Functional analysis of the GmESR1 gene associated with soybean regeneration

Plant regeneration can occur via in vitro tissue culture through somatic embryogenesis or de novo shoot organogenesis. Transformation of soybean (Glycine max) is difficult, hence optimization of the transformation system for soybean regeneration is required. This study investigated ENHANCER OF SHOOT REGENERATION 1 (GmESR1), a soybean transcription factor that targets regeneration-associated genes. Sequence analysis showed that GmESR1 contained a conserved 57 amino acid APETALA 2 (AP2)/ETHYLENE RESPONSE FACTOR (ERF) DNA-binding domain. The relative expression level of GmESR1 was highest in young embryos, flowers and stems in the soybean cultivar ‘Dongnong 50’. To examine the function of GmESR1, transgenic Arabidopsis (Arabidopsis thaliana) and soybean plants overexpressing GmESR1 were generated. In Arabidopsis, overexpression of GmESR1 resulted in accelerated seed germination, and seedling shoot and root elongation. In soybean overexpression of GmESR1 also led to faster seed germination, and shoot and root elongation. GmESR1 specifically bound to the GCC-box. The results provide a foundation for the establishment of an efficient and stable transformation system for soybean.


Introduction
Plant regeneration is a clonal propagation process in vitro, which may involve a variety of processes, such as exogenous plant hormone signaling response, division of quiescent cells, and formation of a meristem or organ primordia [1]. Overexpression of cyclin-related genes in Arabidopsis thaliana could be an important link between cell proliferation in shoot apical meristems and organogenesis [2]. Shaul et al. (1996) [3] suggested that the expression of cyclindependent kinase genes was highly correlated with acquisition of the ability for cell proliferation. In Arabidopsis, the regeneration-associated CLAVATA (CLV) genes may regulate stem cell fate, such as limiting the size of the stem cell population, and the function of CLV3 is dependent on WUSCHEL (WUS) activity in the embryonic shoot meristem [4][5][6] study indicates that changes in molecular patterning scales are associated with organ size in apical stem cell niches in plants [7]. Genes that promote cell proliferation or cell volume increase in soybean have not been studied previously. Plant transformation is achieved by transferring a DNA fragment to the genome of a plant cell, regenerating a shoot from the transgenic cell, and then generating a root system to produce a genetically modified plant [8,9]. Rhizobium radiobacter (Agrobacterium tumefaciens) mediated methods or particle bombardment are the preferred DNA transfer techniques [10]. To achieve high frequencies of plant regeneration, Cheng et al. (1980) [11] used aseptic cotyledon nodes as explants to induce differentiation of multiple shoots clumps. The genetic transformation of soybean (Glycine max) has long been of global interest, with successful transformation dependent on the efficient combination of receptor system and transformation method. The poor reproducibility of soybean regeneration represents a bottleneck for transformation of this important crop. Previous reports on soybean regeneration systems focused mainly on aspects such as genetic constitution, hormone types, culture conditions and explant types. Few studies have investigated the molecular basis of regeneration, especially the underlying mechanisms and the genes involved in this process. Detection of the expression of genes associated with soybean regeneration might help to address this problem.
Many de novo organogenesis systems are potentially useful as model experimental systems to illustrate the complexity of plant cell differentiation. The molecular mechanisms activated by the auxin to cytokinin ratio, however, are not well characterized. Efficient shoot regeneration involves two consecutive incubation steps: the exophyte is first incubated on an auxin-rich callus induction medium, and then on a cytokinin-rich shoot induction medium. The auxin rich callus induction medium initiates the formation of organogenic callus. More explants could improve the conversion efficiency, thus increasing the rate of emergence. Many regenerationassociated genes are related to the cytokinin signaling pathway, of which ENHANCER OF SHOOT REGENERATION 1 (AtESR1) is one gene. AtESR1 plays an important role in the regeneration network [12][13][14][15].
APETALA 2/ETHYLENE RESPONSE FACTOR (AP2/ERF) transcription factors (TFs) are involved in various biological functions, including plant and flower development, fruit and seed maturation, pathogen defense, and responses to damage, high salinity, and drought [16]. The AP2/ERF family belongs to a large group of TFs present in all plant species [17]. In Arabidopsis, the AP2/ERF TF family is divided into five subfamilies: AP2, Related to ABI3/VP1 (RAV), Dehydration Responsive Element Binding (DREB), ERF, and others [16]. These TFs are characterized by a conserved DNA-binding domain, the AP2/ERF domain, which was originally identified in Arabidopsis and is composed of 57-66 amino acids [18][19][20]. The ERF subfamily is the largest subgroup in the AP2/ERF family with members containing one or two AP2/ERF domains with specific DNA-binding motifs [21][22][23]. The three-dimensional structure of the AP2/ERF domain protein showed that the region contains three β fold, which is differences from located in the second β fold in the first 14 (alanine) and 19 (aspartic) amino acid residues, determines the specific binding of such TFs to different cis-acting elements [24]. Gutterson et al. (2004) [25] suggested that ten distinct subfamilies accommodate the structural difference in B subgroup. The ERF VIII-b group genes usually function in the early stages of shoot regeneration [22,26]. The ERF subfamily of TFs participates in biological stress responses through binding to the GCC-box (AGCCGCC) [27]. The role of ERF TFs in the regulation of shoot regeneration is complex; both their biological function, and the ERF-mediated signal transduction pathway, are not well characterized.
The AtESR1 gene (also known as DORNRÖSCHEN; DRN) was identified by screening an Arabidopsis cDNA library. The ESR1 protein contains a domain that shows sequence homology to the AP2/ERF domain [20,28,29]. ESR1 appears to regulate shoot differentiation, with overexpression of ESR1 greatly enhancing the efficiency of shoot regeneration in Arabidopsis tissue culture [14]. ESR1 encodes a TF belonging to the ERF family. The region between the AP2/ERF domain and the ESR motif in ESR1 is indicated to be essential for enhancement of shoot regeneration [30]. ESR1 acts as a transcriptional activator [31,32]. ESR1 binds to the GCC-box in vitro [33]. The GCC-box is an ethylene-responsive element located in the promoter region of many pathogenesis-related genes [34,35]. Using yeast two-hybrid screening, ESR1 has been shown to interact with PHAVOLUTA (PHV), while coimmunoprecipitation and bimolecular fluorescence complementation have shown that ESR1 interacts with BES INTERACTING MYC-LIKE PROTEIN 1 (BIM1), a basic helix-loop-helix (bHLH) protein. BIM1 and PHV also physically interact [36,37]. In Arabidopsis, shoot-related auxin-transport is conducted by ESR1 and ESR2, two partially redundant AP2 TFs that interact during shoot development with PINOID (PID) and PIN-FORMED 1 (PIN1), respectively [38].
An orthologue of AtESR1 was previously isolated from maize. In the maize shoot apical meristem, similar to AtESR1, ZmESR1 transcriptional activity is associated with the anlage of new lateral organs [39]. In the present study we isolated GmESR1 (GenBank accession no. JN590243.1, NCBI protein no. AFO52509.2), an AP2/ERF TF containing an ESR motif, from the soybean cultivar 'Dongnong 50'. In this study we examined one of the two gene copies present in the soybean genome. We present expression patterns of the full-length GmESR1 protein and binding to the GCC-box element and show that the GmESR1 transcript abundance varies in different organs. GmESR1 showed organ-specific expression in soybean. Overexpression of GmESR1 in transgenic soybean and Arabidopsis plants was also investigated. Overexpression of GmESR1 promoted germination and elongation in soybean and Arabidopsis.

Plant materials and cultivation condition
For plant transformation, seeds of soybean (Glycine max (L.) Merr.) cultivar 'Dongnong 50', which shows a high frequency of regeneration, were obtained from the Key Laboratory of Soybean Biology in the Chinese Ministry of Education, Harbin. Seedlings were grown in a growth chamber maintained at 26˚C/18˚C (day/night) under a 16 h photoperiod and light intensity of 350 μmolÁm −2 Ás −1 . Transgenic T 1 soybean seeds were sown under the same conditions. Fifteen days after planting, seedlings at the first-node stage (soybean growth phase V1) [40] were used for phenotype analysis and expression analysis using quantitative real-time PCR (qRT-PCR).
Arabidopsis thaliana ecotype Columbia (Col-0) was used as the wild type (WT). For transgenic Arabidopsis, the T 3 generation was used. Seeds of the mutant Arabidopsis atesr1 T-DNA insertion line (Salk_089567) were obtained from The Arabidopsis Information Resource.

Isolation of GmESR1
To identify GmESR1 and homologs in other plant species, the Phytozome 11.0 (https:// phytozome.jgi.doe.gov/pz/portal.html) database was searched using the AtESR1 gene sequence, extracted from the National Center for Biotechnology Information (NCBI) website, as the query sequence. Total RNA was reverse-transcribed into single-stranded cDNA using the ReverTra Ace1 qPCR RT Kit (TOYOBO, Japan). Using this cDNA as a template, GmESR1 gene-specific primers (GmESR1-F/R) were used to amplify GmESR1. PCR reaction conditions were as follows: 94˚C for 4.5 min, then 35 cycles at 94˚C for 30 s, 58˚C for 30 s, and 72˚C for 1.5 min, with final extension at 72˚C for 10 min. The PCR products were inserted into the pMDTM19-T vector and ligated together (Takara, Japan). The ligation products were then transformed into Escherichia coli DH5α cells (TIANGEN, China) and sequenced (BioMed, China). The sequences were aligned with the GmESR1 sequence using BLAST (http://www.ncbi.nlm.nih.gov/BLAST). The isoelectric point and the molecular weight of the GmESR1 protein were analyzed using the Swiss Institute of Bioinformatics Compute pI/Mw tool (http://web.expasy.org/compute_pi/). Nucleotide and amino acid sequences were compared using the sequence alignment software DNAMAN 6.0 (http://www.lynnon.com/). To predict the GmESR1 protein structure and discover potential domains, the InterPro online portal was used (https://www.ebi.ac.uk/interpro/). Analysis of homologous protein sequence similarity was performed using the algorithm blastp (protein-protein BLAST) (http://www. ncbi.nlm.nih.gov/blast). Phylogenetic analysis of a multiple sequence alignment of the amino acid sequences of GmESR1 and heterologous AP2/ERF members was performed using MEGA 5.2 software (http://www.megasoftware.net). The three-dimensional structure of GmESR1 was predicted using the Phyre 2 online portal (http://www.sbg.bio.ic.ac.uk/phyre2). The RasMol software 2.7.2.1.1 (http://www.OpenRasMol.org/Copyright.html) was used to generate a graphical representation of the protein structure.

Real-time RT-PCR analysis of GmESR1 expression
The expression of GmESR1 was examined with qRT-PCR using SYBR 1 Premix Ex Taq™ II Kit (Tli RNaseH Plus, Takara) according to the manufacturers' instructions (Takara), on an ABI 7500 Real-Time PCR Detection System (ABI, USA). Total RNA was extracted from the pod, root, stem, leaf, flower and immature embryo of soybean 'Dongnong 50' plants using TRIzol 1 Reagent according to the manufacturers' protocol (Invitrogen, China). Genomic DNA was removed, and reverse transcription carried out, using the PrimeScript™ RT Reagent Kit with gDNA Eraser (Takara). Approximately 1 μg of total RNA was used for each reaction. To remove genomic DNA, samples were incubated at 42˚C for 2 min. For reverse transcription, each reaction used 10 μl of the reaction solution from the first step in a total volume of 20 μl. Samples were incubated at 37˚C for 15 min followed by heat shock at 85˚C for 5 s. A standard two-step PCR amplification protocol of 95˚C for 30 s, followed by 45 cycles at 95˚C for 5 s and 60˚C for 40 s, was used. Gene-specific primers (GmESR1-qF/R) for GmESR1, the soybean internal control gene GmACTIN4 (GenBank accession no. AF049106) and the Arabidopsis internal control gene AtACTIN8 (A. thaliana 18S rRNA gene GenBank accession no. X16077) were used. GmACTIN4 and AtACTIN8 were used as reference genes. The 2 −ΔΔCt method was used to determine the relative level of GmESR1 expression in different tissues. Three technical replicates were performed for each real-time RT-PCR experiment.

Expression and purification of recombinant GmESR1 protein
The full-length coding region of GmESRl was amplified using gene-specific primers (GmESR1-1F/R). The PCR products were digested with BamHI and HindIII and were inserted into the pET-29b vector (EMD Millipore, USA). The recombinant vector pET29b-GmESR1 was transformed into BL21 (DE3) competent E. coli cells, which were then grown in Luria broth (LB) with 50 mgÁmL −1 kanamycin at 37˚C to an absorbance of 0.7 at 600 nm. The E. coli liquid medium was induced with 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG). After 4 h induction, the cells were isolated via centrifugation at 5000 ×g for 12 min at room temperature. To purify the recombinant protein, bacteria were resuspended in 15 ml of 1× binding buffer and kept on the ice for 25 min. This was followed by cycles of ultrasonification for 20 s and pause for 20 s until the sample was no longer sticky. The sample was then centrifuged at 2000 ×g for 8 min at room temperature before being recycled and loaded onto a His-bind Resin column (Novagen, BRD). The pure GmESR1 fusion protein was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and quantified based on the pET System manufacturer's protocol (Novagen, BRD).

Electrophoretic mobility shift assay
In soybean and Arabidopsis, members of the ERF family contain a conserved DNA-binding domain (AP2/ERF domain) [41]. A digoxigenin-ddUTP-labeled double-stranded oligonucleotide GCC-box probe has previously been combined with the DNA-binding activity of soybean Ethylene Response Factor 5 (GmERF5) [42]. The sequence of the GCC-box probe and the mutated GCC-box probe are shown in S1 Table. The electrophoretic mobility shift assay (EMSA) was performed as described by Kass et al. (2000) [43].

Identification of transgenic atesr1 plants
Plants homozygous for the T-DNA insert were identified via PCR using a gene-specific primer pair and a T-DNA-specific primer. The left genomic primer (LP), right genomic primer (RP) and the left T-DNA border primer (LB) for atesr1 are shown in S1 Table. After confirmation of the homozygous T-DNA insertion, reduction in relative gene expression level was confirmed using qRT-PCR with a gene-specific primer.

Construction of GmESR1 overexpression vector
To overexpress GmESR1 under the control of the Cauliflower mosaic virus (CaMV) 35S promoter, the pEarleyGate 101 vector, containing the bar r gene, was used via the Gateway cloning system. The full-length open reading frame sequence of GmESR1 was used by designing flanking primers for the BP reaction. The reaction mixture was: 1 μl pGWC, 2 μl buffer, 1 μl T4 ligase, 4 μl gene fragment and water to 10 μl, with the ligation carried out at 16˚C. Next, the fragment with adapters at each end was cloned into the entry vector pGWC, which contains chloramphenicol resistance. The entry clone pGWC-GmESR1 was used to perform the LR reaction. The extracted pEarleyGate 101 plasmid, 1 μl of each entry clone, 1 μl LR enzyme, and water to make up the volume to 5 μl was incubated at 25˚C to facilitate the recombination reaction and clone the desired fragment into the pEarleyGate 101 destination vector. The R. radiobacter strain LBA4404 was transformed with the overexpression vector using the freezethaw method as described by Dang et al. (2007) [44].

Arabidopsis transformation and phenotype analysis
Using the method described by Clough et al.(1998) [45], Arabidopsis was transformed with the overexpression vector. Arabidopsis seeds were vernalized in the dark at 4˚C. The seeds were sterilized in 10% sodium hypochlorite, vortexed for 10 min and washed six times using distilled sterile water. The seeds were sown on Murashige and Skoog (MS) solid medium and, after 3-4 leaves had developed, the seedlings were transplanted into 1:1 sterilized soil and vermiculite mixture. Transformation via infiltration was carried out during flowering. Rhizobium radiobacter cells from a single colony were suspended in 15 ml Yeast Extract Peptone liquid culture medium containing selection antibiotics and incubated at 28˚C, with shaking at 185 rpm, until the absorbance at 600 nm was approximately 0.5. When the absorbance at 600 nm reached 1.6-2.0 the mixture was centrifuged for 15 min at 5000 ×g, and the supernatant was discarded. The flowering Arabidopsis plants were inverted and immersed in the Rhizobium liquid for 30 s. Plants were covered with plastic film and placed in a thermostatic chamber without light for 24 h, then placed upright and left to grow with ambient illumination.
The atesr1 mutant lines, transgenic GmESR1 overexpression lines, and WT Arabidopsis plants were grown in the same growth chamber, maintained at 22˚C with a 16 h/8 h (light/ dark) cycle and light intensity of 350 μmolÁm −2 Ás −1 . After 2.5 d, the rates of germination and elongation in the GmESR1 overexpression lines, atesr1 mutant lines, and WT plants were compared and analyzed statistically. After flowering, the height of Arabidopsis were compared.

Soybean transformation and phenotype analysis
Rhizobium-mediated stable soybean transformation was performed using cotyledonary nodes of soybean 'Dongnong 50' as explants. Following culturing in the dark, shoot regenerative proliferation, shoot elongation induction, root differentiation multiplication, and plantlet regeneration, the regenerated plants were transferred to pots and grown in the greenhouse [46]. Five T 1 GmESR1-overexpressing soybean plants and control lines were grown on soybean seed germination medium. After 5 days, each cotyledon separation and 7-8 wounds were induced in the growing point, to the co-culture medium and dark culture for 3 days, transfer into the bud induction medium; co-culture for 14 d at 25˚C under the same conditions. Then the same size buds of GmESR1-overexpressing and control group was isolated and used for observion by electron microscopy.
Analysis of the homolog of GmESR1 in the soybean genome, based on data obtained from the Phytozome database, indicated that the two genes were clustered in two linkage groups, one each on chromosomes Gm 01 and Gm 02, with one and no introns, respectively. Cladistic analysis of GmESR1 against other ESR1s, representing a range of species including crops, fruits, and vegetables, grouped GmESR1 with members of the plant ESR1 family (Fig 2A). The amino acid sequence of GmESR1 showed 68% and 51% similarity to ESR1 proteins from adzuki bean (Vigna angularis; KOM39777) and chickpea (Cicer arietinum; XP_004489775), respectively ( Fig  2B). The predicted three-dimensional structure of GmESR1, based on data from Phyre 2, indicated that the protein contained a long C-terminal α-helix (α) wrapped in a three-stranded anti-parallel β-sheet (from β1 to β3) ( Fig 2C) and that the AP2/ERF domain was divided into conserved segments (YRG and RAYD) (Fig 2B) [47]. The predicted structure of GmESR1 included a conserved region of 57 amino acid residues (residues 51-107) representing the predicted AP2/ERF DNA-binding domain (Fig 2D). Given that residue 14 of the domain was an alanine and residue 19 an aspartic acid, the gene was classified as a member of the ERF subfamily of AP2/ERF TFs. These two amino acids are crucial for specific binding of ERF TFs to the GCC-box in promoter regions and to activate transcription of target genes [16].

Analysis of GmESR1 expression in various organs
To investigate the potential role of GmESR1, its expression profiles were analyzed in major organs of the soybean plant using qRT-PCR. GmESR1 was expressed in all organs analyzed, with the highest relative expression levels observed in young embryos, flower and stem tissue, and considerably lower relative expression levels observed in the pod, leaf and root (Fig 3). GmESR1 may, therefore, play a role in stem elongation, flower morphogenesis, and embryogenesis.

Purification of the recombinant GmESR1 protein
Expression of the recombinant GmESR1 protein was markedly enhanced after 2-8 h induction with 0.5 mM IPTG at 37˚C, attaining the maximum expression level after 4 h, although the recombinant GmESR1 protein was not detected in the control groups (Fig 4A). The molecular weight of the purified GmESR1 protein was approximately 45 kDa as estimated with SDS-PAGE, consistent with the calculated molecular mass of 42.8324 kDa (Fig 4B).

GmESR1 specifically binds to the GCC-box element in vitro
To confirm binding of GmESR1 to the GCC-box regulatory element in vitro, His-tagged GmESR1 was purified and used in an EMSA alongside a digoxigenin-ddUTP-labeled doublestranded oligonucleotide GCC-box probe. The GCC-box and mGCC-box sequences are Functional analysis of GmESR1 shown in Fig 5A. GmESR1 specifically recognized and bound to the GCC-box, but not to the mGCC-box (Fig 5B). When the ratio of unlabeled to labeled GCC probe was 100:1, the labeled probe was not bound, but when 100-fold unlabeled mGCC probe was used as the competitor, the labeled probe was again bound, confirming the specificity of mobility shift (Fig 5B).

Analysis of Arabidopsis atesr1 mutants and overexpression phenotypes in Arabidopsis plants
Using the "double primer" genomic PCR method described by T-DNA Primer Design (http:// signal.salk.edu/tdnaprimers.2.html), homozygous T-DNA insertion mutants were screened for the presence of the transgene, resulting in identification of plants homozygous and heterozygous for atesr1 (Fig 6A and 6B). All WT and transgenic plants were grown under the same  Functional analysis of GmESR1 conditions. The relative expression level of GmESR1 was examined in WT, mutant atesr1, and three independent GmESR1-ox lines. In the three GmESR1-ox lines, the transcript abundance of GmESR1 was high, whereas no transcripts were detected in the WT or atesr1 Arabidopsis plants (Fig 6C). This finding confirmed that these three lines were overexpression of GmESR1. The germination rates of mutant atesr1 and GmESR1-ox seeds sown on solid medium were compared with those of WT seeds. After 2.5 d, the homozygous mutant atesr1 displayed poor germination. Compared with GmESR1-ox, the germination rates of WT and atesr1 seeds were both reduced, though WT seeds showed slightly better germination rates than atesr1 seeds. All GmESR1-ox seeds successfully germinated (Fig 6D). The germination rates of WT, atesr1 and GmESR1-ox seeds are shown in Fig 6E. The germination rate after 2.5 and 6.5 d was higher in GmESR1-ox seeds compared with WT seeds, and higher in WT seeds compared with atesr1 seeds. This finding indicated that overexpression of GmESR1 promoted and accelerated Arabidopsis seed germination, whereas the atesr1 mutation repressed or delayed germination. The elongation rate of GmESR1-ox plants was significantly faster than that of WT plants, whereas atesr1 mutants failed to germinate (Fig 6F). After 2.5 d GmESR1-ox seedlings had longer roots than the WT seedlings, and atesr1 seeds had not germinated (Fig 6G). These findings indicated that GmESR1 overexpression promoted germination and root elongation, whereas the atesr1 mutation delayed germination. At 30d after transplanting, it was observed that GmESR1 overexpression in transgenic Arabidopsis plants resulted in dwarfism (Fig 6H).

Identification and analysis of transgenic soybean overexpressing GmESR1
The soybean cotyledons were used for stable transformation. T 1 seeds were sown in soil, and transformants were identified at the V1 developmental stage when the first trifoliate leaf appeared. Transgenic soybean plants overexpressing GmESR1 were identified via qRT-PCR as those showing higher relative expression of GmESR1 compared with control soybean plants (Fig 7A). Phenotypic analysis of the transgenic soybean plants showed that GmESR1 overexpression resulted in faster germination and elongation relative to the control soybean plants, indicating that GmESR1 promoted seed germination (Fig 7B). After 15 d, at the seedling stage of development, trifoliate leaves had not yet developed in the control soybean but were fully visible in GmESR1-ox plants (Fig 7C). At the same time point, the root elongation of GmESR1ox soybean plants was faster and the root length was longer than in control soybean plants ( Fig  7D). This finding indicated that GmESR1 promoted shoot and root elongation in soybean. The experimental results showed that GmESR1 overexpression resulted in the increasing number of cells relative to the control soybean clustered bud cells in the same size of the field of vision (Fig 7E). It makes the cell division more exuberant.

Discussion
In this study, we identified GmESR1 in soybean, a novel member of the ERF family of TFs, that promoted soybean germination, and shoot and root elongation when overexpressed. Nakano et al. (2006) [22] investigated phylogenetic relationships, gene structure, and the conserved domain of the ERF gene family in Arabidopsis and rice, but only identified a few members of the ERF family in soybean. At 145 and 420 Mb respectively, Arabidopsis and rice have small genomes compared with the genome of soybean (1115 Mb) [48]. The structure of the AP2/ ERF superfamily are relatively similar in all three species [22]. Presently, 98 genes of known function of the ERF family that contain a complete AP2/ERF domain have been identified in the soybean genome [41]. Only six of these genes have been functionally characterized in soybean [42,47,49,50,51,52].
GmESR1 is the first soybean ERF family TF expressed in response to germination and shoot and root elongation, although little information on the role of GmESR1 in these processes is available. Although Arabidopsis AtESR1 genes were first identified in 2001 [14], little is known about their protein functions in vivo or in vitro. In the present study, GmESR1 was shown to contain an AP2/ERF domain divided into two conserved segments, the YRG and RAYD elements. The amino acid sequence, protein structure, results of soybean transformation and other biological functions of the GmESR1 protein indicated that GmESR1 belongs to the ERF subfamily of AP2/ERF TFs. The GmESR1 sequence analysis indicated that features such as the molecular mass, predicted eukaryotic protein phosphorylation sites, acidic isoelectric point and lack of introns are conserved. GmESR1 is located on chromosome two and contains no introns. AtESR1 and AtESR2 occur as a duplication on chromosome one, with genetic data indicating that they are highly redundant during embryonic patterning [36,53]. GmESR1 was analyzed according to publicly available data (http://soybase.org/GlycineBlastPages/) that indicated that 20 genes were clustered into 20 linkage groups.
In embryogenic shoot growth, the cytokinin-induced regenerative genes AtWUSHEL (AtWUS) triggers TOPLESS (TPL) [54], which weakens auxin signaling by interacting with MONOPTEROS/Auxin Response Factor5 (MP/ARF5) and INDOLE-3-ACETIC ACID INDUCIBLE 12/BODENLOS (IAA12/BDL) [55,56]. Banno et al. (2001) [14] isolated and characterized a novel cDNA of which overexpression promotes ultimate cytokinin-independent shoot regeneration from Arabidopsis explants. Given that the cDNA obtained via screening depended on its overexpression as a substitute of cytokinin essential for shoot regeneration, the cDNA might encode elements involved in cytokinin signaling. Overexpression of AtESR1 in Arabidopsis under the control of the estradiol-inducible XVE system also increases shoot regeneration in the presence of cytokinins [14,57]. The function of AtESR2 is similar to AtESR1, with plants silenced for AtESR2 displaying weaker regeneration in general, and increased shoot regeneration in the presence of extra cytokinins [58]. The present results indicated that GmESR1 is responsible for regulation of stem elongation and embryogenesis in soybean and Arabidopsis. As plant regeneration is a complex process requiring the interaction of multiple genes, a single gene is insufficient to regulate the entire process and, therefore, additional study of the interactions between GmESR1 and other regeneration-associated genes is required. Overexpression of AtESR1 and AtESR2 in Arabidopsis has previously been shown to upregulate CUP-SHAPED COTYLEDON 1 (AtCUC1) expression, with the upregulation of AtCUC1 having a positive effect on shoot regeneration [59]. AtCUC1 overexpression is reported to promote the formation of adventitious buds from callus [60]. AtESR1 is important in the conversion of the young lateral root primordium into a shoot meristem, whereas AtESR2 functions in shoot development [61]. AtCUC1 is involved in the same pathway, and AtESR2 and AtCUC2 are functionally redundant in cotyledon development [58]. A previous study of AtESR1 identified phenotypic changes only during embryonic development, whereas our research analyzed the entire plant growth period. Overexpression of GmESR1 in transgenic soybean promoted germination and elongation, resulting in faster shoot development than in WT soybean plants, suggesting that GmESR1 accelerates soybean development and might be a useful tool for regulation of soybean seedling, stem, and root elongation. Overexpression of GmESR1 in Arabidopsis promoted germination and elongation, resulting in faster shoot development than in WT and mutant atesr1 Arabidopsis plants, suggesting that overexpression of GmESR1 accelerates seedling growth stage in Arabidopsis and might be a useful tool for regulation of germination, seedling development, and promotion of root elongation. Conversely, atesr1 delayed germination and the seedling stage in Arabidopsis. In mature transgenic Arabidopsis plants, overexpression of GmESR1 resulted in a dwarf phenotype, whereas the mutant atesr1 showed reduced stem elongation. We have focused on complementation of the Arabidopsis mutant, the relative experiment is complicated and tedious, and no data have been obtained as yet.
Here we demonstrated that overexpression of GmESR1 in soybean and Arabidopsis plants improved shoot germination and elongation, and that purified GmESR1 protein binds to the GCC-box, which is present in the promoter region of many genes. These observations are further supported by confirmation of the interaction of the AP2/ERF domain of AtESR1 with class III homeodomain-leucine zipper (HD-ZIP) TFs [36]. The binding specificity of the AtESR1 protein can be changed by the interaction of the AP2/ERF domain of AtESR1 and the C-terminal Per/Arnt/Sim (PAS)-like domain of class III HD-ZIP proteins, thus a prolonged sequence containing the GCC-box can be recognized by a combination of ESR1 and class III HD-ZIP TFs [31]. The AP2-type TFs DRN and DRNL interact with the bHLH protein AtBIM1, which supports a role for AtBIM1 in embryonic patterning [37]. The interactions between the soybean homologs of AtBIM1 and AtPID and GmESR1s are currently under investigation. At present, research into GmESR1 is at an early stage and additional investigations are needed to clarify its involvement in regeneration. Furthermore, other reasons should be sought to explain why such a unique proteinic structure is formed. The GmESR1 overexpression resulted in more vacuoles relative to the control soybean clustered bud cells in the same size of the field of vision, so can be divide into multiple cells. The GmESR1 overexpression resulted in larger relative to the control soybean clustered bud cell nucleus. After that, the number of cells increased gradually, which could be showed that soybean plants were growing rapidly at germination and seedling stage. Neighboring cells division increased, and the cells divided repeatedly, and the number of divisions increased linearly [62]. During growth and development, the population of stem cells rapidly proliferates to fill the tissues and organs [63]. The GmESR1 gene is functionally analogous to animal stem cells, and the ability to regenerate can increase the number of cells and increase the volume of cells.
The work reported here may be used to further elucidate the division between the regulation of defense mechanisms and shoot regeneration by the ERF family. In the present study, seed germination, and shoot and root growth of GmESR1-overexpressing transgenic soybean plants were faster than those of non-transgenic soybean plants, suggesting that GmESR1 may be involved in the regulation of seed germination, and shoot and root elongation.

Conclusion
We analyzed the function of the soybean GmESR1 gene. In addition, the relationship between the function of GmESR1 and seed germination, and shoot and root elongation was investigated. In soybean GmESR1 overexpression led to faster seed germination, and shoot and root elongation. And by the observation of cell number under the overexpression of GmESR1, the result support that GmESR1 could promote regeneration. These results indicated that GmESR1 may played an important role in seed germination and elongation of soybean.  Table. Oligonucleotide primers used in this study. (DOC)