Evaluation of Drought Stress-Inducible W<i>si</i>18 Promoter in <i>Brachypodium distachyon</i>

The rice Wsi18 promoter confers drought-inducible gene expression. This property makes it a useful candidate to drive relevant genes for developing drought resistant traits for different monocot crops. In this study, we showed that the Bradi2G47700 gene, the closest homologue to rice Wsi18, was upregulated in Brachypodium distachyon plants exposed to ABA and mannitol. Wsi18: uidA transgenic B. distachyon plants were produced and then subjected to ABA or mannitol treatment. The expression of uidA in three transgenic lines (line 10, 18 and 37) was significantly upregulated in plants exposed to ABA (fold increases of 5.61 ± 0.98, 2.88 ± 0.75 and 9.13 ± 1.96, respectively) compared to the same transgenic plant lines without treatment. The expression of uidA in two transgenic lines (lines 18 and 37) also showed upregulation when treated with mannitol (fold increases of 4.43 ± 1.07 and 8.47 ± 2.90, respectively) compared to the same transgenic plant lines without mannitol treatment. Moreover, GUS histochemical assay showed increased Wsi18 promoter activity in the leaves and stems of transgenic lines upon treatment with ABA or mannitol. This is the first report of the drought inducible rice Wsi18 promoter being active in B. distachyon which is a model plant for molecular biology research of various monocot plants. Taken together, the results indicate that the Wsi18 promoter and its homologue may be explored as a useful tool for drought stress-inducible gene expression in different monocot crops.


Introduction
Abiotic stresses are a seriously threats to crop production. Each year, more than 50% of the global major crop yield is lost due to abiotic stress, with water deficit stresses such as drought and high salinity being major contributors [1]. It is estimated that by the year 2050, the global demand for crop derived calories will have increased by 100%, and thus developing crop varieties with greater tolerance to abiotic stresses will be essential to meeting these needs [2].
The introduction of relevant genes to crops via genetic engineering is an important approach for developing stress tolerance for plants. Several constitutive promoters have been previously used for such purposes, such as Ubi1 from maize, Act1, Rbcs and OsCc1 from rice [3] [4] [5] [6]. However, transgenic plants with strong constitutive promoters often suffer undesirable phenotypes. For example, transgenic rice constitutively over-expressing Ubil: OsNAC6 suffers from growth retardation and low reproductive yields [7]. Transgenic tobacco plants constitutively over-expressing 35S: TPS1 have greater drought tolerance, but suffer stunted growth [8]. This could be due to the cost of the resources needed to constantly overexpress the transgene, or negative interactions with normal cell metabolism [9]. Stress-inducible promoters offer the advantage of expressing the gene only under stress conditions. Thus, plants with stress-inducible promoters could eliminate undesirable phenotypes induced by constitutive transgene expression [10] [11].
Only a few drought inducible promoters have been identified in plants so far.
Wsi18 is a drought-inducible rice gene from the group 3 late embryogenesis abundant (LEA3) family [12]. The promoter has been analyzed in transgenic rice [13] [14] [15] and under normal conditions, the basal expression of Wsi18 was low in transgenic rice tissues, but following drought, NaCl or ABA treatment, the expression of genes driven by Wsi18 was induced in the whole plant body [15]. Following drought conditions or exposure to ABA, Wsi18 had a level of transient expression comparable to that displayed by the strong constitutive promoter Act1 [16]. Despite the research conducted on Wsi18 in rice, its stable expression profile has never been studied in other plant species. Furthermore, it is not clear if the drought-inducible property can function in other monocot crops, and more importantly, be used to drive expression of suitable genes to develop drought tolerance in different monocot crops. In this study, we identified the Bradi2G47700 gene which is the closest homologue to the rice Wsi18 gene and found this gene was upregulated in B. distachyon plants upon exposure to ABA or mannitol. Furthermore, we show that the Wsi18 promoter drives high levels of uidA expression in transgenic B. distachyon plants under ABA and mannitol treatments. Our results demonstrate that the Wsi18 promoter has drought inducible activity in transgenic B. distachyon plants, which is a model research plant closely related to many important monocot crops, and that the Wsi18 promoter might be an effective promoter to drive relevant genes for drought resistance in different monocot crops.

Vector Construction
The Wsi18 promoter region was amplified by PCR using the pWsi18 plasmid as the template (obtained from Ju-Kon Kim at Seoul National University) and the primers of FWsi18 and RWsi18 (Table 1). The 2 × CaMV35S promoter region was amplified by PCR using the pMDC32 plasmid [18] and the primers of F2xCaMV35S and R2xCaMV35S with additional attB sites ( Table 1). The PCR products of the Wsi18 promoter and the 2 × CaMV35S promoter were separately inserted into the pMDC163 plasmid using Gateway® Technology (Thermo Fisher Scientific) [19], generating the constructs of Wsi18:: uidA and 2 × CaMV35S:: uidA respectively. The Wsi18:: uidA and 2 × CaMV35S:: uidA plasmids were separately introduced into Agrobacterium tumefaciens strain AGL1 via electroporation.

Relative Water Content
Relative water content (RWC) was measured for each experimental group to gauge the amount of water loss created by each stress treatment [21]. All of the leaves from two non-transformed plants were collected, and weighed immediately after plants were removed from hydroponic growth medium to obtain the fresh weight (Wfresh). The leaves were floated on ddH 2 O in petri dishes for 6 hours, blotted dry, and reweighed to give the turgid weight (Wturgid). Leaves were then placed in an oven at 50˚C for 48 hours and reweighed to give the dry weight (Wdry). RWC was calculated using the equation: RWC = (Wfresh − Wdry)/(Wturgid-Wdry) × 100.

GUS Histochemical Analysis
GUS histochemical staining was performed as described by Jefferson et al. [23].
Briefly, shoots with leaves of plants in each treatment group were cut off and immediately submerged in GUS staining solution (2 mM 5-bromo-4-chloro-3-indolyl glucuronide (X-gluc), 0.1 M NaPO 4 (pH 7.0), 10 mM EDTA, 1% triton X-100, and 1 mM K 3 Fe(CN) 6 ). X-gluc is a substrate of the β-glucuronidase enzyme, the product of the uidA gene. Each sample was vacuum infiltrated for 30 minutes, then incubated overnight at 37˚C. Samples were submerged in 95% ethanol for 48 hours to remove chlorophyll and make the staining easier to observe.

Statistical Analysis
All statistical analyses were performed using the program "R" version 3.1.3 Copyright© 2015 (The R Foundation for Statistical Computing). RWC measurements were expressed as mean ± standard error of 9 -10 biological replicates for each stress-treated and unstressed control group. The statistical difference between each stress and its corresponding unstressed control group was assessed using a Welch's two sample t-test. Significance was established at p < 0.05. All RT-qPCR results were expressed as the mean ± standard error of 3 biological replicates in each stress treatment group, and 3 biological replicates in each unstressed control group of each transgenic line. The statistical difference between the fold change of each stress treatment group and their corresponding unstressed control group was assessed using a Welch's two sample t-test. Significance was established at p < 0.05.

RWC Decreases in B. distachyon Plants Subjected to Mannitol Treatment
In order to determine the effectiveness of drought treatment, plant relative water  (Figure 1(a)). This was predictable as ABA itself should not cause drought condition, rather, ABA only functions to mediate plant response to drought. On the other hand, the plants grown in hydroponic growth medium containing mannitol had a RWC of 64.81% ± 4.63%, a significant decrease compared to control plants which had a RWC of 90.99% ± 3.87% (Figure 1(b)).
Thus, mannitol treatment was effective to induce water deficit condition in B. distachyon as observed in some other plants such as Sesuvium portulacastrum [24], tobacco [25] and wheat [26].

Bradi2G47700 Gene Expression Is Upregulated in B. distachyon Plants Under Drought Stress
The plant genome database http://www.phytozome.net/ was used to identify possible Wsi18 homologous genes native to B. distachyon. Putative homologues were identified based on amino acid sequence similarity to that of the Wsi18 gene of rice rather than the promoter DNA sequence because coding regions are usually more conserved than non-coding regions [27]. The Bradi2G47700 amino acid sequence was identified as the most similar protein to Wsi18 in B. distachyon and shares 74% amino acid sequence similarity with Wsi18 ( Figure 2). To investigate the expression of the Bradi2G47700 gene in B. distachyon under drought stresses, the non-transformed B. distachyon plants were subjected to ABA and mannitol treatments, and the Bradi2G47700 gene expression was analyzed via RT-qPCR at 8 hours and 18 hours after treatment, respectively. The SamDC gene, which has been shown to express stably under water deficit conditions [22], was used as a reference gene for RT-qPCR to normalize Bradi2G47700 transcript levels between samples. The normalized level of Bradi2G47700 transcripts in stress-treated plants were compared to the transcript level of unstressed  Bradi2G47700 was identified as a putative Wsi18 homologue using www.phytozome.net [30], and the sequence alignment was generated using Clustal Omega [31] [32] [33]. Bra-di2G47700 and Wsi18 proteins share 74% sequence similarity.
control plants to calculate the relative fold change. The Bradi2G47700 gene expression in ABA treated plants was upregulated 794.14 ± 504.91 (P = 0.01914) fold on average compared to the non-stressed control group (Figure 3(a)). The Bradi2G47700gene expression in mannitol treated plants resulted in 101.85 ± 8.17 (P < 0.01) fold increase on average as compared to the non-stressed control group (Figure 3(b)). Yi et al. [15] reported that Wsi18 mRNA abundance increased in leaf tissues or roots exposure to ABA. Interestingly, Joshee et al. [12] reported that the Wsi18 gene can be induced by water stress conditions such as mannitol and NaCl, but not by ABA. Probably Joshee et al. [12] used low dose of ABA (20 M). The dose of ABA as Yi et al. reported (100 M) was five times higher. Taken together, the upregulation of the Bradi2G47700 gene after ABA and mannitol treatments indicates that the Bradi2G47700 gene in B. distachyon is water stress inducible similar to Wsi18. As Bradi2G47700 shares a high level of similarity with Wsi18 and is also inducible for expression under drought treatments, Bradi2G47700 is probably the true homologue of Wsi18. The results indicate that Wsi18 homologues in other monocot crops may be explored for drought stress tolerance development for respective and specific situations.  to kill non-transgenic plants (Supplemental Figure S2). All the lines which were PCR positive germinated normally on medium containing 115 mg/L hygromycin. On the other hand, the plants which did not produce the correct 628 bp product in PCR genotyping either did not germinate in the presence of hygromycin B, or produced only black roots (not shown). PCR genotyping and hygromycin B seed selection provided two levels of plant transformation analyses and only the plants that were PCR positive and showed normal germination and growth under a high level (115 mg/L) of hygromycin B selection were used to evaluate Wsi18 expression. Real-time PCR analysis was used to examine the expression of the uidA reporter gene, which was driven by the Wsi18 promoter, in transgenic lines of Wsi18: uidA following ABA or mannitol treatment. The SamDC gene was used as a reference gene for RT-qPCR to normalize uidA gene expression between samples. Normalized uidA expression levels within each line were compared to the expression level of non-drought-treated control plants of the same transgenic line to determine their relative fold change. Plants that were grown in fresh hydroponic growth medium for the same duration as their respective stress treatment were used as a control. Of the five transgenic lines, the uidA expression in three lines (lines 10, 18 and 37) was significantly upregulated after ABA treatment. Lines 10, 18 and 37 showed fold increases of 5.61 ± 0.98 (P < 0.01), 2.88 ± 0.75 (P < 0.05) and 9.13 ± 1.96 (P < 0.05), respectively (Figure 4(a)). The uidA expression in the other two lines (lines 3 and 12) did not display a significant increase in expression (Figure 4(a)). For mannitol treatment, the uidA expression in two lines (lines 18   and 27 showed an increased upregulation of 2.16 ± 1.27 (P = 0.2298), 6.10 ± 3.52 (P = 0.0938) and 2.96 ± 0.90 (P = 0.0659) fold respectively, there was no significant difference between the control and treated groups (Figure 4(b)). Variation of transgene expression in different transgenic plants has been frequently reported for both constitutive and inducible promoters before [14] [28]. One explanation for the observed differences in uidA expression between the transgenic lines in this study could be the differences in the site of transgene integration into the B. distachyon genome, known as positional effects. Transcriptional regulatory sequences such as enhancers and inhibitors found around the site of integration, as well as the chromatin structure around the site of integration, could all influence the expression of the transgene. Integration at highly expressing loci will produce higher expressing transgenic lines than integration at loci with lower transcriptional activity [34]. Additionally, copy number has been shown to influence transgene expression levels. Integration of multiple transgene copies often produces low levels of expression due to homology-dependent gene silencing [29]. Homology-dependent gene silencing can be the result of transcriptional gene silencing or post-transcriptional gene silencing. Transcriptional gene silencing includes processes such as promoter methylation, which blocks transcription from occurring from the promoter, while post-transcriptional gene silencing causes transcripts to be degraded before they can be translated into protein. If Wsi18 was not water deficit-inducible in B. distachyon, no difference in uidA expression would be expected between the stress-treated and unstressed control plants of any transgenic lines. The mannitol and ABA treatment did show a significant increase in Wsi18 driven uidA expression in certain transgenic lines, suggesting Wsi18 does have water deficit inducible activity in B. distachyon. no detectable GUS activity in any of the control plants ( Figure 5(c)). There was variation in the level of GUS activity between the different transgenic lines, however, GUS activity was more prevalent following stress treatments than in unstressed control plants for each stress treatment. Some unstressed control plants did show small areas of GUS activity, but always to a lesser degree than their stress-treated counterparts. Low level Wsi18 promoter activity in unstressed control plants is consistent with the observations of Wsi18 activity in transgenic rice reported by Yi et al. [15]. The low levels of expression driven by Wsi18 in unstressed control plants may be due to less stringent regulation of gene  Histochemical GUS assays were carried out to observe the pattern of GUS expression following ABA and mannitol treatments. Our results showed that the Bradi2G47700 gene, the closest homologue to rice Wsi18, was upregulated in B. distachyon plants exposed to ABA and mannitol. Further, we showed that the Wsi18 promoter can be induced to express by ABA and the water deficit conditions created by mannitol. Response    Hygromycin B caused the seedlings either to not germinate, or to exhibit black roots and retarded growth. Germination for 12 days on medium containing 60 mg/L hygromycin B was sufficient to disrupt the growth of all wild-type seeds. Higher concentrations of hygromycin B enabled the identification of hygromycin B sensitive seeds over a shorter period of time.