Introduction

Salinity stress negatively impacts agricultural yield throughout the world, affecting production, whether it is for subsistence or economic gain. The plant response to salinity consists of numerous processes that must function in coordination to alleviate both cellular hyperosmolarity and ion disequilibrium (Tester and Davenport 2003). In addition, crop plants must be capable of satisfactory biomass production in a saline environment. Recent progress in the elucidation of salt stress signalling and effector output determinants that mediate ion homeostasis has uncovered some potential biotechnology tactics that may be used to obtain salt-tolerant crop plants, i.e. enhance the yield stability under salinity. However, it has been shown that high constitutive expression of the foreign gene may be detrimental to the host plant, with independent reports of increased sterility, retarded development, abnormal morphology, yield penalty, altered grain composition or transgene silencing (Sinha et al. 1993; Xu et al. 2006). The use of a strong, tissue-specific or inducible promoter to restrict gene expression to the required tissue, at a particular developmental stage and/or in response to a stress, may solve this type of problem (Karim et al. 2007; Pino et al. 2007). Until now, numerous inducible promoters have been isolated from a wide variety of organisms. Among these, the biotic and abiotic stress-inducible rd29A promoter has been widely used to minimise the otherwise negative effect on plant growth of transgene expression in plants such as tobacco, sugarcane and potato (Kasuga et al. 2004; Behnam et al. 2006; Wu et al. 2008).

Signal transduction networks for abiotic stress are divided into three major signalling types: osmotic/oxidative stress signalling involves the generation of reactive oxygen species (ROS)-scavenging enzymes and antioxidant compounds, as well as osmolytes; Ca2+-dependent signalling that leads to the activation of late embryogenesis abundant (LEA)-type genes; and Ca2+-dependent salt overly sensitive (SOS) signalling (Xiong et al. 2002). The SOS signalling pathway, which comprises SOS3, SOS2 and SOS1, is a pivotal regulator of, at least some, key transport systems required for ion homeostasis (Sanders 2000; Zhu 2000). Loss-of-function mutations in SOS3, SOS2 and SOS1 cause hypersensitivity to Na+ (Wu et al. 1996; Zhu et al. 1998). The direct downstream target of this pathway is SOS1, a plasma membrane Na+/H+ antiporter (Shi et al. 2000; Qiu et al. 2002; Quintero et al. 2002). SOS3 is a calcium-binding protein (Liu and Zhu 1998; Ishitani et al. 2000), while SOS2 is a Ser/Thr kinase. Both the catalytic and regulatory domains are essential for SOS2 function in salt tolerance. The C-terminal regulatory domain contains an auto-inhibitory FISL motif that binds to SOS3 or calcineurin B-like 10 (CBL10), thereby releasing SOS2 from auto-inhibition (Liu et al. 2000; Guo et al. 2001; Quan et al. 2007). SOS3–SOS2 and CBL10–SOS2 complexes regulate the plasma membrane Na+/H+ exchange activity of SOS1 through phosphorylation, which releases AtSOS1 from auto-inhibition (Quintero et al. 2002, 2011; Quan et al. 2007). In addition to Arabidopsis, the SOS1 gene has been identified in other plants like rice (Martínez-Atienza et al. 2007), wheat (Xu et al. 2008; Feki et al. 2011), tomato (Olías et al. 2009) and Thellungiella salsuginea (Oh et al. 2009). Despite the demonstrated role of some SOS1 genes in ion homeostasis and in the partitioning of the toxic ion Na+ between plant organs (Shi et al. 2002; Olías et al. 2009), only experimental functional analysis of Arabidopsis SOS1 and Salicornia brachiata SbSOS1 promoter was performed in transgenic Arabidopsis and tobacco plants, respectively (Shi et al. 2002; Goyal et al. 2013). Arabidopsis SOS1, SOS2 and SOS3 promoter regions present the same cis elements. However, the predicted cis elements present in SOS2 promoter are higher than SOS1 and SOS3, with the presence of many small RNA target sites. This in silico analysis indicates that these three SOS promoter regions could be involved in common features of transcriptional regulation, and the SOS2 gene is regulated by several upstream transcription factors and is involved in other outputs besides Na+ transport by SOS1 (Ji et al. 2013). Despite the extensive synteny of the open reading frames (ORFs) and the conservation of gene structures for SOS1 between Arabidopsis and its halophytic relative (Thellungiella parvula), the promoter region of this gene is conserved only between the Thellungiella species (Oh et al. 2010).

It has been shown previously that Arabidopsis SOS1, SOS2 and SOS3 genes have different special expression profiles. The first two genes are expressed in both roots and shoots, while the latter is mainly expressed in root tissues (Liu et al. 2000; Shi et al. 2002; Quan et al. 2007). The expression of these genes is induced by salt stress in roots, but only SOS3 expression displays differential induction in various types of root cells located in different developmental root zones in response to salt stress (Ji et al. 2013). Arabidopsis AtSOS1 is preferentially expressed in epidermal cells at the root tip and in parenchyma cells at the xylem/symplast boundary of roots, stems and leaves (Shi et al. 2002). Concerning SOS1 gene expression, salt challenge induces a clear accumulation of SOS1 mRNA (Martínez-Atienza et al. 2007; Xu et al. 2008; Olías et al. 2009; Tang et al. 2010; Wang et al. 2010). Strikingly, in some cases, salinity produced little or even no alteration of SOS1 transcription (Taji et al. 2004; Kant et al. 2006; Mullan et al. 2007; Wu et al. 2007; Cosentino et al. 2010; Feki et al. 2011). To our knowledge, the expression profile of wheat SOS1 has been analysed under different abiotic stresses, but the wheat SOS1 promoter activity has not yet been experimentally analysed. In this study, we showed that the two isolated wheat SOS1 promoter regions (Pr SOS1-D and Pr SOS1-AB ) are active, age-dependent and organ-specific promoters in Arabidopsis plants. Moreover, we demonstrated that, contrary to Pr SOS1-D , Pr SOS1-AB is an abiotic stress-inducible promoter at different developmental stages (8, 20 and 30 days old). Thus, the Pr SOS1-AB promoter will be useful for specific spatiotemporal targeting and accumulation of proteins conferring tolerance to abiotic stresses in transgenic plants.

Materials and methods

Genomic library screening and isolation of wheat SOS1 promoter regions

The bacterial artificial chromosome (BAC) library from Triticum aestivum (cv. Chinese spring) was screened using polymerase chain reaction (PCR) amplifications as described by Isidore et al. (2005). Amplifications were done using wheat SOS1-specific primers, which are S1 as a sense primer and S2 as a reverse primer (Table 1). The 5′-flanking region of the wheat SOS1 gene was isolated using the inverse PCR method as described by Ochman et al. (1988). PCR reactions were carried out with the corresponding recombinant BAC clone as a template, and with the wheat SOS1-specific primers IS1 and IS2, designed close to the 5′UTR sequence (Table 1). Two promoter sequences, named Pr SOS1-D and Pr SOS1-AB of 2,660 and 2,745 bp, respectively, were obtained by sequencing.

Table 1 Sequences of primers used for genomic library screening, inverse PCR and cloning of the two wheat SOS1 promoter regions in the pCAMBIA1391Z vector
Full size table

In silico analysis of the wheat SOS1 promoters

The search for putative cis elements in the two promoter sequences was carried out using the signal scan search provided by the PLACE (http://www.dna.affrc.go.jp/PLACE/) and the PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) databases (Higo et al. 1999; Lescot et al. 2002).

Construction of the binary vectors and Arabidopsis transformation

Fragments of the two promoters Pr SOS1-D and Pr SOS1-AB were amplified using the two different couples of primers ES1–NS1 and ES2–NS2, respectively. The two forward primers ES1 and ES2 harbour the PstI restriction site, and the two reverse primers NS1 and NS2 harbour the NcoI restriction site. The two resulting fragments were cloned separately in front of the gusA gene into the pCAMBIA1391Z vector (Cambia, Canberra, Australia), using the PstI and NcoI restriction sites. The resulting recombinant binary vectors were named pCAMBIA1391Z–Pr SOS1-D –gusA and pCAMBIA1391Z–Pr SOS1-AB –gusA. Agrobacterium tumefaciens strain GV3101 (Konez and Schell 1986) was transformed separately by the freeze–thaw transformation method (Chen et al. 1994) with these two binary vectors and the binary vector pCAMBIA1301, which were subsequently used for Arabidopsis transformation. The Arabidopsis thaliana transformation was performed using the floral dipping technique (Clough and Bent 1998). Transgenic plants were selected on Murashige and Skoog (MS) agar medium (Murashige and Skoog 1962) containing 20 μg L−1 hygromycin. Seeds of the T2 generation were harvested and the seedlings from the homozygous T3 generation were used for histochemical β-glucuronidase (GUS) staining. The wild-type Arabidopsis and 35S-gusA transgenic plants were used as negative and positive controls, respectively.

Identification of the transgenic Arabidopsis plants

Genomic DNA was extracted as described by Michiels et al. (2003) from the leaves of Pr SOS1-AB –gusA and Pr SOS1-D –gusA transformed Arabidopsis, and was used in PCR amplifications. To detect positive lines, the couples of primers from the gusA gene (GR) and flanking the promoter Pr SOS1-AB region (ABF) or the promoter Pr SOS1-D region (DF) were used (Table 2). The amplified products were resolved on a 1 % agarose gel and visualised by ethidium bromide staining.

Table 2 Sequences of primers used for the analysis of the transgenic Arabidopsis plants by PCR amplifications and semi-quantitative RT-PCR
Full size table

For each Arabidopsis transformation, four positive transgenic lines were randomly selected to analyse the expression of the gusA gene. From 20-day-old seedlings, total RNA was isolated using the Trizol method (Invitrogen), and to remove contaminating DNA, it was then treated with RNase-free DNaseI (Promega) at 37 °C for 15 min and further incubated at 65 °C for 10 min. DNase-treated RNA samples (0.5 μg) were reverse-transcribed using M-MLV reverse transcriptase (Invitrogen). The reverse transcription (RT) reactions were performed at 37 °C for 1 h and using the oligo-dT (18 mer) primer. One microlitre of each cDNA was used as the template for PCR amplification with 2 units of Taq DNA polymerase (Invitrogen), 200 μM dNTPs and 0.5 μM gusA-specific primers (GusF–GusR) (Table 2). An Arabidopsis thaliana β-tubulin gene fragment (GenBank accession no. XM_002863542.1), used as an internal control, was amplified with the forward TubF and the reverse TubR primers (Table 2). The PCR products (5 μl) were separated on 1.5 % agarose gel.

Growth conditions, abiotic stress treatment and gusA expression analysis

The effect of abiotic stress on gusA transcripts accumulation was monitored using seeds of homozygous transgenic lines and the non-transformed plants. Seeds of each T3 homozygous transgenic Arabidopsis lines were surface-sterilised and then grown on MS agar medium under light/dark cycle conditions of 16 h light/8 h dark cycles at 22 °C. Seedlings were grown in MS agar medium for 8, 20 or 30 days and then transferred to MS agar medium containing 100 mM NaCl, 100 mM mannitol or 20 μM abscisic acid (ABA), and kept for 2 days in each treatment. Then, the plants were harvested for analysis by histochemical GUS staining and for gusA expression analysis. Total RNA was extracted from 20- and 30-day-old seedlings, and treated with RNase-free DNaseI. The cDNA was synthesised by means of the M-MLV reverse transcriptase (Invitrogen). PCR amplifications were performed using the primers GusF and GusR (Table 2) and the PCR-amplified products were visualised on ethidium bromide-stained 1.5 % agarose gels and quantified using the Gel DocXR Gel Documentation System (Bio-Rad). This software was used to calculate the average band density, which was recorded and used in graphic analyses. The band density was determined by this software and was given in arbitrary units and graphed using Microsoft Excel. The error bars were determined from three separate biologic replicates. Each of the three biological replicates consisted of pooled plants subjected or not to different stress conditions.

Histochemical GUS staining

GUS activity was assayed histochemically by incubating tissue under vacuum infiltration with GUS staining solution [50 mM Na2HPO4 buffer (pH 7.0), 0.5 mM potassium ferricyanide, 0.5 mM potassium ferrocyanide, 0.1 % Triton X-100 and 1 mg/l X-Gluc (5-bromo-4-chloro-3-indolyl β-D-glucuronide)] for several minutes and then incubated overnight at 37 °C (Jefferson et al. 1987). The pigments and chlorophyll were removed by soaking the Arabidopsis tissues for several hours in 70 % (v/v) ethanol. Three to six stained plants from three independent experiments were observed under binocular loupe and photographed using an Olympus W120 digital still camera, and most of the photos showed similar results.

Results

Isolation and in silico analysis of the two wheat SOS1 promoter regions

A genomic library from bread wheat (Triticum aestivum) was screened by PCR with specific primers of the wheat SOS1 gene, which resulted in the isolation of two different BAC clones, BAC-AB and BAC-D, containing the wheat SOS1 alleles that are localised on genomes A and/or B, and on genome D, respectively. In order to obtain the promoter region of the bread wheat TaSOS1 gene, we performed an inverse PCR using SOS1-specific primers and the two BAC clones obtained as a template. 2,661-bp (Pr SOS1-D ) and 2,745-bp (Pr SOS1-AB ) genomic DNA fragments were isolated from BAC-D and BAC-AB clones, respectively, sequenced and deposited into GenBank NCBI (Pr SOS1-D : accession no. KF169800, Pr SOS1-AB : accession no. KF169799).

During the inspection of the two promoter sequences, we found that 140 nucleotides at the 3′ end of the Pr SOS1-D promoter region exhibited similarity with the sequence upstream of the ATG of the TaSOS1 gene (accession no. AY326952), confirming that this cloned sequence was the upstream region of the wheat SOS1 allele, which is localised on genome D. Moreover, the Pr SOS1-AB promoter region was similar to Pr SOS1-D and contained five supplementary nucleotide sequences (Fig. 1a). Thus, this second SOS1 promoter region was the upstream region of the wheat SOS1 allele, which is localised on genomes A and/or B. This finding allowed us to determine the putative transcription start site (+1), which was located 140 and 133 bp upstream of the ATG codon of the Pr SOS1-D and Pr SOS1-AB promoters, respectively. Two potential TATA boxes (TAAATAA) were identified at −27 and −40 nucleotides upstream of the transcription start site of the Pr SOS1-D and Pr SOS1-AB promoters, respectively (Fig. 1b, c), which was consistent with the regular features of eukaryotic promoters (Ke et al. 1997).

Fig. 1
figure 1

a Schematic presentation of the five gaps in the Pr SOS1-D compared to the Pr SOS1-AB promoter regions. The numbers represent their lengths (positive numbers) and their positions (negative numbers). The ATG codon is indicated by the rectangle. Presentation of the putative transcription start site (designated as +1) and the putative TATA box in the Pr SOS1-D (b) and Pr SOS1-AB (c) promoter regions

Full size image

In silico analysis of the two upstream regions from the transcription start site of the SOS1 allele was performed using the plant promoter databases PLACE and PlantCARE to obtain additional indications about the transcriptional regulation of this gene. In silico analysis revealed similarity in the type but differences in the number and the position of the cis elements among the two wheat SOS1 promoter regions. Many regulatory cis elements were found and some of them are related to abiotic (dehydration and salt), biotic (fungal elicitor) and hormone (ABA) stress responses. Like Arabidopsis and Salicornia brachiata SOS1 promoters (Goyal et al. 2013; Ji et al. 2013), in silico analysis showed also the presence of several potential binding sites for transcription factors such as MYB, DOF and WRKY (Table 3).

Table 3 Some putative cis-acting elements present at 2,661 bp upstream of the ATG of bread wheat SOS1 alleles which are localised on genomes A and/or B (Pr SOS1-AB ) and on genome D (Pr SOS1-D ). Na and Nb: the number of cis-regulatory elements present in Pr SOS1-AB and Pr SOS1-D , respectively; (+): positive strand, (−): minus strand
Full size table

Generation of the transgenic Arabidopsis plants

In order to study the expression of the gusA gene under different abiotic stresses, the two T-DNA of the recombinant binary vectors pCAMBIA1391Z–Pr SOS1-D –gusA and pCAMBIA1391Z–Pr SOS1-AB –gusA were introduced separately by A. tumefaciens-mediated transformation in several independent transgenic Arabidopsis lines. These vectors contain the hptII gene conferring resistance to hygromycin as a selectable marker for plant transformation. After Agrobacterium-mediated transformation of Arabidopsis plants and selection with hygromycin, several transformants were produced. From each transformation event, three transgenic lines were propagated to the T3 generation, from which homozygous plants were isolated for further analyses. The two T-DNA regions of these two binary vectors are schematically presented in Fig. 2a.

Fig. 2
figure 2

Molecular analysis of the different transgenic Arabidopsis plants. a The T-DNA schematic maps of the two binary vectors pCAMBIA1391Z–Pr SOS1-AB –gusA and pCAMBIA1391Z–Pr SOS1-D –gusA used for Arabidopsis transformation. The two wheat SOS1 promoters were inserted separately in this binary vector at the restriction sites PstI and NcoI, upstream of the gusA gene. The HPTII marker is flanked by the CaMV35S promoter (P35S) and terminator region (T35S). GR, ABF and DF indicate the primers used for the detection of the positive Arabidopsis lines. b PCR analysis of the seven and nine T2 transgenic Arabidopsis lines carrying pCAMBIA1391Z–Pr SOS1-AB –gusA (AB lines) and pCAMBIA1391Z–Pr SOS1-D –gusA (D lines), respectively, and the non-transformed Arabidopsis plants (Wt). M: molecular weight marker, (−): negative control without DNA. c RT-PCR analysis of the four AB lines (AB6, 7, 8 and 9) and the four D lines (D1, 4, 7 and 9) using specific primers for the gusA gene. No amplification was detected in the case of the non-transformed plant (Wt). A 0.5-kb β-tubulin gene fragment was amplified by RT-PCR as an internal control

Full size image

T-DNA integration and gusA gene transcription were confirmed in the putative transgenic events by PCR and semi-quantitative RT-PCR, respectively. As expected, PCR products of 600- or 800-bp fragments were detected in the seven putative transgenic Arabidopsis lines carrying the Pr SOS1-AB –gusA construct and the nine putative transgenic plants carrying the Pr SOS1-D –gusA construct, respectively (Fig. 2b). The expression level of the gusA gene was analysed using RT-PCR, performed on young leaves of four AB lines (Arabidopsis carrying the Pr SOS1-AB –gusA construct), four D lines (Arabidopsis carrying the Pr SOS1-D –gusA construct) and non-transformed plants. As a control for cDNA amplification, the constitutively expressed β-tubulin gene was amplified. Semi-quantitative RT-PCR showed that the gusA transcript was expressed in these eight selected transgenic lines, and the expression level was almost similar for the AB and D lines, except the D4 line (Fig. 2c). Two representative transgenic lines AB6 and D9 were selected to further investigate the functional properties of Pr SOS1-AB and Pr SOS1-D , respectively, in T3 homozygous seedlings. Genetic segregation data performed on the selected lines using the hptII gene gave rise to a 3:1 ratio, confirming that this marker segregates as a single copy gene. GUS activity and gusA transcript accumulation were monitored by histochemical staining and RT-PCR, respectively.

Activity of Pr SOS1-AB and Pr SOS1-D promoters in the transgenic Arabidopsis plants

GUS activity was analysed in the transgenic Arabidopsis seedlings carrying the Pr SOS1-AB –gusA (AB6 line) or Pr SOS1-AB –gusA constructs (D9 line) grown in normal MS medium and at different developmental stages. At an early developmental stage (4 days old), clear GUS activity was detected only in the young leaves of these two lines. Histochemical staining of 8- and 12-day-old AB6 and D9 seedlings did not enable the detection of any GUS activity. However, the GUS activity was different between AB6 and D9 lines grown for 20 or 30 days in normal MS medium. Indeed, at these two seedling developmental stages, no GUS activity was observed in the D9 line. Concerning the AB6 line, GUS activity was detected slightly in the roots and leaves of 20-day-old seedlings. In mature plants (30 days old), blue staining was observed in leaves and not in the stems, roots or flowers (Fig. 3). Taken together, these results showed that the Pr SOS1-AB and Pr SOS1-D promoters are active in Arabidopsis plants; their activities were almost the same only at early developmental stages, showing them to be age-dependent and organ-specific promoters.

Fig. 3
figure 3

a GUS activity in transgenic Arabidopsis plants carrying the pCAMBIA1391Z–Pr SOS1-AB –gusA construct (AB6 line) or the pCAMBIA1391Z–Pr SOS1-D –gusA construct (D9 line) at different developmental stages (4, 8, 12, 20 and 30 days old). Wt: non-transformed plants, 35S: transgenic Arabidopsis plants carrying pCAMBIA1301 (positive control). b Binocular observation of GUS staining in flowers of transgenic Arabidopsis plants (AB6 and D9) and of the control plants (335S and Wt)

Full size image

Pr SOS1-AB is an abiotic stress-inducible promoter at different development stages

In a previous work, it has been demonstrated that the expression of the bread wheat TaSOS1 gene is induced by salt stress (Xu et al. 2008). Here, the activities of the two wheat SOS1 promoters were analysed in Arabidopsis under different abiotic stresses and at different developmental stages. For this, GUS activity was examined in the two transgenic AB6 and D9 lines grown for 8, 20 and 30 days in MS agar medium and then transferred to the same medium containing NaCl, mannitol or ABA for 2 days. At an early developmental stage (8 days old), the application of NaCl, mannitol or ABA produced a blue staining in the leaves of the two transgenic Arabidopsis AB6 and D9 lines. Contrary to the D9 line, GUS activity was observed also in the roots of the AB6 line challenged with NaCl, mannitol or ABA (Fig. 4a). Unlike the D9 line, abiotic and hormonal stresses produced a blue staining in a different part of 20-day-old seedlings of the AB6 line. Indeed, NaCl induced deeper blue staining in leaves and roots compared to ABA and mannitol treatment (Fig. 4b). Moreover, NaCl and ABA led to a strong GUS staining in the root tips of the AB6 line compared to mannitol treatment (Fig. 4c). In mature plants (30 days old), blue staining was observed in the stem and root of the AB6 line challenged with NaCl, mannitol or ABA stresses. In the case of the D9 line, no GUS activity was observed after the application of these different abiotic stresses (Fig. 4d). These data show that, contrary to Pr SOS1-D , the activity of Pr SOS1-AB is induced by different abiotic stresses independently of the vegetative stage of the transgenic Arabidopsis plants.

Fig. 4
figure 4

Histochemical GUS staining of (a) 8-, (b) 20- and (d) 30-day-old transgenic seedlings placed for 2 days in MS agar medium containing 100 mM NaCl, 100 mM mannitol or 20 μM ABA. The arrows indicate the blue colouration in 8-day-old seedling roots. c Binocular observation of GUS staining in the root tips of 20-day-old transgenic seedlings subjected or not to different stress conditions. AB6 and D9: Arabidopsis carrying the pCAMBIA1391Z–Pr SOS1-AB –gusA or pCAMBIA1391Z–Pr SOS1-D –gusA constructs, respectively

Full size image

Relation between stress treatment and gusA expression in transgenic Arabidopsis carrying the Pr SOS1-AB –gusA construct

To further ascertain the induced activity of the Pr SOS1-AB promoter in 20- and 30-day-old seedlings of the AB6 and AB7 lines, RT-PCR was used to detect the presence of gusA transcripts under abiotic (salt and osmotic) and hormonal stress (ABA) conditions. Compared to the non-treated plants, an accumulation of gusA transcripts was detected and found to be up-regulated by NaCl, ABA and mannitol treatments in whole 20-day-old seedlings and in the shoots and roots of 30-day-old AB6 seedlings (Fig. 5). These results are in complete agreement with those of the histochemical staining assays.

Fig. 5
figure 5

Analysis of gusA expression in the whole plants of 20-day-old seedlings (a) and in the shoots and roots of 30-day-old seedlings (b), challenged with various abiotic stresses [100 mM NaCl (Na), 100 mM mannitol (M) or 20 μM ABA (a)] for 2 days. NT: non-treated plants. 35S: positive control plant. Wt: non-transformed plant. AB6: the transgenic Arabidopsis plant carrying the pCAMBIA1391Z–Pr SOS1-AB –gusA construct. A 0.5-Kb β-tubulin gene fragment was amplified by RT-PCR as an internal control. (−): negative control without cDNA. The histograms correspond to the band densities in the gels, which are expressed in arbitrary units calculated by the Gel DocXR software. The standard errors were determined from three independent biologic replicates

Full size image

Discussion

The SOS1 protein has been associated to salt stress response, and loss of function of the SOS1 gene results in the hypersensitivity of Arabidopsis to NaCl (Wu et al. 1996). After the description of the first SOS1 gene in Arabidopsis (Shi et al. 2000), other SOS1 genes have been identified in different plants, such as rice, wheat and tomato (Martínez-Atienza et al. 2007; Olías et al. 2009; Xu et al. 2008; Feki et al. 2011). So far, only the expression pattern of Arabidopsis thaliana AtSOS1 and Salicornia brachiata SbSOS1 have been examined using the SOS1 promoter GUS (Shi et al. 2002; Goyal et al. 2013). In this work, we report the expression of Triticum aestivum TaSOS1 in Arabidopsis plants under various abiotic stresses (salt, drought and hormonal). It is worth noting that there is a single SOS1 locus in the genome of the tetraploid durum wheat (Triticum durum L. subsp. durum) (AA BB) (Feki et al. 2011). Thus, it seems that there are at least two SOS1 copies in the hexaploid bread wheat (Triticum aestivum) (AA BB DD), one localised on genomes A or B, and the other one on genome D. Here, we report the isolation of two novel promoter regions Pr SOS1-AB and Pr SOS1-D of the TaSOS1 gene by screening the genomic library of bread wheat (Triticum aestivum). The cloned Pr SOS1-AB and Pr SOS1-D sequences exhibited homology with the 5′ end sequence of SOS1 alleles, which are localised on genomes A or B, and on genome D, respectively. Thus, the isolated sequences were in the upstream region of the two wheat SOS1 alleles.

Sequence analysis of these two promoter regions revealed the presence of the same potential abiotic stress responsive cis elements, transcription factor-binding sites such as DOF and WRKY, ABA responsive element (ABRE) and GT elements (Table 3). DOF factors play an important role in the genes induced by plant hormones and stress signals (Yanagisawa and Sheen 1998). WRKY is required for positive and negative regulatory behaviours of ABA signalling (Zhang et al. 2004). An ABA-responsive cis-acting element named ABRE (C/GACGTGGC) is present in the promoter regions of many ABA-inducible genes. ABA plays a central role as a signalling molecule in stress regulatory networks in plants. Several trans- and cis-acting regulatory elements have been characterised that function in ABA-dependent and/or ABA-independent manners, leading to stress-inducible gene expression. An ABRE-like sequence (ACGTG) is required for the aetiolation-induced expression of the erd1 (early responsive to dehydration) gene in Arabidopsis (Simpson et al. 2003). In addition, ABRE is the most conserved of the dehydration-inducible promoters in Arabidopsis thaliana, rice and soybean, suggesting that the transcriptional regulation of dehydration-inducible genes is similar among these species, with an ABRE-dependent transcriptional pathway (Maruyama et al. 2012). The GT-1 motif plays a role in pathogen- and salt-induced expression of the SCaM-4 promoter (Park et al. 2004). In the same way, in silico analysis of the SOS1 upstream sequence from A. thaliana, Vitis vinifera and Oryza sativa revealed the presence of almost the same cis-regulatory elements (Goyal et al. 2013; Ji et al. 2013).

Many SOS1 genes were isolated and then characterised in mutant yeast (Shi et al. 2002; Martínez-Atienza et al. 2007; Olías et al. 2009; Oh et al. 2009; Feki et al. 2011). So far, only the regulatory mechanism of the Arabidopsis SOS1 gene has been analysed. In Arabidopsis, SOS1 mRNA is unstable under normal growth conditions, and its stability is substantially increased upon salt stress treatment (Shi et al. 2003). Chung et al. (2008) identified the cis element in SOS1 mRNA responsible for the stability-regulation and demonstrated that stress-induced SOS1 mRNA stability is mediated by ROS. As an initial step towards understanding regulatory mechanisms controlling wheat SOS1 gene expression, the expression pattern of the Pr SOS1-AB and Pr SOS1-D promoters was investigated using a gusA reporter gene system in transgenic Arabidopsis seedlings grown under normal or stressed conditions. In a previous work, it has been demonstrated that monocot gene promoters are functional in a dicots plant (Liu et al. 2003; Iwamoto et al. 2004; Tittarelli et al. 2007). In this study, histochemical staining revealed that the monocotyledonous Pr SOS1-AB and Pr SOS1-D promoters are active in this heterologous transgenic system. Moreover, their activities were similar only at early developmental stages (4, 8 and 12 days old). To our knowledge, this is the first isolation and characterisation of two wheat SOS1 promoters. Histochemical staining revealed the ability of these promoters to direct GUS expression with age-dependent and organ-specific patterns. Indeed, GUS activity was detected only in young leaves of transgenic Arabidopsis carrying the Pr SOS1-D -gusA construct (D lines). Despite the presence of gusA mRNA in D lines, GUS activity was never detected in roots and leaves. In an attempt to explain these results, one may speculate that GUS activity is undetectable because of the absence of GUS or an inactive form of the enzyme. Concerning the transgenic Arabidopsis carrying the Pr SOS1-AB -gusA construct (AB lines), GUS activity was absent only in 8- and 12-day-old seedlings.

The over-accumulation of stress regulators by using strong constitutive promoters has, in many cases, improved stress tolerance (Hsieh et al. 2002; Ito et al. 2006). Nevertheless, this enhanced stress tolerance is sometimes conferred at the expense of plant development and growth (Xu et al. 2006). The use of stress-inducible promoters is expected to be optimised for driving candidate abiotic stress tolerance genes (Rai et al. 2009; Zhu et al. 2010; Ben Saad et al. 2011). In this study, we showed that the two wheat TaSOS1 promoters are induced by salt, drought and hormonal stresses. This is consistent with the presence in these promoter regions of the different cis-regulatory elements related to various stresses (Table 3). Contrary to our result, it has been demonstrated that the AtSOS1 and SbSOS1 promoters are induced by salt stress and not by ABA or cold stress (Shi et al. 2000; Goyal et al. 2013). Our data showed that the induction by these different abiotic stresses was different between these two TaSOS1 promoters. Indeed, contrary to the Pr SOS1-D promoter, Pr SOS1-AB is induced by NaCl, ABA and mannitol at different developmental stages (8, 20 and 30 days old). This finding was supported by RT-PCR analysis of steady-state gusA mRNA levels in transgenic Arabidopsis, which revealed that the reporter gene transcription under the control of Pr SOS1-AB was highly induced by the different abiotic stresses. Moreover, GUS activity was significant at root tips under NaCl and ABA stresses. These data are in good agreement with the findings of Shi et al. (2002), which showed that Arabidopsis AtSOS1 is expressed in root epidermal cells, particularly at the root tip and in cells bordering the vascular tissue. Despite the presence of the same cis-regulatory elements in the two wheat SOS1 promoter regions, their activities were different under various abiotic stress conditions. This suggestion could be explained by the difference in the number of cis-regulatory elements within the two isolated promoter regions or by the presence of some cis-regulatory elements in the nucleotide sequences which are absent in the Pr SOS1-D promoter region.

This is the first report on the isolation and characterisation of wheat TaSOS1 promoters (Pr SOS1-AB and Pr SOS1-D ) from Triticum aestivum. Histochemical GUS transient analysis validated that Pr SOS1-AB and Pr SOS1-D are functional, age-dependent and organ-specific promoters. Interestingly, in this heterologous system, only the Pr SOS1-AB promoter is induced by NaCl, mannitol and ABA at different developmental stages. These results will lead to more interest in the Pr SOS1-AB promoter, because it could be an attractive candidate promoter for the development of transgenic crop plants. Moreover, Pr SOS1-AB may be an effective and desirable promoter for controlling stress tolerance candidate genes in terms of driving low constitutive transgene expression under normal conditions and high induction in response to salt (NaCl), hormonal (ABA) and osmotic (mannitol) stresses. For these reasons, using the Pr SOS1-AB promoter could avoid potential harmful effects related to an over-expression of the target gene under the control of constitutive promoters in transgenic plants.