Enhancing trehalose biosynthesis improves yield potential in marker-free transgenic rice under drought, saline, and sodic conditions

Marker-free transgenic lines of rice are developed with enhanced trehalose accumulation that is associated with improved grain yield under salinity, sodicity, and drought stresses.

Abstract

Edaphic factors such as salinity, sodicity, and drought adversely affect crop productivity, either alone or in combination. Despite soil sodicity being reported as an increasing problem worldwide, limited efforts have been made to address this issue. In the present study, we aimed to generate rice with tolerance to sodicity in conjunction with tolerance to salinity and drought. Using a fusion gene from E. coli coding for trehalose-6-phosphate synthase/phosphatase (TPSP) under the control of an ABA-inducible promoter, we generated marker-free, high-yielding transgenic rice (in the IR64 background) that can tolerate high pH (~9.9), high EC (~10.0 dS m^–1), and severe drought (30–35% soil moisture content). The transgenic plants retained higher relative water content (RWC), chlorophyll content, K⁺/Na⁺ ratio, stomatal conductance, and photosynthetic efficiency compared to the wild-type under these stresses. Positive correlations between trehalose overproduction and high-yield parameters were observed under drought, saline, and sodic conditions. Metabolic profiling using GC-MS indicated that overproduction of trehalose in leaves differently modulated other metabolic switches, leading to significant changes in the levels of sugars, amino acids, and organic acids in transgenic plants under control and stress conditions. Our findings reveal a novel potential technological solution to tackle multiple stresses under changing climatic conditions.

Introduction

Salinity/sodicity (alkalinity) and drought stress are the major abiotic stress factors that cause detrimental effects on the growth and yield of many crops, including rice. Globally, about 831 Mha of land has salt-affected soils (Zhang et al., 2018), within which crop losses caused by irrigation-induced salinity are substantial and are estimated to be equivalent to more than US$ 27 billion per year (Qadir et al., 2014). In addition, almost 34 Mha of rain-fed lowland and 8 Mha of upland rice in Asia is estimated to be prone to drought, of which 6.3 Mha of upland and 7.3 Mha of lowland in India are severely affected by drought (Verulkar et al., 2010; Singh et al., 2016). Further, 581 Mha is affected by sodicity, of which 340 Mha is found in Australia (Eskandari et al., 2018). In India, 3.8 Mha (66% of the total saline soil) is affected by sodicity (Gupta Choudhury et al., 2018). Excess exchangeable sodium in sodic soil adversely affects its physical and nutritional properties, leading to substantial yield losses (Singh et al., 2018). Over the last two decades, several transgenic plants with improved tolerance to abiotic stresses have been developed by targeting genes involved in different mechanisms, including transcriptional regulation, signaling, ion homeostasis, biosynthesis of compatible solutes, and antioxidant defense (Singh et al., 2008, 2010; Tripathi et al., 2016; Gupta et al., 2018; Joshi et al., 2018; Soda et al., 2018; Nongpiur et al., 2020), but none of these reports have claimed improved tolerance towards soil sodicity. Among the host of genes that have been utilized by molecular biologists, those contributing to the build-up of osmolytes in the cell have been a favorite choice (Grover et al., 1993; Nadel et al., 2019). Trehalose (α-D-glucopyranosyl-α-D-glucopyranoside), a non-reducing disaccharide, is a principal compatible solute and a potential signal metabolite in plants. It stabilizes dehydrated enzymes, proteins, and lipid membranes during desiccation (Kosar et al., 2019). Plants mostly accumulate trace amounts of trehalose, except highly desiccation-tolerant resurrection plants such as Selaginella lepidophylla (130 mg g⁻¹ DW), and Myrothamnus flabellifolia (34 mg g⁻¹ DW) (Adams et al., 1990; Drennan et al., 1993; Pampurova and Dijck, 2014).

Trehalose biosynthesis is a two-step enzymatic process, where trehalose-6-phosphate (T6P) is synthesized from UDP-glucose (UDPG) and glucose-6-P (G6P) by trehalose-6-phosphate synthase (TPS) and subsequently converted to trehalose by trehalose-6-phosphate phosphatase (TPP). Previous studies have demonstrated stress-inducible expression of TPS in cotton (Wang et al., 2016) and sweet potato (Wang et al., 2019). Ectopic expression of the yeast TPS1 gene in potato, tobacco, and maize results in higher accumulation of trehalose that confers dehydration tolerance (Lee et al., 2003; Kondrák et al., 2012; Liu et al., 2015). Similarly, overexpression of TPS1 isolated from Arabidopsis, cotton, and rice has been shown to improve drought, salinity, and cold tolerance in Arabidopsis and rice, indicating that plant TPS also has a high potential in imparting abiotic stress tolerance (Avonce et al., 2004; Li et al., 2011; Wang et al., 2016). It has been reported that overexpression of bifunctional fusion of the TPS and TPP genes of E. coli under a stress-regulated promoter confers tolerance to salinity, drought, and cold stress in rice seedlings (Garg et al., 2002) and drought and heat stress in tomato (Lyu et al., 2013, 2018). Similarly, TPS and TPP overexpression in rice under the maize ubiquitin promoter (Ubi1) results in higher trehalose production and tolerance to drought, salinity, and cold without stunting growth (Jang et al., 2003). In addition, higher tolerance against drought, freezing, salt, and heat was observed in Arabidopsis transgenic lines overexpressing yeast TPS and TPP gene fusion construct (Miranda et al., 2007). However, none of these reports have assessed the performance of the transgenic plants for their yield output under simulated field conditions.

Despite several studies on transgenic plants with trehalose overexpression, little is known about the metabolic role of trehalose under control and stress conditions. Development of high-yielding, multiple stress-tolerant, marker-free transgenic plants that produce nutritionally stable grains is greatly needed. The antibiotic marker-based selection methods used in the process of generating transgenic plants may have several biosafety concerns related to allergenicity in humans and environmental hazards (Miki et al., 2009; Yau and Stewart, 2013). Therefore, with the aim of better public acceptance and in order to promote commercial deployment, we have developed marker-free transgenic rice (cv. IR64 background) by overexpressing both TPS and TPP as a fusion gene construct and using glucose as the selection agent. IR64 is an early-maturing, high-yielding, disease-resistant mega variety with excellent cooking quality characteristics. However, it is sensitive to abiotic stresses, such as drought and salinity (Mackill and Khush, 2018). We were able to demonstrate that elevated levels of trehalose in the transgenic plants conferred tolerance to multiple abiotic stresses including salinity, sodicity, and drought with reduced yield penalties, as evidenced by various physiological and agronomical parameters. Most importantly, these evaluations were carried out in hydroponics (salinity) and field-simulated microplots (sodicity). In addition, metabolic profiling of the transgenic lines was carried out in order to understand the effects of trehalose overproduction on quantitative alterations of other metabolites that resulted in physiological changes that provide tolerance against salinity, sodicity, and drought. Metabolic profiling of wild-type and TPSP transgenic seeds showed similar nutritional levels, ensuring nutritional equivalence in the transgenic grains. Thus, the present study provides an effective approach for achieving yield and nutritional stability of rice under multiple abiotic stress conditions, including sodicity.

Materials and methods

Generation of marker-free transgenic rice using glucose as a selection agent

To develop the marker-free transgenic lines, the pCAMBIA1300 vector was obtained from Prof. Ray J. Wu (Department of Molecular Biology and Genetics, Cornell University, USA). This vector contains the trehalose biosynthetic fusion gene (TPSP) that contains the coding regions of the E. coli otsA and otsB genes that encode TPS and TPP, respectively, under the transcriptional regulation of an ABA-inducible (ABRC) promoter. The vector was introduced into Oryza sativa subsp. indica rice cv. IR64 by Agrobacterium-mediated transformation. Based on a previous study by Avonce et al. (2004), who reported that AtTPS1 overexpression in Arabidopsis strongly reduces glucose sensitivity, we decided to screen the TPSP-transformed calli on media containing glucose as a selection agent. The sub-lethal and lethal doses of glucose in the callus-induction medium were found to be 7 g l^–1 and 10 g l^–1, respectively, for the untransformed calli, and hence subsequent selection of transformed calli was carried out at 10 g l^–1. Because of this unique metabolite-based selection strategy, any clone not carrying the TPSP fusion gene construct (such as calli transformed with the empty vector) will not survive. Therefore, vector-transformed controls were not included in the present study. Putative transgenics overexpressing TPSP were grown in earthernware pots containing vermiculite under greenhouse conditions at 28±2 °C under 16/8 h day/night conditions.

Confirmation of putative transgenic plants by PCR and Southern blot analysis

Putative transformants were screened through PCR analysis using the genomic DNA of the wild-type (WT) and transgenic lines as the template and ABRC-Fw and TPSP-Re primers (Supplementary Table S3 at JXB online). The plants identified as positive by PCR were further analysed for the integration of the TPSP fusion gene using Southern blot analysis as described previously (Joshi et al., 2016a).

Raising of polyclonal anti-TPS antibodies and western blotting

The polyclonal antibodies against the TPS protein were developed by cloning the TPS gene in the bacterial expression vector pET14b as a 6X-His-tagged fusion. The TPS protein was expressed in BL21(DE3) cells by induction with 0.5 mM IPTG for 6 h at 37 °C. The recombinant protein was purified using Ni-NTA affinity chromatography and used to raise polyclonal antibodies in rabbit as described previously (Singh et al., 2012). Total soluble proteins were isolated from WT and transgenic plants (treated with 100 µM ABA for 24 h) and used for western blotting using anti-TPS antibodies as described previously (Singla-Pareek et al., 2003).

Leaf-strip senescence assay

A rapid, detached-leaf senescence bioassay was performed as described previously (Singla-Pareek et al., 2008). TPSP plants were grown in earthernware pots containing soil under greenhouse conditions at 28±2 °C under 16/8 h day/night conditions. After 60 d of growth, strips were cut from the youngest healthy fully expanded leaf from each plant and floated for 5 d on half-strength Yoshida medium (controls) or on medium containing 200 mM NaCl for salinity, 5% PEG-6000 for drought (Verslues et al., 2006), and NaHCO₃ + Na₂CO₃ for sodicity (pH ∼9.0), after which chlorophyll content, electrolyte leakage, and RWC were measured as described previously (Singla-Pareek et al., 2008). Experiments were performed three times, with three replicates each (n=9).

Germination and seedling growth assays under salinity, sodicity, and drought stress

Seeds from WT and transgenic plants from the T₃ generation were germinated on Yoshida medium supplemented with 200 mM NaCl for salinity, 5% PEG-6000 for drought, and NaHCO₃ + Na₂CO₃ for sodicity (pH ~9.0) for 6 d. Seed germination rates were measured daily, and shoot length and root length were measured at 6 d post-germination, at which point images were taken. Data presented are means (±SD) of three independent experiments with ten seedlings in each case (n=30). The statistical significance of the data was tested by ANOVA.

Fluorescent dye-based detection of intracellular Na+ and cell death in roots

To detect sodium ions in planta, 14-d-old WT and transgenic seedlings that had been grown in a growth chamber at 28±2 °C under 16/8 h day/night conditions, were exposed to 200 mM NaCl in hydroponics for 48 h. Segments of 0.5 cm from the root tip were stained with 10 μM CoroNa Green AM dye and 3 μg ml^–1 propidium iodide (PI; Gupta et al., 2018). Detection of Na⁺ fluorescence was carried out using a confocal laser-scanning microscope (Nikon) with 488 nm excitation and 515 nm emission, and PI fluorescence was visualized at 488 nm excitation and 595 nm emission.

Growth performance and yield attributes under salinity, sodicity and drought stress conditions

To evaluate the salt stress tolerance of the transgenic plants in comparison to the WT at the reproductive stage, salt stress was imposed on 2-month-old plants grown in soil in earthen pots under greenhouse conditions by irrigating them with 200 mM NaCl (Joshi et al., 2018). For sodicity stress, the desired level (pH ∼9.9) was attained by adding sodium bicarbonate (NaHCO₃) and sodium carbonate (Na₂CO₃) to the soil (Krishnamurthy et al., 2016). For drought stress, water was withheld for 25 d, followed by recovery by rewatering until maturity. The soil moisture content was measured using a Field Scout Soil Sensor Reader SM100 (Spectrum Technologies Inc., USA).

Chlorophyll content, K⁺/Na⁺ ratio, and RWC were measured after 25 d of salinity, sodicity, and drought stress, and at 30 d post-recovery after drought stress. Net photosynthetic rate (NPR), electron transport rate (ETR), transpiration rate (TR), stomatal conductance (g_s), and F_v/F_m were also measured using a LI-6400XT portable photosynthesis system (LI-COR). The WT and transgenic lines were harvested after maturity and total biomass, panicle number per plant, filled grains per plant, grain weight per plant, and harvest index were determined as described previously (Joshi et al., 2018).

Microplots/hydroponics study to evaluate yield performance under salinity and sodicity stresses inside transgenic green house

Phenotyping of transgenic lines in relation to salinity and sodicity stress was conducted under biosafety compliant transgenic greenhouse conditions at the ICAR Central Soil Salinity Research Institute, Karnal, India. Screening for salinity stress was conducted in hydroponics with modified Yoshida nutrient solution supplemented with the required amount of salts to create high salinity conditions (ECiw ~10.0 dS m^–1) as detailed by Singh and Flowers (2010). For salinity screening, nursery seedlings were inserted in equidistant holes in thermocol sheets floating on a saline nutrient hydroponic solution in 2×1-m PVC tanks. The solution was refreshed every 5 d and maintained at pH 4.5 until maturity. For non-stress conditions, plants were grown in hydroponics with nutrient solution only. For high sodic stress screening, the TPSP lines were grown under artificially created sodic soil designed to simulate natural sodic conditions (pH ~9.9) by adding NaHCO₃ + Na₂CO₃ to the soil in microplots (6×3×1 m) inside transgenic greenhouse (Ali et al., 2013). . The transgenic and WT plants were transplanted in rows (replicated three times) and maintained until maturity in all the treatments. At maturity (20–25% seed moisture content), the plants were harvested, and various agro-morphological parameters were recorded. Correlation coefficients were determined using PAST (PAlaeontological STatistics, ver.1.89).

GC-MS based metabolite profiling in WT and transgenic lines

The metabolites from leaves and dry seeds (12–14% moisture content) were extracted as described by Soda et al., (2018). The samples were derivatized by adding 100 μl N-Methyl-N-(trimethylsilyl) trifluoroacetamide (MSTFA) and analysed using GCMS-QP-2010 PLUS (Shimadzu). Identification and quantification of trehalose was performed by comparing its retention time with a standard, and other metabolites were identified by comparing their retention time and mass spectra with the entries of the NIST14 (https://www.sisweb.com/manuals/nist.htm) and Wiley 08 mass spectral libraries (https://www.sisweb.com/software/wiley-registry.htm). The data were normalized using ribitol as the internal standard (Soda et al., 2018). Peak identification and spectral deconvolution were done using the GC-MS solution software provided by the manufacturer. Pair-wise comparisons of the metabolite profiles were obtained by comparing the mean standardized peak areas for three replicates. Statistical significance between means was calculated using Student’s t-test.

Results

Ectopically expressing the TPSP gene construct shows stable integration in transgenic rice

To generate TPSP-overexpressing transgenic rice plants, the TPS+TPP fusion construct in the pCAMBIA1300 vector under the ABRC promoter (Fig. 1A) was introduced into indica rice cv. IR64 by Agrobacterium-mediated transformation. The TPSP-transformed calli were selected by adding glucose (10g l^–1) in the growth medium. The integration of the TPSP fusion gene construct into the rice genome was confirmed by PCR analysis of the putative transgenic lines with ABRC-Fw and TPSP-Re primers, resulting in the amplification of a 700-bp region (Fig. 1B). We generated several transgenic lines (at least 12 lines that were PCR-positive for the transgene), which were further checked through Southern hybridization and were found to contain stable integration of the transgene (Fig. 1C) with a single-copy insertion (Fig. 1D). To confirm these findings at the protein level, western blot analysis was performed on the WT (cv. IR64) and the positive transgenic lines using anti-TPS antibodies, and this showed the presence of a protein with the expected apparent molecular mass (corresponding to the TPSP fusion protein) in the transgenic plants (Fig. 1E). We selected the transgenic lines Ta-2 and X-93 for further detailed analysis on the basis that they showed single-copy gene insertion and the highest signals on the western blots.

Fig. 1.

The marker-free TPSP construct and confirmation of transgene integration in rice cv. IR64. (A) Schematic diagram of the binary plasmid containing the trehalose biosynthetic fusion gene (TPSP) that includes the coding regions of E. coli otsA and otsB encoding TPS and TPP, respectively, under the transcriptional control of the ABRC promoter. (B) PCR analysis of putative transgenic rice plants with ABRC-forward and TPSP-reverse primers resulted in amplification of a 700-bp region, which corresponds to the pCAMBIA-TPSP vector used as the template for the positive control (PC). Southern hybridization shows that all four putative transgenic lines contained the TPSP gene construct (C) and a have single-copy insertion of the TPSP gene (D). (E) Western blot analysis with anti-TPS polyclonal antibody indicating accumulation of the TPSP protein after 24 h in the presence of 100 µM ABA. Total soluble protein isolated from transgenic lines was loaded on a polyacrylamide gel and run on SDS-PAGE, and showed an equal loading pattern as stained by CBB dye (lower panel). The corresponding gel was transferred to Hybond-C+ membrane for western blotting analysis (upper panel).

TPSP plants exhibit delayed leaf senescence, higher chlorophyll retention, and maintain cellular homeostasis under abiotic stresses

The transgenic plants were tested for tolerance against salinity, sodicity, and drought stresses using a leaf-strip senescence assay (Fig. 2A; PEG was used for the drought treatment). Significant differences in chlorophyll retention were observed among WT and transgenic plants after 5 d under the stresses (Fig. 2B). Both the TPSP-overexpressing lines maintained higher chlorophyll concentrations compared with the WT under salinity (Ta-2, 78% and X-93, 82%), sodicity (63% and 69%), and drought (60% and 61%). A significant increase in electrolyte leakage was detected in the WT plants relative to the transgenic lines under all three stresses (Fig. 2C). Thus, TPSP-plants showed higher membrane stability under the various stresses in comparison to the WT. Determination of RWC indicated that the transgenic lines retained similar significantly higher water contents than the WT plants under the three stresses (Fig. 2D).

Fig. 2.

Overexpression of TPSP improves the tolerance of rice plants towards multiple abiotic stresses. (A) Leaf-strip senescence assays were carried out to test the ability of the TPSP-overexpressing transgenic rice lines (Ta-2 and X-93) to tolerate salinity (200 mM NaCl), sodicity (pH ~9.0), and dehydration (5% PEG). Two replicates of each line (RepI and RepII) are shown for each treatment. Senescence was assessed after 5 d of stress. Leaf strips floated on half-strength Yoshida medium only acted as control samples. Leaf-strips from the wild-type (WT) and the TPSP-overexpressing lines were further used for determination of (B) total chlorophyll, (C) electrolyte leakage, and (D) relative water content (RWC). Data are means (±SD), n=9. Significant differences for each genotype compared with control conditions were determined using Student’s t-test: * P<0.05.

To further assess the stress tolerance capabilities, WT and TPSP transgenic seeds at the T₃ generation were germinated in the presence of 200 mM NaCl for salinity, 5% PEG-6000 for drought, and NaHCO₃ + Na₂CO₃ for sodicity (pH ~9.0). Under control conditions, almost 100% seed germination was recorded in all the genotypes with no significant differences in growth rate, shoot length, and root length (Fig. 3A, I–K). Under saline conditions, only 15% germination was observed in the WT as compared to 90% in the transgenic lines (Fig. 3B, F), and the transgenic seedlings also exhibited better growth as evidenced by greater shoot and root length (Fig. 3J, K). Similar decreases in germination in the WT were also observed in response to sodicity (Fig. 3C, G) and drought stress (Fig. 3D, H), and similar decreases in growth were also observed (Fig. 3I–K).

Fig. 3.

Seed germination and growth of young seedlings for wild-type (WT) and transgenic rice under salinity, sodicity and drought stress. Seeds of the WT and T₃ seeds of the TPSP-overexpressing transgenic lines Ta-2 and X-93 were germinated for 6 d under (A) control conditions, (B) salinity (200 mM NaCl), (C) sodicity (pH=9.0), and (D) drought (5% PEG). (E–H) Germination rates of the WT and transgenic lines corresponding to treatments shown in (A–D). (I) Growth performance of WT and TPSP-transgenic seedlings. (J) Shoot length and (K) root length of WT and transgenic lines under control, salinity, sodicity, and drought stress conditions. Data are means (±SD), n=30. Significant differences compared to the WT under the same treatment were determined using Student’s t-test: *P<0.05.

TPSP-transgenic plants exhibit less sodium ion accumulation and less damage in roots under salinity stress

In planta accumulation of Na⁺ and cell death in the roots were monitored using CoroNa green and propidium iodide stains, respectively. Fluorescence imaging indicated much higher accumulation of Na⁺ in the roots of the WT compared with the transgenic lines (Fig. 4B) and this was associated with much higher rates of cell death in WT roots (Fig. 4C).

Fig. 4.

Relative Na⁺ content and cell death in the roots of wild-type (WT) and TPSP-overexpressing transgenic rice lines. Seedlings of the WT and two transgenic lines (Ta-2 and X-93) were treated with 200 mM NaCl for 48 h followed by staining with CoroNa dye or propidium iodide (PI). Root tips were then visualized using confocal microscopy. (A) Root tips under bright light (TD), (B) stained with CoroNa Green, and (C) stained with PI.

Overproduction of trehalose promotes tolerance to salinity, sodicity, and drought stresses in transgenic plants

The responses of WT and the TPSP-overexpressing plants to salinity (EC 10 dS m^–1), sodicity (pH ∼9.9), and drought (water withheld to 30–35% soil moisture content) during the pre-flowering (reproductive) stage were examined in pot-grown plants. No differences in growth or phenotype were observed between the WT and the transgenic plants (Fig. 5A). Under salinity stress, WT plants exhibited a severe under control conditions reduction in growth after 40 d, whereas the transgenic lines exhibited continued growth and tolerance up to maturity (Fig. 5B). A similar response was observed for plants exposed to sodicity (Fig. 5C). For both salinity and sodicity, at the end of the treatment period the WT plants appeared to be dead whilst the TPSP-overexpressing plants maintained a healthy phenotype. In contrast, under drought conditions both the WT and transgenic lines showed a reduction in growth after 25 d of stress (Fig. 5D). Interestingly, after 10 d of rewatering, the transgenic plants recovered quickly and unfolded their leaves, whereas the WT plants were not able to recover to the same extent even after 30 d of rewatering. These results indicated greater stress tolerance in the TPSP-overexpressing transgenic plants compared to the WT.

Fig. 5.

Morphological and physiology effects in wild-type (WT) and TPSP-overexpressing rice under salinity, sodicity, and drought stress, and after recovery from drought. (A) WT and TPSP-overexpressing plants (Ta-2 and X-93) at 2 months old were maintained under control conditions by irrigation with water, whilst other plants were subjected to (B) salinity stress (EC 10 dS m^–1), (C) sodicity stress (pH=9.9), or (D) drought stress (water withheld for 25 d, followed by recovery by rewatering). (E) Trehalose content, (F) chlorophyll content, (G) K⁺/Na⁺ ratio, and (H) relative water content (RWC) were measured after 25 d of treatment, and at 30 d of post-drought rewatering. Data shown means (±SD), n=9.

Under control conditions, the transgenic lines showed a 3-fold higher concentration of trehalose in comparison to the WT (Fig. 5E). This difference increased markedly under salinity (19-fold), sodicity (8-fold), and drought (25-fold). Under salinity stress, there was a 72% reduction in total chlorophyll content in the WT, but only a 45% reduction in the transgenic lines (Fig. 5F). Similar reductions in chlorophyll were observed in the WT (64%) and transgenic plants (43%) under sodic stress and under drought stress (81% and 64% for the WT and trangenics, respectively). After rewatering, the chlorophyll content of the transgenic plants was almost fully restored, whereas the WT plants reached only 63% of the control value. These results indicated that higher trehalose levels in rice contribute towards retaining chlorophyll under abiotic stresses.

Similar patterns to chlorophyll were seen in the K⁺/Na⁺ ratio in response to salinity, sodicity, and drought, with greater reductions being observed in the WT compared to the transgenic plants in each case (Fig. 5G). Again, the transgenic plants showed much better recovery than the WT upon rewatering after drought. Relative water content also showed a similar pattern, with the WT being more severely affected by the stresses than the transgenic plants (Fig. 5H). Again, recovery after drought was better in the transgenic plants when they were rewatered.

Transgenic lines with high trehalose contents maintain photosynthetic efficiency under abiotic stresses

Very similar patterns of repsonses in photosynthetic parameters were observed for salinity (Fig. 6A), sodicity (Fig. 6B), and drought (Fig. 6C). In each case, WT plants showed much greater reductions in NPR, g_s, TR, F_v/F_m, and ETR in response to stress that the transgenic lines. For example, NPR in WT plants was reduced by 63%, 57%, and 62% in response to salinity, sodicity, and drought, respectively, whilst the corresponding decreases in transgenic plants were only 27%, 25%, and 37%. In addition, the transgenic lines showed much better recovery than the WT in all parameters when plants were rewatered for 30 d following the end of the drought treatment (Fig. 6D). These results indicated that the transgenic plants retained higher photosynthetic efficiency under stress and were able to recover quickly upon rewatering after drought. This recovery was particularly evident in the two transgenic lines as they showed emergence of new leaves during the recovery period.

Fig. 6.

Evaluation of various photosynthetic parameters showing positive correlations between TPSP-overexpression and abiotic stress tolerance in rice. Wild-type (WT) and TPSP lines (Ta-2 and X-93) at 2 months old were subjected to (A) salinity, (B) sodicity, and (C) drought and measurements were taken after 25 d of stress. The droughted plants were then rewatered for 30 d (recovery) and further measurements were taken (D). The following photosynthetic parameters were determined: net photosynthetic rate (NPR), stomatal conductance, transpiration rate, F_v/F_m, and electron transport rate (ETR). Relative values are plotted in web diagrams, with the WT control set as 100%. Plants irrigated with water acted as the controls. The data represent the means of n=9 values.

Transgenic lines with high trehalose contents maintain yields under abiotic stresses

To determine the effects of overproduction of trehalose on crop yield, we examined panicles in WT and TPSP-overexpressing transgenic plants. Both the WT and transgenic lines showed similar panicle architecture and spikelet number per panicle under control conditions (Fig. 7A); however, under stress conditions, the transgenic lines had more spikelets per panicle and more filled grains per panicle. Salinity, sodicity, and drought all reduced the total biomass of both the WT and transgenic plants, but the reductions were much greater in the WT (Fig. 7B). Similar effects were also seen for panicle number per plant (Fig. 7C), and the critical yield parameters of filled grains per plant (Fig. 7D) and grain weight per plant (Fig. 7E). Importantly, the harvest index of the transgenic plants was almost double those of the WT plants under all of the stresses (Fig. 7F), a result of the decrease in panicle number per plant and number of filled grains per plant in the WT.

Fig. 7.

TPSP-overexpression reduces rice yield penalties under various abiotic stresses. (A) Images of panicles of the wild-type (WT) and the TPSP-transgenic lines (Ta-2 and X-93) under control, salinity, sodicity, and drought stress conditions. At physiological maturity, the following yield components were determined: (B) total biomass per plant, (C) panicle number per plant, (D) filled grains per plant, (E) total grain weight per plant, and (F) harvest index. Data are means (±SD), n=3. Significant differences for each genotype compared with control conditions were determined using Student’s t-test: * P<0.05. (G) Heat-map showing differential accumulation patterns of different metabolites in seeds of the WT and the TPSP-transgenic line X-93 as analysed using GC-MS. The scale includes the minimum, maximum, and mean values of the relative metabolite concentrations. All the identified metabolites from the seeds of the WT and the transgenic lines with their relative abundance are given in Supplementary Table S1.

The sugar metabolome of dry seeds of the WT and TPSP-plants were analysed using GC-MS (Fig. 7G). Significant differences in trehalose and glucose were found between the WT and the transgenic plants, but other carbohydrates remained the same, confirming the improvement of the nutritional quality of transgenic seeds compared to the WT (Supplementary Table S1).

After observing the effects of trehalose in enhancing tolerance to multiple stresses in rice under greenhouse conditions in pots, we then examined responses to salinity and sodicity using microplots/hydroponics that simulated field conditions. Better performance was again seen in the TPSP-transgenic plants compared to the WT, as scored in terms of various growth- and yield-related parameters. Correlation coefficients between these parameters are shown in Tables 1–3. Such inter-trait correlation coefficients are useful for understanding the association dynamics of morphological traits with each other and, in particular, with grain yield, as the latter may be able to form the basis for yield selection criteria. The data indicated that across all three screening environments (non-stress, high salinity, and high sodicity conditions) the initial growth vigour scores were negatively correlated with productive tillers and spikelet fertility (Tables 1–3). In all the three conditions, plant height was positively correlated with tiller number, panicle length, total grains per panicle, and grain yield, while total tiller number was positively correlated with productive tiller number and panicle length. Panicle length was found to be positively correlated with total grains per panicle and grain yield, and spikelet fertility was postively correlated with grain yield.

Table 1.

Correlation coefficients (r) between growth-related and yield-related parameters of the transgenic plants under non-stress conditions

	Vscore	PH	RL	Ttiller	Ptiller	PL	GY	FG	UFG	TG	SF	GW
Vscore		–0.4	–0.12	–0.424	–0.68**	0.17	0.063	0.31	–0.44	–0.31	–0.59*	0.04
PH			0.93***	0.76***	0.58*	0.88***	0.70**	0.41	0.49	0.75***	0.1	0.07
RL				0.78***	0.58*	0.93***	0.57*	0.37	0.66**	0.91***	–0.55	0.13
Ttiller					0.62**	0.58*	0.23	–0.08	0.77***	0.78***	–0.43	0.15
Ptiller						0.3	0.37	0.03	0.55*	0.61**	–0.34	0.15
PL							0.64**	0.56*	0.45	0.78***	0.22	0.15
GY								0.85***	–0.12	0.31	0.63**	–0.13
FG									–0.377	0.12	0.83***	0.17
UFG										0.87***	–0.75***	0.06
TG											–0.36	0.17
SF												0.06

For each parameter, the mean values of three transgenic plants were used. Vscore, vigour score (1–9 scale, 1=best, 9=poorest) (IRRI, 2014); PH, plant height (cm); RL, root length (cm); Ttiller, total tillers per plant; Ptiller, productive tillers per plant; PL, panicle length (cm); GY, grain yield per plant (mg); FG, filled grains per panicle; UFG, unfilled grains per panicle; TG, total grains per panicle; SF, spikelet fertility (%); GW, 1000 grain weight (g). The strength of correlations are indicated as follows: *P≤0.1, **P≤0.05, ***P≤0.01.

Table 3.

Correlation coefficients (r) between growth-related and yield-related parameters of the transgenic plants under high sodicity stress (pH ~9.9)

	Vscore	Gppan	PH	PL	Ttiller	PTiller	TG	GW	SF	GY
Vscore		0.1	–0.8***	–0.79***	–0.85***	–0.85***	–0.78**	–0.82***	–0.81***	–0.69**
Gppan			–0.11	–0.09	–0.21	–0.21	–0.09	–0.14	–0.12	0.04
PH				0.99***	0.97***	0.97***	0.99***	0.98***	0.97***	0.967***
PL					0.97***	0.97***	0.98***	0.99***	0.98***	0.97***
Ttiller						0.97***	0.96***	0.99***	0.98***	0.91***
PTiller							0.96***	0.99***	0.98***	0.91***
TG								0.97***	0.96***	0.97***
GW									0.99***	0.93***
SF										0.93***

For each parameter, the mean values of three transgenic plants were used. Vscore, vigour score (1–9 scale, 1=best, 9=poorest) (IRRI, 2014); PH, plant height (cm); RL, root length (cm); Ttiller, total tillers per plant; Ptiller, productive tillers per plant; PL, panicle length (cm); GY, grain yield per plant (mg); FG, filled grains per panicle; UFG, unfilled grains per panicle; TG, total grains per panicle; SF, spikelet fertility (%); GW, 1000 grain weight (g). The strength of correlations are indicated as follows: *P≤0.1, **P≤0.05, ***P≤0.01.

Interestingly, some traits were found to be significantly correlated with each other only under the two stress conditions. Under both stresses, plant height was positively correlated with spikelet fertility and total grain weight, whereas, the number of productive tillers was positively correlated with grain yield, spikelet fertility, and total grain weight. Similarly, total grain weight was positively correlated with grain yield, total grain number, and spikelet fertility under both salinity and sodicity stress. These correlations were not observed under non-stress conditions, thus indicating the existence of differential genetic and physiological effects between stressed and non-stressed environments. Overall, the correlation coefficients confirmed the favourable roles of plant height, panicle length, and spikelet fertility as candidates for indirect selection for improving grain yield as these three traits exhibited positive correlations with grain yield in saline, sodic, and non-stressed environments. The positive correlations for tiller number and total grain weight with grain yield are only of use for potential selection of the transgenics under the two stress conditions.

Trehalose accumulation in leaves modulates changes in other metabolites that regulate primary and secondary metabolism

To examine the effects of trehalose overproduction on the accumulation of other metabolites, we performed untargeted metabolite profiling in the leaves of the WT and transgenic plants subjected to salinity, sodicity, and drought. The metabolites are listed in Supplementary Table S2 and major effects are shown in Fig. 8. Under control conditions, most of the metabolites showed enhanced levels in the transgenic lines compared with the WT: there were 2-fold increases in fructose, glucose-6-phosphate, ribonate, fructose-6-phosphate, octadecanoic acid, 3PGA, sucrose, proline, and glucose, and 3-fold increases in trehalose, oleic acid, linoleic acid, vanillic acid, and threonine. In response to all three stresses, transgenic lines showed increases in accumulation of proline (3–7-fold), fructose (5–18-fold), and sucrose (2–6-fold) compared with the WT control (Fig. 8A–C). Glucose, pyruvate, and citrate also showed slight increases under stress conditions (Fig. 8D–F). Sorbitol showed high accumulation under drought (8-fold) and sodic (21-fold) stress (Fig. 8G). Even stronger responses for several amino acids were observed in the transgenic lines under stress (Supplementary Table S2). Among the intermediate metabolites, malate and glycine were found to be reduced by stress in the transgenic lines compared with the WT under similar conditions (Fig. 8H, I). Glucose-6-Phosphate (Glc6P), which is an allosteric activator of several other sugars, was also significantly increased in the transgenic lines under stress (Fig. 8J). The transgenic lines also showed significantly higher levels of threonine and valine under all three stresses (Fig. 8K, L). Aspartate showed higher accumulation under drought stress in the transgenic lines, but its levels were not significantly altered under salinity and sodicity (Fig. 8M). The transgenic lines had significantly increased levels of oleic and palmitic acid under all three stresses (Fig. 8N, O). Vanillic acid, a plant phenolic acid, was found to be increased significantly under all three stresses (Fig. 8P).

Fig. 8.

Relative abundances of metabolites in the leaves of wild-type (WT) and TPSP-overexpressing transgenic rice subjected to salinity (S), sodicity (Sd), or drought (D) as determined using GC-MS. Abundances are expressed relative to the WT under control conditions (set as 1). (A) Proline, (B) fructose, (C) sucrose, (D) glucose, (E) pyruvate, (F) citrate, (G) sorbitol, (H) malate, (I) glycine, (J) glucose-6-phosphate, (K) threonine, (L) valine, (M) aspartate, (N) oleic acid, (O) palmitic acid, and (P) vanillic acid. Significant differences relative to the WT under control conditions were determined using Student’s t-test: *P<0.05, **P<0.01, ***P<0.001. All the identified metabolites from control, salinity, sodicity, and drought stress conditions together with their relative abundances are given in Supplementary Table S2.

Discussion

Overproduction of trehalose in various systems including plants has been reported to be beneficial for improving tolerance against salinity and drought. However, its role in tolerance to sodicity has not been reported. Given the increasing problem of soil sodicity worldwide, our first major objective was to raise transgenic rice plants with overproduction of trehalose within the genetic background of a popular high-yielding cultivar, IR64, and to test their suitability for sodic soils, together with salinity- and drought-affected soil. Second, in view of ongoing debates about the presence of selection and reporter marker genes, in order to ensure better consumer acceptance of the transgenic plants, we have for the first time developed marker-free transgenic rice using glucose as the selection agent, which is non-toxic and environment-friendly. Third, most reports so far have tested transgenic plants in earthernware pots under laboratory or greenhouse conditions. We advanced our study of developing trehalose-overproducing transgenic lines by testing them on a relatively large scale using microplots/hydroponics within a transgenic greenhouse. We also performed metabolic profiling of the IR64 wild-type (WT) and transgenic lines under different abiotic stresses and compared them to control conditions in order to determine how the accumulation of trehalose regulates various physiological parameters. In addition, for the first time, we also performed metabolic profiling of the WT and transgenic seeds to check whether the accumulation of trehalose caused any undesired changes in other metabolites that might affect the grain quality. Interestingly, the transgenic seeds showed physiological equivalence to the untransformed WT seeds, thus making them more suitable for public acceptance. Hence, this study reports on the production of transgenic IR64 rice lines that have enhanced tolerance to sodicity, salinity, and drought whilst being marker-free, nutritionally equivalent to WT seeds, and exhibiting lower yield penalties than the WT under stress conditions. Our results demonstrate that ‘trehalose technology’ can be useful for developing high-yielding plants that are tolerant to multiple stresses.

In comparison to the transgenic lines containing the fusion gene coding for trehalose-6-phosphate synthase/phosphatase (TPSP), WT plants showed early leaf yellowing under the three different stress conditions (Fig. 5A). The higher retention of chlorophyll and reduction in membrane damage in the TPSP-transgenic lines (Fig. 2) may have been due to the physicochemical properties of trehalose, which stabilizes lipids and reduces oxidation of lipid membranes during desiccation. Interestingly, the trehalose-overproducing lines showed higher accumulation of lipids under stress conditions as compared to the WT plants (Fig. 8N–P) .Trehalose has previously been found to induce the expression of genes involved in lipid-related pathways (Oszvald et al., 2018). It has also been shown that trehalose combines with lipids to form complexes that act as glycolipid surfactants (Franzetti et al., 2010). These complexes have been reported to serve as immuno-modulators and as cell wall components in microorganisms (Yatsyshyn et al., 2017). Salinity stress causes oxidative stress via generation of reactive oxygen species (ROS), which leads to cell senescence (Jajic et al., 2015). It has previously been reported that pre-treatment of rice and wheat with trehalose prevents formation of superoxide radicals, ROS, and MDA under stress conditions, which suggests that trehalose can scavenge ROS and protect against membrane damage (Theerakulpisut and Gunnula, 2012; Lunn et al., 2014). This has been reported as a mechanism for salt tolerance (Shabala and Cuin, 2008).

Seed germination is the most sensitive stage to saline/sodic stress and several complex mechanisms are known to confer protection against high levels of Na⁺ ions and high pH in higher plants (Wang et al., 2016; Guan et al., 2017). The sequestration of Na⁺ has been considered as one of the key components differentiating between sensitive and tolerant genotypes (Wu et al., 2015). Moreover, the degree of damage increases with stress intensity, which is consistent even under drought stress (Guan et al., 2017). Our studies showed that the transgenic seeds and seedlings were less affected by salinity, sodicity, and drought stress in comparison to the WT, as evidenced by their higher germination rates and better subsequent growth under these stress conditions (Fig. 3). Higher accumulation of trehalose in the transgenic seeds (Supplementary Table S1) correlated well with a previous report that showed that AtTPS1-knockouts in Arabidopsis displayed an embryo-lethal phenotype (Eastmond et al., 2002), suggesting that trehalose plays a key role as a signaling molecule during seed development.

The trehalose-overproducing plants had lower cellular accumulation of Na⁺ in the roots than the WT, as indicated by fluorescence microscopy (Fig. 4). Exogenous trehalose treatment has been found to significantly reduce the accumulation of Na⁺ in the leaves (Garcia et al., 1997), which indicates that it might play a direct or indirect role in determining ion selectivity by modulating cellular exclusion of Na⁺. K⁺ is considered as a beneficial ion that confers an advantage to rice under saline and sodic stress (Shabala and Cuin, 2008). We found that the transgenic lines had higher K⁺/Na⁺ ratios than the WT under both salinity and sodicity stress (Fig. 5).

Trehalose biosynthesis occurs by the conversion of glucose to trehalose-6-phosphate (T6P) by the TPS enzyme, and this is then converted to trehalose by TPP. Accumulation of T6P may lead to aberrations in the plant phenotype if the conversion to trehalose does not take place. Our use of the fusion gene (TPSP) ruled out accumulation of T6P (Karim et al., 2007) and hence no aberrations were seen in our study. Sodicity and salinity are reported to cause severe reductions in growth and productivity of glycophytes due to the toxic effects of Na⁺ or Cl⁻ ions, the restricted uptake of K⁺, Ca²⁺ and NO₃⁻, and osmotic stress (Flowers and Colmer, 2008). The TPSP-transgenic plants exhibited higher tolerance to salinity than WT plants during the vegetative stage, as indicated by enhancement of the chlorophyll content, K⁺/Na⁺ ratio, and relative water content (RWC; Fig. 5F–H). The same was also true in relation to sodic and drought stress. The trehalose concentration in the leaves increased significantly in the transgenic lines under the different abiotic stresses (Fig. 5E), suggesting that its accumulation confers tolerance to multiple abiotic stresses. It has previously been reported that ABA is released from roots and translocated to leaves under drought and salinity stress (Cornish and Radin, 1990), and thus ABA-signaling plays a key role in response to these stresses (Vishwakarma et al., 2017). Han et al. (2019) demonstrated in wheat that under sodic/alkaline stress, the PP2C protein (a negative regulator of ABA signaling) is down-regulated and subsequently modulates the expression of downstream ABA-responsive genes. As the transgene used for transformation in our study was regulated by an ABA-inducible promoter, the differences in trehalose levels observed under the different abiotic stresses may be attributable to the different levels of ABA that were released under these conditions.

In addition to its role as an osmoprotectant, trehalose and its intermediate T6P have also been reported to enhance plant stress tolerance via sugar signaling by allocating and metabolizing carbohydrates (Kolbe et al., 2005; Iturriaga et al., 2009). Interestingly, higher levels of soluble carbohydrates (including trehalose) under stress conditions indicate that trehalose acts as a positive regulator of genes associated with sugar sensing and carbon metabolism, as shown in previous studies (Oszvald et al., 2018). The diverse and complex network of sugar signaling has been identified as a crucial component in responses to abiotic stresses in plants (Hellmann and Smeekens, 2014). We found that our transgenic lines were able to maintain higher RWC under abiotic stresses than WT plants (Fig. 5H). Decreases in RWC under stress conditions are associated with reduced osmotic potential (Qin et al., 2010; Ali and Ashraf, 2011). Reduced RWC drastically decreases a number of physiological processes, including photosynthesis, and stomatal conductance (Alam et al., 2014; Mostofa et al., 2014). Higher accumulation of soluble sugars may function as an osmoprotectant to prevent water loss from plant cells under osmotic stresses (Delorge et al., 2014). Previous studies have shown that supplementation of trehalose in saline medium protects Catharanthus roseus from the inhibitory effects of salt on growth, RWC, and photosynthesis (Chang et al., 2014; Oszvald et al., 2018).

Protection of the photosynthetic machinery significantly contributes towards the ability of a plant to withstand stress. Photosynthesis and the electron transport system are reduced under osmotic stress due to the inhibition of enzymes related to the carbon reduction cycle, glycolysis, and ATP synthesis (Joshi et al., 2016a), and this eventually also inhibits respiration. Photosynthetic and chlorophyll fluorescence parameters in our study indicated that trehalose accumulation mitigated the adverse effects of salinity, sodicity, and drought stress on the photosynthetic components of plants (Fig. 6).

Abiotic stresses can severely limit agricultural productivity in crop species, particularly at the reproductive stage. Trehalose levels significantly affect the regulation of carbon allocation and utilization in plants, resulting in yield improvements under environmental stresses (Paul et al., 2017). In field-simulated micro-plots, overexpression of TPSP resulted in positive correlations between various yield parameters, and hence improved total grain yield under control conditions (Table 1). Under salinity stress (EC 10 dS m⁻¹), root length showed a negative correlation with other yield parameters, indicating that severe damage to the roots eventually affected grain yield (Table 2). We observed root damage through PI staining of roots (Fig. 4). Similarly, under sodic stress (pH 9.9), grains per panicle showed a negative correlation with other yield parameters, reflecting a drastic decline in dry matter partitioning that resulted in reduced grain yield (Table 3). Previous reports suggest that trehalose acts as a positive regulator of stress tolerance in plants (Redillas et al., 2012; Smeekens, 2015; Farooq et al., 2017); however, its explicit function is still unclear, because of its multifaceted contributions in responses to environmental conditions.

Table 2.

Correlation coefficients (r) between growth-related and yield-related parameters of the transgenic plants under high saline conditions (ECiw ~10.0 dS m^–1)

	Vscore	PH	RL	Ttillers	Ptillers	PL	GY	FG	UFG	TG	SF	GW
Vscore		–0.82***	–0.7**	–0.86***	–0.86***	–0.66**	–0.56*	0.44	–0.43	–0.32	–0.56*	–0.65**
PH			0.38	0.89***	0.92***	0.79***	0.75***	0.3	0.75***	0.68**	0.64**	0.71**
RL				0.53**	0.42	0.12	–0.10	–0.64**	–0.09	–0.24	–0.09	0.29
Ttillers					0.88***	0.66**	0.68**	0.11	0.6**	0.51	0.49	0.63**
Ptillers						0.76***	0.76***	0.33	0.71**	0.65**	0.76***	0.83***
PL							0.77***	0.57**	0.76***	0.76***	0.66**	0.61**
GY								0.63**	0.67**	0.7**	0.77***	0.54*
FG									0.71**	0.83***	0.63**	0.34
UFG										0.98***	0.63**	0.77***
TG											0.67**	0.69**
SF												0.65**

For each parameter, the mean values of three transgenic plants were used. Vscore, vigour score (1–9 scale, 1=best, 9=poorest) (IRRI, 2014); PH, plant height (cm); RL, root length (cm); Ttiller, total tillers per plant; Ptiller, productive tillers per plant; PL, panicle length (cm); GY, grain yield per plant (mg); FG, filled grains per panicle; UFG, unfilled grains per panicle; TG, total grains per panicle; SF, spikelet fertility (%); GW, 1000 grain weight (g). The strength of correlations are indicated as follows: *P≤0.1, **P≤0.05, ***P≤0.01.

A higher accumulation of trehalose resulted in increased productivity and biomass under the various abiotic stresses (Fig. 7). TPS is required for both sucrose regulation during deposition of storage reserves and in metabolism during late-embryo development, implying that it has a role in sugar signaling (Eastmond et al., 2002; Martínez-Barajas et al., 2011). Previous studies on rice and maize have also shown that effective partitioning and enhanced growth can be achieved through spatial and temporal regulation of enzymes in sink tissues that hydrolyse T6P (Nuccio et al., 2015; Smeekens, 2017). A mutation in maize RAMOSA3 that results in a defective TPP negatively affects plant development, indicating a regulatory role of T6P in inflorescence architecture (Satoh-Nagasawa et al., 2006). OsTPP1 expression in ear spikelets of maize produces a higher amount of sucrose in comparison to WT plants, suggesting an enhanced sink capacity of the reproductive tissues (Nuccio et al., 2015). These authors also found that expression of TPP under the control of a floral promoter (OsMads6-Tpp1) in developing maize ears resulted in improved kernel set and harvest index, both under control and drought stress.

It has been demonstrated that sucrose is required for spikelet development and that it acts as a signal for preventing starvation-induced abortion (Nuccio et al., 2015). Enhanced production of soluble sugars may also lead to higher starch accumulation in leaves as a temporary carbon reserve and as a primary component of dry matter accumulation (Abdelgawad et al., 2014). The T6P:sucrose ratio has been reported to be important in maintaining sucrose levels for appropriate cell development and plant growth (Paul et al., 2017). Higher glucose levels in the transgenic lines under salinity, sodicity and drought indicate increased sucrose catabolism. Figueroa et al. (2016) demonstrated that total photoassimilates were partitioned into sucrose as well as organic acids. Therefore, the movement of photoassimilates towards sucrose synthesis and the respiratory pathway is the main reason for higher trehalose accumulation. Previous studies have also correlated the molecular mechanisms linking high trehalose levels to post-translational activation of PEPC with a rapid increase in closely related amino acids (e.g., Thr, Ser, and Val) at the expense of sucrose synthesis (Paul et al., 1978; Larsen et al., 1981).

The results from our metabolite analysis were consistent with trehalose acting as a signaling molecule that regulates sucrose-induced changes under salinity, sodicity, and drought (Lunn et al., 2006; Islam et al., 2019). Vicente et al. (2018) showed a hyperbolic correlation between expression of genes in the gluconeogenesis pathway and T6P levels in yeast, which proves that trehalose is a regulatory switch between the rate at which glucose enters the cell and its consumption. Similarly, it has been proposed that T6P and fructose-6-phosphate both act as ‘metabolic flux sensing’ metabolites for rapid changes in carbon assimilation, particularly glucose and sucrose (Huberts et al., 2012; Schluepmann et al., 2012; Lawlor and Paul, 2014). In addition, it has also been reported that changes in the level of T6P can alter TPS activity, which might influence carbon flux through glycolysis by regulating hexose kinase (Romero et al., 1997). Furthermore, it has been demonstrated that T6P inhibits the activity of sucrose non-fermenting related kinase-1 (SnRK1) under stress conditions, which modulates carbohydrate metabolism in plants (Zhang et al., 2009; O’Hara et al., 2013). Higher accumulation of trehalose in the transgenic lines under salinity, sodicity, and drought (Fig. 5) correlated with enhanced accumulation of glucose, pyruvate, and citrate (Fig. 8). Citrate is essential for re-assimilation of ammonium released by photorespiration. Trehalose activates flux towards malate and fumarate, which result in altered glycine decarboxylation that leads to higher glycine accumulation in mitochondria and less in the cytosol. In our previous studies, we demonstrated that salinity-tolerant genotypes accumulate higher levels of proline than sensitive genotypes and concluded that the accumulation of proline is essential for primary metabolism and the prevention of osmotic stress-induced oxidative stress injury (Joshi et al., 2016a, 2016b). The higher levels of proline further reduce free radicals, as shown by previous studies (Obata and Fernie, 2012). Taking together our current results and those of previous studies, it is clear that trehalose accumulation provides a spatial and temporal regulation of carbon metabolism that improves stress tolerance without affecting grain yield and quality in transgenic rice. Our study indicates that accumulation of trehalose improves tolerance against harsh environmental conditions by regulating metabolite allocation at the subcellular levels. Further, our results provide new information on the development of marker-free transgenic lines in a popular rice variety, providing reduced yield penalties under multiple stresses as well as maintaining yield quality. The transgenic lines provide promising material for future development of rice varieties that are tolerant to multiple abiotic stresses.

Supplementary data

Supplementary data are available at JXB online.

Table S1. Metabolites of rice seeds with differential accumulation in the wild-type and transgenic line X-93 under control conditions.

Table S2. Metabolites of rice leaf tissues showing differential accumulation in the wild-type and transgenic line X-93 under salinity, sodicity, and drought conditions.

Table S3. Sequences of different primers used for cloning and screening of putative transgenic plants containing the TPSP fusion construct.

Author contributions

RJ, KKS, AKS, and KA carried out the transgenic analysis; RKG, SLK and PP carried out the microplot/hydroponics experiments; RJ and AKS drafted the manuscript; SLS-P, SKS, and AP conceived and designed the study and finalized the manuscript; all the authors read and approved the final manuscript.