Expression and functional analysis of metal transporter genes in two contrasting ecotypes of the hyperaccumulator Thlaspi caerulescens (original) (raw)

Abstract

Zinc (Zn) hyperaccumulation is a constitutive property of Thlaspi caerulescens, whereas cadmium (Cd) hyperaccumulation varies greatly among different ecotypes. The molecular basis of this variation is unknown. Ecotypic differences in the sequences and expression of four representative ZIP family transporter genes were investigated. Genome analysis indicated the presence of at least two closely related copies of the TcIRT1 gene in both Ganges (high Cd accumulating) and Prayon (low Cd accumulating) ecotypes, with different copies being expressed in each, and, furthermore, the two genes potentially encode different length transcripts. The predominant transcript in Prayon was truncated, missing sequence coding for the putative metal-binding site and the five C-terminal transmembrane helices. The two ecotypes were grown hydroponically ±Fe and Cd, and mRNA abundance determined for four ZIP genes. The four ZIP genes studied (TcIRT1, TcIRT2, TcZNT1, and TcZNT5) were expressed in roots only. TcIRT1 expression (full-length in Ganges, TcIRT1-1G; truncated in Prayon, TcIRT1-2P) was enhanced by Fe deficiency or by exposure to Cd. TcIRT2 expression was induced by Fe deficiency, but was unaffected by Cd exposure. TcZNT5-G showed greater expression in Prayon compared with Ganges. The functions of TcIRT1 from Ganges and Prayon and the Arabidopsis homologue were analysed by heterologous expression in yeast. All three IRT1 genes were able to facilitate growth on low Fe concentrations. Cd sensitivity of yeast was conferred in the order _AtIRT1_>_TcIRT1-1G_>TcIRT1-2P (truncated). Cd uptake after 4 h was only detectable following complementation by AtIRT1. The results suggest that although TcIRT1-G may be involved in Cd hyperaccumulation in the Ganges ecotype of T. caerulescens, the transporter expressed in yeast does not have an enhanced ability to transport Cd compared with AtIRT1. Therefore, the unique Cd-accumulating ability of the T. caerulescens Ganges ecotype must be due to the levels of expression of the protein or to other factors such as interacting proteins.

Introduction

Metal hyperaccumulator is a term used to define plants that are able to accumulate metals (or metalloids) in their above-ground parts to extremely high concentrations, usually 100–1000-fold higher than those found in normal plant species (Baker et al., 2000; McGrath et al., 2002). Hyperaccumulators are highly tolerant to the metals they accumulate. There has been much interest over the last decade in the metal hyperaccumulation phenomenon, mainly because of the potential use of metal hyperaccumulators for phytoremediation or phytomining (Chaney et al., 1997; McGrath and Zhao, 2003). However, many hyperaccumulators are slow-growing, low-biomass plants and therefore not suitable for the purpose of phytoremediation. One solution would be to transfer genes responsible for metal hyperaccumulation and detoxification to high-biomass plants. This approach requires an understanding of the molecular mechanisms of metal hyperaccumulation.

Thlaspi caerulescens J. & S. Presl (Brassicaceae) is a well-known zinc (Zn) hyperaccumulator, with the ability to accumulate and tolerate >25 000 mg Zn kg−1 dry weight (DW) in the shoots (Baker et al., 1994; Shen et al., 1997). Zn hyperaccumulation is a constitutive and species-wide trait in T. caerulescens, although there is a considerable variation between different populations when grown under controlled conditions (Baker et al., 1994; Meerts and van Isacker, 1997; Escarré et al., 2000; Lombi et al., 2000; Assunção et al., 2003_a_; Roosens et al., 2003).

Thlaspi caerulescens is able to accumulate other metals, including Ni and Cd (Baker et al., 1994). When grown under the same conditions, different populations were found to vary widely in Cd accumulation (Lombi et al., 2000; Roosens et al., 2003). In particular, populations from southern France (Ganges ecotype) were far superior in Cd accumulation compared with those tested from other regions of Europe, including the Prayon population from Belgium that is widely used in physiological and molecular studies (Lombi et al., 2000; Roosens et al., 2003). Studies of metal influx showed the _V_max for Cd uptake to be approximately five times higher in Ganges than Prayon (Lombi et al., 2001; Zhao et al., 2002). This suggested that either the transporters for Cd were more highly expressed in Ganges or that those expressed are more active. The Cd influx kinetics for Ganges also showed a saturable component at low Cd concentrations, which was not seen in Prayon, suggesting the existence of a specific transport system with high affinity for Cd, present predominantly in the Ganges ecotype. Cd uptake was significantly enhanced by Fe deficiency in the Ganges but not the Prayon ecotype (Lombi et al., 2002). In contrast, Zn uptake was not affected by the Fe status of the plants of either ecotype (Lombi et al., 2002). Recent studies using inter-ecotypic crosses of T. caerulescens suggest that multiple genes are involved in the accumulation of Zn (Assunção et al., 2003_b_) and Cd (Zha et al., 2004). The molecular mechanisms responsible for the extraordinary ability to hyperaccumulate Cd in the Southern French Ganges ecotype of T. caerulescens have not been elucidated.

A large number of transporter proteins are involved in the uptake and homeostasis of transition metals in plant cells, including members of the families ZIP (ZRT, IRT-like Proteins), NRAMP (Natural Resistance-Associated Macrophage Proteins), P-type ATPases, and CDF (Cation Diffusion Facilitators) (Mäser et al., 2001). Notable examples are several ZIP transporters that mediate uptake of Fe(II) and Zn. Fifteen potential ZIP genes may be identified in the Arabidopsis thaliana genome. Of these, AtIRT1 has been demonstrated to be essential for Fe acquisition from the soil in A. thaliana (Connolly et al., 2002; Vert et al., 2002), whereas AtZIP1 and AtZIP3 may be responsible for uptake of Zn in the roots (Grotz et al., 1998). Expression of several ZIP genes is regulated by the status of Fe or Zn in plants (Guerinot, 2000). Some of the ZIP transporters are not metal-specific, but can transport different metals with varying affinities; for example, AtIRT1 is able to transport several divalent metal ions, including Cd, Co, Mn, and Zn (Eide et al., 1996; Korshunova et al., 1999; Vert et al., 2002). In Fe-deficient A. thaliana there is evidence for a role for AtIRT1 in Cd uptake; A. thaliana plants engineered to overexpress AtIRT1 accumulated greater amounts of Cd than wild-type plants (Connolly et al., 2002).

TcZNT1 was the first ZIP gene isolated from T. caerulescens Prayon, and mediates high-affinity Zn transport, as well as a low-affinity Cd uptake (Pence et al., 2000). TcZNT1 has since been cloned from T. caerulescens Ganges (Lombi et al., 2002) and TcZNT5 from Prayon (AF292029). TcZNT1 is homologous to the A. thaliana gene AtZIP4, with 76% identity at the amino acid level. TcZNT2 is a close homologue of TcZNT1 (Assunção et al., 2001). TcZNT1 and TcZNT2 showed increased gene expression in T. caerulescens roots compared with the non-hyperaccumulator Thlaspi arvense (Pence et al., 2000; Assunção et al., 2001). Expression of both genes was hardly responsive to the Zn status of T. caerulescens, whereas they were expressed in T. arvense roots only when the plants were Zn-deficient. Constitutively high expression of TcZNT1 and TcZNT2 may explain the ability of T. caerulescens to hyperaccumulate Zn (Pence et al., 2000; Assunção et al., 2001). TcIRT1-G, cloned previously from the Ganges ecotype (Lombi et al., 2002), is a close homologue of AtIRT1, with 90.2% identity at the amino acid level. A previous study showed that the high expression of TcIRT1 under iron deficiency in the roots of Ganges and no expression in Prayon corresponded to the ecotypic difference in the Fe deficiency-induced response in Cd uptake (Lombi et al., 2002). This, together with the fact that AtIRT1 can mediate Cd uptake in A. thaliana, implies a role for TcIRT1 in Cd accumulation in T. caerulescens, particularly the Ganges ecotype. However, the functions of TcIRT1 have not been analysed.

In the present study, four representative ZIP genes from both Ganges and Prayon ecotypes are cloned. Expression patterns in response to different Fe and Cd exposures are presented with the aim of identifying candidate genes that could explain the ecotypic difference in Cd accumulation. Furthermore, the functions of TcIRT1 were analysed using heterologous expression in Saccharomyces cerevisiae.

Materials and methods

Plant materials and culture conditions

Seeds of two ecotypes of T. caerulescens J & C Presl were collected in Belgium (Prayon ecotype) and near St Laurent le Minier, Southern France (Ganges ecotype). The seeds were germinated on a mixture of perlite and vermiculite moistened with deionized water and, after germination, with nutrient solution. After 21 d, seedlings were transferred to 1.0 l vessels (five seedlings per vessel) filled with a nutrient solution containing: 1000 μM Ca(NO3)2, 500 μM MgSO4, 50 μM K2HPO4, 100 μM KCl, 10 μM H3BO3, 1.8 μM MnSO4, 0.2 μM Na2MoO4, 0.31 μM CuSO4, 0.5 μM NiSO4, 100 μM Fe(III)-EDDHA [ethylenediamine-di(_o_-hydroxyphenylacetic acid)] and 5 μM ZnSO4, buffered at pH 6.0 with 2 mM MES (2-morpholineoethanesulphonic acid; solution pH was adjusted with 1 M KOH so that the final concentration of KOH was about 1 mM). Nutrient solution was aerated continuously and renewed every 3 d. All plants were grown in a controlled environment with a 16 h day length, 20/16 °C day/night temperature and 60–70% relative humidity.

Iron treatments

Eight days after plants were transferred to the full nutrient solution, Fe treatments were imposed. Control plants (+Fe) were given full nutrient solution, and –Fe plants were given full nutrient solution without Fe(III)-EDDHA. Before applying the –Fe treatment, plant roots were washed in deionized water to remove any Fe remaining from the full nutrient solution. After 8 d of growth under these conditions plants were harvested, and roots and leaves were separated and frozen in liquid nitrogen. Tissue was stored at −80 °C.

Cadmium treatments

Eight days after plants were transferred to full nutrient solution, 0 or 100 μM CdCl2 was added to the nutrient solution. Each treatment was replicated three times. Plants were harvested on day 8 and stored at −80 °C.

Gene cloning

Cloning of T. caerulescens cDNAs, TcZNT1-G (G indicates Ganges) and TcIRT1-1G, from the Ganges ecotype was described previously (Lombi et al., 2002). The T. caerulescens cDNA fragments reported here were cloned using a 3′RACE system for rapid amplification of cDNA ends (Invitrogen, Paisley, UK) or RT-PCR from total RNA extracted from Ganges and Prayon leaf or root tissue. After first strand cDNA synthesis, TcIRT1-2P (P indicates Prayon) cDNA was amplified by PCR using the AUAP (5′-GGCCACGCGTCGACTAGTAC-3′) RACE primer (Invitrogen, Paisley, UK) and the same primer as used for TcIRT1-1G, which included the start codon (PCR conditions: initial denaturing for 3 min at 95 °C, 40 cycles of 30 s at 95 °C, 20 s at 55 °C, and 90 s at 72 °C with a final 5 min at 72 °C for elongation). TcZNT5 was amplified by RT-PCR using oligonucleotide primer sequences designed to the original accession (AF292029) (PCR conditions: initial denaturing of 3 min 95 °C, followed by 40 cycles of 30 s at 95 °C, 20 s at 55 °C, and 80 s at 72 °C, with a final 5 min at 72 °C for elongation). The T. caerulescens gene fragment, TcIRT2, was amplified via RT-PCR from the Prayon and Ganges ecotypes using oligonucleotide primers designed to the AtIRT2 sequence (At4g19680) (Vert et al., 2001) (PCR conditions: initial denaturing: 3 min 95 °C, followed by 40 cycles of 30 s at 95 °C, 20 s at 55 °C, and 60 s at 72 °C, with a final 5 min at 72 °C for elongation). The primers used for RT-PCR were as follows: TcZNT5 primer 1, 5′-CCCTTAATTCAGCCATTTCT-3′, _TcZNT_5 primer 2, 5′-ATTGCATCTTCGGATGCATG-3′, TcIRT2 primer 1, 5′-GACCTCTCCTCTTTTCAGCC-3′; TcIRT2 primer 2, 5′-GAAACATACCACAAGGTGTT-3′.

After gel purification, the cDNA products from PCR were cloned into the pGEM-T Easy vector (Promega, Madison, WI) for transformation into Escherichia coli. Sequencing of the plasmid insert was performed on an ABI Prism 310 genetic analyser using Big Dye™ sequencing reagents (PE-Applied Biosystems, Foster City, CA).

Expression analysis

RNA was extracted from Prayon and Ganges roots or leaves by the method of Verwoerd et al. (1989). Ten micrograms of total RNA from each treatment was separated on a 1% (w/v) agarose gel containing 3% (v/v) formaldehyde and blotted onto a nylon membrane as described in Sambrook et al. (1989). Membranes were hybridized overnight at 65 °C using TcZNT1-G (AJ313521, 1078 bp; Lombi et al., 2002), TcIRT1-G (AJ320253, 1321 bp; Lombi et al., 2002), TcIRT2-G (AJ53844, 645 bp; this manuscript), and TcZNT5-G (1203 bp) cDNA fragments radiolabelled with [α32P]dCTP using the Prime-A-Gene oligonucleotide labelling system (Promega). Membranes were washed to high stringency in 1× SSC (3 M NaCl, 0.3 M sodium citrate, pH 7.0) and 0.1% (w/v) SDS at 65 °C before visualizing on Biomax MS autoradiography film (Eastman-Kodak, Rochester, NY). Between successive hybridizations, probes were stripped from nylon membranes with boiling 0.1% (w/v) SDS. Equal loading of RNA in each lane was verified by ethidium bromide staining.

Genomic DNA extraction

To obtain genomic DNA, 1.0–1.5 g of frozen T. caerulescens shoot tissue was ground to a fine powder and extracted in 0.2 M TRIS-HCl pH 7.5, 0.25 M NaCl, 25 mM EDTA, and 0.5% (w/v) SDS for 20 min at 60 °C with frequent mixing. After centrifugation at 14 000 g for 5 min, phenol/chloroform/isoamyl alcohol (IAA) (25:24:1 by vol.) was added to the supernatant and the aqueous phase separated by centrifugation at 14 000 g for 10 min. A second extraction with chloroform/IAA was performed. An equal volume of isopropanol was mixed with the upper aqueous phase and incubated at 4 °C for 20 min. The genomic DNA was pelleted by centrifugation at 14 000 g for 10 min and washed once with 70% (v/v) ethanol, spun at 14 000 g for 10 min, air-dried for 10–15 min, and resuspended in 50 μl of sterile water. The samples were afterwards treated with RNase A (10 mg ml−1) (2.5 μl per 100 ml) for 1.5 h at 37 °C, extracted with phenol/chloroform/IAA and then with chloroform/IAA and precipitated with 1/10 vols of 3 M Na-acetate and 2.6 vols of absolute ethanol at −20 °C for 1 h. The genomic DNA was collected by centrifugation at 14 000 g for 10 min and washed with 70% (v/v) ethanol, spun at 14 000 g for 10 min, air-dried for 10–15 min, and resuspended in 50 μl of sterile water.

Southern blotting and hybridization

For genomic Southern analysis, 25 μg of genomic DNA was digested with 60 U of _Bam_HI, _Eco_RI, _Xho_I, or _Pst_I, for 15 h at 37 °C. The resulting fragments were separated by electrophoresis on a 1.5% (w/v) agarose TRIS-acetate-EDTA gel, which was then depurinated (0.25 M HCl), denatured (1.5 M NaCl, 0.5 M NaOH) and neutralized (1.5 M NaCl, 0.5 M TRIS-HCl pH 7.5, and 0.001 M EDTA) prior to capillary blotting onto positively charged nylon membrane (Hybond N+, Amersham Biosciences, Buckinghamshire, UK) overnight using 20× SSC. The genomic DNA was UV-crosslinked (UV-Stratelinker 2400, Amsterdam, The Netherlands).

As hybridization probes, a TcIRT1 Ganges full-length DNA (TcIRT1-1G) and a partial 645 bp TcIRT2 Ganges cDNA probe covering nucleotides 185–840 of the corresponding Arabidopsis AtIRT2 were used. Cross-hybridization could not be excluded because of the 72–73% nucleotide sequences identity between the IRT1 and IRT2 probe. For probe preparation, 50 ng of DNA (probe) was denatured at 95 °C for 5 min and labelled with 5 μl of [α-32P]dATP (10 μCi μl−1) as detailed by the manufacturer (RadPRime DNA Labeling System, Invitrogen, Paisley, UK) and purified by using the Nuctrap probe purification column (Stratagene, Amsterdam, The Netherlands) and added directly to the hybridization buffer.

The blots were incubated for 20 min in 10 ml of a hybridization solution containing 0.5 M NaPO4 pH 7.5, 7% (w/v) SDS, and 1% (w/v) BSA at 60 °C and then pre-hybridized at 65 °C for 2 h in hybridization solution containing 100 μg ml−1 sheared salmon-sperm DNA. The blots were then probed with the denatured 32P-labelled probe (see below) added directly to the hybridization solution and the blots incubated at 65 °C for 12–16 h. After hybridization the blots were rinsed twice at 65 °C with 2× SSC and 0.1% (w/v) SDS for 5 min and 20 min, respectively, and then washed with 1× SSC and 0.1% (w/v) SDS and finally in 1× SSC and 0.1% (w/v) SDS for 30 min. After washing, the membranes were placed on X-ray film and exposed at −80 °C for 16 h.

In order to use the same membrane for the two probes, the membrane was washed twice with boiling 0.1% SDS between hybridizations. Superposition of images of the blots was performed by using Adobe Photoshop 7.0.

Genomic sequences

A 1 μg aliquot of genomic DNA of Ganges and Prayon ecotypes was used for PCR amplification of the respective genomic TcIRT1 sequences. The primers were the same as used in Lombi et al. (2002) and the PCR conditions were initial denaturing for 3 min 95 °C, followed by 40 cycles of 30 s at 95 °C, 20 s at 55 °C, and 90 s at 72 °C, with a final elongation step for 5 min at 72 °C. The splicing sites were obtained by the alignment of the cDNA and the present genes using CLUSTAL W and confirmed with NetGene2 (http://www.cds.dtu.dk/services/NetGene2).

Functional analysis

Constructs and transformation

RT-PCR was used to produce full-length TcIRT1-1G and TcIRT1-2P cDNA sequences. The cDNAs were sequenced (four colonies of each in full agreement with the corresponding genomic DNAs) before the cloning to ensure fidelity of clones used in the constructs for yeast transformation. TcIRT1-1G was amplified using the same primers as indicated above, whereas for the _TcIRT1-2_P sequence, a specific primer for the end was used (IRT1-PR-noncod, 5′-GGACACGGGATTGATATGATATG-3′). The resulting PCR products were subcloned into the pGEMT-easy vector (Promega UK, Southampton, UK). Sequencing of the plasmid insert was performed on an ABI Prism 310 genetic analyser using Big Dye™ sequencing reagents (PE-Applied Biosystems, Foster City, CA) to ensure that no PCR errors had been introduced. The _Not_I-_TcIRT1-1G or Not_I-TcIRT1-2P fragments were ligated into the yeast expression vector pFL61 (Minet et al., 1992). The orientation of the insert was established by restriction digestion. pFL61 containing AtIRT1 (pFL61/AtIRT1) was obtained from Professor M-L Guerinot (Dartmouth College, Hanover, USA) and Professor D Eide (University of Wisconsin-Madison, USA).

The yeast strains fet3fet4 DEY1453 (MATa/MATa ade2/+can1/can1 his3/his3 leu2 leu2 trp1/trp1 ura3/ura3 fet3-2::HIS3/fet3-2::HIS3 fet4-1::LEU2/fet4-1::LEU2) and parent strain (wild type) DEY1457 (MATa/MATα ade6 can1 his3 leu2 trp1 ura3) were grown in standard defined media (1.7 g l−1 yeast nitrogen base, 5 g l−1 ammonium sulphate, 5 g l−1D-glucose, 0.83 g l−1 amino acid supplements), supplemented with 0.2 mM FeCl3 in the case of the fet3fet4 mutant. The fet3fet4 strain was obtained from Professor S Thomine (Gif-sur-Yvette, France). These strains were transformed with pFL61, pFL61/TcIRT1-G, pFL61/TcIRT1-P, or pFL61/AtIRT1 constructs by the lithium acetate method (Gietz et al., 1992).

Yeast functional complementation

Minimal medium (SD) lacking uracil (-ura) was supplemented to generate the desired conditions. To assess Cd sensitivity, 2.5, 5, 7.5, or 10 μM CdSO4 was used. For the iron complementation of the fet3fet4 mutant, plates without a supplement of 0.2 mM FeCl3 were tested. Three dilutions of the culture (of optical density at 600 nm of 0.2, 0.02, and 0.002) were spotted onto the plates from the left to the right. Growth was carried out for 4 d at 30 °C.

Cd-uptake in wild type yeast

Overnight cultures of single wild-type (WT) yeast colonies transformed with pFL61/TcIRT1-1G, pFL61/TcIRT1-2P, pFL61/AtIRT1, or the empty pFL61 vector were added to 200 ml volumes of SD-ura to obtain an OD600nm of 0.2 and incubated until an OD600nm of approximately 0.4 was reached (time 0). CdSO4(50 μM) was then added to each culture. Twenty ml volumes were removed from each culture, washed with 20 ml of 1 μM EDTA followed by sterilized deionized water (SDW) and the cells harvested by vacuum filtration 4 h after the initial addition of CdCl2. A sample was analysed in order to obtain the OD600nm. The yeast samples were digested with HNO3 at 160 °C, and Cd concentrations determined by inductively-coupled plasma mass spectrometry (ICP-MS, Agilent 7500ce, Agilent Technologies, Palo Alto, CA, USA).

Results

Gene expression in response to Fe treatments

Expression of four members of the ZIP gene family (TcIRT1, TcIRT2, TcZNT1, TcZNT5) was investigated in root and leaf tissues of the two ecotypes of T. caerulescens under Fe-sufficient or -deficient conditions (Fig. 1).

Fig. 1.

Effect of Fe status on ZIP gene expression in the T. caerulescens Prayon and Ganges ecotypes. mRNA was detected by northern blot analysis of total RNA extracted from root and leaf tissues. Each lane contained 10 μg of total RNA. Expression of (A) TcZNT1-G (AJ313521), (B) TcZNT5-G (AJ937739), (C) TcIRT1-G (AJ320253), and (D) TcIRT2-G (AJ538344) in the roots and leaves of Ganges and Prayon. (E) 18S rRNA was used as a loading control. The approximate sizes of the transcripts are TcIRT1-1G: 1.4 kbp, TcIRT1-2P: 0.5 kbp, TcIRT2: 0.6 kbp, TcZNT1: 1.3 kbp and TcZNT5: 1.3 kbp.

TcZNT1 was expressed in the roots strongly and to a much lesser extent in the leaves of both T. caerulescens ecotypes (Fig. 1A). Little difference in the abundance of TcZNT1 mRNA was observed between the two ecotypes or between the +Fe and –Fe treatments. In contrast, the transcript abundance of TcZNT5 mRNA was higher in the Prayon roots than in the Ganges roots (Fig. 1B). There was little change in the mRNA abundance between the Fe treatments for Prayon; however, for Ganges there was an apparent increase in TcZNT5 expression under Fe deficiency. No expression of TcZNT5 was observed in the leaves of either ecotype.

As previously reported (Lombi et al., 2002), TcIRT1 (TcIRT1-1G) was predominantly expressed in the roots of the Ganges ecotype. In addition, as shown here, a smaller mRNA (c. 0.5 kbp) from Prayon roots cross-hybridizes with the Ganges TcIRT1 probe (TcIRT1-2P) (Fig. 1C). The abundances of mRNAs hybridizing with TcIRT1 in both Ganges and Prayon were greatly increased in root tissue under Fe-deficient conditions when compared with Fe-sufficient conditions.

In contrast to TcIRT1, TcIRT2 was expressed in the roots of both ecotypes (Fig. 1D). The expression was enhanced greatly by Fe deficiency. No difference in the TcIRT2 transcript level was observed between the two ecotypes under either Fe-sufficient or Fe-deficient conditions. There was no expression of either TcIRT1 or TcIRT2 in the leaves.

Analysis of TcIRT1 genes

As indicated by the northerns (Fig. 1), the TcIRT1 transcripts for each ecotype differed in their size. The cDNAs for both of these transcripts were isolated and sequenced. Whereas Ganges expressed the full-size gene (TcIRT1-1G), the transcript of Prayon (TcIRT1-2P) was truncated, resulting in a polypeptide lacking the putative metal-binding site of the ZIP family and the five C-terminal transmembrane helices (Eng et al., 1998) (Fig. 2). The 3′-end non-coding region of Prayon [around 150 bp until the poly (A) tail] showed no similarity to any known genes. To investigate this further, the genomic structure was analysed (Southerns, Fig. 3). Probing genomic DNA with the full-length size TcIRT1-1G, multiple _TcIRT1_-hybridizing bands were apparent in each ecotype (Fig. 3A). To examine cross-hybridization with the closely homologous TcIRT2, the same membrane was re-hybridized with a _TcIRT2_-specific probe, and a single band corresponding to the single copy of this gene in T. caerulescens was observed (Fig. 3B). Superimposing the two images indicates possible cross-hybridization of TcIRT2 and TcIRT1 (marked by an asterisk in Fig. 3C). At least two bands remain which are specific for TcIRT1, suggesting at least two copies of the IRT1 gene in each ecotype, given that none of the sequences contains _Bam_HI or _Eco_RI sites. The multiple band pattern of the Ganges _Eco_RI hybridization may be a sign of other unidentified isoforms which are not apparent in the _Bam_HI hybridization or may be because the genomic DNA was isolated from a population of non-isogenic seedlings. Using specific primers for the full and truncated length sequences (TcIRT1-1G and TcIRT1-2P, respectively), two isoforms of TcIRT1 were amplified in the genomic DNA for each of the two ecotypes, corresponding to the genes for the full (TcIRT1-1G: AM293350 and TcIRT1-1P: AM293351, for Ganges and Prayon, respectively) and truncated forms (TcIRT1-2G: AM293352 and TcIRT1-2P: AM293353, for Ganges and Prayon, respectively) (Fig. 3D). A Vista alignment (http://genome.lbl.gov/vista/index.shtml) indicating homology of the genomic sequence of A. thaliana IRT1 with the different isoforms of T. caerulescens is shown in Fig. 3E. Similarity is high except in the regions of the introns.

Fig. 2.

Alignment of Arabidopsis thaliana IRT1 (accession number NM_118089) and TcIRT1 putative amino acid sequences predicted from the genomic sequences, AM293353 and AM293350 from Prayon and Ganges ecotypes, respectively. Amino acids are shaded grey or black when conserved in two or three sequences, respectively. Transmembrane (TM) domains (Eide et al., 1996) are underlined with a broken line. The putative heavy metal binding site is underlined with a continuous line.

Fig. 3.

Genomic organization of TcIRT1. Genomic DNA isolated from the leaf tissues was digested with _Bam_HI or _Eco_RI (which do not cut in any of the sequences), and hybridized with a 32P-labelled TcIRT1 Ganges full-length probe (TcIRT1-1G) (A) or a Ganges TcIRT2-specific partial (645 bp, TcIRT2 Primers 1 and 2) cDNA probe (B). Superposition of the two probes (IRT2 probe in blue) is shown in (C) and possible cross-hybridizing bands are shown by an asterisk. The IRT1-G probe was 100, 72, 85.5, and 71.8, and the IRT2 probe was 72, 100, 73.2, and 99.2% similar to TcIRT1-1G, TcIRT2-G, TcIRT1-2P, and TcIRT2-P, respectively. The genomic structure of the resulting sequences for Ganges ecotype [full length (TcIRT1-1G): AM293350 and truncated (TcIRT1-2G): AM293352] is shown in (D) from the start codon to the stop, as well as the 3′ non-coding region until the poly(A) site (*). The Prayon structure (data not shown) is very similar [same introns, full-length (TcIRT1-1P): AM293351, truncated (TcIRT1-2P): AM293353]. A Vista (http://genome.lbl.gov/vista/index.shtml) alignment of the genomic sequence of AtIRT1 with the four genomic sequences of TcIRT1 is shown in (E). Percentage similarities were calculated using a 100 bp moving average.

Functional analysis of TcIRT1-G in S. cerevisiae

The A. thaliana IRT1 has been shown to complement the yeast double mutant fet3fet4 (strain DEY1453), lacking the high- and low-affinity iron-uptake systems (Eide et al., 1996). Therefore, in order to investigate the role of TcIRT1 isoforms in Fe assimilation in yeast, fet3fet4 was complemented with the full-length TcIRT1-1G from Ganges and the truncated transcript from Prayon (TcIRT1-2P). Since the fet3fet4 mutant cannot grow under Fe-limited conditions, the ability of the transformants to grow on minimal medium without or with a supplement of Fe (200 μM) was tested. Transformants with the pFL61 control vector were only able to grow in the media supplemented with 200 μM Fe, whereas complementation with IRT1 from A. thaliana and T. caerulescens (including the truncated TcIRT1-2P) allowed growth in the normal minimum medium without addition of 200 μM Fe (Fig. 4A).

Fig. 4.

Yeast functional assays of metal uptake-defective yeast mutants transformed with pFL61, pFL61/AtIRT1, pFL61/TcIRT1-1G (full-length), or pFL61-TcIRT1-2P (TcIRT1 truncated). A growth drop test analysis showing relative growth of yeast at three dilutions (see Materials and methods). (A) Complementation of the fet3fet4 (DEY1453 strain) yeast mutant deficient in low- and high-Fe uptake on synthetic medium (normal) or supplemented with 200 μM FeCl3. (B) Effect of different Cd concentrations on the growth of the wild type (DEY1457). In (A) and (B), 0.2, 0.02, and 0.002 OD600 unit drops of overnight cultures were spotted on synthetic medium (from left to right), and plates were incubated at 30 °C for 4 d. These experiments were repeated three times, each time with two independent colonies, and the data shown are from one representative experiment. (C) Cadmium uptake by the wild-type yeast expressing pFL61, pFL61-AtIRT1, pFL61-TcIRT1-G (full-length), or pFL61-TcIRT1-P (truncated). The growth medium contained 50 μM CdSO4, and Cd uptake by yeast was measured by ICP-MS after 4 h. The data are normalized for a yeast suspension of 1 OD. Four independent transformants were used in this experiment. The error bars are the SD of the mean. Significant difference (P ≤0.05) is shown by an asterisk.

Expression of IRT1 from A. thaliana in wild-type yeast increases the sensitivity to the presence of Cd

To investigate this potential role of IRT1, wild-type yeast was complemented with IRT1 cDNAs of T. caerulescens and A. thaliana. In the presence of 10 μM CdSO4 there was an increase of sensitivity to Cd in yeast transformed with AtIRT1 or TcIRT1-1G but not with TcIRT1-2P. At the lower Cd concentrations used, the presence of the A. thaliana IRT1 increased the sensitivity of the yeast the most (Fig. 4B). These yeast functional assays indicated that TcIRT1 is involved in the transport of Cd as well as Fe, but that in the case of Cd, the full-length transcript was required, suggesting a requirement for the full-length sequence, which includes the histidine-rich putative metal-binding site (see Fig. 2), in order to be able to transport Cd.

The uptake of Cd into yeast was investigated by measuring the amount of Cd taken up by yeast cultures using ICP-MS. Yeast complemented with IRT1 from Arabidopsis showed the greatest uptake after 4 h, whilst little or no uptake above that of the wild type was observed with TcIRT1-1G or TcIRT1-2P (Fig. 4C). These results indicated that TcIRT1-1G had no detectable Cd transport activity under the conditions used and at this time scale, a phenomenon observed by Hassinen et al. (2007) with TcMRP10. Initial Cd uptake rates were assayed using radiolabelled Cd; however, uptake was too low to determine the kinetics reproducibly.

Gene expression in response to Cd treatments

Expression of the ZIP genes was analysed in roots of T. caerulescens ecotypes in the absence or presence of Cd, when plants were supplied with sufficient Fe. Representative blots are shown in Fig. 5. TcZNT1 expression was not affected by Cd and expression levels were similar for both ecotypes. Similar to the Fe treatment experiment, the expression level of TcZNT5 in Ganges was below that seen for Prayon. In both ecotypes, there was no clear effect of Cd treatments on the expression of TcZNT5-G in the roots. Expression of TcIRT1-1G was observed in Ganges roots in the presence of Cd, even though, in this experiment, plants were grown with a sufficient level of Fe. Expression of the truncated TcIRT1-2P was observed in Prayon roots in response to Cd. TcIRT2 transcript levels were not detectable in either ecotype, in the presence or absence of Cd.

Fig. 5.

Effect of Cd treatment (100 μM) on gene expression in T. caerulescens Prayon and Ganges ecotypes. mRNA was detected by northern blot analysis of total RNA extracted from roots. Each lane contained 10 μg of total RNA. Expression of TcZNT1, TcZNT5, TcIRT1, and TcIRT2 in the roots of Ganges and Prayon is as indicated. 18S rRNA was used as a loading control.

Phylogenetic relationship of ZIP genes

It is probable that T. caerulescens homologues exist for most, if not all of the A. thaliana ZIP genes, and T. caerulescens homologues have been cloned from three out of four of the main clusters (shaded, Fig. 6).

Fig. 6.

Phylogenetic tree of ZIP gene transporters based on their amino acid sequences. The plant ZIP genes fall into four main groups represented by each of the shaded areas. The ZIP genes discussed in this report are highlighted in bold. Protein alignments were performed using Clustal X (Higgins and Sharp, 1988). The tree was constructed using njplot (Clustal X), and bootstrap values are shown. The (p) designates that the sequence used was a partial one. A number in parenthesis following a gene abbreviation indicates that two variants were found and each was inserted in the phylogenetic tree. Accession numbers are shown for each sequence as follows: AtZIP11, At1g55910; AtZIP2, At5g59520; AtZIP9, At4g33020; AtIRT3, At1g60960; TcZNT2-LC, AF275752; TcZNT4, AF292370; TcZNT2-G(p), AJ538346; AtZIP4, At1g10970; TcZNT1-LC, AF275751; TcZNT1-P, AF133267; TcZNT1-G, AJ313521; AtZIP12, At5g62160; AtZIP3, At2g32270; TcZNT5-G, AJ937739; TcZNT5-P, AJ937738; TcZNT5, AF292029; AtZIP5(2), NM_202033; AtZIP5(1), NM_100409; GmZIP1, AY029321; AtZIP1, At3g12750; OsZIP2, AAP33800; OsIRT1, AB070226; AtZIP7, AF369911; AtZIP10, At1g31260; PsRIT1, AF065444; LeIRT2, AF136580; LeIRT1, AF136579; ZIP8, AF475143; TcIRT2-P(p), AJ538345; TcIRT2-G(p), AJ538344; AtIRT2(2), NM_001036593; AtIRT2(1), NM_118088; TcIRT1-2P, AM113516 or AM293353; TcIRT1-1G, AJ320253 or AM293350; AtIRT1(2), NM_118089; AtIRT1(1), NM_179215; AtZIP6, At2g30080.

The two TcIRT1 transporters align together along with the Arabidopsis homologue. Of the other T. caerulescens ZIP genes, TcIRT2 was cloned for the first time in the present study: for both Prayon and Ganges, the TcIRT2 cDNA fragment was 645 bp long, spanning nucleotides 189–840 of AtIRT2. The TcIRT2-G and TcIRT2-P fragments showed 99% identity at the nucleotide level and 98% identity at the protein level. They are close homologues of AtIRT2, with 91% identity at the amino acid level. When compared with AtIRT2, TcIRT2 from both T. caerulescens ecotypes has an extra histidine and glycine residue pair at amino acid 114 of accession AJ538344/AJ53844, an area thought to be involved in heavy metal binding (Eide et al., 1996). There is also a four amino acid insert in the Prayon and Ganges TcIRT2 sequences at amino acid 127, when compared with the equivalent position in the A. thaliana homologues.

The phylogenetic analysis indicates that TcZNT1 is a close homologue of AtZIP4 from A. thaliana (previously reported by Assunção et al., 2001) and TcZNT5 is a close homologue of ZIP5. For TcZNT5-P, the resulting cDNA fragment was 1023 bp long, spanning nucleotides 52–1073 of the ZNT5 sequence reported in the database (GenBank AF292029). The Ganges ZNT5 cDNA fragment (TcZNT5-G) was 1014 bp long, spanning the same region of ZNT5 as the Prayon fragment. However, a three amino acid segment was absent from TcZNT5-G at amino acid 20. TcZNT5-P and TcZNT5-G have 98% nucleotide identity. There are some discrepancies between the Prayon cDNA cloned here and that already reported in the database although the nucleotide identity is still over 99%. These small differences may be due to variation within the Prayon ecotype.

The sequence data indicate that ecotypic differences in amino acid sequence exist between the Prayon and Ganges gene homologues. The significance of these differences in terms of metal ion uptake requires investigation through functional analysis.

Discussion

Expression analysis

Physiological differences between the Ganges and Prayon ecotypes of T. caerulescens, in terms of Cd and Zn hyperaccumulation, have been well characterized (Lombi et al., 2001, 2002; Zhao et al., 2002; Roosens et al., 2003; Zha et al., 2004). These studies have provided strong physiological evidence that multiple uptake systems are involved in the root uptake of Zn and Cd in T. caerulescens. A number of ZIP genes have been found to be constitutively highly expressed in T. caerulescens (Pence et al., 2000; Assunção et al., 2001; Hammond et al., 2006) and in another Zn/Cd hyperaccumulator Arabidopsis halleri (Becher et al., 2004; Weber et al., 2004; Talke et al., 2006), when compared with non-hyperaccumulator species. These genes may play a role in the hyperaccumulation of Zn and Cd. In the present study, ecotypic variations in the amino acid sequences and expression patterns of four ZIP genes representing three separate clusters of the ZIP family were investigated (Fig. 6).

The cloning of the respective ZIP gene homologues from both Prayon and Ganges led to the identification of amino acid sequence differences that may contribute to the differing abilities of the two ecotypes to accumulate Cd. The most significant difference was the truncated TcIRT1 transcript in Prayon, TcIRT1-2P. The resulting predicted polypeptide would be less than one-half in length, and lack the putative metal-binding domain (HGHGHGHG) and five of the eight transmembrane domains (Fig. 2). It would be surprising if such a truncated protein could catalyse any transport function.

Other point differences between amino acid sequences were noted. Additional histidine and glycine residues (G116 and H117) were present in TcIRT2 (AJ538344 and AJ538345) but not AtIRT2 (NM_001036593 or NM_118088), and may be of particular interest due to the location in the potential metal-binding site (data not shown). Other amino acid substitutions identified may also be of importance as such substitutions have previously been shown to affect transport by ZIP proteins. For example, a single amino acid substitution in the yeast ZIP gene ZRT1 blocked endocytosis of the ZRT1 protein from the plasma membrane (Gitan and Eide, 2000). In the most comprehensive analysis to date (Rogers et al., 2000), 14 amino acid substitutions in the AtIRT1 sequence were investigated for effects on transport properties including ion selectivity. Six of these substitutions which were in the region deleted in TcIRT1-2P were essential for transport activity, suggesting that the truncated TcIRT1-2P would not have transport function. Altered transport properties were observed for single amino acid substitutions in the N-terminal region of AtIRT1; however, all of these are absolutely conserved in TcIRT1-1G (except an E103D conservative substitution) and TcIRT1-2P.

All four ZIP genes (TcZNT1, TcZNT5, TcIRT1, and TcIRT2) were expressed mainly in the root tissues, with little or no expression in leaves (Fig. 1). This is consistent with previous studies with A. thaliana (Grotz et al., 1998; Vert et al., 2001,, 2002) and with T. caerulescens for TcZNT1 and TcZNT2 (Pence et al., 2000; Assunção et al., 2001), suggesting that the transporters are probably involved in root uptake or xylem loading of metals.

TcZNT1 expression was equally high in the roots of both ecotypes, despite the plants being grown with 5 μM Zn. Furthermore, expression of this gene was not responsive to different Fe or Cd treatments (Figs 1, 5). Pence et al. (2000) showed that TcZNT1 can mediate high-affinity Zn uptake, as well as low-affinity Cd uptake, when the gene was expressed in yeast. The constitutively high expression of TcZNT1 may be an important reason for the hyperaccumulation of Zn by this plant (Pence et al., 2000; Assunção et al., 2001). However, the superior Cd-accumulating ability exhibited by the Ganges ecotype is unlikely to be due to TcZNT1, because of its low affinity for Cd and the similar levels of expression in the two contrasting ecotypes. Although the TcZNT5 sequence has been submitted in the database (AF292029), there is little knowledge regarding the expression of this gene and its function. The expression of TcZNT5 was much higher in Prayon than in Ganges (Figs 1, 5). This ecotypic difference is notable, but is opposite to the pattern of Cd accumulation. This would rule out TcZNT5 as a major Cd transporter involved in the influx of Cd in Ganges roots. However, if TcZNT5 were to mediate Cd efflux from cytoplasm to the apoplast, the observed expression pattern could explain the ecotypic difference in Cd accumulation. TcZNT5 is most closely related to AtZIP5, and to a slightly lesser extent also to AtZIP3 (Fig. 6). No expression or functional data currently exist for AtZIP5, but AtZIP3 is expressed in Zn-deficient A. thaliana roots and can rescue a Zn-uptake defective mutant of yeast (Grotz et al., 1998).

A previous study showed that the expression of TcIRT1-1G was highly inducible under Fe deficiency in the Ganges ecotype, but little or no equivalent transcript was detectable in the Prayon ecotype (Lombi et al., 2002). In addition, a TcIRT1-1G homologue transcript could not be isolated from the Prayon ecotype by RT-PCR. This ecotypic difference was further confirmed in the present study (Fig. 1). The greatly increased transcript abundance of TcIRT1-1G in the Ganges ecotype in response to Fe deficiency corresponded to the greatly increased _V_max for Cd influx and Cd accumulation in the shoots of Ganges under Fe deficiency (Lombi et al., 2002). The expression was enhanced by Cd exposure in an experiment with sufficient Fe supply (Fig. 5). This could be due either to a direct effect of Cd or an indirect effect through Cd-induced Fe deficiency. However, the lack of expression of TcIRT2-G in the roots of both ecotypes in this experiment (see below) suggests that there was no Fe deficiency. The low or zero expression of a full-length TcIRT1 in the Prayon ecotype, even under Fe deficiency, implies that this ecotype must rely upon the truncated TcIRT1 (which appears to be able to transport Fe, see Fig. 4A) or other Fe transporters, such as other ZIP transporters or Nramp transporters, to acquire Fe. For example, it is possible that Prayon relies on TcIRT2 for the acquisition of Fe. Expression of TcIRT2-G was highly inducible under Fe deficiency in both ecotypes, and there was little difference between the ecotypes in TcIRT2-G transcript level (Fig. 1). Unlike TcIRT1-G, expression of TcIRT2-G was not detectable in Fe-replete plants of either ecotype in the Cd exposure experiment (Fig. 5). The expression data suggest that TcIRT2-G is unlikely to play a significant role in the root uptake of Cd. Nor does it explain the ecotypic difference in Cd accumulation.

Gene structure

Quite surprisingly, a truncated transcript lacking the putative metal-binding site and the five C-terminal transmembrane helices has been identified in the low Cd-accumulating ecotype, which is unable to transport Cd as shown by the yeast functional assays (Fig. 4B). Truncated but not full-length transcripts were found in another low Cd-accumulating population (Bizkaia, Spain) of T. caerulescens (data not shown), suggesting that this mechanism is not restricted to Prayon but may be responsible for the high variability of this species for Cd accumulation. The two ecotypes are similar at the genomic level, with each possessing two copies of TcIRT1, one corresponding to the full-length and one to the truncated genes. Differential expression suggests that the promoter regions or transcript stability must be different in the two ecotypes. Examination of promoters of each transcript for each ecotype is necessary to understand fully the mechanisms behind this differential transcription.

In A. halleri, several candidate genes encoding metal transporters have been found to have multiple gene copies, including MTP1, HMA4, ZIP3, ZIP6, and ZIP9, whose expression levels were also much elevated compared with A. thaliana (Dräger et al., 2004; Talke et al., 2006).

Functional analysis

AtIRT1 has been shown to mediate uptake of Fe(II) and Cd across plasma membranes of root cells in A. thaliana (Connolly et al., 2002; Vert et al., 2002). As the expression patterns of TcIRT1-1G were suggestive of a role in Cd uptake, this gene was subjected to functional characterization by heterologous expression in yeast. In agreement with the results reported for AtIRT1 (Eide et al., 1996; Korshunova et al., 1999; Rogers et al., 2000; Connolly et al., 2002), TcIRT1-1G as well as the truncated version were able to complement the Fe-uptake-deficient yeast mutant fet3fet4. The surprising complementation of the fet3fet4 mutant by the truncated version may indicate that the five C-terminal transmembrane helices and the putative metal-binding site are not critical for Fe transport. As kinetics of Fe transport were not examined, it is impossible to say to what extent Fe transport is modified; it may be speculated that the truncated version may allow some permeability of Fe, perhaps in a multimeric conformation. The presence of TcIRT1-1G, at the level of expression achieved, did not apparently facilitate significant Cd uptake at the concentration and at the time scale (4 h) used (Fig 4C); however, an increase in the sensitivity of wild-type yeast strains to Cd was observed (after 3 d), whilst the truncated version did not modify the sensitivity of the wild-type strain to Cd. At low concentrations, the sensitivity of the yeast Wt-pFL61/AtIRT1 was higher than for Wt-pFL61/TcIRT1-1G, suggesting a superior transport function of the Arabidopsis transporter compared with Thlaspi at this low concentration (2.5–7.5 μM Cd in the agar medium), or possibly a higher level of expression (abundance). The yeast functional assays suggest that the putative metal-binding site is involved in Cd transport but not that of Fe in yeast. The medium used (SD) in all the experiments contains 1.23 μM Fe and 2.48 μM Zn, which is unlikely to have any direct competitive effects on Cd uptake (50 μM; Fig. 4C), although post-transcriptional regulation by the presence of these ions has been suggested (Connolly et al., 2002).

Summary

The results from the present study are inconclusive, but not inconsistent with the suggestion that TcIRT1-1G is involved in Cd uptake by the Ganges ecotype, as shown by an increase of Cd sensitivity of the yeast wild type (normally naturally resistant), due to the presence of a full-length cDNA, particularly under Fe deficiency. The fact that TcIRT1-1G was highly expressed in Ganges under Fe-sufficient conditions in the presence of Cd (Fig. 5) may explain the superior Cd-accumulating ability of this plant when compared with Prayon. Conversely, the suppressed expression of a full-length IRT1 transcript in Prayon under any of the observed conditions may be an adaptive strategy to limit metal uptake in this accession. Alternatively, it is possible that an as yet uncloned homologue (as indicated by the Southern data) may be responsible for the observed uptake. In all cases it is likely that patterns or levels of expression or post-transcriptional control rather than a specific characteristic of this transporter contribute to the observed Cd accumulation abilities of the different accessions. Other gene family members still require functional analysis to determine their role in metal ion uptake and hyperaccumulation. In addition, this study demonstrates the need for ecotype-specific studies.

Jonathan Howarth is thanked for valuable discussions and suggestions, and Jianping Xing for assistance with yeast transport measurements. This project was supported partly by Rio Tinto Technology Development Limited. SP was supported by the Swiss National Science Foundation. Rothamsted Research receives grant-aided support from the Biotechnology and Biological Sciences Research Council of the United Kingdom.

References

Differential metal-specific tolerance and accumulation patterns among Thlaspi caerulescens populations originating from different soil types

,

New Phytologist

,

2003

, vol.

159

(pg.

411

-

419

)

A cosegregation analysis of zinc (Zn) accumulation and Zn tolerance in the Zn hyperaccumulator Thlaspi caerulescens

,

New Phytologist

,

2003

, vol.

159

(pg.

383

-

390

)

Elevated expression of metal transporter genes in three accessions of the metal hyperaccumulator Thlaspi caerulescens

,

Plant, Cell and Environment

,

2001

, vol.

24

(pg.

217

-

226

)

Metal hyperaccumulator plants: a review of the ecology and physiology of a biochemical resource for phytoremediation of metal-polluted soils

,

Phytoremediation of contaminated soil and water

,

2000

Boca Raton, FL, USA

Lewis Publishers

(pg.

85

-

107

)

Heavy metal accumulation and tolerance in British populations of the metallophyte Thlaspi caerulescens J&C Presl (Brassicaceae)

,

New Phytologist

,

1994

, vol.

127

(pg.

61

-

68

)

Cross-species microarray transcript profiling reveals high constitutive expression of metal homeostasis genes in shoots of the zinc hyperaccumulator Arabidopsis halleri

,

The Plant Journal

,

2004

, vol.

37

(pg.

251

-

268

)

Phytoremediation of soil metals

,

Current Opinion in Biotechnology

,

1997

, vol.

8

(pg.

279

-

284

)

Expression of the IRT1 metal transporter is controlled by metals at the levels of transcript and protein accumulation

,

The Plant Cell

,

2002

, vol.

14

(pg.

1347

-

1357

)

Two genes encoding Arabidopsis halleri MTP1 metal transport proteins co-segregate with zinc tolerance and account for high MTP1 transcript levels

,

The Plant Journal

,

2004

, vol.

39

(pg.

425

-

439

)

A novel iron-regulated metal transporter from plants identified by functional expression in yeast

,

Proceedings of the National Academy of Sciences, USA

,

1996

, vol.

93

(pg.

5624

-

5628

)

Sequence analyses and phylogenetic characterization of the ZIP family of metal ion transport proteins

,

Journal of Membrane Biology

,

1998

, vol.

166

(pg.

1

-

7

)

Zinc and cadmium hyperaccumulation by Thlaspi caerulescens from metalliferous and nonmetalliferous sites in the Mediterranean area: implications for phytoremediation

,

New Phytologist

,

2000

, vol.

145

(pg.

429

-

437

)

Improved method for high efficiency transformation of intact yeast cells

,

Nucleic Acids Research

,

1992

, vol.

20

(pg.

1425

-

1425

)

Zinc-regulated ubiquitin conjugation signals endocytosis of the yeast ZRT1 zinc transporter

,

Biochemical Journal

,

2000

, vol.

346

(pg.

329

-

336

)

Identification of a family of zinc transporter genes from Arabidopsis that respond to zinc deficiency

,

Proceedings of the National Academy of Sciences, USA

,

1998

, vol.

95

(pg.

7220

-

7224

)

The ZIP family of metal transporters

,

Biochimica et Biophysica Acta

,

2000

, vol.

1465

(pg.

190

-

198

)

A comparison of the Thlaspi caerulescens and Thlaspi arvense shoot transcriptomes

,

New Phytologist

,

2006

, vol.

170

(pg.

239

-

260

)

Isolation of Zn-responsive genes from two accessions of the hyperaccumulator plant Thlaspi caerulescens

,

Planta

,

2007

, vol.

225

(pg.

977

-

998

)

Clustal: a package for performing multiple sequence alignment on a microcomputer

,

Gene

,

1988

, vol.

73

(pg.

237

-

244

)

The IRT1 protein from Arabidopsis thaliana is a metal transporter with a broad substrate range

,

Plant Molecular Biology

,

1999

, vol.

40

(pg.

37

-

44

)

Influence of iron status on cadmium and zinc uptake by different ecotypes of the hyperaccumulator Thlaspi caerulescens

,

Plant Physiology

,

2002

, vol.

128

(pg.

1359

-

1367

)

Cadmium accumulation in populations of Thlaspi caerulescens and Thlaspi goesingense

,

New Phytologist

,

2000

, vol.

145

(pg.

11

-

20

)

Physiological evidence for a high-affinity cadmium transporter highly expressed in a Thlaspi caerulescens ecotype

,

New Phytologist

,

2001

, vol.

149

(pg.

53

-

60

)

et al.

Phylogenetic relationships within cation transporter families of Arabidopsis

,

Plant Physiology

,

2001

, vol.

126

(pg.

1646

-

1667

)

Phytoextraction of metals and metalloids from contaminated soils

,

Current Opinion in Biotechnology

,

2003

, vol.

14

(pg.

277

-

282

)

Phytoremediation of metals, metalloids, and radionuclides

,

Advances in Agronomy

,

2002

, vol.

75

(pg.

1

-

56

)

Heavy metal tolerance and accumulation in metallicolous and non-metallicolous populations of Thlaspi caerulescens from continental Europe

,

Plant Ecology

,

1997

, vol.

133

(pg.

221

-

231

)

Complementation of Saccharomyces cerevisiae auxotrophic mutants by Arabidopsis thaliana cDNAs

,

The Plant Journal

,

1992

, vol.

2

(pg.

417

-

422

)

The molecular physiology of heavy metal transport in the Zn/Cd hyperaccumulator Thlaspi caerulescens

,

Proceedings of the National Academy of Sciences, USA

,

2000

, vol.

97

(pg.

4956

-

4960

)

Altered selectivity in an Arabidopsis metal transporter

,

Proceedings of the National Academy of Sciences, USA

,

2000

, vol.

97

(pg.

12356

-

12360

)

Natural variation in cadmium tolerance and its relationship to metal hyperaccumulation for seven populations of Thlaspi caerulescens from western Europe

,

Plant, Cell and Environment

,

2003

, vol.

26

(pg.

1657

-

1672

)

,

Molecular cloning: a laboratory manual

,

1989

Cold Spring Harbor

Cold Spring Harbor Laboratory Press

Uptake and transport of zinc in the hyperaccumulator Thlaspi caerulescens and the non-hyperaccumulator Thlaspi ochroleucum

,

Plant, Cell and Environment

,

1997

, vol.

20

(pg.

898

-

906

)

Zinc-dependent global transcriptional control, transcriptional deregulation, and higher gene copy number for genes in metal homeostasis of the hyperaccumulator Arabidopsis halleri

,

Plant Physiology

,

2006

, vol.

142

(pg.

148

-

167

)

Arabidopsis IRT2 gene encodes a root-periphery iron transporter

,

The Plant Journal

,

2001

, vol.

26

(pg.

181

-

189

)

IRT1, an Arabidopsis transporter essential for iron uptake from the soil and for plant growth

,

The Plant Cell

,

2002

, vol.

14

(pg.

1223

-

1233

)

A small-scale procedure for the rapid isolation of plant RNAs

,

Nucleic Acids Research

,

1989

, vol.

17

(pg.

2362

-

2362

)

Comparative microarray analysis of Arabidopsis thaliana and Arabidopsis halleri roots identifies nicotianamine synthase, a ZIP transporter and other genes as potential metal hyperaccumulation factors

,

The Plant Journal

,

2004

, vol.

37

(pg.

269

-

281

)

Co-segregation analysis of cadmium and zinc accumulation in Thlaspi caerulescens interecotypic crosses

,

New Phytologist

,

2004

, vol.

163

(pg.

299

-

312

)

Characteristics of cadmium uptake in two contrasting ecotypes of the hyperaccumulator Thlaspi caerulescens

,

Journal of Experimental Botany

,

2002

, vol.

53

(pg.

535

-

543

)

Author notes

*

Present address: Plant Biology, University of Fribourg, Fribourg, CH-1700, Switzerland.

†

Present address: Business and Innovation Unit, BBSRC, Swindon SN2 1UH, UK.

‡

These authors contributed equally to this work.

© 2007 The Author(s).

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.This paper is available online free of all access charges (see http://jxb.oxfordjournals.org/open\_access.html for further details)