Common wheat variety Chinese Spring (CS) was used as material, and the mature seeds were cultivated based on the previous report (Zhang et al., 2014). In brief, the seedings were cultivated in Hoagland solution. The Cd stress treatment at three-leaf stage was conducted using 50 µM CdCl₂ in Hoagland solution. Three biological replicates were set in both treatment and control, and the samples of roots, stems (crown to ligule) and leaves were respectively collected from 24, 48, 72 h treatments and control, and then immediately immerged into liquid nitrogen prior to use.

The expression profiling of TaSBP genes in different organs including roots, stem axis, leaves was detected using the RNA-seq database of wheat (http://www.wheat-expression.com/genes/heatmap?gene_set=RefSeq1.1&genes=TraesCS3A02G422100%2CTraesCS3D02G417500%2CTraesCS3B02G457600). Total RNA isolation, cDNA synthesis and real-time quantitative polymerase chain reaction (RT-qPCR) were based on the previous report (Yu et al., 2016). The specific primers were shown in Table S3 . The expression levels of TaSBP were presented as values relative to the corresponding control samples at the indicated times and conditions after normalization to Ubqutin (UBI) transcript levels.

The TaSBP-A gene clone, vector (16318) construction and leaf protoplast transformation of Chinese Spring were based on previous report (Zou et al., 2020). The specific primers were shown in Table S3 . Confocal laser scanning microscope (Leica TCS SP5, Germany) was used for monitoring the GFP signal and chlorophyll red auto-fluorescence.

Saccharomyces cerevisiae strain DEY1457 (MATα can1 his3 leu2 trp1 ura3 ade6), kindly provided by Prof. Liping Yin, Capital Normal University, was used for heterologous expression of TaSBP-A protein. The full gene CDS of TaSBP-A was cloned in the destination vector pYES2 (Invitrogen). The specific primers were shown in Table S3 . The empty vectors (EV) and combined vectors were respectively transformed into yeast strains DEY1457 following standard procedure (Invitrogen). Single colonies cultured in the exponential phase (OD 2.0) were diluted to six stepped concentration (OD 2, 2 x 10^-1, 2 x 10^-2, 2 x 10^-3, 2 x 10^-4, 2 x 10^-5) and drop on the dextrose-Ura solid medium with/without 30 μM Cd²⁺. Three independent clones were used for the experiments.

Arabidopsis genetic transformation was based on previous report (Li et al., 2017). pCAMBIA1302 with a 3×Flag-tag was used as expression vector, and at least two generations of resistance screening were carried out. All plants were grown in a growing chamber at 21–22°C with cool-white fluorescent light (80–100 µmol m⁻²s⁻¹) in a long day photoperiod (16 h light/8 h dark). Arabidopsis Seedlings were cultivated for 7 days in half-strength Murashige and Skoog (1/2 MS medium) supplemented with 0.6% (w/v) sucrose and 0.7% (w/v) agar, and then transferred to 1/2 MS medium with 0, 75, 150 μM CdCl₂ for another 7 days in growth chamber. The measurement of seedling root length and fresh weight and statistics of survive rate were performed after 7 days treatment in 1/2 MS medium. At the same time, one-week old Arabidopsis plants were cultivated in garden soil (Basic substrate No. 1, Pindstrup Mosebrug A/S, Denmark) without additional fertilizer for Cd stress treatment. Young plants were grown for three weeks in garden soil, and then used for Cd treatment by watering with and without 150 μM CdCl₂ solution, respectively.

Full length of TaSBP-A CSD was cloned into wheat expression vector, pWMB110, with a HA tag under the control of maize UBI promoter. The new vector was transformed into Agrobacterium tumefaciens strain C58C1 by triparental mating, and then further introduced into W48 immature embryos to generate transgenic plants according to the methods described by previous report (Wang et al., 2017a). The mature seeds from three independent stable transgenic lines at T3 generation were geminated in filter paper soaked with distilled water. After 48 h, uniformly germinated seeds were selected to grow in the half strength Hoagland’s nutrient solution. Cd stress treatment was applied to W48 (non-transgenic control) and TaSBP-A overexpressed wheat seedlings at three-leaf stage with 0 and 50 μM CdCl₂ for two weeks. The measurement of the fresh weight and root length were performed after 2 weeks treatment. The seedling leaves and roots were respectively collected for the measurement of Cd content.

Wild-type and transgenic Arabidopsis plants were treated with 150 μM Cd for 4 weeks. The measurement of chlorophyll and malondialdehyde (MDA) content and the detection of superoxide dismutase (SOD) activity were carried out according to previous report (Li et al., 2017). H₂O₂ content was measured using kit (KGT018, KeyGen Biotech, China) based on the manufacturer’s instructions. Chlorophyll fluorescence was detected by using IMAGING‐PAM chlorophyll fluorometer (Walz, Effeltrich, Germany) as previous report (Schreiber et al., 2007). The wild‐type and transgenic plants were treated by 20 photons m^-2•s^-1 (actinic light) after dark-adaptation. And then maximal PSII quantum yield (F_v/F_m) was measured. The equation was used to calculate F_v/F_m : F_v/F_m = (F_m–F₀)/F_m. The effective PSII quantum yield (ΦPSII) was determined by using the formula: ΦPSII = (F_m′-F)/F_m′. The inhibition of PSII quantum yield (Inh) was detected by using the equation: Inh = (ΦPSII control–ΦPSII sample)/ΦPSII control.

Cd-treated and untreated yeast cells and plant leaves were washed with Cd²⁺ free medium or ddH₂O, and dried for 3 d at 55°C, and then put into digestion tank. Pre-digesting was conducted by adding 5 mL 65% HNO₃ (Suprapur; Merck) and 2 mL H₂O₂ (Suprapur; Merck) to the digestion tank for 40 min at room temperature. The samples were digested by microwave digestion instrument (MARS, CEM Corporation, USA) for 2 h. Cd content (ng/g DW) was calculated using inductively couple mass spectrometry (ICP-MS, ELAN DRC-e, PerkinElmer) according to previous report (Maher et al., 2001).

Cd²⁺ fluxes from one-week-old Arabidopsis seedlings cultivated on 1/2 MS were detected. Net fluxes of Cd²⁺ in root hair were measured by the noninvasive micro-test technique (NMT; BIO-001A, Younger United States Science and Technology Corp, Beijing, China) combined with IFLUXES/IMFLUXES 2.0 software (NMT100 Series, Younger USA, Amherst, MA, USA) (Ma et al., 2015). The microelectrodes were calibrated in 0.1 and 0.01 mM Cd²⁺ before the measurements of Cd²⁺ flux. The electrodes with Nernstian slopes were > 29 ± 3 mv/decade. Arabidopsis seedlings were transferred into a measuring chamber containing 10 mL of measuring solution (0.1 mM KCl, 0.03 mM CdCl₂, and 0.3 mM MES, pH 5.8.) and equilibrated for 10 min before measurement.

Visualization of free Cd²⁺ in Arabidopsis roots was conducted in one-week-old seedlings. The Cd probe Leadmium™ Green AM dye (Molecular Probes, Invitrogen, Calsbad, CA, USA) was utilized to detect the distribution of Cd in plant roots pre-treated with 150 μM Cd²⁺ for 0, 6 and 12 h. Cd fluorescence was excited at 488 nm and visualized using Zeiss LSM 780 (Carl Zeiss, Germany).

The mutant (ΔTaSBP-A) with displaced Cys with Gly in the putative metal binding regions was constructed. cDNA in the entry clone was cloned into the destination vectors pEGX-4T-1, used to produce GST-TaSBP-A/ΔTaSBP protein carrying additional 250 amino acids at the N terminus compared to the recombinant proteins. The recombinant plasmids were transformed into E. coli strain BL21. The cell cultures were fostered at 37°C for 3-4 h until the OD = 0.6-0.8, and then 1 mM IPTG was added for induction 16 h at 16°C. The recombinant proteins were isolated by batch purification with glutathione sepharose 4B according to the manufacturer’s instructions (Amersham). Protein concentrations were determined by spectrophotometer (NanoDrop 2000, Thermoscientific) at 280 nm.

Two constructs wild type with the normal CXXC motif (TaSBP-A) and mutant with the GXXG motif (ΔTaSBP-A) were prepared. The recombinant protein (100 μM) of TaSBP-A/ΔTaSBP was incubated with 500 μM CdCl₂ and 500 μM CdCl₂ together with 500 μM EDTA at 4°C overnight. The shift was tested by using 10% SDS-PAGE. The recombinant plasmid of pEGX-TaSBP-A/ΔTaSBP was transformed into E. coli BL21 (DE3) and induced at 16°C for 10 h with 1 mM IPTG. Then, 100 μM CdCl₂ was added to the liquid medium for 10 h culture. The cells were collected and mildly lysed using BugBuster Master Mix (Novagen, 71456-4). The lysis was centrifuged at 16000 g and 4°C for 15 min. The purified proteins were divided into equal two parts: one used for measuring Cd content by ICP-MS, and the other for measuring the protein concentration by DC protein determination kit (Bio-Rad). The stoichiometry of protein binding to Cd was conducted according to Cd and protein concentration (Luo et al., 2019).

Cd-TaSBP-A/ΔTaSBP thermodynamic parameters were detected by calorimetric experiments of the recombinant TaSBP-A/ΔTaSBP proteins. Calorimetric titrations were conducted at 25°C with stirring at 1000 rpm with a filter time constant of 2 s by using a microcalorimeter (Microcal ITC 200 System, GE Healthcare). Negative controls were performed by the injections into the buffer, resulting only in signals from heat of salt dilution.

Secondary structure characterization (modifications) of the recombinant TaSBP-A protein after incubation with Cd²⁺ was performed by circular dichroism (CD). The recombinant protein of TaSBP-A/ΔTaSBP (1 nmol) was incubated with gradient concentration of Cd²⁺ (0, 2, 3, 4 nmol) in 250 μL incubation buffer. Spectra acquisition at 25°C used a spectropolarimeter (J-815, Jasco) at the far UV (200–260 nm). The following parameters were set: 1 nm step, 2 nm band width, and scan speed 200 nm/min with the optical path length 1 mm. The assessment of protein second structure was performed by K2D method based on the linear regression method (Yang et al., 1986).

The wheat protoplasts were cultivated overnight after transformation of plasmids including 16318hGFP, reconstructive TaSBP-A and ΔTaSBP-A, respectively. Then, the viable protoplasts were treated with 50 μM CdCl₂ for 0, 2, 4, 6, 8, 10, and 12 h. To determine cell viability, the GFP fluorescence and chloroplast autofluorescence were observed by confocal laser scanning microscope (Leica TCS SP5, Germany). The transgenic protoplasts were cultivated in 0 μM CdCl₂ as the control (CK). The relative rate of protoplasts viability (treatment group/CK) was counted from 0 to 12 h.

Three protein sequences (TraesCS3A01G422100.1, TraesCS3B01G457600.1 and TraesCS3D01G417500.1) in Chinese Spring Protein database (http://plants.ensembl.org/index.html) were homologous to Arabidopsis SBPs (AtSBP1/2/3); in turn, they respectively located on chromosomes 3A, 3B and 3D. Specifically, TraesCS3A01G422100.1 showed 98.59% and 97.37% identities to TraesCS3B01G457600.1 and TraesCS3D01G417500.1, respectively ( Figure S1 ). This demonstrates that hexaploid wheat had only three SBP copies, and they are named TaSBP-A, TaSBP-B and TaSBP-D. A phylogenetic tree constructed by 11 SBPs from different animal and plant species showed that TaSBP-A/B/D had a close phylogenetic relationship with the SBP56 family ( Figure 1A ).

Structural characterization showed that TaSBPs belonged to the SBP56 family and contained a putative heavy metal binding motif as well as a selenium-binding site CXXC ( Figure 1B ). They also showed a partial similarity with other typical metal binding proteins in the metal binding region CXXC ( Figure 1C ). A three-dimensional model analysis of TaSBP-A was performed by using SWISS-MODEL (https://swissmodel.expasy.org/) to identify the potential Cys residues involved in Cd binding. As the closest homologue, the structure of the hypothetical selenium-binding protein from Sulfolobus tokodaii (SMTL ID: 2ece.1) was used to generate a three-dimensional model with the program Modeler (Global Model Quality Estimation of 0.7; 43.91% sequence identity). As shown in Figure 1D , nine potential Cys residues were located on the surface of the TaSBP-A protein, among which four Cys residues, Cys¹⁰⁰, Cys¹⁰², Cys¹⁶¹, and Cys⁴⁸⁸, showed high conservation. Cys²³ and Cys²⁴ were conserved in all the photosynthetic organisms, which were identified as the selenium-binding sites in Arabidopsis (Schild et al., 2014). The Cys¹⁰⁰ and Cys¹⁰² on the random coil belonged to a CXXC motif related to metal binding (Lin and Culotta, 1995; Yuan et al., 1995; Culotta et al., 1997).

RNA-seq analysis of the three TaSBPs in different organs of wheat showed that all TaSBPs expressed in root, stem and leaf. Of these, TaSBP-A had the highest expression level in three organs, particularly in the plant root ( Figure 1E ). Further RT-qPCR analysis displayed a similar expression pattern in different plant organs ( Figure 1F ). According to the above results, we chose TaSBP-A for further functional survey. TaSBP-A consisted of 1485 bp encoding 495 amino acid residues, and the deduced molecular mass was 54.1 kDa without considering protein modification.

To determine the subcellular location of the TaSBP-A, a TaSBP-A-GFP fusion vector (pTaSBP-A-GFP) driven by two CaMV35S promoter was constructed and transformed into wheat protoplasts. In cells only expressing GFP, the entire cytoplasm and nucleus were diffusely labeled by a laser scanning microscopy. In contrast, the GFP fluorescence of TaSBP-A-GFP was only visible at cytoplasm, which clearly indicates that the TaSBP-A was localized in the cytoplasm ( Figure 1G ).

The TaSBP-A expression in response to Cd stress in the seedling roots of Chinese Spring was detected by both RT-qPCR ( Figure 2A ) and Western blot ( Figure 2B ). The results showed that both transcription and translation expression levels of TaSBP-A were significantly induced by time increasing under Cd stress ( Figures 2A, B ). In particular, both transcription and translation of TaSBP-A arrived at the highest level at 72h compared with control (24h), suggesting that TaSBP-A can be induced and accumulated in roots by Cd stress and may have potential roles in Cd stress response.

Because yeast has no SBP homolog, we overexpressed the TaSBP-A in yeast to detect the Cd tolerance of yeast cells. As shown in Figure 2C , EV and TaSBP-A yeast cells had a practically equal growth rates under normal condition; however, the TaSBP-A cells showed a higher growth rate than did the EV cells under 30 μM Cd²⁺ treatment ( Figure 2C ). Subsequently, we respectively cultured EV and TaSBP-A cells in a liquid medium containing 25 μM CdCl₂; here, the Cd content in yeast was detected by ICP-MS. Interestingly, the TaSBP-A overexpression in yeast could accumulate more Cd than EV significantly, as shown to be by 10.8% ( Figure 2D ). These results indicated that TaSBP-A can enhance the Cd tolerance of yeast cells by binding toxic free Cd ions.

To reveal TaSBP-A function in the response to Cd stress, we further overexpressed it in Arabidopsis. Three stable overexpressed transgenic lines (S2-9, S3-12 and S4-7) were generated using genome PCR and western blot detection ( Figures S2A-C ). The overexpression protein level of TaSBP-A is S4-7 < S2-9 <S3-12 ( Figure S2C ). The seedlings of the TaSBP-A transgenic lines and wild type (WT) were cultivated on 1/2 MS containing 0, 75, and 150 μM CdCl₂ for 7 days. The primary root growth was significantly inhibited and substantial chlorosis of cotyledons occurred in the WT plants ( Figures S2D, E ). However, TaSBP-A overexpressed seedlings maintained a higher fresh weight, root length and survival rate ( Figures S2F-H ), showing that all of lines have a palpable Cd tolerance. The Cd accumulation measurement revealed that the overexpression of TaSBP-A lines could accumulate a greater amount of Cd, specifically 1.13-1.23 times as high as WT ( Figure S2I ).

The Cd tolerance of TaSBP-A overexpressed lines grown in garden soil was further detected under 150 μM CdCl₂ treatment. Both transgenic lines and WT had no clear differences in the growth under normal conditions. However, under Cd stress, the WT plants showed chlorosis leaves and a severe reduction of growth ( Figure 3A ). In contrast, the TaSBP-A transgenic plants displayed a clear resistance to Cd stress even though some chlorosis leaves still occurred. Physiological and biochemical parameter analyses showed that SOD activity in TaSBP-A overexpressed lines was 2.36 times and the content of H₂O₂ and MDA were only 50% and 39% as compared to WT plants ( Figures 3B-D ). These results demonstrated that TaSBP-A transgenic plants could effectively promote SOD activity that would be able to alleviate the oxidative stress triggered by heavy-metal-induced H₂O₂. Cd accumulation in the roots of ICP-MS showed that the TaSBP-A transgenic lines accumulated more Cd, 1.23-1.64 times higher than WT ( Figure 3E ).

We also detected the impairment of photosynthesis as a result of Cd stress in transgenic plants. The results showed that the chlorophyll fluorescence parameters had no clear differences between WT and TaSBP-A overexpressed lines under normal conditions ( Figure 3F ). However, the maximum quantum yield of PSII photochemistry in the dark‐adapted state (F_v/F_m) in the transgenic plant leaves under 150 μM Cd was higher than it was for WT ( Figure 3G ). In particular, the operating efficiency of PSII (ΦPSII) in three overexpression lines under Cd treatments was 22.25-28.80%, which was significantly higher than it was for WT ( Figures 3F, H ). Cd exposure significantly promoted the inhibition of PSII quantum yield (Inh) in the WT plant leaves, but it had no clear effects on transgenic plants ( Figures 3F, I ). Similarly, the chlorophyll content in the leaves had no significant differences between WT and the TaSBP-A transgenic lines after 4 weeks in normal growth conditions, while the TaSBP-A transgenic lines were significantly higher chlorophyll content under Cd stress: 1.95-2.95 times as high as WT ( Figure 3J ). These results indicate that the overexpression of TaSBP-A could significantly alleviated the photosynthesis impairment of plants under Cd stress treatment.

As described above, TaSBP-A overexpression resulted in an increase of Cd accumulation in roots, thus we detected the net fluxes of Cd²⁺ in the root hairs of transgenic Arabidopsis under 30 μM Cd²⁺ using a noninvasive micro-test technique ( Figure 4A ). The results showed that the Cd²⁺ net fluxes in the root hairs of three transgenic lines S2-9, S3-12, and S4-7 were 2.49, 4.11 and 3.20 pmol cm^-2·s^-1 respectively, which is 1.8-2.9 times higher than that of WT ( Figure 4B ). This confirmed that the overexpression of TaSBP-A could enhance the Cd²⁺ accumulation in roots of plant.

Leadmium™ Green AM dye, a Cd probe, was used to detect the state (free or binding) of the accumulated Cd, particularly with regard to the distribution of free Cd²⁺ in plant roots. As shown in Figure 4C , a very low level of green fluorescence was found in the roots of transgenic lines. This indicates that this dye has a high specificity to detect Cd²⁺ and that it does not react with divalent ions such as the Ca²⁺ present in the control roots. In contrast, the green fluorescence signals in WT plants appeared clearly above the hypocotyl after a 6 h pretreatment of 150 μM Cd²⁺; they were more obviously enhanced after a 12 h pretreatment of 150 μM Cd in WT plants. However, the slight green fluorescence signals in the overexpression lines were only present below hypocotyl. These results indicated that overexpressed TaSBP-A played an important role in reducing free Cd²⁺ and that the over-accumulated Cd in the TaSBP-A overexpression lines was present largely in binding state.

Three stable TaSBP-A overexpressed wheat lines, OVEREXPRESSION-1 (OE-1), OE-2 and OE-3, were obtained at T3 generation. The RT-qPCR and Western blot analysis showed both the transcription and translation expression levels of TaSBP-A were significantly higher in TaSBP-A overexpressed wheat seedlings than those in W48 seedlings ( Figures 5B, C ). The growth of W48 and TaSBP-A overexpressed wheat seedlings exposed to 0 and 50 μM CdCl₂ was compared after two weeks. Under normal condition, W48 and TaSBP-A overexpressed wheat seedlings showed a similar morphological characteristics ( Figure 5A ), and significant differences were not found in the fresh weight and root length between W48 and TaSBP-A overexpressed lines ( Figure 5D, E ). When subjected to 50 μM CdCl₂ treatment, a repressed growth was observed in both W48 and TaSBP-A overexpressed wheat seedlings, but W48 seedlings were more sensitive to Cd treatment ( Figure 5A ). The statistical analysis showed that both the fresh weight and root length of TaSBP-A overexpressed seedlings were higher than those in W48 seedlings ( Figure 5D, E ), indicating that TaSBP-A overexpressed wheat seedlings have a higher tolerance to Cd stress. In addition, the measurement of Cd content showed that a significantly higher amount of Cd was accumulated in the roots of TaSBP-A overexpressed wheat seedlings compared with W48 ( Figure 5F ). However, a lower amount of Cd was accumulated in the leaf of TaSBP-A overexpressed wheat seedlings ( Figure 5G ). These results indicated that overexpressed TaSBP-A could improve Cd tolerance of wheat seedlings by inhibiting the transfer of Cd from the root to leaf in wheat seedlings.

The above results indicated that TaSBP-A can enhance Cd tolerance of plants via reducing free Cd²⁺ and inhibiting the transfer of Cd from root to leaf. Thus, it is crucial to determine the Cd binding site in TaSBP-A in order to dissect the molecular mechanism of plant Cd-tolerance. The putative heavy metal binding motif CXXC (TaSBP-A) was mutated to GXXG (ΔTaSBP-A) and then the TaSBP-A and ΔTaSBP-A were respectively expressed in E. coli as a fusion protein with GST ( Figure S3A ). Subcellular localization indicated that the mutant of CXXC had no influence on the location of TaSBP-A ( Figure S3B ). Thus, we further conducted an in vitro Cd²⁺ binding assay according to the principle that ethylenediaminetetraacetic acid (EDTA) can chelate with Cd²⁺ to form Cd (II)-EDTA (Yang and Davis, 1999). As shown in Figure 6A , both recombinant TaSBP-A and ΔTaSBP-A had shifts in SDS-PAGE after incubation with 500 μM CdCl₂; in contrast, the shift disappeared with EDTA and the Cd coexistence in incubation buffer. However, after binding Cd²⁺, it was hard to recognize the shift between TaSBP-A and ΔTaSBP-A. ICP-MS combined with protein quantification was further used to identify the stoichiometry of the interaction between Cd and protein, and the results showed that the stoichiometry of the TaSBP-A:ΔTaSBP-A binding to Cd was about 3: 2 (2.98: 1.59) ( Figure 6B ). This indicates that the ability of TaSBP-A decreased by 47.3% after the CXXC motif mutation. This result further witnessed the interaction between TaSBP-A with Cd, where the CXXC motif served as the heavy metal binding site in the TaSBP-A.

The interaction of Cd²⁺ with TaSBP-A and ΔTaSBP-A was detected by ITC. The ITC curves and thermodynamic parameters during the interaction of Cd²⁺ and TaSBP-A/ΔTaSBP-A are shown in Figure 6C and Table 1 . For each Cd²⁺ injection, a release of heat was observed ( Figure 6C , upper panel); this indicates that a clear binding event between Cd²⁺ and TaSBP-A/ΔTaSBP-A occurred. A one-site binding model was used to fit TaSBP-A/ΔTaSBP-A isotherms ( Figure 6C , lower panel). This can determine an apparent binding affinity constant (K) in the low micromolar range (50 μM). Unexpectedly, the binding enthalpy value of the Cd²⁺ and TaSBP-A interaction was -17.92 kcal/mol, which was significantly higher than that of the Cd²⁺ and ΔTaSBP-A interaction (-13.30 kcal/mol) ( Table 1 ). Therefore, we deduced that the two different strong negative enthalpy values (ΔH) for Cd²⁺ most likely correspond to a combination of a binding event with a specific covalent reaction between TaSBP-A/ΔTaSBP-A and Cd²⁺. However, less heat was released in the interaction of ΔTaSBP-A with Cd²⁺, which suggests that the Cd binding amount of TaSBP-A was palpably diminished after the mutation of the CXXC motif. The CXXC motif also witnessed entropy difference values (ΔS), a measurement for disorder and order in atomic and molecular assemblies (Landsberg, 1984). According to ΔS_TaSBP-A < ΔS_ΔTaSBP-A, TaSBP-A was able to bind more Cd²⁺ than that of ΔTaSBP-A, leading to a dramatic decrease in ΔS. Furthermore, the calorimetric signals needed abnormally large amounts of time to recover to their baseline values after the CXXC motif mutation ( Figure 6C ) suggest that a covalent reaction combined with the binding of the ion was significant restrained due to the CXXC motif mutation in TaSBP-A. According to the K value, K_TaSBP-A was 1.6 times as high as K_ΔTaSBP-A ( Table 1 ), which is consistent with the results of the stoichiometry TaSBP-A/ΔTaSBP-A ( Figure 6B ). This indicates that the ability of TaSBP-A to bind to Cd was heavily suppressed after the loss of CXXC motif, which can reduce the affinity of other binding sites in TaSBP-A to Cd²⁺.

CD analysis was used to further determine whether Cd-binding caused changes of the protein’s secondary structure. The CD spectrum recorded the far-UV with TaSBP-A, ΔTaSBP-A, Cd-bound TaSBP-A and Cd-bound ΔTaSBP-A ( Figure 6D ) to estimate the second structure of proteins and Cd-proteins complexes. According to Table S1 , the mutation of the CXXC motif had no effects on the β-sheet (about 45%) and turn (about 7.4%) content, but it resulted in an increase in random coil and a decrease in α-helix. With the presence of Cd²⁺, the spectrum of TaSBP-A had more significant changes than did ΔTaSBP-A, indicating that the Cd-binding to TaSBP-A and the complex formation led to clear second structure alterations ( Figure 6D ). According to Table S1 , the changes of TaSBP-A were obviously observed in the α-helix, β-sheet, turn and random coil content after binding to Cd²⁺; in contrast, ΔTaSBP-A had only changes in turn content. These results showed that the significant changes of TaSBP-A’s second structure occurred upon Cd²⁺ binding. Moreover, the CXXC motif was a crucial Cd²⁺-binding site which could cause significant alterations of the protein’s second structure as a result of interaction between TaSBP-A and Cd²⁺.

We transformed TaSBP-A and ΔTaSBP-A to wheat protoplasts for further Cd-binding CXXC motif and function verification of TaSBP ( Figure S4 ). Empty vector 16318hGFP and recombinant plasmids containing TaSBP-A and ΔTaSBP-A were transformed into wheat leaves protoplasts, and then they were cultured with protoplasts culture medium with 50 μM Cd²⁺ ( Figure S4A ). The viable protoplasts (yellow light overlapped by chloroplast autofluorescence and GFP green fluorescence) that occurred after transformation were counted ( Figure S4B ). The transgenic efficiency of the empty vector was 27%, and it remained stable during the whole experiment under normal condition (0 μM Cd²⁺). In turn, TaSBP-A/ΔTaSBP-A had a transform efficient of approximately 20%, serving as the basis to explore the relative rate of the protoplasts’ viability (RRPV) under Cd treatment ( Figure S4A ). Compared to TaSBP-A (73%) and ΔTaSBP-A (56%), the number of viable protoplasts dramatically dropped after 12 h of Cd treatment in the empty vector, during which RRPV declined to 25% ( Figure S4B ). The dynamic RRPV was plotted from 0-12 h under cadmium treatment and it was also linearized ( Figure 4B , Table S2 ). The results showed an order of Slope_EV < Slope_ΔTaSBP-A < Slope_TaSBP-A (-0.06306< -0.04225< -0.02226) ( Table S2 ), which indicates that the death rate of protoplasts was EV> ΔTaSBP-A> TaSBP-A. Compared to TaSBP-A, the dynamic RRPV of ΔTaSBP-A transgenic cells during 0-4 h had no significant difference. However, after 4 h under Cd stress, the reduction of ΔTaSBP-A RRPV was more palpable than that of TaSBP-A ( Figure S4 ). These results revealed that the CXXC motif played a key role in TaSBP tolerance to Cd stress, whose mutation can heavily inhibit the ability of TaSBP to interact with Cd²⁺.