Overexpressing the WRKY transcription factor OsWRKY97 improves drought tolerance in rice

Background: WRKY transcription factor (TF) is one of the largest TF families in plants and plays an important role in plant development and stress protection. However, the function of individual WRKY gene in plants are still under investigation. Here, we identified a new member of WRKY TF family, OsWRKY97 , and analyze its role in stress resistance by using a series of transgenic plant lines. Results: OsWRKY97 , which positively regulates drought tolerance in rice. OsWRKY97 was expressed in all examined tissues and could be induced by various abiotic stresses. OsWRKY97 is located in the nucleus. Compared with the wild-type (WT), OsWRKY97 -overexpressing (OE) plants enhanced the tolerance of rice to drought and osmotic stress. In addition, OE plants also showed higher sensitivity to exogenous abscisic acid (ABA). When exposed to drought stress, up-regulation OsWRKY97 increased the accumulation of ABA and reduced water loss. Furthermore, OE plants achieved higher proline content and reduced levels of malondialdehyde (MDA) and reactive oxygen species (ROS). The homeostasis of ROS was maintained via the activity of active oxygen scavenging enzyme, superoxide dismutase (SOD), catalase (CAT) and peroxidase (POD), activated by OsWRKY97 Conclusions: These results indicate that OsWRKY97 plays a crucial role in the response to drought stress and may possess high potential value in improving drought tolerance in rice. seedlings OsWRKY97

sensitivity of ABA mediated stomatal closure and affects its drought tolerance [23]. It is reported that WRKY40 and WRKY60 transcription factors inhibit ABI4 gene expression. In addition, their mutual antagonism not only occurs between WRKY40 and WRKY60, but also between WRKY18 and WRKY40 to balance the repressive functions on the ABI4 gene. However, for ABI5 gene, WRKY60, WRKY18 and WRKY40 inhibited ABI5 independently, but WRKY60 also antagonized the inhibition of WRKY18 and WRKY18-WRKY40 heterodimer to balance the inhibition of ABI5 gene. These findings help to understand the complex mechanism of WRKY mediated ABA signal transduction [24].
The corresponding sequence of OsWRKY97 (LOC_Os12g02420.1) can be downloaded from Rice Genome Annotation Project (RGAP) (http://rice.plantbiology.msu.edu/) and OsWRKY97 belongs to the WRKY family according to the SMART database (http://smart.embl-heidelberg.de/). The OsWRKY97 gene was identified by analyzing rice drought stress microarray data generated by our lab (unpublished). The gene was strongly up-regulated under drought conditions in microarray data. So we think that OsWRKY97 may also play an important role in the response to abiotic stresses. In this study, we found that the expression level of OsWRKY97 in Nipponbare was significantly improved under abiotic stress, and overexpression of OsWRKY97 improved the tolerance of transgenic rice to drought and osmotic stress. OE plants also showed higher sensitivity to exogenous ABA and increased the accumulation of ABA under drought stress. In addition, we found that OE plants had higher activity of active oxygen scavenging enzymes, less membrane lipid peroxidation and higher accumulation of proline under drought conditions. These results suggest that OsWRKY97 positively regulates drought stress tolerance in rice. various stresses (Figure 1). For drought treatment, the transcription level of OsWRKY97 gradually increased until reaching a maximum at 8 h ( Figure 1A). The transcription level of OsWRKY97 reached its maximum after treatment with 50 µM ABA for 12 hours, and similar results were obtained after treatment with 250 mM NaCl ( Figure 1B,C). The expression of OsWRKY97 increased significantly in 12 to 24 h under cold stress ( Figure 1D). However, the transcription level of OsWRKY97 under high temperature treatment was not significantly induced ( Figure 1E).

Subcellular localization of OsWRKY97
To determine the subcellular localization of OsWRKY97, the full-length cDNA sequence of OsWRKY97 was fused to GFP and driven by CaMV35S. The fusion protein and green fluorescent protein (GFP) control were transiently expressed in tobacco cells via Agrobacterium infiltration, and meanwhile, the carrier of the red fluorescent protein nuclear Localization Signal -red fluorescent protein (NLS-RFP) connected with the nuclear localization signal is used as the nuclear localization control. The fluorescence signal of the fusion protein was located in the nucleus (Figure 2A), while the fluorescence signal of the GFP control was located in the nucleus and cytoplasm ( Figure 2B).
OsWRKY97-GFP fusion protein and NLS-RFP were co-located in the nucleus, indicating that OsWRKY97 is a nuclear protein.

Overexpression of OsWRKY97 in rice improved osmotic stress tolerance at the germination and post-germination stages
In this study, we found that OsWRKY97 was strongly induced by osmotic stress ( Figure 1A). To further verify whether OsWRKY97 was involved in regulating the sensitivity of rice to stress, we constructed an OsWRKY97 overexpression vector and transferred it into Nipponbare. The expression level of OsWRKY97 in overexpressed plants was analyzed by qRT-PCR. The results showed that the expression of OsWRKY97 was significantly enhanced in transgenic lines ( Figure 3A). Although we obtained 18 OE lines, only two independent lines, OE-1 and OE-23 were selected for future testing. Under normal conditions, the germination rates of WT and OE lines were not significantly different. However, the germination rate of WT seeds was less than 50%, however, after absorption for 5 days under the osmotic stress, the germination rate of overexpressed seed OE-1 was 76%, and that of OE-23 was 80%. (Figure 3B,C). To further understand the sensitivity of OE lines to osmotic stress at the postgermination stage, seedings of WT and OE lines that were growing under normal conditions for 4 days were selected and transferred to the nutrient solution with or without 20% PEG6000. After 12 days, each material was photographed, and its plant height was measured. Under normal conditions, the seedlings of WT and OE lines were similar in growth. Under osmotic stress, the plant height of WT seedlings was 8.6 cm, while the plant height of overexpression seeding OE-1 was 10.8 cm, and that of OE-23 was 10.6 cm ( Figure 3D,E). These results showed that overexpression of OsWRKY97 in rice did not affect seed germination and seedling growth under normal growth conditions, but significantly attenuated the inhibitory effects on seed germination and seedling under osmotic stress induced by 20% PEG6000.

Overexpression of OsWRKY97 in rice enhanced drought stress tolerance
To further verify the biological function of OsWRKY97, we tested the drought stress tolerance of OE lines by withholding water. The seedings of OE lines and WT in the same pots that were of similar vigor were not watered until the leaves curled and then recovered. Before the pressure was introduced, the growth conditions of all genotypes were similar. After drought treatment, the leaf rolling degree of WT plants was more serious than that of OE plants. After re-watering, The survival rate of overexpressed plant OE-1 was 53%, and that of OE-23 was 55%, however, the survival rate of WT was only 14%-16% ( Figure 4A,B). These results indicated that overexpression of OsWRKY97 enhances the drought tolerance of rice. The relative water loss rate of detached leaves is an important characteristic reflecting drought tolerance [25]. Leaves of 2-week-old transgenic plants and WT seedlings were removed and exposed to water-free air for dehydration. Compared with WT, the leaf water loss of OE lines was significantly slower ( Figure 4C), which means that OsWRKY97 played an active role in improving the water retention capacity of plants under dehydration conditions.

Effects of OsWRKY97 overexpression on related physiological indexes under drought stress
To further clarify the physiological mechanism by which OsWRKY97 confers tolerance to drought, we investigated the possible physiological basis related to the enhancement of drought resistance in OE plants. ROS were generated in plants when they were subjected to abiotic stresses such as drought, salinity ,heat and cold [26]. H 2 O 2 , which is an important second messenger, is one of the most significant of these ROS [27]. Therefore, we measured the accumulation of H 2 O 2 in plants after drought treatment. We found that the accumulation of H 2 O 2 in OsWRKY97-overexpressing plants after drought treatment was much lower than that of WT ( Figure 5A). To explore the potential mechanism of the reduction of active oxygen level in transgenic plants, ROS scavenging enzyme activity was also measured [11]. As shown in Figure 6, there was no significant difference in SOD, CAT and POD activities between transgenic lines and WT under normal conditions, while the enzyme activities of transgenic lines were higher than those of WT lines under drought stress ( Figure  Under drought stress, the proline content of transgenic plants was significantly higher than that of control plants ( Figure 5F). This result clearly shows that overexpression of OsWRKY97 can increase proline synthesis and protect rice plants to better cope with drought stress.

OsWRKY97 may be a positive regulator in ABA signaling under drought stress
In this study, we found that OsWRKY97 was also strongly induced by ABA ( Figure 1B), to verify this expression pattern, the sensitivities of OE plants to exogenous ABA were examined. The results showed that the germination rate of WT plants was higher than OE plants under the treatment of exogenous ABA ( Figure 6A,B). Therefore, we suspect that the drought tolerance of OE transgenic lines may be related to ABA. In order to verify this hypothesis, we measured the ABA content in OE lines and WT respectively under drought stress. The results showed that the endogenous ABA level was not significantly different between WT plants and OE plants under normal conditions. However, under drought stress, the endogenous ABA level in OE plants was significantly higher than WT plants ( Figure  6C). In addition, we also analyzed the transcription level of response genes in ABA signaling pathway in WT plants and OE plants, including OsRAB21OsRD22 , OsRAB16A , OsNCED3 [29,30] . As shown in Figure 6D, the transcription level of these genes in OE plants was significantly higher than WT plants under drought stress. These results indicate that OsWRKY97 may improve drought tolerance via ABA signaling pathway.

Discussion
Plants will suffer from various abiotic stresses in their natural environment, which will affect the normal development of plants and even affect their yield [31,32]. At present, many studies have shown that WRKY TFs are negatively or positively involved in the integration of signaling pathways in abiotic stress responses [33]. However, the functions of many WRKY TFs in plants, especially stress responses, are still unclear. In this study, we examined whether OsWRKY97 participates in the regulation of the response to drought stress and its effect on rice. In addition, we found that OsWRKY97-GFP subcellular localization was in the nucleus of tobacco epidermal cells (Fig. 2), which is consistent with previous studies on other WRKY TFs and may be related to its function [34].
Increasing evidence shows that WRKY TFs play an important role in the drought stress response, as ectopic expression of OsWRKY11 enhanced tolerance to drought stress and induced constitutive expression of drought-responsive genes [35]. Overexpression of GsWRKY20 from Glycine soja L.G07256 in Arabidopsis resulted in increased sensitivity to ABA when stomata were closed and stronger drought tolerance compared with the WT [36]. Similarly, our results showed that OsWRKY97 expression was rapidly activated under drought stress, suggesting that OsWRKY97 may play an important role in drought stress. Therefore, we constructed OE transgenic plants and tested their resistance to drought stress. The results showed that the germination rate of OE at the germination stage was significantly higher, and the growth inhibition of OsWRKY97 at the seedling stage was also weakened compared with that of WT plants under osmotic stress (Fig. 3). These results suggest that OsWRKY97 may positively regulate drought stress tolerance. In addition, this conclusion was supported by the result that the survival rate of OE plants was higher than that of WT plants under drought conditions (Fig. 4A,B).
Plants respond to water loss at physiological, cellular and molecular levels [37]. ABA is an important plant hormone involved in plant developmental process,which is widely considered as the main regulatory factor of plant response to drought. ABA reduces water loss by inducing stomatal closure and induces a number of stress response genes [38,39]. It has been reported that ABA-independent and ABA-dependent regulatory systems both exist in response to drought stress [40]. In this study, First, we found that OsWRKY97 was strongly induced by exogenous ABA (Fig. 1B). Second, the sensitivities of OE plants to exogenous ABA was enhanced (Fig. 6C). Third, the expression level of OsWRKY97 induces ABA accumulation and the expression level of ABA responsive gene under drought stress (Fig. 6A,D), this result was consistent with the capability of OsWRKY97 to reduce the water loss rate of plants under drought conditions (Fig. 4C). These results intimate that OsWRKY97 improves drought tolerance by enhancing water retention of rice through ABA-dependent pathway. Meanwhile the up-regulation of OsWRKY97 expression under drought stress will lead to the up-regulation of ABA biosynthesis and response genes, resulting in ABA accumulation and increased sensitivity to exogenous ABA. There is evidence that H 2 O 2 acts upstream of ABA signaling pathway. Exogenous H 2 O 2 increases ABA catabolism during seed germination by enhancing the expression of CYP707A genes [41,42]. It has also been reported that H 2 O 2 plays an important role as a second messenger in ABA-induced stomatal closure in guard cells [8,9]. It is reported that H 2 O 2 can be induced by ABA, which is mediated by inducing plant gene expression encoding NADPH oxidase to respond to ABA [43,44]. However, the increase of H 2 O 2 content induced by drought stress is more obvious than that caused by exogenous ABA, this shows that the water deficit signal enhances the production of ROS to a greater extent, which will pose a threat to plants [45]. As expected, the level of H 2 O 2 in plants has increased under drought stress. However, the accumulation of H 2 O 2 in WT plants was significantly higher than that in OE plants (Fig. 5A), and the activities of major oxygen scavenging enzymes, including SOD, POD, CAT, were also observably higher than those in OE plants (Fig. 5B-D). SOD, POD and CAT, which maintain the ROS homeostasis, were activated in OE plants.
As a product of ROS-stimulated lipid peroxidation, MDA contents can be used to evaluate the extent of ROS-mediated injuries in plants. In our research, we found that ROS in WT plants accumulated so much that MDA content increased correspondingly compared with OE plants under drought stress (Fig. 5E). In addition, we also found that drought stress led to the accumulation of osmotic compound, proline, in OE plants was more than WT plants (Fig. 5F). Under osmotic conditions, proline, as an important osmotic protective agent, can maintain low water potential of cells so as not to be damaged by active enzymes [46]. These results showed that WT was more seriously damaged than OE plants under osmotic conditions, thus increasing the probability of WT death.

Conclusions
In conclusion, we identified OsWRKY97, a transcription factor located in the nucleus. OsWRKY97 affects the sensitivity of rice to exogenous ABA and the accumulation of ABA content. In addition, OsWRKY97 affects the redox balance and drought resistance of rice. Redox related mechanisms might be involved in OsWRKY97 mediated drought tolerance, which might affect the content of proline and MDA in rice. All these results indicated that OsWRKY97 gene has a high potential in improving rice drought resistance.

Plant material and growth conditions
Japonica rice (Oryza sativa L. japonica. cv. Nipponbare) was selected as the WT in this study. Rice seeds were sterilized with 0.1% NaClO for 30 min before being soaked in distilled water for 2 days in the dark and were then transferred to a culture dish containing Hoagland nutrient solution and grown in a climate chamber (Southeast instrument, Ningbo, China) with the temperature of 28°C,the relative humidity of 70%,and the 14-hour light/10-hour dark photoperiod [47].

Abiotic treatments
To determine the expression pattern of OsWRKY97 under different stress conditions, two-week-old seedlings of Nipponbare rice were subjected to various stress treatments. For drought treatment, seedlings were grown in culture solution containing 20% PEG6000. For salt treatment, NaCl solution was added to achieve a final concentration of 250 mM. ABA treatment was carried out by adding 50 µM ABA to the culture solution. The seedlings were transferred to a 4°C climate chamber for cold treatment. For heat stress, the seedlings were subjected to 42°C heat shock treatment [48,49].
Samples were collected at 0, 1, 2, 4, 8, 12, 16 and 24 h after treatment. Two-week-old seedings of OsWRKY97 overexpression lines were subjected to drought treatment, and samples were collected at 0h and 10h after treatment respectively.

RNA extraction and real-time PCR
The leaves of 14-day-old plants were sampled, and total RNA was extracted with TRIzol reagent (Invitrogen, Nanjing, China) according to the manufacturer's instructions. One microgram of DNasetreated RNA was reverse-transcribed using a RevertAid RT Reverse Transcription Kit (TaKaRa, Beijing, China) according to the manufacturer's protocol. Real-time PCR was performed on a Bio-Rad CFX96 real-time PCR system. Each reaction was performed in triplicate, and the reaction procedure was as follows: 95°C for 10 min and, then 39 cycles of 95°C for 10s and, 60°C for 30s. The data of relative expression level were analyzed by the 2 -△△Ct method [50]. The OsActin rice gene was used as an internal control gene, and relevant primer pairs are listed in Additional file 2: Table S1.

Vector construction and gene transformation
To construct the overexpression vector of OsWRKY97, cDNA from rice (Oryza sativa L. japonica. cv. Nipponbare) was used. Total RNA was used to amplify the open reading frame of OsWRKY97, and the relevant primer pairs are listed in Additional file 2:Table S1. The fragments confirmed by sequencing were digested and cloned into the pCAMBIA1300 vector, and the digestion sites were HindΙΙΙ and SalΙ.
This gene was driven by the CaMV35S promoter (Additional file 3: Figure S2). All the vectors were transferred into Nipponbare plants via Agrobacterium-mediated transformation [51].

Determination of stress-associated physiological indicators
Seedlings growing under normal conditions for two weeks were transferred to a solution with or without 20% PEG6000. Samples were collected at 0 h and 10 h, and then, the corresponding physiological indexes were measured. MDA content was qualified by thecreaction method as described by Gao [52]. H 2 O 2 content was determined according to the method of Zhan [53]. For the water loss rate, two-week-old seedlings were selected, exposed to air under the condition of no water supply and weighed and recorded every half hour until the weight was constant. Water loss rate was calculated by comparing the measured weight from each indicated time with the measurement at time zero. It is expressed as percentage of initial fresh weight [54]. Proline was detected by following the reported methods [55]. The activity of active oxygen scavenging enzymes was determined according to previous methods [56].

Measurement of ABA content
We analyzed the content of ABA in rice by liquid chromatography-mass spectrometry (LC-MS) [57].
The complete homogenate was incubated overnight in darkness at 4°C. After that, they were vortexed and centrifuged at 15,000 rpm, 4°C for 10 minutes, and the crude extract supernatant was collected and dried in in a rotavapor until an aqueous fractions were obtained. Dried samples were resuspended in 1 mL of acetonitrile/water/acetic acid (90:10:0.05, v/v), and filtered through a 0.45 µm PTFE filter (JET BIOFIL, Guangzhou, China). Quantification was done by the standard addition method by spiking control plant samples with ABA solutions.

Subcellular localization of OsWRKY97
The coding region of OsWRKY97 was amplified and cloned into the pAcGFP1-N1 vector to generate the OsWRKY97-GFP (green fluorescent protein) fusion construct, which was inserted into the pCAMBIA1300 vector, and the constructed vector was transferred into Agrobacterium. After sequencing confirmation, the fusion structure and the control (empty body) vector were cotransfected with another NLS-RFP vector respectively into tobacco cells by Agrobacterium mediated transient expression method, and the fluorescence signal was observed by confocal laser scanning microscope [58].

Analysis of stress tolerance
Two independent T2 over-expression lines and WT seeds were soaked in clear water for two days after being sterilized by 0.1% NaClO, and then, the seeds were transferred to culture solution for growth. When the plant height was approximately 1 cm, seedlings with similar growth potential were selected and transferred to nutrient solution containing 20% PEG, and their growth conditions were observed and recorded. To simulate an arid environment, the two-week-old seedlings were transferred to the sand, and watering was stopped after two days of normal irrigation. The mortality rates were determined. In order to test the sensitivity of ABA sensitivity at germination stage, seeds of OE plants and WT plants were germinated on culture solution containing 5 µM ABA, and the germination rate was calculated on the 6th day after germination.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.