The spring wheat variety Longmai 26 (LM26) was used to establish a drought-stress transcriptome and amplify gene sequences (TaERF13-2B). The spring wheat variety LM26, which is characterized by high yield and high environmental adaptability (Zhang et al., 2006), is widely used in wheat breeding programs. Wheat varieties ‘Fielder’, ‘CB037’, and ‘Westonia’ have been applied extensively in wheat genetic transformation, with ‘Fielder’ exhibiting the highest transformation efficiency (Liu et al., 2023). However, most elite wheat varieties are resistant to genetic transformation and typically exhibit lower transformation efficiency (Luo et al., 2021). Therefore, the present study used the spring wheat cultivar ‘Fielder’ (wild-type, WT) to generate transgenic wheat plants. Due to their distinct varietal characteristics, the spring wheat varieties LM26 and Fielder serve as suitable models for studying the gene of interest. The Arabidopsis thaliana ‘Columbia-0’ is a model plant for genetic research and is widely used for gene function validation. In the present study, the WT Arabidopsis was used as the receptor material for TaERF13-2B.

The vectors used in the present study include the pBI121 Arabidopsis overexpression vector (transformation strain GV3101), pGBKT7/PGADT7 vector (transformation strain AH109), pCAMBIA1302-nYFP/cYFP bimolecular-fluorescence complementation (BiFC) vector (transformation strain GV3101), and the pCAMBIA3301 wheat overexpression vector (transformation strain EHA105).

Wheat variety LM26 was cultivated at the two-leaf stage and exposed to drought stress (20% PEG6000). Leaf tissue samples were gathered at 0, 1, 6, and 12 h (Ge et al., 2022; Ye et al., 2021), resulting in 12 wheat samples sent to the Tianjin BMK Company for transcriptome sequencing. The clean reads were quickly and precisely mapped to the reference genome (Version: IWGSC RefSeq v1.1) using HISAT2 software to acquire the localization information of the reads on the reference genome. Subsequently, StringTie was used to assemble the aligned reads and reconstruct the transcriptome for further analysis (SRA: PRJNA1178630). StringTie was used to assemble the transcriptome using a maximum flow algorithm, and fragments per kilobase of transcript per million fragments mapped (FPKM) normalization was applied as an indicator of transcript levels. Differential analysis was conducted using the edgeR software. In the process of identifying differentially expressed genes (DEGs), a fold change ≥ 4 and a false discovery rate < 0.01 were set as the screening criteria. The AP2/ERF family HMM profile (PF00847) from the PFAM was utilized as a query to hunt for and analyze proteins encoded by DEGs using HMMER software (Zhang et al., 2024a). The identified members were assessed through NCBI’s CDD to validate the existence and characterization of the AP2 domain, and candidate ERF members were identified according to the characteristics of the ERF members. Finally, the molecular properties and subcellular localization of each ERF protein were evaluated using the ExPASy and Cell-PLoc software.

All wheat ERF sequence information was obtained from the transcriptome database, whereas the sequence data for Arabidopsis and Brachypodium were retrieved from Ensembl Plants. The ERF amino acid sequences of these three species were aligned using ClustalX for sequence alignment analysis, and a neighbor-joining (NJ) phylogenetic tree was generated using MEGA7 with 1000 bootstrap replicates.

Chromosomal location information from the GFF3 file was visualized using TBtools-II (v.2.121). Coding and genomic sequences were submitted to the GSDS database for gene structure analysis, and the gene structure was visualized using the EvolView tool. The XML file for wheat ERF proteins obtained from the MEME database was visualized using TBtools-II (v.2.121). A 1500-bp upstream segment of the ATG from the TaERF members was analyzed using PlantCARE to identify cis-regulatory elements.

The FPKM values for the wheat ERF genes, extracted from the transcriptome data, were visualized using TBtools-II (version 2.121), resulting in the generation of a heat map. The reliability of the transcriptome information was confirmed using qRT-PCR according to methods detailed in our previous studies (Yu et al., 2023). The relative expression levels of the genes were assessed using the 2^−ΔΔCT method, with Actin-7 (LOC542814) serving as the internal reference gene for wheat. Each experiment was conducted in triplicate. Supplementary Table S1 lists the primers that were used in this study.

The production of TaERF13-2B transgenic plants, as described in our previous studies (Yu et al., 2023), resulted in the acquisition of two homozygous T₃ transgenic Arabidopsis lines. Three-week-old Arabidopsis seedlings were subjected to a 13-day drought period without water, followed by re-watering for 4 days to analyze drought survival rates. The survival rate was assessed by calculating the ratio of surviving plants to the total number of plants following drought treatment. A plant is considered alive if it has green leaves and exhibits normal growth, while a plant is classified as dead if its leaves are wilted, shriveled, or completely necrotic. One replicate contained 40 seedlings of each line. Three independent survival assays were performed. After a 9-day drought period, physical indices were measured, and the expression levels of stress-related genes were assessed in both WT and overexpressing (OE) plants. Malondialdehyde (MDA), proline (Solarbio, Beijing), and chlorophyll contents were quantified as outlined in previous studies (Yu et al., 2023; El-Esawi et al., 2019), with three biological replicates. Finally, qRT-PCR was performed to measure the mRNA levels of stress-modulated genes (AtP5CS1, AtDREB2A, AtRD29A, AtCOR15A, AtMYB15, AtERD10, AtLTI30, AtKIN1, and AtGSTU19), with three biological replicates analyzed in the transcript profiles of genes.

To construct wheat overexpression (OE) lines, the CDS of TaERF13-2B was inserted into the pCAMBIA3301 vector, driven by the maize ubiquitin promoter. Agrobacterium tumefaciens-mediated transformation was employed to generate independent transgenic lines for the TaERF13-2B construct, using the wheat cultivar Fielder as the genetic background (Ishida et al., 2015). The T₀, T₁, and T₂ transgene-positive plants were detected using PCR, and the T₃ generation was detected using PCR and qRT-PCR. Wheat seedlings were grown in a controlled environment with a 16 h light/8 h dark cycle and 22°C during the day and 19°C at night. At the two-leaf stage, drought stress was applied by withholding water for 13 days, during which physiological parameters were measured. Drought stress was extended further to 15 days to photograph the drought phenotypes. The measured physiological parameters included chlorophyll content, MDA levels, O₂ ^.-content, and the of antioxidant enzyme activities (SOD, POD, and CAT), according to the methods reported in our previous studies (Yu et al., 2023). Three biological replicates were performed. Quantitative analysis of stress-related genes, including TaP5CS1, TaERF3, TaDREB1, TaSOD (Fe), TaPOD, and TaCAT, was performed on day 8 of drought stress, with three biological replicates analyzed in the transcript profiles of genes.

To evaluate the transcriptional activity of TaERF13-2B, the CDS of 290 amino acids, along with truncated fragments representing the N-terminal (1 to 139 aa) and C-terminal (140 to 290 aa) regions, were cloned into the pGBKT7 and introduced into AH109 yeast cells. The transformants were then cultured in the dark at 30°C for 3 days on SD medium deficient in tryptophan and SD medium lacking tryptophan, histidine, and adenine plates to assess their transcriptional activity.

A truncated coding region of TaERF13-2B containing the AP2 domain (amino acids 140–290) was ligated into the pGBKT7 and introduced into yeast cells (AH109) for Y2H screening. Following previously established methods (Yu et al., 2023), a Y2H screening was performed using a cDNA library derived from drought-treated wheat. The positive clones were sequenced for further confirmation. Genes screened from the cDNA library contained full-length CDS and were merged with the pGADT7-prey vector. Modified plasmids carrying the bait TaERF13-2B (140–290 aa) and the pGADT7-prey were co-transformed into AH109 yeast cells. The transformants were then nurtured on solid media double-dropout medium (SD lacking tryptophan and leucine) and quadruple-dropout medium (SD lacking tryptophan, leucine, histidine, and adenine) for point-to-point validation of the Y2H interactions.

The complete sequences of TaERF13-2B and TaCIPK9 were individually inserted into pCAMBIA1302-nYFP and pCAMBIA1302-cYFP, respectively. The recombinant vectors were introduced into A. tumefaciens and transiently co-expressed in Nicotiana benthamiana leaves. Fluorescence signals were visualized using an Olympus IX83 confocal microscope (Olympus, Tokyo, Japan). The specific experimental details can be viewed in our previous study (Yu et al., 2022).

Each experiment was conducted with three biological replicates. All data are expressed as mean ± standard error and were statistically analyzed using SPSS 22.0 (IBM SPSS, USA). Group differences were assessed via the Student’s t-test, with statistical significance set at *P < 0.05 and highly significant differences at **P < 0.01.

In a prior study, a transcriptome related to drought stress was established using young leaves from the wheat variety LM26. Transcriptome sequencing was performed on 12 LM26 samples at 0, 1, 6, and 12 h after drought stress, which yielded a large set of DEGs. A local protein database was constructed using the proteins encoded by the DEGs. The HMMER program was used to screen for ERF members, followed by the verification of the presence and integrity of the AP2 domain in the candidate proteins using the NCBI-CDD. This process identified 33 ERFs in wheat ( Supplementary Table S2 ). Based on chromosomal distribution analysis, these TaERF genes were located on 15 of the 21 wheat chromosomes. The largest number of TaERFs (four) were observed on chromosome 1A. Twelve TaERF genes were evenly distributed across chromosomes 1D, 2A, 2B, and 3D, whereas 14 TaERF genes were distributed evenly across chromosomes 1 B, 2D, 3A, 3B, 4A, 4B, and 6D. The remaining three TaERF genes were individually scattered across the other chromosomes ( Supplementary Figure S1 ). Based on their phylogenetic relationships and chromosomal locations, the 33 TaERFs were renamed sequentially from TaERF1-1A to TaERF14-6D. Subsequently, the physical properties of the proteins, including the sequence length, MW, pI, and subcellular localization, were predicted ( Supplementary Table S2 ). TaERF8-4D had the longest sequence, consisting of 342 amino acids, whereas TaERF6-3A and TaERF6-3D had the shortest sequences, with 173 amino acids each. TaERF8-4D had the highest MW (36.51 kDa), whereas TaERF6-3A had the lowest (18.12 kDa). The pI values ranged from 4.6 (TaERF8-4D) to 10.07 (TaERF6-3D). Subcellular localization predictions indicated that 33 TaERF proteins were localized in the nucleus.

An NJ phylogenetic tree was constructed using 33 wheat, 11 Brachypodium, and four Arabidopsis ERFs to explore the phylogenetic ties within the TaERF members ( Figure 1 ; Supplementary Table S3 ). The 33 wheat TaERFs were classified into eight phylogenetic groups: Groups I–VIII, containing six, three, five, three, five, four, three, and four TaERF proteins, respectively.

Based on the NJ phylogenetic tree, ERFs in wheat were classified into eight groups: I–VIII ( Figure 2A ). The intron-exon structure provides insights into the structural features of the ERF family ( Figure 2B ). All ERF members lacked introns, indicating the evolutionary conservation of this family. Transcriptional output may be delayed by the presence of numerous and lengthy introns, potentially leading to the suppression of gene expression under adverse conditions. In contrast, genes with fewer or shorter introns are likely to exhibit more efficient expression when responding to stress environments (Heyn et al., 2015). Thus, the absence of introns in ERFs may facilitate rapid responses to abiotic stress conditions in wheat.

Potential motifs in all TaWRKY proteins were predicted using MEME. Ten conserved motifs (motifs 1–10), ranging in size from 15 to 49 aa, were identified ( Supplementary Figure S2 ). The ERFs belonging to a similar subgroup exhibited similar motif compositions. For example, the six protein sequences of Group I included motifs 1, 2, 3, 4, 5, and 7 ( Figure 2C ). Motifs 1 and 2 were present in all ERFs and encoded the AP2 domain ( Supplementary Figure S3 ). In the AP2 domain, the 14th residue was alanine, and the 19th was aspartic acid ( Supplementary Figure S3 ), confirming that all 33 genes are typical ERFs.

Using PlantCARE, cis-regulatory elements in TaERF promoters were predicted, revealing numerous elements linked to hormone signaling and responses to abiotic stress ( Figure 3 ). The identified hormone-related elements encompassed a variety of responsive motifs, including auxin-regulated elements (TGA element and AuxRR core), abscisic acid-activated elements (ABRE), methyl jasmonate-regulated elements (CGTCA and TGACG), salicylic acid-associated elements (TCA element), and gibberellin-associated elements (GARE-motif and P-box). In addition, several abiotic stress-associated elements were observed, including the mechanical injury response element (WUN-motif), binding sites for MYB, MYC, and WRKY TFs (MYB, MYC, and W-box), defense and stress-associated elements (TC-rich repeats), drought-inducible element (MBS), dehydration-responsive elements (DRE), anaerobic induction elements (ARE), low-temperature-regulated elements (LTR), and hypoxia-specific-associated elements (GC-motif). The results imply that TaERF genes are essential for the regulation of wheat growth and development, as well as its adaptation to environmental factors.

To gain insights into the biological roles of wheat ERF members, we assessed the fragments per kilobase of transcript per million fragments mapped (FPKM) values of 33 ERF DEGs identified from RNA-seq data collected before and following drought stress, and visualized the results ( Figure 4A ). The common wheat (Triticum aestivum L.) is an allohexaploid plant. Most wheat genes exist in three copies and exhibit a high degree of sequence identity within their coding regions, suggesting potential functional similarity. In the present study, under drought stress, most homologous genes displayed similar expression patterns ( Figure 4A ). For example, the TaERF3 gene (TaERF3-2A, TaERF3-2B, and TaERF3-2D) was significantly downregulated under drought stress. Similarly, the TaERF14 gene (TaERF14-6A and TaERF14-6D) exhibited an initial downregulation at 1 h, followed by upregulation at 6 h, and a subsequent downregulation at 12 h. Thus, one homolog from the TaERF1~14 gene group was selected randomly for qRT-PCR validation. As illustrated in Figure 4B , the expression patterns obtained from qRT-PCR closely matched the transcriptome information, confirming the accuracy of the RNA-seq results. Most wheat ERF genes were downregulated markedly under drought conditions. Examples include TaERF1-1A, TaERF2-3B, TaERF3-2D, TaERF4-2A, TaERF5-6B, TaERF6-3D, TaERF8-4B, TaERF9-1A, TaERF10-1A, TaERF11-3D, and TaERF12-1B. In contrast, TaERF7-4A was significantly upregulated after 1 h of exposure to drought stress, recovered to baseline levels after 6 h, and was further downregulated at 12 h. TaERF13-2B was significantly upregulated at 6 h, followed by a return to baseline levels at 12 h. Meanwhile, TaERF14-6D was significantly downregulated after 1 h, recovered to basal levels, and was subsequently downregulated.

Among the 33 ERF DEGs, TaERF13-2B showed the largest drought-induced transcript-level differences, with a 2.83-fold upregulation after 6 hours of drought stress, and its promoter region contained abundant stress-related cis-acting elements, suggesting that TaERF13-2B may serve an essential function in regulating drought stress responses. Therefore, the gene was selected for further investigation. To determine the function of TaERF13-2B, an OE vector for this gene was constructed in Arabidopsis. Two homozygous T₃ transgenic plants were obtained, and their TaERF13-2B expression levels were analyzed using RT-PCR ( Figure 5A ). The findings revealed that TaERF13-2B was not expressed in the WT plants, whereas it was detected successfully in the OE plants, confirming the effective introduction of TaERF13-2B into Arabidopsis. Before the drought stress, no notable differences were recognized in growth between the WT and OE plants ( Figure 5B ). After 13 days of water deprivation followed by 4 days of re-watering, most WT plants resumed growth, albeit slower, whereas most OE lines died, with leaves exhibiting chlorosis. Survival rate analysis further confirmed that the survival rate of OE plants was markedly lower than that of the WT plants ( Figure 5C ). Under normal watering treatments, no significant differences were recognized in chlorophyll content ( Figure 5D ), MDA levels ( Figure 5E ), or proline content ( Figure 5F ) between the WT and OE plants. However, under drought conditions, chlorophyll and proline contents were markedly lower in the OE plants, whereas MDA levels were markedly higher than those in the WT plants.

Under drought conditions, the transcript levels of nine stress-responsive genes—AtP5CS1, AtRD29A, AtDREB2A, AtCOR15A, AtMYB15, AtERD10, AtLTI30, AtKIN1, and AtGSTU19—were assessed in the OE plants using qRT-PCR ( Figure 6 ). In comparison to the WT plants, the transcript levels of these genes were lower in the TaERF13-2B OE plants. The findings reveal that TaERF13-2B OE in Arabidopsis may alter physiological traits by modulating the expression of stress-associated genes, thereby increasing the sensitivity of transgenic plants to drought conditions.

To examine TaERF13-2B function in wheat under drought conditions, the TaERF13-2B gene were overexpressed in wheat and further validated its regulatory role in the drought stress response using stable TaERF13-2B OE lines. The expression levels of TaERF13-2B in the OE wheat plants were measured using qRT-PCR ( Figure 7A ). Compared with the control, the expression levels of TaERF13-2B were upregulated markedly in all six OE plants, with the OE3 line exhibiting the most pronounced increase. We selected the homozygous transgenic lines with higher expression levels of OE1 and OE3 for further studies. Before drought stress, leaf morphology between WT and transgenic lines at the seedling stage was not different ( Figure 7B ). After 15 days under drought conditions, the OE wheat plants wilted severely, whereas drought stress had a comparatively mild impact on the WT ( Figure 7C ). The effect of drought stress on the physiological traits of the transgenic wheat indicated that under drought stress, the chlorophyll ( Figure 7D ) and proline ( Figure 8C ) contents, as well as SOD ( Figure 8D ), POD ( Figure 8E ), and CAT activities ( Figure 8F ), were markedly lower in the OE plants than in the WT ones. Conversely, the MDA ( Figure 8A ) and O₂ ^.- ( Figure 8B ) contents were notably increased notably in the OE plants compared to in the WT plants. In summary TaERF13-2B gene overexpression in wheat under drought conditions improved the sensitivity of OE plants markedly.

We further analyzed the expression levels of TaP5CS1, TaERF3, TaDREB1, TaSOD (Fe), TaPOD, and TaCAT in the TaERF13-2B wheat OE lines following drought stress ( Figure 9 ). Under drought conditions, all stress-related genes were downregulated markedly, with the most pronounced downregulation observed in TaCAT. The results imply that TaERF13-2B may regulate the transcript levels of stress-associated genes in wheat during drought conditions, triggering shifts in the physiological traits of seedlings and ultimately increasing susceptibility of transgenic wheat to drought conditions.

To facilitate subsequent Y2H studies, the transcriptional activity of TaERF13-2B was analyzed. Full-length TaERF13-2B (1–290 aa) and two truncated bait constructs, TaERF13-2B (1–139 aa) and TaERF13-2B (140–290 aa), were generated. These recombinant vectors, along with an empty pGBKT7 control vector, were introduced into yeast cells. Yeast strains carrying pGBKT7-TaERF13-2B (140–290 aa) grew only on the SD medium lacking tryptophan (SD/-Trp) ( Figure 10A ), indicating the absence of transcriptional activation. In contrast, strains transformed with either pGBKT7-TaERF13-2B (1–290 aa) or pGBKT7-TaERF13-2B (1–139 aa) showed robust growth in the three-dropout (-Trp/-His/-Ade) medium ( Figure 10A ). These findings suggest that full-length TaERF13-2B exhibits transcriptional activation activity, with the N-terminal (1–139 aa) region identified as the transcriptional activation domain.

A Y2H screen was performed using a wheat cDNA library to identify interacting proteins. A few yeast colonies grew on double-dropout (-Trp/-Leu) and quadruple-dropout (-Trp/-Leu/-His/-Ade) media. Redundant genes were removed, and the remaining candidate interacting proteins were annotated using the NCBI database. Five potentially interacting proteins were identified: TaVDAC, TaNRX, TaG6PE, TaCIPK9, and TaCAB; detailed annotations are provided in Supplementary Table S4 . The genes encoding these candidate proteins contained complete ORFs and were cloned into the pGADT7 vector to construct prey plasmids for Y2H reverse validation. Each of the five pGADT7 recombinant vectors was co-transformed with pGBKT7-TaERF13-2B (aa 1–334) into the Y2H yeast strains. All transformed strains exhibited growth on both the SD medium lacking tryptophan and leucine (SD/-Trp/-Leu), and the SD medium lacking tryptophan, leucine, histidine, adenine, and supplemented with X-α-Gal (SD/-Trp/-Leu/-His/-Ade/+X-α-Gal) ( Figure 10B ). Overall, the results confirm that TaERF13-2B interacts with five putative proteins in the yeast system. Studies have demonstrated that CIPK9 is crucial in regulating plant adaptations to drought conditions (Song et al., 2018; Shabala et al., 2021). To investigate the interaction between TaERF13-2B and TaCIPK9 in a specific cellular location, both were co-expressed in tobacco plants. YFP fluorescence was detected upon co-expression of TaERF13-2B and TaCIPK9, while no fluorescence was observed in the control groups Figure 10C . This confirmed the interaction between TaERF13-2B and TaCIPK9 in live plant cells through BiFC analysis.