Overexpression of TaMYB4 Confers Freezing Tolerance in Arabidopsis thaliana

Freezing stress is one of the main factors limiting the growth and yield of wheat. In this study, we found that TaMYB4 expression was significantly upregulated in the tillering nodes of the strong cold-resistant winter wheat variety Dongnongdongmai1 (Dn1) under freezing stress. Weighted gene co-expression network analysis, qRT-PCR and protein–DNA interaction experiments demonstrated that monodehydroascorbate reductase (TaMDHAR) is a direct target of TaMYB4. The results showed that overexpression of TaMYB4 enhanced the freezing tolerance of transgenic Arabidopsis. In TaMYB4 overexpression lines (OE-TaMYB4), AtMDHAR2 expression was upregulated and ascorbate-glutathione (AsA–GSH) cycle operation was enhanced. In addition, the expression of cold stress marker genes such as AtCBF1, AtCBF2, AtCBF3, AtCOR15A, AtCOR47, AtKIN1 and AtRD29A in OE-TaMYB4 lines was significantly upregulated. Therefore, TaMYB4 may increase freezing tolerance as a transcription factor (TF) in Arabidopsis through the AsA–GSH cycle and DREB/CBF signaling pathway. This study provides a potential gene for molecular breeding against freezing stress.


Introduction
Wheat (Triticum aestivum) is the second most cultivated cereal crop worldwide. The increased human population combined with modern consumption preferences have led to a proliferation in the demand for wheat over the past 50 years [1]. However, various abiotic stresses severely limit wheat production [2]. Among them, cold stress mainly limits wheat growth and geographic distribution, and affects wheat yield and quality [3]. The winter season is long and cold in the Heilongjiang Province of China. Dongnongdongmai1 (Dn1) is the first winter wheat breed to be successfully overwintered in Heilongjiang Province [4]. A large amount of physiological indices and gene expression data reflect the distinct cold-resistance characteristics of Dn1 [5][6][7]. Nevertheless, the regulatory mechanism of transcription factors (TFs) in Dn1 under cold stress is still poorly understood.
The MYB TF family contributes prominently in terms of plant responses to abiotic stresses [8]. In apple (Malus pumila), MdMYB108L and MdHY5 shape a feedback circuit to regulate CBF3 expression, and overexpression of MdMYB108L improves apple cold tolerance [9]. In Arabidopsis thaliana, the MYB96-HHP module activates the CBF pathway involved in regulating cold tolerance in the plants [10]. In maize (Zea mays), ZmMYB31 expression is upregulated under cold stress, and overexpression of ZmMYB31 elevates SOD and APX activities in Arabidopsis [11]. In rice (Oryza sativa), the regulatory network centered on OsMYB4 enhances cellular antioxidant capacity under cold stress [12]. In wheat, 60 MYB genes have been isolated, of which, 11 gene expressions are affected by cold stress [13]. Recently, we screened TaMYB4 in Dn1, which can respond to freezing stress [14]. Nevertheless, the regulatory mechanism of MYB TFs in wheat under freezing stress is still poorly understood.
Previously, we found that TaMYB4 expression was significantly upregulated in the tillering nodes of Dn1 under freezing stress by transcriptome analysis [14]. To experimentally demonstrate the cold induction of TaMYB4 in the transcriptome data, we detected TaMYB4 expression in the tillering nodes of field-grown Dn1 by qRT-PCR (5 • C as a control). The expression of TaMYB4 in Dn1 at −10 and −25 • C was significantly higher than that at 5 • C, with fold changes of 11.60 and 11.06, respectively ( Figure 1A). However, there was only a two-fold increase in the expression of TaMYB4 at −25 • C in Jimai22 (J22) ( Figure 1B). To exclude the interference of spatiotemporal differences and other environmental factors in the field on the expression of TaMYB4, we subjected Dn1 seedlings to short-term lowtemperature treatment and performed qRT-PCR analysis (22 • C as a control). The TaMYB4 expression pattern in the tillering nodes of greenhouse-grown Dn1 was similar to that of field-grown Dn1, and TaMYB4 expression was significantly upregulated at −10 and −25 • C ( Figure 1C).
Then, we performed a promoter activity assay under cold stress. The promoter of TaMYB4 (ProTaMYB4) contained a low-temperature-responsive element ( Figure S1). The ProTaMYB4 activity was detected using the GUS reporter gene in tobacco leaves transiently transfected with pBI121-ProTaMYB4-GUS. We observed that the blue color of tobacco leaves deepened after cold treatment at 4 • C. After cold stress, the GUS activity of the tobacco leaves transiently transfected with pBI121-ProTaMYB4-GUS was also markedly higher than in leaves without cold stress ( Figure 1D). This suggests that ProTaMYB4 is induced by cold temperatures.

TaMYB4 Co-Expression Network Construction
To explore the TaMYB4 regulatory network under freezing stress, weighted gene co-expression analysis was performed. We found that six members were associated with TaMYB4 ( Figure 2A). These members included zinc finger protein (TraesCS1A02G285900), serine acetyltransferase (TraesCS3D02G286900), RING/U-box superfamily protein (TraesCS4D02G148400), RNA binding protein (TraesCS7A02G398200), kelch-type beta propeller domain containing protein (TraesCS7A02G496600) and monodehydroascorbate reductase (TraesCS7D02G277500). We observed that monodehydroascorbate reductase (MDHAR) gene expression was significantly upregulated in the Dn1 transcriptome under freezing stress, and that its expression was the highest among the six members ( Figure S2). Mean ± SD ( is used to represent values. Significant differences in panels (A-C) were calculated by one ANOVA and marked by asterisks; Significant differences in panel (D) were calculated by Stud t-test and marked by asterisks; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. CK: control c CT: cold treatment.

TaMYB4 Co-Expression Network Construction
To explore the TaMYB4 regulatory network under freezing stress, weighted gen expression analysis was performed. We found that six members were associated TaMYB4 ( Figure 2A). These members included zinc finger protein (TraesCS1A02G285 serine acetyltransferase (TraesCS3D02G286900), RING/U-box superfamily pr (TraesCS4D02G148400), RNA binding protein (TraesCS7A02G398200), kelch-type propeller domain containing protein (TraesCS7A02G496600) and monodehydroasco reductase (TraesCS7D02G277500). We observed that monodehydroascorbate redu (MDHAR) gene expression was significantly upregulated in the Dn1 transcriptome u freezing stress, and that its expression was the highest among the six members (F S2).
Then, we detected TaMDHAR expression in the tillering nodes of field-grown w wheat. TaMDHAR expression in Dn1 was significantly increased at −10 °C and −25 °C the fold change was higher than that in J22 ( Figure 2B,C). To exclude the interferen spatiotemporal differences and other environmental factors in the field on the expre of TaMDAHR, we subjected Dn1 seedlings to short-term low-temperature treatmen performed qRT-PCR analysis (22 °C as a control). The TaMDHAR expression patte greenhouse-grown Dn1 was similar to that of field-grown Dn1 ( Figure 2D). We found TaMDHAR had a similar expression pattern to TaMYB4 in Dn1 under freezing s Therefore, we speculated that there might be an association between TaMYB4 TaMDHAR under freezing stress. Mean ± SD (n = 3) is used to represent values. Significant differences in panels (A-C) were calculated by one-way ANOVA and marked by asterisks; Significant differences in panel (D) were calculated by Student's t-test and marked by asterisks; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. CK: control check, CT: cold treatment.
Then, we detected TaMDHAR expression in the tillering nodes of field-grown winter wheat. TaMDHAR expression in Dn1 was significantly increased at −10 • C and −25 • C, and the fold change was higher than that in J22 ( Figure 2B,C). To exclude the interference of spatiotemporal differences and other environmental factors in the field on the expression of TaMDAHR, we subjected Dn1 seedlings to short-term low-temperature treatment and performed qRT-PCR analysis (22 • C as a control). The TaMDHAR expression pattern of greenhouse-grown Dn1 was similar to that of field-grown Dn1 ( Figure 2D). We found that TaMDHAR had a similar expression pattern to TaMYB4 in Dn1 under freezing stress. Therefore, we speculated that there might be an association between TaMYB4 and TaMDHAR under freezing stress.

Bioinformatics Analysis and Subcellular Localization of TaMYB4 and TaMDHAR
We cloned TaMYB4 and TaMDHAR sequences from Dn1 cDNA. Multiple sequence alignment showed that TaMYB4 contained two typical conserved adjacent repeats in the MYB domain and belongs to the R2R3 subgroup ( Figure S3A). The phylogenetic tree that was generated showed that TaMYB4 clustered with TaMYB-1D and ZmMYB31 ( Figure S3B). Multiple sequence alignments revealed that TaMDHAR is highly similar to MDHAR in other plants and contains a typical Pyr_redox_2 domain ( Figure S3C). The phylogenetic tree constructed from these proteins showed that TaMDHAR clustered with AeMDHAR ( Figure S3D). is used to represent values. One-way ANOVA was used to calculate significant differences, which are marked by asterisks; *, p < 0.05; ***, p < 0.001; ****, p < 0.0001.

Bioinformatics Analysis and Subcellular Localization of TaMYB4 and TaMDHAR
We cloned TaMYB4 and TaMDHAR sequences from Dn1 cDNA. Multiple sequence alignment showed that TaMYB4 contained two typical conserved adjacent repeats in the MYB domain and belongs to the R2R3 subgroup ( Figure S3A). The phylogenetic tree that was generated showed that TaMYB4 clustered with TaMYB-1D and ZmMYB31 ( Figure  S3B). Multiple sequence alignments revealed that TaMDHAR is highly similar to MDHAR in other plants and contains a typical Pyr_redox_2 domain ( Figure S3C). The phylogenetic tree constructed from these proteins showed that TaMDHAR clustered with AeMDHAR ( Figure S3D). Subcellular localization prediction analysis revealed that TaMYB4 might localize to the nucleus, whereas TaMDHAR might localize to the cytoplasm ( Figure S3E,F). Subsequently, we found that the enhanced green fluorescent protein (EGFP) fluorescence signal was mainly detected in the nucleus in tobacco leaves containing the TaMYB4-EGFP plasmid. And the EGFP fluorescence signal was mainly detected in the cytoplasm in tobacco leaves containing the TaMDHAR-EGFP plasmid ( Figure 3). To further determine the localization of TaMYB4 and TaMDHAR, we isolated nuclear and cytosolic proteins from tobacco leaves expressing TaMYB4-EGFP and TaMDHAR-EGFP, respectively. Then, we performed Western blot analysis. The results showed that TaMYB4 and TaMDHAR were detected in the nucleus and cytoplasm using anti-EGFP antibody, respectively ( Figure 4). TaMDHAR expression in greenhouse-grown Dn1. Mean ± SD (n = 3) is used to represent values. One-way ANOVA was used to calculate significant differences, which are marked by asterisks; *, p < 0.05; ***, p < 0.001; ****, p < 0.0001. Subcellular localization prediction analysis revealed that TaMYB4 might localize to the nucleus, whereas TaMDHAR might localize to the cytoplasm ( Figure S3E,F). Subsequently, we found that the enhanced green fluorescent protein (EGFP) fluorescence signal was mainly detected in the nucleus in tobacco leaves containing the TaMYB4-EGFP plasmid. And the EGFP fluorescence signal was mainly detected in the cytoplasm in tobacco leaves containing the TaMDHAR-EGFP plasmid ( Figure 3). To further determine the localization of TaMYB4 and TaMDHAR, we isolated nuclear and cytosolic proteins from tobacco leaves expressing TaMYB4-EGFP and TaMDHAR-EGFP, respectively. Then, we performed Western blot analysis. The results showed that TaMYB4 and TaMDHAR were detected in the nucleus and cytoplasm using anti-EGFP antibody, respectively ( Figure 4).

TaMYB4 Regulates TaMDHAR Expression under Cold Stress
Eight MYB binding sites were present in the isolated 2000 bp promoter of TaMDHAR (ProTaMDHAR) ( Figure S1B). TaMDHAR may be a direct target of TaMYB4. To validate this regulatory relationship, a yeast one-hybrid assay was performed. The eight MYB binding sites in ProTaMDHAR were classified into five categories, including MRE (AACCTAA), MBS (CAACTG), Myb (CAACAG), MYBrs (CCGTTG) and MYBbs (TAACCA). Full-length ProTaMDHAR and three tandem copies of the motif sequence were cloned into the pHis2 vector. Then, the yeast strain Y187 was co-transformed by the pHis2 recombinant plasmid with the pGADT7-TaMYB4 recombinant plasmid. We found that TaMYB4 can recognize five MYB binding sites and has stronger binding activity to the Myb motif ( Figure 5).

TaMYB4 Regulates TaMDHAR Expression under Cold Stress
Eight MYB binding sites were present in the isolated 2000 bp promoter of TaMDHAR (ProTaMDHAR) ( Figure S1B). TaMDHAR may be a direct target of TaMYB4. To validate this regulatory relationship, a yeast one-hybrid assay was performed. The eight MYB binding sites in ProTaMDHAR were classified into five categories, including MRE (AAC-CTAA), MBS (CAACTG), Myb (CAACAG), MYBrs (CCGTTG) and MYBbs (TAACCA). Full-length ProTaMDHAR and three tandem copies of the motif sequence were cloned

TaMYB4 Regulates TaMDHAR Expression under Cold Stress
Eight MYB binding sites were present in the isolated 2000 bp promoter of TaM (ProTaMDHAR) ( Figure S1B). TaMDHAR may be a direct target of TaMYB4. To v this regulatory relationship, a yeast one-hybrid assay was performed. The eigh binding sites in ProTaMDHAR were classified into five categories, including MRE CTAA), MBS (CAACTG), Myb (CAACAG), MYBrs (CCGTTG) and MYBbs (TAA Full-length ProTaMDHAR and three tandem copies of the motif sequence were   Subsequently, we performed a GUS transient expression experiment. We truncated ProTaMDHAR according to the distribution of the MYB binding sites ( Figure 6A). The full-length and truncated sequences of ProTaMDHAR were cloned into the pBI121 vector. Then, tobacco leaves were transiently transfected with PBI121 recombinant plasmid alone or together with 35S-TaMYB4 recombinant plasmid. After 3 d of culture, the tobacco was divided into control and cold-treated (4 • C 2 h) groups. The GUS staining results showed that the blue color of the tobacco leaves gradually became lighter with shortening of the promoter sequence. The P1 and ProMDHAR sequences presented a low-temperature-responsive element. The color of tobacco leaves transiently expressing pBI121-ProTaMDHAR-GUS and pBI121-P1-GUS deepened under cold stress. Moreover, the addition of the 35S-TaMYB4 recombinant plasmid noticeably deepened the color of the tobacco leaves ( Figure 6B). We subsequently assayed GUS activity in tobacco leaves, and the GUS activity change was in line with the tobacco leaf color change ( Figure 6C).  Mean ± SD (n = 3) is used to represent values. One-way ANOVA was u calculate significant differences, which are marked by asterisks; *, p < 0.05; **, p < 0.01; ***, p < ****, p < 0.0001. CK: control check, CT: cold treatment.

Characterization of TaMYB4 and TaMDHAR Function under Freezing Stress
To characterize the functions of TaMYB4 and TaMDHAR under freezing stress, we constructed expression vectors (35S-TaMYB4 and 35S-TaMDHAR) and transformed them into Arabidopsis. After several cycles of kanamycin-resistance screening, TaMYB4 and TaMDHAR overexpression lines (OE-TaMYB4-1, OE-TaMYB4-6, OE-TaMDHAR-3 and OE-TaMDHAR-4) were obtained. We observed the phenotypes of the Arabidopsis lines (WT, OE-TaMYB4 and OE-TaMDHAR) under freezing stress ( Figure 8A). After −10 • C treatment, curling and a deepening in color occurred in the leaves of the three lines, but the WT had a deeper leaf color than the OE-TaMYB4 and OE-TaMDHAR lines. After recovery for 7 d, the OE-TaMYB4 and OE-TaMDHAR lines mostly recovered to a normal growth state, whereas nearly all of the WT died. Ultimately, the OE-TaMYB4 and OE-TaMDHAR lines had higher survival rates than WT ( Figure 8B). Then, we determined the electrical conductivity (EC), malondialdehyde (MDA) and proline (Pro) contents. The results showed that the OE-TaMYB4 and OE-TaMDHAR lines had a lower EC and MDA content and a higher Pro content than WT at 4 • C and −10 • C. Altogether, the OE-TaMYB4 and OE-TaMDHAR lines showed equally excellent physiological properties under freezing stress ( Figure 9). TaMDHAR overexpression lines (OE-TaMYB4-1, OE-TaMYB4-6, OE-TaMDHAR-3 and OE-TaMDHAR-4) were obtained. We observed the phenotypes of the Arabidopsis lines (WT, OE-TaMYB4 and OE-TaMDHAR) under freezing stress ( Figure 8A). After −10 °C treatment, curling and a deepening in color occurred in the leaves of the three lines, but the WT had a deeper leaf color than the OE-TaMYB4 and OE-TaMDHAR lines. After recovery for 7 d, the OE-TaMYB4 and OE-TaMDHAR lines mostly recovered to a normal growth state, whereas nearly all of the WT died. Ultimately, the OE-TaMYB4 and OE-TaMDHAR lines had higher survival rates than WT ( Figure 8B). Then, we determined the electrical conductivity (EC), malondialdehyde (MDA) and proline (Pro) contents. The results showed that the OE-TaMYB4 and OE-TaMDHAR lines had a lower EC and MDA content and a higher Pro content than WT at 4 °C and −10 °C. Altogether, the OE-TaMYB4 and OE-TaMDHAR lines showed equally excellent physiological properties under freezing stress ( Figure 9). TaMDHAR had high homology with Arabidopsis cytoplasmic subtype AtMDHAR2 ( Figure S3D). MYB binding sites were present in the AtMDHAR2 promoter ( Figure S4). OE-TaMYB4 lines showed higher AtMDHAR2 expression and MDHAR activity than WT at 4 °C and −10 °C ( Figure 10). is used to represent values. Student's t-test was used to calculate significant differences, which are marked by asterisks; ****, p < 0.0001. is used to represent values. Student's t-test was used to calculate significant differences, which are marked by asterisks; ****, p < 0.0001.

ROS Scavenging Capacity Is Increased in OE-TaMYB4 and OE-TaMDHAR Arabidopsis under Freezing Stress
To demonstrate that overexpression of TaMYB4 and TaMDHAR confers stronger ROS scavenging capacity in Arabidopsis, we performed the 3,3 -diaminobenzidine (DAB) and nitroblue tetrazolium (NBT) staining. We found that the WT had a deeper leaf color than the OE-TaMYB4 and OE-TaMDHAR lines under freezing stress ( Figure 11A,B). To further demonstrate the ROS changes in the WT, OE-TaMYB4 and OE-TaMDHAR lines under freezing stress, we determined the hydrogen peroxide (H 2 O 2 ) and superoxide anion (O 2 •− ) contents, which were higher in the WT than those in the OE-TaMYB4 and OE-TaMDHAR lines under freezing stress ( Figure 11C,D).     nitroblue tetrazolium (NBT) staining. We found that the WT had a deeper leaf color than the OE-TaMYB4 and OE-TaMDHAR lines under freezing stress ( Figure 11A,B). To further demonstrate the ROS changes in the WT, OE-TaMYB4 and OE-TaMDHAR lines under freezing stress, we determined the hydrogen peroxide (H2O2) and superoxide anion (O2 .-) contents, which were higher in the WT than those in the OE-TaMYB4 and OE-TaMDHAR lines under freezing stress ( Figure 11C,D).

The AsA-GSH Cycle Is Improved in OE-TaMYB4 and OE-TaMDHAR Arabidopsis under Freezing Stress
To further explore whether ROS scavenging was dependent on the AsA-GSH cycle in the OE-TaMYB4 and OE-TaMDHAR lines, we identified the expression and activity of antioxidant enzymes related to the AsA-GSH cycle. The results showed that the OE-TaMYB4 and OE-TaMDHAR lines had a higher expression of AtSOD1, AtSOD2, AtAPX1, AtAPX6, AtDHAR1, AtDHAR2, AtGR1 and AtGR2 than WT at 4 °C and −10 °C. OE-TaMYB4 and OE-TaMDHAR lines also had higher SOD, APX, DHAR and GR activities than WT at 4 °C and −10 °C. At normal and low temperatures, OE-TaMDHAR lines had a higher MDHAR activity than WT (Figures 12 and 13).
We determined the content of antioxidant substances in the OE-TaMYB4 and OE-TaMDHAR lines. The results showed that the OE-TaMYB4 and OE-TaMDHAR lines had Figure 11. Detection of ROS content in the OE-TaMYB4 and OE-TaMDHAR lines under freezing stress. DAB (A) and NBT (B) staining of Arabidopsis leaves; H 2 O 2 (C) and O 2 · -(D) contents in Arabidopsis leaves. Mean ± SD (n = 3) is used to represent values. Two-way ANOVA was used to calculate significant differences, which are marked by asterisks; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

The AsA-GSH Cycle Is Improved in OE-TaMYB4 and OE-TaMDHAR Arabidopsis under Freezing Stress
To further explore whether ROS scavenging was dependent on the AsA-GSH cycle in the OE-TaMYB4 and OE-TaMDHAR lines, we identified the expression and activity of antioxidant enzymes related to the AsA-GSH cycle. The results showed that the OE-TaMYB4 and OE-TaMDHAR lines had a higher expression of AtSOD1, AtSOD2, AtAPX1, AtAPX6, AtDHAR1, AtDHAR2, AtGR1 and AtGR2 than WT at 4 • C and −10 • C. OE-TaMYB4 and OE-TaMDHAR lines also had higher SOD, APX, DHAR and GR activities than WT at 4 • C and −10 • C. At normal and low temperatures, OE-TaMDHAR lines had a higher MDHAR activity than WT (Figures 12 and 13).
We determined the content of antioxidant substances in the OE-TaMYB4 and OE-TaMDHAR lines. The results showed that the OE-TaMYB4 and OE-TaMDHAR lines had higher AsA, DHA, GSH and GSSG contents than WT at 4 • C and −10 • C. The OE-TaMYB4 and OE-TaMDHAR lines also had higher AsA/DHA and GSH/GSSG than WT at 4 • C and −10 • C ( Figure 14). higher AsA, DHA, GSH and GSSG contents than WT at 4 °C and −10 °C. The OE-TaMYB4 and OE-TaMDHAR lines also had higher AsA/DHA and GSH/GSSG than WT at 4 °C and −10 °C (Figure 14).  Mean ± SD (n = 3) is used to represent values. Two-way ANOVA was used to calculate significant differences, which are marked by asterisks (24 • C WT as a control); *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

Cold-Responsive Genes Expression Is Affected in OE-TaMYB4 Arabidopsis under Freezing Stress
AtCBF1, AtCBF2, AtCBF3, AtCOR15a, AtCOR47, AtKIN1 and AtRD29A are typical cold stress marker genes [24]. We examined the expression of these genes in OE-TaMYB4 lines. The results showed that the OE-TaMYB4 lines had a higher expression of AtCBF1, AtCBF2 and AtCBF3 than WT at 24 • C and −10 • C. And the OE-TaMYB4 lines had a higher expression of AtCOR15a, AtCOR47, AtKIN1 and AtRD29A than WT at −10 • C. However, at 4 • C, the expression of these genes in OE-TaMYB4 was not affected or significantly lower than those of WT ( Figure 15). AsA/DHA; (F) GSH/GSSG. Mean ± SD (n = 3) is used to represent values. Two-way ANOVA was used to calculate significant differences, which are marked by asterisks; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

Cold-Responsive Genes Expression Is Affected in OE-TaMYB4 Arabidopsis under Freezing Stress
AtCBF1, AtCBF2, AtCBF3, AtCOR15a, AtCOR47, AtKIN1 and AtRD29A are typical cold stress marker genes [24]. We examined the expression of these genes in OE-TaMYB4 lines. The results showed that the OE-TaMYB4 lines had a higher expression of AtCBF1, AtCBF2 and AtCBF3 than WT at 24 °C and −10 °C. And the OE-TaMYB4 lines had a higher expression of AtCOR15a, AtCOR47, AtKIN1 and AtRD29A than WT at −10 °C. However, at 4 °C, the expression of these genes in OE-TaMYB4 was not affected or significantly lower than those of WT ( Figure 15). (F) and AtRD29A (G) in the OE-TaMYB4 lines under freezing stress. Mean ± SD (n = 3) is used to represent values. Two-way ANOVA was used to calculate significant differences, which are marked by asterisks (24 °C WT as a control); *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

Discussion
In multiple plants, MYB4 has been shown to be able to respond to varying degrees of cold stress. The OsMYB4 expression was upregulated after 5 h of cold treatment (4 °C) [25]. The LcMYB4 expression was consistently upregulated in Leymus chinensis within 24 (F) and AtRD29A (G) in the OE-TaMYB4 lines under freezing stress. Mean ± SD (n = 3) is used to represent values. Two-way ANOVA was used to calculate significant differences, which are marked by asterisks (24 • C WT as a control); *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

Discussion
In multiple plants, MYB4 has been shown to be able to respond to varying degrees of cold stress. The OsMYB4 expression was upregulated after 5 h of cold treatment (4 • C) [25]. The LcMYB4 expression was consistently upregulated in Leymus chinensis within 24 h of cold treatment (4 • C) [26]. Under chilling (4 • C) and freezing stress (−4 • C), MYB4 expression is significantly upregulated in different apple varieties [27,28]. After cold treatment at 4 • C for 2 h, TaMYB4 expression is significantly upregulated in spring wheat [29]. However, the expression and function of TaMYB4 in wheat under freezing stress are still unknown. The emergence of Dn1 provides a possibility to explore the freezing tolerance functions of wheat MYB TFs. In the Dn1 transcriptome, we found that several MYB TFs were differentially expressed under freezing stress. Among them, only the transcriptional level of TaMYB4 was consistently upregulated with the decline in temperature (with the minimum temperature being −25 • C) [14]. In this research, TaMYB4 expression was significantly upregulated in Dn1 under freezing stress ( Figure 1A). Together with the findings of previous studies, this suggests that wheat TaMYB4 may function under both chilling and freezing stress. When the winter field temperature falls to −10 • C, the molecular, physiological and biochemical levels of Dn1 begin to change significantly [5,30,31]. In this research, TaMYB4 expression was initiated at −10 • C in Dn1, functioning sooner than TaMYB4 in J22. TaMYB4 expression was significantly upregulated in field-grown J22 at −20 • C and −25 • C, but the fold change was significantly lower than that of Dn1 ( Figure 1B). This suggests that Dn1 might initiate cold stress defense mechanisms sooner than J22. Our previous study effectively excluded the influence of temporal and spatial differences and other environmental factors on gene expression in Dn1 through short-term low-temperature treatment in the greenhouse [7]. In this research, the TaMYB4 expression pattern in Dn1 under short-term low-temperature treatment in the greenhouse was similar to that of Dn1 in the field ( Figure 1C). This suggests that TaMYB4 functions under both long-term and short-term freezing stress, and that its expression may be mainly induced by freezing stress.
In plants, MYB TFs typically function by regulating downstream gene expression under cold stress. The regulatory mechanisms of AtMYB15 and AtMYB96 in Arabidopsis have been well characterized. MYB15 can either directly inhibit CBF expression or interact with ICE1 to regulate CBF expression [32]. MYB96 positively regulates HHPs expression under cold stress. HHP proteins subsequently activate CBF expression by interacting with ICE1 [10]. The above studies indicate that MYB TFs affect plant cold tolerance by regulating the CBF signaling pathway. However, it is rarely reported that MYB TFs directly regulate antioxidant enzyme genes. In this study, we had several lines of evidence suggesting that TaMYB4 directly regulates TaMDHAR expression in Dn1 under freezing stress. First, WGCNA revealed an association between TaMYB4 and TaMDHAR ( Figure 2A). Second, TaMYB4 and TaMDHAR had similar expression patterns under freezing stress (Figures 1 and 2). Third, TaMYB4 was found to be located in the nucleus and to possess transcription activation properties (Figures 3 and 4). Finally, TaMYB4 was found to recognize and bind to five MYB binding sites in the TaMDHAR promoter (Figures 5-7). Therefore, we speculate that TaMYB4 directly regulates TaMDHAR expression in Dn1 in response to freezing stress.
Extremely low temperatures not only cause freezing damage to plants but also result in the accumulation of intracellular ROS. If the antioxidant defense system fails to maintain the ROS level in a normal state, oxidative damage will occur. Excessive ROS leads to increased MDA content and electrical conductivity, which are biomarkers of cell membrane damage [33]. Overexpressing OsMYB4 can reduce MDA content and electrical conductivity, thereby enhancing the cold tolerance of Arabidopsis, tomato and apple [25,[34][35][36]. The overexpression of pgMYB4 and LcMYB4 can increase Pro content and confer stronger freezing tolerance to Arabidopsis [26,37]. In this study, the OE-TaMYB4 lines had a higher Pro content and lower EC, MDA content and ROS content than WT under freezing stress (Figures 9 and 11). These results coincide with those of previous studies and illustrate that TaMYB4 is able to improve plant freezing tolerance.
In plants, the antioxidant defense system consists of nonenzymatic and enzymatic antioxidant substances, which maintain the homeostasis of ROS. The enzymatic antioxidants comprise APX, MDHAR, DHAR, GR, GPX, PRX, CAT and SOD, while the nonenzymatic antioxidants comprise AsA, GSH, carotenoids, tocopherols, flavonoids, etc. Among them, AsA, GSH, APX, DHAR, MDHAR and GR constitute the well-known AsA-GSH cycle [38]. MDHAR is a momentous enzyme that maintains the reduced pool of AsA in this cycle [15]. In this study, we found that AtMDHAR2 shares the highest homology with TaMDHAR ( Figure S3D) and contains MYB binding sites in its promoters ( Figure S4). OE-TaMYB4 lines had higher AtMDHAR2 expression and MDHAR activity than WT under freezing stress ( Figure 10). Therefore, we speculate that TaMYB4 can control endogenous AtMDHAR2 expression, thereby enhancing the MDHAR activity of Arabidopsis under freezing stress. This suggests that the regulatory relationship between TaMYB4 and MDHAR may be conserved in plants. We found that overexpression of TaMYB4 could promote the expression of antioxidant enzyme genes in the AsA-GSH cycle under freezing stress ( Figure 12). In addition, OE-TaMYB4 lines had higher antioxidant enzyme activities, AsA/DHA and GSH/GSSG in the AsA-GSH cycle than WT under freezing stress. These results were similar to the changes in physiological indices in the OE-TaMDHAR lines under freezing stress (Figures 13 and 14). This suggests that TaMYB4 may regulate MDHAR expression to promote AsA-GSH cycle operation in Arabidopsis. Sun et al. found that the ectopic expression of BrMDHAR confers stronger freezing tolerance by inducing coregulation of the AsA-GSH cycle and enhancing the antioxidant capacity of host plants [39], which was consistent with our findings. The DREB/CBF signaling pathway plays an important role in plant cold resistance [24]. In this study, the OE-TaMYB4 lines had a higher expression of CBF1, CBF2, CBF3 COR15a, COR47, KIN1 and RD29A than WT at −10 • C. And the expression of CBF1, CBF2 and CBF3 in the OE-TaMYB4 lines was also higher than those in WT at 24 • C ( Figure 15). This result was not completely consistent with previous reports, indicating that TaMYB4 has specific regulation on these cold-responsive genes. However, the expression of these cold-responsive genes decreased in the OE-TaMYB4 lines after 3 d of treatment at 4 • C. Previous reports have shown that the expression of these genes peaked after 3 or 24 h of treatment at 4 • C, and then gradually decreased [40,41]. This is consistent with our results. The suppression of the gene expression in the OE-TaMYB4 lines at 4 • C may be to prevent the excessive defense of plants against chilling stress [42]. This suggests that TaMYB4 may affect the freezing tolerance of plants by regulating the expression of cold-responsive genes.
In conclusion, the heterologous expression of TaMYB4 promotes the AsA-GSH cycle operation and is associated with the upregulation of cold-responsive genes, thereby improving the freezing tolerance of transgenic Arabidopsis ( Figure 16). Our findings on the functional role of TaMYB4 will allow additional insight for developing freezing tolerant crops.

Plant Materials
The growing region and cultivation method of winter wheat referred to that used in previous studies [5]. Briefly, Dn1 and J22 (weakly cold-resistant variety) seeds were sown in the experimental field on 10 September 2019. When the average value of the lowest temperature of ten consecutive days reached 5 °C (1 October 2019), 0 °C (1 November 2019), −10 °C (18 November 2019) and −25 °C (31 December 2019), the tillering nodes were sampled. For the greenhouse component of the study, the winter wheat indoor cultivation method from previous studies was improved upon [7]. Briefly, Dn1 seedlings were cultivated in a 22 °C greenhouse (15 d) and subsequently transferred to a 5 °C incubation chamber for cold acclimation (30 d). Then, the temperature was gradually decreased and the seedlings were treated for 1 d each at 0, −10 and −25 °C. The tillering nodes of wheat were sampled at the above temperatures.

Plant Materials
The growing region and cultivation method of winter wheat referred to that used in previous studies [5]. Briefly, Dn1 and J22 (weakly cold-resistant variety) seeds were sown in the experimental field on 10 September 2019. When the average value of the lowest temperature of ten consecutive days reached 5 • C (1 October 2019), 0 • C (1 November 2019), −10 • C (18 November 2019) and −25 • C (31 December 2019), the tillering nodes were sampled. For the greenhouse component of the study, the winter wheat indoor cultivation method from previous studies was improved upon [7]. Briefly, Dn1 seedlings were cultivated in a 22 • C greenhouse (15 d) and subsequently transferred to a 5 • C incubation chamber for cold acclimation (30 d). Then, the temperature was gradually decreased and the seedlings were treated for 1 d each at 0, −10 and −25 • C. The tillering nodes of wheat were sampled at the above temperatures.
The Arabidopsis (Colombia ecotype) cultivation method was described in previous studies [4]. Briefly, Arabidopsis seedlings were cultivated in a greenhouse at 24 • C for 28 d and subsequently transferred to a 4 • C incubation chamber for 3 d of cold acclimation. After cold acclimation, the seedlings were subjected to 2 h of cold treatment at −10 • C. The leaves were sampled at 24, 4 and −10 • C. All plant materials mentioned above were stored at −80 • C. After 7 d of recovery culture, the survival rate was determined.

Quantitative Real-Time PCR
Total RNA was isolated from plants with TRIzol reagent (CWBIO, Beijing, China). The cDNA was subsequently obtained from mRNA with a HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China). Quantitative real-time PCR (qRT-PCR) was used for gene quantification analysis, and the detailed experimental methods referred to those used previous studies [6]. The data were analyzed by the 2 −∆∆CT method.
Subcellular localization prediction, homologous sequence alignment and phylogenetic analysis of TaMYB4 and TaMDHAR were performed by referring to the method of previous studies [4].

Subcellular Localization Analysis
The CDS of TaMYB4 and TaMDHAR were constructed into the pCAMBIA2300-EGFP vector. Tobacco leaves (30 days old) were then injected with Agrobacterium tumefaciens GV3101 harboring pCAMBIA2300-EGFP, pCAMBIA2300-TaMYB4-EGFP or pCAMBIA2300-TaMDHAR-EGFP plasmids. After 72 h of injection, the leaves were harvested to observe the EGFP signal under a fluorescence microscope. Western blot analysis was performed to further analyze the subcellular localization results. The detailed experimental methods refer to those used by the authors in [43]. Anti-EGFP antibody (1:1000, Proteintech, Wuhan, China) was used as the primary antibody to probe EGFP. Horseradish peroxidase (HRP)conjugated anti-rabbit antibody was chosen as the secondary antibody (1:2000, Proteintech).

Yeast One-Hybrid Assay
TaMYB4 was constructed into the pGADT7 vector as an effector. There are five MYB binding sites (MRE, MBS, Myb, MYBrs and MYBbs) in the 2000 bp promoter sequence of TaMDHAR. Three tandem copies of MYB binding sites and the 2000 bp promoter sequence of TaMDHAR were separately constructed into the pHis2 vector as reporters. The protein-DNA interactions were analyzed according to the growth status of the Y187 strain containing reporter and effector on triple dropout synthetically defined (SD) medium lacking Leu, Trp and His containing 60 mM 3-AT.

Transient Expression Assay in Tobacco
To analyze the promoter activity and validate the regulatory relationship between TaMYB4 and TaMDHAR, a transient expression assay for tobacco (Nicotiana tabacum) was performed. The sequence of the TaMYB4 promoter (2000 bp) was constructed into the pBI121-GUS vector. Tobacco leaves (30 days old) were then injected with A. tumefaciens GV3101 harboring the pBI121-ProTaMYB4-GUS plasmid. After 72 h of injection, the tobacco was subjected to 2 h of cold treatment at 4 • C. The leaves were collected for histochemical staining. Then, the GUS gene quantitative detection kit (Coolaber, Beijing, China) was applied for determining GUS activity in tobacco leaves.
The TaMDHAR promoter was truncated according to the MYB binding site distribution, with the 2000 bp promoter sequence and truncation sequences (P1, P2, P3 and P4) constructed into the pBI121-GUS vector as reporters. The TaMYB4 overexpression construct was used as an effector. Tobacco leaves (30 days old) were then injected with A. tumefaciens GV3101 harboring the reporter and effector in a 1:1 ratio. After 72 h of injection, the tobacco was subjected to 2 h of cold treatment at 4 • C. The leaves were collected for histochemical staining and GUS activity assays.
The sequence of TaMDHAR promoter (2000 bp) and truncation sequences were also constructed into the pGreen II 0800-LUC vector. The TaMYB4 overexpression construct was used as an effector. Tobacco leaves (30 days old) were then injected with A. tumefaciens GV3101 (pSoup) harboring the reporter and effector in a 1:1 ratio. After 48 h of injection, the tobacco was subjected to 2 h of cold treatment at 4 • C. The LUC signal was visualized with a CCD system. Then, the Dual Luciferase Reporter Gene Assay Kit (Yeasen, Shanghai, China) was applied for determining LUC and REN activities in tobacco leaves.

Physiological Indices Determination
The content of H 2 O 2 in Arabidopsis leaves was assessed by DAB staining, and the content of O 2 •− was assessed by NBT staining. The detailed experimental methods refer to those used by the authors in [44]. H 2 O 2 and O 2 •− contents were determined using a commercial assay kit (Comin, Suzhou, China), and the specific procedures were carried out as per the manufacturer's instructions. Briefly, the H 2 O 2 content was determined by monitoring the absorbance of the titanium-peroxide complex formed by H 2 O 2 and titanium sulfate at 415 nm [45]. The O 2 •− , hydroxylamine hydrochloride, p-aminobenzenesulfonic acid and α-Naphthylamine reaction generates azo compounds. The O 2 •− content was determined by monitoring the absorbance of azo compounds at 530 nm [46].
The MDA content and EC determination method was carried out in accordance with a previous study [47]. Pro content was determined using a kit (Comin). Briefly, the pro content was determined by monitoring the absorbance of the red substance formed by the reaction of proline and acid ninhydrin at 520 nm [48].
Antioxidant enzyme activity and non-enzymatic antioxidant content were also determined using kits (Comin), and the specific procedures were carried out in accordance with the manufacturer's instructions. Briefly, SOD can remove the O 2 •− in the reaction system. SOD activity was determined by monitoring the absorbance of blue formazan generated by the reaction of the remaining O 2 •− with nitro-blue tetrazolium at 560 nm [49]. APX catalyzes the oxidation reaction of AsA and H 2 O 2 . The APX activity was calculated by measuring the AsA oxidation rate [50]. DHAR catalyzes the reduction reaction of GSH and DHA. DHAR activity was calculated by measuring the rate of DHA reduction [51]. MDHAR catalyzes the reduction reaction of NADPH and MDHA. MDHAR activity was determined by monitoring the absorbance of NADPH at 340 nm [52]. GR catalyzes the reduction reaction of NADPH and GSSG. GR activity was determined by monitoring the absorbance of NADPH at 340 nm [53].
The AsA content was determined by monitoring the absorbance of the substance generated by the reaction of AsA and o-Dianisidine bis (diazotized) zinc double salt at 420 nm [54]. Dithiothreitol reduces DHA to produce AsA. The DHA content was calculated by measuring the rate of AsA formation in the reaction system [55]. The GSH content was determined by monitoring the absorbance of the complex formed by the reaction of 5,5 -Dithiobis-(2-nitrobenzoic acid) and GSH at 412 nm. GR catalyzes the reduction of GSSG to generate GSH in the reaction system. The GSSG content was determined by monitoring the absorbance of the complex formed by the reaction of 5,5 -Dithiobis-(2-nitrobenzoic acid) and GSH at 412 nm [56]. The experimental data were calculated with fresh weight.

Statistical Analysis
Each experiment was performed with at least three biological replicates. All data, mean ± SD (standard deviation), were analyzed using the Student's t-test and ANOVA (analysis of variance) with GraphPad Prism 9.0 at significance levels of * p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001.