Barley metallothionein isoforms, MT2b2 and MT4, differentially respond to photohormones in barley aleurone layer and their recombinant forms show different affinity for binding to zinc and cadmium

Metallothioneins (MTs) are metal-binding proteins that have important roles in the homeostasis of heavy metals. In this study, the two MT genes was studied in response to phytohormones using the barley aleurone layer as a kind of model system. The aleurone layer was isolated from barley embryo-less half grains and was incubated for 24 h with different phytohormones. Based on the results the genes encoding HvMT2b2 and HvMT4 were down-regulated through gibberellic acid (GA), while they were and up-regulated through salicylic acid (SA). Despite this, these two genes were differentially expressed to other hormones. Furthermore, the proteins HvMT2b2 and HvMT4 were heterologous expressed as GST-fusion proteins in E. coli. The HvMT4 and HvMT2b2 heterologous expression in E. coli gives rise to 10- and 3-fold improvements in the accumulation capacity for Zn2+, respectively. Whereas the transgenic E. coli strain that expresses HvMT2b2 could accumulate Cd2+ three-fold higher than control. The expression of HvMT4 did not affect the accumulation of Cd2+. HvMT4 which is known as seed-specific isoform seems to be able to bind to Zn2+ with good affinity and cannot bind Cd2+. In comparison, HvMT2b2 was able to bind both Zn2+ and Cd2+. Therefore HvMT4 could serve a noteworthy role in zinc storage in barley seeds. The expression of HvMT4 is induced by SA 30-fold, concerning the untreated aleurone layer. Such results could provide good insights for the assessment of the effects of phytohormones in the molecular mechanism involved in essential metal storage in cereal seeds.


Introduction
Heavy metals like Cu 2+ and Zn 2+ are necessary for the normal plant growth and development. They important roles as cofactors of various enzymes (Broadley et al. 2007;Ishimaru et al. 2011;Tripathi et al. 2015). Despite this, high concentrations of essential and nonessential heavy metals in the soil may result in toxicity symptoms due to diverse impacts of heavy metals, such as disrupting the protein structure and inhibiting their activity, displacing of an essential element in proteins, and stimulation of the formation of reactive oxygen species (Cho and Park 2000;Van Assche and Clijsters 1990;Sharma et al. 2008;Bhargava et al. 2012). Such toxicity symptoms result in the plant growth deficiency as well as the low yield. However, in the plants, some mechanisms have evolved tolerance in polluted soils which maintain homeostasis of heavy metals in the cells (Tripathi et al. 2015;Clemens 2001;Hossain et al. 2012). Essential metals homeostasis, such as copper, iron, manganese and zinc, needs well-coordinated activities of transporters mediating the import into the cell, in addition to distributing to organelles and exporting from the cell (Tripathi et al. 2015;White and Broadley 2011;Shahpiri et al. 2015a, b). There are also metal chelating peptides and proteins like metallothioneins (MTs) and phytochelatins that are important in the homeostasis and transport of physiologically some essential metals (such as Zn 2+ and Cu 2+ ), metal detoxification (such as Cd 2+ and Hg 2+ ), protection against the possible oxidative stress, intracellular redox balances, cell proliferation and apoptosis regulation, and the need for the protection against neuronal injury as well as degeneration (Cobbett 2000;Wang et al. 2010;Palmiter 1998;Buchanan-Wollaston 1994;Chaudhary et al. 2018;Mierek-Adamska et al. 2019).
MTs can be regarded as a family of low molecular weight (7-10 kDa), Cys-rich proteins that can bind to metals through Cys thiol groups. According to the Cys residues arrangement, MT proteins can be divided into three groups. Class I MTs included 20 conserved cysteine residues, mainly mammalian MTs; they can also be seen in vertebrates. Class II MTs are mostly in plants, fungi and invertebrates, with no strict cysteines arrangement. Class III MTs are mostly in plants; they are non-translationally synthesized polypeptides which are composed of some repeating units of γ-Glu-Cys, such as some phytochelatins (PCs) (Cobbett and Goldsbrough 2002;Rauser 1999). Plant MTs, in turn, may be classified into four groups: MT1, MT2, MT3, and MT4, according to the Cys residue distribution pattern (Freisinger 2011). The members of MT1, MT2 and MT3 subfamilies include two Cys-rich regions in the N-terminal and C-terminal which are separated through a Cys-free linker region having a typical length that is almost between 30 and 45 amino acids (Cobbett and Goldsbrough 2002;Schicht and Freisinger 2009). The six Cys residues which are located at the C-terminus in each subfamily can be arranged in highly conserved CXCXXXCXCXXCXC motifs (X: any amino acid besides Cys) (Freisinger 2011). On the other hand, the number as well as the distribution pattern of Cys residues in the N-terminal region can identify each subfamily; it can be applied as the main distinguishing factor for subfamily discrimination (Freisinger 2011). The typical type 1, type2 and type3 MTs include six, eight and four Cys residues, respectively, at their N-terminal Cys-rich region (Lane et al. 2011). The members belonging to the type 4 subfamily can be described by three Cys-rich regions that are separated through shorter peptide linkers (Peroza and Freisinger 2007).
Plants generally have different MT isoforms that belong to different types of MTs. Different MT types are differentially represented in different plant organs (Hegelund et al. 2012;Zhou et al. 2006). Type 1 MTs can be expressed mostly in roots, whereas type 2 MTs are in the leaves, type 3 MTs are in the fruits and type 4 MTs are in the seeds (Zhou et al, 2006;Hsieh et al. 1995Hsieh et al. , 1996Guo et al. 2003;Mierek-Adamska et al. 2018).
Despite many studies already published on the MT types expression as the response to heavy metals and oxidative stresses, the knowledge still lags behind the MT types expression in responding to plant hormones. Therefore in the present work, the aleurone layer of barley seed was used as a model system to investigate the impact of hormones on the MT types expression (Finnie et al. 2011).
The aleurone layer is composed of one layer of single and unique cells around endosperm in the cereal seeds (Finnie et al. 2011). Upon seed imbibition, the embryo produces the hormones that are received by aleurone layer cells (Bønsager et al. 2007). In response to hormones, the aleurone layer synthesizes and secretes a range of enzymes, such as hydrolases, for depolymerizing endosperm cell walls and degrading endosperm storage carbohydrates and proteins (Caspers et al. 2001;Mundy and Rogers 1986). The barley aleurone layer may be separated from the rest of seed tissues and kept in the culture medium, thus making it possible to address the added signaling molecules in some isolated system (Finnie et al. 2011). Such qualities have encouraged its use to serve as a model system to study plant signaling and germination.
The barley genome encodes a family of 10 metallothioneins (MTs) which belongs to different MT types (Schiller et al. 2013). Previously, Tauris et al. (2009) andHegelund et al. (2012) revealed that the genes encoding HvMT2b2 and HvMT4 and HvMT3 were expressed at an extremely high level in the barley aleurone layer (Tauris et al. 2009;Hegelund et al. 2012). In the present research study, the HvMT2b2 and HvMT4 expression, the members of type2 and type 4, respectively were studied in the response to phytohormones gibberellic acid (GA), abscisic acid (ABA), salicylic acid (SA), kinetin (Kin), indole acetic acid (IAA) and ethephon (ET). Furthermore, to functionally characterize these MT isoforms, the genes which encode HvMT2b2 and HvMT4 were cloned and heterologously represented in Escherichia coli. The ability of strains expressing the recombinant forms of HvMT2b2 and HvMT4 were analyzed for remediation of Zn 2+ and Cd 2+ from the medium. The recombinant HvMT2b2 and HvMT4 were purified and the apo-protein was prepared. The apo-form was exposed to the metal ions and the ability of their binding to HvMT2b2 and HvMT4 were studied by in vitro experiments.

Materials and methods
Preparation and treatment of aleurone layer Seeds obtained from barley cultivar Fajr 30 were got from Isfahan Center for Research of Agriculture Science and Natural Resources. The seeds embryocontaining part was cut away from others seeds with a sharp scalpel. The embryo-less half grains were subjected to sterilizing with 70% ethanol for a period of 1 min; this was followed by washing five times through sterilized water. Subsequently, the seeds were subjected to soaking for a period of 4 days in sterile water with 50 μgml −1 ampicillin and 5 μgml −1 nystatin (Shahpiri et al. 2008;Shahpiri et al. 2015a, b). The removal of the seed coat was done and the endosperm was taken away from the aleurone layers. Accordingly, incubation of 100 mg (fresh weight) of aleurone layers was done in 2 ml buffer (20 mM CaCl 2 , 20 mM Na succinate pH 4.2, 50 μgml −1 ampicillin, 5 μgml −1 nystatin).
For treatment by hormones, addition of either GA (2 μM), ABA (20 μM), SA (1 mM), Kinetin (52 μM) or IAA (10 μM), ET (50 mg l −1 ) to the medium was done. Some aleurone samples were treated with a combination of hormones, GA + ABA and GA + SA. Incubating up to a period of 24 h was done at room temperature through continuous gentle shaking.
Harvesting of Aleurone layers was done following 24 h; washing was then done four times using an incubation buffer with no antibiotics or hormones; then they were frozen in liquid nitrogen and stored at the temperature of − 80° C till use. In addition, for the study of SA, the isolated aleurone layer was incubated for 24 h in different SA concentrations (0, 300, 600, and 1000 µM). The aleurone layer was washed and harvested as explained above.

Real-time PCR
Total RNA was got from the aleurone layer samples by applying the High Pure RNA Isolation Kit (Roche) and then underwent treatment by RNase-Free DNase (Thermo Scientific) for the removal of the genomic DNA contamination. Total RNA (0.1 μg) was reverse transcribed (RT) to synthesize the first strand cDNA by applying M-MLV reverse transcriptase (Thermo Scientific) and oligo-dT primer (Thermo Scientific), based on the manufacturer's recommendations. Regarding the PCR reaction, the designing of specific primers was done by employing the software Invitrogen Oligo primer analysis software V.7. The sequences of forwarding and reverse primers for Amylase (HvAmy), HvMT4, HvMT2b2, and B-Actin (HvACT) were inserted in Table 1. PCR was performed in a total volume of 15 μL containing 7.5 Taq 2X master mix Red (Amplicon), 0.4 μl (10 μM) from each of the primers, 1 μl cDNA (1:15 dilution of the obtained cDNA) as a template and 6 μl RNase free water. Reactions were done using an initial 5 min denaturation at the temperature of 95 °C; this was followed by 40 cycles of 95 °C for 24 S, (54-59 °C dependent on kind of gene) for 25 S, 72 °C for 30 S; after that, there was the final extension at 72 °C for a period of 5 min. Calculation of normalized expression levels was done by the 2 −ΔΔCt method with barley β-Actin (HvACT) as the reference gene (Livak and Schmittgen 2001). This was according to two technical and three biological replicates. Each of the data points indicates the mean value ratio which is obtained from three replicates of samples to the mean value obtained from control samples (untreated aleurone layer for 24 h).

Cloning of the gene encoding HvMT2b2 and HvMT4
Total RNA was got from aleurone layer samples; the removal of the contaminating genomic DNA was done as shown above. Total RNA (0.1 μg) was reverse transcribed (RT) in order to synthesize first strand cDNA by applying AMV reverse transcriptase (Thermo Scientific) and oligo-dT primer (Thermo Scientific) based on the manufacturer's recommendations. The coding sequence of genes encoding HvMT2b2 and HvMT4 were used to design gene specific primers (Table 1). Designing of the primers was done based on the 5′ and 3′ end of CDS with an EcoRI restriction site (underlined) at the 5′ end of forwar primer and a restriction site HindIII at the 5′ end of reverse primers. Four other bases were included at the 5′ end for every oligonucleotide primer. The PCR reaction was performed with pfu DNA polymerase in some 20 µl reaction that contained template (the synthesized cDNA), deoxynucleotides and reaction buffer, as well as the primers. The digestion of the PCR product was done by the corresponding restriction enzymes and purification from the gel. Then the products with sticky ends were ligated into pGEX-4 T-1 as an expression vector followinf linearization with EcoRI and SalI. The obtained plasmids, named as pGEX-HvMT2b2 and pGEX-HvMT4, were subjected to verification by sequencing; after that, they were introduced into E. coli proteasedeficient strain Rosetta (DE3) for proteinexpression. The obtained strains were named as R-MT2b2 and R-MT4. Also, the control strain was also developed through transferring plasmid pGEX with no gene into Rosetta (DE3).

Heterologous expression of HvMT2b2 and HvMT4 in E. coli
The strains Control, R-MT2b2, and R-HvMT4 were grown at the temperature of 37 °C in Luria-Bertani (LB) medium in the volume of 50 ml which was supplemented with 50 µg ml −1 kanamycin and 5 ug ml −1 chloramphenicol to an OD 600 nm of almost 0.6. At this OD, inducing of cultures was done by 100 uM isopropyl β-d-thiogalactopyranoside (IPTG). Supplementation of the culture medium was done by applying 0.6 mM ZnSO 4 .7H 2 O 20 min after adding IPTG. For the confirmation of the protein's heterologous expression, 1 ml samples of the culture medium was harvested through centrifugation 1-4 h after adding IPTG and frozen at the temperature of − 80 °C till application. The obtained frozen pellets were resuspended in 250 µl pre-cold 10 mM Tris-HCl, at the pH level of 8.0, disrupted by mild sonication at the temperature of 4 °C, and centrifuged at 12,000×g, for 20 min. The soluble proteins which were recovered in the supernatant phase were then subjected to analysis through 12% SDS-PAGE and staining by Coomassie Brilliant Blue R-250 (Candiano et al. 2004).
Regarding proteins large-scale production, the cells were grown in 500 ml of medium. Induction was done by IPTG and supplementation with 0.6 mM Cloning of HvMT2b2 ATAT AAG CTT TCA GGC GGT CGA GCC GCA TGA AGC a ZnSO 4 .7H 2 O, as discussed before. The cells obtained from the whole volume were subjected to harvesting 4 h after adding IPTG. The extraction of the soluble proteins was done as previously discussed. Regarding purification, the use was made of the extracted soluble proteins on to GST Trap column (GE, Healthcare) pre-equilibrated with binding buffer (PBS, pH 7.5), and the elution of bound proteins was done by reduced glutathione. The protein fractions aliquots were subjected to analysis through 12% SDS-PAGE and staining through Coomassie Brilliant Blue R-250 (Candiano et al. 2004). Then pure fractions were subjected to pooling and sent into 12 kDa molecular weight cutoff cellulose tubes (Sigma) and dialyzed against 10 mM Tris-HCl, at the pH level of 8.0, at the temperature of 4 °C, overnight for removing imidazole and other salts. Concentrations of protein were determined according to the Beer-Lambert law based on the molar extinction coefficient of 44,537, 46,465, and 44,975 M −1 cm −1 for GST, GST-HvMT2b2, and GST-HvMT4, respectively. The molar extinction coefficient was determined using Expasy's Protparam server (Gasteiger et al. 2005).
Transgenic strains tolerance in response to metals Transgenic strains Tolerance and the control of Zn 2+ and Cd 2+ in the growth medium was investigated by 0.75 mM of Zn 2+ from ZnCl 2 and 0.75 mM Cd 2+ from CdCl 2 ⋅H 2 O. So, 5 ml of the cells overnight cultures was subjected to inoculation in 80 mL of LB medium supplemented by applying the desired antibiotics. The induction of the cultures was done at OD 600 = 0.6 by adding 0.1 mM of IPTG. Following 20 min, addition of the metal salts to the cultures was done. Bacterial growth was investigated up to 12 h according to OD 600 measurements (Nezhad et al. 2013).

Ion accumulation determination in the strains
The cells cell, induction through IPTG, and supplementation by metals were carried out as described before. To analyze Zn 2+ and Cd 2+ , 6 h after adding IPTG, the precipitation of the cells was done; then washing was done with water three times; this was followed by drying overnight at room temperature. The digestion of the cells was done by applying high-pressure microwave method (Van Herck et al. 2001); the metal amount was measured by applying an atomic absorption spectrophotometer (Perkin Elmer AAnalyst). Each of the data points reflects the mean ± SD obtained from two independent replicates.
Preparation of apo-protein and reconstitution by diverse metals Apo-protein and reconstitution with metals were prepared, as shown before, though with slight modifications (Toriumi et al. 2005). Purified GST-HvMT4 and GST-HvMT2b2 aliquots were subjected to acidification with HCl to the pH level of 2.0. Dialysis of the samples was then done against 0.1 N HCl for the removal of the bound ions. The protein's concentration was determined following dialysis. Apo-protein reconstitution with metals was done by adding 10 mol equivalents of Cd 2+ and Zn 2+ ions; this was followed by samples neutralization to pH 8.0 with 200 mM Tris. The removal of unbound metals was done by dialysis against 10 mM Tris-HCl, at the pH level of 8.0, at the temperature of 4 °C, overnight.
UV spectra and DTNB assay for analysis of binding metal to MTs The absorption spectra of the purified and desalted metal-incubated as well as apo-forms of GST-HvMT2b2 and GST-MT4 proteins (200 µg ml −1 ) in the buffer 10 mM Tris-HCl, pH 8.0 was determined between 220 and 350 nm by applying a spectrophotometer (DU 530). For DTNB assay, the wild type and mutant proteins' binding ability towards Cd 2+ , Zn 2+ and was assessed based on the competitive reaction with 5, 50-dithiobis (2nitrobenzoic) acid (DTNB), by applying the Emoto method with slight modifications. A 300 µl reaction mixture which contained 10 mM Tris-HCl, at pH 8.0, and 1.5 nmol of each protein were then put in some quartz cuvette. The reaction started by adding 75 nmol of DTNB. The absorbance occurring at 412 nm was recorded on a Beckman DU-530 spectrophotometer at 1 min intervals for a period of 60 min at room temperature, showing the 5-thio-2-nitrobenzoate anion formation with the molar extinction coefficient being 14,140 M −1 cm −1 . As a blank, 300 µl of 10 mM Tris-HCl solution which contained 75 nmol of DTNB was applied. Each curve's pseudo initial velocity was estimated on the basis of absorbance versus the reaction time.

Statistical analysis
All data related to this study are represented as mean ± standard deviation (SD) from three independent biological replications. The least significant difference (LSD) test was then conducted to compare the means following the analysis of variance (ANOVA). The differences between treatments were considered significant when p ≤ 0.05. The expression of gene encoding α-amylase (Amy) was analyzed beside of HvMT genes due to information available for the effect of hormones on the Amy genes by previous researchers. Here the results showed that whereas GA and Kin significantly induce the expression of Amy, the expression of Amy was suppressed by ABA and SA, with respect to untreated aleurone layer (Fig. 1A). The suppression effects of ABA and SA were also observed when the addition of the GA + ABA and GA + SA mixture was the medium was done. In these conditions, it seems that GA induces the expression of Amy but the induction is not complete because of the inhibitory impact of ABA or SA on the Amy expression. The Amy expression was also enhanced by ET, however, the effect was much lower than the effect of GA and Kin corresponding to previous works. This confirmation showed that the steps of preparation on aleurone layer, aleurone layer culture, the application of treatments have been performed correctly.

Gene
Regarding MT2b2, whereas the expression was induced by SA, IAA, ABA, KT and ET, the expression was suppressed by GA (Fig. 1B). The suppression effect of GA was also confirmed in AL-GA + SA and AL-GA + ABA. Between the inducible hormones, SA has a higher effect on the expression (fivefold concerning untreated samples) than others (up to 3.7 fold concerning untreated samples).
The results related to HvMT4 showed that SA highly induces the expression of HvMT4 in the aleurone layer (30-fold, Fig. 1C). Rather than the hormone GA suppresses the expression of HvMT4 in the aleurone layer. The suppression effect was also observed in Al-GA + SA. The expression of MT4 was not affected by ABA and IAA and was slightly induced by ET and ABA (1.8-fold for ABA and 2.3fold for ET, concerning untreated aleurone layer).
It seems that SA is the hormone that has a considerably high effect on the expression of both MT isoforms in the aleurone layers treated. However, the expression of the genes which encode HvMT2b2 and HvMT4 does not follow the same pattern as the response to other hormones. Because of the high inducible effect of SA on the expression of HvMT4 and HvMT2b2 and the high suppression effect of this hormone on Amy, the aleurone layer was treated with different concentrations of SA. As represented Fig. 2A, Amy expression remarkably decreases by increasing the concentration of SA. On the contrary, the expression of both HvMT4 and HvMT2b2 increase by increasing the concentration of SA (Fig. 2B, C).

HvMT4 and HvMT2b2 can be expressed in E.coli
The HvMT2b2 full-length coding sequence includes 237 bp, which can encode a 78 amino acid protein having the 7.4/4.9 kDa/pI theoretical molecular weight. Like other plants MT type 2, eight Cys residues which are arranged in CCXXXCXCXXX-CXCXXXCXXCXXX are present at this sequence N-terminus (Fig. 3A). Unlike other plants MT type 2 which contain six Cys residues at the C-terminal, the sequence of HvMT2b2 has nine Cys. It should be noted that the sequence of OsMTI-2b in rice has nine Cys at the C-terminal. The HvMT4 full-length coding sequence includes 237 bp which can encode a 78 amino acid protein with 7.5/7.3 kDa/pI theoretical molecular weight. The same as other members of type-4 MTs there are 6 Cys at N-terminal Cys-rich region, 6 Cys at the middle Cys-rich region. However, unlike the other members of type-4 MTs which contain 6 Cys at the C-terminal region, there are 5 Cys residues at this region of HvMT4 (Fig. 3B). (1 mM), Kinetin (52 μM) or IAA (10 μM), ET (50 mg L −1 ) were added to the medium. Expression level is represented as a value relative to that in the control sample (the incubated aleurone layer with no hormones). Each data point represents the ratio of the mean value as obtained from three replicates of samples to the mean value obtained from three control samples (untreated aleurone layer for 24 h). Different letters above the bars represent statistically significant differences by the LSD test Vol:. (1234567890) The coding sequence of HvMT2b2 and HvMT4 were allowed to be subcloned into the expression vector pGEX which contained the coding sequences for an N-terminal fusion partner of glutathione-S-transferase (GST-tag), (Fig. 3C).
After induction through IPTG, some recombinant proteins including GST, GST-HvMT2b2 and GST-HvMT4 were expressed in the soluble fraction of the E. coli cells which carried the pGEX (control strain) and pGEX-HvMT2b2 (strain R-MT2b2) and HvMT4 (strain R-MT4), respectively. GST, GST-HvMT2b2 and GST-HvMT4 theoretical molecular weight was 25, 30.4 and 30.5 kDa, respectively. The SDS-PAGE analysis revealed expected molecular masses sharp protein bands, including GST-HvMT2b2 and GST-HvMT4, as well as GST, expressed in the control strain (Fig. 4A). The sharpness of corresponding bands increases during the time after the addition of IPTG until 4 h. The recombinant GST-HvMT2b2 and GST-Hv-MT4 were purified from the soluble fraction by applying affinity chromatography and the purification extent was measured through SDS-PAGE analysis. (Fig. 4B).
To quantify the protein's solubility, the protein band's intensity was compared in the total fraction (soluble + insoluble), soluble, and insoluble fraction. The GST protein produced by the control strain was highly soluble as predicted and a large portion of GST appeared in the soluble fraction. The protein GST-HvMT4 and GST-HvMT2b2 were also present in the soluble fraction of the large-scale preparation, however, still, a significant portion of the protein was also present in the insoluble fraction (Fig. 4C). It should be noted that the previously several studies have shown that MT proteins without the fusion protein are completely insoluble. In addition, the small size of protein increases the risk of digestion of protein by native proteases.

Tolerance of transgenic strain to Zn 2+ and Cd 2+
The metals minimum concentrations could considerably influence the control strain's growth (E. coli cells harboring empty pGEX vector heterologously expressing GST) were shown to be 0.75 mM for Cd 2+ and Zn 2+ (data not shown). In the culture medium with no metals, the maximum OD 600 (In this work the OD 600 after 12 h) for control, R-HvMT2b2, and R-HvMT4 were similar (OD 600 = 2.5). Despite this, when 0.75 mM ZnSO 4 was present, the maximum OD for control strain reached up to 1 and the OD of both R-HvMT2b2 and R-HvMT4 remained almost constant (Fig. 5). These obtained results, thus, show that both HvMT4 and HvMT2b2 expression confers Fig. 2 Gene expression of A HvAmy, B HvMT2b2 and C HvMT4 in barley aleurone layer responding to different concentrations of SA. Each data point indicates the ratio of the mean value which was obtained from three replicates of samples to the mean value obtained from three control samples (untreated aleurone layer for 24 h). Different letters above the bars represent statistically significant differences by LSD test tolerance to Zn 2+ in transgenic strains. In the presence of CdCl 2 , the maximum OD 600 decreased to 1.7, 1.5, and 1 for control, R-MT2b2, and R-MT4, respectively. Therefore it seems that the presence of Cd 2+ has been toxic for control as well as R-MT2b2 and R-MT4. However, in comparison between R-MT2b2 and R-MT4, it seems that R-MT4 was more sensitive to Cd 2+ (Fig. 5).

Accumulation of metal in transgenic strains
To find if R-HvMT2b2 increased tolerance to Zn 2+ and Cd 2+ and that of R-HvMT4 to Zn 2+ could be related to these ions accumulation in these cells, the recombinant strains, in addition to control, were grown in a liquid medium which was supplemented with either 0.75 mM Zn 2+ or 0.75 mM Cd 2+ (Fig. 6). The amount of metals in the cells was obtained by atomic absorption at 6 h after adding metals to the cultures. While the amount of accumulated Zn 2+ in the control strain was 0.56 mg g −1 DCW, the strain R-MT2b2 and R-MT4 could accumulate 1.8 and 4.5 mg g −1 DCW, respectively. Furthermore, the amount of accumulated Cd 2+ in the control strain was 2.3 mg g −1 DCW. However, the amount of accumulation for R-MT2b2 was significantly higher than control and reached 6 mg g −1 DCW. In contrast, the Cd 2+ accumulation in strain R-MT4 was similar to the control strain. Therefore while the expression of GST-MT2b2 significantly enhanced the Zn 2+ and Cd 2+ accumulation by 3.2 and 2.6 folds more than control strain, the expression of GST-MT4 affected the accumulation of Zn 2+ , eightfold more than control strain and was not effective on the accumulation of Cd 2+ .

The verification of binding metals to recombinant HvMTs in vitro
The recording of the related ligand-to-metal charge transfer bands which were seen in the UV spectra (210-250 nm) was done following the incubation of the recombinant Apo/GST-HvMT2b2 with Cd 2+ and Zn 2+ . In comparison to the apo forms, the proteins Cd 2+ /GST-HvMT2b2 displayed enhanced absorbance at 250 nm, which is common for Cd 2+ -containing proteins. An almost the same increase of absorbance at 240 nm was seen after the Zn 2+ /GST-HvMT2b2 incubation, thus confirming the formation of Cd 2+ and Zn 2+− thiolate clusters in HvMT2b2 (Fig. 7A). In comparison to UV spectra of Apo/GST-HvMT4, the rise of absorbance was observed at 240 nm for Zn 2+ /GST-HvMT4 (Apo/GST-HvMT4 which was incubated with Zn 2+ ), thus indicating the binding ability of GST-HvMT4 to Zn 2+ . Despite this, the spectrum of the proteins which were incubated with Cd 2+ ions was found to be like Apo/HvMT4, indicating that the apo-protein failed to bind Cd 2+ ions (Fig. 7B).

Metal's binding strength to GST-HvMT4 and GST-HvMT2b2
Metal/HvMT2b2 and metal/HvMT4 competitive reaction with DTNB was applied for the purpose of investigating the DTNB accessibility to protein sulfhydryls; this is related to the structure as well as metal's binding strength to proteins ( (Toriumi et al. 2005;Park et al. 2006). The mentioned reaction was then followed by the determination of the TNB the production rate at 412 nm. As can be seen in Fig. 7A, the initial velocity related to the reaction of Cd 2+ /GST-HvMTb2 is much lower than Zn 2+ /GST-HvMT2b2 indicating Cd 2+ bind to GST-HvMT2b2 stronger than Zn 2+ . Conversely the initial velocity of Zn 2+ /GST-HvMT4 is lower than that for Cd 2+ /GST-HvMT4 which suggests the binding strength of Zn 2+ to HvMT4 is higher than Cd 2+ (Fig. 7C, D).

Discussion
MTs are Cys-rich protein serving different roles such as essential metal homeostasis, detoxification of heavy metals, and cellular anti-oxidative defense, in addition to maintaining redox status through binding and releasing metals (mostly zinc under physiological conditions) (Clemens, 2001;Hossain et al. 2012;Tauris et al. 2009;Mierek-Adamska et al. 2018).
While MT members in the plant cells are very important, there is no experimental information on the regulation of plant MTs by phytohormones. A   Fig. 4 SDS-PAGE analysis of GST, GST-HvMT2b2, and GST-HvMT4 overexpressed in E. coli Rosetta (DE3) cells. A soluble proteins extracted from E. coli harboring pGEX, pGEX-HvMT2b2, and pGEX-HvMT4. M, Protein Marker; T0, Soluble protein from cells prior to adding IPTG; T1-T4, Soluble proteins 1, 2, 3, and 4 h following the addition of IPTG. B The verification of the purification of GST-HvMT2b2 and GST-HvMT4. C The study of the solubility of GST, GST-HvMT2b2, and GST-HvMT4, based on comparing their corresponding band intensities in the total protein (T), the insoluble fraction (U), and the soluble fraction extracted from Control, R-MT2b2 and R-MT4 (S) theoretical molecular analysis on the upstream regulatory element as well as exon/intron organization of every rice MT isoforms demonstrates the presence of GA-, ABA-, IAA-, ET-and metal-responsive elements in the upstream of the which gene encodes MT isoforms (Zhou et al. 2006). However, the distribution and the presence of these elements differ between different MT genes. Separation of the barley (Hordeum vulgare) aleurone layer from the other seed tissues could be done and maintained in culture, thus leading to a special system useful for analyzing responses to plant hormones (Finnie et al. 2011). The considered aleurone layer experimental system, which was isolated, can help to analyze intracellular proteins as well as the release of hydrolytic enzymes which are accumulated in the culture supernatant (Shahpiri et al. 2008;Finnie et al. 2011]. Using this model system it has been very well documented that the expression of gene encoding Amy is induced as the response to GA (Shahpiri et al. 2015a, b;Chrispeels and Varner 1967;Siapush and Shahpiri 2017;Varty et al. 1983) but the GA-induced Amy is inhibited by ABA and SA via a pathway involving WRKY gene (Xie et al. 2007). Based on our previous study ET has no effect on the expression of Amy but the secretion of this enzyme is induced by ET (Siapush and Shahpiri 2017).
In this research study, the isolation and hormonal treatment of the aleurone layer were verified by the expression profile of HvAmy responding to hormones due to the corresponding results of the previous studies (Gómez-Cadenas, et al. 2001;Bethke et al. 1997;Shahpiri et al. 2015a, b;Siapush and Shahpiri 2017). Therefore the treated aleurone samples were used for the analysis of HvMT2b2 and HvMT4, the only two MT genes expressed in the barley aleurone layer. In contrast to HvAmy, the expression of gene encoding HvMT2b2 was suppressed by GA but instead its expression was highly induced with SA and to a lesser extent by ABA, IAA, KT, and ET. Similarly, the expression of HvMT4 was suppressed by GA and highly induced by SA. The expression of HvMT4 was not affected with ABA, IAA, and ET. However, its expression was induced slightly by KT. Therefore in spite of the similar localization of these two isoforms, their regulation does not follow a similar profile in response to different hormones.
The hormone SA seems to be an inducible effect on the expression of both HvMT4 and HvMT2b2 in the aleurone layer. SA increases the expression of HvMT4 30-fold, concerning control (untreated aleurone layer). Nevertheless, the expression of HvMT2b2 was enhanced 5.3 fold related to its corresponding control. The relation between external application of phytohormones and plant's tolerance to heavy metals and bioremediation has also been studied recently for the accumulation of phytochelatins as metal-binding peptides, (Khan et al. 2015;Bücker-Neto et al. 2017). For example, rice plants putrescine pre-treatment, when followed by cadmium stress, enhances the hormone polyamines which subsequently Exogenously applied auxin on green alga Acutodesmus obliquus increased the phytochelatin synthase activity as well as phytochelatins and their precursor's accumulation, which are important for the sequestration of Pb. Cytokinins could be described by an opposite effect on PCS when consider in association to auxin (Piotrowska-Niczyporuk et al. 2020).
It should be noted that developing cereal plants with high essential metal ion contents in their seeds have two major bottlenecks. The first one is the uptake of the essential minerals from the soil via root and shoot and the second one is the movement of the metal ions to seed during seed filling. The metal transporters, metal-ligand complex, and yellow stripe-like transporter proteins are examples of protein families which are involved in the metal uptake from root and transport to xylem epidermal cells (Palmgren et al. 2008;Stanton et al. 2021). A roadmap for trafficking zinc in the barley grain, starting from phloem unloading and continuing till deposition, has proposed the involvement of some members of protein families including P1B-type ATPase, vacuolar iron transporter, zinc-induced facilitator, cation exchanger (White and Broadley 2011;Tauris et al. 2009). Finally, for the metal ions storage in the grains, it seems that low molecular weight chelators MTs, PCs, and nicotinamines have important roles during seed development.
The expression of HvMT4 was previously analyzed during seed filling after pollination (Hegelund et al. 2012). The transcripts of this gene did not appear before full seed development. After maturation, detection of it was done in the embryo and aleurone layer. The functional analysis illustrated a strong preferential Zn 2+ binding, as compared to Cu 2+ and Cd 2+ . Here we also heterologously expressed both HvMT2b2 and HvMT4 in E. coli. GST served as an N-terminal fusion partner as it had good solubility and stability against proteolytic degradation (Hegelund et al. 2012;Tauris et al. 2009). Whereas the strain R-MT4 (expressing GST-HvMT4) showed significant tolerance to Zn 2+ , the expression of HvMT2b2 did not impact the transformed E. coli tolerance to Zn 2+ . Correspondingly the strain R-MT4 accumulated Zn 2+ tenfold more than that of the control strain. In comparison, the strain R-MT2b2 accumulated Zn 2+ three-fold higher, as compared to the control strain. When Cd 2+ was present, the strain R-MT2b2 was more tolerant than R-MT4. According to these results, the R-MT2b2 strain could accumulate three-fold more than the control strain. However, the R-MT4 strain could not accumulate Cd 2+ .
The production as well as purification of HvMT2b2 and HvMT4 in noteworthy amounts could lead to testing the capability of Zn 2+ and Cd 2+ binding to HvMT2b2 and HvMT4. It was shown by DTNB assay and UV absorption spectra that the Zn 2+ /HvMT4 and Cd 2+ /HvMT2b2 complex formation in vitro confirming the high preference of HvMT2b2 for binding to Cd 2+ and high preference of HvMT4 to Zn 2+ .
The isoforms HvMT4 and HvMT2b2 are known as two highly expressed genes in seed (Tauris et al. 2009). In another study, HvMT4 was known as a seed-specific MT with stable binding to Zn 2+ that is proposed to be a storage protein with an important function in the zinc accumulation in seed throughout development (Hegelund et al. 2012). However, the other barley HvMTs including HvMT2b2 and HvMT3 which are also localized in different organelles are not known as specific Zn 2+ -binding and are able to bind other metals like Cd 2+ . Therefore it seems that the enhancement of expression of HvMT4 during seed development may raise the zinc accumulation in the seed. Here we found that SA is a signal with high efficiency for enhancement of expression of HvMT4 in the aleurone layer when SA is applied as exogenous hormones. Subsequently, this may increase the storage of zinc in barley seeds. However since the maternal plant supplies zinc for storage in the grains, an experimental study for the application of exogenous SA on the barley during seed development is required to prove this hypothesis.

Conclusion
Barley aleurone layer was used as a model system to address the possible impacts of plant hormones on the way two isoforms MT, HvMT2b2 and HvMT4 are expressed. The specific affinity of HvMT4 protein for Zn 2+ binding compared to HvMT2b2, which could bind both Zn 2+ and Cd 2+ , may indicate that the HvMT4 isoform plays a key role in zinc storage in barley grains. Because SA induces HvMT4 expression 30-fold in the aleurone layer, the application of exogenous SA may affect the amount of zinc in the grains. However, the zinc in the seeds is supplemented from the mother plant. Therefore, to reach a definitive conclusion about the effect of SA on zinc storage, further studies are needed on the application of SA on the mother plant during seed growth.
Author contributions ZP performed the experiments in the lab and analyzed the data. AS managed the work and wrote the manuscript text. PG performed in statistical analysis of data.
Funding Iran National Science Foundation is thanked for financial support (Grant No. 97000792).

Conflict of interest
The authors declare that there are no conflicts of interest.