SbSI-1 protein localization was examined in vivo using Gateway technology (Walhout et al., 2000). RNA was isolated from S. brachiata and cDNA was synthesized using a SuperScript RT III first-strand cDNA synthesis kit. The SbSI-1 CDS was amplified with AccuPrime™ Pfx DNA polymerase using SbSI-1LF and SbSI-1LR primers. Subsequently, the amplified CDS was cloned into a pENTER/D-TOPO entry vector (Invitrogen, USA). LR recombination reaction was performed between an attL-containing Entry clone pENTER/D-TOPO-SbSI-1 vector and an attR-containing destination vector pSITE-4CA using Gateway LR Clonase II enzyme mix (Invitrogen, USA). Integration of SbSI-1 gene in destination vector was confirmed by sequencing. The control vector and expression cassette (RFP:SbSI-1) (Supplementary Figures 1A,B) were transformed into onion epidermal cells using gene gun (PDS-1000/He Biolistic, Biorad, USA). The transformed epidermal cells were observed for transient expression of RFP with an epifluorescence microscope (Axio Imager, Carl Zeiss AG, Germany) after an incubation of 24 h on Murashige and Skoog's (MS) basal medium (Murashige and Skoog, 1962). Further the electrophoretic mobility shift assay (EMSA) of SbSI-1 protein was performed following Uguru et al. (2005) to check DNA binding property. The DNA samples was incubated with recombinant protein. After incubation the sample were electrophoresed in 6% native PAGE. The gel was silver stained and documented to record electrophoretic mobility shift.

The SbSI-1 gene was amplified using S. brachiata cDNA with AccuPrime™ Pfx DNA polymerase using forward (5′-TCCGAGCTCATGCCTAATAAACATATCATGG-3′) and reverse (5′-CGCGGATCCTTAACGGTTCCCTTGTTTC-3′) primers containing Sac I and BamH 1 sites, respectively. The SbSI-1 amplicon with BamH 1/Sac I sites was cloned into pRT101 vector (Töpfer et al., 1987). Further, the cassette containing the CaMV35S constitutive promoter, SbSI-1 gene and terminator was cloned into the pCAMBIA2301 vector at Pst I site (Supplementary Figure 1C) and subsequently mobilized into Agrobacterium tumefaciens (LBA 4404). Tobacco (Nicotiana tabacum cv. petit havana) was transformed following the standard protocol (Horsch et al., 1985). Putative transgenic shoots regenerated in the presence of 50 mgl⁻¹ kanamycin were elongated, rooted, acclimatized and grown under containment facility to harvest the T₀ seeds.

The putative transgenic (T₀ and T₁) lines were checked for transformation through histochemical GUS assay (Jefferson, 1987) using β-glucuronidase reporter gene staining kit (Sigma, USA) and PCR amplification of uidA and SbSI-1 gene (Supplementary Table 1). The number of homologs of SbSI-1 gene in S. brachiata and the transgene integration in transgenic tobacco was confirmed through southern blot analysis. To determine the homologs of SbSI-1, 30 μg genomic DNA from S. brachiata was digested with EcoR 1 and Xho 1 restriction enzymes. To determine the transgene integration 20 μg genomic DNA from WT and transgenic tobacco were digested with Hind III. The digests were separated on 0.8% agarose gel by electrophoresis. Subsequently these were transferred onto the Hybond N⁺ membrane (Amersham Pharmacia, UK) using buffer (0.4 N NaOH with 1 M NaCl) to generate DNA blot. The blot was hybridized with PCR-generated probe for the SbSI-1 gene labeled with DIG-11-dUTP. Pre-hybridization and hybridization were carried out at 42°C overnight in DIG EasyHyb buffer solution (Roche, Germany). The hybridized products were detected using CDP-Star chemi-luminescent as substrate, following the manufacturer user guidelines (Roche, Germany) and signals were visualized and documented on X-ray film after 30 min. Transgene expression in transgenic tobacco was studied through semi-quantitative reverse transcriptase (RT) PCR analysis using actin as an internal control.

To study the involvement of SbSI-1 gene in abiotic stress tolerance, qRT PCR (Bio-Rad IQ5, Bio-Rad, USA) was performed in WT and transgenic tobacco under different (10% PEG, 200 mM NaCl, 20 μM ABA and 10 μM SA) stresses using QuantiFast Kit (Qiagen, USA). The expression pattern of NtSOD, NtCAT, and NtAPX genes encoding different antioxidant enzymes (Huang et al., 2011) and NtDREB2 and NtAP2 encoding transcription factors were analyzed (Supplementary Table 1). The qRT-PCR specificity was monitored by melt curve analysis and fold expression rate was calculated by CT method (Livak and Schmittgen, 2001).

Seeds of wild-type (WT) plant and transgenic (L₈, L₂₂, and L₃₃) lines overexpressing SbSI-1 were germinated on MS medium supplemented with 200 mM mannitol or 200 mM NaCl (20 seeds per replicate). The germination was carried out under a culture room conditions [26 ± 2°C, 55–60% relative humidity (RH), 12 h photoperiod with a light intensity of 35 ± 5 μmol m⁻² s⁻¹ SFP]. Seed germination was recorded after 21 days (d) and germination percent was calculated. Uniform WT and kanamycin-positive T₁ transgenic seedlings were cultured vertically onto MS medium containing 200 mM mannitol or 200 mM NaCl for 21 d. The different growth parameters (plant height, shoot/root length, fresh and dry weight) and ion contents (Na⁺ and K⁺) were measured in these plants and compared with WT.

The seeds were germinated on kanamycin-supplemented medium for functional characterization. Uniform WT and kanamycin-positive T₁ transgenic seedlings were cultured hydroponically in the ½ strength of Hogland salt for 45 d under culture room conditions. These were subjected to 10% polyethylene glycol (PEG) and 200 mM NaCl for 24 h; thereafter samples were harvested for morphological (growth parameters) and physio-biochemical (relative water contents, electrolyte leakage, membrane stability index, tolerance index, H₂O₂ estimation, lipid peroxidation, ROS, and starch accumulation, osmotic adjustment, estimation of antioxidant enzyme activity and determination of sugar, polyphenols, and free amino acid contents) analysis in T₁ transgenic tobacco along with WT.

Uniform leaf disks (5–6 mm diameter) punched from leaf lamina of 45 d old WT and transgenic lines were used for the leaf senescence assay, chlorophyll estimation, and cell viability assay. Leaf disks (n = 11) of WT and transgenic lines were floated in 5 ml ½ strength of Hoagland salt (control) supplemented with 100, 150, 200 mM NaCl, and 10% PEG for 7 d. The retention of greenish color by leaf disks was recorded as phenotypic effects of stress treatments. Leaf disks at the end of stress period were used for photosynthetic pigment estimation (Inskeep and Bloom, 1985; Chamovitz et al., 1993) and 2,3,5-triphenyltetrazolium chloride (TTC) assay for cell viability estimation (Towill and Mazur, 1975; Hema et al., 2014).

Simultaneous gas exchange and chlorophyll fluorescence parameters were measured in leaves of control and 24 h stress treated WT and transgenic tobacco using LI-6400XT portable photosynthesis system (LICOR, Lincoln, NE, USA). Plants were dark-adapted for 45 min and fluorescence parameters were recorded. Steady state fluorescence yield was achieved by exposing plants to ambient light (1,200 μmol m⁻²s⁻¹) for 60 min in a plant growth chamber (PGC-105, Percival Scientific, US). The gas exchange measurement conditions were 1,200 μmol m⁻² s⁻¹ PPFD, ambient atmospheric CO₂ (380 μmol⁻¹ mol⁻¹) and RH (60–65%) and 26°C block temperature. The net photosynthetic rate (P_N; μmol m⁻²s⁻¹), stomatal conductance (g_s; mmol H₂O m⁻²s⁻¹), intercellular CO₂ (C_i) concentration (μmol m⁻²s⁻¹), transpiration rate (E; mmol m⁻²s⁻¹), PSII operating efficiency (ΦPSII), electron transport rate (ETR), photochemical quenching (qP), non-photochemical quenching (NPQ) were determined.

Chlorophyll a fluorescence transient (Strasser and Strasser, 1995) in leaves was measured on dark adapted stressed and unstressed plant using plant efficiency analyser (Hansatech Instruments Ltd., England). Chlorophyll fluorescence was recorded thrice (sub-replicate) at different regions of 03rd leaf of 03 plants (biological replicate) from each treatment (n = 9).

Leaf disks were punched from leaf lamina of stressed and unstressed plants and fresh weight (FW) was recorded. Subsequently turgid weight (TW) was recorded after submerging these disks in deionized water for 12 h. Dry weight (DW) was recorded after drying the disks at 80°C for 48 h in hot air oven. The relative water content (RWC) was calculated as: RWC (%) = (FW–DW/TW–DW) × 100.

The leaf samples from stressed and unstressed tobacco plants were washed with deionized water. These were immersed in 10 ml deionized water in a closed vial and incubated at 26°C on a gyratory shaker for 24 h. The electrical conductivity (EC) of the solution (EC₁) was measured using conductivity meter (SevenEasy, Mettler Toledo AG 8603, Switzerland). These samples were autoclaved at 121°C for 15 min, cooled up to 26°C and EC (EC₂) was measured. The electrolyte leakage (EL) was calculated as (EC₁/EC₂) × 100.

The uniform leaf disks from stressed and unstressed tobacco were immersed in 10 ml deionized water in close vials in two separate sets. A set of vials was incubated at 40°C for 30 min while the second set of vials at 100°C for 10 min. EC of both sets (EC₄₀ for 40°C and EC₁₀₀ for 100°C) was recorded and membrane stability index (MSI) was calculated (Sairam, 1994) as [1-(EC₄₀/EC₁₀₀)] × 100.

Uniform leaf samples from control and treated plants were dried and DW was recorded. The tolerance index (TI) was determined as DW of the plant under stress condition × 100/DW of the plant under control condition (Tuteja et al., 2013) and compared with that of WT plants.

Hydrogen peroxide and O2- radical accumulation were detected in vivo in leaves of stressed and unstressed tobacco. The leaves were immersed in nitro-blue tetrazolium (NBT) solution (1 mg mL⁻¹ in 10 mM phosphate buffer; pH 7.8) at RT for 2 h, thereafter exposed to high irradiance for 12 h until blue spots showing accumulation of O2- appeared. The presence of H₂O₂ was detected by immersing the leaf samples in 3,3-diaminobenzidine (DAB) solution (1 mg mL⁻¹ in 10 mM phosphate buffer; pH 3.8) at RT for 6 h in the dark, thereafter exposing to the light until brown spots showing accumulation of H₂O₂ appeared. Before documentation, the samples were washed with ethanol to bleach out the chlorophyll. Uniform leaves from stressed and unstressed tobacco were processed for in vivo detection of starch following Cheng et al. (2014).

Lipid peroxidation was determined by estimating malondialdehyde (MDA) concentration following Hodges et al. (1999). Leaf samples were extracted with 0.1% trichloroacetic acid (TCA), and 0.2 ml extract was reacted with 0.8 ml of TBA reagent (0.5% TBA in 20% TCA) and subsequently boiled at 95°C for 30 min. Samples were ice cooled, centrifuged at 10,000 g for 5 min and absorbance was read at 440, 532, and 600 nm.

The leaf tissue from treated and untreated plants were ground in liquid nitrogen and extracted in protein extraction buffer (50 mM potassium buffer (pH 7.0), 1 mM EDTA, 0.05% (w/v) triton × −100, 5% (w/v) polyvinylpolypyrrolidone). The protein concentration in the extract was determined following Bradford method (Bradford, 1976) and it was used for determination of activity of superoxide dismutase (SOD), ascorbate peroxidase (APX), and catalase (CAT). The SOD activity was assayed by monitoring the inhibition of photochemical reduction of nitro blue tetrazolium (NBT) following Beyer and Fridovich (1987). The amount of enzyme required for 50% inhibition of NBT reduction as monitored at 560 nm was considered as one unit of SOD activity. The APX activity was assayed by monitoring oxidation of ascorbate (Nakano and Asada, 1981). The decline in absorbance at 290 nm was recorded and 2.8 mM⁻¹ cm-1 was taken as extinction coefficient (Δe). The CAT activity was assayed by monitoring the disappearance of H₂O₂ (Miyagawa et al., 2000) and taking 43.6 M⁻¹ cm⁻¹ as Δe at 240 nm (Patterson et al., 1984).

To determine the H₂O₂ contents the leaf samples were extracted with 0.1% TCA. A 100 μl extract was reacted with 1 ml of reaction buffer containing 0.25 mM FeSO₄, 0.25 mM (NH₄)₂SO₄, 25 mM H₂SO₄, 1.25 mM xylenol orange and 1 mM sorbitol (He et al., 2000) at RT for 1 h and absorbance was read at 560 nm.

The accumulation of O2- radicals were studied through XTT assay (Schopfer et al., 2001) by demonstrating the reduction of XXT into formazans (Able et al., 1998). Uniform leaf disks were incubated with 1 ml of XTT solution (500 mM XTT in 20 mM K-phosphate buffer, pH 6.0) in darkness at 25°C on a shaker. The increase in absorbance (A₄₇₀) of the incubation medium was recorded (Sutherland and Learmonth, 1997).

The fresh leaf tissue from control and treated plants were frozen, thawed and centrifuged at 10,000 g to extract the sap. The solute concentration was determined using Vapro Pressure Osmometer (model-5600; Wescor, Logan UT, USA). The solute potential (Ψ_s) of the sap was calculated as —nRT/V; where n represent a number of solute molecules; R is universal gas constant; T is the temperature in ° K, and V is volume in liter.

The treated and untreated WT and transgenic seedlings were oven dried, acid [perchloric acid and nitric acid solution (3:1 v/v)] digested and heated to dryness. The digest was dissolved in deionised water and ion contents (Na⁺ and K⁺) were estimated by inductively coupled plasma optical emission spectrometer (Optima2000DV, PerkinElmer, Germany).

The proline content in leaf tissues (200 mg) of stressed and unstressed tobacco were extracted with 3% sulphosalicylic acid and estimated following Ringel et al. (2003). The leaf tissues (1,000 mg) were extracted in 70% ethanol, and the residue left after evaporation was dissolved in milliQ water for estimation total soluble sugar, reducing sugar, polyphenol and free amino acid contents. The total soluble sugar and reducing sugar contents in leaves were measured following anthrone-sulphuric acid (Dubois et al., 1956) and DNS method (Miller, 1959), respectively. The polyphenol contents in leaves were determined following Chandler and Dodds (1983) using Folin-Ciocalteau's reagent. Total free amino acid contents in leaf tissues were determined following Sugano et al. (1975).

Each experiment was performed three times with three biological replicates for physio-chemical estimations and with 10 replicates for growht measurments. The data recorded were subjected to one-way ANOVA for analysis of variance to determine the significance among mean values of WT and transgenic plants among treatments. LSD was used for comparisons of means. The data was presented as mean ± SD and significance in responses of transgenic lines against control was indicated by classical asterisks (^*<0.05; ^**<0.01; ^***<0.001). The multivariate analysis (Principal Component Analysis; PCA) was carried out to study the correlation between different variables among multidimensional datasets. Plants grown under stresses and different parameters measured were considered as observations and variables, respectively. PCA was performed and analyzed using SigmaPlot (version 13) and Origin (version 15) for individual (for growth, physiological health, osmotic adjustment and photosynthetic parameters), and integrated (all parameters together) responses.

The transient expression assays of RFP and RFP:SbSI-1 fusion construct showed SbSI-1 protein as a nuclear-localized protein. Onion cells transformed with RFP alone exhibited an even distribution of red fluorescence signals in the entire cell region, whereas transformed cells with RFP:SbSI-1 fusion construct showed fluorescence signals in the nucleus only. DAPI stained (blue spot) nucleus showed overlapping with an RFP stained (red spot) nucleus (Figure 1A). In merged image fluorescence was observed only in the nucleus and not from any region in the cell confirming the nuclear localization of the SbSI-1 protein. The southern blot analysis confirmed single copy of SbSI-1 gene (Figure 1B) in S. brachiata. The EMSA further showed DNA binding property of SbSI-1 protein (Supplementary Figure 1D).

At the end of transformation experiment 42 putative transgenic lines were obtained showing positive GUS assay (Supplementary Figures 1E, 2A). The transgenic tobacco plants (putative transgenic plants and T₁ transgenic tobacco lines) were confirmed by GUS assay and PCR amplification of both SbSI-1 and uidA gene (Supplementary Figures 2A,B). Based on seed germination (data not shown), GUS expression assay and PCR analysis (Figure 1C) line L₈, L₂₂, and L₃₃ were selected for functional validation. Semi-quantitative RT-PCR showed expression of SbSI-1 gene in transgenic tobacco whereas no expression was detected in WT plants (Supplementary Figure 2C; Figure 1D). The southern hybridization confirmed single copy gene integration to the selected transgenic lines (Figure 1E).

The seeds of WT and transgenic lines germinated equally under unstressed conditions (Supplementary Figure 3A) while transgenic lines overexpressing SbSI-1 exhibited significantly higher seed germination under stressed conditions (Figure 2A). The transgenic lines exhibited significantly improved growth attributes (Supplementary Figures 3B–F) and grew healthier (Figure 3) than WT under drought and salt stress. In senescence assay, the degree of stress-induced bleaching in leaf disks (Figure 4) and retention of photosynthetic pigments (Figure 2B; Supplementary Figures 4A–C) at the end of stress period showed better endurance in transgenic lines against salt and drought stress. In agreement with these results in TTC assay the transgenic lines exhibited higher cell viability as compared to WT under stress conditions (Figure 2C). The in vivo staining of starch showed more accumulation of starch in transgenic lines than WT under stressed conditions indicating better photosynthetic performance by transgenic lines (Figure 5A). Transgenic lines exhibited significantly higher tolerance index (TI) than WT under stressed conditions (Figure 5B). These results clearly evidenced that overexpression of SbSI-1 enable transgenic tobacco to perform better and maintain the higher cell viability during the stress period.

The abiotic stress leads to the generation of ROS and the plant cope up with excess ROS through its enhanced antioxidant defense system. The activity of SOD, APX and CAT was found higher in both WT and transgenic lines under stress conditions (Figures 6A–C). However, the transgenic lines exhibited significantly higher SOD and APX activities under stress conditions as compared to WT. The transgenic lines exhibited lower accumulation of O2- and H₂O₂ than WT tobacco under stressed conditions (Figures 5C,D). In XTT assay the decreased absorbance in transgenic lines as compared to WT under stress conditions confirmed lower generation O2- in transgenic lines overexpressing SbSI-1 gene (Supplementary Figures 4D). Further the H₂O₂ was quantified lower in transgenic lines than WT under stress conditions (Supplementary Figure 4E). These results confirmed that overexpression of SbSI-1 alleviates the build-up of ROS (Supplementary Figures 4D,E). In agreement with these results the accumulation of MDA was significantly lower in transgenic lines than WT under stressed conditions (Figure 7A).

The EL an indicator of membrane permeability increased significantly under stressed conditions however it was significantly lower in transgenic lines than WT (Figure 7B). Similarly, the MSI was recorded significantly higher in transgenic lines than WT under stressed conditions (Figure 7C). The polyphenols accumulated significantly higher in leaves of transgenic lines than WT under stressed conditions (Figure 7D). The higher MSI, lower EL, and lower accumulation of MDA in transgenic lines could be considered as proof for curtailment of oxidative damage. The transgenic lines and WT had similar RWC under control conditions while under stressed conditions transgenic lines had significantly higher RWC than WT (Supplementary Figure 4F). These results clearly evident better physiological health of the transgenic tobacco overexpressing SbSI-1 than WT under stressed conditions.

Further to study the involvement of SbSI-1 in curtailment of ROS induced damages expression pattern of antioxidant genes and transcription factors was examined. The transcript analysis of WT and transgenic lines revealed that expression of the NtSOD, NtAPX, and NtCAT genes was relatively higher in transgenic lines under different stresses (Supplementary Figures 5A–C). The results indicated higher activity of SOD, APX and CAT in transgenic lines under stressed conditions (Figures 6A–C). Further up-regulation of NtDREB2 and NtAP2 in transgenic lines under stress conditions indicated the role of SbSI-1 in stress tolerance (Supplementary Figures 5D,E).

The P_N, g_s, and C_i decreased in WT and transgenic lines under stressed conditions however the transgenic lines exhibited significantly higher P_N, g_s, and C_i than WT under drought and salt stress (Figures 8A–C). Similarly the transgenic lines maintained quite higher E than WT (Figure 8D). Transgenic lines exhibited better ΦPSII and higher ETR than WT both under controlled and stressed conditions (Figures 8E,F). In agreement with ΦPSII and ETR, qP was found quite higher in transgenic lines (Figure 8G). The NPQ was comparatively lower in transgenic lines than WT under stressed conditions (Figure 8H) however reduction was non-significant under salt stress. The better P_N, g_s, C_i, E, ΦPSII, ETR, and qP along with the higher accumulation of starch in transgenic lines evidenced better endurance to the drought and salt stress in transgenic tobacco overexpressing SbSI-1 gene.

Chlorophyll a fluorescence transient analysis was performed to corroborate the photosynthetic performance of transgenic lines under stressed conditions (Figure 9). The area over the transient curves (an indicator of plastoquinone pool) increased under drought stress. Under salt stress, it reduced significantly in WT (20.68%) while the reduction was non-significant in transgenic lines. The T_FM (time taken to achieve maximum fluorescence) improved comparatively in transgenic lines under drought stress. Under salt stress, T_FM decreased, however, lines L₈ and L₂₂ showed improvement. The chlorophyll fluorescence intensity in transgenic lines when all PSII reaction centers are open (F_o) and closed (F_m) was comparatively higher than WT under stressed conditions. F_o/F_m (non-photochemical de-excitation) increased under drought and salt stress, however, transgenic lines had relatively lower F_o/F_m than WT. The F_v/F_m (maximum quantum yield of PSII) reduced under stressed conditions and the transgenic lines had relatively higher F_v/F_m than WT. The F_v/F_o (the activity of the water-splitting complex on PSII donor side) reduced under stress conditions while transgenic lines showed relatively better F_v/F_o. The TR_o/ABS (maximum yield of primary photochemistry) and ET_o/ABS (electron transport flux) reduced in WT under stress conditions while transgenic lines maintained relatively higher TR_o/ABS and ET_o/ABS. The PI_Total (performance index) decreased significantly in WT under drought (18.22%) and salt (61.02%) stress while it was significantly higher in transgenic lines than WT under stressed conditions.

Transgenic and WT plants accumulated proline under stressed conditions however the accumulation was significantly higher in transgenic tobacco than WT plants (Figure 10A). Similarly the soluble sugar and reducing sugar accumulated under stressed conditions and the accumulation was significantly higher in transgenic tobacco than WT (Figures 10B,C). The free amino acids were similar in WT and transgenic tobacco under control conditions. These reduced significantly in WT plants under stress conditions while transgenic tobacco maintained significantly higher accumulation of free amino acids under stress conditions (Figure 10D).

The accumulation of Na⁺ and K⁺ varied in both WT and transgenic tobacco under stress conditions. The transgenic tobacco showed significantly lower accumulation of Na⁺ and higher accumulation of K⁺ compared with that of WT under salt stress (Supplementary Figures 3G,H). Transgenic tobacco significantly accumulated higher K⁺ than WT under drought stress. Similarly, transgenic tobacco had significantly higher K⁺/Na⁺ ratio under stressed conditions (Figure 11). The osmotic potential in WT and transgenic tobacco was recorded similar under control conditions. In agreement with accumulations of organic osmolites and inorganic ions, it increased negatively and the transgenic tobacco showed better osmotic potential than WT (Figure 10E) under stressed conditions.

PCA was carried out both individually for the comparative study of the growth responses, physiological health, osmotic adjustment, and photosynthetic responses; and integrated responses of WT and transgenic tobacco under unstressed and stressed (drought and salt) conditions. Individually growth responses, physiological health, osmotic adjustment and photosynthetic responses PCA revealed 79.47, 83.76, 82.87, and 78.34% variations, respectively (Supplementary Figures 6A–D). The WT and transgenic tobacco showed comparable response under control and stressed conditions as revealed by clustering of transgenic tobacco plants under particular stress in the individual bi-plot analysis. Clustering of transgenic tobacco plants under particular stress showed comparable responses to avoid the harmful effect of stresses. The integrated PCA revealed 68.36% variations (Supplementary Figures 6A–D). Both the individual and integrated PCA showed the possible correlations of plant response to different variables and stress condition (Supplementary Figures 6E). Overall, the PCA showed a statistical difference among growth responses, physiological health, osmotic adjustment and photosynthetic responses of WT and transgenic tobacco under control and stress conditions.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.