Introduction

Drought is one of the most devastating abiotic stresses limiting plant growth and crop productivity (Boyer 1982; Valliyodan and Nguyen 2006), and the plant response to drought is consequently complex. Because many genes are involved in plant responses to drought, it is important to study the functions of these stress-related genes for understanding the molecular mechanisms of stress tolerance (Zheng et al. 2010).

Phospholipase D (PLD) is a major family of lipid-hydrolyzing enzymes. Increasing evidence suggests that PLD is activated in response to various stresses, such as high salinity (Darwish et al. 2009), water deficits (Mane et al. 2007), freezing (Welti et al. 2002; Li et al. 2004), wounding (Lee et al. 1997), ethylene and abscisic acid (ABA) (Fan et al. 1997). Several PLD genes have been isolated from various plant species, such as castor bean, Arabidopsis, rice and maize, and their molecular characterization has highlighted their role in early signaling events (Wang 2000). In Arabidopsis, there are 12 identified PLDs that are classified into six types, PLDα(3), -β(2), -γ(3), -δ(1), -ε(1) and ζ(2) (Qin and Wang 2002). Different AtPLDs have distinct functions, which explain in part the different effects of PLD in stress responses (Wang et al. 2006). For example, PLDδ promotes production under rapid dehydration and high salinity (Katagiri et al. 2001) and is also involved in freezing tolerance (Li et al. 2004), PLDζ is involved in the plant response to phosphate deficiency (Cruz-Ramirez et al. 2006), and PLDε plays a role in nitrogen signaling (Hong et al. 2009).

Previous studies have revealed that the PLDα genes in plants play important roles in various stresses. The Arabidopsis PLDα group has three members, PLDα1, -α2 and -α3. AtPLDα3 positively mediates plant responses to hyperosmotic stresses (Hong et al. 2008). Although AtPLDα2 appears to be a constitutive form of PLD, its functions in plant response to stresses have not yet been reported (Wang 2000). PLDα1 has been reported to promote ABA-mediated stomatal closure which decreases water loss in response to water deficit (Jacob et al. 1999; Sang et al. 2001; Zhang et al. 2004; Mishra et al. 2006; Hong et al. 2010). PLDα1 also plays a role in superoxide production in plants through the generation of PA as a lipid messenger (Sang et al. 2001; Welti et al. 2002). In addition, PLDα1 is involved in the regulation of the jasmonic acid (JA) signaling pathways (Wang 2000). Altogether, these results indicate that the PLDα1 gene has important and active roles in various stresses and is a key step in the ABA signaling pathway, underscoring the significance in cloning and studying the PLDα1.

Foxtail millet is native to China and is regarded as an elite drought-tolerant crop (Cheng and Liu 2003) Therefore, cloning and characterization of PLDα1 from foxtail millet may have potential benefits for genetic improvement of other cereal crops, such as rice, wheat and maize. In our previous work, we isolated a group of water stress up-regulated ESTs from foxtail millet seedlings using suppression subtractive hybridization and cDNA microarray (Zhang et al. 2007). Two of those ESTs are up-regulated by dehydration stress and aligned well with two regions of the same rice gene OsPLDα1. Here, we report the cloning, characteristics and biological functions of SiPLDα1 from foxtail millet in order to provide the molecular basis for potentially exploiting this gene in genetic engineering.

Materials and methods

Plant materials, stress treatments

Foxtail millet cv. Mar51 was selected as a typical drought variety according to our previous evaluation (Zhang et al. 2005). Seeds were surface-sterilized in 3% sodium hypochloride for 20 min and rinsed 15 times (1 min/each) in distilled water. After 3 days of germination, the seeds were planted in pots with a mixture of peat/forest and vermiculite (1:1 v/v). Plants were grown in the greenhouse (28°C day/20°C night, 14 h photoperiod, natural lighting, 70% relative humidity). When the seedlings were 5 weeks old, the stress treatments were performed as described by EI Maarouf et al. (1999), with minor modifications. In dehydration experiments, leaves were cut off and dehydrated on filter paper as described by Yamaguchi-Shinozaki and Shinozaki (1994) at ambient temperature (26°C) and under light for 0, 0.5, 1, 2, 4 and 6 h. The effect of ABA on the detached leaves was tested by plunging them into a 100 μM ABA solution for 0, 0.5, 1, 2, 4 and 6 h. For NaCl stress, the detached leaves were submerged into a 250 mM NaCl solution for 0, 0.5, 1, 2, 4 and 6 h. The treated leaves were sampled at the indicated times and stored at −80°C for RNA extraction.

Cloning of the Setaria italica PLDα1 full-length cDNA

In a previously established PEG-induced subtracted cDNA library, two ESTs sequences, which encode the same PLDα1 protein, RCRST0_005783 (near 5′ end, Accession No. EC612614) and CL185Contig1 (near 3′ end, Accession No. EC612075), were shown to be induced by PEG treatment (Zhang et al. 2007). Here, the RCRST0_005783 and CL185Contig1 sequences were used as templates for designing RACE amplification primers. Amplification of 5′ RACE and 3′ RACE was performed using the FirstChoice RLM-RACE Kit (Ambion). The 5′- and 3′-nested PCR primers are listed in Supplemental Table S1. The amplified RACE products were purified and cloned into the pGEM-T Easy Vector (Promega) and sequenced. After assembling the 5′ RACE and 3′ RACE fragments of SiPLDα1 cDNA, the ORF sequence was amplified with the specific primers, which are listed in Supplemental Table S1. The PCR products were cloned into the pGEM-T Easy Vector and sequenced. The complete sequence was deposited in GenBank under the accession number (GU480026).

Vector construction and Arabidopsis transformation

The full-length cDNA of the SiPLDα1 gene was amplified using the SiPLDα1 5′ forward primer 5′-GGATCCTCTAGACTCCCGATCCTCAACAGCCTCT-3′ with a BamHI restriction site and the SiPLDα1 3′ reverse primer 5′-GTGAGCTCCAGTCTGCCCTCTATGTGGTGA-3′ with a SacI restriction site. The SiPLDα1 gene and the 35S promoter were subcloned into the plant expression vector pBI121 using the BamHI and SacI restriction sites. The resultant construct, named pBISiPLDα1, contains nptII as the selectable marker, which can be used for screening transgenic plants by exploiting its resistance to kanamycin. These constructs were mobilized into Agrobacterium tumefaciens GV3101 and then introduced into wild-type Arabidopsis ecotype (Col-0) by floral dip (Clough and Bent 1998). Homozygous transgenic lines were selected in the T3 generation and used for further assays.

Germination study and stress treatment

Arabidopsis thaliana, ecotype Col-0, was used in this study as the transformation receptor. Arabidopsis seeds were sterilized and kept for 3 days at 4°C in the dark to break dormancy. After 1 week, the seedlings were transferred to a pot filled with a mixture of peat/forest soil and vermiculite (1:1 v/v) in a greenhouse at 22°C, with a light intensity of 50 μmol m–2 s–1 and 70% RH under a 16 h-light photoperiod.

For germination analysis, approximately 30 seeds each from the wild-type and transgenic lines were surface-sterilized and plated on MS medium supplemented with different concentrations of ABA, NaCl, or mannitol, incubated at 4°C for 3 days and then transferred to 22°C under long-day light conditions. Germination (emergence of radicles) and green cotyledons were scored at the third and seventh day, respectively. All experiments were performed in triplicate, and the SD is shown as an error bar. For dehydration stress, the water solution was supplemented with mannitol to a final concentration of 350 mM, and the plants were placed in a container with a mannitol solution for 0, 2 and 4 h to simulate dehydration stress. For drought-stress treatment, 3-week-old plants were transferred to a growth chamber at 22°C, with no further addition of water for 12 days. Samples were collected and stored at −80°C for RNA extraction. The biomass of transgenic and wild-type plants was estimated by measuring fresh weights of plants.

Drought resistance evaluation

For drought resistance evaluation, 30 seedlings each of transgenic and wild-type plants were grown in pots (10-cm diameter) with normal watering every 3 days and a commercial fertilizer every 2 weeks before water was withheld. After 2 weeks, the plants were divided into two groups for stress treatments. One group was subjected to drought stress by withholding water, and a control group was watered normally. After 2 weeks of drought treatment, all of the pots were rewatered simultaneously, and the survival rate was scored 7 days later. Plants were considered dead if there was no growth 7 days after watering.

Relative water content measurements

The relative water content (RWC) was essentially determined as reported by Gaxiola et al. (2001). The RWC values of transgenic and WT plants were determined during a 10-day drought-stress experiment. Leaves of drought- and non-stressed transgenic and WT seedlings were carefully excised, and their fresh weight was immediately scored. After being floated on deionized water at 4°C for overnight, their rehydrated weight was determined. Finally, they were dried in an oven at 80°C overnight and weighed. The RWC was calculated as RWC = (fresh weight − dry weight)/(rehydrated weight − dry weight).

Measurement of ion leakage

To measure the ion leakage, leaves were detached and rinsed with distilled water and then immersed in 15 ml of distilled water in glass tubes. After degassing under vacuum for 30 min to remove air bubbles on the leaf surface, samples were incubated with gentle agitation for 3 h (Fan et al. 1997). The initial conductivity was measured with a conductivity meter, and then the samples were boiled in a water bath for 20 min. The total conductivity was measured again after cooling to room temperature. The ion leakage is expressed as a percentage of the initial conductivity over the total conductivity.

RNA isolation, semiquantitative RT-PCR and quantitative real-time PCR

Total RNA was extracted by the Trizol reagent (Invitrogen, Carlsbad, CA, USA) from 100 mg of young leaves of foxtail millet and Arabidopsis seedlings, and then treated with RNase-free DNase to remove contaminating DNA. For the RT-PCR reaction, 1 μg of DNase-treated RNA samples from each sample was used for the reverse transcription reaction. Subsequently, the first-strand cDNA (1 μl) was used as templates for PCR amplification. The gene-specific primers for semiquantitative RT-PCR were 5783 3′ RACE primer 5′-CATGATGACTTCCACCAGCCA-3′ and CL185 5′ RACE primer 5′-CAGGGAAGCCAAAAGCAGCACCA-3′. Real-time PCR analysis was performed and statistically analyzed as described by Livak and Schmittgen (2001). Triplicate quantitative assays were performed on 1 μl of each cDNA dilution with SYBR Green Master Mix and an ABI 7300 sequence detection system. The gene-specific primers for SiPLDα1 in the foxtail millet seedlings were 5′-CATGATGACTTCCACCAGCCA-3′ and 5′-CAGGGAAGCCAAAAGCAGCACCA-3′. The gene-specific primers for the real-time RT-PCR in Arabidopsis plants are listed in Supplemental Table S2. Actin from foxtail millet and His2A from Arabidopsis were used as internal controls to quantify the relative transcript levels of each target gene in our assays.

Results

Isolation and characterization of full-length cDNA of SiPLDα1

From the subtracted foxtail millet cDNA libraries, as described above, two ESTs (RCRST0_005783 and CL185Contig1) were isolated as differentially expressed sequences in response to PEG stress (Zhang et al. 2007). According to the BLAST results, RCRST0_005783 and CL185Contig1 align to two regions of the same rice gene OsPLDα1, leading us to speculate that the two fragments may belong to the same cDNA of foxtail millet. The two regions were separated by a 4-bp nucleotide sequence. Using the specific 5783 3′ RACE primer 5′-CATGATGACTTCCACCAGCCA-3′ and CL185 5′ RACE primer 5′-CAGGGAAGCCAAAAGCAGCACCA-3′, a 292-bp fragment was amplified by RT-PCR, which covered partial sequences from the two ESTs and the gap sequence between them. Then, 5′ RACE and 3′ RACE were carried out to obtain the 5′ and 3′ end sequences of the PLDα1 cDNA. Finally, the open reading frame of this gene was amplified by RT-PCR. Since this gene was the first member of PLDα family from Setaria italica, we named it as SiPLDα1. The SiPLDα1 gene (Accession No: GU480026) had a 2,938-bp full-length cDNA with an open reading frame of 2,436 nucleotides, which encodes 811 amino acids with a predicted molecular mass of 92.0 kDa and a pI of 5.41 (Fig. 1). A putative conserved domain detection tool (http://www.ebi.ac.uk/Inter-ProScan/) indicated that SiPLDα1 includes a C2 domain (Calcium/lipid-binding domain) and two conserved HXKXXXXD domain (X represents any amino acid). The SiPLDα1 has 92, 90, 89 and 78% identities with ZmPLDα, OsPLDα1, TaPLDα and AtPLDα1, respectively (Fig. 1).

Fig. 1
figure 1

Alignment of amino acid sequences of SiPLDα1 with PLDs from other plant species. The predicted SiPLDα1 protein sequence was compared with PLD sequences from rice (OsPLD), maize (ZmPLD), wheat (TaPLD) and Arabidopsis (AtPLD). Alignment of the amino acid sequences was performed with the CLUSTALW(1.83) program on the EBI server. Species abbreviations are: Si, Setaria italica; At, Arabidopsis thaliana; Os, Oryza sativa; Zm, Zea mays; Ta, Triticum aestivum

Full size image

Phylogenetic analysis of SiPLDα1 gene

To analyze the closeness of SiPLDα1 gene with related PLD genes, a phylogenetic tree was constructed based on their amino acid sequences. Based on this phylogram, the SiPLDα1 gene is most similar to the ZmPLDα1 gene among all of the plant PLDs. In the 12 Arabidopsis PLD genes, SiPLDα1 is more similar to AtPLDα1 than the others (Fig. 2).

Fig. 2
figure 2

Phylogenetic tree analysis of the predicted relationships between the SiPLDα1 protein of foxtail millet and those of other plant species. The full-length amino acid sequences of each protein were aligned using ClustalW. Fifteen rice PLD genes (OsPLD), five tomato genes (LePLD), one wheat PLD (TaPLD), one maize PLD (ZmPLD) gene and two resurrection plant PLD (CpPLD) genes were compared with SiPLDα1. Si, Setaria italica; At, Arabidopsis thaliana; Le, Lycopersicon esculentum; Os, Oryza sativa; Zm, Zea mays; Ta, Triticum aestivum; Cp, Craterostigma plantagineum

Full size image

Expression of the SiPLDα1 in response to various abiotic stresses

Real-time quantitative RT-PCR was used to analyze the expression pattern of SiPLDα1 under various stress treatments. The results showed that SiPLDα1 expression in shoots is strongly up-regulated by ABA, NaCl and dehydration stresses. As shown in Fig. 3a, the transcript of SiPLDα1 began to increase after 0.5 h of drought treatment and gradually accumulated up to the highest level after 6 h of treatment. In the case of ABA treatment, expression of the SiPLDα1 gene was induced at 1 h of ABA treatment. The accumulation of the SiPLDα1 transcripts increased up to 6 h, but a lower expression level at 4 h was observed in comparison to that after 2 h of treatment (Fig. 3b). Under NaCl treatment, the SiPLDα1 expression level also continuously increased with the stress time, except there was less induction after the 2 h treatment compared to that after 1 h (Fig. 3c). Furthermore, expression of the SiPLDα1 gene under the dehydration treatment was induced more sharply than under the ABA and NaCl treatments (Fig. 3). The SiPLDα1 transcription level was very low in normal growth conditions, suggesting that the SiPLDα1 gene plays an active role in response to drought and osmotic stress in the shoots of foxtail millet.

Fig. 3
figure 3

Real-time RT-PCR analysis of SiPLDα1 gene expression in foxtail millet shoots treated with ABA, NaCl and dehydration stress. a SiPLDα1 expression profiles in foxtail millet shoots treated with dehydration stress. b SiPLDα1 expression profiles of foxtail millet shoots treated by ABA. c SiPLDα1 expression profiles of foxtail millet shoots treated by NaCl. Actin was used as an internal control. Data represent mean ± SD (n = 3 experiments)

Full size image

Molecular characterization of the SiPLDα1-transgenic Arabidopsis lines

The full-length ORF of the foxtail millet SiPLDα1 was cloned into the pBI121 vector under control of the 35S promoter (Fig. 4a). The SiPLDα1 gene was transformed into wild-type Arabidopsis thaliana by the floral dip method. More than ten 35S::SiPLDα1 transgenic plantlets were selected on culture medium with 50 μg/ml of kanamycin. Six independent T3 homozygous lines showed high expression of SiPLDα1 by semi-quantitative RT-PCR of young leaves (Fig. 4b). A quantitative real-time PCR analysis was carried out to determine the expression level of the SiPLDα1-transgenic lines. The transgenic events L2 and L32, respectively, representing low and high levels of SiPLDα1 gene expression, were selected for further stress-tolerance tests (Fig. 4c). Genetic analyses of both L2 and L32 also showed that SiPLDα1 was co-segregated with the kanamycin resistance in a 3:1 ratio, indicating T-DNA insertion in the two lines at a single locus.

Fig. 4
figure 4

Expression analysis of SiPLDα1 in the transgenic and WT plants. a Schematic diagrams of the pBISiPLDα1 construct. The SiPLDα1 gene was inserted between the 35S CaMV promoter and the nopaline synthase promoter terminator (NOS-T), and the nptII gene was flanked by the nopaline synthase promoter (NOS-P) and terminator (NOS-T). b Analysis by semi-quantitative RT-PCR of transgenic SiPLDα1 lines. Specific 292-bp PCR products were detected in six transgenic lines, 1, L23; 2, L27; 3, L33; 4, L32; 5,L26; 6, L2; 7,WT. His2A was amplified by RT-PCR as an internal control. c Analysis by real-time PCR of transgenic SiPLDα1 lines. Relative expression levels of SiPLDa1 in six transgenic lines were calculated according to 2−∆CT methods based on Arabidopsis His2A as internal controls

Full size image

Overexpression of SiPLDα1 enhances the sensitivity of Arabidopsis to ABA and osmotic stress

To investigate whether SiPLDα1 overexpression affects the ABA response, we conducted analyses of germination and early seedling growth (i.e., radicle emergence) of 35S::SiPLDα1 and wild-type Arabidopsis plants. There was no difference in seed germination between the wild-type and transgenic plants under ABA treatment (data not shown). However, altered ABA sensitivity was observed during post-germination growth. In the presence of exogenous ABA, the seedling growth of both wild-type and 35S::SiPLDα1 were inhibited significantly, but transgenic seeds were inhibited to a greater extent (Fig. 5). For example, on 0.5 μM ABA MS medium, the 35S::SiPLDα1 seeds revealed only 13% green cotyledon rates, whereas the wild-type seeds displayed approximately 30% with green cotyledons, demonstrating that the growth of the aerial parts of 35S::SiPLDα1 seedlings was more sensitive to ABA.

Fig. 5
figure 5

Early seedling growth of transgenic and WT seeds in response to ABA, NaCl and mannitol. Seeds were germinated on MS agar plates supplemented with 0.5 μM ABA, 100 mM NaCl and 300 mM mannitol, and the green cotyledons were scored 7 days after germination. Bars indicate SD (n = 3). * and ** indicate significant and highly significant differences from the wild type at P < 0.05 and P < 0.01, respectively, by Student t tests

Full size image

To investigate the response to osmotic stress, germination and early growth in media supplemented with 100 mM NaCl or 300 mM mannitol were performed for 35S::SiPLDα1 transgenic and wild-type lines. No significant differences in germination rates were observed between the wild-type and 35S::SiPLDα1 seeds grown on MS medium with 100 mM NaCl and 300 mM mannitol. In contrast, the post-germination seedling growth of wild-type and transgenic lines was affected significantly by 100 mM NaCl and 300 mM mannitol (Fig. 5). Transgenic lines showed severe inhibition of early growth by osmotic stress than the wild-type seeds. As shown in Fig. 5, the numbers of green cotyledon 35S::SiPLDα1 seeds were significantly lower than those of wild type (Fig. 5).

Overexpression of SiPLDα1 improves the drought tolerance of Arabidopsis

Under normal growth conditions, the transgenic plants did not show any obvious morphological or developmental abnormalities (Fig. 6a). However, when subjected to drought stress for 2 weeks, as shown in Fig. 6, the wild-type plants showed severe wilting and chlorosis on the rosette leaves, whereas the transgenic plants were turgid and their leaves remained green; at the same time, the transgenic lines had higher biomasses than the wild type under drought stress (Fig. 6b). The leaves of transgenic lines also had higher RWC and lower relative electrolyte leakage under drought stress (Fig. 6c, d). These results indicated that overexpression of SiPLDα1 in Arabidopsis improved their drought tolerance. Moreover, 14-day-old plants grown on soil were deprived of water for 2 weeks. After re-watered, the transgenic plants showed a stronger growth recovery and higher survival rate than the wild-type plants (Table 1).

Fig. 6
figure 6

Drought stress tolerance of 35S::SiPLDα1 transgenic plants. a Drought-tolerant test after 3-week-old plants was stopped watering for 12 days. b Biomass comparison of wild-type and transgenic plants under drought stress. c Relative water content of wild-type and transgenic plants under water-deficit stress. Each of the ten wild-type and transgenic plants were stressed for 12 days. One leaf per plant (three plants of each group) was removed from the seedlings for analysis. d Relative electrolyte leakage of SiPLDα1 transgenic plants and wild type in response to drought stress. The leaves were harvested for measurement. Bars indicate SD (n = 3); * and ** indicate significant and highly significant differences from the wild type at P < 0.05 and P < 0.01, respectively, by Student t tests

Full size image
Table 1 Survival rates of transgenic and WT plants under severe drought stress
Full size table

Overexpression of SiPLDα1 affects the expression of drought stress-related genes

To elucidate the molecular mechanism of drought tolerance, quantitative RT-PCR experiments were conducted to compare the transcript levels of nine drought stress-related genes between WT and 35S::SiPLDα1 lines using gene-specific primers (Supplemental Table S2). The results revealed that the ABA and dehydration stress-inducible genes (i.e., RD29A, RD29B, RD22 and RAB18) showed higher expression levels in both the wild type and transgenic plants under dehydration, but the expression levels of these genes in transgenic seedlings were significantly higher than in WT (Table 2). For example, under dehydration, the relative expression level of RD29B drastically increased, reaching 237.22 folds in transgenic seedlings, but only 124.82 folds in wild type after 2 h of dehydration. However, the expression levels of P5CS1, which is the rate-limiting enzyme in the biosynthesis of Proline, showed no significant difference between WT and transgenic plants. Except for RD29B, we observed a consistent trend that the expression levels of transgenic plants were significantly lower than WT under normal conditions.

Table 2 Relative expression levels of stress-responsive genes in response to dehydration treatment
Full size table

Discussion

Overexpression of SiPLDα1 in Arabidopsis increased the drought tolerance

Previous studies have revealed that the PLDα genes in plants play an important role in various stresses. Here we reported the cloning of the SiPLDα1 full-length cDNA from foxtail millet and analysis of its expression pattern in various stresses. The results showed that the expression of SiPLDα1 is significantly induced by dehydration, NaCl and ABA treatments, which is consistent with previous reports in other species (Ryu and Wang 1995; EI Maarouf et al. 1999; Darwish et al. 2009). The results also showed that the overexpression of SiPLDα1 in Arabidopsis significantly increased the drought tolerance of transgenic plants, which was revealed by less chlorosis and growth inhibition, higher biomass and RWC in transgenic plants than in the wild type.

Generally, increased drought tolerance accompanies hypersensitivity to ABA during seed germination and early seedling growth (Hu et al. 2006; Ko et al. 2006; Dai et al. 2007). The plant hormone ABA is a principal stress hormone and is involved in many aspects of plant responses to abiotic stresses, such as drought and salt stress (Yamaguchi-Shinozaki and Shinozaki 1994). Several reports have shown that osmotic stress delays seed germination and arrests early seedling development, primarily through ABA action. Usually, the ABA level is increased under osmotic stress, and transgenic lines are generally more sensitive to ABA than the wild type (Saez et al. 2004, 2006). In this study, we observed the hypersensitive phenotypes of SiPLDα-transgenic plants conferred under ABA, NaCl and mannitol treatment (Fig. 4). Therefore, we suggest that the increased drought tolerance in 35S::SiPLDα1 transgenic plants is carried out by their high sensitivity to the ABA signal.

To elucidate the molecular mechanism of SiPLDα1 in the drought response, we analyzed the expression levels of nine drought stress-related genes involved either in stress tolerance or in ABA biosynthesis and signaling processes. It is well known that RD29A, RD29B, RD22 and RAB18 are induced by dehydration- and ABA-responsive genes, and that the higher expression of stress-responsive genes improves plant tolerance under stress (Shinozaki and Yamaguchi-Shinozaki 1997; Zhu 2002). In our study, these genes showed up-regulation in both wild-type and transgenic plants under dehydration, but their expression levels in 35S::SiPLDα1 seedlings are significantly higher than those in wild type (Table 2). The transcript levels of ABA biosynthesis and signaling genes (NCED3 and ABI1) were higher in 35S::SiPLDα1 seedlings than in wild type after drought treatment, which may suggest that SiPLDα1 is an upstream gene regulating the ABA biosynthesis and signaling transduction. Stress-responsive genes, including the alcohol dehydrogenase gene ADH1 and KIN2, were found up-regulated by ABA and drought stress, and their transcript levels were also higher in the 35S::SiPLDα1 transgenic seedlings than in the wild-type plants under drought stress. P5CS1 is the rate-limiting enzyme in the biosynthesis of proline and it is up-regulated by drought stress (Hu et al. 1992; Yoshiba et al. 1999), but the expression of P5CS1 in our transgenic seedlings did not display obvious change. Interestingly, the nine drought stress-related genes except for RD29B were down-regulated in transgenic lines to various degrees under normal conditions, suggesting that the expression levels of these genes are partially repressed by SiPLDα1, similar to previously reported results for ABF2 (Uno et al. 2000; Kim et al. 2004). However, the mechanism of this is still unknown.

The application of SiPLDα1 in crop improvement for drought tolerance

In this study, we transferred the SiPLDα1 from foxtail millet into Arabidopsis and certified that the SiPLDα1 confer drought tolerance in Arabidopsis plants, a feature that can be attributed to the faster/stronger activation of signaling pathways downstream of SiPLDα1. Unlike other transcription factors, such as DREB factors (Kasuga et al. 1999) and ABFs (Kim et al. 2004), constitutive overexpression of SiPLDα1 had no obvious adverse effects on the growth and development of transgenic plants. These results indicate that SiPLDα1 has great potential in the genetic improvement of plants for drought tolerance. It will be important to investigate the overexpression of the PLDα1 gene in other plants, such as maize, wheat and rice, with the aim of improving crop tolerance to multiple environmental stresses. A stress-inducible promoter may be necessary to drive SiPLDα1 overexpression, because constitutive expression of SiPLDα1 is likely to compromise the tolerance of crop plants (Agarwal et al. 2006).