• Open Access

Expression of an Arabidopsis molybdenum cofactor sulphurase gene in soybean enhances drought tolerance and increases yield under field conditions

Authors

  • Yajun Li,

    1. State Key Laboratory of Plant Physiology and Biochemistry, College of Agronomy and Biotechnology, China Agricultural University, Beijing, China
    Search for more papers by this author
    • These two authors contributed equally to this work.
  • Jiachang Zhang,

    1. State Key Laboratory of Plant Physiology and Biochemistry, College of Agronomy and Biotechnology, China Agricultural University, Beijing, China
    Search for more papers by this author
    • These two authors contributed equally to this work.
  • Juan Zhang,

    1. State Key Laboratory of Plant Physiology and Biochemistry, College of Agronomy and Biotechnology, China Agricultural University, Beijing, China
    Search for more papers by this author
  • Ling Hao,

    1. State Key Laboratory of Plant Physiology and Biochemistry, College of Agronomy and Biotechnology, China Agricultural University, Beijing, China
    Search for more papers by this author
  • Jinping Hua,

    1. State Key Laboratory of Plant Physiology and Biochemistry, College of Agronomy and Biotechnology, China Agricultural University, Beijing, China
    Search for more papers by this author
  • Liusheng Duan,

    1. State Key Laboratory of Plant Physiology and Biochemistry, College of Agronomy and Biotechnology, China Agricultural University, Beijing, China
    Search for more papers by this author
  • Mingcai Zhang,

    Corresponding author
    • State Key Laboratory of Plant Physiology and Biochemistry, College of Agronomy and Biotechnology, China Agricultural University, Beijing, China
    Search for more papers by this author
  • Zhaohu Li

    1. State Key Laboratory of Plant Physiology and Biochemistry, College of Agronomy and Biotechnology, China Agricultural University, Beijing, China
    Search for more papers by this author

Correspondence (Tel (+8610) 6273 3049; fax (+8610) 6273 2567; email zmc1214@163.com)

Summary

LOS5/ABA3 gene encoding molybdenum cofactor sulphurase is involved in aldehyde oxidase (AO) activity in Arabidopsis, which indirectly regulates ABA biosynthesis and increased stress tolerance. Here, we used a constitutive super promoter to drive LOS5/ABA3 overexpression in soybean (Glycine max L.) to enhance drought tolerance in growth chamber and field conditions. Expression of LOS5/ABA3 was up-regulated by drought stress, which led to increasing AO activity and then a notable increase in ABA accumulation. Transgenic soybean under drought stress had reduced water loss by decreased stomatal aperture size and transpiration rate, which alleviated leaf wilting and maintained higher relative water content. Exposed to drought stress, transgenic soybean exhibited reduced cell membrane damage by reducing electrolyte leakage and production of malondialdehyde and promoting proline accumulation and antioxidant enzyme activities. Also, overexpression of LOS5/ABA3 enhanced expression of stress-up-regulated genes. Furthermore, the seed yield of transgenic plants is at least 21% higher than that of wide-type plants under drought stress conditions in the field. These data suggest that overexpression of LOS5/ABA3 could improve drought tolerance in transgenic soybean via enhanced ABA accumulation, which could activate expression of stress-up-regulated genes and cause a series of physiological and biochemical resistant responses.

Introduction

Soybean (Glycine max L.) is the major economic oilseed crop worldwide and provides the largest source of vegetable oil and protein for human and animal. Soybean is subjected to many environmental stresses, such as drought, salt and cold, which adversely affect their growth and development depending on the plant stage at onset, duration and intensity of stress (Cutforth et al., 2007). Drought is recognized as the most devastating environmental stresses, which can inhibit soybean growth and reduce about 40% yield (Manavalan et al., 2007). Moreover, soybean is sensitive to drought stress, especially during critical periods of plant development (Liu et al., 2003). As water availability for crop production becomes more limited in the world, the development of drought-resistant crops becomes of great importance. Also, it is essential that improvement in drought tolerance in soybean would be the most promising and effective approach to improve soybean productivity during water shortage.

Plants adapt themselves to the changing environment through a series of molecular responses. The physiological and metabolic basis for these molecular responses involves perception of stress signals, generation of secondary messengers and subsequent signal transduction (Valliyodan and Nguyen, 2006). Previous studies have revealed multiple signal transduction pathways, such as protein kinase signalling, calcium signalling, ABA signalling and sugar signalling, that occupy a central place in these transduction networks (Yamaguchi-Shinozaki and Shinozaki, 2006). Among these signal transduction pathways, ABA regulates expression of stress-induced genes and activates signal transduction pathways that are essential for plant survival and productivity (McCourt and Creelman, 2008).

ABA is well known to be produced in the vegetative tissues under drought stress and to be an essential messenger in the plant response to drought (Wilkinson and Davies, 2002). ABA plays a crucial role in modulating plant water status through regulating stomatal closure and inducing expression of genes that encode enzymes and other proteins involved in cellular dehydration tolerance (Zhu, 2002). Signal of drought stress is associated with a reduction in stomatal conductance in different species (Earl, 2003). Regulation of ABA biosynthesis by genetic engineering has increased plant drought tolerance (Thompson et al., 2007; Xiao et al., 2009). For example, overexpression of 9-cis-epoxycarotenoid dioxygenase (NCED) gene in tobacco and tomato resulted in increasing ABA accumulation and water-use efficiency (Qin and Zeevaart, 2002; Thompson et al., 2007). Development of specific transgenic crops has improved drought tolerance and increased yield under drought (Nakashima et al., 2007), but there has been few report in soybean to date (Manavalan et al., 2007). Overexpression of the Arabidopsis P5CR gene in soybean enhanced drought tolerance by increasing free proline and relative water content (RWC) and reducing reactive oxygen species (ROS; Kocsy et al., 2005). Accordingly, control of stomatal conductance by ABA accumulation under drought is a promising physiological trait for developing drought resistance in soybean (Manavalan et al., 2007).

Aldehyde oxidase (AO), which converts abscisic aldehyde to ABA in the last step of ABA biosynthesis pathway, needs the sulphurylated form of a molybdenum cofactor (MoCo) for its activity (Bittner et al., 2001; Seo et al., 2004). MoCo sulphurase, identified in genetic screens from Arabidopsis, is correlated with reduced ABA levels, and LOS5/ABA3 gene encodes MoCo sulphurase involved in regulation of ABA biosynthesis (Bittner et al., 2001; Xiong et al., 2001). Expression of Arabidopsis LOS5/ABA3 gene in rice under drought stress has increased drought tolerance and yield (Xiao et al., 2009).

In most cases, a transgene driven by a constitutive promoter expresses constitutively with no change under stress condition. However, some exceptions have been reported that the expression of stress response transgene is up-regulated by abiotic stress although transcription of the transgene is controlled by a constitutive promoter. For example, the SOS1 transcript driven by a constitutive promoter was markedly up-regulated by salt stress and increased salt tolerance in the transgenic lines (Shi et al., 2003). Also, this phenomenon was observed in rice expressing SaVHAC1 gene (Baisakh et al., 2012) and maize expressing betA gene (Wei et al., 2011) under salt or drought stress. Recently, we demonstrated that constitutive overexpression of LOS5/ABA3 in tobacco and cotton increased drought tolerance, and the expression level of LOS5/ABA3 was up-regulated by drought stress under greenhouse conditions (Yue et al., 2011, 2012). These suggested that a stress response transgene under constitutive promoter in homologous or heterologous is capable of being induced by environmental stresses and increases the tolerance. Post-transcriptional control of transcript accumulation is an important mechanism for gene regulation under stress and has been seen previously for some stress-regulated genes (Cohen et al., 1999; Shi et al., 2003). However, little is known about the effect of overexpression of LOS5/ABA3 on soybean, particularly its effect under drought stress, although its role in drought stress response has been indirectly shown in los5 mutant (Xiong et al., 2001).

In the present study, to develop novel drought-tolerant soybean and to explore the mechanism whereby the LOS5/ABA3 gene improved drought tolerance, we used the constitutive overexpression approach in soybean to provide evidence of the biological role LOS5/ABA3 gene against dehydration and drought stress under growth chamber and field conditions. Expression level of LOS5/ABA3 in transgenic plants was significantly induced by drought stress, which enhanced the activities of AO, and led to the accumulation of ABA. LOS-5/ABA3-overexpressing plants exhibited high drought tolerance, which displayed lower water loss and membrane damage. Also, the LOS5/ABA3 overexpression modulates the expression of stress-up-regulated genes in transgenic plants under drought stress. Moreover, LOS5-/ABA3-overexpressing plants exhibited more biomass and seed yield under drought stresses in the field. These suggested that expression of LOS5/ABA3 in soybean functions as regulator of AO involved in ABA biosynthesis and played an important role in improving soybean drought tolerance.

Results

LOS5/ABA3 overexpression enhances drought tolerance

We generated transgenic soybean lines that overexpressed LOS5/ABA3 to explore whether overexpression increased drought tolerance. The pCAMBIA1300 -LOS5/ABA3 vector (Figure 1a) was introduced into soybean via Agrobacterium-mediated transformation. The transformations were confirmed by PCR analysis (Figure 1b), and a total of 57 independent soybean transgenic lines were generated. All independent transgenic lines were exposed to drought stress, and ten independent representative lines exhibited relatively high drought tolerance compared with WT based on leaf water status and biomass (data no shown). Then, Southern blotting analysis revealed that four T3 transgenic lines (S-1, 2, 3 and 4) appeared to have one copy of the transgene (Figure 1c). Homozygous transgenic plants for two lines, S-2 and S-3, which had a single-copy transgene integration and similar fresh weight of shoot and root, height and total leaf area compared with WT under well-watered conditions (Table 1), were selected for drought tolerance analysis.

Table 1. Morphological characteristics in overexpression LOS5/ABA3 transgenic soybean
LinesFresh weight (g per plant)Height (cm)Total leaf area (cm2)
ShootRoot
  1. 21-day-old seedlings of T4 transgenic and WT soybean were used to analyse the morphological characteristics under well-watered conditions. The data point was the mean of two independent biological experiments, and each experiment comprises five samples. Error bars indicated denote the standard deviation values.

WT5.253 ± 0.7823.274 ± 0.68515.9 ± 1.962.5 ± 4.9
S-25.391 ± 0.1643.172 ± 0.44116.2 ± 1.563.9 ± 5.6
S-35.284 ± 0.2813.256 ± 0.16215.8 ± 2.662.7 ± 5.8
Figure 1.

Schematic structure of the T-DNA and molecular analysis of LOS5/ABA3-expressing soybean. (a) T-DNA region of the vector pCAMBIA1300-LOS5/ABA3. LB, left T-DNA border; RB, right T-DNA border; pSuper, ‘Super Promoter’; Hygromycin, hygromycin phosphotransferase II gene; CaMV, cauliflower mosaic virus. (b) Four independent T3 transgenic lines were confirmed by PCR using LOS5-/ABA3-specific primers. (c) Southern blot analysis of transgene copy number in T3 transgenic soybean and WT. Genomic DNA was digested with Eco RI. The Hyg probe was used for hybridization. P, pCAMBIA1300 as positive control. (d) The relative expression level of LOS5/ABA3 under well-watered and drought stress conditions. RNA was extracted from the first trifoliolate leaves of transgenic and WT soybean. WW, well-watered; D1, moderate drought stress (60% normal water supply) for 5 days; D2, severe drought stress (40% normal water supply) for 5 days; RW, rewatered for 2 days. Values were the mean ± SD (n = 6).

To investigate the role of overexpressed LOS5/ABA3 in transgenic soybean under drought stresses, real-time PCR analysis was used to monitor its LOS5/ABA3 transcript level in leaves under different drought stresses. Under well-watered condition, the LOS5/ABA3 gene was transcribed in leaves of transgenic plants but not in WT plants. However, drought stresses enhanced LOS5/ABA3 expression levels (Figure 1d).

LOS5/ABA3 overexpression promotes ABA accumulation to drought stress

The LOS5/ABA3 gene encoding MoCo sulphurase was involved in regulation of AO, which regulated the last step of ABA biosynthesis, so a native protein gel assay was used to analyse the AO activity of leaf extracts from transgenic (S-2 and S-3) and WT soybean (Figure 2a). Under D1 treatment, transgenic lines S-2 and S-3 exhibited 106% and 61% higher AO activity than that of WT, respectively. Also, in D2 treatment, S-2 and S-3 soybean had 44% and 28% higher AO activity than that of WT soybean, respectively. However, there was no significant difference in AO activities between transgenic and WT plants under well-watered treatment or after rewatering.

Figure 2.

Aldehyde oxidase (AO) activity and ABA content in transgenic LOS5/ABA3 and WT soybean under well-watered and drought stress conditions. (a) AO activity of transgenic and WT soybean. Values of AO activity presented percentage of the relative intensity of formazan bands of plants explored to different water treatments compared with that of WT under well-watered condition. (b) Changes in ABA content in transgenic and WT soybean. Drought stress treatment was imposed as described in Figure 1(c) above. Values were the mean ± SD (n = 6), and asterisks indicated a significant difference (< 0.05) compared with the corresponding controls.

To determine whether overexpressed LOS5/ABA3 under drought stress increased ABA levels in transgenic soybean, ABA content was detected by an ABA immunoassay kit. Under D1 treatment, transgenic lines S-2 and S-3 exhibited 76% and 67% higher ABA levels than those of WT, respectively (Figure 2b). Also, in D2 treatment, leaves of S-2 and S-3 had 88% and 77% higher ABA content than that of WT, respectively. Otherwise, ABA concentrations were similar between transgenic and WT plants under well-watered or rewatering conditions.

LOS5/ABA3 overexpression prevents water loss and modulates stomatal closure

To assess the effect of LOS5/ABA3 overexpression on drought tolerance, 21-day-old seedlings were deprived of water for 10 days, and then, watering was resumed in growth chamber. After 2 days without water, leaves of WT plants showed slight wilting, but transgenic lines S-2 and S-3 had normal turgid leaves (Figure 3a). After 5 days of without water, leaves of WT plants were severely wilted and leaves of transgenic lines S-2 and S-3 were moderately wilted. After 10 days without water, leaves of transgenic line S-2 showed less wilting than that of line S-3, but leaves of WT plants were completely wilted. When wilted plants were rewatered, transgenic lines S-2 and S-3 were revived, but WT plants did not recover.

Figure 3.

Phenotype, water loss and stomatal assay of transgenic LOS5/ABA3 and WT soybean leaves under well-watered and drought stress conditions. (a) Drought-stress-tolerant phenotype of transgenic and WT soybean. Drought stress treatment was applied to 21-day-old seedlings of transgenic and WT soybean by completely withholding irrigation for 10 days, then re-watering for 2 days. 21d, 21-day-old seedling; D 2 days, completely withholding irrigation for 2 days; D 5 days, completely withholding irrigation for 5 days; D 10 days, completely withholding irrigation for 10 days; R 2 days, rewatered for 2 days. (b) Water loss from detached leaves of transgenic and WT soybean. Values were the mean ± SD (n = 4). (c) Stomatal apertures of transgenic and WT soybean. Drought stress treatment was imposed as described in Figure 1(c) above. Values were the mean ± SD (n = 15). (d) Transpiration rate in transgenic and WT soybean. Drought stress treatment was imposed as described in Figure 1(c) above. Values were the mean ± SD (n = 6). Asterisks indicated a significant difference (< 0.05) compared with the corresponding controls.

When the leaves were detached from the 21-day-old seedlings, transpirational water loss from transgenic lines S-2 and S-3 was 30% and 16% less than that from WT soybean, respectively (Figure 3b). Reduced water loss by transgenic plants indicated that stomatal action was regulated by overexpressed LOS5/ABA3. Under well-watered condition, stomatal apertures of transgenic lines were similar to WT plants. However, stomatal apertures of S-2 and S-3 under D1 treatment were 39% and 32% lower than those of WT soybean, respectively (Figure 3c). Exposed to D2 treatment, stomatal apertures of S-2 and S-3 were 53% and 70% that of WT, respectively. After rewatering, stomatal apertures of transgenic lines were significantly larger than those of WT plants.

Changes in transpiration rate of transgenic and WT plants followed a similar pattern as stomatal apertures (Figure 3d). For example, the transpiration rate of S-2 and S-3 was 22% and 19% less than that of WT under D1 treatment and was 39% and 27% less than that of WT under D2 treatment, respectively. Under well-watered or rewatering conditions, values of transgenic lines were higher than those of WT.

To further characterize the drought response, leaf water potential and RWC of transgenic and WT soybean leaves were evaluated in growth chamber and field. Under well-watered condition, values of leaf water potential for S-2 and S-3 were similar to those of WT plants (Figure 4a,b). Drought stress caused the leaf water potential of transgenic and WT soybean to decline, but leaf water potential of transgenic lines was much more than that of WT. After rewatering, leaf water potential of transgenic and WT soybean was restored well-watered values. RWC of transgenic lines was higher than that of WT under drought stresses in growth chamber and field (Figure 4c,d). Such as in growth chamber, RWC of S-2 and S-3 was 12% and 9% higher than that of WT under D1 treatment and 17% and 9% higher than that of WT under D2 treatment, respectively. RWC between transgenic and WT plants under well-watered condition or after rewatering was similar.

Figure 4.

Leaf water status assay of transgenic LOS5/ABA3 and WT soybean exposed to drought stresses. Changes in leaf water potential and relative water content of transgenic and WT soybean subjected to drought stresses in growth chamber (a, c) and field condition (b, d). Drought stress treatment in growth chamber was imposed as described in Figure 1(c) above. And in the field, transgenic plants were subjected to 21 days of drought stress starting at R1 stages, thus followed by 7 days of rewatering. Values were the mean ± SD (n = 6), and asterisks indicated a significant difference (< 0.05) compared with the corresponding controls.

LOS5/ABA3 overexpression reduces low cell membrane damage

Drought stress usually is associated with increased production of activated oxygen species, which leads to damaged cell structures. To assess whether LOS5/ABA3 overexpression reduced activated oxygen damage, nonenzymatic and enzymatic antioxidants were measured. Electrolyte leakage in transgenic lines under drought stress was maintained lower than that in WT in growth chamber or field (Figure 5a,b). For example, under D1 treatment, electrolyte leakage of S-2 and S-3 was 15% and 11% lower than that of WT and under D2 treatment was 12% and 9% lower than that of WT, respectively. Also, similar results were shown in the field. The electrolyte leakage was similar between transgenic and WT plants under well-watered condition or after rewatering.

Figure 5.

Cell membrane damage and nonenzymatic antioxidants assay of transgenic LOS5/ABA3 and WT soybean under well-watered and drought conditions. Changes in malondialdehyde content, electrolyte leakage and proline content in plants exposed to drought stress in growth chamber (a, c and e) and field conditions (b, d and f), respectively. Drought stress treatment was imposed as described in Figure 4 above. Values were the mean ± SD (n = 6), and asterisks indicated a significant difference (< 0.05) compared with the corresponding controls.

When exposed to drought stresses, transgenic lines S-2 and S-3 significantly produced 32% and 27% less malondialdehyde (MDA) content compared with WT under D1 treatment and produced 22% and 18% less MDA content than that of WT under D2 treatment, respectively (Figure 5c). At the same time, transgenic lines S-2 and S-3 exhibited less MDA accumulation compared with WT plants with drought stresses enhancing under field condition (Figure 5d). However, MDA production in well-watered or rewatering treatments was similar for transgenic and WT plants.

Proline accumulation is one positive response to drought by avoiding dehydration in many plant species. WT and transgenic lines under well-watered condition had similar proline contents (Figure 5e,f). Proline contents increased with increasing drought stresses and were markedly higher in transgenic lines than in WT plants under both D1 and D2 treatments. After rewatering, proline contents of transgenic and WT plants were similar and had returned to levels similar to well-watered plants.

To study further the difference between transgenic and WT soybean under drought stress, activities of superoxide dismutase (SOD), peroxidase (POD) and catalase (CAT) in leaves of transgenic and WT plants were determined. Activities of SOD, POD and CAT under well-watered condition were similar between transgenic and WT plants (Figure 6). However, drought stress greatly increased the activities of SOD, POD and CAT in all plants, and the activities of SOD, POD and CAT were higher in transgenic lines than in WT. After rewatering, activities of SOD, POD and CAT in transgenic and WT plants decreased rapidly and were similar between transgenic and WT soybean.

Figure 6.

Antioxidant enzyme activities assay in transgenic LOS5/ABA3 and WT soybean explored to drought stresses. The changes in superoxide dismutase, peroxidase and catalase activities in plants subjected to drought stresses in growth chamber (a, c and e) and field conditions (b, d and f), respectively. Drought stress treatment was imposed as described in Figure 4 above. Values were the mean ± SD (n = 6), and asterisks indicated a significant difference (< 0.05) compared with the corresponding controls.

LOS5/ABA3 overexpression enhances expression of drought stress-related genes

Overexpression of LOS5/ABA3 leads to drought tolerance in soybean, so we determined the phenotypic changes related to changes in the expression patterns of stress-responsive genes in transgenic lines. In real-time PCR analysis under well-watered condition, expression of abiotic stress-related genes, such as GmDREB1, GmRR34, GmNAC3 and GmP5CS, was similar for transgenic and WT soybean (Figure 7). However, expression of GmDREB1 under drought stresses in S2 or S3 was higher than that of WT (Figure 7a). Expression of GmRR34 and GmNAC3 of transgenic and WT soybean followed a pattern similar to GmDREB1 (Figure 7b,c), and expression of these genes under drought stress was higher in transgenic lines than in WT. The increase in GmP5CS expression under drought stresses was observed in both transgenic and WT plants, but the expression was greater in transgenic lines than in WT (Figure 7d).

Figure 7.

Expression patterns of GmDREB1 (a), GmRR34 (b), GmNAC3 (c) and GmP5CS (d) stress-responsive genes in transgenic LOS5/ABA3 and WT soybean under well-watered and drought conditions. The expression level of transgenic lines was shown relative to the expression of WT plants grown under well-watered condition. Drought stress treatment was imposed as described in Figure 1(c) above. Values were the mean ± SD (n = 6), and asterisks indicated a significant difference (< 0.05) compared with the corresponding controls.

LOS5/ABA3 overexpression enhances biomass accumulation and produced higher yield in the yield

Transgenic and WT plants were subjected to D2 treatment for 5 days and then were returned to normal water supply for 5 days. LOS5/ABA3 overexpression in soybean under drought stress promoted root growth and root system that were considerably larger than in WT. LOS-5/ABA3-expressing plants produced significantly more dry mass than WT soybean (Table 2). For example, S-2 and S-3 produced 18% and 14% more dry shoot mass and 42% and 35% more dry root mass than WT. After drought stresses, root/shoot ratio of S-2 and S-3 was 20% and 19% greater than that of WT, respectively. However, dry mass and root/shoot ratio under well-watered conditions were similar for transgenic and WT plants.

Table 2. Biomass assay in transgenic and WT soybean
LinesShoot dry weight (g per plant)Root dry weight (g per plant)Root/shoot
Well-wateredDroughtWell-wateredDroughtWell-wateredDrought
  1. Dry mass of shoots and roots of 21-day-old WT and transgenic soybean were exposed to severe drought stress (D2) by withholding water for 5 days and restored water for 5 days. Values were the mean ± SD (n = 10), and * indicated a significant difference (P < 0.05) compared with the corresponding controls.

WT2.723 ± 0.1071.384 ± 0.0940.652 ± 0.0260.362 ± 0.0120.239 ± 0.0210.262 ± 0.012
S-22.735 ± 0.1401.638 ± 0.102*0.662 ± 0.0340.515 ± 0.023*0.242 ± 0.0320.314 ± 0.016*
S-32.718 ± 0.1051.575 ± 0.075*0.639 ± 0.0170.488 ± 0.031*0.235 ± 0.0130.310 ± 0.011*

Field trials were conducted in a rain shelter to evaluate the performance of LOS5/ABA3 transgenic soybean plants, and WT and two independent LOS5/ABA3 lines S-2 and S-3 were grown at the Wuqiao Experimental Farm in Hebei in the summer of 2011. The yields of soybean plants subjected to 21 days of drought stress starting at the R1 stage in the field were shown in Figure 8. The upground dry biomass in transgenic lines S-2 and S-3 was 21% and 16% higher than that in WT plants, and root dry biomass was 24% and 19% higher than that in WT plants under drought stress condition (Figure 8a,b). Thus, the LOS5/ABA3-overexpressing soybean lines S-2 and S-3 produced 36% and 29% more pots number per plant and 32% and 21% more seed yield per plant than those of WT plants under drought stress condition (Figure 8c,d). Again, LOS5/ABA3-overexpressing soybean plants outperformed WT plants with respect to seed yield in the field and produced 29% and 21% more seed yield than wild-type plants (Figure 8e). However, LOS5/ABA3-overexpressing soybean lines S-2 and S-3 showed similar values in biomass, pods and seed weight per plant, and yield compared with WT plants under well-watered conditions.

Figure 8.

Biomass and yield assay of overexpression LOS5/ABA3 transgenic soybean in the field. (a) Upground dry biomass of transgenic and WT soybean plants; (b) root dry biomass of transgenic and WT soybean plants; (c) total pods per plant of transgenic and WT soybean plants; (d) seed yield per plant of transgenic and WT soybean plants; (e) seed yield of transgenic and WT soybean plants. WW, normal irrigation; DS, nonirrigation starting at R1 stages for 21 days. Values were the mean ± SD (n = 50) in (a) to (d), and values of seed yield were mean ± SD (n = 3). Asterisks indicated a significant difference (P < 0.05) compared with the corresponding controls.

Discussion

Although soybean is one of the worldwide major food crops, there are few reports of genetic engineering to improve or manipulate drought stress in soybean (Manavalan et al., 2007). Here, we report the identification and characterization of transgenic LOS5/ABA3 soybean. LOS5/ABA3 encodes a MoCo sulphurase that catalyses production of a sulphurylated MoCo required by AO and functions indirectly in ABA biosynthesis (Bittner et al., 2001; Xiong et al., 2001). The LOS5/ABA3 gene is expressed at relatively low levels in several plant parts at normal conditions, but the expression level in Arabidopsis increases in response to drought (Xiong et al., 2001). In the present work, relatively low expression of LOS5/ABA3 was observed in leaves of transgenic soybean under well-watered condition, and transgenic plants were similar to WT plants in morphology and ABA content. Although transcription of the transgene was driven by the constitutive promoter (Figure 1d), the expression of LOS5/ABA3 was greatly up-regulated in transgenic S-2 or S-3 soybean explored to drought stress. This observation has been reported in other events (Baisakh et al., 2012; Shi et al., 2003; Wei et al., 2011), and stresses can cause a post-transcriptional stabilization of the transcript. Post-transcriptional control of transcript accumulation is an important mechanism for gene regulation under stress (Cohen et al., 1999; Shi et al., 2003). It would be interesting to identify the factor or factors that mediate LOS5/ABA3 transcript stabilization in transgenic soybean under drought stress.

As has been suggested by Liu et al. (1998) and Sakuma et al. (2006), DREB factors activity in activating stress-responsive genes may require post-transcriptional modifications, which depended on ABA-regulated molecules or numerous other ABA-responsive regulatory factors (Merlot et al., 2001; Rock, 2000). GmDREB1 is an element-binding protein specifically bound to the dehydration-responsive element, thereby inducing expression in response to drought and salt stresses in leaves of soybean seedlings (Li et al., 2005). GmRR34, containing dehydration-responsive ABRE and MYBR motifs in the promoter regions, is a dehydration-responsive TCS gene involved in the regulation of drought stress response in soybean (Mochida et al., 2010), and the ABRE complex in these promoters mediates gene induction by ABA (Nakashima et al., 2009). Tran et al. (2009) found NAC genes from soybean were induced by drought and other stresses such as high salinity, cold and ABA treatment, and GmNAC3 is a plant-specific NAC transcription factor that is markedly induced by dehydration (Le et al., 2011). In our studies, the expression of GmDREB1, GmRR34 and GmNAC3 was enhanced by drought stress in transgenic and WT plants, but overexpressing LOS5/ABA3 significantly increased their expression levels compared with WT plants under drought stresses (Figure 7a–c). Due to the promoter region of LOS5/ABA3 containing putative ABRES and DRE-/CRT-like elements, the LOS5/ABA3 gene may be regulated by ABA and drought stress in a manner similar to other stress-responsive genes (Xiong et al., 2001). Thus, the role of LOS5/ABA3 in the osmotic stress regulation of gene expression is mediated by ABA, and ABA up-regulation of LOS5/ABA3 expression is a positive feedback regulation of ABA biosynthesis. Taken together, our results demonstrated that LOS5/ABA3 overexpression enhanced drought tolerance via ABA signalling that manipulated expression of stress-regulated genes.

LOS5/ABA3 encodes a MoCo sulphurase which catalyses production of a sulphurylated MoCo required by AO and xanthine oxidase (XHD) to gain enzymatic activity, which are involved in ABA biosynthesis and degradation of purines, respectively (Bittner et al., 2001; Xiong et al., 2001). Although transcription of LOS5/ABA3 was driven by a constitutive super promoter, physiological parameters such as AO activity and ABA content showed no significant difference between transgenic and WT plants under well-watered condition. However, LOS5/ABA3-overexpressing plants had higher AO activity and ABA content compared with WT plants under drought stress (Figure 2). Based on the model for stress induction of ABA biosynthesis (Xiong et al., 2002), the limited amount of newly synthesized ABA would then stimulate the expression of AAO (Seo et al., 2004) and LOS5/ABA3, and this coordinated increase in the transcription of all ABA biosynthesis genes would result in a more rapid and sustained increase in ABA biosynthesis. Our results suggested that constitutive expression of LOS5/ABA3 in soybean could not promote ABA accumulation under well-watered or rewatering conditions and might regulate XHD. The effect of LOS5/ABA3 overexpression on regulating XHD requires further investigation.

ABA plays an important regulatory role in plant water status through guard cells and growth as well as by inducing expression of drought and desiccation tolerance genes in cellular dehydration tolerance (Zhu, 2002). Transgenic soybean under drought conditions had smaller stomatal apertures and lower transpiration rate compared with WT plants, which reduced water loss in transgenic plants. Also, LOS5/ABA3 overexpression enhanced water-retaining ability (higher RWC) of transgenic plants in response to drought. Our results are consistent with NCED overexpression in tomato or tobacco, which leads to increased ABA production and reduced leaf transpiration under drought conditions (Thompson et al., 2000, 2007).

Drought causes a rapid and excessive accumulation of ROS in plant cells, and MDA is a secondary end product of oxidation of polyunsaturated fatty acids, as an indicator for the degree of oxidative stress. Aldehyde dehydrogenases play a major role in the detoxification processes of aldehydes generated in plants when exposed to abiotic stresses, and ALDH genes are up-regulated by dehydration and ABA (Kirch et al., 2001; Sunkar et al., 2003). LOS5/ABA3 overexpression reduced production of MDA and decreased electrolyte leakage under drought stress, which could alleviate membrane damage in transgenic soybean (Figure 5). ROS are signalling molecules that regulate plant-protective stress responses including ABA-induced activation of stomatal closure and induction of defence gene expression (Desikan et al., 2001). Water-stress-induced ABA accumulation regulated ABA-/stress-responsive gene expression including ROS network genes such as SOD, APX and CAT (Jiang and Zhang, 2002; Yamaguchi-Shinozaki and Shinozaki, 2006). Our results show that the activities of SOD, CAT and POD in LOS5/ABA3-expressing soybean are higher compared with WT plants under drought stress (Figure 6). Concurrently, LOS5-/ABA3-expressing soybean exposed to drought stress promotes accumulation of proline, and GmP5CS expression in transgenic plants is 1.7- to 2.1-fold higher compared with WT soybean. P5CS, a key regulatory enzyme in proline biosynthesis, plays a key role in biosynthesis under osmotic stress (Savouré et al., 1995; Yoshiba et al., 1995). Strizhov et al. (1997) and Xiong et al. (2001) reported that stress-induced P5CS gene expression under osmotic stress required ABA. It is possible that LOS5/ABA3-overexpressing plants under drought stress could accumulate proline by overproducing ABA. Our results clearly indicate that LOS5/ABA3 overexpression enabled soybean to detoxify ROS efficiently and to enhance drought stress tolerance via ABA accumulation to mobilize ROS-scavenging enzymes.

Well-watered transgenic tobacco overexpressing PvNCED1, a rate-limiting enzyme in ABA biosynthesis, was indistinguishable from WT plants in both vegetative and reproductive phases, but reduced stomatal conductance and delayed seed germination (Qin and Zeevaart, 2002). Similar results were obtained when orthologous genes LeNCED1 (Thompson et al., 2000) and AtNCED3 (Iuchi et al., 2001) were overexpressed in tomato, tobacco and Arabidopsis, respectively. Conversely, our results exhibit that LOS5/ABA3 overexpression did not change stomatal apertures under well-watered conditions and increased stomatal apertures after rewatering and does not delay seed germination in transgenic soybean (data not shown). Otherwise, we observed that LOS5/ABA3-overexpressing soybean after rewatering grew much better than WT. The dry shoot and root mass of LOS5/ABA3-expressing soybean seedlings grown under different drought stress conditions were markedly higher than those of WT (Table 2). Thus, overexpression of LOS5/ABA3 in soybean promotes root system development and growth leading to enhanced drought stress tolerance. Similar results were shown in field trials, and overexpression of LOS5/ABA3 in soybean (S-2 and S-3) exhibited more upground and root dry biomass and higher pods and seed yield compared with WT plants under drought conditions (Figure 8).

In summary, LOS5/ABA3 overexpression in soybean under drought stress increased drought tolerance by regulating AO activity to promote ABA accumulation. ABA accumulation of transgenic soybean exposed to drought stress reduced water loss and activated expression of stress-regulated genes that alleviate membrane damage. These data provide important insights into utilization of ABA-related biosynthesis genes and significantly further our understanding of stress gene regulation and stress tolerance in plants.

Experimental procedures

Construction of the binary vector, plant transformation and selection

LOS5/ABA3 cDNA was cloned as a XbaI–KpnI fragment downstream of the constitutive super promoter, which consists of three copies of the octopine synthase enhancer in front of the mannopine synthase promoter in the modified pCAMBIA 1300 (Figure 1a).The super promoter was cloned as a SalI–XbaI fragment into the pCAMBIA 1300 binary vector containing a hygromycin-resistant selectable marker (Yang et al., 2009). The recombinant plasmid was introduced into Agrobacterium tumefaciens strain EHA105, which was used to transform soybean. Transformation of soybean cv. Zhonghuang 20 via the Agrobacterium-mediated cotyledonary node method was performed as described by Olhoft et al. (2003). Transgenic soybean harbouring LOS5/ABA3 was screened on Murashige and Skoog (MS) agar medium containing 10 mg/L hygromycin, and 57 independent transgenic lines were produced. All the homozygous lines were assayed for drought tolerance. Ten independent representative lines exhibited relatively high drought tolerance compared with WT, based on leaf water status and biomass response to drought stress. Then, four single-copy transgene integrations were obtained by Southern blotting analysis (Chen et al., 2012). Two independent representative lines (S-2 and S-3), with single-copy transgene integration and similar morphological characteristics to WT under well-watered conditions, were executed detailed analysis to explore drought tolerance mechanism.

Transgenic lines with hygromycin resistance were transplanted into the (15 × 15 × 20 cm deep) filled with a mixture of vermiculite and sand (1 : 1; v/v) and grown in growth chamber. The presence and integrity of the transgene were further confirmed by PCR analysis using LOS5/ABA3-specific primers (forward: 5′-CCTGATGGCTCTTGGTTTGGCTAC-3′/reverse: 5′-TTCCACTGACGACGGTTCCATTCC-3′) and genomic DNA obtained from soybean leaves following the cetyltrimethylammonium bromide method (Saghai-Maroof et al., 1984). Furthermore, the copy numbers in T3 generation of transgenic lines were confirmed by Southern blotting (Chen et al., 2012). Genomic DNAs from wild-type and T3 transgenic lines were digested with the restriction enzyme EcoRI at 37 °C overnight. Digested DNA was separated on 0.8% (w/v) agarose gel and blotted onto Hybond N+ nylon membrane. The membrane was hybridized with a DIG-labelled Hyg-specific probe (Hyg-61F 5΄-TACACAGCCATCGGTCCAGA-3΄ and Hyg-905R 5΄-TAGGAGGGCGTGGATATGTC-3΄) at 68 °C for 16–18 h. The hybridized membrane was washed and detected according to the protocol of DIG Nucleic Acid Detection Kit (Roche corporation, Basel, Switzerland).

Drought experiments in growth chamber

Seeds of WT and T4 transgenic soybean (independent lines S-2 and S-3) were planted into the pots (15 × 15 × 20 cm deep) filled with a mixture of vermiculite and sand (1 : 1; v/v) and grown in growth chamber with a 14-h photoperiod at 25/30 °C night/day temperature cycle, 400 μmol/m2/s irradiance (enhanced with high-pressure sodium lamps) and a relative humidity of 60%.

Drought stress was induced by completely withholding irrigation to 21-day-old transgenic and WT soybean seedlings (at V3 stage) for 10 days. Drought-stress-tolerant phenotypes of transgenic lines were observed, the number of wilted plants was scored, and photographs were taken. Then, the wilted plants were rewatered to resume the growth, and drought-stress-tolerant phenotypes of transgenic soybean after 2 days were recorded.

Drought experiments were conducted with transgenic and WT soybean grown in the pots as above. Plants were watered to capacity daily by providing about 400 mL water per pot. Uniform plants were divided randomly into four groups: well-watered group, moderate drought group (D1, 60% normal water supply), severe drought group (D2, 40% normal water supply) and rewatered group (rewatered after D2 treatments). For the drought treatments, D1 and D2 irrigations were with 240 and 160 mL water per pot daily, respectively, for 5 days. Then, the rewatered group following D2 treatment was supplied with normal water regime for 2 days, and some plants were further cultured for another 5 days for biomass analysis.

At each harvest, the plant was separated into shoots and roots. Shoots were cut at the cotyledon node and fresh weight determined. Roots were measured by pulling pots from the ground and the root mass was soaked in water and then manually stirred and poured into a sieve (0.25-mm2 mesh). The sieve was suspended in a large water bath and shaken continuously until roots were washed free of soil. Soil materials remaining on the sieve were removed manually. The separated root fractions were collected to determine fresh weight. Then, all samples were killed at 105 °C for 30 min and dried at 70 °C to determine the shoot and root dry weight.

Water loss measurements

The first trifoliolate leaves of transgenic and WT soybean were detached to measure transpirational water loss as described by Chen et al. (2006). Stomatal bioassay was performed as described by Pei et al. (1997) with slight modifications. Leaves were carefully cut into 8-mm-long and 4-mm-wide strips, and the strips were immediately incubated in FAA fixative liquid (38% formaldehyde, acetic acid and 50% alcohol, 5:5 : 90). Stomata were observed under a scanning electron microscope (S-570; Hitachi, Tokyo, Japan), and the width and length of stomatal apertures were measured using image analysis computer software (Scion Image; Scion Corp., Frederick, MD).

Transpiration, membrane damage and ABA analysis of transgenic plants

The first trifoliolate leaves of transgenic and WT soybean were taken on 0 day (21-day-old seedlings) and 5 days after initiation of different drought treatments and 2 days after rewatering as described above. A pressure chamber (Model 3000; Soil Moisture Equipment Corp., Santa Barbara, CA) was used to measure the leaf water potential, with one leaf per plant from six plants for each treatment. The RWC was measured as described by Gaxiola et al. (2001). Transpiration rate were measured using a portable infrared gas analyzer-based photosynthesis system (LI-6400; Li-Cor Inc., Lincoln, NE).

Membrane damage was assayed by measuring electrolyte leakage from leaf discs according to Shou et al. (2004). The extent of lipid peroxidation was estimated by measuring the amount of MDA according to Quan et al. (2004). Proline content was measured as described by Bates et al. (1973). Endogenous ABA content was measured by an indirect enzyme-linked immunosorbent assay as described by Yang et al. (2001).

Analysis of enzyme activity

Fresh leaf segments from transgenic and WT soybean from different drought treatments were crushed into fine powder in a mortar and pestle under liquid nitrogen. Soluble protein content was determined following the method of Bradford (1976). Total SOD activity was assayed as described by Giannopolitis and Ries (1977). POD activity was determined by the guaiacol oxidation method (Nakno and Asada, 1981). CAT activity was measured as described by Aebi (1984).

Protein extraction and AO active staining were performed as described by Porch et al. (2006) with minor modifications. After activity attaining, the gels were scanned to quantify the relative intensity of formazan bands, which were directly proportional to enzyme activity (Zdunek and Lips, 2001) using the Quantity One computer software in Bio-Rad ChemiDoc SRS (Bio-Rad, Hercules, CA). Native PAGE was carried out with a protein IIxi Cell (JunYi, Beijing, China).

Real-time quantitative PCR (RT-qPCR) analysis

Fresh leaves of 21-day-old transgenic and WT soybean from different drought treatments were collected in liquid nitrogen before isolation of RNA. Total RNA was isolated using TRIZOL® reagent (Invitrogen, Carlsbad, CA) and purified using Qiagen RNA easy columns (Qiagen, Hilden, Germany) according to the instructions of the manufacturer. Reverse transcription was performed using Moloney murine leukaemia virus (M-MLV; Invitrogen) according to the method described by Le et al. (2011). Real-time quantitative RT-PCR was performed on a 7500 real-time PCR system (Applied Biosystems, Foster, CA) using SYBR® Premix Ex Taq™ (TaKaRa Code: DRR041A, Dalian, China). Primer Express program 3.0 (Applied Biosystems) was used to design the primers of chosen genes: LOS5/ABA3, forward: 5΄-GGGAAAGGGTGGAGGAGT-3΄/reverse: 5΄-GTAGCCAAACCAAGAGCC-3΄; GmNAC3, forward: 5΄-GTGTTATTGATTCCCGCTTG-3΄/reverse: 5΄-TCAACCGTCCTTCATCTTGT-3΄; GmRR34, forward: 5΄-GAGGCAACAAAGGAACTTCG-3΄/reverse: 5΄-TCCGTACAGCGTGATGATACA-3΄; GmDREB1a, forward: 5΄-CGACCAGGAGGGCAGTGAT-3΄/reverse: 5΄-GCTTTTCGGCGAATGGAAT-3΄; GmP5CS, forward: 5΄-TGTCTCTCAGATCAAGAGTTCCAC-3΄/reverse: 5΄-CAGCCTGCTGGATAGTCTATTTTT-3΄; Gm18SrRNA, forward: 5΄-AAACGGCTACCACATCCAAG-3΄/reverse: 5΄-CCTTCAATGGATCCATCGTT A-3′. Fold changes in RNA transcripts were calculated by 2-∆∆Ct method (Livak and Schmittgen, 2001). The relative expression levels of transgenes were calculated using the 18SrRNA gene as an internal control.

Drought treatment of soybean plants in field trials

A field experiment was performed in the CAU Experimental Station in Wuqiao, Hebei Province (116 : 22E, 37 : 37N), under a shelter (the shelter was rolled down on rainy days). The trial plots in nearly every case were in a randomized complete block design with three replications. Each transgenic line and WT was sown in the equal row plot in early May. The plot was 20 m2 and 440 plants were planted in each plot. Plants were subjected to drought stress starting at the R1 stage for 21 days in the field. Nonirrigated treatment was designed as drought stress and normal irrigation as well-watered treatment. After the drought stress, the plants were rewatered. Seed yields were determined from the four centre rows from the 10-row plots (5.0-m-long and 0.4-m row spacing). The pod and seed numbers were recorded after harvest from fifty randomly selected plants for each plot.

Physiological and biochemical estimations were carried out using the second upmost expanded leaves. Samples were collected 0, 7, 14 and 21 days after the water treatment. Fresh samples of all treatments were used for immediate assays or frozen in liquid nitrogen and stored at −80 °C for physiological and biochemical analysis.

Statistical analysis

Results are based on two independent experiments with at least three replicate tissue samples from transgenic and WT soybean in each treatment. All data in figures and tables were compared using Tukey–Kramer honestly significant difference test (< 0.05), and data were presented as mean ± standard deviation (SD).

Acknowledgements

This work was supported by National Nature Science Funds for Distinguished Young Scholars (No. 30825028) and the Ministry of Agricultural of China for transgenic research (No. 2011ZX08004-002 and -005). We thank Dr Jiankang Zhu (Purdue University, USA) and Dr Zhizhong Gong (China Agricultural University, Beijing) for supplying LOS5/ABA3 gene and excellent technical assistance. The authors also thank Dr Calvin G. Messersmith (North Dakota State University, USA) and Dr Jinsheng Lai (China Agricultural University, Beijing) for technical improvement to the manuscript.

Ancillary