solution supplemented with 100 mM NaCl. The sampling time points were 0 h, 1 h, 3 h, 6 h, 12 h, and 24 h after the start of salt treatments. Three individual plants under the same treatment were pooled into one biological sample.
The total RNA was extracted using RNAiso PLUS reagent (TaKaRa Biotechnology Co. Ltd. Dalian, China) and the cDNA was synthesized using PrimeScript TM RT reagent Kit with gDNA Eraser (TAKARA, Dalian, China) according to the manufacturer's instructions. Quantitative reverse-transcription PCR (qRT-PCR) was performed using SYBR Premix Ex Taq TM II (Tli RNaseH Plus) (TAKARA, Dalian, China) with the CFX96 Touch TM Real-Time Detection System (Bio-Rad, Hercules, CA, USA). GmELF1b, AtACT2 and NtL25 were used as internal reference genes, respectively for soybean, Arabidopsis and tobacco BY-2 cells [14,40,41]. The gene specific primers are listed in Table S1. All the reactions were performed with three replicas. Relative gene expression was quantified using the 2 −∆∆Ct method [42].

Molecular Cloning and Sequence Analysis of Gs5PTase8
First-strand cDNAs were synthesized with the RNA extracted from the soybean treated by salt stress, using a PrimeScript TM II 1st strand cDNA Synthesis Kit (TaKaRa, Dalian, China) following the manufacturer's protocol. Phusion High-Fidelity DNA polymerase (ThermoFisher Scientific, USA) was employed for amplifying the coding sequences of Gs5PTase8 with the gene specific primers (Table S1) [35]. The thermocycling conditions for PCR were as follows: 98 • C 30 s; 35 cycles of 98 • C 10 s, 58 • C 30 s and 72 • C 60 s; 72 • C 10 min and 12 • C for hold. The amplicons with expected size were gel-purified and cloned to the pBluescript II KS (+) vector between the sites of XbaI and XhoI. The sequences of cDNAs were sequenced (BGI, Shenzhen) for verification. Then the cDNA was subcloned to the same restriction enzyme sites of the binary vector pTEV8 at the downstream of cauliflower mosaic virus 35S promoter [43]. The recombinant construct was named as V7-Gs5PTase8. After sequence verification, they were transformed into Agrobacterium rhizogenes strain K599, Agrobacterium tumefaciens strain GV3101 or LBA4404 respectively by a freeze-thaw method [44].
The BLASTP searches were performed to find the homologous sequences (https://blast.ncbi.nlm. nih.gov/Blast.cgi). Multiple sequence alignment was performed with the full length sequences of protein using the CLUSTALX2.1 program [45]. The phylogenetic tree was constructed using MEGA version 6 by the neighbor-joining method with a bootstrap value of 1000 repeats [46].

Cell Viability Assay of Transgenic BY-2 Cells
Tobacco (Nitcotiana tabacum) BY-2 cells [37] were transformed with A. tumefaciens stain LBA4404 containing the recombinant plasmid, V7-Gs5PTase8 by a co-cultivation method [47]. The positive transformants were selected on the MS medium containing 50 mg/L kanamycin and the expression of the transgene was confirmed by qRT-PCR.
Four-day-old BY-2 cell suspension cultures were treated with 100 mM NaCl for 20 h. The salt-treated and untreated cells were stained with 0.4% trypan blue (Sigma-Aldrich, St. Louis, MO, USA) for 5 min, and then were observed under a light microscope (Nikon Eclipse 80i, Tokyo, Japan). The images were captured by a CCD camera. A total of 200-300 cells were counted for each sample. Three replicates were performed for each sample.

Assays of Hairy Roots of Transformed Soybean Cotyledons under Salt Stress
Hairy root transformation was performed as previously described with minor modifications [48]. The wild type of A. rhizogenes stain K599 (CK), K599 containing the empty vector (V), and K599 containing the recombinant construct V7-Gs5PTase8 (Gs5PTase8) were employed to transform cotyledons of any salt-sensitive soybean accessions (such as C08 and Williams 82) for hairy root induction. A. rhizogenes stains were grown in lysogeny broth (LB) medium containing 50 mg/L kanamycin and 200 µM acetosyringone at 28 • C for 16 h. The soybean seeds were germinated with water at room temperature for 2-4 days. The cotyledons were cut off by a scalpel and scored by a syringe with needle. The wounds were infected with A. rhizogenes K599 for 15 min and then transferred to moist filter paper for co-cultivation in the dark at 25 • C for 5 days. Subsequently, the infected cotyledons were transferred to a growth chamber under 12 h light/12 h dark cycle at 25 • C. When the hairy roots were sprout, the infected cotyledons with the hairy roots of similar lengths (around 1 cm) were selected and treated with 1/2 Hoagland solution with or without 150 mM NaCl for 10 days. The hairy roots were photographed by a single lens reflex camera D7100 (Nikon Imaging (China) Sales Co., Shanghai, China) and the fresh weight of roots was measured. DNA was extracted respectively from each hairy root and the PCR reaction was performed with the NPTII primers (kanamycin selection marker of the binary vector pTEV8) to verify the positive transformation. The successful transformants of the empty vector and the recombinant construct would have a PCR product of 400 bps. The fresh weight data of those negative transformants were removed.

Generation of Transgenic Arabidopsis Lines
A. thaliana (ecotype: Col-0) plants were transformed by A. tumefaciens stain GV3101 containing the recombinant construct V7-Gs5PTase8 using the floral dip method [49]. The transgenic seeds were screened on MS medium with 50 mg/L kanamycin for positive transformants and at least 3 independent T 1 lines were selected out. The homozygous T 2 transgenic lines were obtained by selecting on MS medium with kanamycin and their progenies were used for further analysis. The expression of Gs5PTase8 was verified in the transgenic lines by qRT-PCR.

Phenotype Analyses of Transgenic Arabidopsis Lines
For germination analyses, Arabidopsis seeds were sowed on normal 1/2 MS plates or the medium supplemented with 175 mM NaCl concentrations or varied ABA concentrations (0.25 µM, 0.5 µM, 0.75 µM, 1 µM, 2 µM, 5 µM, or 10 µM) after surface-sterilization. The seeds were stratified at 4 • C in dark for 3 d before shifting to the growth chamber of 21 • C. Then, the germinated seeds were counted every 12 h in 6 days. To determine the best concentration for ABA sensitivity assay, around 150 seeds were used for each sample. In the other germination analyses, a total of around -500 seeds were used for each sample from three independent experiments.
For survival assays, the 3-day seedlings were transferred to the fresh 1/2 MS media with or without 175 mM NaCl for one week. A total of 75 plants were used for each sample from three replicates. The survival rate was calculated as the percentage of the survival plants after salt treatment. The chlorophyll content was extracted from 0.05 g of rosette leaves from the survival plants by being soaked in 1 mL 95% ethanol at 4 • C in dark for 24 h. The absorbance of the extraction was checked at the wavelength of 649 and 665 nm. Total chlorophyll concentration was calculated as the following equation: C total = C a + C b = ((OD 665 nm × 13.95−OD 649 nm × 6.88) + (OD 649 nm × 24.96−OD 665 nm × 7.32))/(sample weight) [50].

Assays for Antioxidant Enzyme Activities
The 10 d-seedlings were transferred to the 1/2 MS media with or without 150 mM NaCl for 24 h. Briefly, the fresh leaves of 0.1 g were ground in 2 mL of 0.2 M phosphate buffer saline (PBS, pH 7.8) containing 1% polyvinylpyrrolidone (PVP) on ice. The homogenate was centrifuged at 4000 rpm for half an hour at 4 • C and the supernatant was stored at 4 • C for the activity detection of catalase (CAT) [51], ascorbate peroxidase (APX) [52], superoxide dismutase (SOD) [53] and peroxidase (POD) [54]. The CAT activity was determined by monitoring decreases in absorbance at 240 nm due to the decomposition of H 2 O 2 . The APX activity was assessed by the decline of absorbance at 290 nm for the oxidation of ascorbate. The SOD activity was measured by detecting the absorbance at 560 nm for inhibiting the reduction of nitro blue tetrazolium (NBT) in light. The POD activity was assayed by detecting the absorbance at 470 nm due to the oxidation of guaiacol by H 2 O 2 . Each experiment was carried out at least three times.

Subcellular Localization of Gs5PTase8
The full length coding sequence of Gs5PTase8 without the stop codon was amplified and cloned into the XbaI and BamHI restriction enzyme sites of the binary vector pTEV8 at the downstream of 35S promoter. The GFP coding sequence was cloned behind Gs5PTase8 between the sites of BamHI and XhoI. The constructs were coated with gold particles and bombarded into the epidermal cells of onion (Bio-Rad PDS-1000/He system, Hercules, CA, USA). The onion sections were imaged at room temperature using a Leica SP8 X inverted confocal microscope with an Argon laser (Leica, Wetzlar, Germany). Green fluorescence protein (GFP) is excited at 488 nm and the emitted light is captured at 510-525nm. The images were captured digitally and processed using the Leica Application Suite Advanced Fluorescence Lite (LAS AF version: 2.6.3 build 8173). Plasmolysis was performed by incubating the samples in 25% sucrose for 5 min.