Overexpression of AaWRKY1 Leads to an Enhanced Content of Artemisinin in Artemisia annua

Artemisinin is an effective component of drugs against malaria. The regulation of artemisinin biosynthesis is at the forefront of artemisinin research. Previous studies showed that AaWRKY1 can regulate the expression of ADS, which is the first key enzyme in artemisinin biosynthetic pathway. In this study, AaWRKY1 was cloned, and it activated ADSpro and CYPpro in tobacco using dual-LUC assay. To further study the function of AaWRKY1, pCAMBIA2300-AaWRKY1 construct under 35S promoter was generated. Transgenic plants containing AaWRKY1 were obtained, and four independent lines with high expression of AaWRKY1 were analyzed. The expression of ADS and CYP, the key enzymes in artemisinin biosynthetic pathway, was dramatically increased in AaWRKY1-overexpressing A. annua plants. Furthermore, the artemisinin yield increased significantly in AaWRKY1-overexpressing A. annua plants. These results showed that AaWRKY1 increased the content of artemisinin by regulating the expression of both ADS and CYP. It provides a new insight into the mechanism of regulation on artemisinin biosynthesis via transcription factors in the future.


Introduction
A. annua is an important traditional Chinese herbal plant according to the Chinese materia medica [1]. Artemisinin, extracted from herbal plants A. annua, is an effective component of drugs against malaria. It was discovered by Miller and Su [2]. Artemisinin-based combination therapy (ACT) is considered the most effective method against malaria [3]. In recent years, other functions of artemisinin have been reported. For instance, artemisinin might have effects on human cancers [4]. Artemisinin can potentially serve as a drug against hepatitis B virus [5].
Although artemisinin is widely used, the yield of artemisinin cannot meet market demand [23,24]. Plant metabolic engineering is promising to produce artemisinin. Overexpression of one or more genes in artemisinin biosynthetic pathway increases the yield of artemisinin in transgenic A. annua. Overexpression of HMGR resulted in enhancement of the artemisinin content in transgenic A. annua [25,26]. FPS overexpressing lines exhibited 3.6-fold higher content of artemisinin than wild plants [27,28]. Overexpression of DBR2 significantly improved the content of artemisinin to 1.59-2.26 times compared with that in control [29]. Overexpression of HMGR and ADS led to 7.65-fold higher content of artemisinin compared with nontransgenic plants [30]. Stacked overexpression of FPS, CYP, and CPR resulted in 3.6-fold higher content of artemisinin compared with the controls [31]. Downregulation of enzymes competitive with artemisinin biosynthesis also led to an improved yield of artemisinin. The content of artemisinin was increased by downregulation of squalene synthase (SQS), a key enzyme of sterol pathway [32,33]. The artemisinin content of transgenic A. annua by downregulation of -caryophyllene synthase (CPS) exhibited an increase of 54.9% compared with the wild plants [34].
Transcription factors are also used in plant metabolic engineering. WRKY transcription factors are one of the largest families of regulatory proteins in plants [35]. WRKY transcription factors contain their DNA binding domain, WRKY domain, which binds to W box (TTGAC[C/T]) [35][36][37]. WRKYs have many different functions, including biotic stress, abiotic stress, and trichome development [35]. One  WRKY transcription factor may regulate several different processes. For example, HvWRKY38 is involved in plant development through gibberellin signaling and abiotic stress [38][39][40]. LtWRKY21 can activate abscisic acid (ABA) signaling pathway [41]. Overexpression of OsWRKY45 improved drought tolerance of transgenic Arabidopsis [42]. Heat and drought tolerance was enhanced in transgenic rice plants by overexpressing OsWRKY11 [43]. Overexpression of AtWRKY25 or AtWRKY33 increased salt tolerance and ABA sensitivity in transgenic Arabidopsis [44]. Overexpression of AtWRKY25 also exhibited enhanced heat tolerance [45]. Three WRKY transcription factors, GmWRKY13, GmWRKY21, and GmWRKY54, exhibited differential tolerance to abiotic stresses [46]. BhWRKY1 was involved in the dehydration by binding to the promoter of galactinol synthase [47]. AtWRKY63 was involved in plant responses to  ABA and drought tolerance in transgenic Arabidopsis [48]. AtWRKY15 functioned as a negative regulator of osmotic stress responses by mitochondrial retrograde regulation [49]. Many transcription factors were found to regulate key enzymes in artemisinin biosynthetic pathway. They are also quite important in plant metabolic engineering. Both AaERF1/2 and AabZIP1 can bind with the promoter of ADS and CYP and regulate their expression [50][51][52]. TAR1 could interact with ADS and CYP and further regulate the biosynthesis of artemisinin [53]. Ma et al. found that AaWRKY1 could bind to the W boxes of ADSpro and activate the expression of ADS in transient expression systems [54].
In this study, pCAMBIA2300-AaWRKY1 fusion expression vectors were constructed under the drive of 35S promoter and transformed into A. annua. Transgenic plants overexpressing AaWRKY1 were analyzed. Overexpression of AaWRKY1 led to great enhancement of artemisinin content.

Plant Materials.
The seeds of low artemisinin-yielding A. annua were obtained from our lab. Seeds were sterilized in 75% ethanol for 2-3 min, followed by 1% sodium hypochlorite solution for 8 min, and then washed with sterilized distilled water several times. Seeds were sown on Murashige and Skoog (MS) medium under a photoperiod of 16 h light/8 h dark at 22 ± 1 ∘ C. The seedlings were transferred to the soil after 10 days in greenhouse. The young leaves of A. annua plants were collected for RNA extraction. The leaves of plants were collected for DNA extraction 2 months after being transferred to the soil in the growth chamber. Tobacco (Nicotiana benthamiana) seeds were sown on soil in pots in greenhouse. The leaves of 4-week-old tobaccos were prepared for dual-luciferase assay.  transcribed with 0.5 L 50 M oligo-dT primers. The 1ststrand cDNA was used as template for quantitative real-time RT-PCR (qPCR). Expression of AaWRKY1 and other enzymes in A. annua was analyzed by qPCR using the fluorescent intercalating dye SYBR-Green (Tiangen Biotech, Beijing) in a detection system (Opticon3, MJ Research). The qPCR was performed as previously described [21,55]. First single-stranded cDNA was used as the template in 20 L reaction mixture including 10 L SYBR Premix Ex Taq6 and 10 pmol of each primer. The primers used were listed in Table 1. The qPCR was performed first at 95 ∘ C (30 sec) with 40 cycles at 95 ∘ C (5 sec) and 56 ∘ C (30 sec) and finally a dissociation stage at 95 ∘ C (15 sec), 56 ∘ C (30 sec), and 95 ∘ C (15 sec). The data were analyzed by 2 −ΔΔCT method [56]. The transcript level of A. annua ACTIN was used as the control for qPCR analysis.

DNA Sequence
Analysis. AaWRKY1 was isolated by amplifying the total cDNA with the specific primers AaWRKY1-up and AaWRKY1-down. The acquired sequence was cloned to pJET1.2 vector according to the instructions (Invitrogen, USA). AaWRKY1 sequences were analyzed using Vector NTI software (Invitrogen, USA). Bioinformatics analysis of AaWRKY1 was performed online at the NCBI database (http://www.ncbi.nlm.nih.gov/) and EBI database (http://www.ebi.ac.uk/Tools/msa/clustalw2/).

Dual-Luciferase Assay.
The reporter and the effectors constructs were performed according to the protocols described before [52,57]. In this study, 2.9 kb and 2 kb Table 1: Primers used in this study.

Primers
Primer

Construction of Overexpression Vectors and A. annua
Transformation. The pCAMBIA2300 vector was used for the transformation of A. annua. AaWRKY1 was amplified with specific primers AaWRKY1-up and AaWRKY1-down (Table 1) and then was cloned to the pJET1.2 vector. The pJET1.2-AaWRKY1 construct was digested with Kpn I and Sac I. The small fragment was extracted and was inserted into pCAMBIA2300 vector digested with the same restriction enzymes. The acquired construct pCAMBIA2300-AaWRKY1 was introduced into E. coli DH10b and then was grown in LB medium containing kanamycin. The acquired construct was introduced into A. tumefaciens strain EHA105 by the freeze and thaw method. The seedlings were collected for stable A. annua transformation when they grew to 5 cm in length. The seedlings were cut into 0.5 cm diameter discs for the cocultivation. The plant transformation was carried out as previously described [21,32,58]. The A. tumefaciens strain harboring AaWRKY1 was cocultivated with the leaves discs at 28 ∘ C in the dark for 2 days. The explants were washed with water several times after cocultivation. The explants were screened in selective shootinduction medium MS1 (MS0 + 2 mg/L N 6 -benzoyladenine + 0.2 mg/L naphthalene-1-acetic acid + 50 mg/L kanamycin). The plantlets acquired were transferred into rooting medium MS2 (MS0 + 0.3 mg/L naphthalene-1-acetic acid) for root induction. When roots were formed, regenerated A. annua plants were transferred into soil in the growth chamber. After genomic DNAs were extracted by CTAB method [55,59], using primers 35S and AaWRKY1-down (Table 1), AaWRKY1 transgenic plants were identified by PCR analysis.

Artemisinin
Quantification. The 10th leaves from apical meristem on the main stem of five A. annua plants were collected. The leaves were immerged in 5 mL chloroform in 15 mL tubes. The mixture was shaken for 5 min. The leaves from the tubes were collected and weighed when they dried completely in drying oven at 50 ∘ C. The extract was also dried completely in fume hood and then dissolved in 3 mL methanol, followed by ultrasonic treatment with 35 W at 40 ∘ C. The final supernatant was filtered through a 0.25 m filter.
For each sample, 20 L of the filtrate was injected into an injection port of HPLC. The samples were analyzed using a Waters Alliance 2695 HPLC system coupled with a Waters 2420 ELSD detector (Milford, USA). The filtrate was separated on a 5 m C18 column with 1 mL/min flow rate. A mixture of 60% (v/v) methanol in water was used as mobile phase. The ELSD conditions were optimized at a nebulizer gas pressure of 345 kPa and a drift tube temperature of 45 ∘ C. The content of artemisinin was indicated as mg/g dry weight (DW). Three biological repeats were measured for each sample.

Cloning of AaWRKY1.
According to the previous study of AaWRKY1 (GenBank accession number FJ390842), AaWRKY1 gene sequence was amplified by PCR using A. annua cDNAs as templates, reversely transcribed from total RNAs. The sequence was acquired from low artemisininyielding A. annua var. 38 obtained from our lab, using primers AaWRKY1-up and AaWRKY1-down. Then, it was cloned into pJET1.2 vector and sequenced.
The acquired WRKY gene contained 936 bp open reading frame (ORF) sequence and encoded a protein of 311 amino acids. Alignment of the protein sequences of the acquired WRKY with the other two cloned WRKYs showed that they had high similarity on protein sequences. Four amino acids are different between the acquired WRKY and AaWRKY1, and five amino acids are different between the acquired WRKY and another AaWRKY1 cloned by Han et al. in 2014 [60]. The putative WRKY motif and the amino acid residues which form the zinc finger were conserved in all the three WRKYs (data not shown). Although there are some differences in these sequences, the similarity among these three genes was more than 97% and the WRKY motif and the amino acid residues which formed the zinc finger were identical. Therefore, we consider that they are alleles of AaWRKY1 from different ecotypes of A. annua.

AaWRKY1
Was Induced by MeJA. Methyl jasmonate (MeJA) is an effective elicitor in artemisinin biosynthesis [61]. The expression of ADS and CYP, the two key enzymes of artemisinin biosynthetic pathway, was activated by MeJA. The expression level of ADS was increased rapidly and peaked in 1 h after treatment, while the expression of CYP was induced slowly and peaked in 9 h after treatment [51]. In this study, the expression of AaWRKY1 and AaWRKY3 (accession number GU299481) previously cloned in our lab was tested. The expression of AaWRKY1 was increased rapidly and peaked in 1 h after treatment with 100 M MeJA (Figure 1(a)), while the expression of AaWRKY3 was not changed at the same condition (Figure 1(b)). Ma et al. reported that the expression of ADS was rapidly induced by 0.3 mM MeJA in 0.5 h and peaked by semiquantitative RT-PCR.
JA is linked with artemisinin biosynthesis and trichome initiation and development. JA promoted the formation of both glandular and filamentous trichomes [61]. JA also increased the expression of artemisinin biosynthesis genes [61][62][63]. Many transcription factors induced by JA have been reported to regulate the artemisinin biosynthesis. AaERF1/2 could bind with the promoters of ADS and CYP and increased the accumulation of artemisinin in overexpression plants [51]. AaORA positively regulated the expression of ADS, CYP, and DBR2 and significantly increased the content of artemisinin in overexpression plants [50]. These transcription factors above are strongly induced by JA rapidly. In this study, the expression of AaWRKY1 was also increased rapidly by JA. These results suggested that AaWRKY1 might regulate artemisinin biosynthesis via MeJA.

AaWRKY1 Activated the Expression of ADS and CYP in
Tobacco. In order to investigate the function of AaWRKY1, a dual-luciferase (dual-LUC) assay was performed in transient tobacco expression system. The promoters of ADS and CYP, named ADSpro and CYPpro, were ligated to the pGREEN0800-LUC vector as the reporters, respectively. For the negative control, a CFP protein under the 35S promoter was coinfiltrated with reporters. AaWRKY1 activated the expression of the ADS and CYP promoters by showing higher value of LUC/REN compared with the CFP control in tobacco ( Figure 2). The expression of AaWRKY1 plus ADSpro sample was increased to 9.3-fold compared with the CFP control (Figure 2(a)). The expression of AaWRKY1 plus CYPpro sample was increased to 2.4-fold compared with the CFP control ( Figure 2(b)). These results showed that AaWRKY1 activated both ADS and CYP promoters in plants.
AabZIP1 can bind with the promoters of ADS and CYP in tobacco and can improve the content of artemisinin in overexpressing transgenic plants [52]. AaWRKYs might have the same function in overexpressing transgenic plants. Ma et al. showed that AaWRKY1 binds with ADSpro and activates the expression of ADSpro in tobacco [54]. The promoter of CYP, the second key enzyme in artemisinin biosynthetic pathway, was cloned in 2012 [18]. Two putative W boxes were found in the promoters of ADS and CYP [17,18]. This indicates that AaWRKY1 might bind with the promoters of ADS and CYP on W box site. In this study, AaWRKY1 activates the expression of ADS and CYP promoters in vivo. This suggests that AaWRKY1 might regulate the content of artemisinin in plants by regulating the expression of ADS and CYP.

Stable Transformation and PCR Analysis.
To investigate the function of AaWRKY1 in artemisinin biosynthesis, pCAMBIA2300-AaWRKY1 expression vector was constructed ( Figure 3). The acquired construct was stably transformed into A. annua by A. tumefaciens-mediated leaf-disc transformation method. The leaf discs were used as explants (Figure 4(a)) and were cocultivated at 28 ∘ C in the dark for 48 hours. The shoots were regenerated after changing the selective shoot-induction medium every two weeks (Figures 4(b)-4(c)). The roots were formed in rooting medium every two weeks (Figure 4(d)). Then, the rooted plantlets were transferred into soil for further growth (Figure 4(e)). Seventysix plants acquired were analyzed. Thirty-five transgenic A. annua plants overexpressing AaWRKY1 on kanamycinselective medium were obtained and confirmed by PCR (data not shown). Four independent transgenic lines with high expression level of AaWRKY1 were used for further analysis ( Figures 5 and 6(a)).

Expression of Enzymes and the Content of Artemisinin in
Transgenic Plants. In AaWRKY1-overexpressing transgenic plants, the expression of AaWRKY1 was greatly increased by about 50-to 90-fold compared with that in the nontransgenic control plants (Figure 6(a)). There were no obvious differences in plant height and morphology between transgenic plants and wild type plants.
The expression of the key enzymes in artemisinin biosynthetic pathway was also measured. Compared with that of nontransgenic control plants, the expression of ADS was increased by 1.5-to 3.0-fold, while the expression of CYP was increased by 4.4-to 14.0-fold ( Figure 6(b)).
Furthermore, the content of artemisinin in the transgenic plants overexpressing AaWRKY1 was measured. Artemisinin content in transgenic plants was significantly higher than that in nontransgenic control plants. The artemisinin content was increased by 1.3-to 2.0-fold compared with nontransgenic control plants (Figure 7).
Previous studies showed that AaWRKY1 could bind to the promoter of ADS and activate the expression of enzymes in artemisinin biosynthetic pathway using transient expression systems [54]. Under the 35S promoter, the content of artemisinin in transgenic plants containing AaWRKY1 was slightly increased in the background of high artemisininyielding A. annua var. Chongqing [60]. The content of artemisinin in transgenic plants containing AaWRKY1 was greatly increased in the background of low artemisininyielding A. annua var. 38 in this study. At the same time, only the expression of CYP was increased in transgenic plants containing AaWRKY1 in the background of high artemisinin-yielding A. annua var. Chongqing, while both expression levels of ADS and CYP were increased in transgenic plants containing AaWRKY1 in the background of low artemisinin-yielding A. annua var. 38. This may explain why the artemisinin yield was dramatically increased in low artemisinin-yielding A. annua var. 38 in this study. It indicates that some other coactivators existed in low artemisininyielding A. annua var. 38 or corepressors existed in high artemisinin-yielding A. annua var. Chongqing to regulate the expression of key enzymes in artemisinin biosynthetic pathway together with AaWRKY1. Alternatively, the differences of AaWRKY1 or promoters between high artemisinin-yielding A. annua var. Chongqing and low artemisinin-yielding A. annua var. 38 could also lead to the different content in transgenic plants. Therefore, AaWRKY1 could regulate ADS and CYP promoters, so it could greatly improve the content of artemisinin in low artemisinin-yielding A. annua var. 38.

Conclusions
In sum, AaWRKY1 could regulate the expression of both ADS and CYP in low artemisinin-yielding A. annua var. 38 and greatly increase the yield of artemisinin. These results may provide some new insight into the research on plant metabolic engineering in the future.