Two wheat varieties (ZhongMai175 and Xiaobaimai) and one rice variety (Oryza sativa L. japonica. cv. Nipponbare) were utilized in this research. ZhongMai175 is a high Zn wheat variety that allowing for analysis of the expression of TaZIP genes in wheat (He et al., 2015). This line was planted at the experimental station of Northwest A&F University, Yangling, China (34°20'N, 108°24'E). Each row was 1m wide with four duplicates, while the row spacing was 0.25m and the plant spacing in each row was 0.05m. At the 7days after flowering stage (7 DAF), we collected grain sample from one plant in each row which was mixed with liquid nitrogen. One week later, the second sample was obtained (14 DAF). Four samples were collected in this study; there are 7, 14, 21, and 28 DAF.

Xiaobaimai is a landrace, drought-tolerant wheat variety found in PingYao city (ShanXi Province, China). This variety contained low zinc and iron in grain. Culturing in a glass petri dish with two layers of filter paper containing 1/4 Hoagland solution and place at climate chambers (RXZ-500D-LED, Ning Bo) at a light/dark cycle of 16/8h at 24°C for 10days (Seedling two leaf). Then for analyzed the expression under Fe and Zn stress, wheat treated with Hoagland medium containing different concentrations of ZnSO₄ and FeCl₃ solutions for 1h. In this study, the concentration of ZnSO₄ and FeCl₃ solutions was 0.05, 0.5, 50μmol/L. For investigated the expression under Zn- and Fe- deficient conditions, we treated it with Hoagland medium lacking ZnSO₄ (Zn-deficient), Fe (III)-EDTA (Fe-deficient).

The sequence of the wheat proteins was downloaded from the Ensembl Plant database,¹ after which the HMM profile for the ZIP DNA-binding domain (PF02535) was downloaded from the Pfam v31.0 database² to search against the plant protein sequences using a threshold of E<1e⁻⁵ (Finn et al., 2016). Blast and manual corrections were then performed to remove alternative events and redundancy. The OsZIP genes were downloaded from the NCBI database,³ according to the methods used by Chen and Tiong (Chen et al., 2008; Tiong et al., 2014). The NJ phylogenetic tree was constructed with MEGA 7 and EvolView⁴ based on the wheat and rice protein sequences, with 1,000 bootstrap replicates. Putative TMHs of TaZIPs were predicted using the TMHMM Server v.2.0 (Krogh et al., 2001). Subcellular location was predicted by the WoLF PSORT.⁵

Gene structure was analyzed by GSDS.⁶ Protein conserved motifs were predicted using the MEME Suite web server,⁷ with the number of motifs set to 10, at a width range from 5 to 200 amino acids.

The miRNA target to the TaZIP genes was searched using the psRNATarget tool (Dai and Zhao, 2011), and the TaZIP cascade transcript was submitted in the miRBase. The cytoscape tool⁸ was used to visualize the regulatory network of the-miRNA and TaZIP genes.

For wheat tissue expression pattern analysis, the expression pattern data were downloaded from the RNA-seq database.⁹ For the expression pattern at grain filling stage and under Zn and Fe stress assay, RT-qPCR was used. Total RNA was isolated from the wheat grains and wheat, rice leaves using an RNAprep Pure Plant Kit (Tiangen, Beijing, China) and from the seedlings with TRIZOL (Takara, Dalian, China). cDNA synthesis was performed in a 20μl reaction mixture containing 1μg of total RNA and a mixture of TIANscript RT Kit (Tiangen, Beijing, China). The real-time PCR mixture contained 1μl cDNA, 1μl forward and reverse primers, and 17μl SYBR Green (Tiangen, Beijing, China). Real-time qPCR was performed in an ABI7300 (Thermo Fisher Scientific, United States) Real-TimeThermal Cycler and repeated three times. The actin of the wheat genes (Gene ID: AB181991) was used as a control. The 2^−∆∆Ct method was used for fluorescence quantitative data analysis (Livak and Schmittgen, 2001).

The miRNAs of the wheat seedling under ZnSO₄ and FeCl₃ stresses were extracted using a miRcute miRNA Isolation Kit (Tiangen KR211, Beijing, China). miRNA-cDNA synthesis was performed with miRcute and miRNA First-Strand cDNA Kit (Tiangen KR211, Beijing, China). The real-time reaction mixture was performed with miRcute Plus miRNA qPCR Kit (SYBR Green; Tiangen FP411, Beijing, China) with three biological replicates. The qPCR reaction conditions were 95°C for 15min, followed by 45cycles of 94°C for 20s, and 58–60°C for 34s. Data of miRNA-qPCR were analyzed using the 2^−∆∆Ct method.

The CDS and ORF sequences were obtained from the Wheat Sequence database.¹⁰ The primers were designed with Oligo 7, with three genes cloned. In this step, the RNA of Xiaobaimai and Oryza sativa L. seedling was used for cDNA synthesis. The PCR reaction solution was 50μl and contained the following: 5μl cDNA as an amplification templet, 2.5μl forward and reverse primers, 25μl 2×master Mix, and 15μl Nuclease free water (NEB, United States). The reaction solution was performed on a DNA amplification machine (Thermo Fisher Scientific, United States). PCR amplification procedures were as follows: initial denaturation at 98°C for 30s, followed by 35cycles of denaturation at 98°C for 10s, annealing at 60°C for 20s, extension at 72°C for 30s, while the final extension was at 72°C for 2min. After amplification, we added 7μl purple 2-Log Ladder (NEB, United States) to the PCR products and separated them on 1.5% agarose gel for 30min at 120V. After they were separated, the PCR products were purified using a Universal DNA Purification Kit (Tiangen, Beijing, China) and connected to the cloning vector pLB (Tiangen, Beijing, China), and sequenced.

Specific primers were designed for PCR amplification and constructing the expression vector. The PCR procedure is the same as above, except for the annealing, which took place at 70°C for 20s. The PCR products were spread on agarose gel and linked to the BamH I site of the yeast expression vector pDR195 (PLASMID, China). They were subsequently sequenced and transformed into yeast competent cells. The yeast competent cells were prepared according to the methods used by Gietz and Schiestl (1995). Three yeast strains were used in this experiment: DY1455 (MATa ade6 can1 his3 leu2 trp1 ura3), fet3fet4 DEY1453 (MATa/MATa ade2/+ can1/can1 his3/his3 leu2/leu2 trp1/trp1 ura3/ura3 fet3-2::HIS3/fet3-2::HIS3 fet4-1::LEU2/fet4-1::LEU2), and zrt1zrt2 ZHY3 (MATa ade6 can1 his3 leu2 trp1 ura3 zrt1::LEU2 zrt2::HIS3; Li et al., 2013). The pDR195-TaZIPs were converted to DEY1453 and ZHY3 with the lithium acetate conversion method used by Gietz and Schiestl (1995). To verity the experiment was performed properly, OsIRT1, OsZIP3, and OsZIP5 as well as converted to yeast competent cells as positive controls, the wild-type strain DY1455 harboring pDR195 was also used as a positive control. The empty vector pDR195 was used as a negative control converted to two yeast mutants. Transformed cells were coated on the selective SD-URA solid medium without corresponding amino acids. To verify the gene function, we diluted the yeast liquid OD₆₀₀ to 1, 0.1, 0.01, 0.001, and dropped 10μl yeast liquid onto a different medium. The yeast strain of zrt1zrt2 ZHY3 was grown on an SD/–ura medium (pH 4.4) supplemented with 0.4mM EDTA or 300μM ZnSO₄. The yeast strain of fet3fet4 DEY1453 was grown on SD/–ura medium (pH 5.8) containing 50mM 2-(4-Morpholino) ethanesulfonic acid (MES) supplemented or 200μM FeCl₃.

Specific primers were used for vector construction, the expression vector PCAMBIA1302 and the Nco I site were used for gene construction. After sequencing, the overexpression plasmid of PCAMBIA1302-TaZIP13 was transferred into the GV3101 (Agrobacterium tumefaciens) strain and transformed into Arabidopsis. The transgenic lines were cultured in a light temperature incubator until they were propagated for the third generation. The homozygous plants of the T3 progeny and WT were used for further study.

For the germination assays, the seeds of wild type and transgenic lines were surface sterilized and kept at 4°C for 72h in the dark before germination. About 25 seeds of every genotype were sown on the same plate containing different concentration of FeCl₃ and ZnSO₄ solution MS medium at 22/20°C (day/night) with a photoperiod of 16/8h (day/night) for 7days. Each day germinated seeds with protruded radicles were counted. The concentration of FeCl₃ and ZnSO₄ solution was 0, 50, 200, 300μmol/L. After germinated seeding were counted, four Arabidopsis lines were cultured until they were at six leaf stage, then the root length of 30 seedings from each line was measured and photographed. On the other hand, four lines from MS medium were transplanted into soil and treated with different concentrations of FeCl₃ and ZnSO₄ to analyze the tolerance. In this study, the concentration of ZnSO₄ and FeCl₃ was 200, 300, 400μmol/L. After 2weeks, a 0.15g of each sample leaf was collected and determine the chlorophyll content according to the methods used by Richardson et al. (2010). And three leaves of each line were collected and taken the midsection epidermis of the leaves to observe the stomas.

Four lines were treated with water until maturity. Then seeds from the siliques located in the basis of a major inflorescence were selected for observation. Seeds from WT and transgenic lines were randomly selected, then photographed using stereomicroscope (Olympus mzx7) and a test instrument (WAN SHENG, HiCC-A). For the Zn and Fe content assays, 0.15g plant shoots, roots, and seeds of Arabidopsis were collected and digested in 2ml HNO₃ overnight, then 2ml H₂O₂ was added and completed the digestion by microwave treatment, after which the digests were diluted with Millipore-purified water and filtered. The volume was then adjusted to 25ml, was ICP-OES analyses using an ICAP 6000 Series spectrometer (Thermo-Fisher; Hansen et al., 2013). For metal content measurements, three samples were conducted for each line. Each sample (shoots, roots, and seeds, respectively) was a mix of 15 plants.

The length of roots, length and width of seeds were counted by ImageJ software (Rueden et al., 2017). Data were analyzed and graphs were drawn using Excel 2019 (Microsoft Corporation, United States). In all graphs, error bars indicate standard deviation, and significant differences are indicated with ^*p<0.05 or ^**p<0.01.

A total of 58 ZIP genes were identified using a whole-genome search (Figure 1; Supplementary Table S1; Supplementary Figures S1, S2), of which 44 wheat ZIP genes were found to share an orthologous relationship with rice. These were named according to rice-related nomenclature (Figure 1). The remaining genes were named from TaZIP17 to TaZIP30, based on their location on the chromosome, from 1A to 7D (Supplementary Figure S2). Our results demonstrated that the TaZIPs were unevenly distributed on the chromosomes, and no ZIP genes located on the fifth chromosome group. The length of the TaZIP amino acids ranged from 185 to 577 and contained between 3 and 13 transmembrane domains. Most contained between 7 and 9 TM, while the length between TM-3 and TM-4 varied. Subcellular localization of the TaZIP genes was found to be on the plasma membrane and nucleus (Supplementary Table S1).

We constructed the phylogenetic relationship of the wheat ZIPs with rice, maize, and Arabidopsis ZIP proteins (Figure 1 and Supplementary Tables S1 and S2). The results revealed that these ZIP proteins were classified into three groups. Group ZIPI and group ZIPIII included all species proteins, suggesting that these ZIP proteins have a conserved function in monocots and dicots. Only three monocot species were included in group ZIPII, including Arabidopsis, suggesting that these proteins are unique to monocot. The wheat proteins are consistent with those of rice and maize, except for ZmZIP5 and ZmZIP1. Six wheat ZIP genes showed a close relationship with OsIRT1, OsIRT2, AtIRT1, and AtIRT2, suggesting that these six TaZIPs shared a similar function with OsIRT1, OsIRT2, AtIRT1, and AtIRT2, all of which are involved in Fe transport in wheat. The wheat proteins are evenly distributed on three branches.

To comprehensively understand the function of TaZIPs, we analyzed their gene structure and conserved motifs (Figure 2). ZIP size ranged from 836 to 14,494bp (Supplementary Table S1). Of these, the TaZIP28 gene was the shortest and the TaZIP27 gene was the longest. The number of introns varied from 0 to 11, and the number of exons ranged from 1 to 12. TaZIP27 had the largest size, TaZIP16-A, B, D and TaZIP13-A, B, D had the most exons (up to 12). In addition, TaZIP28 has only one exon. Genes sharing a closer phylogenetic relationship had a more similar gene structure.

Using the MEME tool, 10 conserved motifs were identified in the wheat ZIPs (Figure 2 and Supplementary Table S3). The conserved motifs of the same group were similarly organized. Almost all TaZIP proteins possess motif 3 because it contains a histidine residue that binds to metal ions for transmembrane transport. All TaZIP proteins in group ZIPI had motif 1 through motif 7, except for four truncated genes that lacked some of these motifs (TaZIP5-A, TaIRT1-D, TaZIP27, and TaZIP28). In group ZIPII, three TaZIP proteins (TaZIP14-A, B, D) only had motif 3, while other members have the same motif. In group ZIPIII, motif 8 was detected in TaZIP11-A, B, D, and motif 5 was only detected in TaZIP16-A, B, D. Group ZIPI had the most motifs, while group ZIPII and ZIPIII had different quantities of motifs.

The putative miRNA-targeted TaZIP genes were analyzed to assess the network of miRNA and TaZIP genes. Our results demonstrated that 20 miRNAs were predicted to target 30 TaZIP genes, while 28 TaZIP genes were not targeted by miRNA. This could be due to the current limitations on wheat miRNA (Supplementary Table S4). Based on the target relationship, 57 miRNA-TaZIP interactions were constructed (Figure 3). The wheat ZIP genes were inhibited by miRNA via translation (57.89%), while the rest of the genes were inhibited via cleavage (42.11%). Additionally, the miRNAs are primarily targeted in the CDS region but ahead of the ZIP domain of the TaZIP genes, silencing gene expression.

Furthermore, we constructed the co-expression regulatory network to detect the interaction between the TaZIP genes and miRNAs using a dataset of 173 RNA-seq, based on the weighted correlation of their expression.¹¹ tae-miR164 and tae-miR5384-3p had major target genes, 10 and 6, respectively, while the TaZIP25 and TaZIP27 genes were targeted by the greatest number of tae-miRNAs (five tae-miRNAs; Figure 3 and Supplementary Table S4). Other genes targeted by a major number of tae-miRNAs were TaZIP20 (targeted by four tae-miRNAs), TaIRT2-D, TaZIP13-A, and TaZIP14-D (targeted by three tae-miRNAs). Three genes of TaIRT2-A, TaZIP14-B, and TaZIP13-B were targeted by two tae-miRNAs. Figure 3 demonstrates that miRNA164 targeted TaIRT2-A, miRNA9660-5p targeted TaZIP13-A, and tae-miR5084 targeted TaZIP14-B.

Overexpression of some ZIP genes may enhance zinc and iron content, thus enhancing grain and fruit quality. RNA-Seq data were downloaded for analyzed the tissue expression pattern. To investigate the expression pattern of the TaZIP genes in wheat grain, we designed 39 primers for the fluorescence quantification of wheat ZIP genes (Supplementary Table S5). In total, 39 ZIP genes are expressed in wheat four tissues. Most were highly expressed in the roots but less in the grain (Figure 4A and Supplementary Table S6). Four genes (TaZIP4-B, TaZIP4-A, TaZIP6-D, and TaZIP14-B) were highly expressed in the stem and three genes (TaZIP19, TaZIP6-A, and TaZIP2-A) were highly expressed in the leaves. Except for TaZIP16-A and TaZIP16-B, group ZIPIII was highly expressed in four tissues, particularly in the grain (Figure 4). There were also differential expressions pattern between homoeologs genes: TaIRT2-A and TaIRT2-B displayed high expression in wheat grain, while TaIRT1-D displayed low expression in wheat grain (Figure 4A and Supplementary Table S6).

We utilized gene expression levels at seven DAF as a control to better understand the expression pattern of TaZIP genes during the grain filling stage. The results showed that TaZIPs have significantly different expression levels (Figure 4B and Supplementary Table S6). Thirty-one TaZIP genes displayed downregulated expression, with the lowest expression levels at 28 DAF. Eight TaZIP genes were highly expressed, but the expression trend was diverse, TaZTP7-B had highest expression in 28 DAF. At the grain filling stage, TaIRT2-A, D were also highly expressed. TaZIP14-B and TaZIP14-A were unique genes in that their expression was upregulated at the grain filling stage. The expression level of TaZIP14-B greatly increased, while TaZIP14-A was expressed moderately. Both TaZIP13-B and TaZIP13-D were highly expressed, but their expression patterns were different. The expression levels of TaZIP13-D increased rapidly while TaZIP13-B displayed moderate expression (Figure 4B and Supplementary Table S6).

We then analyzed the expression patterns of ZIP genes under different concentrations of ZnSO₄ and FeCl₃ solutions. Under low concentrations of ZnSO₄ stress, almost all ZIP genes were upregulated, but their expression trends differed. Sixteen ZIP genes were highly expressed (value more than 1) under 0.05μmol/L ZnSO₄ and decreased under 0.5 and 50μmol/L conditions (Figure 5A). The expression of these genes was suppressed when Zn concentration increased. Eighteen genes were highly upregulated under the concentration of 0.50μmol/L ZnSO₄. Moreover, the expression of nine ZIP genes (TaZIP21, -TaZIP5-D, TaZIP5-B, TaZIP8-B, TaIRT1-D, TaIRT2-D, TaIRT2-A, TaZIP10-B, and TaZIP10-A) increased in 50μmol/L ZnSO₄. Especially TaZIP14-B, which displayed high levels of expression in the 0.5μmol/L solution, but low expressed in the 50μmol/L ZnSO₄ solution. The TaZIP13-B gene was moderately expressed in the 0.05μmol/L solution, while it was upregulated in other concentrations of ZnSO₄ solution (Figure 5A and Supplementary Table S6).

The expression pattern of TaZIPs in FeCl₃ was similar to ZnSO₄. At low concentrations of FeCl₃ solution, the expression of all ZIP genes was increased, but of which 16 genes were further suppressed or moderately upregulated under 0.5 and 50μmol/L FeCl₃ treatment (Figure 5B and Supplementary Table S6). Furthermore, nine genes were highly expressed in the 0.05 and 0.5μmol/L FeCl₃ solutions. Eleven genes were highly expressed under 0.05 and 0.5μmol/L of FeCl₃ treatment. The genes TaZIP14-B, TaZIP13-B, and TaZIP7-A, B, D were all upregulated under high concentrations of FeCl₃ solution. The TaIRT1-D gene is key to transporting iron ions; however, its expression level under FeCl₃ treatment is lower than under ZnSO₄ treatment. Additionally, the expression pattern of TaIRT2-A, D was similar to that of TaIRT1-D (Figure 5B and Supplementary Table S6). We also performed Fe and Zn starvation treatment that wheat seedlings at two leaf stage were treated with Hoagland nutrient solution as CK and Hoagland medium lacking ZnSO₄ (Zn-deficient) or Fe (III)-EDTA (Fe-deficient). After treated by 6 and 12h, six genes were upregulated in roots and shoots and mainly expressed in roots (Supplementary Figure S3). TaIRT2-A, D were upregulated slightly under Zn-deficiency while highly expressed under Fe-deficiency. TaZIP13-B, D and TaZIP14-A, B were highly expressed under Zn-deficiency condition.

To investigate whether miRNA degraded the ZIP genes, we analyzed the expression pattern of three miRNAs in wheat under ZnSO₄ and FeCl₃ solutions. These three miRNAs target TaIRT2-A, TaZIP14-B, and TaZIP13-B (Supplementary Tables S4 and S5).

Three miRNAs were slightly downregulated under 0.05 and 0.5μmol/L of the ZnSO₄ solution compared with the control (Figure 6). tae-miR164 was also slightly downregulated in 50μmol/L ZnSO₄ solution, while tae-miR5084miR5084 and tae-miR395a were slightly upregulated under 50μmol/L of the ZnSO₄ solution.

tae-miR164 displayed low levels of expression in the FeCl₃ solution compared with control, with the lowest expressed in 0.5μmol/L FeCl₃ solutions. tae-miR395a and tae-miR5084 had similar expression patterns in the FeCl₃ solution. These two miRNAs were downregulated in the 0.05 and 0.5μmol/L FeCl₃ solutions, with the lowest expression in the 0.05μmol/l FeCl₃ solution. However, under 50μmol/L of FeCl₃, these two miRNAs were upregulated slightly and did not differ from the control (Figure 6). The expression pattern of miRNAs contrary to the targeted genes.

After demonstrating that the three genes were upregulated when exposed to Zn and Fe stress, we also revealed the biological function of three ZIP genes (TaZIP14-B, TaZIP13-B, and TaIRT2-A) by yeast complementation analysis (Supplementary Table S5). OsZIP3, OsZIP5, and OsIRT1 were chosen as positive controls, all of which have been demonstrated to be involved in Zn and Fe transport in rice (Chen et al., 2008; Tiong et al., 2014).

Three yeast strains wild-type DY1455, the Saccharomyces eviscerate zrt1zrt2 mutant (ZHY3), and the fet3fet4 mutant (DEY1453) were used, to verify that the three wheat TaZIP genes were capable of restoring the ability to transport zinc and iron in the mutant yeast. The full-length cDNA of both the wheat and rice genes were inserted and expressed in the two mutants. The transformed ZHY3 with TaZIP genes were grown on an SD medium with 0.4mM EDTA and the transformed DEY 1453 were grown on a SD medium with 50mM MES. The results demonstrated that the growth of the ZHY3 yeast with pDR195 was inhibited under zinc-deficient conditions in a normal SD-Ura medium, while the mutant with TaZIP genes and rice genes successfully recovered from the growth defect (Figure 7A). The TaZIP13-B gene reversed the growth defect. When the ZHY3 yeast was exposed to a 200μM ZnSO₄ medium, the growth of ZHY3 was not inhibited. The growth of DEY1453 was similar to ZHY3 (Figure 7B). Under Fe-deficient conditions, the growth of DEY1453 containing a vector was severely inhibited, while the growth was reversed during the expression of TaZIP and rice genes. Once a sufficient amount of FeCl₃ was supplied, growth recovered. TaZIP13-B demonstrated the strongest propagation under Fe-deficient conditions. These results revealed that TaZIP14-B, TaZIP13-B, and TaIRT2-A could effectively complement the zinc transporter mutant zrt1zrt2 and the iron transporter mutant fet3fet4, suggesting they could successfully transport Zn and Fe.

The TaZIP13-B gene was upregulated both at the grain filling stage and under Fe/Zn stress. The yeast complementation experiment proves that TaZIP13-B can transport Fe and Zn in a yeast mutant. Therefore, TaZIP13-B was transformed into Arabidopsis to verify this gene function. We selected lines OE-1, OE-2, and OE-3 with high expression levels for further analyses (Figure 8A).

On the MS, MS-Zn, and MS-Fe medium, the germination rate of the transgenic lines OE1, OE2, OE3, and WT approached 100%, with no significant difference between transgenic lines and WT (Figures 8B,C). When treated with 50μmol/L of Zn and Fe, the germination rate of four lines decreased, and no significant difference was observed between them (Figures 8B,C). When exposed to 200μmol/L of Zn and Fe MS medium, the germinations of the OE1-3 line were significantly higher than the WT line. The germination rate of OE1 was 85.3%, while that of OE2 was 82.7%, OE3 was 93.3%, and WT was 46.7% in 200μmol/L Fe MS medium. In 200μmol/L Zn MS, the germination rates of OE1-3 were 89.3, 90.7, and 85.3%, respectively, while that of WT was 74.7%. Compared with OE1 and OE3, the transgenic line OE2 had the highest germination on 300μmol/L Fe MS (up to 58.7%), while WT had a germination rate of only 22.7%. All three transgenic lines had significantly higher germination rates than WT line (Figures 8B,C).

Root length is another index used to evaluate plant tolerance to Zn and Fe stresses. We found no significant difference in root length between the three transgenic lines and WT on MS, Zn-, Fe-, and 50μmol/L Zn/Fe MS medium (Figures 8D,E); however, the number of roots in Zn-, Fe-, and 50μmol/L Zn/Fe MS medium increased. When transgenic lines and the WT line were placed on 200μmol/L Fe MS medium, the root lengths were significantly inhibited, but the root lengths of the three transgenic lines were significantly longer than WT line. The root length of OE2 was 1.78cm, which was the longest root length on the 200μmol/L Fe MS medium. On the 200μmol/L Zn MS medium, the three transgenic lines were significantly longer than the WT line. On the 300μmol/L Fe MS, the root lengths were further inhibited compared to the 200 Fe MS medium, and the root lengths of the three transgenic lines were significantly longer than the WT line. The root lengths were also inhibited in 300 MS Zn MS, though the root lengths of the three transgenic lines were significantly longer than that of the WT line (Figures 8D,E).

To further evaluate the tolerance of the three transgenic lines and WT to Zn and Fe stress, we treated the transgenic and WT lines with different concentrations of a ZnSO₄ and FeCl₃ solution. Under 200μmol/L FeCl₃, the WT leaves wilted and yellowish-brown spots appeared on a few leaves, though this did not occur on the OE1, OE2, and OE3 lines (Figure 8F). Under 300 and 400μmol/L of the FeCl₃ solution, most WT leaves were brown while transgenic lines were normal. Almost all of the WT petiole browned, particularly under 400μmol/L FeCl₃ solution treatment (Figure 8E). We then measured the chlorophyll content of four lines. As the concentration of the FeCl₃ solution increased, the chlorophyll content of the four lines decreased (Figures 8G,H). However, the chlorophyll content in the three transgenic lines was significantly higher than in the WT line. The three transgenic lines exhibit greater FeCl₃ resistance than the WT line, according to observations of the stomata in all four lines (Figure 8I). The stomata opened on all four lines when treated with water. Under 200 FeCl₃ μmol/L treatment, the stomata of the WT line closed slightly, while the stomata of the three overexpression lines opened. When treating the four lines with 300μmol/L FeCl₃ solutions, the stomata of the WT closed while OE1, OE2, and OE3 remained open (Figure 8I). Under 400μmol/L FeCl₃ solution, the stomata of all lines closed, and the stomata of the WT died. When treated with 200μmol/L ZnSO₄ solution, the leaves of the WT line turned gray, while the leaves of the three transgenic lines remained green (Figure 8G). The results were similar to the 200μmol/L ZnSO₄ treatment when they were treated with the 300 and 400μmol/L ZnSO₄ solution. However, when treated with the 300 and 400μmol/L ZnSO₄ solutions, part leaves in the transgenic lines turned gray and wilted (Figure 8G). The chlorophyll content of both of the transgenic and WT plants was analyzed. Under treatment with three different concentrations of ZnSO₄ solution, the chlorophyll content in the WT line was significantly lower than in the transgenic lines (Figure 8H). The stomata phenotype of all lines under the ZnSO₄ treatment was similar to that of the FeCl₃ treatment (Figure 8I).

Zn is important to photosynthesis in plants, and photosynthesis is related to crop yield. Therefore, we analyzed whether the TaZIP genes affect seed size. When the four lines were treated with water, the seed width and length of the three transgenic lines were longer than that of the WT line (Figures 8I,J). When treated with 200mol/L FeCl₃ and ZnSO₄ solution, all lines’ seed size shrank, but the three transgenic lines’ seed breadth and length remained somewhat longer than the WT line (Figures 8J,K). Our results indicate that transferring the wheat TaZIP13-B gene into Arabidopsis increases seed size and might increase production.

We used 0.15g samples of the roots, shoots, and seeds to measure the metal contents of tissues in this study. Compared with the WT, the three overexpression lines accumulated more Fe (23.6–38% higher) and Zn (7.5–33% higher) in the roots, while overexpression lines also accumulated more Fe (17.3–20.3%) and Zn (10.3–26.0%) in the shoots than the WT line. Compared with the WT, the seeds of the transgenic lines have a higher Fe content and a higher Zn content (Figure 8L). These results indicate that transgenic lines can absorb more Fe and Zn from the soil, enriching the Fe and Zn in seeds.