Metabolomic and antioxidant enzyme activity changes in response to cadmium stress under boron application of wheat (Triticum aestivum)

Boron (B) has previously been shown to inhibit cadmium (Cd) uptake in wheat. Here, we investigated the physiological response of external B application (C for no B added, B for B added, B+Cd for B and Cd added, B/Cd for B 24 h pretreatment before Cd added, B and Cd were 46.2 μM and 5 μM, respectively) on wheat growth under Cd stress. The results showed that the wheat growth was significantly weaker under Cd treatment, while B application did not significantly improve the wheat growth under Cd stress. However, B application decreased Cd concentrations and malondialdehyde (MDA) concentrations of shoot and root. The key enzyme activities including superoxide dismutase (SOD) and peroxidase (POD) significantly increased under Cd treatments while decreased under B treatments. Further, a total of 198, 680 and 204 of the differential metabolites were isolated between B and C treatment, Cd and C treatment and B+Cd and Cd treatment, respectively. The metabolites with up-accumulation in B application (B+Cd) roots were mainly galactaric acid, citric acid, N6-galacturonyl-L-lysine, D-glucose, while the metabolites with down-accumulation were mainly threoninyl-tryptophan and C16 sphinganine. The differential metabolic pathways were mainly concentrated in linoleic acid metabolism, galactose metabolism, sphingolipid metabolism, glycolysis/gluconeogenesis, propanoate metabolism in diabetic complications between B+Cd treatment and B treatment. The results indicate that B alleviates Cd toxicity in winter wheat by inhibiting Cd uptake, increasing antioxidant enzyme activity and changing metabolites.


Introduction
It is well known that cadmium (Cd) is one of the heavy metals and has high toxicity for human and plant health. For human, Itai-itai is the best known Cd-induced disease, and other diseases such as renal dysfunction, osteoporosis and bladder cancer have been reported to be related to Cd exposure (Feki-Tounsi and Hamza-Chaffai 2014). The most important cause of these diseases is due to long-term consumption of Cd-contaminated food. Studies show that food as the main source of Cd exposure contributed about 90% of the total Cd intake in the general population (Clemens et al. 2013;He et al. 2013). Therefore, it is an important measure for human food safety to prevent the absorption of Cd in crops and its migration to edible parts. In plant, Cd can enter root and be transported to the ground through the various Cd transport systems, such as transporters for uptake in roots (OsNramp5, HvNramp5, OsIRT1, OsIRT2, OsNramp1 and OsCd1), transporters for vacuolar sequestration in roots (OsHMA3), transporters involved in rootto-shoot translocation (OsHMA2, OsHMA3, OsZIP7 and OsCAL1) and transporters for distribution to the grains (OsHMA2, OsLCT1, OsZIP7 and OsCCX2) (Ma et al. 2021). Cd affects plant growth and development due to its toxic effects on different metabolic and physiological processes, such as causing leaf chlorosis, necrotic lesions, wilting, inhibited root elongation and reduced biomass (Qin et al. 2020a;Rizwan et al. 2016). Therefore, reducing the Cd uptake or migration to edible sites is the key factor Responsible Editor: Gangrong Shi * Peng Zhao zhpddy@163.com 1 to ensure no Cd pollution in grains in polluted soil and stop the migration to the food chain. Many agronomic strategies have been shown to reduce Cd absorption in crop, such as soil removal and replacement, soil turnover and dilution, soil dressing, chemical washing, phytoremediation and inorganic amendments (Wang et al. 2011;Hseu et al. 2010;Uraguchi and Fujiwara 2012;Rizwan et al. 2016). Among these strategies, inorganic amendments were also reported as a low-cost but effective approaches. For example, there is an efficient competition between Cd with iron (Fe), manganese (Mn), zinc (Zn) and calcium (Ca) in uptake, and the application of these micronutrients commonly decreases the uptake and accumulation of Cd (Qin et al. 2020a, b;Meda et al. 2007;Astolfi et al. 2012;Duan et al. 2018). Silicon (Si) fertilizer or nana-silicon application could reduce Cd absorption and alleviate the toxicity to crop (Ma et al. 2015;Wang et al. 2015). It was also reported that selenium (Se) fertilization could reduce Cd uptake and mitigates Cd toxicity in crop (Hu et al. 2014;Lin et al. 2012). The underlying mechanism of effects of these elements on Cd toxicity and accumulation may include (1) competitive uptake via the same transport system (Fahad et al. 2015;Qin et al. 2020a), (2) increase the antioxidative capacity and photosynthetic characteristics (Lin et al. 2012;Qin et al. 2018a, b), (3) induce the formation of phytochelatin and cysteinerich peptides (Jasinski et al. 2003;Semane et al. 2007).
Boron (B) is an essential micronutrient for plant growth. It is essential for the structural and functional integrity of the cell wall and membranes, photosynthesis, cell division and elongation, nitrogen and carbohydrate metabolism, etc. (Shireen et al. 2018). In recent years, it has been found that B plays an important role in alleviating abiotic stress in plants. For example, B alleviates aluminum (Al) toxicity by promoting root alkalization in transition zone via polar auxin transport , and B alleviates Cd accumulation and toxicity by improving oxidative stress and suppressing Cd uptake and transport (Chen et al. 2019). Likewise, our previous work showed that B could inhibit Cd uptake by regulating gene expression in wheat (Triticum aestivum L.) (Qin et al. 2020b). However, the physiological mechanism why B inhibits Cd absorption is yet to be investigated. To understand how B influences the uptake of Cd and Cd toxicity, antioxidant substances and antioxidant enzyme activity were studied, and the different metabolites and metabolic pathways were analyzed.

Plant material and experimental conditions
Hydroponic experiment was carried out on Henan Agricultural University, Zhengzhou, China. Wheat seeds (Triticum aestivum L. cv. Zhengmai 379) were surface-sterilized in 0.5% Na-hypochlorite for 15 min, and seeds were then rinsed carefully with deionized water. The seeds were germinated on a plastic net floating on deionized water, and 7-day-old wheat seedlings with similar size were transferred to the incubators to grow in a controlled chamber with photoperiod of 16  O. The solution pH was adjusted to 6.5 using 1 M HCl or NaOH. Half-strength nutrient solution was applied for the first 7 days and changed to full strength solution for another 4 days. Each treatment was replicated four times. Five treatments was set as follow: (1) C (without B); (2) B (B was added after 7 days of culture); (3) Cd (Cd was added after 7 days of culture); (4) B+Cd (B and Cd were added after 7 days of culture); (5) B/Cd (B 7-day pretreatment before Cd exposure). Cd was applied as CdCl 2 solution with a final Cd concentration of 5 μM according to Qin et al. (2020b). After 23-day Cd treatments, root and leaf samples rinsed in distilled water were harvested and used fresh or kept frozen in liquid N 2 for various biochemical assays.

Cd concentration analysis
Plants were divided into root and shoot fractions and ovendried at 70 °C to constant weight. These fractions were then digested with mixed acid (HNO 3 :HClO 4 (4:1, v:v)) in a Microwave Digestion System. Cd concentration was determined by graphite furnace atomic absorption spectrometry (GFAAS). The translocation factor (TF) was calculated as follows: TF = shoot C Cd / root C Cd .

Malondialdehyde (MDA) determination
The level of lipid peroxidation is estimated as the amount of MDA determined by the thiobarbituric acid (TBA) reaction as described by Qin et al. (2018a, b). Briefly, 500 mg of fresh root or shoot samples was homogenized with 2 mL 5% v/v trichloroacetic acid (TCA) and centrifuged at 10,000×g for 10 min. Then, 1 mL 0.5% v/v TBA in 20% v/v TCA was added to 1 mL supernatant, and the homogenate was boiled for 30 min at 95 °C. The mixture was followed by an immediate cooling on ice to stop the reaction and centrifuged at 10,000×g for 5 min. The absorbance was determined at 450 nm, 532 nm and 600 nm, and MDA concentration was estimated by the formula: MDA (μmol/g FW) = 6.45(OD 532 − OD 600 ) − 0.56 OD 450 .

Antioxidant enzyme activities
Antioxidant enzyme activities (SOD, CAT and POD) in roots and leaves were determined spectrophotometrically according to previous literatures and make some modifications (Molins et al. 2013;Wu et al. 2017a, b). Fresh tissues (0.5 g) were ground in liquid nitrogen to a fine powder using a mortar and pestle. The powder was transferred to a pre-cooled (4 °C) mortar and pestle with 5 mL of 50 mM phosphate buffer (pH 7.8), containing 0.12 mM EDTA and 2% (w/v) polyvinylpolypyrrolidone. The homogenate was centrifuged at 12,000×g at 4 °C for 20 min. The supernatant was used for assays of the antioxidant enzyme activities. The assay of SOD was carried out briefly as follows: the assay mixture (total 3 mL) for SOD contained 0.05 M phosphate buffer, containing 12 μM EDTA and 13 mM L-methionine, 75 μM nitroblue tetrazolium (NBT), 2 μM riboflavin in 5 mM KOH and 0.05 mL enzyme extract. Reaction was initiated by placing the glass test tubes in between two fluorescent tubes. By switching the light on and off, the reaction was started and terminated, respectively. The increase in absorbance due to formazan formation was read at 560 nm. The activity was expressed as ΔA 560 (U mg −1 protein min −1 ). CAT activity was assayed. The reaction mixture for CAT comprised 1.5 mL of 25 mM phosphate buffer (pH 7.0), 0.3 mL of 0.1 M H 2 O 2 and 0.2 mL of enzyme extract. The change of absorbance value within 210 s was measured at 240 nm with the blank setting of deionized water. The CAT activity was expressed as U mg −1 protein min −1 . Guaiacol peroxidase activity was determined spectrophotometrically by measuring the oxidation of guaiacol to tetraguaiacol at 470 nm. The reaction mixture contained 1 mL of 50 mM phosphate buffer (pH 5.5), 1 mL of 0.3% H 2 O 2 , 0.95 mL 0.2% (v/v) guaiacol and 0.05 mL enzyme extract. The change rate of absorbance value was recorded at 470 nm within 210 s, and the number of reads was recorded every 30 s. A peroxidase activity unit (U) was determined by △A 470 changing by 0.01 per minute, and the activity was expressed as U mg −1 protein min −1 . Soluble protein in root and leaf tissues were assayed as described by Qin et al. (2017).

Metabolite extraction and metabolite profiling analysis
The identification of differential metabolites was performed by Shanghai Lu Ming Biological Technology Co. Ltd. Eighty milligram accurately weighed sample was transferred to a 1.5-mL Eppendorf tube. Two small steel balls were added to the tube. Twenty microliters of 2-chloro-lphenylalanine (0.3 mg/mL) dissolved in methanol as internal standard, and 1 mL mixture of methanol and water (7/3, v/v) was added to each sample. Samples were placed at −20 °C for 2 min, grinded at 60 HZ for 2 min and ultrasonicated at ambient temperature for 30 min after vortexed, then placed at −20 °C for 20 min. Samples were centrifuged at 13,000 rpm, 4 °C for 10 min. In a freeze concentration centrifugal dryer, 0.3 mL of supernatant in a brown and glass vial was dried. To each sample, 0.4 mL mixture of methanol and water (1/4, v/v) was added. Samples were vortexed for 30 s, then placed at 4 °C for 2 min. Samples were centrifuged at 13,000 rpm, 4°C for 5 min. The supernatants (150 μL) from each tube were collected using crystal syringes, filtered through 0.22-μm microfilters and transferred to LC vials. The vials were stored at −80 °C until LC-MS analysis. QC samples were prepared by mixing aliquots of the all samples to be a pooled sample. An ACQUITY UHPLC system (Waters Corporation, Milford, USA) coupled with an AB SCIEX Triple TOF 5600 System (AB SCIEX, Framingham, MA) was used to analyze the metabolic profiling in both ESI positive and ESI negative ion modes. An ACQUITY UPLC BEH C18 column (1.7 μm, 2.1×100 mm) was employed in both positive and negative modes. The binary gradient elution system consisted of (A) water (containing 0.1% formic acid, v/v) and (B) acetonitrile (containing 0.1% formic acid, v/v) and (B) acetonitrile (containing 0.1% formic acid, v/v), and separation was achieved using the following gradient: 0 min, 5% B; 2 min, 20% B; 4 min 25% B; 9 min 60% B; 14 min 100% B; 18 min 100% B; 18.1 min 5% B and 19.5 min 5% B. The flow rate was 0.4 mL/min, and column temperature was 45 °C. All the samples were kept at 4 °C during the analysis. The injection was 10 μL. Data acquisition was performed in full scan mode (m/z ranges from 70 to 1000) combined with IDA mode. The QCs were injected at regular intervals (every 10 samples) throughout the analytical run to provide a set of data from which repeatability can be assessed.
The acquired LC-MS raw data were analyzed by the progqenesis QI software (Waters Corporation, Milford, USA). The internal standard was used for data QC (reproducibility). Metabolites were identified by progenesis QI (Waters Corporation, Milford, USA) Data Processing Software, based on public databases such as http:// www. hmdb. ca/; http:// www. lipid maps. org/ and self-built databases. The positive and negative data were combined to get a combine data which was imported into R ropls package. Principle component analysis (PCA) and (orthogonal) partial leastsquares-discriminant analysis (O)PLS-DA were carried out to visualize the metabolic alterations among experimental groups, after mean centering and Pareto variance (Par) scaling, respectively. The Hotelling's T2 region, shown as an ellipse in score plots of the models, defines the 95% confidence interval of the modeled variation. Variable importance in the projection (VIP) ranks the overall contribution of each variable to the OPLS-DA model, and those variables with VIP > 1 are considered relevant for group discrimination. The differential metabolites were selected on the basis of the combination of a statistically significant threshold of variable influence on projection (VIP) values obtained from the OPLS-DA model and P values from a two-tailed Student's t test on the normalized peak areas, where metabolites with VIP values less than 0.05 were considered as differential metabolites.

Statistical analysis
Data were analyzed by one-way analysis of variance (ANOVA) using the statistical package SPSS, version 20.0 (SPSS, Chicago, IL, USA). All data are the means ± SE of four independent replicates; LSD's multiple range test was used to determine the significant differences between means (P < 0.05).

Effect of different treatments on wheat growth and Cd accumulation
From the perspective of growth, the wheat growth under Cd treatment (Cd treatment, B+Cd treatment and B/Cd treatment) was significantly weaker than that without Cd treatment (C treatment and B treatment) (Fig. 1a). B application improved the wheat growth under Cd stress (Fig. 1b). Similar trend was observed in B/Cd treatment compared with B+Cd treatment (Fig. 1c). In addition, the fresh weight (FW) of root and shoot was decreased under Cd treatment compared with without Cd treatment (Table 1). However, the root FW increased by 15.22% and 56.52% under B+Cd and B/Cd treatment compared with Cd treatment, and the shoot FW increased by 10.59% and 45.88% under B+Cd and B/ Cd treatment compared with Cd treatment, respectively. Cd treatment significantly reduced the root and shoot water content, while little effect on water content with B treatment. Compared with Cd, shoot Cd concentrations decreased by 8.78% and 3.99% under B+Cd and B/Cd treatment, where root Cd concentrations decreased by 26.34% (P < 0.05) and 26.30% (P < 0.05), respectively. However, the transport coefficients of Cd transport from roots to shoots increased 24.02% (P < 0.05) and 30.88% (P < 0.05) under B+Cd and B/Cd treatment compared with CK+Cd treatment.

Effect of different treatments on MDA concentration and antioxidant enzyme activities
Lipid peroxidation measured in terms of MDA increased in all the Cd stressed plants comparing with control. Compared with C treatment, MDA concentration significantly increased with 120.8% (P < 0.05), 62.00% and 122.68% (P < 0.05) in root, and 188.86% (P < 0.05), 122.25% (P < 0.05) and 221.72% (P < 0.05) in shoot at Cd, B+Cd and B/ Cd treatments, respectively (Fig. 2). The MDA concentration decreased 26.62% in root and 23.06% in shoot respectively under B+Cd treatment compared with Cd treatment. However, the MDA concentrations have no significant difference between B/Cd treatment and Cd treatment.
In the present study, the key enzymes including SOD, POD and CAT exhibited differential responses to Cd treatments in shoots and roots of wheat (Fig. 3). It was found that SOD activities significantly increased by 84.11% (P < 0.05), 54.19% (P < 0.05) and 48.44% (P < 0.05) at Cd, B+Cd and B/Cd compared with C treatment in root respectively (Fig. 3a), while no significant differences in shoot (Fig. 3b). However, SOD activities of root decreased by 16.25% and 19.37% at B+Cd treatment and B/Cd treatment compared with Cd treatment, respectively. Similarly, POD activities under Cd treatment showed an increased tendency, especially increased by 359.52% (P < 0.05), 300.49% (P < 0.05) and 366.21% (P < 0.05) at Cd, B+Cd and B/Cd compared with C treatment in shoot (Fig. 3d). However, there was no significant difference in POD activities between B+Cd and B/Cd compared with Cd treatment. Different from SOD and POD activity, the CAT activity of the Cd-treated plants was decreased compared with the C treatment in both shoot and root ( Fig. 3e and f). CAT activities decreased 4.40%, 56.82% and 85.38% (P < 0.05) at Cd, B+Cd and B/Cd compared with C treatment in root, respectively. In shoot, the CAT activities decreased 57.30% (P < 0.05), 65.70% (P < 0.05) and 29.34% at Cd, B+Cd and B/Cd compared with C treatment, respectively.

Effect of different treatments on root extract metabolite
To better explore the differential metabolites in B that alleviate the Cd toxicity metabolites, the differential metabolites of wheat under C, Cd and B+Cd were analyzed by LC-MS.  Our data showed that the parameters of the OPLS-DA for the classification were expressed by the R2Y (cum) of 0.866 and Q2 (cum) of 0.368 for B and C treatment (Fig. 4a), 0.983 and 0.941 for Cd and C treatment (Fig. 4b), 0.81 and 0.0107 for B+Cd and B treatment (Fig. 4c), respectively. In addition, methods of 7-fold cross validation and response permutation testing (RPT) were used to investigate the quality of the model, indicating that there was no overfitting in the measurement model. These parameters representing the R2Y and Q2 exhibited a better stability and predictability and effectively reflected the metabolic differences between treatment. The volcano plot were used to visualize the P value and the fold change value to screen the differential metabolites under positive ion mode and negative ion mode. As shown in the Figure 4, the red origin represents the significantly upregulated differential metabolites in the experimental group, the blue origin represents the significantly down-regulated differential metabolites and the gray point represents the non-significantly different metabolites. Using VIP screening (variable importance in the projection) >1 at a level of P < 0.05, a total of 198 of the differential metabolites were isolated between B and C treatment, in which a total of 91 of the differential metabolites were up-regulated (red dots, Fig. 4d) and a total of 107 of the differential metabolites were down-regulated (blue dots, Fig. 4d) in the B treatment compared with CK treatment, respectively. In the same way, a total of 680 of the differential metabolites were isolated between Cd and C treatment, in which a total of 341 of the differential metabolites were up-regulated (red dots, Fig. 4e) and a total of 339 of the differential metabolites were down-regulated (blue dots, Fig. 4e) in the Cd treatment compared with C treatment, respectively. A total of 204 of the differential metabolites were isolated between B+Cd and Cd treatment, in which a total of 132 of the differential metabolites were up-regulated (red dots, Fig. 4f) and a total of 72 of the differential metabolites were down-regulated (blue dots, Fig. 4f) in the B+Cd treatment compared with Cd treatment, respectively. Furthermore, using screening conditions with more than 1 of VIP and fold change at a level of P less than 0.05, we isolated out a total of top 50 of the differential metabolites as shown in Table 2. Analyses showed that the B application (B+Cd) not only obviously  a and b), peroxidases (POX) (c and d) and catalase (CAT) (e and f) activity in root and shoot respectively under different treatments. Data are means ± SE. Bar with different letters is significantly different at P < 0.05 in multiple comparison by LSD's test up-regulated accumulation of galactaric acid, citric acid, N6-galacturonyl-L-lysine, D-glucose, 3,4,5-trihydroxy-6-{[4-methoxy-6-(3-oxoprop-1-en-1-yl)-2H-1,3-benzodioxol-5-yl]oxy}oxane-2-carboxylic acid, N-acetyldjenkolic acid and 2-methylthiophene, but also down-regulated the accumulation of threoninyl-tryptophan and C16 sphinganine in the roots compared with Cd treatment (VIP>5).

The pathway enrichment analysis of differential metabolites
The pathway enrichment analysis of differential metabolites was carried out according to the Kyoto Encyclopedia of Genes and Genomes (KEGG, https:// www. kegg. jp/) database. Pathway enrichment analysis showed that the differential metabolites (top 50) were involved in glycerophospholipid metabolism, linoleic acid metabolism, one carbon pool by folate, glycosylphosphatidylinositol (GPI)-anchor biosynthesis, autophagy-other, pentose phosphate pathway, alpha-linolenic acid metabolism, galactose metabolism, aminoacyl-tRNA biosynthesis, AGE-RAGE signaling pathway in diabetic complications between B treatment and C treatment (Fig. 5a); linoleic acid metabolism, pyrimidine metabolism, glycerophospholipid metabolism, purine metabolism, one carbon pool by folate, beta-alanine metabolism, alphalinolenic acid metabolism, cyanoamino acid metabolism, aminoacyl-tRNA biosynthesis, sphingolipid metabolism for Cd treatment and C treatment (Fig. 5b); linoleic acid metabolism, galactose metabolism, sphingolipid metabolism, glycolysis/gluconeogenesis, propanoate metabolism, fructose and mannose metabolism, AGE-RAGE signaling pathway in diabetic complications, pyrimidine metabolism, phenylalanine metabolism, arachidonic acid metabolism for B+Cd treatment and B treatment (Fig. 5c).

Discussion
Cadmium (Cd) is a toxic heavy metal that has been known causing many physiological disorders such as a decrease in photosynthesis, sugars and soluble proteins and antioxidant enzyme activities, and also increased the production of reactive oxygen species (ROS) which caused oxidative stress in plants (Abbasi et al. 2015;). This would result in a retardation of plant growth, reduction in biomass and grain yield. Boron (B) is an essential micronutrient for the structural and functional integrity of the cell wall and membranes, photosynthesis, cell division and elongation, nitrogen and carbohydrate metabolism, etc. (Shireen et al. 2018). Meanwhile, studies have also found that B plays a crucial role in alleviated Cd accumulation and toxicity (Chen et al. 2019;Qin et al. 2020b). Here, we determined the responses of wheat against Cd stress under B supply or  pre-supply, with respect to their biomass, ROS scavenging mechanism and differential metabolite mechanism. For growth parameters like biomass and length, the previous studies showed that B significantly alleviated Cdinduced growth inhibition and improved the growth parameters. Moreover, B application decreased Cd accumulation in rice and wheat, especially in the roots (Chen et al. 2019;Qin et al. 2020b). Similar results were also found in the present study. Moreover, B pre-supply can improve (nonsignificant) the wheat growth better under Cd stress. In addition, the transport coefficients of Cd transport from root to shoot increased under B+Cd and B/Cd treatment compared with Cd treatment. These results suggest that B can inhibit the absorption of Cd and promote its migration from root to aboveground, where B pretreatment is more effective in inhibiting Cd absorption and promoting Cd migration.
Oxidative stress is inevitably related to Cd stress in plants, and the antioxidant defense machinery protects plants against oxidative stress damages (Gill and Tuteja 2010). For example, malondialdehyde (MDA) represents the degree of lipid peroxidation and reflects the degree of cell damage. Antioxidant enzymes can remove reactive oxygen species (ROS) from plants and reduce the abiotic stress caused by Cd. In our study, MDA concentration significantly increased at Cd-added treatments, while exogenous addition of B inhibited the increase in MDA concentration in plant tissues under Cd stress. This result indicates that B could reduce the accumulation of MDA to alleviate Cd toxicity in wheat. In order to better understand the role of B in antioxidant enzymes, the antioxidant enzyme activity of superoxide dismutase (SOD), peroxidase (POD) and catalase (CAT) was detected. SOD is a key enzyme that constitutes the first line of defense against ROS, which converts superoxide radicals into less toxic agents, producing H 2 O 2 (Gill et al. 2010). CAT and POD are important enzymes that scavenge excess of H 2 O 2 by catalyzing it into water and divalent oxygen for ROS detoxification (Garg and Manchanda 2009). Here, it was found that SOD and POD activity significantly increased at Cd treatment. However, lower SOD and POD activity was found under B+Cd and B/Cd compared with C treatment. In addition, lower activity of CAT was also observed at B application compared with Cd treatment in root. The results suggested that SOD and POD activity has been greatly activated under Cd stress condition, and almost all the antioxidant enzymes moderately reduced after B application.
Some small molecular compounds play an important role in plant transport heavy metals and resistance to heavy metal stress. For instance, Zn hyperaccumulation was associated with enhanced xylem transport and phloem remobilization of Zn by binding to organic acid and citric acid in Sedum alfredii . For Cd, studies show that glutathione (GSH), metallothioneins (MTs) and phytochelatins (PCs) are the most important ligands involved in Cd detoxification in plants (Gallego et al. 2012;Nakamura et al. 2020). In addition, phytosiderophores, organic acids, amino acids, peptides and proteins have also been discovered as important ligands in inducing Cd precipitation by chelation (Bali et al. 2020). Within plant cells, Cd can bind to these small molecule compounds to protect the cytosol from free Cd ions and sequester Cd in the vacuole to increase the tolerance of plants to Cd (Schat et al. 2002;Zimeri et al. 2005;Ernst et al. 2008). In the present study, a total of 198, 680 and 204 of the differential metabolites were isolated between B and C treatment, Cd and C treatment and B+Cd and Cd treatment respectively. B application (B+Cd) not only obviously up-regulated accumulation of galactaric acid, citric acid, N6-galacturonyl-L-lysine, D-glucose, 3,4,5-trihydroxy-6-{[4-methoxy-6-(3-oxoprop-1-en-1-yl)-2H-1,3-benzodioxol-5-yl]oxy}oxane-2-carboxylic acid, N-acetyldjenkolic acid and 2-methylthiophene but also down-regulated the accumulation of threoninyl-tryptophan and C16 sphinganine in Fig. 5 Metabolic pathway enrichment diagram of differential metabolites between B and C treatment (a), Cd and C treatment (b) and B+Cd and B treatment (c). The P value of the metabolic pathway is the significance of the enrichment of the metabolic pathway. The red line indicates a P value of 0.01, and the blue line indicates a P value of 0.05. It is significant for the signal pathway when the top of the bar is higher than the blue line the roots compared with CK+Cd (VIP>5). Furthermore, differential metabolic pathways were mainly concentrated in linoleic acid metabolism, galactose metabolism, sphingolipid metabolism, glycolysis/gluconeogenesis, propanoate metabolism, fructose and mannose metabolism, AGE-RAGE signaling pathway in diabetic complications between B+Cd treatment and B treatment. A previous study has also shown that B nutrition could change the organic carbon structure of cell walls and enhanced the contents of amino acid, cellulose, phenols and lignin in the cell wall (Dong et al. 2016;Wu et al. 2018). In addition, B nutrition could induce changes in pectin composition and architecture of components in root cell walls, which have been shown to play an important role in inhibiting Cd absorption (Wu et al. , 2017a. Therefore, we hypothesize that these differential metabolites, including carbohydrate and amino acid compounds, play a key role in mitigation of Cd toxicity with B application. Nevertheless, other metabolites and metabolic pathways may also be involved in B mitigation of Cd toxicity in wheat, which still needs further verification.
Author contribution SQ and PZ conceived and designed the research. SQ, ZN, CL and YX perform all the experiments, supervised the study, implemented the method and performed data analysis, interpreted the data and wrote the manuscript. WG and HL supervised the study and checked the results. All authors read and approved the manuscript.
Funding This research was funded by the Chinese State Natural Science Foundation (32002128), the Scientific and Technological Key Projects of Henan Province (212102310979,202102110213) and Key Scientific Research Project of the Higher Education Institutions of Henan Province (20A210018). We thank Shanghai Lu Ming Biotech Co., Ltd. (Shanghai, China) for assistance with LC-MS metabonomics analysis.
Data availability All data generated or analyzed during this study are included in this published article (and its supplementary information files).

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication Not applicable.
Competing interests The authors declare no competing interests.