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


Author for correspondence: Steve P. McGrathTel: +44 1582 763 133 Fax: +44 1582 760 981


• Uptake kinetics and translocation characteristics of cadmium and zinc are presented for two contrasting ecotypes of the Cd/Zn hyperaccumulator Thlaspi caerulescens, Ganges (southern France) and Prayon (Belgium).

• Experiments using radioactive isotopes were designed to investigate the physiology of Cd and Zn uptake, and a pressure-chamber system was employed to collect xylem sap.

• In contrast to similar Zn uptake and translocation, measurements of concentration-dependent influx of Cd revealed marked differences between ecotypes. Ganges alone showed a clear saturable component in the low Cd concentration range; maximum influx Vmax for Cd was fivefold higher in Ganges; and there was a fivefold difference in the Cd concentration in xylem sap. Addition of Zn to the uptake solution at equimolar concentration to Cd did not decrease Cd uptake by Ganges, but caused a 35% decrease in Prayon.

• There is strong physiological evidence for a high-affinity, highly expressed Cd transporter in the root cell plasma membranes of the Ganges ecotype of T. caerulescens. This raises evolutionary questions about specific transporters for non-essential metals. The results also show the considerable scope for selecting hyperaccumulator ecotypes to achieve higher phytoextraction efficiencies.


Phytoextraction of heavy metals has been thought to be an environmentally friendly, cost-effective solution to remediate soils polluted by heavy metals (Chaney, 1983; McGrath et al., 1993). Two strategies are being developed in the field of phytoextraction (Salt et al., 1998). The first one is based on the use of hyperaccumulator plants with exceptional metal-accumulating capacity (natural phytoextraction). The second strategy makes use of high biomass crops which are induced to take up large amounts of metals when the mobility of metals in soil is enhanced by chemical treatments (chemically enhanced phytoextraction).

Hyperaccumulator plants possess several unusual characteristics, such as the ability to take up and translocate exceedingly large amounts of metals to their shoots and hypertolerance to the metals. These characteristics are essential for a successful phytoextraction strategy. However, the main constraint of natural phytoextraction is the generally low biomass of metal hyperaccumulator plants, with the exception of some nickel (Ni) hyperaccumulators. To develop a successful remediation technology based on hyperaccumulator plants, the shoot biomass of these plants needs to be increased. Recently, Brewer et al. (1999) reported the results of somatic hybridization of the Zn hyperaccumulator Thlaspi caerulescens with the high biomass crop Brassica napus. The hybrids obtained were taller than T. caerulescens and also produced more biomass. They were able to accumulate Zinc (Zn) and Codmium (Cd) at levels that are toxic to B. napus, but below the levels commonly associated with hyperaccumulation. Another possibility is to transfer genes conferring the hyperaccumulation characteristics to high biomass plants using genetic engineering. However, the basic mechanisms responsible for hyperaccumulation are not yet fully understood and this lack of information is one of the factors limiting the application of molecular engineering techniques.

T. caerulescens represents a model metal hyperaccumulator species. This plant is able to accumulate extremely large concentrations of Zn in its shoots (Baker et al., 1994; Brown et al., 1995; Shen et al., 1997). Several physiological characteristics related to Zn hyperaccumulation in T. caerulescens have become clearer in recent years. Lasat et al. (1996) used radiotracer techniques to demonstrate that the ability of T. caerulescens to take up large amounts of Zn is related to the high density of Zn transporters per unit membrane area in the roots of this plant. Genes encoding a high affinity Zn transporter, named ZNT1, were recently cloned and characterized in T. caerulescens (Lasat et al., 2000). Salt et al. (1999) studied the ligands responsible for complexing Zn in root, xylem sap and leaves of T. caerulescens. In the root, Zn was mainly coordinated with histidine, in the xylem sap it was mainly transported as free hydrated cations and in the leaves it was primarily complexed with organic acids. Küpper et al. (1999) investigated the cellular compartmentation of Zn in leaves of T. caerulescens using X-ray microanalysis and single-cell sap extraction. They found that Zn was accumulated more in the epidermal cells than in mesophyll cells and concluded that compartmentation into the vacuoles is probably involved in detoxification of Zn.

T. caerulescens is also considered to be a Cd hyperaccumulator (Brooks, 1998; Baker et al., 2000), although much less is known about Cd hyperaccumulation. In a recent paper, we showed that four populations of T. caerulescens had similar ability to hyperaccumulate Zn but were markedly different in terms of Cd accumulation (Lombi et al., 2000). Two populations from southern France accumulated much more Cd than the populations from Prayon (Belgium) and Whitesike (UK), under both controlled environment and field conditions. In a hydroponic culture, the Southern French population was able to accumulate >10 000 mg kg−1 Cd in the shoots without showing any symptoms of phytotoxicity. These results indicate that, contrary to the general belief, the mechanisms of Cd and Zn hyperaccumulation are not identical in this species. The populations from southern France and from Prayon or Whitesike are clearly different ecotypes of T. caerulescens. These ecotypes offer highly interesting and valuable contrasts for studies of the physiology, molecular biology and genetics of Cd hyperaccumulation (Krämer, 2000).

The present work aims to characterize Cd and Zn uptake kinetics physiologically in two contrasting ecotypes of T. caerulescens and to investigate the effect of Zn on Cd uptake, using radiotracer techniques. Furthermore, translocation of Cd and Zn from root to shoot is studied using a pressure chamber method to extract xylem sap from plants of the two ecotypes investigated.

Materials and methods

Seeds of T. caerulescens J & C Presl were collected in Belgium (Prayon ecotype) and in the Cevennes region in southern France (Ganges ecotype). The Ganges ecotype was called ‘French A’ in our previous work (Lombi et al., 2000). The seeds were germinated on a mixture of perlite and vermiculite moistened with deionized water for the first week and then with nutrient solution. After 30 d seedlings were transferred to vessels filled with a nutrient solution containing (in µM) 1000 Ca(NO3)2, 500 MgSO4, 50 K2HPO4, 100 KCl, 10 H3BO3, 1.8 MnSO4, 0.2 Na2MoO4, 0.31 CuSO4, 0.5 NiSO4, 50 Fe(III)-EDDHA (ethylenediamine-di(o-hydroxyphenylacetic acid)), and 5 ZnSO4 (Shen et al., 1997). Solution pH was maintained at around 6.0 with 2000 µM MES (2-morpholinoethanesulphonic acid, 50% (mol/mol) as potassium salt). The plants were grown and the experiments were conducted in a controlled environment with the following conditions: 16 h d length with a light intensity of 350 µmol photons m−2 s1 supplied by fluorescent tubes, 20°C : 16°C day : night temperature, and 60–70% rh. The hydroponic solutions used were continuously aerated.

Radiotracer 109Cd and 65Zn experiments

Seedlings were grown in hydroponic vessels (3 seedlings in each 50 ml vessel) with full nutrient solution for 10 d. The nutrient solution was then replaced with a pretreatment solution containing 2000 µM MES (pH 6.0) and 500 µM CaCl2 (Lasat et al., 1996). After 24 h pretreatment, the seedlings were used for the different experiments as described in the next paragraph. All the experiments were carried out using vessels filled with an uptake solution identical to the pretreatment solution, plus either Cd or Zn.

Concentration-dependent kinetics of 109Cd and 65Zn influx

Eight different concentrations of Cd (0.2–50 µM) or Zn (0.5–100 µM) were used to study the influx kinetics of Cd and Zn. The isotopes 109Cd or 65Zn were added to the uptake solutions to give 1 µCi l−1. Each concentration was replicated three times. After 20 min uptake, the seedlings were quickly rinsed with the unlabelled pretreatment solution, and then transferred to vessels containing ice-cold desorption solutions (2000 µM K-MES, 5000 µM CaCl2, and 50 µM CdCl2 or 100 µM ZnCl2 for the Cd and Zn uptake series, respectively). Lasat et al. (1996) showed that this desorption step was effective in removing most of the Zn adsorbed on cell walls of T. caerulescens and T. arvense roots. After desorption, the seedlings were separated into roots and shoots, blotted dry and weighed. Roots and shoots were transferred into radioactivity counting vials, to which 4 ml of 5 M HNO3 were added. The samples were left to stand for at least 3 d to allow solubilization of cell contents, then 109Cd and 65Zn were assayed by gamma spectroscopy (Canberra Packard Auto Gamma 5780).

Uptake of 109Cd and 65Zn from ice-cold solutions

This experiment was performed to study the involvement of apoplastic adsorption in the apparent uptake of Cd and Zn by the two ecotypes of T. caerulescens. Uptake by active mechanisms into the symplastic pathway should be minimal from ice-cold solutions, and the apparent uptake after desorption can be mainly attributed to Cd and Zn remaining adsorbed in the root apoplast. Seedlings of the two ecotypes were transferred into ice-cold radiotracer labelled solutions containing 2000 µM K-MES, 500 µM CaCl2, and 10 µM CdCl2 or 10 µM ZnCl2. Four replicates of three seedlings each were prepared. After 20 min, the seedlings were removed from the uptake solutions, and the roots were rinsed and desorbed as described in the previous paragraph. Roots and shoots were then separated, blotted dry and weighed, prior to radio-assay of 109Cd and 65Zn.

Zn competition and 109Cd uptake

The effect of Zn on 109Cd uptake was investigated in both ecotypes of T. caerulescens. Seedlings were transferred from the pretreatment solution to uptake solutions containing either 0.2 or 10 µM CdCl2. To investigate the effect of Zn on the uptake of 109Cd (1.5 µCi l−1), a control treatment without Zn in the uptake solution and a treatment with 10 µM Zn were compared. The experiment had four replicates per treatment (three seedlings per replicate). After 20 min of uptake time the apoplastic Cd was desorbed as previously described. Then the seedlings were harvested and the radioactivity of shoots and roots measured as previously described. The results were analysed using a 2-way ANOVA of the log-transformed data.

Long-term redistribution of 109Cd and 65Zn in plants

Roots of intact seedlings of the two ecotypes were immersed in an uptake solution containing 5 µM CdCl2 and 5 µM ZnCl2, labelled with 1.5 µCi l−1109Cd and 65Zn. After 1 h the roots were rinsed and transferred to a desorption solution containing both 50 µM Cd and 100 µM Zn. In this experiment, the desorption was carried out at 20°C to minimize the stress imposed on the seedlings. After desorption the seedlings were transferred to a full nutrient solution containing 5 µM unlabelled Cd and Zn. Seedlings were harvested at 0, 24, 48, 72 and 96 h after the pulse labelling of radioactive 109Cd and 65Zn. Three replicates were included for each time point. After harvesting, roots and shoots were separated, blotted dry and weighed. 109Cd and 65Zn were measured in both roots and shoots as previously described.

Cd and Zn in the xylem sap

This experiment was conducted to study the root to shoot translocation of Cd and Zn in the two ecotypes of T. caerulescens. A pressure chamber technique was adapted from the work of Hsu et al. (1990) and Sicbaldi et al. (1997), who used this technique to study the translocation of pesticides in the xylem. This method allowed us to monitor changes in Cd and Zn concentration in the xylem sap over time. Three-month-old plants grown in hydroponics were used in the experiment. Plants of each ecotype were cut at the base of the rosette. The stem, a few mm long, was sealed to a silicon tubing using Xantopren (Heraeus Kulzer, Dormagen, Germany). The roots attached to the tubing were immersed in a 250 ml beaker inside a pressure chamber. The beaker contained a continuously aerated nutrient solution plus Cd and/or Zn. Three treatments were compared: 0 µM Cd and 100 µM Zn; 100 µM Cd and 100 µM Zn; 100 µM Cd and 5 µM Zn. The chamber was then tightly sealed and a hydro-static pressure of 0.3 MPa was generated using compressed air containing 7% oxygen. After a few minutes, the xylem sap began to flow and was collected in separate vials at hourly intervals for 5 h. Each treatment was replicated six times. Cd and Zn in the xylem sap were measured, after dilution, using either inductively coupled plasma atomic emission spectrometry (ICP-AES) or graphite furnace atomic absorption (GFAA). ATP levels were determined in root extracts following the method described by Thore (1979), to monitor the metabolic functionality of the roots. The ATP content was measured in roots kept in the pressure chambers for up to 15 h and compared to roots kept in the same solution but at atmospheric pressure.


Concentration-dependent uptake kinetics of 109Cd and 65Zn

Because 109Cd and 65Zn uptake were measured over a short period only (20 mins), the results mainly represent unidirectional influxes. The concentration-dependent uptake kinetics for 109Cd and 65Zn showed a saturable (hyperbolic) component and a nonsaturable linear component for both ecotypes (Fig. 1). These curves were mathematically resolved using SigmaPlot (Chicago, IL, USA). The best fit for the following equation (Eqn 1), that adds a linear component to the Michaelis–Menten model, was calculated for each curve:

Figure 1.

Concentration-dependent kinetics of Cd (a) and Zn (b) uptake in roots of the Ganges (open circles) and Prayon (closed circles) ecotypes of Thlaspi caerulescens. Data points and error bars represent mean values ± SE. Solid lines represent the best fit of the data using the Michaelis–Menten plus linear model.

inline image

([Me], the concentration of Cd2+ or Zn2+). Lasat et al. (1996) and Cohen et al. (1998) used a similar approach to resolve the concentration-dependent kinetics of Zn influx in T. caerulescens and Cd influx in pea seedlings, respectively. Table 1 summarizes the parameters obtained for Cd and Zn in the two ecotypes.

Table 1.  Parameters of the linear component and Michaelis–Menten model as in Eqn 1
EcotypeVmax (nmol g−1 root f. wt h−1)Km (µM)a (nmol g−1 root f. wt h−1 µM−1)
  1. SE in parentheses.

Ganges160 (16.2)0.18 (0.10)11.2 (0.58)
Prayon 33 (11.0)0.26 (0.41) 6.0 (0.37)
Ganges176 (27.1)0.30 (0.29) 8.6 (0.48)
Prayon239 (106)0.92 (1.56)10.0 (1.69)

In the case of Cd uptake, the Ganges ecotype showed a marked saturable component at low Cd concentrations, whereas in the Prayon ecotype this component was very weak (Fig. 1a). At larger concentrations the curves were dominated by a linear component for both ecotypes. Saturable Cd2+ influxes were characterized by similar Km values of 0.18 (± 0.10) and 0.26 (± 0.41) µM for the Ganges and Prayon ecotype, respectively. The maximal influx (Vmax) for Cd2+ was significantly different between the two ecotypes. The value of Vmax for the Ganges ecotype was five-fold greater than that for the Prayon ecotype. In addition, the slope for the linear component was higher for the Ganges than for the Prayon ecotype.

In contrast to Cd uptake, both ecotypes showed similar concentration-dependent uptake kinetics for Zn (Fig. 1b). Also in this case a saturable component dominated at low Zn concentrations in the uptake solution. The apparent Km values for this component were not significantly different between the two ecotypes, considering the relatively large standard errors associated with the estimated Km (Table 1). The Ganges and Prayon ecotypes also showed similar values of Vmax for Zn uptake, as well as similar slopes for the linear component.

Uptake from ice-cold solutions

The apparent uptake of Cd and Zn from ice-cold solution was similar for both metals and for both the ecotypes studied (Fig. 2). This is in contrast with the results for Cd and Zn uptake at 20°C where large differences between the ecotypes existed in term of Cd uptake (Fig. 2). In comparison with uptake at 20°C, there was an 80–85% reduction in apparent Zn uptake from the ice-cold solution in both ecotypes. At 20°C and with 10 µM Cd, the Ganges ecotype absorbed more than twice as much Cd as the Prayon ecotype. The former also showed a larger decrease (81%) in the apparent Cd uptake than the latter (60%) when the temperature was lowered from 20°C to ice-cold.

Figure 2.

Cd and Zn uptake after 20 min by the Ganges and Prayon ecotypes of Thlaspi caerulescens from ice-cold solutions (columns with horizontal rules) compared with uptake from solutions at 20°C (columns with verticle rules). Values are means + SE, n= 4.

The effect of Zn on 109Cd influx

Fig. 3 shows the effect of Zn on the uptake of 109Cd. At low Cd concentration in the uptake solution (0.2 µM), the addition of 10 µM Zn significantly reduced the apparent Cd uptake by 49% and 33% for the Ganges and Prayon ecotype, respectively. When Cd and Zn were present in the uptake solution at the same molar concentration (10 µM), the uptake of Cd was not affected in the Ganges ecotype, but was significantly reduced in the Prayon ecotype (33% reduction).

Figure 3.

Effect of Zn (0 µM, columns with horizontal rules; 10 µM, columns with verticle rules) on apparent uptake after 20 min of 109Cd as a function of Cd concentration and Thlaspi caerulescens ecotype. Values are means + SE, n = 4.

Distribution of 109Cd and 65Zn between roots and shoots

Redistribution of 109Cd and 65Zn taken up during an 1-h pulse labelling period was followed over a 4-d chase period. The Ganges ecotype took up significantly more Cd (21.3 ± 6.3 nmol g−1 f. wt h−1 than Prayon (5.6 ± 1.7 nmol g−1 f. wt h−1) during the 1-h uptake period. By contrast, the uptake of Zn was very similar in both ecotypes: 10.2 (± 6.3) and 11.0 (± 4.0) nmol g−1 f. wt h−1 for Ganges and Prayon, respectively. However, there were no significant differences between the two ecotypes of T. caerulescens in terms of the proportion of each element translocated from the roots to the shoot over the following 4-d chase period (Fig. 4). Four days after the pulse of radioactive 109Cd labelling, 33(± 5)% and 45(± 8)% of the initial 109Cd present in the roots was translocated to the shoot of plants of the Ganges and Prayon ecotype, respectively. The root to shoot transfer of 65Zn was c. 54(± 7)% in Ganges and 52 (± 2)% in Prayon. As a percentage of total metal uptake, slightly more Zn than Cd was transferred to the shoot.

Figure 4.

Translocation time-course of Cd (a) and Zn (b) from roots to shoots in the two Thlaspi caerulescens ecotypes (Ganges, closed circles; Prayon, open circles). Data are expressed as a percentage in the shoots of total Cd and Zn uptake after 1 h pulse using 109Cd and 65Zn. Values are means ± SE, n= 3.

The concentrations of Cd and Zn in xylem sap

There were no differences in the ATP content between the roots that were exposed to a pressure of 0.3 MPa inside the pressure-chamber and those that were not (data not shown), suggesting that roots inside the pressure-chamber maintained normal metabolic activities. The Cd concentration in the xylem sap was four–six-fold larger in the Ganges ecotype than in the Prayon plants, independent of the concentration of Zn in the nutrient solutions (Fig. 5). No significant differences in the concentration of Cd in the xylem sap were observed when the concentration of Zn in the nutrient solution was increased from 5 to 100 µM. The concentration of Zn in the xylem sap was similar in both ecotypes, ranging from 30 to 80 µM (Fig. 6). The Ganges ecotype showed slightly larger, but not always significant, concentrations of Zn in the xylem sap than the Prayon ecotype. The concentration of Zn in the xylem sap did not increase significantly with an increase in Zn concentration in the nutrient solution from 5 to 100 µM. The presence of 100 µM Cd in the nutrient solution seemed to decrease the concentration of Zn over time in the Prayon ecotype but had no effect on the Ganges ecotype. However, due to the fairly large standard errors, the differences in Zn concentrations between treatments were not significant. The flow rate of xylem sap was larger in plants of the Ganges ecotype compared with the Prayon ecotype (8.7 and 3.5 ml (g−1 root f. wt h−1, respectively).

Figure 5.

Concentration of Cd in the xylem sap of two Thlaspi caerulescens ecotypes (Ganges, closed circles; Prayon, open circles) at (a) 100 µM Cd + 5 µM Zn and (b) 100 µM Cd + 100 µM Zn treatments. Values are means ± SE, n= 6.

Figure 6.

Concentration of Zn in the xylem sap of two Thlaspi caerulescens ecotypes (Ganges, closed circles; Prayon, open circles) subjected to three Cd and Zn treatments. (a) 0 µM Cd + 100 µM Zn, (b) 100 µM Cd + 5 µM Zn and (c) 100 µM Cd + 100 µM Zn. Values are means ± SE, n= 6.


The present work is a continuation of the study that we conducted to compare different ecotypes of T. caerulescens (Lombi et al., 2000). In that study we carried out a series of experiments in hydroponic, glasshouse and field conditions to demonstrate that different populations of T. caerulescens differed markedly in Cd accumulation and tolerance. In this paper we characterize the physiological aspects of uptake and translocation of Cd and Zn in the two most contrasting ecotypes identified in our previous work.

The concentration-dependent kinetics of Cd and Zn influx showed a saturable component and a linear component (Fig. 1). The linear component is probably due to 109Cd and 65Zn ions that remain bound to cell walls after desorption. For divalent cations, such as Cd2+ and Zn2+, it is very difficult to completely remove metals adsorbed by the cell walls without causing significant efflux of the ions from the symplasm (Reid et al., 1996; Cohen et al., 1998). In this study we used the desorption procedure developed by Lasat et al. (1996). They showed that this procedure removed c. 80% of the Zn adsorbed on the cell walls of T. caerulescens roots. In our experiment, apparent uptake rates of Cd and Zn at the ice-cold temperature (c. 2°C) following desorption were much lower than those at 20°C (Fig. 2). Since active mechanisms of uptake should be minimal at 2°C, we conclude that the apparent uptake observed in this experiment was due to 65Zn and 109Cd that remained bound to the cell walls after desorption. The results from this experiment indicate that the amounts of 65Zn and 109Cd bound to the root cell walls after desorption were similar, and the Ganges and Prayon ecotypes had similar cell wall binding characteristics. The saturable component represents true Zn and Cd transport across the plasma membrane as concluded by Lasat et al. (2000). In the case of Cd uptake kinetics, the Ganges ecotype from southern France clearly showed a marked saturable component in the low concentration range, which was much less evident in the Prayon ecotype from Belgium (Fig. 1a). Judging from the Km values obtained, this component has a high affinity for Cd. The Km values for the two ecotypes appeared to be similar, although the fairly large error associated with Km for the Prayon ecotype makes it difficult to make a direct comparison between the ecotypes. The main difference between the two ecotypes was in Vmax, with the value obtained for the Ganges ecotype being five times larger than that for Prayon. This implies that plants of the Ganges ecotype have a higher density of a Cd transporting system on the root cell membranes than Prayon or that transport systems which are expressed are more active. The difference between the two ecotypes in the values of Vmax for Cd is similar to that observed for Zn between the hyperaccumulator T. caerulesens (Prayon ecotype) and the nonhyperaccumulator Thlaspi arvense (Lasat et al., 1996). Cohen et al. (1998) found a similar saturable component for Cd influx in iron-deficient pea roots. They concluded that Cd may be transported by a high-affinity Fe transporter, such as IRT1. However, our results show that a high affinity transporter for Cd is more active in the Ganges ecotype than in Prayon, in Fe-sufficient conditions.

No differences in terms of Km and Vmax for Zn influx were observed between the two ecotypes. This is in agreement with our previous finding that four populations of T. caerulescens, including the two ecotypes used here, had similar Zn accumulation characteristics (Lombi et al., 2000). In both ecotypes, the saturable component of the kinetics of Zn influx suggested the presence of high-affinity Zn transporters in the roots. Lasat et al. (2000) have recently cloned a high-affinity Zn transporter that is over expressed in T. caerulescens Prayon. This transporter, named ZNT1, is a member of a family of transporters responsible for the transport of micronutrients which includes the IRT1 Fe transporter (Eide et al., 1996) and the ZIP and ZRT1 Zn transporters (Zhao & Eide, 1996; Grotz et al., 1998). In their study using T. caerulescens Prayon, Lasat et al. (2000) demonstrated that ZNT1 encodes for a transporter that mediates Zn uptake as well as that of Cd albeit with a much lower affinity, and concluded that Cd may be transported in plants via the Zn transporters. Whereas this may be true for the low Cd accumulation ecotype from Prayon, it certainly cannot explain the extraordinary ability of the Ganges ecotype to take up Cd. Our results provide strong physiological evidence of a transporter with high affinity for Cd that is highly active in the root cell membranes of the Ganges ecotype of T. caerulescens from southern France. Direct evidence for such a transporter will have to come from molecular biology studies.

The experiment concerning competition between Cd and Zn for root uptake showed that at low Cd concentrations, Zn at a 50-fold higher concentration was effective in reducing the apparent Cd uptake. However, when Cd and Zn were present in the solution at equimolar concentrations (10 µM), the uptake of Cd was reduced only in the Prayon ecotype. At a low Cd concentration (0.2 µM) the presence of Zn at a much larger concentration (10 µM) may result in a reduction of either Cd absorption or Cd adsorption on the cell walls, for both ecotypes. At a larger Cd concentration in the uptake solution, the effect of Zn is significant only for the ecotype Prayon. At 10 µM Cd, there was a significant interaction (P < 0.05) between ecotypes and the Zn treatments in the 2-way ANOVA, indicating that the effect of Zn on Cd uptake was different in the two ecotypes. This is in agreement with the results obtained in a previous pot experiment (Lombi et al., 2000). In that experiment we increased the concentration of Zn in the substrate from 100 to 2000 mg kg−1, but kept the concentration of Cd at 250 mg kg−1 in both treatments. As a result, the amount of Cd in the ecotype Prayon decreased whereas the ecotype Ganges accumulated more Cd, possibly as a consequence of an increase in Cd bioavailability in the medium. These results are consistent with the hypothesis that Cd uptake is possibly mediated by Zn transporters in the Prayon ecotype, but mainly by a different transporter with high affinity for Cd in the Ganges ecotype.

Hyperaccumulator plants are characterized by a highly efficient translocation of heavy metals from roots to shoots (Baker et al., 1994; Lasat et al., 1996; Shen et al., 1997). The question arises whether the Ganges ecotype of T. caerulescens is more efficient than the Prayon ecotype in translocating Cd from roots to shoots. The experiment that used pulse labelling of radioactive 109Cd and 65Zn followed by chasing with nonlabelled Cd and Zn showed that no substantial differences existed between the two ecotypes in the distribution of 109Cd and 65Zn between roots and shoots. Therefore, the large difference between the two ecotypes in Cd accumulation is related mainly to the differences in terms of metal uptake, rather than to a difference in metal sequestration in root cell vacuoles or efficiency in xylem loading. This is in contrast with observations made for Zn in the nonhyperaccumulator T. arvense. In this species, sequestration of Zn in root vacuoles was shown to be one of the reasons for the reduced translocation of Zn to the shoots in comparison to T. caerulescens (Lasat et al., 1998).

The results from the xylem sap analyses were in agreement with the kinetics of Cd influx. In the xylem sap collected from plants of the Ganges ecotype, the concentration of Cd was 5 times larger than that from the Prayon ecotype. This difference matches the five-fold difference in Vmax for Cd between the two ecotypes. This experiment also indicated that Cd was readily loaded into the xylem after absorption. In fact, the concentration of Cd in the xylem sap was already constant 1 h after the addition of Cd to the uptake solution.


The present work demonstrates that substantial differences exist between the Ganges (southern France) and Prayon (Belgium) ecotypes of T. caerulescens in terms of Cd influx across the root plasma membranes. One (or more) high-affinity transport system(s) for Cd, different from that involved in Zn uptake, seems to play a key role in the high capacity of the Ganges ecotype of T. caerulescens to accumulate Cd. The comparison between ecotypes showed that the Cd transporter(s) is probably different from the ZNT1 Zn transporter demonstrated to be responsible for the enhanced Zn uptake by this plant. The two ecotypes investigated did not appear to differ in other characteristics involved in Cd and Zn hyperaccumulation, such as sequestration of the metals in root cell vacuoles or efficiency of translocation of the metals from roots to shoots. Further studies are needed to characterize the transporter(s) involved in Cd uptake at a molecular level. This will be a key step toward the improvement of plants to be used in phytoremediation technologies and will provide useful information concerning the mechanisms of Cd accumulation in plants.

The results from this study are significant in two aspects. First, there is considerable scope to screen for and select different ecotypes of metal hyperaccumulator plants to achieve higher phytoextraction efficiency for multiple heavy metals. Second, if there is indeed a transporter with high affinity for Cd, as suggested by the physiological evidence obtained here, then an interesting question arises as to the evolutionary advantages of having a specific transporter for a metal that has not been proved to be essential or beneficial for terrestrial higher plants.


We gratefully acknowledge financial support from DG XII of the European Commission for the PHYTOREM Project. We thank Andrew Tye, Sarah J. Dunham and Giovannella Giucani for their valuable help in the laboratory. IACR-Rothamsted receives grant-aided support from the Biotechnology and Biological Sciences Research Council of the United Kingdom.