HMMER 3.0 software was used to search the soybean genome database for genes containing the structural domain of hexokinase (PF03727 and PF00349) and 17 soybean hexokinase genes were retrieved finally. The HXK gene sequences of G. max was downloaded from Ensembl Plants (http://plants.ensembl.org/index.html). The physical and chemical parameters of the proteins, including molecular weight (MV) and theoretical isoelectric point (PI), were computed by ExPASy (http://web.expasy.org/protparam/). The presence of a chloroplast transit peptide (cTP) in the protein sequence was predicted using the ChloroP 1.1 Server (http://www.cbs.dtu.dk/services/ChloroP/). The existence of transmembrane helices (TMHs) in the protein was predicted using TMHMM Server v. 2.0 (http://www.cbs.dtu.dk/services/TMHMM-2.0/). Chromosomal localizations of GmHXKs were mapped in Map Gene2Chromosome v2 (http://mg2c.iask.in/mg2c_v2.0/).

The sequences of the GmHXKs proteins obtained from the G. max genome were aligned using DNAMAN 7.0 software to search for conserved domains by inspection using sites present in AtHXK1 as a reference. To compare evolutionary relationships, the putative HXKs from G. max, A. thaliana, Solanum lycopersicum, O. sativa and Nicotiana tabacum were used to construct the phylogenetic tree using MEGA-X with the neighbor-joining (NJ) method and 1,000 bootstrap replicates (Kumar et al., 2018). Expression data on GmHXK gene family members at different developmental stages and in different tissues under normal conditions were downloaded from the Soybase (https://www.soybase.org/). Data on differential expression for only 14 members were eventually obtained and used for subsequent analysis.

To analyze expression pattern of soybean seedlings under salt stress, soybean seedlings were grown in a growth chamber under greenhouse conditions of 28°C under a16-h light/8-h dark cycle. Three-week-old seedlings were treated with 0.5% NaCl (salt stress) or drought treatment (10%PEG 6000). The root samples of the seedlings were collected after treatment for 2-h, 8-h, 24-h, and 72-h. Then, different samples were frozen quickly in liquid nitrogen, and stored at −80°C for RNA extraction and analysis. Total RNA was isolated using the Plant RNA Kit (CWBIO, Beijing, China), and its concentration and purity were determined by Nanodrop2000 nucleic acid analyzer (Thermo, America). First-strand cDNA was synthesized from 0.5 µg of total RNA using the HiFi-MMLV cDNA Kit (CWBIO, Beijing, China), and then used as a template for qRT-PCR analysis using gene-specific primers (Supplementary Table S1). Data analysis of RT-qPCR was performed using 2^−ΔΔCT method (Livak and Schmittgen, 2001).

Gene structures were analyzed using Gene Structure Display Serve 2.0 (https://gsds.cbi.pku.edu. cn/) to investigate the exon-intron organizations of GmHXK genes based on their information given in the Ensembl Plants. The novel motifs of GmHXKs were searched using MEME (http://meme-suite. org/tools/meme) (Bailey et al., 2009). The parameters were set as follows: the site distribution was set to any number of repetitions (anr), the number of motifs was set to 10, and all other optional parameters remained default (He et al., 2019). The combination of gene structures, motifs, and phylogenetic tree was then generated using TBtools. In addition, the cis-acting regulatory elements in the 2000-bp genomic sequence upstream of the coding GmHXK gene sequences were investigated using the online PlantCARE databases. (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/).

GmHXK2 was obtained by PCR using the soybean genome DNA as a template with the primes, which are specific for GmHXK2 gene. The PCR was performed using standard conditions: initial denaturation at 94°C for 2 min followed by 35 cycles of 94°C for 30 s, 56°C for 30 s and 72°C for 1 min, followed by a final extension of 72°C for 5 min. The PCR products of GmHXK2 were gel purified using gel purification kit (TaKaRal MiniBEST Agarose Gel DNA Extraction Kit Ver.4.0) and used to construct the entry vector (Invitrogen, United States of America) by TOPO cloning reaction according to the manufacturer’s instructions. The entry vector containing the gene of the correct orientation and sequence was used to construct the target vector pMDC83 via the LR Clonase II enzyme mediated gateway cloning reaction according to manufacturer’s protocol. The recombinant plasmid pMDC83 with the hygomycin phosphotransferase gene under the regulation of double CaMV 35S promoter was generated, with GFP fused at the C-terminus of GmHXK2. Then, sequencing was performed to confirm the insertion into the correct gene in pMDC83. Next, the integrated vector was introduced into A. tumefaciens strains GV3101 by the liquid nitrogen freeze-thaw method. The A. tumefaciens transformation was transformed into WT Arabidopsis using the floral dipping technique. Transformed Arabidopsis plants were grown in a greenhouse under a 16/8 h light/dark cycle at 24/22°C with 70% relative humidity. Seeds harvested from the transformed plants (T0) were grown on 1/2 MS medium containing 20 mg L-1 hygromycin under the same growth conditions. Homozygous T3 progeny WT35S::GmHXK2 derived from T2 population were selected and confirmed by PCR for further salt tolerance analysis.

Genomic DNA of from the putative T3 transgenic Arabidopsis was isolated and analyzed by PCR amplification using specific primers. The amplification fragments were monitored in transgenic Arabidopsis with 35S:GmHXK2-GFP in pMDC83. No transformed seedlings were used as control.

The roots from 4-week-old seedlings expressing GFP in the transgenic Arabidopsis transformed with 2 × 35S::GmHXK2-GFP in pMDC83 were monitored using confocal laser-scanning microscopes. Images were captured at an excitation of 480 nm and emission between 515 and 565 nm for GFP. The subcellular localization of GmHXK2-GFP fusion proteins was confirmed.

WT and WT35S::GmHXK2 transgenic Arabidopsis plants were grown on 1/2 MS medium containing 20 mg L-1 hygromycin. After 15 days of cultivation, total proteins of the seedlings were extracted from WT and transgenic Arabidopsis seedlings with a buffer consisting of 50 mM Tris/HCl (pH 8.0), 150 mM NaCl, 1 mM EDTA, and 0.2% (w/v) Triton X-100, 4% β-mercaptoethanol, 1 mM dithiothreitol (DTT), and 1% (v/v) protease inhibitor cocktail of which were then used for protein quantification using the BCA Protein Quantitative Kit (Boster). The protein samples (200 μg amounts) were electrophoresed in 8% SDS-PAGE and the gels were transferred to nitrocellulose membranes. The membranes were blocked with TBST buffer (10 mM Tris/HCl, pH 7.5, 150 mM NaCl, and 0.05% Tween-20) supplemented with 5% non-fat milk for 2 h and incubated with primary antibodies (Anti-GFP antibody, abcam, diluted at 1:1,000) in TBST buffer with 5% BSA overnight at 4°C. Afterwards, the membranes were washed three times (10 min each) with TBST buffer and incubated with the secondary antibodies (Goat Anti-Mouse IgG H&L (HRP), abcam, dilution at 1:1,000) for 2 h. After washing three times with TBST buffer, the membranes were incubated with a chromogenic agent using the Enhanced HRP-DAB Chromogenic Substrate Kit (Boster).

Seeds of WT and WT35S::GmHXK2 transgenic Arabidopsis plants of homozygous T3 generation were sterilized using 75% ethanol and 5% sodium hypochlorite for 1 min and 10 min, respectively. The seeds were washed six to ten times in sterilized water and were sown on half-strength MS (1/2 MS) medium with 0 and 100 mM NaCl. They were then transferred to a culture room after 3 days of vernalization at 4°C. Finally, root lengths and fresh weight of the germinated seeds were measured after planting for 7 days to detect salt tolerance in transgenic Arabidopsis overexpressing GmHXK2.

To analyze salt tolerance of transgenic Arabidopsis plants at the seedlings stage, treatment of plant samples were as followed. After the seeds were sterilized, they were sown on MS medium after 3 days of vernalization at 4°C, then they were transferred to a culture room and grown for 9 days. Next, Arabidopsis seedlings were transplanted into 1/2 MS medium containing 0 mM, 100 mM and 150 mM NaCl with or without 100 mM glucose (Glc). Finally, fresh and dry weight, length of the roots, malondialdehyde, chlorophyll and proline contents of the seedlings were measured after 6 days.

Chlorophyll was extracted using ethanol as solvent (Matschinsky et al., 2006). 100 mg of leaves was pulverized with 2 mL 95% ethanol, and the sample supernatant was measured with a spectrophotometer UV-1800PC at 665 nm and 649 nm. The chlorophyll extraction solution was calculated using the following formula: C a = 13.95 A 665 − 6.8 A 649 C b = 24.96 A 6665 − 7.32 A 649 C T = C a + C b

The chlorophyll content per unit fresh weight of leaf was calculated using the following formula: Chlorophyll content (mg g-1) = (CT×extraction solution volume×dilution ratio)/(fresh  weight  of  leaves).

Fresh seedlings (0.1 g) were immediately homogenized with liquid nitrogen and were then mixed with 5 mL of 10% trichloroacetic acid (TCA) and centrifuged at 4,000 rpm for 10 min. The supernatant (2 mL) and 2 mL 0.6% thiobarbituric acid (TBA) were pipetted into a new tube. The mixture was incubated in a water bath at 100°C for 15 min and then cooled on ice and centrifuged at 5,000 rpm for 10 min. The absorbance of the supernatant was measured at 532 and 450 nm by spectrophotometer (Zhang et al., 2008). The MDA content was calculated using the formula: MDA (nmol g-1) = (6.45× 10⁻⁶×A532–0.56×10⁻⁶×A450)×V/W. V = volume of supernatant(L), W = weight of seedlings (g).

The proline content of Arabidopsis seedling was determined according to the method described by Jaemsaeng et al. (Zhang et al., 2008). Seedlings (0.05 g) were homogenized in 5 mL of sulphosalicylic acid (3%) and centrifuged for 10 min at 12,000 rpm. 2 mL of glacial acetic acid and 2 mL of acid ninhydrin were added to the supernatant (2.0 mL). The mixture was boiled in a water bath at 100°C for 30 min. After cooling, extraction was done with 4 mL of toluene. The absorbance was measured at 520 nm using toluene as a blank. The proline content (mg*g^-1) = (Y × 5/2)/W. Y = content of proline in 2 mL supernatant. W = Weight of seedlings (g).

Seedings (0.1 g) were homogenized with 1 mL phosphate buffer (100 mM, ph = 7.8) and centrifuged at 12000 rpm for 30 min at 4°C. Then 0.1 mL of supernatant and phosphate buffer were added to the reaction solution (50 mM, ph = 7.8 Na₃PO₄,130 mM MET, 750uM NBT, 100uM EDTA, 20uM riboflavin) respectively. The reaction solution in which phosphate buffer was added served as the light control. The reaction was placed under fluorescent light for 15 min and then terminated in the dark. The absorbance value was measured at 560 nm. The SOD was calculated using the formula: SOD(U)=(OD_L-OD_S)*V/(0.5*OD_L*V_S*T*M). OD_L = Absorbance value of light control. OD_S = Absorbance value of samples. V = Total volume of sample extracts (ml). V_S = Volume of sample extract for determination (ml). T = Light reaction time (min). M = Weight of seedings(g) (Donahue et al., 1997).

The conductivity of the sample (0.2 g) immersed in 20 mL of deionized water for 2 h at room temperature was measured as C1. Then the sample was boiled in a water bath for 15 min and cooled to room temperature. The conductivity measured was C2. Conductivity of deionized water as a blank control (C0). The electrolyte leakage (%)=(C1-C0)/(C2-C0) (Jungklang et al., 2017).

The specific fragments from CDS of GmHXK2 and GmPDS were amplified using primers. The PCR products and the virus vector pTRV2 were digested with XbaI and BamHI(TAKARA),respectively. The products were ligated to obtain pTRV2-HXK2 and pTRV2-PDS vectors. For VIGS experiment, plasmids of pTRV1, pTRV2 and pTRV2 recombinant vectors (pTRV2-PDS and pTRV2-HXK2) were transformed into A. tumefaciens strain GV3101 cells by using the freeze-thaw method. The empty plasmid pTRV2 was used as control (TRV:00). The Agrobacterium strains were inoculated into 50 mL of LB medium as above on a shaker at 180 rpm at 28°C for 12–16 h to an OD600 of 1.2–1.5. The Agrobacterium cells were centrifuged at 4000 g for 10 min at room temperature and washed twice, resuspended with the infiltration solution (10 mM MgCl₂, 10 mM MES, and 200 μM acetosyringone) to a final OD600 of 0.8–1.0 and placed at room temperature in darkness for 3 h. The infiltration solution of the Agrobacterium strain containing pTRV1 was mixed with the infiltration solution of pTRV2 or Agrobacterium carrying the constructs in a 1:1 ratio (v/v) and 20–40 washed soybean seeds were added respectively (Zhao et al., 2020b). Soybean seeds that had been soaked for 24 h in darkness at room temperature were removed and planted in nutrient soil. The silencing efficiency of soybean seedlings was determined by qRT-PCR when the first true leaf was fully expanded.

The results of root length, fresh weights, and dry weights were from five independent experiments, and results of Chlorophyll, MDA, proline contents, EL and SOD were from three independent experiments. GraphPad Prism 5 was utilized for all statistical analysis. Data significance analysis was performed using Student’s t-test (Yang et al., 2022). Data are presented as mean ± SE. Single, double and three asterisks denote significant differences compared with the values of WT at p < 0.05, p < 0.01 and p < 0.001, respectively.

In total, 17 HXK genes were identified in the G. max genome, and they were designated as GmHXK1-17. The name, Ensembl Plants accession number, mRNA and CDS length of nucleotide sequence as well as the length of amino acid sequence, MW, PI, chromosome location, cTP, and number of TMHs, are summarized in Table 1. As depicted in Table 1, the length of the CDS varied from 471 to 1,812 bp, encoding 156 to 504 amino acids and corresponding to molecular weights ranging from 17.42 to 55.40 kDa. The theoretical PIs of these proteins ranged from 5.18 to 8.76. There were three genes on chromosomes 1, 3 genes on chromosomes 17, and 2 genes on chromosomes 5, 7, 8, and 11. GmHXK2, GmHXK10 and GmHXK4 were distributed on chromosomes 9, 12 and 14, respectively. GmHXK1, GmHXK2 and GmHXK11 contained no cTP or TMHs. GmHXK5, GmHXK6, GmHXK12–17 contained cTP and one TMH. GmHXK3 and GmHXK4 contained cTP but no TMHs.

To better study the structural difference between GmHXKs and their possible functional differences, the amino acid sequences of G. max were aligned and analyzed (Figure 1). Conserved domain analysis showed that most GmHXKs were multidomain proteins, which contained two phosphate sites (I and II), two connect sites (I and II), one α-helix site, one adenosine binding site, and one sugar binding site. However, GmHXK1 contained just three conserved domains, α-helix site, adenosine binding site, and sugar binding site. Additionally, GmHXK2 contained phosphate site II, two connect sites (I and II), one α-helix site, one adenosine binding site, and one sugar binding site.

To further reveal the evolutionary relationship of hexokinase family, the protein sequences of 56 hexokinases from five different plant species, including G. max, A. thaliana, S. lycopersicum, O. sativa and N. tobacum were used to construct phylogenetic tree using MEGA-X. As shown in Figure 2, all hexokinase genes were divided into three subfamilies (I–III). Most of the HXK genes of five different plant species were sorted into Cluster III. Cluster II contains fewer genes. It indicated the close relationship between the five plant species aforementioned. There were none GmHXKs in Cluster I, which contains only some hexokinase genes in S. lycopersicum.

Generally speaking, the conserved regions of proteins frequently distinguish them from other proteins and determine the basis of their functions. To reveal the structural diversity and functional characteristics of the GmHXK family members, their conserved motifs were analyzed by MEME and mapped in the phylogenetic tree. In total, 10 conserved motifs were identified (Figure 3B). The detailed sequences and conserved motifs are shown in Supplementary Table S2. The identified motifs ranged from 29 to 50 amino acids in length. Among them, motifs 3, 6 and 8 were found to encode the hexokinase_1 domain, while motifs 1, 2, four and 5 encode the hexokinase_2 domain. Nevertheless, the functions of motif 7, 9 and 10 are unknown. GmHXK7-10, 12, 14–17 were found to have all 10 motifs, while other GmHXK family members had only portion motifs (Figure 3B). GmHXK3 and GmHXK4 contained all motifs except motif 2, and GmHXK5 and GmHXK6 contained all motifs except motif 9.

Different combinations of exons and introns can lead to diverse gene function. To explore the structure diversity of GmHXK genes, Gene Structure Display Server 2.0 was employed to analyze the distribution of exon-intron structure based on the corresponding genome and coding sequences (Figure 3C). The results showed that most GmHXK genes contained nine exons and eight introns. GmHXK2 and GmHXK5 contained 10 exons and nine introns. GmHXK1, contained only three exons and two introns.

The identification of the cis-regulatory elements in the promoter part of the gene is important for functional and regulatory studies. To investigate the cis-regulatory elements of the 17 GmHXK genes, we analyzed the sequence 2,000 bp upstream of the start codon ATG. The cis-elements of the GmHXK genes were classified into three categories, involved in plant growth and development, stress responses and hormone-induced response (Supplementary Figure S1). The plant growth and development category contained seed-specific regulation (RY-element) and meristem expression (CAT-box) cis-elements. In the stress-responsive category, the elements included dehydration-responsive (DRE), anaerobic induction (ARE and GC-motif), low-temperature-responsive (LTR) and MYB-binding sites involved in drought inducibility (MBS), as well as defense and stress responsiveness (TCA). Seven types of phytohormone responsive cis-elements were detected, including auxin-responsive (AuxRR-core), abscisic acid-responsive (ABRE), methyl jasmonate-responsive (CGTCA-motif), ethylene-responsive (ERE), gibberellin responsive (GARE-motif), salicylic acid-responsive (TCA-element) and heat shock, osmotic stress, low pH, nutrient starvation-responsive (STRE). It is worth noting that the cis-elements MYB and MYC associated with phytohormone and abiotic stress were present in 17 GmHXK genes.

We analyzed the expression patterns of 14 GmHXK family members in different tissues including young-leaf, flower, pod, pod shell, seed, root and nodule using data in soybase (Figure 4A). It showed that GmHXK1,5,6,13,14 was expressed at a very low level or not expressed in seven tissues above. The level of expression of GmHXK4,7,8,9,17 was not obvious in majority of the plant tissues except for that in a certain tissue. GmHXK3,12,15,16 were expressed obviously in most tissues.

And we analyzed also the expression of GmHXKs under salt and drought stress after treatments for different times by qRT-PCR. By analyzing expression data after treatment with NaCl or drought stress for 0, 2,8,24,72 h, we found that the expression of GmHXK 2,3,6,9,13,15 under treatment with NaCl (Figure 4B) and GmHXK2,4,6,7,9,15 under drought treatment with drought stress (Figure 4C) were increased after treatments within 72 h. The expression of other GmHXKs under salt or drought stress was obvious at the initial stage of treatment but decreased over time within 72 h. Owing to the increase in expression of GmHXK2,6,9,15 under both salt and drought stress, we selected GmHXK2 to study further its function under the salt stress.

GmHXK2 gene (Gene ID: GLYMA_09G144600) fragment containing introns without stop codon, was isolated by PCR with primers using genome DNA as a template. Fragments of size 2,033 bp were detected (Figure 5A). Then, the PCR product was inserted into a pMD18-T vector and sequenced. Sequence analysis suggested that the fragment was identical to the known GmHXK2 gene. The fragments were cloned into the entry vector TOPO by TOPO cloning reaction and subsequently into gateway destination vector pMDC83. Products amplified by PCR using genome DNA from T3 transgenic Arabidopsis WT35S::GmHXK2 seedlings of expected size 1,357 bp were detected (Figure 5B). Our results show that the GmHXK2 gene was integrated into the genome of WT.

To further explore the role of GmHXK2 in plant salt tolerance, we constructed a TRV-VIGS vector by selecting specific sequence in the CDS region of the GmHXK2 gene and identified the efficiency in silencing of GmHXK2-silenced plants by qRT-PCR, which were 30%–50% of control plants (Supplementary Figure S2). The silenced plants showed different phenotype from the control. Margins of silenced plants in first true leaf were discolored and wilted (Figure 6B) compared with the control (TRV:00). The validity of silenced plants was further determined by the PDS phenotypes (Figure 6C). Finally, we selected the plants with the highest efficiency of silencing for the subsequent experiment on salt stress.

After treatment with salt stress, all leaves in soybean plants had been wilt, curled, and were yellow to some extent, and the leaves of GmHXK2-silenced plants were more severely damaged than control (Figure 6A).

To investigate the potential physiological mechanisms by which GmHXK2 enhances plant tolerance, the proline, chlorophyll content, EL and SOD in TRV:00 and TRV:HXK2 plants were measured under normal and salt stress conditions. Our results showed that the EL of TRV:HXK2 (26.18%) was significantly higher than that of TRV:00 (11.33%) (Figure 7B) under salt conditions. This indicated that TRV:HXK2 plants were damaged more seriously than the control plants. The content of proline, chlorophyll and SOD of TRV:HXK2 (0.60 mg/g, 3.06 mg/g and 12.61 U, respectively) were decreased compared with that of TRV:00 plants (1.11 mg/g, 4.31 mg/g and 16.80 U, respectively) (Figures 7A, C, D). These results suggested that capability of TRV:HXK2 plants in resisting salt stress and scavenging ROS generated intracellularly diminished compared with control plants.

The plasma membrane Na⁺/H⁺ antiporters GmSOS1 (salt overly sensitive) (Ma et al., 2020), the high-affinity potassium transporters (GmHKTs) (Sun et al., 2021), GmbZIP44 (Zhao et al., 2020a), and GmSALT3 as one of cation/H exchangers, which play a vital role in response to environmental salinity (Xu et al., 2022a), were known to be involved in the regulation of salt tolerance in soybean. So we investigated the expression of these Salt-Responsive genes to ascertain the relationship of GmHXK2 and Na⁺ homeostasis.

Under control conditions, expressions of both GmSALT3 (Figure 8A) and GmSOS1 (Figure 8D) were decreased in silenced plants compared with WT, while expression of GmbZIP44 (Figure 8B) and GmHKT1 (Figure 8C) were not significantly different. However, the expression of all four genes in GmHXK2-silenced plants under salt stress was decreased significantly than WT (Figure 8). These results suggested that silence of the GmHXK2 gene downregulated the expression of salt stress-related genes involved in Na⁺ homeostasis, implying that GmHXK2 had an important role in remaining homeostasis of Na⁺ and K⁺.

As shown in Figure 9A, signals of GmHXK2-GFP were distributed on cell wall and cell membrane, which revealed that GmHXK2 was predominantly localized on both vacuolar membrane and cell membrane. To examine GmHXK2 protein expression, transgenic Arabidopsis seedlings WT35S::GmHXK2 were further analyzed via immunoblotting with an antibody specific to GFP by Western blot. The results showed that T3-T6 transgenic lines expressed the integrated protein with an expected molecular mass of 59.92 kDa (Figure 9B), which corresponds to the fusion protein, while no proteins of this size were observed in that of WT seedlings.

As shown in Figure 10A, the WT and WT35S::GmHXK2 transgenic Arabidopsis grew well in the control group. After 100 mM NaCl treatment, the growth of the Arabidopsis was inhibited, but the growth of WT was more suppressed than that of WT35S::GmHXK2 transgenic Arabidopsis.

Under control condition, the root length of WT35S::GmHXK2 was significantly higher than that of WT, and fresh weight of WT35S::GmHXK2 were increased by 47% compared with WT. After NaCl treatment, the root length and fresh weight of Arabidopsis were decreased, and the root length and fresh weight of WT35S::GmHXK2 were significantly higher than that of WT. (Figures 10B, C). Hence, the expression of GmHXK2 promoted the growth of Arabidopsis.

To verify whether GmHXK2-expressing Arabidopsis contributes to salt tolerance, we observed the phenotype of transgenic Arabidopsis and WT after 6 days of NaCl and glucose treatments and then measured the physiology and biochemistry indexes. Under salt stress, the growth of Arabidopsis seedlings were obviously inhibited, the leaves of all plants gradually became yellow and shriveled in addition to roots becoming shorter. In addition, the growth of WT was more suppressed than that of WT35S::GmHXK2. After adding 100 mM glucose, the growth condition of seedlings was much better compared to those at 0, 100 and 150 mM NaCl treatments. (Figure 11).

After NaCl treatment, the fresh and dry weight, root length and chlorophyll content decreased, though the MDA and proline contents increased (Figure 12). As shown in Figures 12A, B, the fresh weight and dry weight of WT35S::GmHXK2 transgenic Arabidopsis were higher than that of WT under 0, 100 and 150 mM NaCl treatment. After salt treatment, root length, chlorophyll and proline contents of WT35S::GmHXK2 seedlings were significantly increased compared to WT, but the MDA content of WT35S::GmHXK2 seedlings was much lower than that of WT (Figures 12C–F).

Exogenous glucose counteracted the effect on salt stress. The fresh weight and dry weight, root length, chlorophyll and proline contents of WT and WT35S::GmHXK2 were increased compared to those at the 0, 100 and 150 mM NaCl stress. And the application of glucose under salt treatment inhibited the increase of MDA content under salt stress. (Figure 12).

These results revealed the expression of GmHXK2 could enhanced the salt tolerance of plants during the seedling stage. In addition, 100 mM exogenous glucose alleviated the inhibition of salt stress on the growth of Arabidopsis plants.