GmNF‐YC9‐mediated CCAAT‐box transcriptional complex can activate the expression of GmSQE1

To elucidate the molecular mechanism underlying GmNF‐YC9‐mediated stress tolerance in soybean, we conducted RNA‐seq assays on GmNF‐YC9‐OE and WT plants. The analysis revealed upregulation of numerous stress‐related genes in GmNF‐YC9‐OE plants. Gene Ontology (GO) term and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis identified significant enrichment in genes associated with sterol metabolism, including biosynthetic processes, transporters, and responses to steroid hormones (Figure 3a,b, Figure S3a,b). Among these genes, squalene monooxygenase (GmSQE1), which encodes the enzyme responsible for the rate‐limiting step in sterol synthesis, displayed significant induction in GmNF‐YC9‐OE plants under stress conditions (Figure 3b). This suggested that the GmNF‐YC9 transcription factor regulated the sterol synthesis pathway by modulating GmSQE1 activity. Our qRT‐PCR results further confirmed elevated transcript levels of GmSQE1 in GmNF‐YC9‐OE plants (Figure S3c), which were significantly induced by drought and salt stress treatments (Figure S3d). Notably, GmSQE1 exhibited higher expression in GmNF‐YA2 and GmNF‐YB24 transgenic hairy roots (Figure S4a,b), suggesting that it could be a potential target gene of GmNF‐YC9‐mediated CCAAT‐box transcriptional complex.

Open in a separate window

Figure 3

Downstream target gene and its regulation pathway analysis of GmNF‐YC9 transcription factor. (a) The Venn diagram shows gene co‐expression analysis of WT and GmNF‐YC9‐OE soybean plants under drought conditions. (b) GO enrichment analysis of differentially expressed genes (DEGs) in GmNF‐YC9‐OE soybean plants. Red asterisk indicates squalene monooxygenase (GmSQE1). (c, d) Interaction analysis between the NF‐Y transcription factor and GmSQE1 promoter by LUC activity assay. (e) The module of SQE‐mediated catalysation to squalene. (f) Phenotype analysis of 7‐day‐old soybean seedling plants after 10days of SQ (600μg/L), 2,3‐OSQ (300μg/L), mannitol (200mM), and terbinafine (100μM) treatments, respectively. (g, h) Total root length (g) and fresh weight (h) of soybean seedlings under different treatments. (i) The inhibition of terbinafine on SQE‐mediated catalysation to SQ. (j) Phenotype of 7‐day‐old soybean seedlings after 10days of 100mM NaCl, 100mM mannitol, and 25μM terbinafine treatments. (k–n) Fresh weight (k, l) and total root length (m, n) of 7‐day‐old soybean seedlings after 10days of 100mM NaCl, 100mM mannitol, and 25μM terbinafine treatment conditions. (o) Squalene content analysis of soybean seedling roots under different treatment conditions (25μM terbinafine and 100mM mannitol). (p, q) Cycloartenol (p) and beta‐amyrin (q) content analysis of soybean seedling roots under 100mM mannitol treatment conditions. AM, beta‐amyrin; CY, cycloartenol.

To validate the regulatory role of GmNF‐YC9 on GmSQE1 expression, luciferase (LUC) activity analysis assays were conducted. These assays revealed a substantial increase in GmSQE1 promoter‐induced LUC activity in the presence of GmNF‐YA2, GmNF‐YB24, or GmNF‐YC9 proteins (Figure 3c,d). Electrophoretic mobility shift assays (EMSA) further demonstrated the binding of the GmNF‐YA2/GmNF‐YB24/GmNF‐YC9 transcription complex to the CCAAT‐box of the GmSQE1 gene promoter (Figure S4c). These findings conclusively established that GmSQE1 was a downstream target gene of a trimeric nuclear transcriptional activation complex. Moreover, enzyme‐linked immunosorbent assay (ELISA) was employed to analyse SQE enzyme activity and its catalytic product, 2,3‐oxidosqualene (2,3‐OSQ), in both GmNF‐YC9‐OE and WT plants. The results demonstrated a significant increase in SQE enzyme and 2,3‐OSQ content in salt and drought‐treated GmNF‐YC9‐OE plants compared to WT plants (Figure S4d–g), indicating that the regulation of GmSQE1 activity by GmNF‐YC9 can affect the sterol synthesis pathway.

Squalene and squalene‐2, 3‐epoxide, two key components in SQE‐mediated catalytic pathway, can affect plant stress tolerance

To elucidate the contribution of enhanced GmSQE1 expression to plant stress tolerance, we investigated the correlation between stress tolerance and two key components of the SQE‐mediated catalytic pathway: squalene (SQ) and 2,3‐oxidosqualene (2,3‐OSQ) (Figure 3e). 7‐day‐old soybean seedlings were subjected to different treatments. After a treatment period of 10days, we observed the phenotypes of the soybean seedlings under different treatment conditions. The results revealed that exogenous application of SQ (600μg/L) and 2,3‐OSQ (300μg/L) promoted soybean growth under normal conditions (Figure 3f–h). Moreover, these compounds significantly enhanced soybean stress tolerance under 200mM mannitol conditions (Figure 3f–h). In contrast, terbinafine, a specific inhibitor of SQE widely used to inhibit sterol biosynthesis pathways (Figure 3i), conditionally impeded the development of normal root structures in soybean seedlings exposed to mannitol‐induced stress (Figure 3f). Notably, this inhibitory effect was alleviated by providing exogenous concentrations of 2,3‐OSQ at various levels, which partially prevented the occurrence of short‐thick root phenotype (Figure 3f–h, Figure S5a–d). These results underscored the crucial role played by SQE‐mediated plant sterol synthesis pathways in both growth and stress tolerance.

Furthermore, we employed the SQE inhibitor terbinafine to investigate the significance of squalene production in plant stress tolerance. Similarly, 7‐day‐old soybean seedlings were exposed to various stress treatments. After a duration of 10days, the phenotypes of soybean seedlings were observed under different treatments. Our findings clearly demonstrated that the application of terbinafine compromised soybean's tolerance to stress (Figure 3j–n). Additionally, we performed liquid chromatography‐mass spectrometry (LC–MS) analysis to examine changes in key metabolites from the SQE‐mediated plant sterol synthesis pathway in soybean roots under 100mM mannitol‐induced stress. These analyses revealed a significant increase in squalene content in soybean roots upon exposure to 100mM mannitol (Figure 3o, Figure S5e). Interestingly, treatment with terbinafine led to a substantial accumulation of squalene in soybean roots (Figure 3o, Figure S5e). Moreover, there was a notable elevation in the levels of cycloartenol and β‐amyrin in soybean roots‐ two crucial products derived from 2,3‐OSQ – after treatment with 100mM mannitol (Figure 1p,q, Figure S5f), further highlighting their role in plant stress tolerance. Collectively, these results underscored the indispensable function of SQE in both plant growth and stress tolerance.

Overexpression of the NF‐Y target gene GmSQE1 in soybean improves drought and salt stress tolerance

To assess the potential role of squalene‐derived compounds in enhancing plant stress tolerance, we generated three distinct GmSQE1 overexpression lines (GmSQE1‐OE1, GmSQE1‐OE3, and GmSQE1‐OE6) (Figure S6a–d). The impact of GmSQE1 overexpression on drought and salt stress tolerance paralleled the observations made for GmNF‐YC9‐OE plants during the seedling stage (Figure S6e–j). Under stress conditions, GmSQE1‐OE plants exhibited enhanced characteristics such as increased biomass and longer hypocotyl lengths compared to their WT counterparts (Figure S6e–j). These findings were further supported by pot experiments, highlighting the robust adaptation of GmSQE1‐OE plants to drought and salt stresses (Figure 4a). Moreover, elevated proline and chlorophyll contents along with reduced malondialdehyde (MDA) levels were observed in GmSQE1‐OE lines compared to WT plants (Figure 4b), providing evidence for the enhanced stress resilience conferred by overexpressing GmSQE1.

Open in a separate window

Figure 4

Stress tolerance identification analysis of GmSQE1 overexpressing soybean plants. (a) Phenotype of WT and GmSQE1‐OE seedling soybean plants under drought and 200mM NaCl treatment conditions. (b) Proline, MDA, and chlorophyll content analysis of WT and GmSQE1‐OE soybean plants under different stress conditions. (c) Phenotype analysis of GmSQE1‐OE soybean plants at adult stage under drought conditions in the field in the year 2020. (d, e) Phenotype (d) and relative content (e) of different seed types in GmSQE1‐OE soybean plants under drought conditions in the field in the year 2020. (f) Phenotype analysis of GmSQE1‐OE soybean plants at adult stage under drought conditions in the field in the year 2021. (g) Relative contents of different seed types in GmSQE1‐OE soybean plants under drought conditions in the field in the year 2021. (h–l) Agronomic trait analysis of WT and GmSQE1‐OE soybean plants under drought conditions in the field in the year 2021. (m) Oxidative stress tolerance analysis of GmSQE1 transgenic soybean plants under 50μM MV treatment. (n) Fresh weight analysis of GmSQE1 transgenic soybean plants under 50μM MV treatment. (o) Oxidative stress tolerance analysis of GmNF‐YC9 transgenic soybean plants under 50μM MV treatment. (p) Fresh weight analysis of GmNF‐YC9 transgenic soybean plants under 50μM MV treatment. (q, r) 200mM H₂O₂‐induced cell death analysis of GmSQE1 (q) and GmNF‐YC9 (r) transgenic soybean plant roots by Pi staining.

Furthermore, the stress tolerance effects of GmSQE1 overexpression were also observed during the flowering stage, as evidenced by increased biomass in GmSQE1‐OE and GmNF‐YC9‐OE plants compared to WT plants under drought and salt stress conditions (Figure S6k–p). These results confirmed that the high expression of GmSQE1 and GmNF‐YC9 in soybean provided benefits for enhancing plant stress tolerance. To validate the durability of these effects, we conducted drought tolerance tests in field conditions over two seasons. Under normal irrigation conditions, no discernible differences in agronomic performance were observed between GmSQE1‐OE, GmNF‐YC9‐OE, and WT plants (Figure S7). However, under drought stress, both GmSQE1‐OE and GmNF‐YC9‐OE plants exhibited significantly improved agronomic performance compared to WT plants. This was evident through the increases in stem base circumference, grain count per plant, and grain weight per plant (Figure 4c–l, Figure S8a–t). Furthermore, an analysis of seed traits indicated that under drought conditions, GmSQE1‐OE and GmNF‐YC9‐OE plants produced more high‐quality seeds and fewer low‐quality seeds than WT plants (Figure 4d–g, Figure S8l–t, Table S1). Importantly, GmSQE1‐OE and GmNF‐YC9‐OE plants maintained these drought tolerance characteristics and improved agronomic traits at the adult stage under drought stress conditions (Figure 4c–l, Figure S8).

GmSQE1 and GmNF‐YC9 can enhance tolerance to abiotic stress‐induced oxidation

A recent study has demonstrated that the depletion of SQE in cholesterol auxotrophic lymphomas leads to an accumulation of squalene, thereby preventing oxidative cell death (Garcia‐Bermudez et al., ²⁰¹⁹). This finding suggests that the involvement of SQE in mediating cellular antioxidant process. However, our results and previous studies indicate that abiotic stress induces increased levels of reactive oxygen species (ROS) in plants (Figure S9a,b). Consequently, we hypothesised that GmSQE1 may enhance plant stress tolerance by improving soybean's resilience to oxidative stress. To test this hypothesis, GmSQE1‐OE, GmNF‐YC9‐OE, and WT plants were subjected to methyl viologen (MV) and hydrogen peroxide (H₂O₂). After 7days of treatment, MV induced wilting in WT plants, whereas GmSQE1‐OE and GmNF‐YC9‐OE plants displayed varying degrees of resistance to this effect (Figure 4m–p). These transgenic plants exhibited a higher biomass compared to WT plants under MV treatment conditions (Figure 4n,p). Additionally, after a 1.5‐h exposure to 200mM H₂O₂, cell damage was evaluated using propidium iodide (PI) staining. While WT plant root cells exhibited PI red fluorescence in their cytoplasm, only a few GmSQE1‐OE and GmNF‐YC9‐OE plant root cells displayed such fluorescence (Figure 4q,r). This result indicated that GmSQE1‐OE and GmNF‐YC9‐OE plants experienced less cell damage compared to WT plants under 200mM H₂O₂ treatment conditions (Figure 4q,r). Moreover, high‐temperature (43°C) stress, known to induce ROS production (Pospíšil, ²⁰¹⁶), was applied to four‐week‐old seedlings. Following a recovery period after 12h of exposure, both GmSQE1‐OE and GmNF‐YC9‐OE plants survived at higher rates and accumulated more shoot biomass than the control group, which further highlighted their enhanced tolerance towards oxidative stress (Figure S9c–h). Interestingly, the application of terbinafine compromised soybean's oxidative stress tolerance (Figure S10a–c). Collectively, these results suggested that overexpressing both GmSQE1 and GmNF‐YC9 significantly improved plant responses to oxidative stress across various contexts.

Analysis of stress tolerance on transgenic soybean plants

For 150mM NaCl and 200mM mannitol treatments, T₃ homozygous GmNF‐YC9 and GmSQE1 transgenic soybean seeds were surface‐sterilised with chlorine for 6h, following the method described by Chen et al. (²⁰¹⁸) and Cheng et al. (²⁰²¹). The sterilised seeds were then transferred to a B5 vitamin solid medium for growth under normal conditions. After 7days of growth, the seedlings were transferred into a B5 vitamin solid medium supplemented with 150mM NaCl and 200mM mannitol for stress tolerance analysis. After 7days of treatment, the seedlings were analysed.

For stress tolerance analysis during the soybean seedling stage, T₃ homozygous GmNF‐YC9 and GmSQE1 transgenic soybean seeds were sown in flowerpots filled with equally weighted nutrient soil for growth under normal conditions. Four‐week‐old soybean seedlings were treated with either water controlling or 250mM NaCl. After 14days of drought or 7days of salt treatment, the stress tolerance phenotypes of transgenic soybean seedlings were analysed. The method was expounded upon by Wang et al. (²⁰²⁰) and Yu et al. (²⁰²¹).

For drought tolerance analysis at soybean adult stage, we utilised rain‐proof installations to systematically investigate the effects of water stress on transgenic soybean plants. Every year at the end of May (Beijing, China), we planted transgenic soybean seeds in the field for growth under natural conditions. Around the middle of August, soybean plants began to enter periods of pod filling. We thus stopped irrigation to control water at the beginning of August and utilised rain‐proof installations to prevent weather from affecting drought treatment (Figure S15). After 1 month of water controlling treatment, the soil water content was measured using the method described by Yu et al. (²⁰²¹) (Table S3), and the phenotype under drought conditions was recorded. In the middle of October, soybean plants began to be harvested, and agronomic traits were recorded for analysis.

Bimolecular fluorescence complementation (BiFC) assay

The ORFs of bait protein and candidate protein were cloned into the pUC‐SPYNE and pUC‐SPYCE vectors, respectively. Subsequently, the recombinant vectors were co‐transformed into Arabidopsis protoplasts through PEG‐mediated transfection, followed by incubation in the dark at 25°C. The detailed method was described by Yoo et al. (²⁰⁰⁷). YFP fluorescence of the protoplasts was observed using a confocal laser scanning microscope (LSM 700; Zeiss, Oberkochen, Germany).

Luciferase complementation imaging (LCI) assay

The bait protein and candidate protein genes were individually inserted into pCAMBIA1300‐nLUC and pCAMBIA1300‐cLUC vectors. Subsequently, Agrobacterium GV3101 harbouring the nLUC and cLUC derivative binary recombinant plasmids were separately introduced via co‐injection into leaves of 4‐week‐old Nicotiana benthamiana. The firefly luciferase complementation imaging (LCI) transient expression assay was performed following the protocol established by Fujikawa and Kato (²⁰⁰⁷).

In vitro GST pull‐down assay

The open reading frames (ORFs) of bait protein and candidate protein genes were cloned into the pCold™ TF vector (TaKaRa, Bio) and pGEX‐4T‐1 vector (GE Healthcare, Chicago, IL), respectively. Subsequently, the recombinant plasmids were introduced into the E. coli strain transetta (TransGen, Biotech). Expressions of the corresponding proteins were induced overnight using 0.5mM isopropyl β‐D‐thiogalactoside (IPTG) (Inalco SPA, San Luis Obispo, CA) at 16°C, and then a pull‐down assay was performed. The experimental details were described by Yu et al. (²⁰²¹).

In vitro EMSA binding assay

The genes encoding three subunit proteins of the NF‐Y transcription factor were cloned into expression vectors harbouring the tag proteins (His, GST, and MBP), respectively, and subsequently transferred into competent cells of the E. coli strain Rosetta. These recombinant proteins were employed for an EMSA using a biotin end‐labelled duplex DNA probe. The method was previously described by Yu et al. (²⁰²¹).

Effect of fucosterol and soyasaponin II on crop growth under different treatments

Three‐week‐old of soybean, wheat, foxtail millet, and maize seedlings grown under normal conditions were subjected to controlled water treatment. During the treatment, the seedlings were sprayed with aqueous solutions of mock (40mL distilled water containing 200μL of 40mM fatty alcohol polyoxyethylene ether (APE), which was a leaf surfactant), fucosterol (20μL of fucosterol at a concentration of 600mg/L was added into 40mL distilled water containing 200μL of 40mM APE), and soyasaponin II (20μL of soyasaponin II at a concentration of 600mg/L was added into 40mL distilled water containing 200μL of 40mM APE). The sprays were applied four times each week, with at least 1 day between each spraying. The changes in crop phenotypes were observed and recorded during controlled water treatment. The leaf samples of crops were collected at the onset of phenotypic variations in order to quantify physiological and biochemical alterations.

Five‐week‐old of soybean, wheat, foxtail millet, and maize seedlings cultivated under normal conditions were subjected to osmotic stress treatment by irrigating them with 200mM mannitol aqueous solutions. During osmotic stress treatment, the seedlings were sprayed with mock (mock, 40mL distilled water containing 200μL of 40mM APE), fucosterol (fucosterol aqueous solutions, 20μL fucosterol (600mg/L) into 40mL distilled water containing 200μL of 40mM APE), and soyasaponin II aqueous solutions (soyasaponin II aqueous solutions, 20μL soyasaponin II (600mg/L) into 40mL distilled water containing 200μL of 40mM APE). The sprays were applied four times each week, with at least 1 day between each spraying. The phenotype changes of the crops were observed and recorded during controlled water treatment. Leaf samples of the crops were taken when differences in the leave phenotype began to appear in order to measure physiological and biochemical changes.

Metabolites extraction, UHPLC–MS–MS analysis, and data processing

The samples of stress induced WT and transgenic soybean seedlings (GmSQE1‐OE6 and GmNF‐YC9‐OE8 plants) were freeze‐dried, and then crushed with a mixer mill. A 50mg aliquot of each individual samples were precisely weighed and transferred to Eppendorf tubes, which were used for metabolites extraction and UHPLC–MS analysis (Allwegene Technology Co., Ltd, Beijing, China). The methods of plant metabolites extraction were described in detail by Vos et al. (²⁰⁰⁷).

The UHPLC separation was performed using an EXIONLC System (Sciex). The mobile phase A was 0.1% formic acid in water, while mobile phase B was acetonitrile. The column temperature was set at 40°C, and the auto‐sampler temperature was set at 4°C with an injection volume was 2μL. A Sciex QTrap 6500+ (Sciex Technologies) was applied for assay development. Typical ion source parameters were as follows: IonSpray Voltage, +5500/−4500V; Curtain Gas, 35psi; Temperature, 400 °C; Ion Source Gas, 1:60psi; Ion Source Gas 2, 60psi; DP, ±100V (Allwegene Technology Co., Ltd, Beijing, China).

SCIEX Analyst Work Station Software (Version 1.6.3) was employed for MRM data acquisition and processing. MS raw data (.wiff) files were converted to the TXT format using MSconventer. An in‐house R program and database were applied to peak detection and annotation (Allwegene Technology Co., Ltd, Beijing, China).

This research was financially supported by the National Key R & D Program of China (2022YFF1001600), China Postdoctoral Science Foundation (2022M720495), the National Natural Science Foundation of China (31900744 and 32372131), Hainan Seed Industry Laboratory (B21HJ0215), Beijing Natural Science Foundation (6234044) and the Key Research and Development Program of Hainan Province (ZDYF2022XDNY135). We are grateful to Drs. Li‐Juan Qiu and Shi Sun of the Institute of Crop Science, CAAS for kindly providing soybean seeds and Dr. Wen‐Sheng Hou of the Institute of Crop Science, CAAS for providing soybean transforming technology.