AaORA, a trichome-specific AP2/ERF transcription factor of Artemisia annua, is a positive regulator in the artemisinin biosynthetic pathway and in disease resistance to Botrytis cinerea

Authors

  • Xu Lu,

    1. Plant Biotechnology Research Center, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China
    Search for more papers by this author
  • Ling Zhang,

    1. Plant Biotechnology Research Center, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China
    Search for more papers by this author
  • Fangyuan Zhang,

    1. Plant Biotechnology Research Center, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China
    Search for more papers by this author
  • Weimin Jiang,

    1. Plant Biotechnology Research Center, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China
    Search for more papers by this author
  • Qian Shen,

    1. Plant Biotechnology Research Center, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China
    Search for more papers by this author
  • Lida Zhang,

    1. Plant Biotechnology Research Center, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China
    Search for more papers by this author
  • Zongyou Lv,

    1. Plant Biotechnology Research Center, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China
    Search for more papers by this author
  • Guofeng Wang,

    1. Plant Biotechnology Research Center, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China
    Search for more papers by this author
  • Kexuan Tang

    Corresponding author
    • Plant Biotechnology Research Center, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China
    Search for more papers by this author

Author for correspondence:

Kexuan Tang

Tel: +86 21 34206916

Email: kxtang@sjtu.edu.cn

Summary

  • Six transcription factors of APETALA2/ethylene-response factor (AP2/ERF) family were cloned and analyzed in Artemisia annua. Real-time quantitative polymerase chain reaction (RT-Q-PCR) showed that AaORA exhibited similar expression patterns to those of amorpha-4,11-diene synthase gene (ADS), cytochrome P450-dependent hydroxylase gene (CYP71AV1) and double bond reductase 2 gene (DBR2) in different tissues of A. annua.
  • AaORA is a trichome-specific transcription factor, which is expressed in both glandular secretory trichomes (GSTs) and nonglandular T-shaped trichomes (TSTs) of A. annua. The result of subcellular localization shows that AaORA is targeted to the nuclei and the cytoplasm.
  • Overexpression and RNA interference (RNAi) of AaORA in A. annua regulated, positively and significantly, the expression levels of ADS, CYP71AV1, DBR2 and AaERF1. The up-regulated or down-regulated expression levels of these genes resulted in a significant increase or decrease in artemisinin and dihydroartemisinic acid. The results demonstrate that AaORA is a positive regulator in the biosynthesis of artemisinin.
  • Overexpression of AaORA in Arabidopsis thaliana increased greatly the transcript levels of the defense marker genes PLANT DEFENSIN1.2 (PDF1.2), HEVEIN-LIKE PROTEIN (HEL) and BASIC CHITINASE (B-CHI). After inoculation with Botrytis cinerea, the phenotypes of AaORA overexpression in A. thaliana and AaORA RNAi in A. annua demonstrate that AaORA is a positive regulator of disease resistance to B. cinerea.

Introduction

Artemisinin, an important secondary metabolite of Artemisia annua, is currently the best therapeutic agent against both drug-resistant and cerebral malaria-causing strains of Plasmodium falciparum (Weathers et al., 2006). Artemisinin and its derivatives have also been shown to be effective in the treatment of several cancers and viral diseases (Efferth, 2007; Efferth et al., 2008). However, the artemisinin content of A. annua is so low (0.01–0.8% dry weight) that it greatly limits the commercialization of artemisinin.

Recently, the artemisinin biosynthetic pathway has been almost completely resolved by several groups (Fig. 1). In plants, there are two independent pathways producing isopentenyl diphosphate (IPP). One is the classical cytosolic mevalonic acid (MVA) pathway regulated, to a large extent, by 3-hydroxy-3-methylglutaryl-CoA reductase (HMGR); the other is the methylerythritol phosphate (MEP) pathway regulated, to a large extent, by 1-deoxyxylulose 5-phosphate synthase (DXS) and 1-deoxyxylulose 5-phosphate reductoisomerase (DXR; Weathers et al., 2011). Then, IPP and dimethylallyl diphosphate (DMAPP) as substrates are converted to farnesyl diphosphate by farnesyl diphosphate synthase (FPS). The first committed step in the artemisinin-specific biosynthetic pathway is the conversion of farnesyl diphosphate to amorpha-4,11-diene by amorpha-4,11-diene synthase (ADS; Bouwmeester et al., 1999; Mercke et al., 2000). In the following step, amorpha-4,11-diene is hydroxylated to yield artemisinic alcohol, which is catalyzed by a cytochrome P450-dependent hydroxylase (CYP71AV1) and NADPH:cytochrome P450 oxidoreductase (CPR) as the native redox partner (Ro et al., 2006; Teoh et al., 2006). These two enzymes can also oxidize artemisinic alcohol to artemisinic aldehyde and then further to artemisinic acid (Ro et al., 2006). Recently, experimental evidence has been presented that dihydroartemisinic acid is the direct precursor of artemisinin (Brown & Sy, 2004). This route is supported by the cloning and characterization of double bond reductase 2 (DBR2; Zhang et al., 2008), and the cloning of aldehyde dehydrogenase 1 (ALDH1), which catalyzes the oxidation of artemisinic and dihydroartemisinic aldehydes (Teoh et al., 2009). The conversions of dihydroartemisinic acid to artemisinin and artemisinic acid to arteannuin B are suggested to be the enzyme-independent reactions (Brown & Sy, 2004, 2007).

Figure 1.

Artemisinin structure and artemisinin biosynthetic pathway. HMGR, 3-hydroxy-3-methylglutaryl-CoA reductase; DXS, 1-deoxyxylulose 5-phosphate synthase; DXR, 1-deoxyxylulose 5-phosphate reductoisomerase; FPS, farnesyl diphosphate synthase; ADS, amorpha-4,11-diene synthase; CYP, cytochrome P450; CYP71AV1, cytochrome P450-dependent hydroxylase; CPR, cytochrome P450 oxidoreductase; DBR2, double bond reductase 2; ALDH1, aldehyde dehydrogenase 1; DXP, 1-deoxy-d-xylulose 5-phosphate; FPP, farnesyl diphosphate; G3P, glyceraldehyde 3-phosphate; PYR, pyruvate; HMG-CoA, 3S-hydroxy-3-methylglutaryl-CoA; DMAPP, dimethylallyl diphosphate; IPP, isopentenyl diphosphate; MEP, methylerythritol phosphate; MVA, mevalonic acid.

Artemisinin, dihydroartemisinic acid and a number of other terpenoids are produced in glandular secretory trichomes (GSTs) present on aerial surfaces of the plant (Olofsson et al., 2011). The key enzymes of the artemisinin biosynthetic pathway are ADS, CYP71AV1 and DBR2, which are specifically expressed only in GSTs (Olofsson et al., 2012). Trichomes are divided into nonglandular T-shaped trichomes (TSTs) and GSTs. Nonglandular trichomes have a more specialized role in plant defense and act as deterrents to insect oviposition and feeding (Levin, 1973). GSTs have the capacity to synthesize, store and sometimes secrete large amounts of specialized metabolites which are related to demonstrated or presumed roles in plant defense (Levin, 1973; Schilmiller et al., 2008).

AP2/ERF transcription factors are one of the most important families of transcription factors involved in the plant response to biotic and abiotic stresses, and in the regulation of metabolism and developmental processes in various plant species (Agarwal et al., 2006). The overexpression of the AP2/ERF transcriptional regulator ORCA3 resulted in the enhanced expression of several metabolite biosynthetic genes and, consequently, in increased accumulation of terpenoid indole alkaloids in Catharanthus roseus (Van der Fits & Memelink, 2000). Recently, the overexpression of AP2/ERF transcriptional factors AaERF1 and AaERF2 has been shown to elevate the transcript levels of both ADS and CYP71AV1, resulting in increased accumulation of artemisinin and artemisinic acid in A. annua (Yu et al., 2012). Therefore, the AP2/ERF transcription factors in A. annua, which are specifically expressed in trichomes, may play important roles in the regulation of the artemisinin biosynthetic pathway and in plant defense.

In this article, six AP2/ERF transcription factors were cloned and analyzed in A. annua. After a series of filtrations by real-time quantitative polymerase chain reaction (RT-Q-PCR), AaORA was the only AP2/ERF transcription factor which exhibited similar expression patterns to those of ADS, CYP71AV1 and DBR2 at different positions of the leaves and in different tissues in A. annua. AaORA was analyzed in detail, including promoter analysis and subcellular localization. We also performed Agrobacterium tumefaciens-mediated transformation to generate transgenic plants expressing the AaORA promoter-β-glucuronidase (GUS), AaORA-overexpressing and AaORA RNA interference (RNAi) vectors, and reported their phenotypic as well as biological and biochemical characterization with respect to the biosynthesis of artemisinin and disease resistance to Botrytis cinerea.

Materials and Methods

Isolation and characterization of AaERFs

A cDNA library of A. annua L. was constructed using the leaves of Artemisia annua as the source of mRNA. Randomly picked clones were sequenced by the Beijing Genomic Institute Shenzhen, China. Five full-length AP2/ERF transcription factors were obtained by this method.

The sequence of C. roseus ORCA3 (EU072424, Octadecanoid-derivative Responsive Catharanthus AP2-domain protein 3) was used as a query sequence for a BLAST search of GenBank. The search returned an A. annua EST (expression sequence tag; EY074653) which was also from the normalized leaf library. Based on the sequence, a pair of primers was designed, synthesized and used to amplify the partial sequence of a putative ERF gene from A. annua. Then, 3′ end and 5′ end sequences of the AP2/ERF transcription factor were obtained by the rapid amplification of cDNA ends (RACE) method, according to the manufacturer's instructions (Invitrogen). The AP2/ERF transcription factor, which had the closest evolutionary relationships to ORCA2 and ORCA3, was named AaORA (Supporting Information Fig. S1). The novel nucleotide sequences published here have been deposited in the EMBL/DDBJ/GenBank databases under accession numbers AaORA (JQ797708), AaERF4 (JQ797709), AaERF5 (JQ797710), AaERF6 (JQ797711), AaERF7 (JQ797712) and AaERF8 (JQ797713).

Comparative and bioinformatic analyses of all six AaERFs were carried out online at the websites http://www.ncbi.nlm.nih.gov and http://cn.expasy.org. Sequence analysis was performed using DNAMAN software (Lynnon Biosoft, Vaudreuil, Quebec, Canada) and Vector NTI software (Invitrogen). The phylogenetic analysis of the six AaERF proteins and ERFs from other species was carried out by alignment with Clustal X1.81 (Thompson et al., 1997) using default parameters. A phylogenetic tree was constructed by the neighbor-joining method (Saitou & Nei, 1987) using the software MEGA version 3.1 (Kumar et al., 2001).

Plant materials

The seeds of A. annua were obtained from the School of Life Sciences, Southwest University in Chongqing, China. Seeds of A. annua and A. thaliana were surfaced sterilized in 70% ethanol for 1 min and then with a 0.5% sodium hypochlorite solution with 0.1% Tween 20 for 10 min. After rinsing five times thoroughly with sterile distilled water, the seeds were placed on Murashige and Skoog (MS) medium (Sigma-Aldrich, St Louis, MO, USA) with 88 mM sucrose and 0.7% agar (pH 5.8). Tobacco (Nicotiana benthamiana) seeds were seeded directly on the soil. Different tissues and leaves from different positions of A. annua were collected for RNA extraction using plant RNA isolation reagent (Tiangen Biotech, Beijing, China) following the manufacturer's instructions. Samples were collected from the meristems (Leaf0), Leaf1, Leaf2, Leaf3, Leaf5, Leaf7 and Leaf9 counted from the apical top of the main stem of A. annua plants, which were grown in a glasshouse for 1 month to a height of 35–45 cm. The different tissues (roots, stems, leaves, Bud0, Bud1 and flowers) of the plants were collected from five plants grown in the field for 5 months. All the tissues from the five plants were separately pooled for each determination. The experiments were repeated in triplicate.

Expression pattern analysis of AaERFs by RT-Q-PCR

The expression patterns of AaERFs in various tissues of A. annua were analyzed using RT-Q-PCR. AaORA overexpression and AaORA RNAi in transgenic A. annua, as well as AaORA overexpression in A. thaliana, were also analyzed by this method. All RNA samples were digested with DNase I (RNase-free) before use. Aliquots of 0.4 μg of total RNA were employed in the reverse transcriptase reaction using random hexamer primers for the synthesis of first-strand cDNA. The amplification reactions of RT-Q-PCR were performed on a Real-Time PCR Machine (Bio-Rad, Watford, Hertfordshire, UK) with gene-specific primers, and a SYBR ExScript RT-PCR kit (Takara, Shiga, Japan) to confirm changes in gene expression. The RT-Q-PCR procedures were performed as described previously (Lu et al., 2012). All the primers used in RT-Q-PCR are listed in Table S1.

Cloning and analysis of AaORA promoter and GUS expression in transgenic A. annua

The upstream region of AaORA was amplified from genomic DNA using the Genome Walker Kit (Clontech, Mountain View, CA, USA). The AaORA-specific primers (AaORA-sp1 and AaORA-sp2), Adaptor Prime1 and Adaptor Prime2 were used, following the manufacturer's recommended procedures. The final reaction products were electrophoresed in 1% agarose gel, and the 1193-bp fragment was eluted from the gel and cloned into pMD18-T-simple vector. The insert DNA was sequenced by Beijing Genomic Institute Shenzhen, China.

To generate the AaORA promoter-GUS construct, the 5′-flanking DNA of the AaORA coding region was amplified with AaORA-PF and AaORA-PR. The 1193-bp PCR fragment was cloned into the pCAMBIA1391Z vector. The construct was transformed into A. annua plants, as described previously (Zhang et al., 2009). Histochemical staining for GUS activity in transgenic plants was performed as described previously (Jefferson et al., 1987). Plants transformed with pCAMBIA1391Z were used as a parallel negative control.

Subcellular localization of AaORA–yellow fluorescent protein (YFP) fusion proteins

The open reading frame (ORF) of AaORA was recombined into the pEarlyGateway104 vector by the Gateway LR recombination reaction (Invitrogen) to generate pEarlyGateway104-AaORA. The plasmid was used for transient expression in tobacco (Nbenthamiana) epidermal cells, as described previously (Voinnet et al., 2003). We observed and recorded the results of the 48 h after transformation. The YFP fluorescence was imaged using a Leica TCS SP5 laser scanning confocal microscope (Leica Microsystems, Wetzlar, Germany). A ×20 objective was used for confocal imaging. Fluorescence was detected at 530–600 nm for YFP.

AaORA overexpression in A. annua

The full length of the AaORA coding sequence was amplified with primers AaORAfull5 and AaORAfull3 using PrimeSTAR™ HS DNA polymerase (Takara) and subcloned into pMD18-T simple vector. The pMD18-T-AaORA vector was digested by BamHI and SacI. The full-length ORF of AaORA was cloned into the BamHI and SacI sites of the pCAMBIA2300+ vector under the 35S promoter to generate pCAMBIA2300-35S::AaORA::NOS. The construct was transferred into Agrobacterium tumefaciens EH105, and then introduced into A. annua (Zhang et al., 2009). Leaf5 was used to analyze the expression of genes in AaORA-overexpressing and empty vector transgenic A. annua plants.

AaORA RNAi in A. annua

The 274-bp fragment of AaORA, corresponding to AaORA cDNA from nucleotides 552 to 825 bp, was cloned from A. annua. In reverse transcription-PCR, AaORAi-F (with XbaI and XhoI sites) and AaORAi-R (with BamHI and HindIII sites) were used as the forward and reverse primers, respectively. The amplified fragment was cloned into pMD18-T simple vector and sequenced. After confirmation by sequencing, the fragment was placed in forward and reverse orientation on the two ends of the GUS intron in pBluescript SK+ to construct the hairpin (hp) structure. Then, the expression cassette was excised with SacI and KpnI from pBluescript SK+ containing the hp structure of AaORA and ligated into the expression vector pCAMBIA2300+ to obtain the final vector pCAMBIA2300-35S::hairpin AaORA::NOS. pCAMBIA2300+ vector containing only nptII (neomycin phosphotransferase gene conferring resistance to kanamycin) was used as the control vector in the transformations. The pCAMBIA2300-35S::hairpin AaORA::NOS and pCAMBIA2300+ vectors were then transferred into Agrobacterium tumefaciens strain EHA105 by the conventional freezing and melting method, and the resulting strains were used in the transformation of A. annua. The transformation of A. annua was performed following the protocol of Zhang et al. (2009). The meristems were used to analyze the expression of genes in AaORA RNAi and empty vector transgenic A. annua plants.

High-performance liquid chromatography (HPLC)

Leaves of A. annua were dried at 45°C and then ground. The dried leaf powder (0.1 g per sample) was extracted twice with 2 ml methanol using a Shanghai Zhisun Instrument Co. Ltd model JYD-650 ultrasonic processor (Shanghai, China), and then centrifuged for 10 min at 1699 g to remove the suspended particles. The final supernatant was filtered through a 0.25-μm pore-size filter.

The samples were analyzed using a Waters Alliance 2695 HPLC system coupled with a Waters 2420 ELSD detector (Milford, MA, USA). The HPLC conditions for artemisinin were as follows: column, Waters C18 (YMC-Pack ODS-A; particle size, 5 μm; pore size, 12 nm; column size, 4.6 × 250 mm2); mobile phase, water–methanol (40 : 60, v/v); flow rate, 1 ml min−1. The ELSD conditions were optimized at a nebulizer gas pressure of 345 kPa and a drift tube temperature of 45°C, and the gains were set at 7 min.

The HPLC conditions for dihydroartemisinic acid were as follows: column, Waters C18 (YMC-Pack ODS-A; particle size, 5 μm; pore size, 12 nm; column size, 4.6 × 250 mm2); mobile phase, acetonitrile–0.1% aqueous acetic acid (pH 3.2; 60 : 40, v/v); flow rate, 1 ml min−1. The ELSD conditions were optimized at a nebulizer gas pressure of 345 kPa and drift tube temperature of 45°C, and the gains were set at 13 min. Artemisinin from Sigma and dihydroartemisinic acid from Guangzhou Honsea Sunshine Bio Science and Technology Co. Ltd (Guangzhou, Guangdong, China) were used as standards. For each sample, the injection volume was 20 μl, and the results were analyzed using Empower (Waters' chromatography data software).

AaORA overexpression in A. thaliana

The plasmid pCAMBIA2300-35S::AaORA::NOS was transferred into Agrobacterium tumefaciens GV3101, and then introduced into A. thaliana (ecotype Columbia) plants using the floral dip method (Zhang et al., 2006). Transgenic plants were selected on MS plates containing 50 μg ml−1 kanamycin. PCR was performed to verify the transgenic status of the screened plants. The phenotypic effects were analyzed in the T2 generation.

Pathogen infections

The procedure of pathogen infection was performed as described previously with some modifications (Gao et al., 2011). Botrytis cinerea was grown on potato dextrose agar plates for 2 wk at 26°C. Spores were collected from 2-wk-old cultures and washed twice with sterile water. Washed spores were suspended in 10 ml of sterile water, and the suspension was filtered through Miracloth to remove mycelia (Huaxiweicai Co. Ltd, Xinxiang, Henan, China). Four-week-old A. thaliana and 2-month-old A. annua were spray inoculated with spore suspensions (2 × 105 spores ml−1) and maintained under high humidity until disease assessment. Inoculated plants were scored on the basis of the presence of any disease symptoms 4 d after inoculation with B. cinerea, including chlorosis, curling and necrosis of the leaves. The data presented are representative of three separate experiments.

Results

Isolation and characterization of AaERFs

A cDNA library of A. annua was constructed using leaves as the source of mRNA. In total, 405 randomly picked clones were sequenced. Five full-length AP2/ERF transcription factors were obtained, which were subsequently confirmed by sequencing. One gene, which was named AaORA, was homologously cloned by RACE. It contained an 1101-bp ORF encoding 366 amino acids. The coding region was followed by a 3′-untranslated region that was 78 bp long downstream from the stop codon including the poly (A).

The results of the BLAST-Protein (BLASTP) online (http://www.ncbi.nlm.gov/blast) showed that AaERFs shared a highly conserved AP2/ERF domain with other ERF proteins, including ORCA3 (C. roseus), Arabidopsis AtERF1, ORA59 and AaERF1 (Fig. S2). The highly conserved WLG motif was found in the AP2 domain (Sakuma et al., 2002) (Fig. S2). From the phylogenetic tree, which was constructed by the neighbor-joining method, AaORA had the closest evolutionary relationships to ORCA3 and ORCA2 (Fig. S1).

Expression patterns of AaERFs at different positions of the leaves

To dissect the expression dynamics of key genes in the artemisinin biosynthetic pathway and the nine AP2/ERF transcription factors of A. annua (including AaERF1, AaERF2 and AaERF3), the expression of these genes was analyzed by RT-Q-PCR at different positions of the leaves. The results showed that AaORA was highly expressed in very young leaves (Leaf0, Leaf1 and Leaf2), and expression reduced quickly during leaf development. The expression pattern of AaORA was similar to those of ADS, CYP71AV1 and DBR2 (Fig. 2). The other eight AaERFs showed different patterns from those of ADS, CYP71AV1 and DBR2. The expression of these genes increased gradually during leaf development (Fig. 2).

Figure 2.

Expression patterns of Artemisia annua ethylene-response factors (AaERFs), amorpha-4,11-diene synthase (ADS), cytochrome P450-dependent hydroxylase (CYP71AV1) and double bond reductase 2 (DBR2) at different positions of the leaves in Artemisia annua. The expression levels of ADS, CYP71AV1, DBR2, AaORA, AaERF1, AaERF2, AaERF3, AaERF4, AaERF5, AaERF6, AaERF7 and AaERF8 at Leaf0, 1, 2, 3, 5, 7 and 9 were measured by real-time quantitative polymerase chain reaction (RT-Q-PCR). Values indicate the mean fold relative to sample Leaf0. Data represent means ± SE from three replicates. Actin was used as a control for normalization.

Expression patterns of AaERFs in different tissues

The expression patterns of the key genes in the artemisinin biosynthetic pathway and of the nine AP2/ERF transcription factors of A. annua were investigated in different tissues by RT-Q-PCR. The key enzyme genes ADS, CYP71AV1 and DBR2 of artemisinin biosynthesis were highly expressed in Bud0 and Bud1, lower in stem, leaf and flower and poorly expressed in root. The expression pattern of AaORA was similar to those of ADS, CYP71AV1 and DBR2. All four genes showed high expression in Bud0 and Bud1. The other AaERFs showed different expression patterns (Fig. 3). AaERF4 and AaERF8 were highly expressed in root, whereas AaERF1, AaERF2, AaERF5 and AaERF6 were expressed almost at the same level in all of the tissues. AaERF3 was only highly expressed in Bud0. AaERF7 was expressed more highly in Bud0 than in other tissues (Fig. 3).

Figure 3.

Expression patterns of Artemisia annua ethylene-response factors (AaERFs), amorpha-4,11-diene synthase (ADS), cytochrome P450-dependent hydroxylase (CYP71AV1) and double bond reductase 2 (DBR2) in different tissues of Artemisia annua. The expression levels of ADS, CYP71AV1, DBR2, AaORA, AaERF1, AaERF2, AaERF3, AaERF4, AaERF5, AaERF6, AaERF7 and AaERF8 in young buds (Bud0), buds not fully blossomed (Bud1), fully florescent flowers (Flower), mature leaves (Leaf), stems (Stem) and roots (Root) were measured by real-time quantitative polymerase chain reaction (RT-Q-PCR). Values indicate the mean fold relative to sample Leaf. Data represent means ± SE from three replicates. Actin was used as a control for normalization. The expression of AaERF3 in different tissues has been published previously by our laboratory (Lu et al., 2012) and the data were used to compare with other AaERFs in this figure.

AaORA is a trichome-specific transcription factor in A. annua

To examine the expression pattern of AaORA in detail, we cloned an 1193-bp promoter sequence (JQ797714) of AaORA by genomic walking. The transcription start site (TSS) of the cloned promoter was predicted using TSSP software (http://linux1.softberry.com/berry.phtml). A putative TSS of AaORA (labeled +1 in Fig. S3) was predicted 187 bp upstream of the translation initiation ATG codon. Putative cis-acting elements of the promoter were predicted using PLANTCARE software (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/). Some important binding sites, such as G- and W-boxes, were marked in the promoter sequence (Fig. S3).

After the generation of AaORA promoter-GUS transgenic plants, GUS staining was used to investigate the expression pattern of AaORA in detail. Three independent lines of transgenic A. annua plants showed similar patterns. There are two types of trichome on aerial organs of A. annua, that is GSTs and nonglandular TSTs (Wang et al., 2011). GUS staining was only observed in GSTs and TSTs of stem (Fig. 4a), leaves (Fig. 4b), leaf primordia (Fig. 4c) and flower buds (Fig. 4d). No signals were observed in the negative control plants transformed with pCAMBIA1391Z empty vector (Fig. S4). The results show that AaORA is a trichome-specific transcription factor, which is expressed in both GSTs and TSTs of A. annua.

Figure 4.

β-Glucuronidase (GUS) staining of transgenic Artemisia annua transformed using the pAaORA-GUS plasmid: (a) stem; (b) leaf; (c) leaf primordial; (d) flower bud. GST, glandular secretory trichome; TST, T-shaped trichome.

Targeting of AaORA protein to both the nuclei and the cytoplasm

To examine the subcellular localization of AaORA protein in vivo, the fused expression vector pEarlyGateway104-AaORA was constructed and used to perform a transient expression assay in N. benthamiana epidermal cells. Ten epidermal cells of N. benthamiana with the pEarlyGateway104-AaORA, pEarlyGateway104-AaERF1 and pEarlyGateway104 vectors were analyzed. The AaORA–YFP fusion protein was localized in both the nuclei and the cytoplasm, whereas the control YFP was distributed throughout the cells and the AaERF1–YFP fusion protein was exclusively localized to the nuclei (Fig. 5a–c). Although AaORA protein was localized in both the nuclei and cytoplasm, the protein showed a much stronger expression in the nuclei. Therefore, AaORA should have its main regulatory function in the nuclei.

Figure 5.

Subcellular localization of AaORA and AaERF1 in transiently transformed tobacco (Nicotiana benthamiana) leaves: (a) empty vector; (b) AaERF1; (c) AaORA. YFP, yellow fluorescent protein.

AaORA overexpression in A. annua causes a significant increase in artemisinin and dihydroartemisinic acid

Independent transgenic plants were selected in kanamycin-containing medium and further confirmed by genomic PCR using the 35S forward primer, AaORA reverse primer and reverse primer of the kanamycin-resistant gene (Fig. S5a,b). Three independent transgenic lines were chosen for further analysis. Compared with the control level, the transcript levels of AaORA were increased 4–14-fold, whereas the expression of ADS, CYP71AV1, DBR2 and AaERF1 was increased 6.3–12.9-, 5–17-, 3.6–12.7- and 4.2–12.5-fold, respectively, in AaORA-overexpressing transgenic lines (Fig. 6a). The statistical analysis showed that the observed differences were statistically significant. The expression levels of HMGR, DXS, DXR, FPS, CPR, ALDH1 and AaERF2 were barely or only slightly changed (Fig. 6a). The artemisinin and dihydroartemisinic acid contents in 5-month-old A. annua plants were analyzed by HPLC. Compared with the control, the artemisinin and dihydroartemisinic acid contents were increased by 40–53% and 22–35%, respectively, in AaORA-overexpressing transgenic plants (Fig. 6b,c).

Figure 6.

Expression analysis by real-time quantitative polymerase chain reaction (RT-Q-PCR) and the measurements of artemisinin and dihydroartemisinic acid by high-performance liquid chromatography (HPLC) in AaORA-overexpressing transgenic Artemisia annua. (a) Expression analysis of AaORA, the key enzyme genes of the artemisinin biosynthetic pathway, AaERF1 and AaERF2 in empty vector and three independent AaORA-overexpressing transgenic A. annua by RT-Q-PCR. (b) Artemisinin content in empty vector and three independent AaORA-overexpressing transgenic A. annua. (c) Dihydroartemisinic acid content in empty vector and three independent AaORA-overexpressing transgenic A. annua. Data represent means ± SE from three replicates. Actin was used as a control for normalization. Statistical significance was determined by Student's t-test (**, < 0.01; *, < 0.05). Asterisks indicate the difference between empty vector and AaORA-overexpressing transgenic A. annua.

AaORA RNAi in A. annua causes a significant reduction in artemisinin and dihydroartemisinic acid

To further prove the function of AaORA in the regulation of the artemisinin biosynthetic pathway, the expression of AaORA was down-regulated by RNAi. The transgenic plants were first confirmed by PCR using the 35S forward primer, AaORAi reverse primer and the reverse primer of the kanamycin-resistant gene (Fig. S5c,d). Three independent transgenic lines were chosen for further analysis. In the RNAi transgenic lines, the transcript levels of AaORA were suppressed to 42–68% of the control level, whereas the expression of ADS, CYP71AV1, DBR2 and AaERF1 was reduced to 53–59%, 61–73%, 67–75% and 20–77%, respectively, of the control levels (Fig. 7a). The statistical analysis showed that the observed differences were statistically significant. The expression levels of HMGR, DXS, DXR, FPS, CPR, ALDH1 and AaERF2 in the RNAi transgenic lines were barely or only slightly changed (Fig. 7a). The artemisinin and dihydroartemisinic acid contents in 5-month-old A. annua plants were analyzed by HPLC. Compared with the control, the artemisinin and dihydroartemisinic acid contents were decreased to 64–52% and 63–41%, respectively, in AaORA RNAi plants (Fig. 7b,c). Considering that the genome sequence of A. annua is not known, the RNAi construct used to silence AaORA expression in A. annua may potentially silence other related genes (yet unknown but possibly similar sequences).

Figure 7.

Expression analysis by real-time quantitative polymerase chain reaction (RT-Q-PCR) and the measurements of artemisinin and dihydroartemisinic acid by high-performance liquid chromatography (HPLC) in AaORA RNAi transgenic Artemisia annua. (a) Expression analysis of the key enzyme genes of the artemisinin biosynthetic pathway, AaORA, AaERF1 and AaERF2 in empty vector and three independent AaORA RNAi transgenic A. annua by RT-Q-PCR. (b) Artemisinin content in empty vector and three independent AaORA RNAi transgenic A. annua. (c) Dihydroartemisinic acid content in empty vector and three independent AaORA RNAi transgenic A. annua. Data represent means ± SE from three replicates. Actin was used as a control for normalization. Statistical significance was determined by Student's t-test (**, < 0.01; *, < 0.05). Asterisks indicate the difference between empty vector and AaORA RNAi transgenic A. annua.

AaORA functions as a positive regulator of disease resistance to B. cinerea

AaORA is a trichome-specific transcription factor, which is expressed in both GSTs and TSTs of A. annua. As the trichomes of plants have close relationships with disease resistance (Levin, 1973; Schilmiller et al., 2008), we inferred that AaORA may have a function in disease resistance. The analysis of three independent AaORA-overexpressing Arabidopsis lines showed that the expression level of AaORA was increased significantly (Fig. 8a). Correspondingly, B-CHI was shown to be elevated by 2.1–5.8-fold in AaORA-overexpressing lines (Fig. 8b). Compared with the wild-type plants, the transcript levels of PDF1.2 and HEL were elevated by 30–245- and 10.2–30.5-fold, respectively (Fig. 8c,d). The statistical analysis showed that the observed differences were statistically significant. The AaORA-overexpressing lines were observed following inoculation with B. cinerea. For each of the AaORA-overexpressing lines, disease symptoms were reduced significantly in inoculation experiments. Four days after inoculation with B. cinerea, all the leaves of wild-type plants showed infected symptoms, whereas only between 33% and 63% of leaves from AaORA-overexpressing lines were symptomatic (Fig. 8e,f). The statistical analysis showed that the observed differences were statistically significant. Six days after inoculation with B. cinerea, the wild-type plants were dry and dead, whereas most of the AaORA-overexpressing plants (L2 line) were growing well (Fig. 8g). AaORA RNAi lines in A. annua were also used to analyze the pathogen infection. Six days following inoculation with B. cinerea, all three independent RNAi lines showed more serious disease symptoms than the control (Fig. 8h,i). Six days after inoculation, AaORAi-11 and AaORAi-35 had even died (Fig. 8i). All the results demonstrate that AaORA is a positive regulator of disease resistance to B. cinerea in A. annua.

Figure 8.

AaORA functions as a positive regulator of disease resistance to Botrytis cinerea. (a) Expression analysis of AaORA in Col-0 and three independent AaORA-overexpressing transgenic Arabidopsis lines. (b) Expression analysis of BASIC CHITINASE (B-CHI) in Col-0 and three independent AaORA-overexpressing transgenic Arabidopsis lines. (c) Expression analysis of PLANT DEFENSIN1.2 (PDF1.2) in Col-0 and three independent AaORA-overexpressing transgenic Arabidopsis lines. (d) Expression analysis of HEVEIN-LIKE PROTEIN (HEL) in Col-0 and three independent AaORA-overexpressing transgenic Arabidopsis lines. (e) The numbers of infected leaves 4 d after inoculation with B. cinerea in Col-0 and three independent AaORA-overexpressing transgenic Arabidopsis lines. (f) The phenotypes of Col-0 and AaORA-overexpressing transgenic Arabidopsis (L2) 4 d after inoculation with B. cinerea. (g) The phenotypes of Col-0 and AaORA-overexpressing transgenic Arabidopsis (L2) 6 d after inoculation with B. cinerea. (h) The phenotypes of empty vector and three independent AaORA RNAi transgenic Artemisia annua lines, without inoculation with B. cinerea. (i) The phenotypes of empty vector and three independent AaORA RNAi transgenic A. annua lines 6 d after inoculation with B. cinerea. Data represent the means ± SE from three replicates. Actin was used as a control for normalization. Statistical significance was determined by Student's t-test (**, < 0.01; *, < 0.05). Asterisks indicate the difference between empty vector and AaORA-overexpressing transgenic Arabidopsis.

Discussion

The relationships of AaORA, artemisinin and dihydroartemisinic acid

The artemisinin and dihydroartemisinic acid contents have been measured previously at different positions of the leaves in our laboratory (Zhang et al., 2012). The results suggested that dihydroartemisinic acid was accumulated rapidly in Leaf0, Leaf1, Leaf2 and Leaf3, and ADS, CYP71AV1, DBR2 and AaORA were highly expressed at the same time. Subsequently, the content of dihydroartemisinic acid declined quickly during the development of Leaf5, Leaf7 and Leaf9, and the expression of ADS, CYP71AV1, DBR2 and AaORA was reduced simultaneously. However, the artemisinin content increased step by step from Leaf0 up to a maximum in Leaf7 and Leaf9. Dihydroartemisinic acid is a precursor of artemisinin, which can easily be converted photo-oxidatively into artemisinin (Wallaart et al., 1999). Therefore, the expression patterns of ADS, CYP71AV1, DBR2 and AaORA, which are similar to the pattern of dihydroartemisinic acid at different positions of the leaves, imply that AaORA plays an important role in the regulation of the artemisinin biosynthetic pathway.

The artemisinin and dihydroartemisinic acid contents were measured in different tissues (Fig. S6). The results showed that artemisinin and dihydroartemisinic acid were highly accumulated in Bud0 and Bud1, less so in stem, leaf and flower, and almost none in root. The expression patterns of ADS, CYP71AV1, DBR2 and AaORA were similar to the content patterns of artemisinin and dihydroartemisinic acid in different tissues. All the above results imply that AaORA is an important regulator of the artemisinin biosynthetic pathway.

The GUS staining of AaORA promoter-GUS transgenic plants showed that AaORA is a trichome-specific transcription factor, which is expressed in both GSTs and TSTs of A. annua. These results explain why AaORA exhibited very similar expression patterns to those of ADS, CYP71AV1 and DBR2 in different tissues and different positions of the leaves. As artemisinin is produced in GSTs (Olofsson et al., 2011), AaORA, which is a trichome-specific transcription factor, should play an important role in the regulation of the artemisinin biosynthetic pathway.

AaORA is a positive regulator of the artemisinin biosynthetic pathway

In AaORA-overexpressing transgenic lines, dihydroartemisinic acid and artemisinin contents were increased by 22–35% and 40–53%, respectively (Fig. 6b,c). In AaORA RNAi transgenic lines, dihydroartemisinic acid and artemisinin contents were decreased to 63–41% and 64–52%, respectively, of the control level (Fig. 7b,c). Consequently, AaORA is a positive regulator of the artemisinin biosynthetic pathway.

As the transcript level of AaORA increased in AaORA-overexpressing transgenic lines, the expression of ADS, CYP71AV1 and DBR2 also increased to the same extent. In AaORA RNAi transgenic lines, the transcript level of AaORA decreased, and the expression of ADS, CYP71AV1 and DBR2 also decreased. ADS, CYP71AV1 and DBR2 are specifically expressed in GSTs (Olsson et al., 2009; Wang et al., 2011, 2013; Olofsson et al., 2012). As AaORA is also specifically expressed in trichomes of A. annua, we inferred that AaORA may regulate directly the promoters of ADS, CYP71AV1 and DBR2 in GSTs of A. annua.

RT-Q-PCR was also used to analyze the expression of AaERF1 and AaERF2 in AaORA-overexpressing and AaORA RNAi transgenic lines. The results showed that the expression of AaERF1 exhibited a positive relationship with that of AaORA. The expression of AaERF2 showed little relationship with that of AaORA. AaERF1 is expressed ubiquitiously in all organs and can activate artemisinin biosynthesis through ADS and CYP71AV1 (Yu et al., 2012), whereas the key genes ADS, CYP71AV1 and DBR2 of the artemisinin biosynthetic pathway are specifically expressed in GSTs (Olsson et al., 2009; Wang et al., 2011, 2013; Olofsson et al., 2012). Therefore, in addition to the function of regulating the artemisinin biosynthetic pathway, we inferred that AaERF1 should have other functions which result in the normal expression pattern of AaERF1 being different from those of ADS, CYP71AV1, DBR2 and AaORA in A. annua. The results of RT-Q-PCR showed that ADS, CYP71AV1, DBR2 and AaORA were highly expressed in very young leaves (Leaf0, Leaf1 and Leaf2). Although the expression level of AaERF1 increased from Leaf0 to Leaf9, the results of RT-Q-PCR showed that the expression level of AaERF1 was also quite high in very young leaves (Leaf0, Leaf1 and Leaf2). Therefore, although the normal expression pattern of AaERF1 was different from that of AaORA, there was a possibility that AaORA activated artemisinin biosynthesis through AaERF1. Considering that the expression of AaERF1 exhibited a positive relationship with that of AaORA in transgenic A. annua, we inferred that AaORA might activate artemisinin biosynthesis through AaERF1.

All the overexpression and RNAi data in A. annua showed that AaORA could positively regulate the expression levels of ADS, CYP71AV1, DBR2 and AaERF1 in A. annua. Furthermore, the up-regulated or down-regulated expression levels of these genes resulted in an increase or decrease in artemisinin and dihydroartemisinic acid. All the results demonstrate that AaORA is an important positive regulator of the artemisinin biosynthetic pathway in A. annua.

AaORA functions as a positive regulator of disease resistance to B. cinerea in A. annua

PDF1.2, HEL and B-CHI genes are marker genes to several fungi in A. thaliana, including B. cinerea (Reymond & Farmer, 1998; Solano et al., 1998; Berrocal-Lobo et al., 2002; Lorenzo et al., 2003). RT-Q-PCR results showed that the transcript levels of PDF1.2, HEL and B-CHI were increased significantly in AaORA-overexpressing A. thaliana. After the inoculation with B. cinerea, wild-type plants dried out and died, whereas most of the AaORA-overexpressing plants (L2 line) grew well. The AaORA RNAi lines in A. annua were also used to study pathogen infection. The results showed that, compared with empty vector transgenic plants, disease resistance to B. cinerea was decreased significantly in AaORA RNAi A. annua. The results of RT-Q-PCR showed that AaORA could activate defense marker genes in A. thaliana and result in an increase in disease resistance to B. cinerea. Therefore, the decrease in disease resistance to B. cinerea in AaORA RNAi A. annua may be caused by the reduced expression of unknown defense-related genes that are potentially regulated by AaORA during pathogen attack. All the data in Arabidopsis and A. annua demonstrate that AaORA is a positive regulator of disease resistance to B. cinerea and, possibly, other pathogens.

In conclusion, six AP2/ERF transcriptional factors were cloned and analyzed in A. annua. RT-Q-PCR showed that AaORA was the only transcription factor which exhibited similar expression patterns to those of ADS, CYP71AV1 and DBR2 at different positions of the leaves and in different tissues of A. annua. The AaORA promoter-GUS transgenic plants showed that AaORA is a trichome-specific transcription factor, which is expressed in both GSTs and TSTs of A. annua. The up-regulation or down-regulation of AaORA in A. annua positively regulated the expression levels of ADS, CYP71AV1, DBR2 and AaERF1, which resulted in an increase or decrease in artemisinin and dihydroartemisinic acid in A. annua. Consequently, AaORA is an important positive regulator of the artemisinin biosynthetic pathway. The inoculation experiments with B. cinerea in A. thaliana and A. annua showed that AaORA is a positive regulator of disease resistance to B. cinerea. All these data demonstrate that AaORA is a valuable AP2/ERF transcription factor, not only for the regulation of the artemisinin biosynthetic pathway, but also for the enhancement of the disease resistance to B. cinerea in A. annua.

Acknowledgements

This work was funded by the China ‘863’ Program (grant no. 2011AA100605) and China Transgenic Research Program (grant no. 2011ZX08002-001), Ministry of Education.

Ancillary

Advertisement