Identification of Two Novel Wheat Drought Tolerance-Related Proteins by Comparative Proteomic Analysis Combined with Virus-Induced Gene Silencing

Drought is a major adversity that limits crop yields. Further exploration of wheat drought tolerance-related genes is critical for the genetic improvement of drought tolerance in this crop. Here, comparative proteomic analysis of two wheat varieties, XN979 and LA379, with contrasting drought tolerance was conducted to screen for drought tolerance-related proteins/genes. Virus-induced gene silencing (VIGS) technology was used to verify the functions of candidate proteins. A total of 335 differentially abundant proteins (DAPs) were exclusively identified in the drought-tolerant variety XN979. Most DAPs were mainly involved in photosynthesis, carbon fixation, glyoxylate and dicarboxylate metabolism, and several other pathways. Two DAPs (W5DYH0 and W5ERN8), dubbed TaDrSR1 and TaDrSR2, respectively, were selected for further functional analysis using VIGS. The relative electrolyte leakage rate and malonaldehyde content increased significantly, while the relative water content and proline content significantly decreased in the TaDrSR1- and TaDrSR2-knock-down plants compared to that in non-knocked-down plants under drought stress conditions. TaDrSR1- and TaDrSR2-knock-down plants exhibited more severe drooping and wilting phenotypes than non-knocked-down plants under drought stress conditions, suggesting that the former were more sensitive to drought stress. These results indicate that TaDrSR1 and TaDrSR2 potentially play vital roles in conferring drought tolerance in common wheat.


Introduction
As a result of the continuous growth of the global population and the adverse effects of environmental change on food production, food security has become a major problem worldwide [1]. Wheat (Triticum aestivum L.) is one of the most important food crops globally. At present, wheat production in many parts of the world is still dependent on excessive use of water, which is ecologically and economically unsustainable [2][3][4]. The shortage of water resources in many regions of the world greatly hinders crop productivity and sustainability. It is estimated that wheat production is likely to decline (W5DYH0 and W5ERN8) were selected and dubbed TaDrSR1 (drought stress response 1) and TaDrSR2 (drought stress response 2), respectively, for further functional analysis using VIGS.

Comparison of Drought Tolerance Between XN979 and LA379
As shown in Figure 1, 16 days after sowing, there were no significant phenotypic changes in drought-tolerant wheat varieties XN979 under drought stress (DS) conditions compared to no stress (NS) conditions. However, leaf drooping appeared in drought-sensitive wheat variety LA379 under drought stress conditions but did not under NS ( Figure 1A). The RWC of LA379 dropped significantly by 40.39% but only dropped by 8.42% in XN979 ( Figure 1B). As proline is an important osmotic regulator, we measured the content of proline in the leaves of both varieties under NS and DS conditions. Results showed that the content of proline significantly increased in both varieties: By 241.59% in XN979 under DS conditions relative to that under NS and 103.81% in LA379 ( Figure 1C). MDA content and electrolyte leakage are important indices of membrane injury. The content of MDA in LA379 significantly increased by 169.64% under DS conditions relative to that under NS but only by 74.02% in XN979 ( Figure 1D). Similarly, the electrolyte leakage was 73.05% higher than under NS in LA379 but only 20.05% higher than under NS in XN979 ( Figure 1E). tolerance variety XN979 were obtained through a comparative proteomic analysis. Two DAPs (W5DYH0 and W5ERN8) were selected and dubbed TaDrSR1 (drought stress response 1) and TaDrSR2 (drought stress response 2), respectively, for further functional analysis using VIGS.

Comparison of Drought Tolerance Between XN979 and LA379
As shown in Figure 1, 16 days after sowing, there were no significant phenotypic changes in drought-tolerant wheat varieties XN979 under drought stress (DS) conditions compared to no stress (NS) conditions. However, leaf drooping appeared in drought-sensitive wheat variety LA379 under drought stress conditions but did not under NS ( Figure 1A). The RWC of LA379 dropped significantly by 40.39% but only dropped by 8.42% in XN979 ( Figure 1B). As proline is an important osmotic regulator, we measured the content of proline in the leaves of both varieties under NS and DS conditions. Results showed that the content of proline significantly increased in both varieties: By 241.59% in XN979 under DS conditions relative to that under NS and 103.81% in LA379 ( Figure 1C). MDA content and electrolyte leakage are important indices of membrane injury. The content of MDA in LA379 significantly increased by 169.64% under DS conditions relative to that under NS but only by 74.02% in XN979 ( Figure 1D). Similarly, the electrolyte leakage was 73.05% higher than under NS in LA379 but only 20.05% higher than under NS in XN979 ( Figure 1E).

Protein Identification and DAPs Analysis
An iTRAQ-based comparative proteomic analysis was performed to identify wheat drought stress tolerance-related proteins. A total of 227076 spectra were generated and 38070 were matched to known spectra. These identified spectra were assigned to 17416 peptides, with 12126 unique peptides. Ultimately, 5369 proteins were identified ( Figure 2A, Table S1). In the XN979_NS-XN979_DS comparison, a total of 482 proteins showed more than a 1.2-fold change (P < 0.05) in their respective abundances and were classified as DAPs (Table S2). Among the DAPs, 199 proteins were up-regulated and 283 proteins were down-regulated in XN979_DS compared to those in XN979_NS ( Figure 2B,C). A comparison of LA379_NS and LA379_DS enabled the identification of 600 DAPs; these comprised 301 up-regulated and 299 down-regulated proteins ( Figure 2B,C; Table S3). Among the DAPs identified above, 335 DAPs were exclusively identified in drought-tolerant variety XN979, but not in drought-sensitive variety XN979 ( Figure 2B,C; Table S4). These 335 DAPs may be involved in the conferring drought stress tolerance in wheat.

Protein Identification and DAPs Analysis
An iTRAQ-based comparative proteomic analysis was performed to identify wheat drought stress tolerance-related proteins. A total of 227076 spectra were generated and 38070 were matched to known spectra. These identified spectra were assigned to 17416 peptides, with 12126 unique peptides. Ultimately, 5369 proteins were identified ( Figure 2A, Table S1). In the XN979_NS-XN979_DS comparison, a total of 482 proteins showed more than a 1.2-fold change (P < 0.05) in their respective abundances and were classified as DAPs (Table S2). Among the DAPs, 199 proteins were up-regulated and 283 proteins were down-regulated in XN979_DS compared to those in XN979_NS ( Figure 2B,C). A comparison of LA379_NS and LA379_DS enabled the identification of 600 DAPs; these comprised 301 up-regulated and 299 down-regulated proteins ( Figure 2B,C; Table  S3). Among the DAPs identified above, 335 DAPs were exclusively identified in drought-tolerant variety XN979, but not in drought-sensitive variety XN979 ( Figure 2B,C; Table S4). These 335 DAPs may be involved in the conferring drought stress tolerance in wheat.

Pathway Analysis of the DAPs Involved in Drought Stress Tolerance
To explore the metabolic pathways that the present DAPs were involved in, the 335 proteins were further investigated using the web-based tool KOBAS 3.0. In total, 12 pathways were significantly enriched in the 335 DAPs (Table S5, Figure 2D). The top four significantly enriched pathways were photosynthesis (24.7%), carbon fixation (17.4%), glyoxylate and dicarboxylate metabolism (14.9%), and photosynthesis-antenna proteins (13.6%). Other pathways, such as carbon

Pathway Analysis of the DAPs Involved in Drought Stress Tolerance
To explore the metabolic pathways that the present DAPs were involved in, the 335 proteins were further investigated using the web-based tool KOBAS 3.0. In total, 12 pathways were significantly enriched in the 335 DAPs (Table S5, Figure 2D). The top four significantly enriched pathways were photosynthesis (24.7%), carbon fixation (17.4%), glyoxylate and dicarboxylate metabolism (14.9%), and photosynthesis-antenna proteins (13.6%). Other pathways, such as carbon metabolism, oxidative phosphorylation, and oxidative phosphorylation, may also be involved in drought stress tolerance (Table S5, Figure 2D).

Real-Time PCR Verification
To verify the iTRAQ data and investigate the correlation of the abundance of proteins with their corresponding mRNA level, we randomly selected 4 XN979-specific DAPs (B1P766, W5DYH0, W5ERN8, and W5EI90), 5 LA379-specific DAPs (A0MA43, M8BCN0, M8BCR3, M8BTH4, and T1N5G8), and 3 common responsive proteins in both varieties (W5H6J0, V9QGR5, and Q8VYX1) for analysis of their RNA levels. TaActin and TaGAPDH were used as reference genes to normalize the expression level of target genes, respectively. Results showed that the expression levels of most DAPs were consistent with their protein expression levels ( Figure 3). In most cases, the results of verification of expression using two different reference genes (TaGAPDH and TaActin) were similar. Only a few DAPs exhibited poor correlation between mRNA and protein expression levels ( Figure 3).

Real-Time PCR Verification
To verify the iTRAQ data and investigate the correlation of the abundance of proteins with their corresponding mRNA level, we randomly selected 4 XN979-specific DAPs (B1P766, W5DYH0, W5ERN8, and W5EI90), 5 LA379-specific DAPs (A0MA43, M8BCN0, M8BCR3, M8BTH4, and T1N5G8), and 3 common responsive proteins in both varieties (W5H6J0, V9QGR5, and Q8VYX1) for analysis of their RNA levels. TaActin and TaGAPDH were used as reference genes to normalize the expression level of target genes, respectively. Results showed that the expression levels of most DAPs were consistent with their protein expression levels ( Figure 3). In most cases, the results of verification of expression using two different reference genes (TaGAPDH and TaActin) were similar. Only a few DAPs exhibited poor correlation between mRNA and protein expression levels ( Figure  3). Relative mRNA expression analysis of twelve differentially abundant proteins (DAPs) using real-time PCR under drought stress (DS) and no stress (NS) conditions. Each bar shows the mean ± standard errors (SE) of three biological replicates. Two independent trials were conducted using TaGAPDH (GAPDH) and TaActin (ACT) as reference genes, respectively. The relative expression levels of each gene were calculated using the formula 2 −∆∆Ct (* P ≤ 0.05; ** P ≤ 0.01, Duncan's multiple range test). Relative mRNA expression analysis of twelve differentially abundant proteins (DAPs) using real-time PCR under drought stress (DS) and no stress (NS) conditions. Each bar shows the mean ± standard errors (SE) of three biological replicates. Two independent trials were conducted using TaGAPDH (GAPDH) and TaActin (ACT) as reference genes, respectively. The relative expression levels of each gene were calculated using the formula 2 −∆∆Ct (* P ≤ 0.05; ** P ≤ 0.01, Duncan's multiple range test).

VIGS of TaDrSR1 and TaDrSR2
As mentioned above, the DAPs that were exclusively identified in drought tolerance variety XN979 are most likely involved in conferring drought stress tolerance. Therefore, we selected two functional uncharacterized proteins (W5DYH0 and W5ERN8) from the four qRT-PCR verified, XN979-specific DAPs (B1P766, W5DYH0, W5ERN8, and W5EI90) for further functional analysis using VIGS technology. Here, the drought stress-tolerant variety XN979 and another Chinese wheat variety ZM9023 were used for viral infection in the VIGS experiment independently. Firstly, we investigated the expression levels of TaDrSR1 and TaDrSR2 in four independent BSMV TaDrSR1 -and BSMV TaDrSR2 -infected plants; results showed that the transcription levels of the TaDrSR1 and TaDrSR2 genes were significantly knocked down compared to those of the negative control (BSMV 0 ) and DS plants (Figure 4). As mentioned above, the DAPs that were exclusively identified in drought tolerance variety XN979 are most likely involved in conferring drought stress tolerance. Therefore, we selected two functional uncharacterized proteins (W5DYH0 and W5ERN8) from the four qRT-PCR verified, XN979-specific DAPs (B1P766, W5DYH0, W5ERN8, and W5EI90) for further functional analysis using VIGS technology. Here, the drought stress-tolerant variety XN979 and another Chinese wheat variety ZM9023 were used for viral infection in the VIGS experiment independently. Firstly, we investigated the expression levels of TaDrSR1 and TaDrSR2 in four independent BSMVTaDrSR1and BSMVTaDrSR2-infected plants; results showed that the transcription levels of the TaDrSR1 and TaDrSR2 genes were significantly knocked down compared to those of the negative control (BSMV0) and DS plants ( Figure 4). Relative expression In XN979, ten days after infection, slight chlorosis was observed in all the BSMV construct-infected plants; this was attributed to the plant's response to viral infection. The BSMVPDS-infected plants (positive control) exhibited bleached leaves ( Figure 5A), indicating that the infection was successful [44]. Twenty days after infection, a substantial level of leaf drooping and wilting was observed in both BSMVTaDrSR1-and BSMVTaDrSR2-infected plants. However, no significant sagging and withering of the leaves was observed in plants infected with BSMV0 (negative control) In XN979, ten days after infection, slight chlorosis was observed in all the BSMV construct-infected plants; this was attributed to the plant's response to viral infection. The BSMV PDS -infected plants (positive control) exhibited bleached leaves ( Figure 5A), indicating that the infection was successful [44]. Twenty days after infection, a substantial level of leaf drooping and wilting was observed in both BSMV TaDrSR1 -and BSMV TaDrSR2 -infected plants. However, no significant sagging and withering of the leaves was observed in plants infected with BSMV 0 (negative control) and DS plants ( Figure 5B-F). The leaf RWC in both BSMV TaDrSR1 and BSMV TaDrSR2 -infected plants reduced by 59.35% and 50.93% compared to that under NS but only reduced by 16.43% and 12.68% in the DS and BSMV 0 -infected plants ( Figure 6A). The relative electrolyte leakage rate in both BSMV TaDrSR1 -and BSMV TaDrSR2 -infected plants increased by 371.75% and 322.62% compared to that under NS, respectively, which was much higher than that in the DS and BSMV 0 -infected plants ( Figure 6B). MDA content in both BSMV TaDrSR1 -and BSMV TaDrSR2 -infected plants increased by 314.08% and 277.34% compared to that under NS respectively, which was significantly higher than that in the DS and BSMV 0 -infected plants ( Figure 6C). Proline content in both BSMV TaDrSR1 -and BSMV TaDrSR2 -infected plants increased by 114.89% and 98.43% compared to that under NS, respectively, which was significantly lower than that in the DS and BSMV 0 -infected plants ( Figure 6D). Similar results were obtained for ZM9023 ( Figure 5G-K and Figure 6E-H). under NS, respectively, which was much higher than that in the DS and BSMV0-infected plants ( Figure 6B). MDA content in both BSMVTaDrSR1-and BSMVTaDrSR2-infected plants increased by 314.08% and 277.34% compared to that under NS respectively, which was significantly higher than that in the DS and BSMV0-infected plants ( Figure 6C). Proline content in both BSMVTaDrSR1-and BSMVTaDrSR2-infected plants increased by 114.89% and 98.43% compared to that under NS, respectively, which was significantly lower than that in the DS and BSMV0-infected plants ( Figure  6D). Similar results were obtained for ZM9023 ( Figures 5G-K and 6E-H).

Discussion
Drought stress occurs frequently worldwide and is a major limiting factor for plant growth and productivity. It has been reported that wheat can adapt in response to drought stress and exhibit significant differences in drought tolerance among different genotypes [45]. The elucidation of the molecular mechanisms and investigation of drought tolerance-related genes/proteins is crucial for the genetic improvement of plant drought tolerance.
Comparative proteomic analysis is a powerful tool for the study of plant stress response [31,46]. The changes in protein expression profiles under drought conditions have been investigated in several plants, including wheat, maize, rice, peanut, and soybean, and many drought-responsive proteins have been characterized [32][33][34][35][36][37]. However, despite this progress, most of these proteins have not been functionally verified. The molecular mechanism underlying plant drought tolerance is still far from fully elucidated, especially in hexaploid wheat. Here, a comparative proteomic analysis of two wheat varieties with contrasting drought tolerance enabled the identification of 335 DAPs specific to the drought-tolerant variety XN979, and 147 common responsive DAPs in both varieties (Table S4, Figure 2B). The XN979-specific drought-responsive proteins may participate in pathways mediating drought stress tolerance in wheat. Pathway enrichment analysis indicated that most of these proteins were involved in photosynthesis, carbon fixation, glyoxylate and dicarboxylate metabolism, and some other pathways ( Figure 2D). In our study, 19 DAPs involved in the photosynthesis pathway were significantly enriched in drought-tolerant variety XN979. Previous studies have shown that the photosynthesis pathway is indeed affected by drought stress. The photosynthetic rate and the assimilation product are reduced, thereby further reducing the material

Discussion
Drought stress occurs frequently worldwide and is a major limiting factor for plant growth and productivity. It has been reported that wheat can adapt in response to drought stress and exhibit significant differences in drought tolerance among different genotypes [45]. The elucidation of the molecular mechanisms and investigation of drought tolerance-related genes/proteins is crucial for the genetic improvement of plant drought tolerance.
Comparative proteomic analysis is a powerful tool for the study of plant stress response [31,46]. The changes in protein expression profiles under drought conditions have been investigated in several plants, including wheat, maize, rice, peanut, and soybean, and many drought-responsive proteins have been characterized [32][33][34][35][36][37]. However, despite this progress, most of these proteins have not been functionally verified. The molecular mechanism underlying plant drought tolerance is still far from fully elucidated, especially in hexaploid wheat. Here, a comparative proteomic analysis of two wheat varieties with contrasting drought tolerance enabled the identification of 335 DAPs specific to the drought-tolerant variety XN979, and 147 common responsive DAPs in both varieties (Table S4, Figure 2B). The XN979-specific drought-responsive proteins may participate in pathways mediating drought stress tolerance in wheat. Pathway enrichment analysis indicated that most of these proteins were involved in photosynthesis, carbon fixation, glyoxylate and dicarboxylate metabolism, and some other pathways ( Figure 2D). In our study, 19 DAPs involved in the photosynthesis pathway were significantly enriched in drought-tolerant variety XN979. Previous studies have shown that the photosynthesis pathway is indeed affected by drought stress. The photosynthetic rate and the assimilation product are reduced, thereby further reducing the material basis of leaf growth [26,[47][48][49]. The photosynthetic rate and the chlorophyll a content in a drought-tolerant variety SR3 were higher than in the leaves of its parent JN177. This higher capacity to maintain photosynthesis under stress may be achieved through a combination of a more robust cellular homeostasis and a more effective means of removing ROS and other toxic by-products [26]. Therefore, the 19 DAPs involved in the photosynthesis pathway may play important roles in drought tolerance in wheat. Other pathways, such as carbon fixation and photosynthesis-antenna proteins, are also associated with plant photosynthesis [50]. The other enriched pathways, such as carbon metabolism, pyruvate metabolism, amino acids metabolism, and oxidative phosphorylation, are mainly involved in the metabolism and allocation of carbohydrates. These results indicate that the regulation of the photosystem and carbohydrates metabolism play important roles in wheat drought tolerance.
Two DAPs (W5DYH0 and W5ERN8) were selected and dubbed TaDrSR1 and TaDrSR2, respectively, for further functional analysis using VIGS technology. The ortholog of TaDrSR1 in rice, r40c1, is up-accumulated in the roots of DREB1A transgenic plants, which may play an important role in the generation of drought-tolerant plants [51,52]. TaDrSR2 encodes a functional unknown protein with 71 amino acid residues. Therefore, we know very little about the functions of both genes. We analyzed the chromosomal locations of TaDrSR1 and TaDrSR2 according to Yang et al. [53] and found that TaDrSR1 and TaDrSR2 were anchored on chromosome 4A and 2D, respectively. Interestingly, the chromosomal regions also have a number of QTLs for agronomic traits under drought stress conditions [54][55][56]. For example, a QTL for the root number under water-limited environments, qRN.qgw-4A, is linked with the TaDrSR1 gene [54]. This chromosomal region also harbors QTLs for the drought stress susceptibility index (SSI) and drought stress tolerance (TOL) [56]. TaDrSR2 is also linked with QTLs for plant height, spike length, spikelets per spike, and kernels per spike under drought stress conditions [55]. This evidence further confirms that the chromosomal regions in which TaDrST1 and TaDrST2 are located are important for wheat drought stress tolerance. As RWC, relative conductivity, MDA, and proline content are important physiological indicators reflecting plant drought tolerance [14,[57][58][59], we measured these indices in the TaDrSR1and TaDrSR2-knock-down plants and negative controls. Our results showed that the relative electrolyte leakage rate and MDA content significantly increased, while the RWC and proline content significantly decreased in the VIGS-mediated TaDrSR1and TaDrSR2-knock-down plants compared to that in the controls (BSMV 0 and DS) under drought stress conditions ( Figure 6). The TaDrSR1and TaDrSR2-knock-down plants exhibited severe drooping and wilting phenotypes relative to the negative control (BSMV 0 -infected plants) and non-infected plants under drought stress conditions, indicating that they were more sensitive to drought stress ( Figure 5B-F). We performed the VIGS trial using another variety (ZM9023) as receptors for viral infection and obtained similar results ( Figure 5G-K and Figure 6E-H). Previous studies have shown that the higher the MDA content of plants, the greater the level of membranous peroxidation and permeability. Further, the cytoplasm undergoes extravasation under drought stress, resulting in an increase in relative conductivity [14]. Under drought stress, proline reduces the osmotic potential, plays a role in osmotic regulation, and protects cell membrane structure; these changes represent an adaptive response of plants to adverse environments [60][61][62]. Our results, together with those of previous studies, strongly suggest that TaDrSR1 and TaDrSR2 potentially play vital roles in conferring drought tolerance in common wheat.
A comparative proteomic analysis approach can identify hundreds of wheat drought-tolerant protein candidates. For example, in this study, we identified 335 drought tolerance-related proteins, which were involved in 12 pathways ( Figure 2D; Table S5). Some of them may play important roles, while others may only serve as an aid. The verification the function of these proteins is a big challenge in hexaploid wheat. VIGS has the advantages of simple operation and short test period [39,40]. It is suitable for preliminary functional verification on a large number of candidates. Therefore, comparative proteomic analysis combined with VIGS is a good choice for identifying novel genes for drought tolerance in wheat, although subsequent transgenic validation is necessary.
In summary, we identified 335 DAPs involved in wheat drought tolerance. Most of them were involved in the pathways of photosynthesis, carbon fixation, glyoxylate and dicarboxylate metabolism, carbon metabolism, and oxidative phosphorylation. Two DAPs TaDrSR1 and TaDrSR2 were selected for further functional analysis using VIGS. The TaDrSR1and TaDrSR2-knock-down plants exhibited severe drooping and wilting phenotypes under drought stress conditions. Further physiological and molecular analyses indicated that these knock-down plants were more sensitive to drought stress, suggesting that TaDrSR1 and TaDrSR2 potentially play vital roles in conferring drought tolerance in common wheat. Our results also showed that comparative proteomic analysis combined with VIGS is an efficient way for identification of novel drought tolerance-related proteins in hexaploid wheat.

Plant Materials
XN979 and LA379, two wheat varieties with contrasting drought tolerance, were used as materials to conduct comparative proteomic analysis. XN979 exhibits far higher tolerance to drought stress than LA379. The drought stress-tolerant variety XN979 was used for viral infection in the VIGS experiment. A previous study reported that ZM9023 is susceptible to viral infection in VIGS trials [39]. To further confirm the results of VIGS in XN979, ZM9023 was also selected as a receptor for viral infection in this study.

Plant Growth Conditions and Sampling
Pot culture was used in this trial. Firstly, seeds were disinfected with 1% H 2 O 2 for 10 min and then rinsed with distilled water for three times. Before sowing, substrate (Pindstrup Substrate Peat) and vermiculite were well mixed (volume ratio 1:2). Seeds were germinated for 16 h at 22 • C; then, 24 germinated seeds were sown in each pot with soil water content at 90% field capacity (FC). The pots were divided into two groups for two watering treatments, respectively: (i) A group of 20 pots (10 pots per cultivar) was grown under well-watered conditions (maintained at 80-90% FC, NS), and (ii) another 20 pots (10 pots per cultivar) were subjected to drought stress (DS, no watering after sowing). The temperature ranged from 22-23 • C in the daytime and 18-20 • C at night during the pot culture trial. Wheat plants were thinned to 20 plants per pot after emergence. Sixteen days after sowing (about 44% FC under DS conditions), fresh leaves of XN979 and LA379 under NS and DS conditions were sampled, respectively, and used for the measurement of physiological indices (proline, MDA content, and electrolyte leakage), with at least three biological replicates. The leaves of fifteen individual plants of XN979 and LA379 under NS and DS conditions were mixed respectively, fast-frozen in liquid nitrogen, and stored in a −80 • C freezer for protein extraction. This collection was repeated for the second biological replicate. The samples of two biological replicates of XN979 and LA379 under NS and DS conditions were named as XN979_NS-1, XN979_NS-2, LA379_NS-1, LA379_NS-2, XN979_DS-1, XN979_DS-2, LA379_DS-1, and LA379_DS-2, respectively. For RNA extraction and real-time PCR verification, three biological replicates were analyzed. Twenty days after sowing (about 40% FC under DS conditions), the leaves were collected for measurement of RWC.

The Measurement of Physiological Indices
The RWC, which was measured as described by Flexas et al. [63], was calculated based on the following formula: RWC (%) = [(FW − DW)/ (TW − DW)] × 100; FW represents fresh weight, TW refers to turgid weight, and DW represents dry weight. Electrolyte leakage was assayed according to Yan et al. [43]. Proline was extracted and determined according to the method of Bates et al. [64]. MDA content was measured following the methods of Hodgeset al. [65].

Strong Cationic Exchange (SCX) Fractionation and LC-ESI-MS/MS Analysis
iTRAQ labeled peptides were combined and dried under vacuum. Strong cationic exchange (SCX) chromatography was performed to fractionate the labeled peptides using the AKTA Purifier system (GE Healthcare) as previously described, with minor modifications [68,69]. In brief, the labeled peptide mixture was acidified using 10 mM KH 2 PO 4 in 25% of ACN (buffer A, pH 3.0) and eluted at a flow rate of 1 mL/min with a gradient of 0-8% 500 mM KCl, 10 mM KH 2 PO 4 in 25% of ACN (buffer B, pH 3.0) for 22 min, 8-52% buffer B for 22-47 min, 52%-100% buffer B for 47-50 min, 100% buffer B for 50-58 min, and buffer B was reset to 0% after 58 min. The elution was monitored by absorbance at 214 nm, and fractions were collected every 1 min. Samples were reconstituted with trifluoroacetic acid and stored at −80 •C for LC-MS/MS analysis.
Liquid chromatography-electrospray ionization tandem MS analysis was performed using a Q Exactive mass spectrometer coupled to an Easy nLC (Proxeon Biosystems) according to previously reported literature with minor modifications [68,69]. The peptide mixture was loaded onto a reverse phase trap column connected to the C18-reversed phase analytical column in 0.1% formic acid (buffer A) and separated with a linear gradient of 0.1% formic acid and 84% acetonitrile (buffer B) with a 300 nL/min flow rate. The linear gradient was 0-35% buffer B for 0-50 min, 35-100% buffer B for 5 min, and hold in 100% buffer B for another 5 min. The most abundant precursor ions were dynamically selected from the survey scan (300-1800 m/z) to acquire MS data. The instrument was run with the peptide recognition mode enabled [68,69].

Data Analysis
MS/MS spectra were searched using MASCOT engine (Matrix Science, version 2.2) embedded into Proteome Discoverer 1.4 and run against the UniProt_Poaceae database and the decoy database (http://www.uniprot.org).
The Mascot search parameters were set according to previous studies [68,69]. Mascot search parameters were set as follows: Enzyme: Trypsin; max missed cleavage: 2; fixed modification: Carbamidomethyl (C), iTRAQ8plex(N-term), iTRAQ8plex(K); variable modification: Oxidation (M), iTRAQ8plex (Y); peptide mass tolerance: ± 20 ppm; fragment mass tolerance: 0.1 Da; peptide false discovery rate (FDR) ≤ 1%. All peptide ratios were normalized by the median protein ratio. The abundance of each protein was calculated as the median of unique peptides of the protein, and the fold change was defined based on the abundance of the protein under drought relative to their respective level under control. Protein species with an abundance ratio fold change of at least 1.2 and a P-value < 0.05 were defined as DAPs [70,71].

Pathway Enrichment Analysis
Pathway enrichment analysis was performed using a web-based tool KOBAS 3.0 (http://koba s.cbi.pku.edu.cn/anno_iden.php) [72]. The adjusted P-values with Benjamini-Hochberg correction under a threshold of 0.05 were considered to represent statistically significant differences.

Quantitative Real-Time PCR
Total RNA from the leaves was extracted using the total RNA kit (TaKaRa, Dalian, China). The Two-Step Prime-Script TM RT Reagent Kit (Perfect Real Time; TaKaRa) with gDNA Eraser was used for the reverse transcription reactions following the manufacturer's instructions. Twelve DAPs were randomly selected for RNA-level examination. The specific primers were designed using Primer 5.0 (Premier Biosoft, Palo Alto, CA, USA). The cDNA samples were used as templates and mixed with primers and SYBR Premix ExTaq (TaKaRa) for real-time PCR analysis. Real-time PCR was conducted using a BioRad IQ5 Real-time PCR Detection System (Bio-Rad, Hercules, CA, USA). The temperature settings were 95 • C for 5 min followed by 40 cycles of 95 • C for 15 s, 60 • C for 15 s, and 72 • C for 15 s.
TaActin and TaGAPDH were used as reference genes to normalize the expression level of target genes, respectively. Relative gene expression was computed using the 2 −∆∆Ct method [73]. All primers used in this study are listed in Table S6.

Vector Construction for VIGS
To further study the functions of DAPs, we constructed VIGS vectors of TaDrSR1 and TaDrSR2. Firstly, 118 bp-and 107 bp-fragments of the coding regions of TaDrSR1 and TaDrSR2 were cloned, respectively. Vector construction was performed as previously described [74]. Linearized plasmids were used as templates to synthesize α, β, and γ RNAs of the BSMV genome with the RiboMAX TM Large Scale RNA Production System-T7 (Promega, Madison, WI, USA) [75]. In vitro transcripts of each RNA fragment were mixed in an equimolar ratio and added to triturated FES buffer [40]. Then, the mixture was diluted and treated with an equal volume of DEPC. Each of the silencing constructs consisted of BSMV α, β, and γ with the target gene insertion. The original BSMV 0 was constructed from α, β, and γ derived from the original empty pSL038-1 vector and used as the negative control. BSMV PDS (in which the wheat gene encoding phytoene desaturase, GenBank: FJ517553.1, was silenced) was used as a positive control of VIGS [41]. BSMV TaDrSR1 was constructed from α, β, and γ with the insertion of the target gene TaDrSR1. BSMV TaDrSR2 was constructed from α, β, and γ with the target gene TaDrSR2's insertion.

Infection with VIGS Vectors
The drought stress-tolerant variety XN979 and virus-susceptible variety ZM9023 were selected as receptors for infection with VIGS vectors. Two-leaf-stage plants of XN979 and ZM9023 were used for infection, which was performed according to previously described procedures [40]. After the infection, the incubator temperature was set at 23 ± 2 • C, with darkness for 24 h, followed by a 16 h light/8 h dark photoperiod. Twenty days after the infection, the third and fourth leaves were collected and stored at −80 • C for measurement of physiological indices (proline and MDA content) and real-time PCR analysis. Twenty-four days after the infection, the leaves of the remaining plants were collected for the measurement of RWC and rate of relative electrolyte leakage.

Data Analysis
The means, standard errors (SE), and ranges of each measured morphological trait were analyzed using IBM SPSS statistics 21 software. For each morphological trait and gene expression level, P = 0.05 and P = 0.01 were used as thresholds to identify significant and extremely significant differences, respectively (Duncan's multiple range test). Venny 2.1 (a web-based tool) was used to generate Venn diagrams (http://bioinfogp.cnb.csic.es/tools/venny/index.html).