The Brassica juncea BjCdR15, an ortholog of Arabidopsis TGA3, is a regulator of cadmium uptake, transport and accumulation in shoots and confers cadmium tolerance in transgenic plants


Author for correspondence:
Antonella Furini
Tel: +39 045 8027950


  • A bZIP transcription factor from Brassica juncea (BjCdR15) was isolated by the cDNA-amplified fragment length polymorphism technique after cadmium treatment. Sequence analysis indicated high similarity between BjCdR15 and Arabidopsis TGA3. In Arabidopsis, TGA3 transcription is also induced by cadmium; hence, we investigated whether BjCdR15 is involved in cadmium tolerance and whether it can functionally replace TGA3 protein in Arabidopsis tga3-2 mutant plants.
  • BjCdR15 expression was detected mainly in the epidermis and vascular system of cadmium-treated plants, and increased in roots and leaves after cadmium treatment. The overexpression of BjCdR15 in Arabidopsis and tobacco enhanced cadmium tolerance: overexpressing plants showed high cadmium accumulation in shoots. Conversely, Arabidopsis tga3-2 mutant plants showed high cadmium content in roots and inhibition of its transport to the shoot.
  • We demonstrated that BjCdR15 can functionally replace TGA3: in 35S::BjCdR15-tga3-2 plants, the long-distance transport of cadmium from root to shoot was restored and these plants showed an increased cadmium content in shoots compared with all other assays. In addition, BjCdR15/TGA3 regulated the synthesis of phytochelatin synthase and the expression of several metal transporters.
  • The results indicate that BjCdR15/TGA3 transcription factors play a crucial role in the regulation of cadmium uptake by roots and in its long-distance root to shoot transport. BjCdR15/TGA3 may thus be considered as useful candidates for potential biotechnological applications in the phytoextraction of cadmium from polluted soils.


Cadmium is a heavy metal commonly released into the environment from industrial processes and farming practices (Hüttermann et al., 1999). It is highly toxic and is readily taken up by plants, and so its accumulation in the food chain may pose serious threats to human health (Buchet et al., 1990). Cadmium accumulation in plants affects general root and shoot growth: it causes a reduction in photosynthesis, diminishes water and nutrient uptake (Sanità di Toppi & Gabrielli, 1999) and inhibits enzyme activities with subsequent disruption to cell transport processes and disturbance of cellular redox control (Clemens, 2001; Schützendübel & Polle, 2002). However, a number of plants, endemic to metal-rich soils, have evolved naturally selected metal hypertolerance, and can thrive in soil contaminated by metals. Some of these plants are also hyperaccumulators of cadmium and other heavy metals, which they can accumulate in shoots to a level orders of magnitude higher than that normally found in plants (Baker & Brooks, 1989).

In view of the risks caused by cadmium, and other heavy metals, to the environment and humans, there has been an interest in developing the use of plants to extract these metals from contaminated soils. Indeed, the identification of more than 400 species that can tolerate and accumulate high concentrations of heavy metals in their above-ground tissues (Reeves & Baker, 2000) suggests that the genetic potential for phytoextraction exists. Unfortunately, most of the hyperaccumulator species identified are small and slow growing, severely limiting their use for large-scale soil decontamination (Ebbs et al., 1997). A suggested means to overcome this limitation, making phytoremediation a feasible technology, is to transfer the genes responsible for the tolerance and hyperaccumulation of heavy metals to fast-growing, high-biomass plants (Meagher, 2000). Although progress has been made towards the achievement of this goal, and transgenic plants with enhanced tolerance to and accumulation of cadmium and other heavy metals have been produced (Zhu et al., 1999; Domínguez-Solis et al., 2001; Gong et al., 2003; Song et al., 2003), our molecular knowledge about metal hyperaccumulation remains limited, and investigation is still needed to understand the genetic mechanisms responsible for heavy metal accumulation and detoxification. Genes belonging to different functional categories have been characterized in relation to their role in cadmium tolerance and accumulation. For instance, the ABC transporter ATM3 from Arabidopsis thaliana contributes to cadmium resistance and accumulation probably by mediating the transport of cadmium–glutathione conjugates across the mitochondrial membrane (Kim et al., 2006). The P1B-type ATPase HMA4 plays a role in cadmium detoxification, being involved in its transport from root to shoot (Mills et al., 2003, 2005; Verret et al., 2004, 2005). A major role of HMA4 in zinc hyperaccumulation and cadmium and zinc hypertolerance has recently been demonstrated in the hyperaccumulator species Arabidopsis halleri (Hanikenne et al., 2008). Phytochelatins have also been studied extensively for their role in heavy metal detoxification (Goldsbrough, 1998; Clemens et al., 1999; Cobbett, 2000), and phytochelatin synthases (PCSs) have been overexpressed in plants, although the results obtained are contradictory. For instance, ectopic expression of AtPCS1 in A. thaliana does not lead to increased cadmium tolerance and accumulation (Lee et al., 2003), whereas, in tobacco and Brassica juncea, its overexpression increases cadmium tolerance (Pomponi et al., 2006; Gasic & Korban, 2007).

The characterization of the heavy metal-binding protein Cdl19 indicated that this protein plays a role in metal homeostasis and confers cadmium tolerance when overexpressed in Arabidopsis (Suzuki et al., 2002). In Thlaspi caerulescens, quantitative trait loci mapping allowed the identification of quantitative trait loci for cadmium accumulation in roots and shoots (Deniau et al., 2006).

The presence of high concentrations of heavy metals in soil is perceived by plant cells as a stress signal and transduced in the expression of different classes of proteins. This signal transduction pathway is controlled by an intricate interaction of genes in which transcription factors play crucial roles. Hence, modifications in the expression of genes coding for transcription factors strongly affect the plant stress response. Among transcription factor families, bZIP genes are well represented in the Arabidopsis genome, and molecular studies of several of them have shown that they are involved in various biological processes, plant defence and environmental challenges (Jacoby et al., 2002). TGA factors in Arabidopsis represent a subfamily of bZIP transcription regulators involved in the expression of pathogenesis-related genes and in the induction of systemic acquired resistance (Zhang et al., 2003). Early studies have indicated that these transcription factors bind to as-1-type elements that induce transcription in response to salicylic acid and xenobiotic stress cues (Liu & Lam, 1994; Qin et al., 1994), and have also suggested that this type of element may confer cadmium-sensitive promoter activity (Kusaba et al., 1996). A functional role for TGA factors in cadmium-responsive gene expression has never been reported. Furthermore, it has been demonstrated that yeast bZIP proteins mediate cadmium resistance (Wu et al., 1993), and that CgAP1, a bZIP gene from Candida glabrata, is involved in multidrug resistance, and the disruption of CgAP1 induces hypersensitivity to cadmium chloride (Chen et al., 2007).

In an effort to identify cadmium-modulated genes, we used the cDNA-amplified fragment length polymorphism (cDNA-AFLP) technique on B. juncea (Fusco et al., 2005). Our previous analysis had shown that BjCdR15 from B. juncea is up-regulated in plants treated for 6 h with cadmium (Fusco et al., 2005), and it is a putative ortholog to Arabidopsis TGA3. We therefore investigated whether this gene is involved in cadmium tolerance and accumulation, and whether the loss-of-function tga3-2 mutant can be replaced by BjCdR15. In this study, we showed that the constitutive overexpression of BjCdR15 in Arabidopsis and tobacco increased cadmium accumulation in the shoot; these transgenic plants showed higher shoot biomass and less chlorosis than control plants when treated with cadmium, indicating that this gene is involved in cadmium tolerance. Furthermore, we showed that, in the tga3-2 mutant, the TGA3 function in terms of cadmium uptake and transport to the shoot is restored by BjCdR15.

Materials and Methods

Plant material

Plants of Brassica juncea (L.) Czern (cv. Aurea) were cultured in hydroponic solution, treated at different time points with 10 μm Cd(NO3)2 and maintained in glasshouse conditions, as described previously (Fusco et al., 2005). For in vitro tests, seeds of Nicotiana tabacum (cv. Petit Havana SR1) and A. thaliana (Col-0) were cultured in vitro on Murashige and Skoog medium (Murashige & Skoog, 1962) under a 16 h light : 8 h dark regime at 22°C : 18°C.

Identification of the tga3-2 mutant

An Arabidopsis tga3 mutant line was found in the SALK collection (tga3-2, SALK_088114) (; Alonso et al., 2003). The homozygous mutant was identified by PCR using a combination of gene- and T-DNA-specific primers, as described previously ( The knockout mutation was confirmed by PCR analysis on cDNA obtained from homozygous mutant plants.

In situ hybridization

In situ hybridization experiments were conducted as described by Varotto et al. (2003). Roots and leaves of B. juncea plants treated with cadmium for 6 h were fixed. Sections (7–10 μm) were cut using a microtome RM 2135 (Leica, Nussloch, Germany). A 470 bp fragment of BjCdR15 cDNA, containing 110 bp of the 5′ untranslated region and part of the translated region, was cloned into the pBluescript vector (Stratagene, La Jolla, CA, USA). To obtain DIG-UTP (Roche Applied Science, Mannheim, Germany), in vitro transcription of labeled sense and antisense RNA probes was performed using T7 and T3 polymerases. DIG detection and signal visualization were performed using nitro blue tetrazolium chloride and 5-bromo-4-chloro-3-indolyl phosphate stain following the manufacturer’s instructions (Roche Applied Science). Images were acquired using a DC 300F camera (Leica).

Rapid amplification of cDNA ends (RACE)

One microgram of total RNA, isolated from leaves of 4-wk-old B. juncea maintained in hydroponic culture and treated for 6 h with 10 μm Cd(NO3)2, was used for the synthesis of 5′-RACE-ready cDNA and 3′-RACE cDNA (SMART™ RACE cDNA Amplification Kit, Clontech, Palo Alto, CA, USA). The primers 5′-GAGAATCAATCACTCTACATC-3′ (forward) and 5′-GCTCTGAGTTCTCTTTGGG-3′ (reverse), designed on a 288 bp transcript-derived fragment (accession no. DT317671) identified and isolated by a cDNA-AFLP approach in B. juncea (Fusco et al., 2005), were used for the 5′ and 3′ ends, respectively. All steps were performed according to the manufacturer’s protocol.

Cloning and generation of BjCdR15-overexpressing plants and complementation analysis of the tga3-2 mutant

BjCdR15 cDNA was amplified by PCR using the specific primers 5′-AGCTTCGTCCTTTCGATCTTC-3′ and 5′-CCTAGCTAACATCATTCGCG-3′.

The amplified fragment was ligated into the binary vector pBI121 (Clontech), under the control of the cauliflower mosaic virus 35S promoter (35S::BjCdR15-WT), and used to transform Arabidopsis (Col-0) and tobacco. To verify whether BjCdR15 can complement the function of TGA3, BjCdR15 cDNA was inserted under the control of the cauliflower mosaic virus 35S promoter into the pCambia 1302 binary vector, and the pCambia-35S::BjCdR15 construct was used for the transformation of the Arabidopsis tga3-2 mutant (35S::BjCdR15-tga3-2). As control lines for each experiment, Arabidopsis and tobacco wild-type plants were transformed with the pBI121 empty vector. Three independent lines for each construct, used to transform Arabidopsis and tobacco, were then considered for further experiments.

In addition, the effect of BjCdR15 under the promoter of TGA3 in the tga3-2 mutant background was tested. For this analysis, the construct pTGA3::BjCdR15::NosT was prepared. BjCdR15 cDNA was amplified from B. juncea using two specific primers (5′-ACTCTAGAATGGAGATGATGAGCTCTTC-3′ and 5′-GTGTCGCCTAAGGCGTTCTCGTGGAC-3′) and cloned into pCambia 1304 upstream from the previously inserted nopaline synthase terminator. A fragment of 1700 bp of the TGA3 promoter was obtained by PCR on A. thaliana genomic DNA using specific primers (5′-CACCAAGGTTTATTACGAAATTTTAGAC-3′ and 5′-AAGCTTGAGATGGATCCTG ATAAGCAGAGAAGAAATTCAT-3′), and cloned into pENTR/D-TOPO (Invitrogen, Carlsbad, CA, USA). The fragment BjCdR15::NosT, derived from pCambia 1304, was finally cloned downstream from pTGA3 in pENTR/D-TOPO, and the entire cassette pTGA3::BjCdR15::NosT was transferred via LR reaction to the pHGW Gateway Vector following the procedure reported in the user manual (Invitrogen).

Nicotiana tabacum grown in vitro was used for plant transformation, as described by Horsch et al. (1985), whereas Arabidopsis was transformed by floral dipping (Clough & Bent, 1998). The presence and expression of the transgene were confirmed by PCR and reverse transcriptase-PCR (Fig. S4, see Supporting Information).

The values for two lines of each genotype are reported in the figures, whereas the entire dataset of three independent lines is presented in Supporting Information (Tables S1–S4).

RNA isolation, cDNA synthesis and quantification of transcription by real-time PCR

Poly (A)+ RNA was prepared for both cadmium-treated and untreated whole B. juncea plants by chromatography on oligo dT-cellulose (Bartels & Thompson, 1983).

Total RNA was isolated from whole tobacco and Arabidopsis plants, or from separately harvested shoots and roots of Arabidopsis and B. juncea, using Trizol Reagent (Invitrogen). cDNAs were prepared using ImProm-II™ Reverse Transcriptase, as indicated by the manufacturer (Promega, Madison, WI, USA). Real-time PCRs were performed using the ABI Prism sequence detection system (Applied Biosystems, Foster City, CA, USA) with the Platinum SYBR Green qPCR SuperMix UDG (Invitrogen). The analysis was conducted separately on leaves and roots of B. juncea plants grown in hydroponic culture and treated with 10 μm Cd(NO3)2, or with either 10 μm Pb(NO3)2 or 10 μm Ni(NO3)2. For all experiments, real-time PCRs were performed as triplicates on three different RNA samples isolated independently from each tested condition. The data were organized according to the 2–ΔΔCT method for relative gene expression analysis (Livak & Schmittgen, 2001).

Intracellular protein localization

BjCdR15 cDNA was fused 5′ to the gene coding for the red fluorescent protein from the reef coral Discosoma (dsRED) in the pGJ1425 vector (Jach et al., 2001), and the construct was used to transfect tobacco protoplasts. Mesophyll protoplasts were isolated from in vitro-grown tobacco plants, as in Negrutiu et al. (1987). Freshly isolated protoplasts (4 × 104 ppt μl−1) were transfected with 10 μg of plasmid DNA by polyethylene glycol-mediated DNA uptake (Walden et al., 1994). Five hours before the analysis, Hoechst 33258 (Sigma-Aldrich) was added (5 μg ml−1) to the protoplast culture in order to stain the nuclear DNA. Transfection with the empty vector (pGJ1425) was used as control.

Measurements of cadmium content and assays of cadmium sensitivity

Plants were grown in hydroponic culture and treated with 10 μm Cd(NO3)2 for 3 wk. The cadmium content was then determined, as described previously (Fusco et al., 2005). Seeds of transgenic Arabidopsis and tobacco plants were sown in vitro on Murashige and Skoog medium. After 3 wk, the plantlets were transferred to the same medium supplemented with 200 or 400 μm Cd(NO3)2 for Arabidopsis and tobacco, respectively, to observe shoot development and leaf morphology, and to measure shoot fresh weight and chlorophyll content. The latter was determined as described in DalCorso et al. (2008).

Microscopic analyses

Roots of Arabidopsis and tobacco plants grown in vitro were observed with an inverted Olympus IX70 microscope (St. Louis, MO, USA), and images were acquired with a JVC KI-58 CCD camera. The transfected tobacco protoplasts were observed with a Leica TCS SP2 laser confocal microscope. Images were collected frame by frame with the acousto-optical tuneable filter using argon/krypton and helium/neon lasers.

Western blot

Total proteins were extracted by grinding in liquid nitrogen leaves of 3-wk-old A. thaliana plants treated with cadmium (20 μm) at different time points (0, 24, 72 and 120 h). The powder was resuspended in extraction buffer [2% sodium dodecylsulfate (w/v), 62.5 mm Tris-HCl (pH 6.8), 10% glycerol (v/v), 2.86 mβ-mercaptoethanol] and centrifuged for 20 min at 10 000 g. The supernatant was boiled for 10 min and the proteins were separated by sodium dodecylsulfate-polyacrylamide gel electrophoresis and transferred to poly(vinylidene difluoride) membranes, as described previously (Sambrook et al., 1989). After blocking, membranes were probed with anti-AtPCS1 antibodies. Signals were visualized on X-ray film by means of chemiluminescence (Amersham ECL Plus Detection kit, Uppsala, Sweden). AuroDye (GE Healthcare, Amersham, Buckinghamshire, UK) staining was used as loading control. Densitometric analyses of immunoblot signals were performed using Bio-Rad image software (Quantity one 4.4.1).

Hormonal and stress treatments

Three-week-old plants of Arabidopsis and B. juncea grown on Murashige and Skoog agar medium were transferred onto sterile Whatman 3 MM paper saturated with liquid Murashige and Skoog medium for 2 d. Plants were then subjected to several treatments for 5 h: abscisic acid (0.1 mm abscisic acid); high salinity (250 mm NaCl); high temperature, 42°C; cold temperature, 4°C; 50 μm methyl-jasmonic acid; 1 mm salicylic acid; and drought. Untreated plants were used as a control. After treatments and RNA extraction, cDNAs were prepared and real-time PCRs were performed as described above.

Primers for real-time PCRs

Forward and reverse primers were designed on cDNA sequences of TGA3 (5′-GCTCTCTCTCAAGGCTTAGAT-3′ and 5′-TGCTTGAAGATTCTCCATGGC-3′) and BjCdR15 (5′-TTTAGGAGAGTACTTCCACAG-3′ and 5′-TTGACTCTTCTTCGTTGTGCTT-3′). To examine the expression of AtNramp3, AtHMA4, AtMRP3, AtPDR8 and AtATM3 transporters in roots of Arabidopsis plants following different exposure times to cadmium (20 μm), 3-wk-old plants were treated at different time points. Real-time PCRs were performed using the following specific primers: AtNramp3, 5′-GTAAGCGTAGTAGCTATGTCT-3′ and 5′-CAAAACCAGAGCTGCTTTATAA-3′; AtHMA4, 5′-AAATGTTCCAGCAAAGGCAGT-3′ and 5′-GGAATTGCAATACATAAAGTACT-3′; AtMRP3, 5′-AGCCACAACACTAGCTGGAT-3′ and 5′-GGGTTCAACAGTAGTAGAAGC-3′; AtPDR8, 5′-TCTCTCAAACAACAAAGTCTCT-3′ and 5′-ATGACACTACAATACATATTTAGA-3′; AtATM3, 5′-TCTCATTGATTTTCTATGGAGC-3′ and 5′CTCGTGTTAAAATTATATCATTCA-3′. Quantifications of transcript were normalized to the expression of β-actin cDNA (primers 5′-GAACTACGAGCTACCTGATG-3′ and 5′-CTTCCATTCCGATGAGCGAT-3′ for Arabidopsis; primers 5′-TGTTCTGGACTCTGGTGATG-3′ and 5′-AGGATCTTCATGAGGTAATCAG-3′ for B. juncea; primers 5′-ATCCCAGTTTGCTGAGAATAC-3′ and 5′-GGCCCGCCATACTGGTGTGAT-3′ for tobacco).

Statistical analysis

Data were obtained in at least three replications and analyzed using analysis of variance with a threshold P value of 0.05 (*), 0.01 (**) or 0.001 (***). Means were compared by the Student–Newman–Keuls’ method using COSTAT software (CoHort Software, Monterey, CA USA).


Cloning and sequence analysis of BjCdR15 from B. juncea

A transcription-derived fragment of 288 nucleotides (DT317671) was isolated by the cDNA-AFLP technique from B. juncea treated for 6 h with 10 μm Cd(NO3)2 (Fusco et al., 2005). PCR-based methods, including 5′- and 3′-RACE, allowed us to clone a 1511 bp cDNA of BjCdR15. Sequence analysis revealed a predicted full-length protein corresponding to an open reading frame of 1161 bp (National Center for Biotechnology Information accession no. EU110098). The predicted protein sequence, 386 amino acids, contains a bZIP signature (Fig. 1a). BjCdR15 sequence alignments revealed high sequence identity with Arabidopsis TGA3 (At1g22070) bZIP transcription factor (nucleotide identity of 89% and amino acid identity of 86%, with similarity of 92%) and TGA4 from Zea mays (84% and 85% for nucleotide and amino acid identity, respectively) (Fig. 1a). In addition, a phylogenetic tree showing the relationships between bZIP transcription factors points to BjCdR15 as a putative ortholog of TGA3 (Fig. 1b).

Figure 1.

 Sequence comparison between BjCdR15 and other bZIP proteins. (a) Amino acid sequence alignment of BjCdR15, TGA3 (Arabidopsis thaliana) and TGA4 (Zea mays). The NLS box indicates the nuclear localization signal and the bZIP-D box represents the signature domain of the bZIP transcription factors, class D, of A. thaliana (Jacoby et al., 2002). *Amino acids of the leucine zipper domain. (b) Phylogenetic tree (based on the unweighted pair group method with arithmetic averaging) showing the relationships between BjCdR15 and other bZIP transcription factors identified in nine plant species. The bootstrap percentages based on 500 replications are indicated; phylogenetic analyses were conducted in MEGA4 (Tamura et al., 2007).

BjCdR15 expression pattern and localization

To verify the data obtained by the cDNA-AFLP technique, RNA blot analysis was performed on whole B. juncea plants using BjCdR15 cDNA as probe. It was confirmed that, after a 6 h treatment with 10 μm Cd, BjCdR15 was up-regulated and the level of transcription decreased with prolonged cadmium exposures (Fig. 2a). The expression pattern of BjCdR15 was also monitored by real-time PCR in roots and shoots at different time points during the first 9 h after cadmium addition. The transcript level increased in both leaves and roots, and the highest BjCdR15 mRNA amount was detected after 2 and 6 h in leaves and roots, respectively. The up-regulation of BjCdR15 was also observed in roots after 6 h of nickel and lead treatment (Fig. 2b).

Figure 2.

BjCdR15 expression analysis. (a) RNA blot analysis of poly(A)+ RNA (2 μg per lane). Poly(A)+ RNA was isolated from whole plants after cadmium exposure for 0 h (1), 6 h (2), 24 h (3) and 6 wk (4). (b) Analysis of the expression pattern of BjCdR15 (open bars, roots; filled bars, leaves) at different time points of cadmium treatment by real-time PCR. BjCdR15 expression was also tested after treating plants for 6 h with nickel and lead. Bar, SE. (c) In situ hybridization of BjCdR15 mRNA on leaf and root tissues of cadmium-treated Brassica juncea plants. Cross-section of leaves (1 and 2) and root (3). Blue staining represents the hybridization signal. Bars: 1 and 3, 100 μm; 2, 50 μm. (d) Fluorescence field of a mesophyll tobacco protoplast transfected with the dsRED protein (control) showing the cytosolic localization (panel 1). Nuclear localization of BjCdR15–dsRED fusion protein (panel 2). Fluorescence generated by the nuclear DNA dye Hoechst 33258 (panel 3). Bright-field view (panel 4).

In situ hybridization was performed to determine the precise localization of BjCdR15 expression in leaves and roots of B. juncea treated for 6 h with cadmium. Leaf cross-sections showed BjCdR15 mRNA in the epidermis and in the internal and external phloem veins (Fig. 2c, panels 1 and 2), whereas, in cross-sections of mature roots, the hybridization signal was detected in the epidermis and vascular cylinder (Fig. 2c, panel 3). In untreated plants, BjCdR15 was faintly expressed in leaves and undetectable in roots (data not shown).

To gain an insight into the subcellular localization of BjCdR15 protein, the BjCdR15 complete coding sequence was fused to the red fluorescent protein (dsRED) sequence and transfected into tobacco protoplasts. Cells harboring the control dsRED protein showed red fluorescence throughout the cytoplasm (Fig. 2d, panel 1). By contrast, protoplasts transfected with the fusion construct exhibited red fluorescence mainly in the nucleus, and the signal overlapped with the fluorescence generated by the Hoechst nuclear DNA dye (Fig. 2d, panels 2 and 3). This result, together with the high sequence similarity with well-known TGA factors (Fig. 1a), demonstrates that the BjCdR15 product is a nuclear-localized protein.

Cadmium induction of TGA3 and Arabidopsis tga3-2 mutant identification

Given the high similarity between BjCdR15 and TGA3 proteins, we tested whether the expression of TGA3 is modulated by cadmium treatment. In Arabidopsis, transcription of TGA3 is increased in the shoots and roots after 2 h of cadmium treatment (Fig. 3a). It should be noted that the expression of TGA3 in A. thaliana is lower than that of BjCdR15 in B. juncea, although a similar trend is maintained.

Figure 3.

Arabidopsis thaliana TGA3 expression analysis. (a) Analysis of the expression pattern of TGA3 at different time points of cadmium treatment by real-time PCR. Open bars, TGA3 in roots; filled bars, TGA3 in leaves; ± SE. (b) Left: diagram of the T-DNA insertion mutant tga3-2, mapped inside the 5′ untranslated region, 166 bp upstream of the start codon (the T-DNA insertion is not drawn to scale). Right: reverse transcriptase PCR analysis on wild-type (WT) and tga3-2 mutant cDNA using TGA3-specific primers and Actin1-specific primers as a control.

We identified an Arabidopsis mutant line (tga3-2, Fig. 3b, left) that harbored a T-DNA in the TGA3 sequence (during the preparation of the manuscript the characterization of another tga3 mutant, tga3-1, was reported by Kesarwani et al., 2007). Segregation analysis on selective medium and Southern blot experiments revealed the presence of a single insertion locus (not shown). No TGA3 transcripts were detected by reverse transcriptase PCR in homozygous tga3-2 plants (Fig. 3b, right), indicating the complete loss of gene function. Under glasshouse growth conditions, mutant plants showed no morphological or developmental abnormalities.

BjCdR15 functionally replaces Arabidopsis TGA3

The cadmium induction of BjCdR15 prompted us to investigate whether this transcription factor could be involved in cadmium tolerance and accumulation; thus, wild-type Arabidopsis and tobacco plants were transformed with the construct harboring the 35S::BjCdR15 cassette. As a result of the high similarity between BjCdR15 and TGA3, and given that TGA3 is also induced by cadmium, the tga3-2 mutant line was also transformed with a construct carrying the 35S::BjCdR15 cassette. For both Arabidopsis and tobacco, wild-type plants transformed with the empty vector (pBI121) were considered as control lines.

We measured the cadmium content in both the shoots and roots of control and BjCdR15 overexpressing plants grown in hydroponic culture and treated with cadmium for 3 wk. In 35S::BjCdR15-WT plants, the average cadmium content in the shoots was c. 13% higher than that in controls. Shoots of tga3-2 plants showed significantly lower cadmium content than both control plants (< 0.001) and 35S::BjCdR15-WT plants (< 0.001). Ectopic expression of BjCdR15 in the tga3-2 mutant caused a significantly higher increase in cadmium accumulation in shoots than in all other assays (< 0.001) (Fig. 4a, left). Expression of the TGA3 coding sequence in the tga3-2 background, under the control of the 35S promoter of cauliflower mosaic virus, restored the control plant phenotype with regard to cadmium accumulation in tga3-2 plants (data not shown), confirming that the observed phenotypes are caused by the interruption of the TGA3 gene.

Figure 4.

 Measurements of cadmium content. (a) Shoots and roots of Arabidopsis lines treated with 10 μm Cd. Values shown are the average of 50 plants per three replicates (+ SE). (b) Shoots and roots of tobacco plants treated with 10 μm Cd. Values are the average of 25 plants per three replicates (+ SE). Significant differences from control plants are indicated by: *< 0.05; ***< 0.001.

Notably, the accumulation of cadmium in the roots of Arabidopsis plants was similar in all tested conditions, except for the tga3-2 mutant, in which the cadmium content was consistently higher (< 0.001) (Fig. 4a, right).

These observations suggest that the BjCdR15 gene is an ortholog of Arabidopsis TGA3, as its product functionally replaces TGA3 activity in tga3-2 mutant plants and restores control plant cadmium contents in the tga3-2 background. To reinforce this hypothesis, tga3-2 knockout plants were transformed with BjCdR15 under the control of the TGA3 promoter, and it was observed that the cadmium content in shoots and roots was similar to the cadmium accumulation in control plants (Fig. S1).

Similar results were obtained with tobacco plants: shoots of 35S::BjCdR15 plants showed significantly higher cadmium content (< 0.05) than the shoots of control plants. There was no significant difference in cadmium content in roots (Fig. 4b). These results indicate that BjCdR15-expressing plants transport more cadmium to the shoot than do control plants.

BjCdR15 overexpression enhances cadmium tolerance

To evaluate the cadmium tolerance of the transformants, 3-wk-old Arabidopsis and tobacco plants were grown on Murashige and Skoog medium containing 200 and 400 μm Cd(NO3)2, respectively, for a further 3 wk. In the medium without cadmium, the growth of tga3-2 knockout plants, BjCdR15 overexpressing lines and control plants was similar (data not shown). In the presence of cadmium, Arabidopsis and tobacco lines overexpressing BjCdR15 grew better than control plants (Fig. 5a,b). Leaves were greener and broader, and showed less chlorosis (Fig. S2). Control plants also flowered earlier, apparently induced by the stress conditions. After 3 wk of cadmium treatment, control plants were severely stressed and highly chlorotic, whereas symptoms of cadmium toxicity were barely noticeable in plants constitutively expressing BjCdR15 (Fig. 5a,b). Grown in cadmium-supplemented medium, the tga3-2 mutant showed a phenotype similar to control plants (Fig 5a).

Figure 5.

 Phenotypes of plants grown in vitro on cadmium-containing medium. (a) Arabidopsis plants were cultivated in vitro on 200 μm Cd-containing medium for 3 wk. Light micrographs of the roots are depicted and the white arrows highlight the lateral primordia. (b) Tobacco plants were cultivated in vitro on 400 μm Cd-containing medium for 3 wk. Light micrographs of the roots are depicted.

Systematic inspection of the roots of plants grown in cadmium medium showed, in Arabidopsis control plants, an abundance of root hairs, accompanied by the presence of numerous lateral root primordia that did not develop further (Fig. 5a). A similar phenotype was observed in cadmium-treated tga3-2 plants (Fig. 5a). In control tobacco plants, the induction of numerous root hairs was observed, particularly in the region behind the root tip (Fig. 5b). Interestingly, the cadmium treatment of both Arabidopsis and tobacco plants overexpressing BjCdR15 did not induce any particular phenotype, and the root morphology was similar to that of BjCdR15-overexpressing plants not treated with cadmium (data not shown).

BjCdR15 overexpression enhances chlorophyll content and biomass

Plants grown in medium without cadmium showed similar chlorophyll contents. In the presence of cadmium, BjCdR15-overexpressing Arabidopsis (wild-type background) showed a significantly higher chlorophyll content than did control plants (< 0.001). Conversely, the tga3-2 mutant displayed a similar chlorophyll content to control plants. The expression of BjCdR15, in the tga3-2 background, showed a significant increase compared with mutant plants (< 0.001) (Fig. 6a, left). A similar effect of the transgene was observed in cadmium-treated tobacco plants, in which the chlorophyll level of BjCdR15-overexpressing plants was higher than that in the controls (< 0.01) (Fig. 6a, right).

Figure 6.

 Chlorophyll content and shoot fresh weight of plants grown in vitro on cadmium-containing medium. (a) Chlorophyll content in Arabidopsis plants (left) cultivated in vitro on 200 μm Cd-containing medium and tobacco plants (right) cultivated in vitro on 400 μm Cd-containing medium, for 3 wk. The mean values (+ SE) of four plants in each line, grown with and without cadmium, are reported. (b) Shoot fresh weight in Arabidopsis (left) and tobacco (right) plants. Values shown are the average of three replicates (10 plants each) for Arabidopsis and three replicates (three plants each) for tobacco.

Biomass accumulation, measured as shoot fresh weight, of controls, BjCdR15-overexpressing lines and the tga3-2 mutant, was similar in the absence of cadmium, but significant differences were found after cadmium treatment. In Arabidopsis, the biomass of BjCdR15-overexpressing lines increased (< 0.001) compared with control and tga3-2 plants (Fig. 6b, left). Similar results were observed for tobacco (< 0.05) (Fig. 6b, right).

BjCdR15 and TGA3 affect the expression of Arabidopsis PCS (AtPCS1)

AtPCS1 protein abundance was evaluated in shoots of cadmium-treated plants by Western analysis. In a time-course study with 3-wk-old Arabidopsis plants, in the absence of cadmium, a basal expression level was found for all genotypes tested, although BjCdR15-overexpressing lines showed a higher level of AtPCS1, indicating that the ectopic expression of BjCdR15 enhances the synthesis of PCS. An increase in the enzyme level was observed in control plants following cadmium addition (Fig. 7a,b). Notably, tga3-2 mutant plants showed the lowest amount of AtPCS1, and were cadmium insensitive for the accumulation of AtPCS1 protein, and the expression of BjCdR15 in the tga3-2 line restored a higher level of AtPCS1 protein independent of cadmium addition (Fig. 7).

Figure 7.

 Expression analysis of AtPCS1 protein. (a) Immunoblot analysis using anti-AtPCS1 on leaf total protein extracts from Arabidopsis plants at different time points of cadmium treatment (0, 24, 72 and 120 h). The lower panel shows AuroDye staining as loading control. (b) Densitometric analysis of the immunoblot signals depicted in (a). A similar trend was observed for all three independent biological repetitions performed.

Transporters are regulated by BjCdR15/TGA3

As various types of transporter have been shown to be involved in heavy metal detoxification mechanisms, transcripts of AtNramp3, AtHMA4, AtMRP3, AtPDR8 and AtATM3 were quantified by real-time PCR in the roots of Arabidopsis control plants, 35S::BjCdR15-overexpressing lines and tga3-2 mutant plants following cadmium treatment. A similar expression modulation was observed for each transporter in all genotypes during cadmium treatment (Fig. 8), with the exception of AtATM3, which was not affected by the loss of function of TGA3, and was similarly up-regulated in all genotypes.

Figure 8.

 Expression analysis of metal transporters. AtNramp3, AtHMA4, AtMRP3, AtPDR8 and AtATM3 transcript levels in the roots of Arabidopsis plants at different time points during cadmium treatment (0, 24, 120 h) by real-time PCR. Data are the means of two independent biological experiments (± SE).

AtNramp3, AtHMA4, AtMRP3 and AtPDR8 were shown to be induced by cadmium in control and BjCdR15-overexpressing plants, whereas their expression was cadmium insensitive in the tga3-2 mutant. Overexpression of the BjCdR15 protein in the tga3-2 mutant restored the cadmium induction of these genes. AtPDR8 was an exception to this behavior; indeed, in the absence of cadmium (i.e. 0 h Cd), its expression was not restored by BjCdR15 overexpression (see Discussion).

Expression of BjCdR15 and TGA3 mRNA in response to various treatment regimes

The expression of BjCdR15 and TGA3 was further investigated by comparing, in B. juncea and Arabidopsis, respectively, mRNA levels in response to different hormone and stress treatments by real-time PCR. As shown in Fig. S3, both bZIP transcription factors were induced by cold temperature (4°C) and almost unaffected by high temperature (42°C). These genes were also induced by other abiotic stresses, such as salt (NaCl) and drought. Abscisic acid treatment revealed that both BjCdR15 and TGA3 were responsive to this factor, although TGA3 was more sensitive than BjCdR15 to this hormone. In addition, it was observed that these genes were not induced by methyl jasmonate and only TGA3 was up-regulated by salicylic acid.


Molecular genetic studies of some Arabidopsis TGA factors have shown that they are involved in plant development, as well as in stress signaling (Jacoby et al., 2002). It has been demonstrated recently that TGA3 is required for basal defense and the induction of pathogenesis-related genes (Kesarwani et al., 2007). It is also known that TGA3 strongly interacts with NPR1, conferring disease resistance to bacterial and fungal pathogens (Zhou et al., 2000). In this study, we describe the characterization of a cDNA encoding a B. juncea bZIP transcription factor, BjCdR15, orthologous to TGA3, up-regulated by cadmium (Fusco et al., 2005). We provide evidence that implicates BjCdR15 in cadmium tolerance, accumulation and translocation to the shoot. Firstly, using RNA analyses and real-time PCR, it was shown that, in B. juncea, BjCdR15 is induced by cadmium and other heavy metals and is mainly expressed in the vascular system and epidermal cells. Secondly, it was demonstrated that Arabidopsis and tobacco plants overexpressing BjCdR15 grow better and are significantly more tolerant to cadmium exposure than are control plants; indeed, overexpressing plants show higher chlorophyll content and shoot fresh weight and accumulate more cadmium in the shoot than do control plants. Thirdly, it was demonstrated that the BjCdR15 protein plays a role in long-distance root to shoot cadmium transport and in the induction of several metal transporters. In Arabidopsis, the TGA3 protein plays a similar role, as shown by the complementation with BjCdR15 of the tga3 knockout mutant.

A role for BjCdR15 as a transcription factor was deduced from its nuclear localization and its sequence similarity with other Arabidopsis TGA transcription factors. By northern and real-time PCR analyses, BjCdR15 transcription was shown to increase shortly after cadmium exposure and to decline rapidly thereafter. A similar expression pattern was observed for TGA3, although the attenuated transcription may reflect differences in promoter regulatory regions between A. thaliana and B. juncea (see further discussion below).

This early induction on cadmium exposure and rapid shut off, together with the phenotype of BjCdR15-overexpressing plants, suggest a key role for this transcription factor in the gene regulation involved in cadmium tolerance. Furthermore, the in situ localization of BjCdR15, mainly in epidermal cells of cadmium-treated B. juncea plants, points to a possible role of this bZIP factor in controlling both cadmium absorption and storage. The root epidermis layer is the primary contact site between plant and soil, and is where cadmium gains access, whereas the leaf epidermis is a major cadmium storage site in B. juncea (Salt et al., 1995).

In Arabidopsis, lateral root formation was inhibited under cadmium stress in control plants and tga3-2 mutants, whereas it was promoted in BjCdR15 overexpressors. Cadmium exposure also induced the formation of numerous root hairs in both Arabidopsis and tobacco control plants, as well as in the tga3-2 mutant line, whereas the formation of root hairs in BjCdR15-expressing plants was comparable with that of cadmium-untreated plants. As cadmium has inhibitory effects on water and mineral nutrient uptake (Deckert, 2005; Meda et al., 2007), the normal development of roots in plants overexpressing BjCdR15 may represent an adaptive response, in part accounting for the observed increase in cadmium tolerance, maximizing the ability of the root system to uptake water and mineral nutrients even in the presence of cadmium. Conversely, the root structural changes shown by control and tga3-2 plants exposed to cadmium may be interpreted as a symptom of cadmium toxicity. However, the observation that cadmium-treated tga3-2 plants did not differ in root phenotype from control plants does not allow speculation, at this point in the investigation, about the role of the TGA3 transcription factor in lateral root and root hair development.

In the tga3-2 mutant, cadmium treatment did not affect the phenotype, shoot biomass and chlorophyll content more severely than in control plants. This result is not unexpected as the lack of TGA3 protein in the mutant prevents, at least in part, the translocation and accumulation of cadmium to the shoot. Remarkably, tga3-2 plants accumulated c. 1000 mg kg−1 more cadmium in the roots than did control and BjCdR15-overexpressing plants, whereas the cadmium transported to the shoot was only c. 200 mg kg−1 lower than in the control lines. These findings suggest that cadmium uptake by roots is altered in tga3-2 plants. Interestingly, once the function of the TGA3 protein is restored by BjCdR15, cadmium transport to the shoot is reactivated. Indeed, the cadmium content increased in the shoots of overexpressing lines, whereas it was similar to that of control plants in the roots. The effect of BjCdR15 overexpression was different in wild-type and tga3-2 backgrounds. In the latter, BjCdR15 protein caused a higher cadmium loading in the shoots. These data thus indicate that BjCdR15/TGA3 contribute towards maintaining a steady cadmium content in the roots by acting on both cadmium uptake and long-distance root to shoot transport. This is consistent with the in situ BjCdR15 expression, predominantly localized in the vascular system of roots and leaves. Moreover, it has been shown recently that the phloem may play a major role in the long-distance ‘source to sink’ transport of cadmium (Mendoza-Cózatl et al., 2008). The apparent contradiction between increased cadmium accumulation and tolerance has already been noted (Verret et al., 2004), and it has been observed that the major part of Arabidopsis biomass is formed by the shoot, which has large apoplastic and vacuolar compartments to enhance metal sequestration.

High cadmium accumulation in the roots has been reported for the Arabidopsis phytochelatin-deficient mutant cad1-3. It has been proposed that a phytochelatin (PC)-dependent ‘overflow protection mechanism’ would contribute towards keeping the cadmium accumulation in the roots low by causing extra cadmium transport to the shoots (Gong et al., 2003). A possible role of BjCdR15/TGA3 in the cellular mechanisms of cadmium detoxification (PC synthesis and PC–Cd2+ complexation) and transport to the shoot was therefore supposed, and the expression of AtPCS1 was measured. In the absence of cadmium, AtPCS1 protein in control plants was less abundant than in BjCdR15 overexpressors, indicating that the ectopic expression of this bZIP factor perturbs AtPCS1 synthesis. Cadmium addition causes an increase in AtPCS1 expression in control plants, as already noted by Heiss et al. (2003); conversely, the overexpression of BjCdR15 induces a higher AtPCS1 level, independent of cadmium treatment. This higher expression may contribute towards explaining, at least in part, the increased cadmium accumulation in the shoots of BjCdR15 overexpressors. By contrast, the AtPCS1 protein is poorly abundant in tga3-2 plants and does not seem to be affected by cadmium treatment. This result is consistent with the higher amount of cadmium found in the roots of this mutant. These findings together suggest that BjCdR15/TGA3 could be involved in the regulation of PCS synthesis.

Several studies have indicated that ABC transporters could be involved in cadmium efflux, transport, sequestration and/or redistribution (Bovet et al., 2003, 2005; Kim et al., 2007). The P-type ion pump HMA4 has been studied extensively (Verret et al., 2004, 2005; Mills et al., 2005; Talke et al., 2006; Courbot et al., 2007; Hanikenne et al., 2008). It has also been proposed that AtNRAMP3 functions as a metal transporter and can modulate heavy metal toxicity in plants (Thomine et al., 2003). In our work, an increased expression of the metal transporters tested was observed in control plants on cadmium treatment. In the loss-of-function tga3-2 mutant, the absence of this induction, under cadmium supply (with the exception of AtATM3), and the restored upregulation in 35S::BjCdR15-tga3-2 plants, demonstrated that BjCdR15/TGA3 regulate the expression of transporters under cadmium stimulus. However, we cannot exclude the possibility that long-term cadmium treatment induces other mechanisms, such as iron deficiency, which, in turn, contribute to the overexpression of metal transporters (Thomine et al., 2000; Yoshihara et al., 2006). Moreover, the expression of several transporters, mainly in the vascular tissues of roots and leaves and in the epidermis (Thomine et al., 2003; Verret et al., 2004; Kim et al., 2007), is consistent with BjCdR15 expression. With regard to the expression of AtPDR8, BjCdR15 complemented TGA3 function only on cadmium treatment. This observation led us to conclude that TGA3 and BjCdR15 may respond to cadmium by inducing redundant, but distinct, pathways. Noticeably, BjCdR15/TGA3 affect the expression of genes that play different roles as transporters: both AtPDR8 and AtHMA4 contribute to cadmium detoxification as cadmium extrusion pumps (Mills et al., 2005; Kim et al., 2007), whereas the characterization of AtNRAMP3 demonstrates its role in iron and cadmium transport and cadmium sensitivity, probably by driving cadmium efflux from the vacuole to the cytoplasm (Thomine et al., 2003). Experimental data have shown that the AtATM3 transporter has a mitochondrial localization, and is capable of contributing to heavy metal detoxification (Kim et al., 2006).

It should be noted that AtHMA4, AtMRP3, AtPDR8 and AtATM3 genes possess, in their upstream promoter regions, at least one as-1 binding motif (TGACG) sufficient for TGA3 transcription factor recognition (Xiang et al., 1997), whereas AtNramp3 is induced by BjCdR15/TGA3, but does not have the TGA binding sequence in its upstream region. These findings suggest that several metal transporters could be activated directly by TGA3, whereas others may be regulated indirectly. An early report has indicated that the TGACG motif may confer cadmium-sensitive promoter activity (Kusaba et al., 1996), although no relationship between cadmium responsiveness and TGA transcription factors has so far been reported.

It is noteworthy that AtPDR8, a member of the pleiotropic drug resistance subfamily of ABC transporters, is involved in both cadmium resistance and pathogen resistance (Kobae et al., 2006; Stein et al., 2006), even though no direct link between pathogen resistance and heavy metals has yet been found (Kim et al., 2007). It has been suggested that AtPDR8 can recognize a very wide range of compounds that may be toxic to the plant cell when accumulated at high concentration, and transport them out of the plasma membrane (Kim et al., 2007). In this study, we have reported, for the first time, that members of the TGA transcription factor family, TGA3 and BjCdR15, induce the expression of genes responsible for root to shoot cadmium transport and are implicated in the extrusion of cadmium into the apoplastic and vacuolar compartments. These findings, together with the observation that TGA3 interacts with components that mediate disease resistance (Kesarwani et al., 2007), point to the role of TGA3 in regulating the network of plant defense and other stress-induced signaling transduction pathways.

Our results further indicate that regulator sequences of BjCdR15 and TGA3 may have, at least in part, a different response to different stress-associated signaling molecules. Real-time PCR results suggest that the promoter regions which control the transcription of TGA3 in Arabidopsis contain, as reported previously, the specific elements inducible by salicylic acid, a well-known inductor of systemic acquired resistance (Pontier et al., 2001; Kesarwani et al., 2007). Conversely, the transcription of BjCdR15 in B. juncea is not modulated by salicylic acid or methyl jasmonate, suggesting that it is driven by inducible promoter elements, mainly responsive to abiotic stresses. Thus, as TGA3 and BjCdR15 proteins share high sequence identity, are implicated in cadmium transport and accumulation in the shoot and are induced by abiotic stresses, they could be orthologous genes which, during speciation, may have acquired differences in promoter elements that confer them with typical responsiveness to stress situations. Such variability may reflect the divergent evolution of the related species A. thaliana and B. juncea.

In summary, the results reported here show that BjCdR15/TGA3 regulate the expression of the genes responsible for cadmium tolerance and accumulation. The overexpression of BjCdR15 in heterologous systems, such as Arabidopsis and tobacco, enhances cadmium translocation from root to shoot; therefore, because the harvesting of vegetable parts is standard practice, this transcription factor could be a useful candidate for potential biotechnological applications in the phytoextraction of cadmium from polluted soils.


The authors wish to dedicate this paper to their colleague and friend, Dr Lorena Borgato, in recognition of her invaluable contribution to the research before her untimely death. We thank Dr G. Peltier (CNRS-CEA, Aix Marseille II, France) for providing the AtPCS1 polyclonal antibodies, Dr A. Pera (Institute for the Study of Ecosystems-CNR, Pisa, Italy) for the statistical analysis and Dr P. Pesaresi (Department of Vegetable Crops, University of Milan/Tecnoparco Foundation, Lodi, Italy) for helpful suggestions and critical comments on the manuscript.