Introduction

Wheat (Triticum aestivum L.) is an agriculturally important staple food crop for human worldwide. A reliable, efficient, and reproductive regeneration system is a prerequisite for improvement of various economic traits by genetic engineering approaches. However, wheat shows a strong recalcitrance in tissue culture process. Wheat immature embryos have been considered as the best responding explant in tissue culture due to their great potential to generate embryogenic callus, readily followed by the production of large number of genetically fertile transformed plants [1].

The induction of high-quality embryogenic calli is an essential step for the plant regeneration from wheat immature embryo culture. To date, accumulated investigations have been performed to improve the embryogenic callus production. Researchers have optimized several parameters, including genotypes [2–4] and environmental conditions, such as growth conditions of donor plants [5], culture media [3], and incubation conditions [6]. Plant hormones, particularly auxins, were demonstrated to be one of the most important factors for the induction of somatic embryogenesis (SE) [2, 7, 8]. For the in vitro culture of calli, exogenous application of high concentration of auxins was effective for plant regeneration [9, 10]. The addition of 2 mg/l−1 2, 4-dichlorophenoxyacetic acid (2, 4-D) in callus induction medium (CIM) was usually applied in wheat immature embryo culture [11–14]. During the past decade, a traditional protocol, which contains a 4-week culture on the media containing 2, 4-D under darkness before differentiation on 2, 4-D free media under light condition, was reported by many research groups [6–8, 12, 14, 15].

However, the regeneration potential of wheat immature embryo has been proved to be genotype-dependent and the traditional culture protocols were inefficient for most genotypes [16–18]. For example, Fennell et al. [6] obtained a regeneration rate ranging from 2 to 94 % using immature embryos among 48 elite CIMMYT bread wheat cultivars following the traditional protocol.

The difference of embryogenic competence in plants or genotypes can be partly caused by different concentrations of endogenous phytohormones [9]. Embryogenic calli normally had higher concentration of endogenous auxins than non-embryogenic calli in wheat [19]. Pasternak et al. [20] discovered that endogenous auxin levels in the responsive explants of alfalfa (Medicago sativa subsp. varia A2) can also be increased through exogenous addition of 2, 4-D. Nevertheless, prolonged culture in inductive conditions led to the reduction of endogenous auxins levels, which resulted in a reduction of embryogenic capacity [19]. Furthermore, continuous culture in 2, 4-D-containing media inversely hampered the establishment of the polar auxin gradient and hence inhibited embryogenic development in carrot [21]. Therefore, we assumed that the removal of 2, 4-D from the culture regime timely may contribute to the formation of embryogenic calli during wheat tissue culture.

Many researchers have studied the relationship between oxidative stresses and the induction of SE [10, 22–26], and all suggested that oxidative stresses were possibly involved in somatic embryo development. Active oxygen species (AOS) could be one possible link between oxidative stress and plant regeneration in tissue culture [27]. 2, 4-D has been confirmed as stress-related substances for acquisition of embryogenic competence by plant cells [10]. Many antioxidative stress-related genes were identified to be involved in the induction of SE during the application of exogenous 2, 4-D, including catalase (CAT), superoxide dismutase (SOD), ascorbate peroxidase (APX), and peroxidase (POX) [28, 29]. In addition, regenerative calli showed higher induction of these antioxidant enzymes than non-regenerative ones [25]. Nevertheless, the relationship between the AOS and the callus differentiation is still unknown in wheat.

In this study, we developed an improved method to improve the regeneration potential of the immature embryos of twelve wheat accessions by removal of 2, 4-D from CIM earlier than the traditional culture regimes. Furthermore, 16 cDNA/ESTs highly relevant to phytohormone-regulated pathways were screened to investigate the candidate gene(s) responsible for wheat regeneration capacity. Our study provides valuable information for the improvement of in vitro regeneration of wheat, which will potentially facilitate the enhancement of the efficiency of stable wheat transformation.

Materials and Methods

Plant Materials

Twelve wheat accessions were used in this study. Based on our previous research, two groups were assigned in terms of their regeneration capacity [30], including Bobwhite, CB037, KN199, LX987, and XC9 with high regeneration capacity and J411, J771, JM22, Z8601, NC4, NC16, and CS with low-regeneration potential. T. aestivum L. Bobwhite seeds were kindly provided by Dr. Tom Clemente at the University of Nebraska-Lincoln, USA. All the other accessions were maintained in the germplasm collections in Chinese Academy of Agricultural Sciences (CAAS).

Immature Embryos Culture

Wheat immature seeds at 12–14 days post anthesis (DPA) were collected and surface sterilized with 70 % ethanol for 2 min, 20 % sodium hypochlorite (NaClO) for 15 min, and rinsed three times with sterile water. Immature embryos of 1.0–1.2 mm in size were aseptically excised from immature seeds after the embryogenic axes were carefully removed. The immature embryos were then incubated with scutellum upwards on CIM-2 in 90-mm petri dishes in the dark at 25 ± 1 °C. The CIM-2 contained MS [31] inorganic, 1.0 mg/l−1 vitamin B1, 150 mg/l−1 asparagine, 2.0 mg/l−1 2, 4-D, 30 g/l−1 sucrose, and 2.4 g/l−1 gelrite (pH 6.0). In total, about 800 immature embryos for each wheat accession were incubated on 10 petri dishes in each treatment. Each treatment was performed in triplicate. For the traditional method, the immature embryos were put on CIM-2 for 22 days. For our improved method, the immature embryos were put on CIM-2 for 3, 5, 8, and 10 days, separately, then moved onto CIM-0 (CIM-2 subtracting 2, 4-D) for 3–12 days with 3-days interval, followed by another 2 days on CIM-2. The culture protocol was shown in Fig. 1. The granular and compact embryogenic calli were counted, and transferred onto shoot forming medium (SM) at 25 ± 1 °C, with a 16/8 h photoperiod and a photosynthetic photon flux density of 400 μmol/m−2/s−1 at a relative humidity of 45 %. After 2 weeks, regenerated green shoots with normal shape were counted, and statistical analysis was conducted based on methods of Chen et al. [32] and Taguchi-Shiobara et al. [33]. The efficiency of the improved method versus the traditional method of regeneration was compared by calculating EIE (the number of embryogenic calli/the number of incubated immature embryos × 100 %), GDE (the number of calli with green spots/the number of incubated immature embryos × 100 %), PRE (the number of plants regenerated/the number of incubated immature embryos × 100 %).

Fig. 1
figure 1

Culture protocol of 2, 4-D-removal treatment versus traditional treatment for immature embryos of wheat. a Traditional culture condition. The immature embryos were put on CIM-2 for 22 days for callus induction, followed by 2 weeks on SM. b Immature embryo culture on CIM-2 for 3, 5, 8, and 10 days before removing 2, 4-D from CIM-2. c Immature embryo culture on CIM-0 for 3, 6, 9, and 12 days, separately, after incubation on CIM-2 for 8 days. Black and white boxes represented culture on CIM-2 under the dark condition and SM under the light condition, respectively. The gray box showed the incubation on CIM-0 under the dark condition. The time durations of each culture stage were shown above the boxes

Full size image

RNA Extraction, cDNA Synthesis, and Semi-Quantitative RT-PCR Assay of 16 ESTs/cDNAs

Total RNA was isolated from wheat calli induced by two different methods using TRNzol (Tiangen, Beijing, China) according to manufacturer’s manual with some modifications. In brief, the frozen tissues were pulverized in liquid nitrogen and mixed thoroughly in 1 ml TRNzol buffer. After incubation for 20 min at 65 °C, the mixture was centrifuged at 12,000 g for 30 min at 4 °C. The supernatant was mixed with 200 μl chloroform, and the RNA in the aqueous phase was precipitated with isopropanol at −20 °C, and dissolved in diethylpyrocarbonate (DEPC)-treated water following washing twice in 70 % ethanol. The first-strand cDNA was synthesized after the removal of genomic DNA using one Primed cDNA Synthesis Kit (TaKaRa, Japan) according to manufacturer’s instructions.

In this study, 16 ESTs/cDNAs relevant to auxin metabolic pathways were investigated to explore the regeneration-related candidate gene(s) in the immature embryo culture of wheat accessions with high regeneration capacity (CB037 and Bobwhite) and low-regeneration capacity (NC4 and CS) by semi-quantitative RT-PCR (Table S1). Two of them (RT15 and RT16) were just derived from wheat. The other 14 genes were chosen based on the alignment of amino acid sequences between wheat and mustard by blastp. Calli from the same developmental stage with different regeneration potentials were assessed to ascertain the correlation of regeneration trait with gene expression. For semi-quantitative RT-PCR, 1 μl from 50 × diluted cDNA stocks was used as template and mixed with 0.2 μM of each primer, 100 μM dNTP and 0.15 U Taq DNA polymerase (Transgene, China). PCR reactions were performed in C1000 thermal cycler (Bio-Rad, USA) using the following protocol: 1 cycle of 5 min at 94 °C followed by 25 cycles of 30 s at 94 °C, 30 s at 58–65 °C, and 30 s at 72 °C, and a final extension cycle at 72 °C for 5 min. The amplified products were resolved by electrophoresis using a 1.5 % (w/v) agarose gel. Wheat actin gene (GenBank accession no. AB181991) was used as an internal control to normalize the quantity of cDNA template. All the amplifications were performed in triplicate. The primer sets used for semi-quantitative RT-PCR were listed in Table S1.

Measurement of H2O2 Content

According to the protocol described by Zhang et al. [34], H2O2 contents in callus tissues on SM at different time points (5, 10, 15, 20, 25, and 30 days) were measured from ten wheat accessions including Bobwhite, CB037, LX987, KN199, XC9, CS, JM22, J411, J771, and NC4. The standard curve was made by using seven known concentrations of H2O2 at 0, 1.6, 3.2, 6.4, 9.6, 12.8, and 16 μ/mol. H2O2 content (μ/mol/g−1) was calculated using the formula: \( {\raise0.7ex\hbox{${C*Vt}$} \!\mathord{\left/ {\vphantom {{C*Vt} {FW*V1}}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{${FW*V1}$}} \). (C: H2O2 concentration based on standard curve; Vt: total volume of sample extraction; V 1: sample volume for measurement; FW: fresh weight.)

Isolation of 5′ Flanking Sequence

Inverse PCR (I-PCR) was conducted to isolate 5′ flanking sequence of TaCAT1 according to Ochman et al. [35]. In brief, EcoRI, SacI, and BamHI were employed to digest wheat genomic DNA followed by self-ligation to form circular DNA molecules and then used as templates in three rounds of extension. Two nested pairs of primers (F1/R1 and F2/R2, Table S2) were applied in each round. The PCR products derived from F1/R1 primer pairs were 50× fold diluted and then used as the template for amplification using F2/R2 primer pairs. PCR products were recovered from 1 % agarose gel and ligated into pMD18-T (TaKaRa, Japan) for sequencing. Putative cis-acting elements were identified based on online database (PlantCARE: plant cis-acting regulatory elements at http://bioinformatics.psb.ugent.be/webtools/plantcare/html/, Lescot et al. [36]). TSSP software was used to predict TATA-box and transcriptional start site (TSS) (http://linux1.softberry.com/berry.phtml?topic=tssp&group=programs&subgroup=promoter).

Results

Optimization of Auxin-Removal Treatment in the Immature Embryo Culture of Wheat

To improve the regeneration potential of immature embryo in wheat, we modified the callus induction treatment by removing 2, 4-D for some days from the CIM. The immature embryos of Z8601 showed very low-regeneration potential in immature embryo culture [30] in vitro and thus was selected for optimization of auxin-removal treatment aiming at improving regeneration potential of wheat immature embryo culture. Before 2, 4-D-removal, we put the immature embryos on CIM-2 for 3, 5, 8, and 10 days, and found that 8-day (d) treatment on CIM-2 followed by 2, 4-D-removal for 12 days resulted in the best results with 82.5 % of EIE, 91.5 % of GDE, and 286 % of PRE (Fig. 2a). Compared to 3-day induction on CIM-2, other induction time can result in higher values of EIE, GDE, and PRE except for EIE at 10-day induction. However, extended culture time on CIM-2 seemed not to have positive effect on regeneration rate since incubation on CIM-2 for 10 days can result in lower EIE, GDE, and PRE compared to 8-day induction on CIM-2. The effect of different culture time on CIM-0 after culture on CIM-2 has been also tested. The result showed that 12-day withdrawal led to better regeneration efficiency of wheat immature embryos than 3-, 6-, and 9-day incubation on CIM-0 (Fig. 2b). Based on the results in Fig. 2, 8-day induction on CIM-2, 12-day on CIM-0 and another 2-day on CIM-2 were chosen as the improved culture protocol.

Fig. 2
figure 2

Effects of different culture time on CIM-0 and CIM-2 on callus induction and differentiation efficiency of immature embryos from Z8601 with low regeneration efficiency. a Effect of different time on CIM-2 for regeneration rates of immature embryos. Before the calli were moved into CIM-0, they were kept on CIM-2 for 3, 5, 8, and 10 days, separately (Fig. 1b). EIE, GDE, and PRE were measured for callus induction and differentiation efficiency at each time point. b The effect of 2, 4-D-removal from CIM on the callus induction and differentiation efficiency. After 8-day culture on CIM-2, the calli induced from immature embryos were moved into CIM-0 for 3, 6, 9, and 12 days, separately. EIE, GDE, and PRE were measured for callus induction and differentiation efficiency in each time point of 2, 4-D-removal. Significant differences between the mean values of H2O2 content were determined using the general linear model (GLM) procedure in SAS programs. One or two asterisks over the histogram represent significant differences at the 0.05 or 0.01 probability levels, respectively. Statistical analysis was performed based on the comparison between 3-day culture on CIM-2 and other culture time in (a) and 0-day on CIM-0 and other culture time in (b)

Full size image

To further compare the improved culture method with the traditional culture method, we chose two wheat accessions with high regeneration potential (XC9 and KN199) and three accessions with low-regeneration potential (CS, NC4, and NC16) for regeneration experiments. The traditional method was depicted as in Fig. 1a. The improved protocol led to the induction of more embryogenic calli with pale, smooth, and compact nodules, whereas the conventional culture protocol led to the production of calli of cream color with soft watery texture (Fig. 3). In terms of EIE, NC4, NC16, CS, and XC9 showed significantly higher values (3–5 fold increase) than the traditional culture method (Table 1). The improved culture method also resulted in higher GDE and PRE than the traditional culture method. On average, the regeneration efficiency using the improved culture method was increased by 2–8 folds in PRE, showing significant difference over the traditional culture method in the five genotypes tested (Table 1). However, no significant difference was found in EIE and GDE for KN199, which may be partially caused by its relatively easier regeneration potential compared to other wheat genotypes. Overall, the result demonstrated that culture regime on CIM-2 for 8 days followed by 12 days on CIM-0, another 2 days on CIM-2 and 2 weeks on SM resulted in greatly improved regeneration potential of the freshly isolated immature embryos.

Fig. 3
figure 3

The effect of 2, 4-D-removal on regeneration efficiency of wheat immature embryos. a, b traditional culture method as depicted in Fig. 1 a, c, d the improved culture protocol including culture on CIM-2 for 8 days, CIM-0 for 12 days, CIM-2 for 2 days, and SM for 2 weeks in turn. a, c T. aestivum cv. CS with low regeneration potential; b, d T. aestivum cv. XC9 with high-regeneration potential

Full size image
Table 1 The effect of new culture method on the regeneration efficiency of 5 wheat accessions
Full size table

Expression Profile of 16 Selected ESTs/cDNAs

In this study, semi-quantitative RT-PCR was performed to investigate the expression of 16 ESTs/cDNAs relevant to auxin metabolic pathways in the immature embryo culture using wheat accessions with high regeneration capacity (CB037 and Bobwhite) and low-regeneration capacity (NC4 and CS). The similarities of the 14 ESTs/cDNAs between wheat and mustard (B. juncea) ranged from 29 to 91 % (Column 5 in Table S1). The results showed RT11 which encodes CAT had higher expression level in the calli with high regeneration capacity than in those with low-regeneration capacity at the same developmental stage based on the improved culture method, especially in the calli from CS (Fig. 4a). In contrast, the expression profiles of other genes displayed inconspicuous differences among the tested wheat accessions and between different developmental stages (Fig. S1).

Fig. 4
figure 4

Expression profile of RT11 from five wheat accessions during induction and regeneration stage. a The names of four wheat accessions (Bobwhite and CB037 showing high-regeneration potentials and NC4 and CS showing low regeneration potentials in immature embryo culture) and internal control (actin) were shown on the left. The pictures at the bottom represent the calli from 0-, 5-, and 10-day (d) culture on SM with low- and high-regeneration potential marked by the number 1, 3, 5, and 2, 4, 6, respectively. b: RT11 expression profile in common wheat cv. Z8601 calli subjected to 2, 4-D withdrawal for 0, 3, 6, 9, and 12 days separately after 8 days incubation on CIM-2

Full size image

RT11 was deposited in GenBank with accession No. GU984379, which encodes a deduced 492-amino acid (aa) protein. Based on homologous alignment of amino acid sequences, it showed 81.5 % identity with maize cat1 (X12538), 85.5 % with maize cat2 (J02976), and 67.8 % with maize cat3 (X12539). In addition, RT11 shared 73.1, 81.5, and 91.3 % identities with catA (D29966), catB (D64013), and catC (D86611) from O. sativa L., respectively. This candidate gene was designated as TaCAT1. Furthermore, there were three amino acid substitutions between high- and low-regeneration wheat accessions, viz, Y278 versus H278, E280 versus D280, and S317 versus A317 (Fig. S2), indicating that the three amino acids may be the key amino acids accounting for different regeneration potentials among different wheat genotypes.

To further investigate the effect of 2, 4-D-removal on the expression patterns of TaCAT1 in the immature embryos-derived calli, expression profile was constructed using Z8601 as materials (Fig. 4b) under the same PCR conditions as shown in Fig. 4a. The expression of RT11 increased gradually from 0 to 9 days, with a peak at the 9-day 2, 4-D-removal treatment and then a drop at 12 days, indicating the regulatory role of 2, 4-D-removal treatment on TaCAT1 transcript accumulation.

Changes in H2O2 Content During SE on SM

To further explore the changes of H2O2 followed by 2, 4-D-removal treatment, H2O2 content was measured in the immature embryos-derived calli from the improved culture method and the traditional approach on SM using 10 wheat accessions, encompassing the high-regeneration and low-regeneration wheat genotypes. The H2O2 accumulation in both methods increased gradually from day 5 to 15, then a sharp drop happened at day 20, after that it had some increases at day 25 and 30 (Fig. 5). H2O2 content was significantly higher in the improved method than the traditional method for most of the accessions tested from day 5 to 30 at the 0.05 level (Fig. 5a, b). Furthermore, H2O2 content was a little higher in plantlets from high-regeneration wheat genotypes than low-regeneration ones on all the checked time points, suggesting that 2, 4-D-removal from CIM stimulate H2O2 accumulation in the following regeneration process.

Fig. 5
figure 5

Changes on H2O2 content during SE at different time points. H2O2 content was measured from the calli on SM for 5, 10, 15, 20, 25, and 30 days, separately. a Improved method including culture on CIM-2 for 8 days, CIM-0 for 12 days, CIM-2 for 2 days. b Traditional culture regime as depicted in Fig. 1a. Wheat accessions with high-regeneration capacity: Bobwhite, CB037, LX987, KN199, XC9; wheat accessions with low regeneration capacity: CS, JM22, J411, J771, NC4. Significant differences between the mean values of H2O2 content were determined using the GLM procedure in SAS programs. Asterisks over the histogram represent significant differences on H2O2 content between improved method and traditional method at the 0.05 probability level

Full size image

Isolation and Cis-Acting Element Analysis of TaCAT1 Promoter

Based on the overlapping sequences caused by I-PCR method, a 3,500-bp promoter region upstream of the TaCAT1 start codon ATG was isolated from genomic DNA of wheat cultivar CB037 based on three rounds of extension from EcoRI, SacI, and BamHI digested genomic DNA successively (1.4, 1.9 and 1.8 kb in Fig. S3 A, B, C, respectively). The promoter sequence of TaCAT1 has been deposited in GenBank with accession no. HM989895. Since the correct TATA-box site could not be predicted using PlantCARE, the TSSP database was employed to define TATA-box and TSS. As a result, a most likely TATA-box was located 35 bp upstream from TSS and appropriately 250 bp downstream of the CAAT box with linear discriminant function (LDF) value of 19.49, which met the classical location of TATA-box [37]. The TSS was situated −13 bp upstream from ATG with LDF of 0.15. In total, 43 cis-acting elements were found, including 21 on sense strand and 22 on antisense strand (Fig. S4). Of them, two regulatory elements (TGA-element) relevant to auxin were found located at −597 and −746 bp from start codon ATG.

Discussion

Plant regeneration is a process under genetic as well as physiological control [38]. Correspondingly, transfer of key regeneration genes to recalcitrant cultivars or species seems more direct and convenient to improve the regeneration capacity of plants than traditional backcrosses. Furthermore, an efficient in vitro regeneration system and a reliable procedure for gene transfer are two important aspects contributing to the success in plant genetic transformation and overcoming the limiting factors for quality improvement of most cereal recalcitrant species [11]. Nevertheless, the information for key genes controlling plant regeneration is limited, which greatly restrain the improvement of plant regeneration ability in a large scale, especially in monocotyledonous cereal crops, such as maize, rice, and wheat.

Auxins, as a key group of phytohormones, play an important role in recuperating the embryogenic potential of the mitotically quiescent or differentiated somatic cells on their exogenous application in the induction media [39]. The acquisition of embryogenic competence largely relies on proliferative dedifferentiation due to the presence of auxin which prevents precocious germination (for review, Fehér et al. [40] and Caliskan et al. [41]). Furthermore, based on the investigation on auxin distribution and transport during in vitro induction of SE in wheat, Fischer-Iglesias et al. [42] found that during the induction phase of embryo development (also known as radially symmetrical embryos development stage), inclusion of auxin transport inhibitor 2, 3, 5-triiodobenzoic acid (TIBA) in culture media can inhibit dedifferentiation and exacerbate the formation of morphologically abnormal embryos in culture, since the embryos do not yet have organs required for auxin biosynthesis. However, once competent calli have been induced, the role of exogenous auxin changed since endogenous auxin was produced following some auxin-producing organs formation (such as coleoptile, suspensor, and etc.) [43]. These indicate that timely removal of exogenous auxin application may contribute to SE. In this study, proper timing of 2, 4-D-removal during induction stage greatly improved regeneration potential of wheat immature embryos. In addition, Okubara et al. [12] pointed out that application of low concentrate of 2, 4-D during the transition from induction to regeneration stage contributed to wheat plantlet regeneration. However, we found that fully removing 2, 4-D from CIM after 8 days on CIM-2 during callus induction stage can produce the same good results. This may be due to the stimulation on the high accumulation of endogenous auxins triggered by the withdrawal of 2, 4-D during the induction stage.

On the other hand, Caliskan et al. [41] reported that auxin treatment can stimulate rapid accumulation of “germin-like” oxalate oxidase (gl-OXO) mRNA. To identify gene(s) accumulation during SE, 16 ESTs/cDNAs were selected based on the studies of Singla et al. [39] and Gong and Pua [44]. Although they have investigated those genes expression profile and grouped them into a subgroup relevant to auxin metabolic pathways by screening and sequencing cDNA library and mRNA differential display methods during auxin-induced SE in wheat and mustard (Brassica juncea), they had not dissected those genes’ expression in immature embryo-derived calli from different wheat accessions or genotypes. In view of the above, semi-quantitative RT-PCR analysis was utilized to identify the candidate ESTs/cDNAs during SE using wheat immature embryos as explants. We successfully found the correlation of TaCAT1 expression with the calli of high-regeneration potential (Fig. 4a). Further investigation showed the tight link of TaCAT1 expression with the 2, 4-D-removal treatment during induction stage (Fig. 4b). Further extension of 2, 4-D withdrawal treatment on CIM (such as 12 days on CIM-0) led to the low expression of TaCAT1. Besides, measurement on H2O2 content was also performed in this study (Fig. 5). Gradual increase of H2O2 content in the first 2 weeks at the beginning of SM, following the withdrawal of 2, 4-D on CIM may in part account for the above SE processes. The increase of H2O2 content with a concomitant decrease of CAT activity suggested that H2O2 produced in excess due to the disruption of oxidative balance may promote the expression of some genes responsible for the induction of morphogenic processes [23]. Tian et al. [45] reported H2O2 production coincided with emergence of meristemoid and formation of bud primordium in morphogenic calli during shoot organogenesis of strawberry, and high O2 − and low H2O2 level were detected in calli possessing low-organogenesis capacity. In contrast, Papadakis et al. [5] reported that H2O2 content was 2-fold lower in totipotent tobacco protoplasts compared with non-totipotent tobacco protoplasts. However, in the current study, the accumulation of H2O2 on SM in the improved method (Fig. 5) and expression of TaCAT1 in calli with high-regeneration potential (Fig. 4a) were both higher. This indicates that further investigation is essential to unravel the relationship of SE with H2O2 production and CAT activity during the whole wheat embryogenic callus regeneration process.

Plant regeneration process requires the perception and transduction of specific signals and the expression of genes blocked in non-regenerative plants [24]. In silico analysis on the promoter of TaCAT1 revealed many cis-regulatory elements. Of those, two auxin-responsive elements (TGA-element) were found, which may contribute to the induced expression of TaCAT1 by the application of exogenous auxin (Fig. S4). Besides, three ABREs (ABA-responsive elements, located at −80, −167, and −3480 bp from start codon ATG) were also identified. Zhang et al. [34] discovered that ABA promoted SE in Larix leptolepis, induced production of H2O2 and other AOS, and mediated CAT, SOD, and APX gene expression during SE in L. leptolepis. Jiménez and Bangerth [19] found higher ABA levels in the competent wheat genotypes. Since ABA biosynthesis was known to be increased after various abiotic stresses, it has been proposed that increased levels of ABA might induce SE (review by Jiménez [9]). Based on promoter analysis of TaCAT1, further investigation should focus on the expression of TaCAT1 stimulated by the two phytohormones. Based on the I-PCR results, we also isolated and aligned the promoter sequence of TaCAT1 from the ten wheat accessions. However, there is no obvious difference in the 5′ flanking sequence of TaCAT1 between high- and low-regeneration wheat genotypes except some minor differences in SNPs. Considering the three amino acid substitutions in TaCAT1 protein from wheat genotypes with high-and low-regeneration potentials, it is essential to discern what is more important in improving wheat regeneration trait, the promoter or the TaCAT1 itself?