INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Arsenic may play an essential role in animal nutrition (Uthus, 1992, 1994), perhaps in Met metabolism, but there is no doubt that the element is principally renowned for its toxicity (National Research Council, 1977). Indeed, arsenic toxicity in humans has recently become evident on a very large scale in Bangladesh (Dhar et al., 1997), and the National Research Council has recently recommended that the maximum contaminant level standard for drinking water in the U.S. be lowered from the current value of 50 µg L¹ (National Research Council, 1999). Arsenic is also toxic to plants and microorganisms and has been used in pesticides, herbicides, preservatives, and pharmaceuticals (National Research Council, 1977). Many of these uses continue today, and therefore it is important to remediate past contamination (Dutre et al., 1998). In this paper we address arsenate uptake by Indian mustard (Brassica juncea) plants growing hydroponically. Our data suggest that arsenate (As^V) enters the roots as a phosphate analog and is promptly reduced to As^III. Little arsenic is transported to the aboveground tissues. The addition of dimercaptosuccinate to the hydroponic growth solution caused significant amounts of arsenic to move into the shoot, perhaps offering a way of removing arsenate from contaminated soils.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plant Growth

Indian mustard (Brassica juncea [L.] Czern. variety 426308) (Kumar et al., 1995) plants were grown under microbiologically controlled conditions such that their roots were maintained axenically. Seeds were surface-sterilized in 2.6% (w/v) sodium hypochlorite for 30 min, rinsed four times in autoclaved de-ionized water, and transferred onto sterile 1.2% (w/v) agarose plates containing 3.0% (w/v) Suc. Plates were held vertically and the seeds allowed to germinate and grow in the dark at 22°C for 72 h. Etiolated seedlings not showing microbial contamination on the agarose plates were transferred individually into small glass vials (29 × 65 mm) containing 23 mL of sterile nutrient solution. Soft styrofoam stoppers used to cap the vials were incised radially to provide support for the hypocotyls. The nutrient solution contained 0.7 mM Ca²⁺, 1.5 mM K⁺, 0.5 mM Mg²⁺, 0.25 mM NH₄⁺, 2.9 mM NO₃, 0.25 mM H₂PO₄, 0.5 mM SO₄², 4.75 µM ferric tartrate, 0.075 µM Cu²⁺, 0.2 µM Zn²⁺, 1.25 µM Mn²⁺, 11.5 µM H₃BO₃, and 0.025 µM MoO₃ at pH 5.5. When applicable, arsenate was added to the hydroponic solution as sodium arsenate. Vials were agitated on an orbital shaker (Lab-Line Instruments, Melrose Park, IL) at 60 rpm to provide aeration and mixing, and the nutrient solution was replaced weekly. Plants were cultivated in a growth chamber with a 10-h light period, with light provided by fluorescent and incandescent lamps at an illuminance of 17,200 lux. All plants were maintained at a constant temperature of 22°C and a relative humidity of 50%, during both day and night. After 11 d, plants were transferred under microbiologically controlled conditions to 23 mL of fresh nutrient solution containing various treatments and exposed to the conditions described above.

Arsenic Quantitation

In experiments in which non-radioactive arsenic was used (Table II), arsenic was analyzed as follows. The plant tissue was dried at 70°C and then wet ashed using nitric and perchloric acids according to standard methods (Jones and Case, 1990). The resulting solution was analyzed for arsenic content by inductively coupled plasma spectrometry (Fisons Accuris, Fisons Instruments, Beverly, MA). Certified National Institute of Standards and Technology plant standards (peach leaves) were carried through the digestions and analyzed as part of the QA/QC protocol. Reagent blanks and internal standards were used where appropriate to ensure accuracy and precision in the analysis. For experiments employing ⁷³As (Tables I and V), the radioactive isotope was added in the sodium arsenate form and measured in fresh tissue using a gamma counter (model C5002, Packard Instruments, Meriden, CT).

X-Ray Absorption Spectroscopy

Indian mustard root and leaf tissue samples for x-ray absorption spectroscopy were frozen in liquid nitrogen immediately after harvesting and kept frozen until after measurements were completed. Immediately prior to analysis, tissues were carefully ground and packed into 2-mm-pathlength lucite sample holders under liquid nitrogen. Xylem sap from Indian mustard seedlings was collected by decapitating the plant just above the root and collecting the sap that was exuded under root pressure (Salt et al., 1995). Sap was usually collected for up to 1 h after decapitation. After collection, the sap was immediately frozen in liquid nitrogen. Prior to analysis, sap was thawed and 30% (v/v) glycerol was added as a glassing agent. For analysis, the sample was loaded into 2-mm-pathlength lucite sample holders and frozen in liquid nitrogen.

Solutions of arsenic model compounds had a concentration of between 5.0 and 7.5 mM after the addition of 30% (v/v) glycerol as a glassing agent. They were pipetted into 2-mm-path-length lucite sample holders and frozen in liquid nitrogen. The arsenic(III)-tris-glutathione and arsenic-dimercaptosuccinate complexes were made by adding a 10-fold molar excess of glutathione or dimercaptosuccinate to a solution of sodium arsenite to give a pH 5.5 solution.

X-ray absorption spectra were measured at the Stanford Synchrotron Radiation Laboratory (SSRL) on Beamline 7-3 using a double crystal monochromator (Si220), with an upstream vertical aperture of 1 mm. Harmonic rejection was accomplished by detuning one monochromator crystal to approximately 50% off peak, and no specular optics were present. The incident x-ray intensity was monitored using a N₂-filled ionization chamber and arsenic K-edge x-ray absorption spectra were measured by monitoring the arsenic K fluorescence using a 13-element Ge detector (Canberra Industries, Meriden, CT). During data collection, samples were maintained at a temperature of 10 K using a flowing liquid helium cryostat (Oxford Instruments, Concord, MA). The absorption of an elemental arsenic foil was measured simultaneously by transmittance to calibrate the monochromator for each spectrum; the first energy inflection was assumed to be 11867 eV.

X-ray absorption spectra were collected using the program XAS Collect (M.J. George, unpublished) and spectra were analyzed using the EXAFSPAK suite of programs (http://ssrl.slac.stanford.edu/exafspak.html) according to established methods. Near-edge spectra were quantitatively analyzed using a total curve-fitting method in which the near-edge spectra of plant samples were fit to those of arsenic solution models. All small components (<1%, or where 95% confidence limits exceeded value) were excluded from final fits. Examples of the fits achieved are shown in Figure 1. Extended x-ray absorption fine structure (EXAFS) spectra were quantitatively curve-fit according to standard methods (e.g. see Pickering et al., 1999) using ab initio phase-shift and total amplitude functions calculated using the program Feff v7.02 (Rehr et al., 1991).

View larger version (19K): [in this window] [in a new window]   Figure 1.   Edge fitting of selected arsenic K near-edge spectra from Indian mustard. Sample spectra were fitted with combinations of spectra from various model compounds. The results of these fits are presented as percentages. A, Roots from plants treated with 250 µM arsenate for 5 d fitted with 97% AsIII-tris-glutathione at pH 5.5 (a) and 3% arsenate at pH 9.0 (b). B, Xylem sap exudate from plants treated with 25 µM arsenate for 5 d fitted with 41% arsenate at pH 9.0 (a) and 59% arsenite at pH 5.5 (b). C, Leaves from plants treated with 250 µM arsenate and 250 µM dimercaptosuccinate for 2 d fitted with 33% AsIII-tris-glutathione at pH 5.5 (a) and 67% AsIII-dimercaptosuccinate at pH 5.5 (b). In all panels the dots, solid line, and dotted line below show the data, the best fit, and the residual, respectively.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Table I shows the uptake of arsenate by 2-week-old Indian mustard seedlings exposed to 250 µM arsenate for 2 and 5 d in axenic hydroponic solution. The roots appear to accumulate the majority of the arsenic, with only very low levels of arsenic detected in the stem and leaves. This uptake is slightly inhibited by phosphate (Table II), suggesting that phosphate and arsenate are transported by the same uptake system. This inhibition is not very strong, however, since the phosphate concentration was twice that of the arsenate in the experiment shown in Table II, yet arsenate uptake into roots was only inhibited by 72%.

X-ray absorption spectroscopy provides a unique tool for studying the chemical form of an element with minimal pretreatment of the sample. The technique is particularly valuable for studying the movement of metal and metalloid ions in plants (Salt et al., 1995, 1997, 1999; Kramer et al., 1996; De Souza et al., 1998; Lytle et al., 1998; Orser et al., 1998/1999; Pilon-Smits et al., 1998, 1999; Zayed et al., 1998) and fungi (Sarret et al., 1998), because it detects all forms of the element under study. In favorable cases, it is possible to identify the principal chemical components of mixtures of species if the spectra of appropriate model compounds are available. Figure 2 compares the x-ray absorption near-edge spectra of aqueous arsenite and arsenate as a function of pH. The spectra are strongly pH dependent, which is expected because of the ionization of the oxygen atoms.

View larger version (22K): [in this window] [in a new window]   Figure 2.   Arsenic K near-edge spectra of aqueous solutions of arsenite and arsenate. Arsenite is shown at pH 5.5 (solid line) and pH 12 (dashed line), and arsenate at pH 4.5 (solid line) and pH 9 (dashed line).

The pH values for arsenite and arsenate were specifically chosen with reference to their pK_a values to maximize the abundance of a single ionized species. Thus, in the case of arsenite, the first pK_a is 9.2 (Greenwood and Earnshaw, 1990), and the spectra of Figure 2 correspond to the solution species As(OH)₃ and [As(OH)₂O] at pH 5.5 and 12, respectively. The As(OH)₃ spectrum has a sharper near-edge peak due to increased degeneracy of the valence p-levels allowed by higher symmetry. In the case of arsenate, the acid is tribasic, with pK_a values of 2.2, 6.9, and 11.5 (Greenwood and Earnshaw, 1990), and we thus expect the spectra at pH 4.5 and 9.0 to correspond to [As(OH)₂O₂] and [As(OH) O₃]², respectively. The latter, more symmetric species gives the sharper near-edge peak, again, as would be expected from increased orbital degeneracy. Also as expected, at lower pH values (data not shown) the EXAFS spectra of arsenate showed evidence of polymerization. In any case, it is important that the correct form be used in attempting to fit the spectra of plant material. Figure 3 shows the absorption spectra of the arsenite and arsenate species that are the most abundant at neutral pH and compares them with spectra of some other biologically relevant arsenic compounds.

View larger version (20K): [in this window] [in a new window]   Figure 3.   Arsenic K near-edge spectra of AsIII-tris-glutathione at pH 5.5, arsenite at pH 5.5, dimethylarsinate at pH 9.0, and arsenate at pH 9.0.

Figure 4 compares the spectra of roots and shoots of Indian mustard plants 5 d after 25 µM arsenate was added to the hydroponic culture solution. The spectra of aqueous arsenate and the As^III-tris-glutathione complex are also included in this figure. It is clear that the spectra of roots and shoots are essentially identical and very similar to that of the arsenic(III)-tris-glutathione complex. The spectra of roots and shoots from plants grown with 250 µM arsenate in the hydroponic culture solution were also collected. Best-fit analysis suggested the compositions shown in Table III; the vast majority of arsenic is well modeled by the As^III -tris-glutathione complex, but the roots at the higher concentration do show a minor component of arsenite (see also Fig. 1A).

View larger version (16K): [in this window] [in a new window]   Figure 4.   Arsenic K near-edge spectra of root and leaf of Indian mustard treated with 25 µM arsenate for 5 d. The spectra are compared with those of AsIII-tris-glutathione at pH 5.5 [AsIII(SG)3] and arsenate at pH 9.0.

The similarity between the plant spectra and that of As^III-tris-glutathione is confirmed in the EXAFS analysis (Fig. 5; Table IV), which shows that arsenic in both the plant roots and shoots and in the arsenic(III)-tris-glutathione complex has three sulfur ligands at 2.25 ± 0.01 Å.

View larger version (31K): [in this window] [in a new window]   Figure 5.   Arsenic K-edge EXAFS (A) and corresponding Fourier transforms (B) of Indian mustard and arsenic model compounds. Top to bottom: Root of Indian mustard treated with 25 µM arsenate for 5 d; leaves of Indian mustard treated with 250 µM arsenate for 5 d (note that leaves from the 25 µM arsenate treatment were too weak for EXAFS); AsIII-tris-glutathione at pH 5.5 [AsIII(SG)3]; dimethylarsinate at pH 9.0 (DMA); arsenite at pH 5.5; and arsenate at pH 9.0. Solid lines show the data and the dashed lines the best fit (Table IV). Fourier transforms have been phase-corrected for the first shell (As-S or As-O). In A, the ordinate zero is indicated for each spectrum by a dotted line. Thus, the spectra are offset vertically but are plotted with the same relative scale.

View this table: [in this window] [in a new window]   Table IV.   Results of fitting arsenic K-edge EXAFS of Indian mustard and arsenic solution species Coordination number (N), interatomic distance (R), Debye-Waller factors (2), and energy offset to nominal threshold value of 11885 eV (E0). The fit-error is defined as k6(obscalc)2/k6obs2. Nos. in parentheses after a value indicate three times the estimated SD of the last digit(s) of the value.

The x-ray absorption spectrum of xylem sap collected from plants grown on 25 µM arsenate for 5 d is shown in Figure 1B. Although the concentration of arsenic is extremely low (approximately 1 µM), it is clear that the spectrum can be modeled very well by summing the arsenate and arsenite species that are the most abundant at neutral pH. The best fit of the near-edge data is shown in Figure 1B and Table III. The arsenic concentration was too low for EXAFS data to be collected.

As shown in Tables I and II, the amount of arsenic transported to the leaves is only a small fraction of that absorbed into (or adsorbed onto) the roots. Adding the chelating agent dimercaptosuccinate to the growth medium dramatically alters this distribution and allows the arsenic to become distributed throughout the plant (Table V). The EXAFS spectra of roots and shoots of plants harvested after either 5 d of exposure to 250 µM arsenate or after 2 d of exposure to 250 µM arsenate and 250 µM dimercaptosuccinate in the hydroponic solution (Table IV, spectra not displayed) show that the arsenic is coordinated by three sulfurs at 2.25 Å. The spectra are essentially the same as those of As^III-dimercaptosuccinate and As^III-tris-glutathione complexes in water, which are themselves almost indistinguishable. However, the near-edge spectra of the same samples (Fig. 6) do show subtle differences in the region beyond the first strong peak, allowing the speciation to be determined. Fitting the plant spectra with the spectra of known model compounds yields the distribution shown in Table VI. As expected, no As^III-dimercaptosuccinate was observed when spectra acquired from plants exposed to arsenate alone were analyzed (Table VI).

View larger version (21K): [in this window] [in a new window]   Figure 6.   Arsenic K near-edge spectra of root and leaf of Indian mustard treated with both 250 µM arsenate and 250 µM dimercaptosuccinate for 2 d compared with those of AsIII-tris-glutathione at pH 5.5 and AsIII-dimercaptosuccinate at pH 5.5. The inset shows the region around 11,880 eV on an expanded scale.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The data presented here show that arsenate is accumulated by roots of Indian mustard, a result similar to that reported for arsenite by Carbonell-Barrachina et al. (1994) with tomato roots. In Indian mustard roots (and in tobacco, data not shown), the arsenate (As^V) is reduced to As^III, and coordinated by three sulfur ligands, which can be modeled as the As^III-tris-glutathione complex. In aqueous solution, thiols such as glutathione will reduce arsenate to arsenite (Delnomdedieu et al., 1994; Carter, 1995), with concomitant formation of the disulfide form of glutathione. The high affinity of arsenite for thiols is well known, and in the presence of excess glutathione the As^III-tris-glutathione complex will form. It is thus possible that the initial reduction of arsenate in Indian mustard is by glutathione, but the presence of a mixture of arsenate and arsenite in the xylem sap may argue against this. The reduction and coordination observed in Indian mustard is reminiscent of the arsenic resistance system in Leishmania tarentolae (Dey et al., 1996), which expels As^III-tris-glutathione. In plant cells it is likely that the arsenic is transported into and stored within the cell vacuole, but this remains to be confirmed. If it is indeed stored within the vacuole, the coordinating thiols may be those of phytochelatins. In support of this, a recent finding indicates that a mutant Arabidopsis lacking the ability to synthesize phytochelatins is much more sensitive to arsenate than the wild-type plant (Ha et al., 1999). Also, deletion of the gene required for phytochelatin synthesis in the fission yeast Schizosaccharomyces pombe was found to be sufficient to confer arsenate sensitivity in the deletion mutant (Ha et al., 1999). In contrast, glutathione does not appear to be involved in arsenic resistance in bacteria (Silver, 1996; Latinwo et al., 1998).

The finding that arsenate uptake was inhibited by phosphate at concentrations in the range of the low-affinity phosphate transport system was not unexpected, as it has been reported in other species (Meharg et al., 1994; Cox et al., 1996). Under some circumstances, however, phosphate can actually stimulate arsenate uptake by plants by increasing the bioavailability of arsenate (Peryea, 1998), so attempts to use plants to remove arsenic from soils (phytoremediation) need to take the multiple affects of phosphate into consideration.

While arsenate was complexed to sulfur and stored within the root tissue of hydroponically grown Indian mustard plants, a small proportion of arsenic was translocated in the xylem sap to the shoots. The arsenic in the xylem sap was present as the oxyanions arsenate and arsenite and was not coordinated by sulfur. An analogous difference in the chemical form between stored and transported forms of Cd has been observed in Indian mustard (Salt et al., 1995).

A prerequisite for successful phytoremediation (Salt et al., 1998) of arsenic-contaminated soils by Indian mustard is increased transport of the metalloid into the harvestable aboveground tissues. One approach is to apply chemical chelators to the soil, and this has been effective at enhancing the plant accumulation of lead (Huang and Cunningham, 1996; Blaylock et al., 1997; Huang et al., 1997), uranium (Huang et al., 1998), and gold (Anderson et al., 1998). We have shown that total arsenic translocation to the shoot increases when the As^III chelator dimercaptosuccinate is added to the hydroponic solution (Table V). In both roots and leaves, arsenic was mainly distributed between As^III-tris-glutathione and As^III-dimercaptosuccinate, but the leaves had the greater proportion of the latter (Table VI). This is in agreement with the well-known affinity of As^III for dithiols. Indeed, Delnomdedieu et al. (1993) have found that dimercaptosuccinate will displace glutathione from the As^III-tris-glutathione complex. It is noteworthy that the total amount of arsenic accumulated by the plants was not substantially increased by the presence of dimercaptosuccinate. However, the As^III-dimercaptosuccinate complex appears to be more effectively translocated from root to shoot, causing a redistribution of As^III from the roots to the harvestable shoots, a very important prerequisite for phytoremediation. To our knowledge, this is the first time arsenic chelates have been shown to enhance arsenic translocation in plants. Elucidation of the detailed mechanism of action of dimercaptosuccinate must await further study, but nevertheless provides a potentially useful amendment in stimulating the removal of arsenic from contaminated soil.