GmSAL1 Hydrolyzes Inositol-1,4,5-Trisphosphate and Regulates Stomatal Closure in Detached Leaves and Ion Compartmentalization in Plant Cells

Inositol polyphosphatases are important regulators since they control the catabolism of phosphoinositol derivatives, which are often signaling molecules for cellular processes. Here we report on the characterization of one of their members in soybean, GmSAL1. In contrast to the substrate specificity of its Arabidopsis homologues (AtSAL1 and AtSAL2), GmSAL1 only hydrolyzes inositol-1,4,5-trisphosphate (IP3) but not inositol-1,3,4-trisphosphate or inositol-1,4-bisphosphate.The ectopic expression of GmSAL1 in transgenic Arabidopsis thaliana led to a reduction in IP3 signals, which was inferred from the reduction in the cytoplasmic signals of the in vivo biomarker pleckstrin homology domain–green florescent protein fusion protein and the suppression of abscisic acid-induced stomatal closure. At the cellular level, the ectopic expression of GmSAL1 in transgenic BY-2 cells enhanced vacuolar Na+ compartmentalization and therefore could partially alleviate salinity stress.

Gain-of-function experiments showed that the overexpression of At5PTase1 resulted in a reduction of the stomatal response toward light and ABA treatment, presumably due to a lowered IP 3 level [8]. The overexpression of At5PTase2 also showed decreased sensitivity toward ABA inhibitory effects on seed germination [10]. These two pieces of evidence point to the regulatory role of inositol 5-phosphatases in ABA signaling. On the other hand, the overexpression of At5PTase7 in A. thaliana enhanced salt tolerance while the knock-out mutant of At5PTase7 aggravated salt damage due to a loss in the ability to induce reactive oxygen species that are required to trigger the expression of ABA-responsive genes [13].
The in vivo functions of AtSAL1 on stress responses are still controversial. The ectopic expression of the Arabidopsis gene AtSAL1 in yeast conferred lithium tolerance, similar to the effects of overexpressing the endogenous ScHal2 gene in yeast [9]. It was an expected result since AtSAL1 and ScHal2 are homologues. However, it was subsequently reported that the overexpression of AtSAL1 in A. thaliana did not elevate NaCl tolerance [15]. In fact, AtSAL1 is a negative regulator of drought tolerance in A. thaliana, since a mutation in the AtSAL1 gene led to enhanced drought tolerance [16].
In this work, we identified the coding sequence of the SAL1 homologue in soybean, GmSAL1. We also characterized its substrate specificity and demonstrated its effects on various stress responses in plant cell through its function as an inositol polyphosphatase.

Cloning of GmSAL1
Soybean (Glycine max L. Merr.) plants were grown in a greenhouse. For experiments leading to the cloning of GmSAL1, the seeds were first germinated in sand irrigated with water. After the opening of the first trifoliate, the seedlings were irrigated with modified Hoagland's solution [17]. NaCl treatment was performed using 150 mM NaCl for 3 d.
Total RNA samples were obtained using a modified phenol extraction protocol [18]. The first-strand cDNA was then obtained from the total RNA by reverse transcription using the Moloney murine leukemia virus-reverse transcriptase (Gibco BRL, Grand Island, NY, USA) according to the manufacturer's manual.
The GmSAL1cDNA clone covering the entire coding region was amplified from the first-strand cDNA samples using the primers 5'CGCCGCTGACACTAATCGTTT3' and 5'CGAGCCGACAACAAAGTTAGC3'. The DNA sequence information of GmSAL1 was deposited into GenBank (accession number: EF637045).

DNA sequencing and sequence analysis
DNA sequencing was performed using the ABI PRISM dRhodamine Terminator Cycle Sequencing Ready Reaction kit (PerkinElmer, Waltham, MA, USA) and analyzed by the Genetic Analyzer ABI Prism 3100 system, according to the manufacturer's protocol. Homologue searches were performed with Position-Specific Iterated Basic Local Alignment Search Tool (PSI-BLAST) (http://www.ncbi.nlm.nih.gov/BLAST/). Multiple alignments were performed using the ClustalW program [19] in the BioEdit package (ver. 7.0.5.3).

Gene expression under stress
Soybean seeds were germinated in vermiculite with water in a greenhouse. After one week, seedlings were transferred to hydroponic cultures with half-strength Hoagland's solution [20]. Just after the emergence of the first trifoliate, the plants were transferred to half-strength Hoagland's solution supplemented with 60mM, 125mM, and 185mM NaCl, and 10%, 14%, and 16.5% (w/v) polyethylene glycol (PEG)-6000 for 24h. Treated sample tissues were harvested and frozen in liquid nitrogen for total RNA extraction. A total of 20μg RNA for each sample was used for northern blot analysis. The osmolarity of NaCl and PEG solutions was measured by Advanced TM Micro Osmometer (Model 3300; Advanced Instruments, Inc., Norwood, MA, USA).

Construction of GmSAL1transgenic A. thaliana lines
Transgenic A. thaliana ectopically expressing GmSAL1 was constructed according to a previous report [22]. The cDNA of GmSAL1 was cloned into a binary vector [25] downstream of the cauliflower mosaic virus 35S promoter. Six-week-old A. thaliana (Col-0) plants were transformed by the vacuum infiltration method [26] using the Agrobacterium tumefaciens strain GV3101 (pMP90) transformed with the GmSAL1 construct. The expression of GmSAL1in the transformed A. thaliana was verified by real-time PCR ( Figure S1 in File S1).

Construction of A. thaliana lines expressing both PH PLCd -GFP and GmSAL1
Reciprocal crosses were performed between the PH PLCd -GFP and the GmSAL1 transgenic lines. Five-to six-week-old plants grown on soil were used. Mature flowers were detached from the pollen donor parent. Sepals, petals, and stamens of the flower buds of the pollen recipient parent were removed with a pair of fine forceps. Pollens of the donor parents were transferred to the stigma of the recipient flower bud. About 2-3 weeks after artificial crossing, seeds were harvested. After a few generations of self-fertilization, double homozygous lines were screened by PCR. For the line PH PLCd -GFP/ GmSAL1-1,GmSAL1 was the pollen donor. For the line PH PLCd -GFP/GmSAL1-2, PH PLCd -GFP was the pollen donor.

Expression and purification of the GmSAL1 protein in E. coli
The coding sequence of GmSAL1 was amplified by PCR using Pfx polymerase (Invitrogen, Carlsbad, CA, USA) with the following primers: 5'CCCCAGATCTATGCCTTACGAGAAGGAATTC3' and 5'CCCGCAATTGTCACAAGGATGAAACTTTC3'. The amplified GmSAL1cDNAwas subcloned into pGEX-2T vector (GE Healthcare, Chalfont St Giles, UK) to form a fusion protein with the glutathione S-transferase (GST). The GST-GmSAL1 construct was then introduced into the E. coli strain BL21 (DE3) cells. The expression of the recombinant protein was induced by the addition of 0.1mM IPTG to the E. coli culture followed by incubation for 4h, before the cells were washed and resuspended in the lysis buffer (50mM Tris-HCl pH 7.5, 100mM NaCl and 1mM phenylmethylsulfonyl fluoride). The soluble protein fraction was obtained by sonication and subsequent centrifugation. Soluble GmSAL1 protein was affinity-purified by GST-Trap column (Amersham Biosciences, Piscataway, NJ, USA) and then dialyzed overnight in the enzyme assay buffer (25mM Tris-HCl pH 7.5, 1mM MgCl 2 ) with 2mM dithiothreitol.

Enzyme assays and determination of the K m values
Phosphatase assays were performed according to previous reports [9,27,28] with slight modifications. A 100μl reaction mixture containing the recombinant protein and substrate in the assay buffer (25mM Tris-HCl pH7.5 and 1mM MgCl 2 ) was incubated at 37°C for 30min and the released inorganic phosphate was quantified at 650nm using a 96-well microtiter plate reader (Tecan Group Ltd., Seestrasse, Männedorf, Switzerland). Protein concentrations were determined by the Bradford method [29]. The K m for IP 3 hydrolysis was determined by measuring the rate of hydrolysis at the following substrate concentrations: 0.0125mM, 0.025mM, 0.05mM, 0.1mM, and 0.2mM.The K cat of IP 3 was calculated using the K m , protein concentration, and molecular weight of the recombinant protein. All substrates used were from Sigma-Aldrich Co. (St Louis, MO, USA) except I(1,3,4)P 3 and I(1,4)P 2 (Echelon Biosciences Inc., Salt Lake City, UT, USA).

Relative in vivo IP 3 levels in guard cells
The microscopic analysis of relative in vivo IP 3 level in guard cells was according to a previous report [30]. The lower epidermal of rosette leaves of 4-week-oldA. thaliana grown on soil at 22°C (16h light-8h dark cycle) was peeled off. The epidermal peels were immersed in buffer containing 50μM CaCl 2 , 5mM KCl, 10mM MES-Tris, (pH 6.15) for 2 h under constant light, before subjected to confocal microscopic analysis. Images were collected using Olympus FV1000(Ex: 488nm; Em: 510-525nm). The fluorescence signals were analyzed using the ImageJ program (ver. 1.371.44p) [31].

Stomatal aperture assay
The stomatal aperture assay was performed according to a previous report [8]. Leaves of 4-week-oldA. thaliana grown on soil at 22°C (16h light-8h dark cycle) were detached and incubated in a perfusion solution (50mMKCl, 10mM MES, pH 7.0) without supplements for 2h, followed by incubation in a perfusion solution with supplements for another 2h. In the control experiment, the perfusion solution was supplemented with 0.1% (v/v) MeOH (solvent of ABA). In the other two sets of experiments, perfusion solutionscontaining100μM ABA with or without 5mM CaCl 2 were used. The concentration of ABA employed was according to a previous report [8]. All incubations were conducted at 22°C under constant light. The differential interference contrast (DIC) images of guard cells were captured using a light microscope (Nikon Eclipse 80i). The stomatal aperture was measured using a digital ruler available in the software SPOT Advance (ver. 4.6, Diagnostic Instruments, Inc.).

Seed germination assay
The seed germination assay was performed according to a previous report [8].A. thaliana seeds were surface-sterilized, placed on half-strength MS agar plate (1% (w/v) sucrose supplemented with one of the following: 0.1% (v/v) MeOH alone, 2.5μM ABA in 0.1% (v/v) MeOH, or 4μM ABA in 0.1% (v/v) MeOH) and kept at 4°C in the dark for 2 d. Seeds were then allowed to germinate at 25°C under continuous light. The germination rate was calculated using 138 to 211 seeds from three independent experiments.

Establishment of transgenic tobacco BY-2 cell lines
The same recombinant construct used to transform A. thaliana was transformed into the tobacco BY-2 cells [32] using Agrobacterium (strain LBA4404) by a co-cultivation method [33]. After selecting the transformants on antibiotic-containing medium, PCR screening using gene-specific primers was performed to verify the successful integration of the transgene into the genome and real-time PCR was performed to verify the expression of the transgene in the transformed cell lines. Cells were grown in a liquid MS medium [32] at room temperature in the dark with mild agitation.

Microscopic analysis of Na + compartmentalization
For Na + compartmentalization studies, BY-2 cells were harvested 4 d after subculture and used for all microscopic analyses. After the cells had been treated with 150mM NaCl in MS medium, they were incubated with shaking at room temperature for 1h. SodiumGreen TM indicator (S6901; Invitrogen, Carlsbad, CA, USA was used to visualize the intracellular contents of Na + [17], and the confocal images were captured using Olympus FV1000 (Ex: 488nm; Em: 510-525nm). The fluorescence signals were analyzed using the ImageJ program (ver. 1.44p) [31]. The total fluorescence intensity in the pixels was divided by the total area to obtain the average pixel fluorescence intensity. Background fluorescence intensity was measured in the same field and was subtracted. DIC images of cells were obtained by excitation with a red diode. Two replicates of each experiment were performed.
For real-time image capturing, 3-day-old cells were preincubated with 10μM Sodium Green TM indicator for 30 min prior to the 200mMNaCl treatment. Images of cells were captured at 10-sec intervals after NaCl treatment for a total period of 50min, using the Bio-Rad Radiance 2100 system (Ex: 514nm, filter set HQ545/40). The first reading was taken ~20s after NaCl treatment. Cell sizes were measured by the ImageJ program (ver. 1.37) and the % change was reported. Na + content was measured as the intensity of the Sodium Green TM signal per unit area using the same program. The images were collated and converted to an MPEG video and attached as a supplementary file.

Microscopic analysis of cell viability
For cell viability assays, 4-day-old BY-2 cell suspension cultures were remained untreated or treated with one of the following: 150mM NaCl in MS medium, 150mMNaCl with 1μM IP 3 in MS medium, or 13.3% (w/v) PEG-6000 in MS medium (near-isotonic to 150mM NaCl), for 24 h with shaking in an orbital shaker. After treatment, cells were stained with 0.4μg/μl Trypan blue (Sigma Aldrich Co., St Louis, MO, USA). The images of stained cells (around 150 cells in each experiment) were captured using the CCD camera attached to the light microscope (Nikon Eclipse 80i). A total of 91-247 cells were counted from 4-12 fields. Two replicates of each experiment were performed.

Determination of total chlorophyll in A. thaliana
The determination of total chlorophyll in A. thaliana was performed as described previously [34]. Leaf tissue of less than 0.02 g was immersed in 0.8 ml N, N-dimethylformamide (DMF) followed by incubation at 4°C overnight [36]. The absorbance at 603, 647 and 664 nm was recorded. The amount of total chlorophyll was calculated using a formula published previously [37].

Statistical analysis
Statistical analysis was performed using the Statistical Package for Social Sciences (version 16.0; SPSS Inc., Chicago, IL, USA).

The expression of GmSAL1 was induced by NaCl but not near-isotonic PEG
We obtained the full-length coding region ofGmSAL1 (the soybean homologue ofAtSAL1) by PCR using degenerate primers followed by RACE.Basic Local Alignment Search Tool (BLAST) analysis showed that the overall amino acid sequence identity of GmSAL1 (GenBank accession No.: EF637045) to the closest homologues in A. thaliana, AtSAL1 (GenBank accession No.: Q42546) and AtSAL2 (GenBank accession No.:NP_201205), is 77% and 63% respectively. Multiple alignments were performed on GmSAL1, AtSAL1 and AtSAL2 (Figure 1). The consensus sequences [38] for inositol-and phosphate-binding are all conserved in GmSAL1.
Since the yeast homologue, ScHal2, is known to be a saltstress determinant [39],we tested the expression of GmSAL1 when the plants were subjected to different concentrations of NaCl ( Figure 2). It was found that the levels of GmSAL1 transcripts in both leaves and roots were induced by treating the plants with NaCl ( Figure 2). Since NaCl treatment consists of two stress components: ionic stress and osmotic stress, we therefore used near-isotonic PEG treatments to control for the osmotic stress. The difference in GmSAL1 expression levels between each isotonic pair of NaCl and PEG treatments showed that GmSAL1 could only be induced by ionic stress GmSAL1 Regulates IP₃ Signals PLOS ONE | www.plosone.org Multiple alignments were performed for GmSAL1, AtSAL1, and AtSAL2 (Genbank numbers EF637045, Q42546, and NP_201205, respectively), using the ClustalW program [19] in the BioEdit package (ver. 7.0.5.3). Identical amino acid residues were shaded black and similar amino acid residues were shaded grey. The conserved motifs involved in substrate-and metal-binding and nucleophilic water activation were marked as A, B and C [38]. (due to NaCl) but not osmotic stress (due to PEG) ( Figure 2). PEG treatment actually led to a repression of GmSAL1gene expression ( Figure 2).

GmSAL1 hydrolyzed IP 3
The amino acid sequence alignment suggested that GmSAL1 may possess inositol polyphosphatase activities( Figure 1). We expressed and purified the GmSAL1 protein from Escherichia coli in order to determine its substrate specificity in vitro. Surprisingly, we found that GmSAL1 used IP 3 readily as the substrate but had no effect on I(1,3,4)P 3 , I(1,4)P 2 , inositol 1-monophosphate (I(1)P) or IP 6 . The K m and K cat values were also determined ( Table 1). The GmSAL1 substrate specificity is therefore completely different from that of AtSAL1, an inositol 1-phosphatase and the closest homologue of GmSAL1 in A. thaliana.
To study the physiological significance of the enzymatic activities of GmSAL1 in vivo, we employed the biosensor construct Pleckstrin homology domain-green florescent protein (PH PLCd -GFP). PH PLCd is a protein domain which binds to PI(4,5)P 2 (on the plasma membrane) and IP 3 (in the cytosol) [30]. The PH PLCd -GFP construct was successfully employed to indicate the cytoplasmic IP 3 levels [30].
We first generated homozygous lines of transgenic A. thaliana expressing GmSAL1 and confirmed the expression of the transgene ( Figure S1 in File S1). Two independent GmSAL1 transgenic lines were crossed to an A. thaliana line expressing the PH PLCd -GFP construct. Homozygous transgenic lines containing both the GmSAL1 and PH PLCd -GFP constructs were obtained and the expression of GmSAL1 was validated ( Figure S1B in File S1).
Since the expression of PH PLCd -GFP was reduced in the GmSAL1/PH PLCd -GFP double transformants ( Figure S1D in File S1), instead of comparing the total GFP signals, we examined the percentage of signals localized in the cytoplasm of guard cells.
Compared to the original PH PLCd -GFP transgenic line, the proportion of signals in the cytoplasm was much lower in the GmSAL1/PH PLCd -GFP double transformants (Figure 3).   Table S1 in File S1). Two-week-old soybean seedlings grown hydroponically were placed in fresh half-strength modified Hoagland's solution without (untreated) or with various NaCl or PEG supplements. Leaf and root tissues were harvested 24 h after treatment. 20μg total RNA from each sample was used for northern blot analysis. The experiment was performed twice and similar results were obtained. Together with the in vitro enzymatic data, this in vivo evidence supports the function of GmSAL1 to down-regulate the level of cytosolic IP 3 .

Ectopic expression of GmSAL1 negated the effects of ABA on stomatal closure and seed germination
IP 3 plays a key role in mediating the ABA signaling in guard cells to control the stomatal aperture [8]. To test whether the ectopic expression of GmSAL1 will also affect IP 3 -mediated stomatal closure, a stomatal aperture assay was conducted. Detached A. thaliana leaves were treated in a buffer containing 0.1% (v/v) MeOH, with or without 100μM ABA. Under ABA treatment, the stomatal apertures in the wild type leaves (WT) and the empty vector-transformed control were much reduced compared to no ABA treatment (Figure 4). On the other hand, the stomatal apertures in the leaves of the GmSAL1 lines were significantly larger than those in the controls under the same ABA treatment (Figure 4).
The model of ABA-induced stomatal closure postulates that ABA increases the cytosolic IP 3 level in guard cells, which in turn leads to an increase in cytosolic calcium [Ca 2+ ] cyt , resulting in the differential activation and inactivation of K + channels on the plasma membrane and the tonoplast [1]. The net result is the efflux of K + (and subsequently water) out of the cytosol and the vacuole, which leads to the loss of turgidity in guard cells and, consequently, stomatal closure [1,3,40]. To investigate whether the effect of GmSAL1 on stomatal aperture is Ca 2+dependent, 5mM Ca 2+ was also included in the medium in addition to 100μM ABA. External Ca 2+ leads to the elevation in [Ca 2+ ] cyt and stomatal closure [41]. Our results indicated that the suppressing effect of GmSAL1 on stomatal closure under 100μM ABA was mitigated by the addition of external Ca 2+ (Figure 4).In the medium containing 100μM ABA and 5mM Ca 2+ , the stomatal aperture of wild type, the empty-vector line, and the GmSAL1transgenic lines show no significant differences ( Figure 4).The effect of GmSAL1 on the stomatal opening may hence be a result of its hydrolytic activities toward cellular IP 3 .
Besides controlling the stomatal aperture, another important function of ABA in plants is the inhibition of seed germination. Germination rate is a common strategy to study ABA sensitivity [8,10]. The effects of ABA on the seed germination rate of the wild type A. thaliana, empty-vector transgenic control and theGmSAL1transgenic lines were compared. Under 2.5μM ABA and 4μM ABA treatments, the germination rate of GmSAL1transgenic lines was significantly higher than the controls ( Figure 5) with the effects being more pronounced under 2.5μM ABA than 4μM ABA treatment, indicating that GmSAL1 can reduce the sensitivity of plants toward ABA.

Ectopic expression of GmSAL1 enhanced vacuolar Na + compartmentalization in BY-2 cells under salinity stress
Next, we examined the effects of GmSAL1 on plant cells in general. A previous study of AtSAL1 showed that its ectopic expression in yeast cells could increase salinity tolerance [9].   Since the expression of GmSAL1 is salt-inducible (Figure 2), the effect of GmSAL1 on salt tolerance in plant cells was investigated. The survival rates of GmSAL1 transgenic BY-2 cells under NaCl stress ( Figure 6) and PEG-induced osmotic stress ( Figure S2 in File S1) were studied. The expression of GmSAL1 in the transgenic cells was validated ( Figure S1 in File S1). Trypan blue was used to stain the dead cells. NaCl (salinity stress) significantly increased the number of dead cells. The expression of GmSAL1 could alleviate NaC1induced salinity stress ( Figure 6) but not PEG-induced osmotic stress ( Figure S2 in File S1). The percentage of survival in the GmSAL1 transgenic BY-2 cell lines under NaCl stress was significantly higher than in other lines (Figure 6), whereas the supplementation of 1μM IP 3 could negate the protective effects of GmSAL1 ( Figure 6). We also traced the cellular compartmentalization of Na + using the fluorescence dye Sodium Green TM . Our results indicated that under NaCl treatment, GmSAL1 transgenic BY-2 cells exhibited enhanced vacuolar compartmentalization of Na + , as reflected by the higher fluorescence intensity in the vacuole when compared to the wild type BY-2 cells and empty-vector transgenic control (Figure 7). Similar to the results of the cell survival test, the effects of expressing GmSAL1 was much reduced by the supplementation of 1μM IP 3 (Figure 7).
To better visualize the changes of BY-2 cells under NaCl stress, we captured time-series images of a single cell after NaCl treatment. Two major differences were observed when comparing the wild type BY-2 cell to the GmSAL1 transgenic cell. Firstly, when NaCl was added, the size of the protoplast in the wild type cell decreased (Figure 8; Video S1), probably due to the efflux of water from the cell. Under the same treatment, the GmSAL1 transgenic cell exhibited an initial shrinkage in protoplast size followed by a recovery after about 15 min ( Figure 8; Video S1). Secondly, while there was no significant elevation of Na + compartmentalization into vacuoles over time in the wild type BY-2 cell under NaCl treatment, an obvious increase in vacuolar Na + was observed in the GmSAL1 transgenic cell under the same conditions. The vacuolar Na + was maintained at a higher level than before the NaCl treatment when the transgenic cell gradually recovered from shrinkage ( Figure 8; Video S1).

Ectopic expression of GmSAL1did not enhance the tolerance of A. thaliana to NaCl and PEG stress
We also investigated the effect of ectopic expression of GmSAL1 at whole plant level. Wild type (WT), transgenic empty vector (Empty vector), GmSAL1 transgenic (GmSAL1-1 and GmSAL1-2) A. thaliana were treated with NaCl and nearisotonic PEG-6000. In contrast to the protective effects of GmSAL1 on BY-2 cells, ectopic expression of GmSAL1 in A. thaliana did not confer obvious protection under salt stress and osmotic stress ( Figure S3 in File S1).

Discussion
It is common to classify enzymes that can act on I(1,4)P 2 and I(1,3,4)P 3 as inositol 1-phosphatase and those that can act on IP 3 as inositol 5-phosphatase [5]. While GmSAL1 showed strong sequence homology to AtSAL1 and AtSAL2 which were reported to be inositol 1-phosphatases that have no activities toward IP 3 , GmSAL1 employed IP 3 as the preferred substrate and is inactive toward I(1,4)P 2 and I(1,3,4)P 3 (Table 1).
There are two possible explanations for this observation. GmSAL1 may be an inositol 1-phosphatase like AtSAL1, but differs from AtSAL1 in substrate specificity. Another possibility is that GmSAL1 possesses inositol 5-phosphatase activities that act on the 5'-phosphate of IP 3 . Inositol polyphosphates without a 5'-phosphate such as I(1,4)P 2 and I(1,3,4)P 3 are therefore non-substrates. Detailed sequence analysis revealed that GmSAL1 exhibits a low degree of homology to the two consensus domains found in inositol 5-phosphatases ( Figure  S4 in File S1). In this research, we focus on the consequence of the IP 3 hydrolytic activities exhibited by GmSAL1.
The K m value of GmSAL1 toward IP 3 was found to be about 10μM, which is similar to the K m value of human inositol 5phosphatase that also acts on IP 3 [5]. This value is at least two folds higher than the cellular IP 3 level that is needed to affect K + transport [42].Therefore, GmSAL1 activity inside the cell may not be at maximum velocity under normal conditions. The IP 3 level required to induce Ca 2+ is at the μM level [42]. While mainly located in the cytosol, IP 3 may be able to bind to receptors such as Ca 2+ channels on the plasma membrane [30,43]. On the other hand, stress will increase the level of cytosolic IP 3 [44][45][46]. For instance, NaCl treatment could increase the IP 3 level up to 15 folds in A. thaliana [2]. GmSAL1 may therefore play a role in the fine adjustment of the cytosolic IP 3 concentration under stress.
The expression of GmSAL1 in its native host was responsive to NaCl (salinity stress) but not near-isotonic PEG treatment (osmotic stress) (Figure 2). This may be tied to its physiological roles. The ability of GmSAL1 to reduce the ABA-induced stomatal response (Figure 4), by lowering the IP 3 level, is apparently not a protective mechanism against long-term osmotic stress. The expression of GmSAL1 in transgenic BY-2 cells also showed no improvement in the tolerance toward PEG treatment ( Figure S2 in File S1). On the other hand, GmSAL1 can help to combat salinity stress at the cellular level by enhancing the vacuolar compartmentalization of Na + ( Figure 6) and such an effect was not observed when IP 3 was added. It is possible that under such experimental conditions, the protective function of GmSAL1 is brought forth by reducing the IP 3 below a threshold level.
Using stomatal closure ( Figure 4) and seed germination rate ( Figure 5) as parameters, we showed that GmSAL1 can lower the plant's sensitivity toward ABA treatments, presumably due to the reduction of IP 3 signals. Such effects were also observed in inositol 5-phosphatases which use IP 3 as their substrate [8,10].
The cytosolic IP 3 offers protection against water loss in planta via inducing the closure of stomata by activating the tonoplast and cell membrane K + channels that remove K + from the vacuole and the cytosol and inactivating K + channels that increase uptake [1]. These K + channels are reported to be nonspecific and can also transport Na + [47,48]. Consistent with this observation, it was previously reported that the addition of NaCl could lead to stomatal opening, a phenomenon that could be reversed by ABA [49].
The classical model that IP 3 is the direct signaling molecule inducing cytosolic Ca 2+ influx in the guard cells [50] has been challenged by some recent findings. IP 6 was found to be a much more potent signalling molecule controlling Ca 2+ influx and the effect of IP 3 on Ca 2+ influx might be due to its conversion to IP 6 [51]. However, GmSAL1 does not use IP 6 as the substrate (Table 1) and hence the GmSAL1 effect on stomatal opening is via regulation of the cytosolic IP 3 levels.
The level of cellular IP 3 increases under stress [44][45][46]. If the effect of IP 3 on vacuolar cation channels also occurs in cells other than the guard cells, a higher level of IP 3 will decrease vacuolar Na + compartmentalization. In contrast, the hydrolysis of IP 3 will enhance the accumulation of vacuolar Na + under NaCl treatments. It was indeed what we observed using the Four-day-old BY-2 cells grown in MS medium were used, including cells of the untransformed wild type (WT), empty-vector transgenic control (Empty vector), and three independent GmSAL1 transgenic lines (A3-1, B3-1, D2-3). Vacuolar Na + compartmentalization was visualized with the use of Sodium Green TM . Cells pre-washed with MS medium were transferred to fresh MS medium containing no supplements (untreated), MS medium supplemented with 150mM NaCl or 150mM NaCl+ 1μM IP 3 for 1h with shaking. They were then washed with fresh MS medium and stained with 5μM Sodium Green TM , followed by confocal microscopic analysis. A: Typical photos showing the relative levels of vacuolar Na + using the fluorescent signal of Sodium Green TM (represented by the pseudo-green color). Scale bar= 50μm.B: Statistical analysis. The relative fluorescence intensity of 17-34 cells (from 4 fields) was determined for each data point. Error bar: standard error. ** indicates a significant difference (p<0.01) between the transgenic lines and WT, based on ANOVA followed by the posthoc Tukey's test.
BY-2 cell model. GmSAL1 produced in the transgenic BY-2 cells hydrolyzed IP 3 and hence increased Na + compartmentalization in the vacuole (Figure 7), resulting in a higher survival rate for the transgenic cells under NaCl stress than for the wild type ( Figure 6).
Compartmentalization of Na + in the vacuole is an effective way to protect the plant cell against salinity stress [52][53][54]. On one hand, the toxic Na + is removed from the cytosol. At the same time, the accumulation of Na + in the vacuole sets up an osmotic gradient to enable the plant cell to uptake water from an environment with low osmotic potential [55,56]. This is supported by our observations that severe plasmolysis occurred in the untransformed wild type BY-2 cells upon NaCl treatment and the GmSAL1 transgenic cells could accumulate Na + in the vacuole more effectively and could therefore partially restore the protoplast size, presumably through increased water intake following ion compartmentalization in the vacuole (Figure 8; Video S1). The detailed mechanism of how GmSAL1 and cytosolic IP 3 level regulate Na + compartmentalization into vacuole is still unclear at this point.
Since ABA is the hormone which induces stomatal closure to protect the stressed plant from water loss [57] and GmSAL1 GmSAL1 Regulates IP₃ Signals reduces ABA-induced stomatal response, it is not surprising that the protective effect of GmSAL1 on NaCl or PEG stress was not obvious at the whole plant level ( Figure S4 in File S1).
In summary, we conclude that GmSAL1 is a novel soybean SAL1 homologue that hydrolyzes IP 3 and plays differential roles at the whole plant level versus at the cellular level in response to salinity stress.

Supporting Information
File S1. A combined file containing one supplemental table and four supplemental figures as follows: Table S1, Osmolarity of near-isotonic solutions; Figure S1, Validation of transgene expression; Figure S2, Ectopic expression of GmSAL1 in BY-2 cells did not enhance their survival rates under PEG stress; Figure S3, Ectopic expression of GmSAL1 in A. thaliana did not enhance their tolerance toward NaCl or PEG stress; Figure S4, Multiple alignments of GmSAL1 with inositol 5-phosphatases.

(DOCX)
Video S1. An MPEG video is provided to show the change in cell size and vacuolar Na + in the GmSAL1 transgenic and wild type BY-2 cell lines under 200 mM NaCl treatment. The images were taken over a 50-min period. (MPG)