Response of wheat DREB transcription factor to drought stress based on DNA methylation

Background: The growth and development of wheat are seriously inuenced by drought stress, and the research on drought resistance mechanism of wheat is very important. Dehydration responsive element binding protein (DREB) plays an important role in the response of plant to drought stress, but epigenetic regulation for gene expression of DREB transcription factor is less studied, especially the regulatory role of DNA methylation has not been reported. Results: In this research, DREB2, DREB6 and Wdreb2 were cloned from wheat, their CDS sequences were composed of 732 bp, 837 bp or 1035 bp, respectively, and one 712 bp intron was found in DREB6. Although AP2/EREBP domain of DREB2, DREB6 and Wdreb2 had 73.25% identity, they belong to different types of DREB transcription factor, and the expression of Wdreb2 was signicantly higher, yet was the lowest in DREB2. Under drought stress, the expression of DREB2, DREB6 and Wdreb2 could be induced, but had different trends along with the increase of stress time, and their expression had tissue specicity, was obviously higher in leaf. Promoter of DREB2, DREB6 and Wdreb2 in leaf was further studied, some elements related to adverse stress were found, and the promoter of DREB2 and Wdreb2 was slightly methylated, but DREB6 promoter was moderately methylated. Compared with the control, the level of promoter methylation in DREB2 and DREB6 decreased as stressed for 2 h, then increased along with the increase of stress time, which was opposite in Wdreb2 promoter, the status of promoter methylation in DREB2, DREB6 and Wdreb2 also had signicant change under drought stress. Further analysis showed that promoter methylation of DREB6 and Wdreb2 was negatively correlated with their expression, especially was signicant in Wdreb2. Conclusions: DREB2, DREB6 and Wdreb2 might function differently in response to drought stress, and promoter methylation had more signicant effects on gene expression of Wdreb2 and DREB6. this study, expression level of genes, methylation ratio of promoters were tested by signicance level, ANOVA and multiple comparisons of Duncan’s multiple range, and the correlation between gene expression and promoter methylation was analyzed by Pearson coecient r of SPSS software.

of plants to adverse stress [6]. For example, Arabidopsis DREB1/CBF can regulate the expression of rd29A, erd10, cor6.6, cor15a, rd17 and other stress-resistant genes to drought, low temperature and so on [8], overexpression of DREB1A in transgenic Arabidopsis can enhance the expression of downstream target genes, and drought tolerance of transgenic Arabidopsis would signi cantly increase [9]. At present, many genes encoding DREB transcription factor have been cloned from Arabidopsis, Maize, Soybean, Sesame, etc., its expression could be induced and would increase rapidly in a short time under abiotic stress [10], however the mechanism on expression regulation of DREB transcription factor is less studied, especially epigenetic regulation has not yet been reported. DNA methylation is a major epigenetic modi cation, and plays vital function in the growth and development of plants [11], but DNA methylation of plants is easily affected by physiological status, developmental stage and environmental factors [12]. Under drought stress, the state of DNA methylation would change in plants, for example, methylation level improved and methylation pattern was signi cantly different at different development stages of Rice [13], while methylation level of Ryegrass decreased and the expression of demethylated related genes was up-regulated [14]. Fan et al. also found that methylation level of Dendrobium huoshanense decreased, and methylation polymorphism gradually increased along with the increase of drought stress [15]. Some studies have shown that DNA methylation plays an important role in the response of plants to adverse stress, could regulate the expression of stress-resistant genes by changes of DNA methylation, and then would improve the resistance of plants to adverse stress [16]. For example, the physiological processes of Rice in response to drought stress were related to DNA methylation [13], the change of DNA methylation status was closely connected with drought resistance of trees [17], furthermore, methylation or demethylation of gene in plants would lead to the difference of gene expression under drought stress [18], and this changes of gene expression mediated by DNA methylation would make plants escape or endure drought stress [19].
Wheat (Triticum aestivum L.) belongs to Gramineae, is rich in starch, protein, sugar and other substances, and is one of main food crops. In recent years, the growth and development of wheat has been seriously in uenced by drought stress which is the signi cant reason restricting the sustainable increase of wheat production [20], however studies on the response of wheat to drought stress are usually con ned to phenotype, structure, physiology and biochemistry, stress-resistant genes and other studies [21], epigenetic regulation of wheat response to drought stress is rarely involved, especially the regulatory role of DNA methylation in DREB transcription factor response to drought stress. In this study, main members of DREB family in wheat were identi ed, the expression and promoter methylation of DREB gene were analyzed under drought stress, which would be helpful to reveal the regulatory mechanism of DNA methylation in the response of plants to drought stress.

Results
Cloning and sequence analysis of DREB As shown in Fig. S1, the CDS sequence of DREB2, DREB6 and Wdreb2 in wheat AK58 was 732 bp, 837 bp and 1035 bp respectively, DREB2 and Wdreb2 had no intron, but one 712 bp intron was found in DREB6. CD-search analysis indicated that the typical AP2/EREBP conserved domain was found in amino acid sequence of DREB2, DREB6 or Wdreb2 (Fig. S1), was composed of YRG and RAYD conserved modules with three β folds and one α helix, simultaneously, valine (V) and glutamate (E) were very conserved at 14 th or 19 th of AP2/EREBP domain ( Fig. 1, a). The nucleotide sequences or amino acid sequences of DREB2, DREB6 and Wdreb2 were further compared by DNAMAN, the similarity among amino acid sequences was low with only 33.24% identity (Fig. 1, b), but AP2/EREBP domains had 73.25% identity, even reached to 83.93% between AP2/EREBP domains of DREB6 and Wdreb2 (Fig. 1, a).
Homologous sequences of DREB2, DREB6 and Wdreb2 from wheat AK58 were analyzed and compared ( Table 1, Fig. S2), the similarity of wheat DREB2 was 95% with Aegilops tauschii ERF, and was about 60% with TINY from Oryza sativa, Sorghum bicolor or Zea mays, AP2/EREBP domain of DREB2 was the same to that of Aegilops tauschii ERF and Zea mays TINY (Fig. S2, a). As listed in Table 1, the similarity of wheat DREB6 and some sequences was higher and was 98% or so, such as Thinopyrum elongatum AP2/EREBP, Aegilops biuncialis DREB2, Leymus multicaulis DREB2, etc. AP2/EREBP domain of DREB6 was the same to that of Thinopyrum elongatum AP2/EREBP, Aegilops biuncialis DREB2 and Agropyron mongolocum AP2/EREBP (Fig. S2, b). In addition, the similarity of wheat Wdreb2 with Aegilops tauschii DREB2B reached up to 99%, was also higher and was about 95% with Aegilops speltoides DREB1, Triticum turgidum DRF or Triticum dicoccoides DREB (Table 1). Furthermore, AP2/EREBP domain of Wdreb2 was the same to that of Aegilops tauschii DREB2B and Aegilops speltoides DREB1 (Fig. S2, c). The expression pattern of DREB in wheat As shown in Fig. 2, under normal condition, the expression level of DREB2, DREB6 and Wdreb2 in leaf was obviously higher than that in root, which was especially signi cant in Wdreb2 (P<0.05). Compared with that of DREB6, the expression level of Wdreb2 was signi cantly higher, yet the expression level of DREB2 was lower. Under drought stress, the expression level of DREB2, DREB6 and Wdreb2 in leaf was also higher than that in root (P<0.05), compared with the control, the expression of DREB2, DREB6 and Wdreb2 altered, but this change was different along with the increase of stress time.
Under drought stress, the expression level of DREB2 increased, and reached to the highest level as stressed for 2 h, which was signi cantly higher than the control (P<0.05), however the expression level of DREB2 decreased along with the increase of stress time, and was lower as stressed for 8-10 h, which was still higher than the control (P<0.05) (Fig. 2, a). The expression level of DREB6 was also the highest as stressed for 2 h, and was signi cantly higher than the control (P<0.05). Subsequently, along with the increase of stress time, the expression level of DREB6 gradually decreased, was signi cantly lower than the control as stressed for 10-12 h (P<0.05) (Fig. 2, b). As shown in Fig. 2 (c), the expression level of Wdreb2 in root signi cantly increased under drought stress, was obviously higher the control as stressed for 2 h, and also signi cantly increased in leaf when stressed for 6-8 h, especially stressed for 12 h (P<0.05).

Promoter analysis of wheat DREB
In this study, the promoter of DREB2, DREB6 and Wdreb2 was cloned, was respectively 1735 bp, 1792 bp or 649 bp, and was submitted to GenBank (MT974473, MT974471, MT974472). As shown in Fig. 3 and Table S1-S3, the promoter of DREB2, DREB6 and Wdreb2 contained basic regulatory element, such as TATA-box, CAAT-box, and there were 26, 18 and 5 TATA-boxes in the promoter of DREB2, DREB6 or Wdreb2, respectively. Many elements related to adverse stress were also found in the promoter of DREB2, DREB6 and Wdreb2, such as drought response element DRE/CRT, low temperature response element LTR, abscisic acid response element ABRE, light response element GAG-motif, drought-induced element MYB binding sites, etc (Fig. 3, Table S1-S3).
Further analysis found that there were some unique elements in the promoter of DREB2, DREB6 or Wdreb2, for example, the promoter of DREB2 had specially light response element MNF, leaf development element HD-ZIP and meristem speci city element OCT ( Fig. 3, a; Table S1). A series of speci c functional elements were also found in the promoter of DREB6, such as ethylene response element ERE, fungal elicitor response element Box-W1, MeJA regulatory element CGTCA-motif, and gibberellin response element P-box ( Fig. 3, b; Table S2). Moreover, the promoter of Wdreb2 had root speci city element as1, zein metabolism regulation element O2-site, light response element C-box, and CE3 element involved in ABA and VP1 reactions (Fig. 3, c; Table S3).

Methylation analysis of DREB promoter
The distribution of CpG island in the promoter of DREB2, DREB6 and Wdreb2 was predicated and analyzed by MethPrimer and EMBOSS CpG Plot, one CpG island with 234 bp was found in the promoter of DREB2 (Fig. S3, a). As shown in Fig. S3 (b), four CpG islands located respectively in 507-644 bp, 826-960 bp, 1149-1584 bp or 1631-1735 bp of DREB6 promoter, and one CpG island with 559 bp existed in the promoter of Wdreb2 (Fig. S3, c). Furthermore, there were also functional elements in above CpG islands, such as abscisic acid response element, light response element, low temperature response element, and so on (Fig. 3, Table S1-S3).
Some CpG island regions predicted in DREBs promoter were further examined from wheat leaf by bisul te sequencing PCR (BSP), and the CpG island region examined was respectively located in 1589-1904 bp (Fig. S3, a), 1135-1617 bp (Fig. S3, b) or 280-615 bp (Fig. S3, c) at the promoter of DREB2, DREB6 or Wdreb2. As shown in Fig. 4 and Table 2, there were more CHH sites and less CHG sites in the promoter region of DREB2, DREB6 and Wdreb2, but methylation rate of CG was the highest. In the promoter region of DREB2, CHH sites were not methylated, methylation rate of CG and CHH was 2.38% or 1.03%, and belonged to mild methylation (<20%) (Fig. 4, a; Table 2). As shown in Fig.4 (b) and Table 2, in the promoter region of DREB6, methylation rate of CG was 88.08% and was severely methylated (>60%), methylation rate of CHG was 51.36% and was moderately methylated (>20%), but methylation rate of CHH was only 4.93% and belonged to mild methylation (< 20%). Furthermore, in the promoter region of Wdreb2, methylation rate of CG, CHG or CHH was 1.89%, 1.0% and 0.29%, respectively, which were all mildly methylated (Fig.4, c; Table 2).

Methylation level of DREB promoter under drought stress
Under drought stress, cytosine methylation altered in the promoter region of DREB2, DREB6 and Wdreb2 from wheat leaf (Fig. 5). Compared with the control, methylation rate of CG in the promoter region of DREB2 decreased obviously (P<0.01), was 0.5% or 1.42% as stressed for 2 h and 10 h, but methylation rate of CHG and CHH increased signi cantly as stressed for 10 h (P<0.01). Further analysis showed that methylation level of DREB2 promoter was obviously lower or higher than the control when stressed for 2 h or 10 h, and this difference was signi cant (P<0.05) (Fig. 6, a).
As shown in Fig. 6 (b), methylation level of DREB6 promoter changed under drought stress, and was signi cantly higher than the control when stressed for 12 h (P<0.05). Compared with the control, methylation rate of CG and CHG was obviously lower or higher as stressed for 2 h and 12 h, although methylation rate of CG and CHG was signi cantly lower as stressed for 2 h, the promoter region of DREB6 was still heavily CG cytosine methylated (>60%) and moderately CHG cytosine methylated (>20%). As stressed for 2 h or 12 h, methylation rate of CHH was higher than the control, but this change was less than that of CG and CHG (P<0.05).
Furthermore, methylation level of Wdreb2 promoter also changed under drought stress, was signi cantly higher or lower than the control when stressed for 2 h and 12 h (P<0.01) (Fig. 6, c). Methylation rate of CG, CHG and CHH was respectively 2.16%, 1.5% or 1.02% as stressed for 2 h, and was obviously higher than the control (P<0.01), however was signi cantly lower than the control as stressed for 12 h (P<0.01).
Methylation status in DREB promoter under drought stress As listed in Table 3, methylation status in the promoter region of DREB2, DREB6 and Wdreb2 had signi cant change under drought stress. Along with the increase of stress time, the number of hypermethylation sites signi cantly increased in DREB2 promoter, for example, there were 1 CG site and 2 CHH sites in hypermethylation status as stressed for 2 h, but were 2 CG sites, 3 CHH sites and 1 CHG site as stressed for 10 h, furthermore, there were 3 CG sites and 1 CHH site in demethylation status under drought stress. Under drought stress, the number of hypermethylation and demethylation sites also changed in DREB6 promoter (  (Table 3). Along with the increase of stress time, the number of hymethylation sites had hardly changed in Wdreb2 promoter, but demethylation sites increased, and the change of methylation status was signi cant in CHH site, after stressed for 2 h, 2 CHH sites were respectively hypermethylated and demethylated, there were 1 CHH site in hypermethylation status and 2 CHH sites in demethylation status as stressed for 12 h (Table 3).

Correlation analysis between promoter methylation and expression of DREB
In order to explore the correlation between promoter methylation and expression of wheat DREB2, DREB6 or Wdreb2, under drought stress for different times, their relative expression levels in wheat leaf and methylation rates of CG, CHG or CHH in their promoter regions were respectively analyzed by SPSS software. As listed in Table S4, Pearson coe cient r between expression of Wdreb2 and methylation rate of CG, CHG or CHH was respectively -0.986, -0.973 and -0.878, indicating that signi cant negative correlation existed between promoter methylation and gene expression of Wdreb2, similarly, promoter methylation and gene expression of DREB6 was negatively correlated (Table S4). Although the signi cant negative correlation existed between expression of DREB2 and methylation rate of CG or CHG (Table S4), but promoter methylation of DREB2 had no negative correlation with its expression as stressed for 10 h (Fig. 2, a; Fig. 6, a).

Discussion
DREB transcription factor plays an important role in the response of plant to drought stress, could speci cally bind to DRE/CRT element in the promoter of stress-responsive gene and then would enhance the response or tolerance of plant to adverse stress [6]. AP2/EREBP domain of DREB transcription factor is composed of about 60 amino acid residues, has two conserved regions of YRG and RAYD [7]. In this study, DREB2, DREB6 and Wdreb2 were cloned from wheat AK58, one 712 bp intron was found in DREB6, AP2/EREBP domain of DREB2, DREB6 and Wdreb2 had 73.25% identity, the amino acid at 14 th or 19 th of AP2/EREBP domain was V and E, respectively. However, the similarity was lower among nucleotide sequences or amino acid sequences of DREB2, DREB6 and Wdreb2, BLASTP results further showed that DREB2, DREB6 and Wdreb2 were different types of DREB transcription factor and might respectively belong to DREBA-4 class, DREB-2 class or DREB-1 class, which was also found in other research [22,23].
Under abiotic stresses, such as drought, low temperature, high salt, etc., the expression of DREB transcription factor would alter [24,25]. In this study, the expression of DREB2, DREB6 and Wdreb2 could be induced under drought stress, and generally reached to the highest level after stressed for 2 h, but showed different trends along with the increase of stress time. The expression levels of DREB2, DREB6 and Wdreb2 were also different, as stressed for 2 h, the expression of Wdreb2 was signi cantly higher, but was the lowest in DREB2, Lopato et al also found that the expression of DREB2 was very low [26]. Further analysis showed that the expression of DREB2, DREB6 and Wdreb2 had tissue speci city, and was obviously higher in leaf than that in root, which was similar in other research [27], the expression of DREB in Daucus carota also showed tissue speci city, DcDREB-A1-1 and DcDREB-A1-2 had main role in leaf or root, respectively [28].
It is well known, the cis-acting regulatory elements in the promoter provide the possibility for transcription and expression of gene [29], there are some cis-acting elements related to adverse stress in plant promoter, such as DRE/CRT, EREH, ABRE, LTR and so on [30]. Except typical regulatory element TATA-box and CAAT-box, the promoter of DREB2, DREB6 and Wdreb2 in wheat AK58 contained DRE/CRT, LTR, ABRE, and drought-induced MYB binding site, etc, con rming that the expression of DREB2, DREB6 and Wdreb2 may be in uenced by adverse stress. Furthermore, in the promoter of DREB2, DREB6 and Wdreb2, CpG island with a variety of cis-acting elements was detected by MethPrimer and EMBOSS CpG Plot, some studies found that DNA methylation could regulate the expression of stress-responsive genes, and play an important role in the response of plant to adverse stress [16], especially promoter methylation had more signi cant effect on gene expression [31]. BSP analysis showed that there were more CHH sites and less CHG sites in the promoter region of DREB2, DREB6 and Wdreb2, but the methylation rate of CG sites was the highest.
Many studies have found that degree and state of DNA methylation in plant would change under drought stress, low temperature, high salt and other conditions [32,33], especially the change of methylation state in the promoter of gene [34]. Under drought stress, methylation level altered in the promoter region of DREB2, DREB6 and Wdreb2, compared with the control, methylation level in DREB2 and DREB6 promoter decreased after stressed for 2 h, then increased along with the increase of stress time, which was opposite in Wdreb2 promoter. Furthermore, methylation status in the promoter region of DREB2, DREB6 and Wdreb2 had signi cant change under drought stress, such as demethylation and hypermethylation, Zilberman also found that gene expression could be respectively promoted or inhibited by demethylation and hypermethylation of promoter [35].
Further analysis showed that promoter methylation of DREB6 and Wdreb2 was negatively correlated with their expression by Pearson coe cient, especially was signi cant in Wdreb2, this negative correlation was also found in other studies [35,36]. Although the promoter of DREB2 and Wdreb2 with low methylation level was both slightly methylated, the expression of Wdreb2 was signi cantly higher than that of DREB2, indicating that promoter methylation might have little effect on gene expression of DREB2, and its promoter possibly belongs to low CpG-contain promoter. Similarly, the promoter of z1B4 and z1B6 in Zea mays was almost not methylated [37], DNA methylation was not found in the promoter of some genes in Arabidopsis or tomato and only occurred in their coding regions [38,39]. In addition, one CpG island was also predicted in the coding region of DREB2, DREB6 and Wdreb2, and the CpG island almost covered the whole coding region of DREB2. However, it is unclear to the relation between DNA methylation in the coding region and gene expression of wheat DREB, the mechanism of DNA methylation regulating the expression of wheat DREB needs to be further studied.

Conclusions
In this study, DREB2, DREB6 and Wdreb2 were cloned and identi ed from wheat, and one 712 bp intron was found in DREB6. Under drought stress, the expression of DREB2, DREB6 and Wdreb2 would be induced, was obviously higher in leaf, but had different trends along with the increase of stress time. In the promoter region of DREB2, DREB6 and Wdreb2, some elements related to adverse stress were also found, further analysis showed that promoter methylation of DREB6 or Wdreb2 was negatively correlated with their expression, especially was signi cant in Wdreb2. Therefore, DREB2, DREB6 and Wdreb2 in wheat might function differently in response to drought stress, and promoter methylation had more signi cant effects on gene expression of Wdreb2 and DREB6, which would be helpful to reveal the regulatory mechanism of DNA methylation in plant response to drought stress.

Experimental materials
In this study, seeds of wheat AK58 were kindly provided by Xinxiang Academy of Agricultural Science, Henan, China. The tolerance of wheat AK58 is strong to drought stress, and its yield is generally high and stable. Primers and their sequences used in this study were listed in Table S5, and all primers were synthesized by Yingjie Ji Trade Co., Ltd. (Shanghai, China).

Cultivation and treatment of wheat seedlings
Cultivation of wheat seedlings was performed according to methods and conditions used by Duan et al. [33], wheat seeds were rstly surface-sterilized for 10 min by 0.1% HgCl 2 , then were washed for 50 min by sterile water. Subsequently, sterilized seed were sown in pots (diameter of 15 cm) containing nutrition soil and vermiculite (1:1), were cultured at 24 ± 1 ℃ with 45% relative humidity and 14 h photoperiod of 50 μmol m −2 s −1 light intensity, and were irrigated with 5 ml distilled water every two days.
At the three-leaf stage, wheat seedlings were irrigated with 15% PEG 6000 solution, roots and leaves of wheat seedlings were collected at 0 h (just before drought stress), and 2 h, 6 h, 8 h, 10 h or 12 h after subjected to drought stress in light condition, and immediately freezed with liquid nitrogen and then store at -80 ℃. In addition, there were three biological replicates for each experiment group in this study.

Extraction of genomic DNA
Genomic DNA was extracted from root or leaf of wheat seedlings by cetyltriethyl ammonium bromide (CTAB) method [40], the yield and purity of genomic DNA were determined at 260 nm by microspectrophotometry, and the integrity of genomic DNA was detected by 0.8% agarose gel electrophoresis.
Subsequently, genomic DNA from wheat seedlings was stored at -20 °C.
Isolation and reverse transcription of RNA Total RNA in root and leaf of wheat seedlings was respectively extracted by RNAiso Plus (TaKaRa, Japan) according to the instructions. In order to remove DNA, DNase/RNase-free treatment and phenolchloroform extraction were performed in this research, RNA was dissolved in RNase-free dH 2 O and was stored at -80 °C. Furthermore, the integrity of total RNA was veri ed by 1.0% agarose gel electrophoresis, the yield and purity of total RNA was determined by UV spectrophotometer.
In addition, cDNA was synthesized by reverse transcription of the extracted total RNA from wheat seedlings, the method of reverse transcription was referred to the introduction of HiScript II Q RT SuperMix for qPCR (+gDNA wiper) kit (Vazyme, China).

Cloning and analysis of DREB gene
In order to clone DREB genes from wheat AK58, speci c primers were designed according to the sequence of wheat DREB2 (GU785008), DREB6 (AY781361) and Wdreb2 (AB193608), and were listed in Table S5, furthermore, genomic DNA and cDNA of wheat AK58 were respectively used as the ampli cation template to obtain DNA or cDNA sequence of DREB genes.
In this experiment, PCR reaction system was composed of 2.0 μl DNA template, 1.0 μl each primer (10 μM), 10.0 μl 2x Taq Mix and 6.0 μl ddH2O. PCR procedure was at 95° C for 5 min, followed by 35 cycles of 94 °C for 30 s, 55 °C for 30 s and 72 °C for 1 min, and nally extended at 72 °C for 5 min. After PCR ampli cation products were detected by 1.0% agarose gel electrophoresis, target fragments were obtained by gel extraction and recycling, and then were sequenced in Vazyme (Nanjing, China).
In addition, analysis of target sequences was performed in the following, the extron, intron and ORF of DREB gene in wheat was analyzed with ProtParam, the conserved domain and amino acid sequence encoded by wheat DREB gene were analyzed by CD-search in NCBI, BLASTP was used to search similar amino acid sequences of wheat DREB, the domain or homologous sequences of wheat DREB were respectively compared with DNAMAN.
Fluorescence quantitative real-time PCR The expression of DREB gene in wheat was studied by uorescence quantitative real-time PCR (qRT-PCR), the internal reference gene was β-Actin, these primers for qRT-PCR were listed in Table S5. qRT-PCR was performed in LightCycler 96 Real-time PCR instrument, and cDNA synthesized by reverse transcription of total RNA was used as the template in qRT-PCR.
In addition, the relative expression level of wheat DREB under drought stress was normalized and analyzed by the comparative Ct ( 2-ΔΔct ) method [41]. The calculation formula was as follows: Relative expression level =2 -∆∆Ct , ∆∆Ct (target gene) =∆Ct (treatment group) -∆Ct (control group), ∆Ct (target gene) =Ct (target gene) -Ct (reference gene). Furthermore, three biological replicates were set up, and each qRT-PCR experiment was repeated three times.

Isolation and analysis of promoter sequence
The promoter region was cloned to further analyze expression pattern of DREB gene in wheat AK58, speci c primers were designed according to promoter sequence of wheat DREB2 (GU785008), DREB6 (HG670306.1) or Wdreb2 (KF731666), and were listed in Table S5.
PCR reaction system of DREB promoter was 20 μl, consisted of 2.0 μl DNA template, 10.0 μl 2x Taq Mix, 1.0 μl each primer (10 μM) and 6.0 μl ddH2O. The reaction conditions of PCR procedure was at 95 °C for 5 min, followed by 40 cycles of 94 °C for 30 s, 55 °C for 30 s and 72 °C for 1.5 min, nally extended at 72°C for 5 min. PCR ampli cation products were separated with 1.0% agarose gel electrophoresis, and the target fragments were obtained by gel extraction and recycling, then were sequenced in Vazyme (Nanjing, China). Furthermore, PlantCARE and PLACE were used to analyze cis-acting elements in the promoter sequence of wheat DREB.

Methylation analysis of promoter
CpG island (Island size > 100, GC Percent > 50.0, Obs/Exp > 0.6) in the promoter of DREB was predicted and analyzed by MethPrimer and EMBOSS CpG Plot. According to the analysis of CpG island, ampli cation primers of bisul te sequencing PCR (BSP) were designed by MethPrimer, Methyl Primer Expressv1.0 and Primer Premier5.0 (Table S5), and the CpG island of DREB6 promoter was ampli ed in two parts (region I and region II) because of the limited length of BSP ampli cation.
In this study, genomic DNA from leaf of wheat seedlings was rstly treated with EZ DNA Methylation-Lightning TM Kit (Zymo Research, America), then was used as template in BSP ampli cation of DREB promoter. BSP reaction system was 30.0 μl, and composed of 2.0 μl bisul te-treated DNA, 1.0μ l each primer (10 μM), 3.0 μl 10×buffer (Mg 2+ ), 1.0 μl dNTP, 1.0 μl Relia™ hot-start Taq polymerized aes and 21.0 μl dH2O. PCR ampli cation procedure was pre-denaturation at 95 °C for 4 min followed by 40 cycles (94°C for 30 s, 55 °C for 30 s, 72 °C for 40 s), and nal extension at 72 °C for 5 min. PCR ampli cation products were detected with 1.0% agarose gel electrophoresis, found that only target fragments were ampli cated, subsequently the target fragments were obtained by gel extraction and recycling, and were sequenced in GENEray (Shanghai, China).
In addition, at least 10 clones of per target fragment were sequenced and three biological replicates were set up in this study, statistics analysis on methylation site, methylation type and methylation rate was performed with CyMATE and Kismeth.

Statistical analysis
Statistical analysis of data was performed in this study, expression level of genes, methylation ratio of promoters were tested by signi cance level, ANOVA and multiple comparisons of Duncan's multiple range, and the correlation between gene expression and promoter methylation was analyzed by Pearson coe cient r of SPSS software. Declarations