Thiourea and hydrogen peroxide priming improved K+ retention and source-sink relationship for mitigating salt stress in rice

Plant bioregulators (PBRs) represent low-cost chemicals for boosting plant defense, especially under stress conditions. In the present study, redox based PBRs such as thiourea (TU; a non-physiological thiol-based ROS scavenger) and hydrogen peroxide (H2O2; a prevalent biological ROS) were assessed for their ability to mitigate NaCl stress in rice variety IR 64. Despite their contrasting redox chemistry, TU or H2O2 supplementation under NaCl [NaCl + TU (NT) or NaCl + H2O2 (NH)] generated a reducing redox environment in planta, which improved the plant growth compared with those of NaCl alone treatment. This was concomitant with better K+ retention and upregulated expression of NaCl defense related genes including HAK21, LEA1, TSPO and EN20 in both NT and NH treated seedlings. Under field conditions, foliar applications of TU and H2O2, at vegetative growth, pre-flowering and grain filling stages, increased growth and yield attributes under both control and NaCl stress conditions. Principal component analysis revealed glutathione reductase dependent reduced ROS accumulation in source (flag leaves) and sucrose synthase mediated sucrose catabolism in sink (developing inflorescence), as the key variables associated with NT and NH mediated effects, respectively. In addition, photosystem-II efficiency, K+ retention and source-sink relationship were also improved in TU and H2O2 treated plants. Taken together, our study highlights that reducing redox environment acts as a central regulator of plant’s tolerance responses to salt stress. In addition, TU and H2O2 are proposed as potential redox-based PBRs for boosting rice productivity under the realistic field conditions.

www.nature.com/scientificreports/ (CO 2 ) is fixed in chloroplasts via the Calvin cycle to yield triose phosphates (triose-P). Triose-P is transported to cytosol by triose-P/phosphate transporter for the synthesis of sucrose. In cytosol, aldolases catalyse the formation of fructose 1,6-bisphosphate (F1,6BP), which is further metabolized to yield sucrose by the combined action of fructose 1,6-bisphosphatase and sucrose phosphate synthase (SPS). Sucrose as a final product of photo assimilation is translocated from its synthesis site (leaf) to various non-photosynthetic tissues 15 (sink tissues) via phloem tissue. Sucrose translocated to sink is metabolized by cytosolic neutral invertase (NI), sucrose synthase (SuSy), and vacuolar acid invertase (AI), providing the hexose pool for the synthesis of structural and non-structural carbohydrates. The invertases (INV) catalyse irreversible hydrolyzation of sucrose into its hexose monomers (glucose and fructose), whereas SuSy catalyses reversible cleavage of sucrose using UDP to yield fructose and UDP-G. Salt stress negatively affects the sugar dynamics of the source leaf and developing sink tissue by inhibiting the synthesis, redistribution and utilisation of sucrose 16 . Rice is the staple food crop which feeds more than half of the world population. Among the different abiotic stresses, soil salinity poses a major constraint for rice productivity. For rice, critical salinity level is estimated to be 6.9 dS m −1 which leads to 50% yield loss 17 . Salt sensitivity also varies with age, from being moderately tolerant at seedling stage to highly susceptible at reproductive phase 18 . External application of chemicals and biomolecules, referred to as plant bioregulators (PBRs), has been shown to minimize the salt-stress induced yield losses in multiple crops. Though PBRs may act differently, a unified redox/ROS dependent action has recently been proposed 19 . In the present study, two redox based PBRs such as thiourea 20 (TU, non-physiological thiolbased ROS scavenger) and H 2 O 2 21,22 (ROS signalling molecule), which chemically drive ROS levels in opposite direction, were assessed whether they will have an overlapping or independent effect(s). To this end, a short-term study was conducted at seedling stage to understand the impact of nutrient medium-supplemented TU and H 2 O 2 on redox and ionic equilibrium. Further, the effects of foliar supplemented TU and H 2 O 2 on field-grown plants were also studied in terms of source-sink relationship and yield attributes. The findings revealed that both TU and H 2 O 2 maintain reduced redox status that could act as "core" regulator for improving K + retention ability, photosynthetic efficiency and source-sink strength of the plants. These changes were ultimately reflected in the form of improved growth and yield under both control and NaCl stress conditions.

Materials and methods
Plant material, growth conditions and stress treatment. The study was performed on Indian rice (O. sativa) var. IR-64. The seeds were surface sterilized using 30% ethanol for 3 min followed by repeated washing with distilled water to remove traces of ethanol. The surface-sterilized seeds were germinated for 48 h and hydroponic cultures were established, as per the method described previously 23 . The ameliorative potential of TU and H 2 O 2 towards NaCl stress was evaluated using two independent approaches. In the first approach, 14 days old hydroponically grown seedlings were subjected to different treatments including control (Yoshida medium), NaCl (50 mM), TU (7.5 µM), NaCl (50 mM) + TU (7.5 µM), H 2 O 2 (1 µM) and NaCl (50 mM) + H 2 O 2 (1 µM). Hereafter, NaCl + TU and NaCl + H 2 O 2 treatments were denoted as NT and NH, respectively. A pre-treatment of 7.5 µM TU (TU and NT) and 1 µM H 2 O 2 (H 2 O 2 and NH) was also given for 24 h. The pre-treatment strategy has already been demonstrated to maximize the impact of TU-mediated amelioration of AsV in rice 23 . In shoots, activities of antioxidant enzyme and Na + and K + accumulation were quantified in a time-course manner ranging from 1, 4, 24 and 48 h post-stress. In addition, 6 h after the onset of treatments, expression levels of selected saltresponsive genes were analyzed. At 7 days post-stress, phenotypic parameters, both qualitative and quantitative along with antioxidant capacity were recorded. In the second approach, four healthy 30 days old hydroponically grown seedlings were transferred to plastic pots in six groups (5 pots/group), under the net-house experimental facility of Bhabha Atomic Research Centre, Mumbai (India). The fertigation and agronomic protocols were followed as previously described 24 . Group-1 plants were treated with NaCl (11 g/pot) dose twice, at 42-and 57-days post-transplantation. The NaCl dose (22 g NaCl/per pot; equivalent to ~ 62 mM) was calculated considering the total water holding capacity of 14 kg paddy soil (4.6 L) and the top-water (1.4 L). The group-2 and -3 plants were given foliar applications of TU (6.5 mM containing 0.01% Tween-20) and H 2 O 2 (1 mM containing 0.01% Tween-20), respectively. A total of three foliar applications were given at vegetative, early anthesis and grain filling stages that corresponded to 40, 55 and 72 days post-transplantation, respectively. Group-4 and -5 plants were given the combined treatment of NaCl + TU (NT) and NaCl + H 2 O 2 (NH) treatments, respectively. Group-6 plants were foliar-sprayed with water (three times at 40, 55 and 72 days post-transplantation) and served as control. At 5 days post 3rd-foliar spray, various morphological traits (plant height, flag leaf length and width, tiller number, panicle number and length, chlorophyll content) were recorded and biochemical attributes (superoxide radical imaging, GR activity, ASA/DHA ratio and photosynthetic efficiency) were quantified from the flag leaves. Additionally, parameters of plant source-sink relationship and ion accumulation were quantified in three different tissues such as youngest flag leaf, old leaf from the bottom and developing inflorescence.

Measurement of antioxidant capacity and activities of antioxidant enzymes.
The non-enzymatic antioxidant status of leaf tissues was analyzed according to the Oxygen Radical Absorbance Capacity (ORAC) method 25 . For the measurement of antioxidant enzyme activities, total protein was extracted from liquid N 2 ground plant material (~ 250 mg) using the pre-chilled buffer [(was extracted using 100 mM chilled potassium phosphate buffer (pH 7.0) containing 0.1 mM EDTA and 1% polyvinyl pyrrolidone (w/v)]. The samples were centrifuged at 15,000×g for 15 min at 4 °C. The supernatant was separated and used for the measurement of superoxide dismutase (SOD), catalase (CAT), ascorbate peroxidase (APX) and glutathione reductase (GR) activities 26 . The protein content in the supernatant was quantified as per Bradford method 27 .
Quantification of plant growth and yield parameters. The various plant growth parameters viz shoot length, flag leaf length, leaf width and panicle length were quantified using a meter scale. The number of panicles per plant and 1000 seed weight were quantified manually. The leaf chlorophyll content was calculated as SPAD value using Chlorophyll Meter SPAD-502 plus-konica Minolta, representing an average value of five different points in the same leaf.
Measurement of stomatal conductance and PS-II stability. The gas exchange of leaves was measured using an Infrared Gas Analyzer, GFS-3000 (Walz, Germany). The photosynthetic photon flux density (PPFD) was fixed at 1000 μmol m −2 s −1 after optimization with a light curve. The photosynthetic efficiency was quantified using following parameters; cuvette air flow 750 mL min −1 , chamber temperature (25 °C), Relative humidity (60%) and atmospheric CO 2 concentration (400 ppm CO 2 ). Net photosynthetic rate, stomatal conductance, and transpiration were recorded simultaneously. Water use efficiency (WUE) was calculated as the ratio between net photosynthesis and transpiration. Using the differential minimum (F o ) and maximum fluorescence (F m ) signals of from the open and close PS II centers, maximum quantum efficiency of PSII (F v / F m ) was quantified. A 20 min prior dark adaptation is a prerequisite for the quantification of F v /F m, and can be defined as (F m − F o )/F m . Similarly, maximum fluorescence of dark (F m )and light adapted leaf (F m' ) were used for the quantification of Non photochemical quenching (NPQ); NPQ = Fm/F′m − 1. Further steady state chlorophyll fluorescence (Fs) was also quantified for the measurement of ETR and actual quantum efficiency of PSII (Φ PSII ); ΦPSII = (F′m − Fs)/F′m 28 .
Histochemical detection of superoxide radical and quantification of ascorbate pool. Superoxide radicals were detected in situ using nitroblue tetrazolium (NBT) staining 29 . NBT stain intensity was quantified using image J software (version 1.53d; https ://image j.nih.gov). The levels of ascorbate (ASA) and dehydroascorbate (DHA) contents were quantified using α-α′-bipyridyl-based colorimetric method 30 and the results were presented as ASA/DHA ratio.
Quantification of sucrose and starch levels. Lyophilized leaf sample (10 mg) was extracted in 15 mL of 80% ethanol. The extract was boiled for 10 min and then subjected to centrifugation at 15,000g for 15 min at room temperature. The sucrose and starch were quantified in the supernatant and pellet respectively using sucrose quantification kit (SCA-20; Sigma) and starch assay kit (STA-20; Sigma Aldrich), respectively according to the manufacturer's protocol.
Measurement of the activities of source-sink homeostasis related enzymes. For the quantification of enzymatic activities related to source sink homeostasis, total protein was extracted from liquid N 2 ground plant material (~ 300 mg) using the pre-chilled buffer [(containing 100 mM chilled MOPS (50 mM; pH 7.5), MgCl 2 (15 mM), EDTA (1 mM), poly-vinylpyrrolidone (2%; w/v), and phenyl methyl sulfonyl fluoride (2 mM)]. The samples were centrifuged at 12,000×g for 15 min at 4 °C. The supernatant was separated and used for the measurement of SPS, SuSy 31 and FPBase 32 . Additionally, Neutral invertase (NI) and acid soluble invertase (AI) activity in the plant samples were measured spectrophometrically following the extraction and assay methods 31 . The protein content in the sample was measured as per the Bradford method. All the enzyme activities were represented as units mg −1 protein which corresponds to μM of the product formed mg −1 protein min −1 .

Statistical analysis.
The experiments were conducted as randomized block design using three biological replicates. One-way analysis of variance (ANOVA) was performed with the whole dataset to confirm the variability of data and validity of results, and Duncan's multiple range test (DMRT) was performed to determine the significant difference between treatments. Different letters indicate significantly different values (DMRT, p ≤ 0.05). Principal component analysis (PCA) was performed with datasets of the source leaf and developing sink, using Origin 2016 (Origin Lab, Northampton, MA, USA), and the first two components (PC1 and PC2) explaining the maximum variance in the datasets were used to make biplots.

Results
TU and H 2 O 2 ameliorate NaCl stress through enhanced antioxidant capacity. There was a considerable growth reduction in terms of leaf drying and fragile stem in the NaCl-treated seedlings (Fig. 1A). Both root and shoot lengths were decreased by 14.75 and 16.36% (Fig. 1B,C) and fresh biomass was reduced by 35.03% (Fig. 1D) in NaCl-treated seedlings compared with that of control. Significant growth restoration was observed in NT and NH treatments respectively, in terms of shoot length (16.96 and 13.11%; Fig. 1B), root length (21.75 and 17.46%; Fig. 1C), and biomass (20.59 and 22.55%; Fig. 1D), compared to NaCl-treated  Fig. 2A-C). Additionally, SOD, CAT and GR activities were also increased under TU and H 2 O 2 alone treatments as compared with those of control; except for SOD activity in TU-treated leaves. In general, increase in antioxidant enzyme activities was higher in H 2 O 2 than TU alone treatment ( Fig. 2A Expression levels of early NaCl stress responsive genes under different treatments. The topranked eight early NaCl stress responsive genes were selected on the basis of published transcriptome of rice seedlings 33 and their expression levels were measured in the shoot tissue under different treatments. The results revealed that both TU and H 2 O 2 treatments led to upregulated expression under control as well as NaCl stress . After 7 days of treatment, differential phenotyping was observed qualitatively (A) and also quantified in terms of shoot length (B), root length (C), seedlings fresh weight (D). In addition, the total antioxidant capacity as trolox equivalent was also quantified from shoot part (E). For NT and NH, 24 h pretreatment of 7.5 µM TU and 1 µM H 2 O 2 , respectively was also given. All the values are mean of triplicates ± SD. Different letters indicate significantly different values (DMRT, p ≤ 0.05).    (Fig. 4A). The major differences were seen in terms of panicle number, panicle length and leaf width which were increased by 24.39, 13.46 and 13.33% in NT and 24.39, 10.10 and 6.67% in NH, respectively compared with those of NaCl treatment (Table 1A). Although the harvest index remained unchanged (Fig. 4B), 1000 seed weight and seed yield/plant were increased by ~ 30-34% (Fig. 4C) and ~ 25-27% (Fig. 4D) under NT and NH treatment, respectively compared to NaCl treatment. The major impact of TU and H 2 O 2 alone treatments was seen in terms of panicle number and leaf length respectively which were increased by 14 and 18.3%, compared with those of WS control (Table 1A). In addition, the harvest index, 1000 seed weight and seed yield/plant were also increased by 13 (Fig. 5A). The activity of GR and ASA/DHA ratio were increased and decreased by 28.5 and 30.7%, in NaCl-treated seedlings compared with those of control. Both these parameters were further increased by 27.94 and 47.6% (NT) and 55 and 67% (NH), respectively compared with that of NaCl treatment (Fig. 5B,C). Although GR activity remained unchanged in TU, it was increased by 66.6% under H 2 O 2 alone treatment (Fig. 5B). No significant change was observed for the ASA/DHA ratio under TU and H 2 O 2 alone treatment, compared to control (Fig. 5C).
Under NaCl stress, maximum Na + accumulation was observed in young leaves (YL), where it increased by 403.8%, followed by old leaves (OL) and developing inflorescences (DI), where it increased by 159.94 and 70.21%, respectively, compared with WS-treated plants (Supplementary table 3A). In contrast, under NT treatment, Na + accumulation from YL and DI organs wasecreased by 21.48 and 53.75%, respectively; however, in OL, it was increased by 15.97%, compared with those of NaCl-treated plants. In NH-treated plants, the Na + accumulation was reduced in all the three tested organs including YL, OL and DI by 38.57, 21.48 and 50%, respectively compared with those of NaCl-treated plants. No significant change in Na + accumulation was observed in TU while H 2 O 2 alone treatments resulted in 34.18% decrease and 45.91% increase the Na + accumulation in the YL and OL respectively compared to those of WS control (Supplementary table 3A). The increased Na + levels in . NT and NH denote combined treatment of NaCl + TU and NaCl + H 2 O 2 treatments, respectively. A total of three foliar applications were given at vegetative, early anthesis and grain filling stages that corresponded to 40, 55 and 72 days post-transplantation, respectively. At the time of maturity, the differential phenotype was recorded (A) and representative panicles were shown (B). In addition, the yield parameters such as harvest index (C), seed yield per plant (D) and 1000 seed weight (E) were also quantified.

Photosynthetic responses under TU and H 2 O 2 supplementation.
The overall process of photosynthesis was negatively affected in NaCl-treated plants. The major impact was seen in terms of photosynthetic rate (PR), water use efficiency (WUE) and electron transport rate (ETR) which were decreased by 69.35, 61.90 and 53%, respectively compared with those of WS control. In contrast, significant photosynthetic recovery was noticed under both NT and NH treatments as these parameters were increased by 78.95, 62.5 and 101.23% in NT and 105.26, 75 and 72.84% in NH compared with that of NaCl treatment. In addition, quantum yield of photosystem II (PS-II yield) and non photochemical quenching (NPQ) were also increased specifically under NT (56.32%) and NH (31.25%), respectively compared with those of NaCl treatment (

Modulation of source-sink homeostasis under TU and H 2 O 2 supplementation.
Akin to photosynthesis, the activities of key enzymes determining source (SPS and FBPase) and sink (AI, NI and SuSy) strength were also negatively impacted in NaCl-treated plants; except, for SuSy which was increased by 108.75%, in old leaves compared with WS control. The ameliorative effects were observed under both NT and NH treatments, at the level of source as well as sink. Besides, differential nature of NT and NH was also seen, especially in the source leaves. For instance, the SPS activity was increased by 93.47, 29.75 and 40.37% in NH-treated YL, ML and DI organs, respectively compared with those of NaCl treatment. However, under NT treatment, increased SPS activity was limited to YL (74.7%) and DI (22.36%) organ only. Unlike SPS, FBPase activity was significantly increased under both NT and NH, in all the three tested organs, compared with those of NaCl treatment. The sucrose degradation pathway was also activated under NT and NH treatment conditions. In source organs like YL and OL, the NI activity was increased 66.43 and 40.44% in NT and 59.05 and 120.97% in NH, respectively compared with those of NaCl treatment (Fig. 6C). Besides, in YL organ, NI and SuSy activities were also increased by 59.21 and 96.05% in NT and NH, respectively compared with those of NaCl treatment (Fig. 6D,E). In sink (DI), the activities of AI and SuSy were increased respectively by 381.99 and 141.38% in NT and 391.74 and 151.54% in NH compared with those of NaCl treatment (Fig. 6D,E). Under TU and H 2 O 2 alone treatments, well-coordinated sucrose biosynthesis, mediated by SPS and FBPase, was observed in YL (Fig. 6A,B). In parallel, SuSy mediated sucrose breakdown was also observed under YL and DI; while AI was found to be activated in OL organ (Fig. 6C-E). Understanding treatment-variable interactions through PCA based clustering. PCA was performed on the entire data sets to identify the key variables associated under various treatment conditions. In source leaves, different treatments were grouped into three categories (Fig. 7A). The first category contained photosynthesis and sucrose biosynthesis and breakdown related parameters and was found to be associated with WS and TU and H 2 O 2 treatments. The second category had only one attribute (GR activity) which was associated with NT as well as NH treatments. The third category represented NaCl stress and this was not associated with any of the variables (Fig. 7A). Similarly, three major categories were also identified in sink organs. The first category included plant growth and yield related parameters which was associated with WS, TU and H 2 O 2 treat-   www.nature.com/scientificreports/ ments. The second category included source-sink homeostasis related enzymes/metabolites, associated with NT and NH treatments. The third category of sucrose and Na + content was associated with NaCl treatment (Fig. 7B).

Discussion
Redox homeostasis is an essential constituent for sustained maintenance of plant growth and survival under stress conditions. The present study has evaluated widely used redox modulators like TU and H 2 O 2 for mitigating salt stress in rice. In our previous studies, we have successfully demonstrated TU-mediated responses at multiple levels of organization. At the physiological level, TU improved source-to-sink relationship leading to increased crop yield 34 while at the molecular level, it improved cellular energetics 35 , co-ordinated calcium and abscisic acid (ABA) signaling events 36 , maintained plant-water homeostasis 37 , enhanced antioxidant defense 20 and improved www.nature.com/scientificreports/ sulphur metabolism 23 . Additionally, TU effectiveness has also been demonstrated under other types of abiotic stresses 38 , like drought, heat, UV radiation, metal stress. Similarly, H 2 O 2 is another broad range effective PBR whose ameliorative potential has been demonstrated against various abiotic stresses 39 . Considering their chemical nature, TU (a ROS scavenger) and H 2 O 2 (a biological ROS) are expected to have a contrasting effect on the cellular redox state, which is described as an integrated ratio of reduced to oxidized form of all the redox couples present inside the cell. At whole plant level, the redox state is regulated by ROSscavenging/producing enzymes and antioxidant metabolites 40 . Initially, the post-germination phenotyping was performed at seedling stage with variable doses of NaCl (ranging from 50 to 100 mM) and 50 mM was identified as the IC50 dose at which plant biomass is reduced by 48.15% compared with those of control ( Supplementary  Fig. 1). Similar dose-dependent analysis was also performed with TU and H 2 O 2 (ranging from 1 to 100 μM). On the basis of biomass, 7.5 and 1.0 μM were identified as optimum doses of TU and H 2 O 2 which were selected for all the medium supplementation studies (Supplementary Fig. 2). In spite of having a contrasting chemistry, in planta supplementation of both TU as well as H 2 O 2 increased the redox capacity and improved the plant growth potential under NaCl stress conditions (Fig. 1). Such an overlapping response clearly indicated that H 2 O 2 supplementation also generates reducing redox environment of the plants, which is shown by the significant increase in activities of antioxidant enzymes like SOD, CAT and GR under both control and NaCl stress conditions ( Fig. 2A-C)  www.nature.com/scientificreports/ various crops like soybean 41 , rice 42 and wheat 43 . Unlike H 2 O 2 , the enzymatic antioxidants were less pronounced in TU-treated control plants, indicating that non-enzymatic antioxidants were involved in maintaining reducing redox environment. Earlier, TU-mediated activation of non-enzymatic antioxidants has been demonstrated in Brassica juncea 44 . Since most of the NaCl-induced damages including biomass reduction as well as decreased root and shoot length were associated with ROS-induced redox imbalance 45 , therefore, the NaCl ameliorative potential of TU and H 2 O 2 could largely be associated with their ability to generate reducing redox environment inside the plants.
In addition to redox balance, improved K + retention was also observed under both NT-and NH-treated plants ( Fig. 2F; Supplementary table 2), which might have supported the improved plant phenotype under NaCl stress conditions (Fig. 1). A positive correlation between salt tolerance and higher K + accumulation has been demonstrated in crops like rice and barley 46 . The NaCl-induced K + -leakage is associated with ROS production which activates K + -efflux channels including guard cell outward rectifying K + channel (GORK) and stelar K + outward rectifier (SKOR) 47 . The higher K + retention ability of TU and H 2 O 2 further substantiates our hypothesis that both these modulators have an overlapping capacity to generate reducing redox environment. This was also supported by TU and H 2 O 2 dependent upregulated expression of HAK21 (Fig. 3A), which is known to maintain K + ion homeostasis under NaCl stress conditions 8 . Other NaCl stress responsive genes, especially LEA 1 , dehydrin 7 , TSPO 10 and EN20 11 , known for imparting NaCl tolerance in redox dependent manner, were also upregulated suggesting their association in the tolerant phenotype observed under NT and NH treatments. Incidentally, their redox-dependent regulation has already been demonstrated [48][49][50] , further justifying their regulation through TU and H 2 O 2 dependent manner. In addition, the expression of most of the tested NaCl tolerance related genes were also increased constitutively even under stress-free conditions (Fig. 3). This result indicated that during TU/H 2 O 2 pre-treatment phase, the seedlings were better equipped to face the ensuing NaCl stress exposure.
Similar to nutrient medium supplementation, foliar-applied TU and H 2 O 2 was effective in maintaining the ROS and redox homeostasis as evident by decreased superoxide load under NT and NH treatments compared to NaCl (Fig. 5). Further H 2 O 2 mediated increase in NBT staining under control conditions (Fig. 5A) may be considered as pro-oxidant behavior by imposing mild oxidative stress. Foliar-application of TU and H 2 O 2 was also found effective in enhancing plant growth and yield under field conditions in both control and NaCl stress conditions (Fig. 4), substantiating their agronomic feasibility. Considering the difference in plant size and possible degradation under the natural sunlight, we have selected higher doses of TU (6.5 mM) and H 2 O 2 (1 mM) for foliar supplementation, which were also used previously in various other crops 19 . The significant reduction in PS-II efficiency under NaCl treatment was reflected by a drastic reduction in plant growth as well as yield ( Fig. 4; Table 1). The redox balancing, better K + retention and improved photosynthesis were identified as key attributes responsible for mediating TU and H 2 O 2 dependent amelioration under NaCl stress conditions. Of these, K + retention and dynamic ROS production in leaf mesophyll cell have been shown to be well-correlated with salt tolerance in rice at reproductive stage, in both greenhouse as well as field conditions 51 . Besides, K + and ROS are considered as major signaling regulators and hence, have potential capacity to alter plant growth 52,53 . In line with K + retention, we also observed reduced Na + accumulation under both NT and NH treatments, especially in YL and DI which represent the main organs for photosynthesis and reproduction, respectively (Supplementary table 3).
The Na + dependent inhibition of photosynthesis has been reported in various crops like rice 54 , barley 55 and wheat 56 . The lower Na + accumulation as observed in our study could have facilitated the improved photosynthesis (Table 1B) and overall plant growth and vigor. Although, there are multiple ways by which Na + imposes toxicity; imbalance in Na + /K + ratio is a major contributor of Na + mediated reduction in plant photosynthesis 57 . Most of the PS-II efficiency components (photosynthesis rate, WUE, Fv/Fm, ETR and quantum yield of PSII) were equally improved; however, NPQ was identified as NH-specific parameter (Table 1B). The excess energy dissipation through NPQ protects the photosystem from ROS mediated photoinhibition. This is supported by H 2 O 2 dependent regulation of xanthophylls 58 , which represent the major components for maintaining higher NPQ and low ROS under salt stress conditions 59 .
In plants, sucrose and starch are considered as key indicators of source (leaves) and sink (developing inflorescence) strengths, respectively 60 . The higher and lower levels of sucrose and starch in source and vice versa in sink, clearly indicated a synchronized source-sink relationship in both NT-and NH-treated plants (Supplementary Table 4). Most of the Calvin cycle enzymes are regulated in redox dependent manner 61 . For example, FBPase (a sucrose biosynthesis enzyme) is active under the reducing redox environment 62 . Thus, owing to the generation of reducing redox environment (Fig. 5C), the overall source strength was high resulting in the improved growth potential of NT-and NH-treated plants. This was also supported by PCA in source leaf, wherein NT-and NH-treatments were found to be associated with GR activity (Fig. 7A). GR is the major enzyme for regenerating GSH and hence, is responsible for maintaining reducing redox conditions in the plant 63 . Similarly, overexpression of SPS (a rate-limiting enzyme of sucrose biosynthesis 64 has been achieved in sugarcane transgenic lines which accumulated more sucrose with significant increase in plant height and stalk number compared to non-transgenic control 65 . In addition, de-regulation between FBPase and SPS activities was also observed, suggesting the onset of senescence in NT-treated OL organ. This is in contrast with NH wherein, increased AI and NI activities avoided the feedback inhibition and hence, sucrose biosynthesis was found to be active in both YL and OL organs (Fig. 6A,B). In both NT-and NH-treated sink organs, SuSy and AI were preferred (over NI) for breaking down the sucrose metabolites (Fig. 6C-E). The higher starch content will not only improve sink strength but also, the overall fitness of the plant 66 . In contrast, NI activity was also observed in OL organ, which can provide hexoses, as they are necessary for restricting ROS level 67 and also to fulfill energy demand of plants under stress conditions 68 . Similar to our results, SuSy overexpression 69 or mutation in AI 70 have been shown to increase or decrease the grain weight, respectively, in transgenic lines of rice. Taken together, the approach of www.nature.com/scientificreports/ using TU and H 2 O 2 can be seen as an alternative to the genetic methods of enhancing source-sink strength in plants under NaCl stress conditions. A significant enhancement in growth and yield was also noticed under TU and H 2 O 2 alone treatments ( Fig. 4; Table 1). Although the absolute levels of sucrose and starch remained unchanged (Supplementary table 4), the higher activities of SPS and FBPase in YL and SuSy in both YL and DI organs clearly indicated that source-sink strength was boosted in TU and H 2 O 2 treated plants (Fig. 6). Further, PCA also indicated an overlapping response as most of the growth and yield related attributes were grouped together with between TU and H 2 O 2 alone treatments (Fig. 7). Thus, the positive impact of TU and H 2 O 2 under both control and NaCl stress conditions greatly increases their versatility to be applied under the realistic field scenario.

Conclusions
In conclusion, the study highlights that despite having contrasting redox chemistry, both TU and H 2 O 2 impart comparable level of NaCl stress tolerance in rice to a comparable extent. Both TU and H 2 O 2 upregulated the expression of NaCl stress responsive genes in a constitutive manner, without showing any significant growth reduction under control conditions. In addition, reducing redox status was maintained along with better K + retention and upregulated expression of NaCl stress tolerant genes like HAK21, LEA1, dehydrin, TSPO and EN10 under both NT and NH treated seedlings, resulting in improved growth. Under both control and NaCl stress conditions, foliar-supplemented TU and H 2 O 2 improved the growth and yield attributes of the plants. The ameliorated phenotype under NT and NH was associated with reduced Na + accumulation, improved photosynthesis efficiency and better source-sink relationship. Taken together, the results of this study suggest that the maintenance of reduced redox status acts as a "central" regulator for mitigating NaCl-induced damages. Besides, it also extends the concept of using redox modulators for improving rice crop productivity under the realistic field scenario.