Reduced Cd Accumulation in Zea mays: A Protective Role for

Cadmium is a nonessential trace metal and a frequent soil contaminant. Because plants vary in their accumulation of Cd, an understanding of the specif...
0 downloads 0 Views 87KB Size
Environ. Sci. Technol. 2002, 36, 5363-5368

Reduced Cd Accumulation in Zea mays: A Protective Role for Phytosiderophores? KORALIE A. HILL AND LEONARD W. LION School of Civil and Environmental Engineering, Cornell University, Ithaca, New York 14853 BETH A. AHNER* Biological and Environmental Engineering, Cornell University, Ithaca, New York 14853

Cadmium is a nonessential trace metal and a frequent soil contaminant. Because plants vary in their accumulation of Cd, an understanding of the specific processes that control uptake could reveal methods for reducing Cd levels in edible plant tissues and conversely increasing Cd accumulation in plants used for phytoremediation. Phytosiderophores are iron chelators excreted by graminaceous plants under conditions of iron limitation, but they also complex other metals including cadmium. Here we examine the influence of Cd exposure on phytosiderophore production by hydroponically grown maize. Cd increased the rate of the phytosiderophore 2′-deoxymugineic acid (DMA) release under both Fe-sufficient and Fe-limiting conditions ((Fe). In addition, the -Fe plants released more DMA while taking up less Cd than the +Fe plants. In other shortterm Cd uptake experiments, plants exposed to Cd in the presence of root exudates in which the DMA-Cd complex was likely the dominant Cd species displayed reduced Cd accumulation in root tissue, and plants similarly exposed to strongly chelated Cd in the presence of EDTA (employed as a positive control) contained the least Cd. Collectively, these results indicate that Cd stress causes Fe deficiency symptoms that result in greater DMA production by maize roots, and then the DMA appears to reduce Cd accumulation.

Introduction Cadmium is a nonessential trace element that is toxic to plants and animals. Background levels of cadmium in natural systems including agricultural soils have increased dramatically because of human activity (1, 2). Cd occurs as an impurity in phosphate fertilizer and can also contaminate soil via atmospheric deposition from smelting and refining industries and through the application of biosolids to cropland (2, 3). Since Cd primarily enters plant tissues from the soil and one of the major sources of human exposure to Cd is through food ingestion (3, 4), an understanding of the mechanisms by which Cd is transported into the plant tissue is critical to predicting accumulation and could result in altered field practices that minimize the Cd in plant tissues. The amount of Cd accumulated by different plant species varies widely (5, 6). The reasons for differences in metal uptake * Corresponding author phone: (607)255-3199; fax: (607)255-4080; e-mail: [email protected]. 10.1021/es020695z CCC: $22.00 Published on Web 11/14/2002

 2002 American Chemical Society

by plants are not well understood, although speciation in the soil solution is likely to be very important (7, 8). Plants and their associated rhizospheric bacteria can have a large affect on this speciation, and metal nutrition (such as Fe deficiency) may be a key driver (9). There are several mechanisms by which Cd can accumulate in plant tissues. Experiments with wheat (10) and peas (11) demonstrated that Cd uptake is comprised of linear and saturable components. The saturable component can be attributed to active transport of Cd across the root-cell plasma membrane mediated by nutrient metal carriers, several of which have been cloned, sequenced, and characterized (12, 13). The linear component is attributed to irreversible cell wall binding of dissolved Cd(II). Cohen et al. (11) demonstrated that conditions of Fe deficiency stimulated greater Cd accumulation in pea plants, suggesting that Cd is being taken up by an Fe(II) transporter. Elevated Fe(II) transporter (IRT1) transcription accompanied increased Cd accumulation in these plants (11) and in the Cd hyperaccumulator Thlaspi caerulescens (14) under Fe limitation. Another mechanism that may influence cadmium uptake is the exudation of organic chelators by plant roots. Plant root exudates can change the chemistry of the rhizosphere and alter metal speciation. While organic acids may lower pH and act as nonspecific metal chelators in soils (15, 16), phytosiderophores are iron-selective chelators produced by graminaceous plants (such as maize, grasses, and barley) in response to iron deficiency (17). Phytosiderophores are released from the root apex into the rhizosphere where they solubilize metals such as Fe (18). Specific transporters in the plasma membrane of root cells retrieve the phytosiderophore complex (19, 20). Evidence exists to suggest that phytosiderophores may be involved in the transport of metals other than Fe(III). Phytosiderophores have relatively strong binding constants with other metals (21, 22) and have been shown to mobilize Cd and other metals such as Cu, Zn, and Mn in soils (23-25). An iron-efficient genotype of oats, one producing enough phytosiderophores to be resistant to chlorosis (a condition caused by Fe deficiency), accumulated more Cd in its shoots than an Fe-inefficient genotype that was susceptible to chlorosis when grown in metal-enriched soil (26). The additional phytosiderophore production of the Fe-efficient variety was hypothesized to be responsible for the greater Cd uptake, but the mechanism was not determined. Phytosiderophore-mediated Zn transport was demonstrated in experiments that utilized a phytosiderophore transportdeficient mutant of maize (27). The mutant accumulated significantly less zinc in the presence of Zn-phytosiderophore complexes and absorbed much less of the whole zincphytosiderophore complex, suggesting that the wild-type absorbed the Zn complex via the Fe-phytosiderophore membrane transporter. It is thus possible that Cd-phytosiderophore complexes are also transported in this way into plant tissues. Conditions of Fe and Mn limitation can be produced by biosolid amendments to soil (28). As a consequence, it can be anticipated that plant production of phytosiderophores will increase in sludge-amended soils. Since biosolids commonly contain elevated levels of toxic metals including Cd (3), this combination of conditions (i.e., Fe/Mn deficiency, increased production of phytosiderophores, and the presence of toxic metal species) may exacerbate plant uptake of Cd in sludge-amended soils. Therefore, the specific role of phytosiderophores in plant Cd uptake warrants investigation. VOL. 36, NO. 24, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5363

TABLE 1. Total Concentrations (µM) of Added EDTA, Fe(III), and Cd in Experimental Medium medium +Fe

-Fe

no added Cd pCd ) 9.6 pCd ) 8.6 pCd ) 7.6 no added Cd pCd ) 9.6 pCd ) 8.6 pCd ) 7.6

[EDTA]T

[Fe(III)]T

[Cd]T

50 60 150 1050 30 40 130 1030

20 20 20 20 0a 0a 0a 0a

0 10 100 1000 0a 10 100 1000

a Zero added metal; however, trace amounts could be present in the reagents employed.

The aim of this study was to determine whether 2′deoxymugineic acid (DMA), the predominant phytosiderophore produced by maize, is directly involved in Cd accumulation. Using hydroponically grown maize seedlings, we examined the effects of Cd on phytosiderophore release under Fe-deficient and Fe-sufficient conditions and measured cadmium absorption by maize in the presence of root exudates containing DMA.

Materials and Methods Seedling Cultivation. At all steps, aseptic techniques were used to minimize bacterial contamination of plants. Maize seeds (Zea mays cv. Dekalb GR) were lightly abraded to promote germination and then soaked in 90% ethanol for 10 min, rinsed, and then soaked in 5% hypochloride for another 10 min. The seeds were rinsed 6 times with autoclaved deionized (Milli-Q, Millipore) water and placed in Petri dishes with nutrient agar, 3 or 4 seeds per dish, to provide moist conditions and to test the sterility of the seeds while germinating. These dishes were then transferred to a sterile, covered, opaque tray lined with moistened paper towels. The tray was incubated at 30 °C for 3 d, the average time needed for the seeds to germinate. The germinated seeds were transferred using heat- and alcohol-sterilized tweezers to 25 mm diameter foam plugs. Each seedling was inserted into a slit cut halfway into a plug. The plugs were transferred with the seedling roots pointing down into autoclaved 25 × 150 mm culture tubes containing a nutrient solution with macronutrients KH2PO4 (0.6 mM), NH4NO3 (0.4 mM), KNO3 (2.5 mM), Ca(NO3)2 (3 mM), and MgSO4 (1 mM) and micronutrients H3BO3 (12.5 µM), (NH4)6MO7O24 (0.1 µM), CuSO4 (2.0 µM), MnCl2 (3.0 µM), ZnSO4 (1.0 µM), CoCl2 (0.2 µM), and Na-EDTA (30 µM). The pH was adjusted to 5.7 by the addition of small amounts of KOH. Iron-deficient solutions (-Fe) contained no added iron, while solutions with added iron (+Fe) contained FeCl3 (20 µM) with added Na-EDTA (20 µM). Solutions used to apply a cadmium stress to the plants contained varying amounts of cadmium added as a Cd-EDTA complex so that the speciation of other metals was not altered (i.e., free ion concentrations of all other metals remains constant). Metal speciation was determined by MINEQL (29), a chemical equilibrium program. Free ion concentrations, [Cd2+], for the solutions containing cadmium were determined to be 10-9.6, 10-8.6, and 10-7.6 M (hereafter reported as pCd ) -log[Cd2+], i.e., 9.6, 8.6, and 7.6) and pFe ) 17.9 in all medium. In Cd-containing media, the concentration of the dominant inorganic cadmium chloride species (CdCl+) was roughly 20% of the free Cd2+ concentration. A summary of total Cd, Fe, and EDTA concentrations is provided in Table 1. The tubes containing the seedlings and nutrient solution were capped, and the bottoms of the tubes were covered with black plastic to simulate dark soil conditions and to encourage root growth downward. The root zone was not 5364

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 24, 2002

aerated to minimize bacterial contamination. The plants were grown in a controlled environment growth chamber (15/20 °C, 8 h dark/16 h light). Exudate Collection. At the beginning of the 15th day, plants were removed from the nutrient solution culture tubes using heat- and alcohol-sterilized tweezers and rinsed sequentially in two tubes containing 40 mL of sterile deionized water to remove metals and EDTA-containing growth medium. They were then transferred to tubes containing 25 mL of sterile deionized water and a stir bar. These tubes were returned to the growth chamber (early in the light cycle). DMA collection was performed over a time interval of 56 h, with sampling every 4-6 h for the first 24 h and twice over the remaining 32 h. Prior to sampling, the contents of the tubes were well mixed by holding the tubes over a stir plate. Samples (400 µL) were taken with ultrathin sterile electrophoresis pipette tips. Short-Term Cd Uptake. Additional experiments were conducted to measure cadmium uptake in the presence of root exudates including phytosiderophores. Exudates were collected from 10 2-week-old -Fe seedlings by placing them for 3 d in sterile deionized water as described above. The exudate-containing solutions were pooled, and the concentration of DMA was measured to be 7 µM (as described below). For the uptake experiment, a different set of plants was grown in +Fe conditions. After 14 d, the plants were rinsed as described above and transferred in triplicate to 25 mL each of three different experimental media. One treatment was the pooled root exudates amended with 0.3 µM CdCl2 and 5 mM HEPES buffer with pH adjusted to 6. The other two, also containing the same Cd and buffer, served as controls; one had no added chelator, and the other had 50 µM EDTA. One additional test tube was prepared for each treatment that did not contain a plant. The plants and tubes were incubated at 20 °C for 12 h in the light. The concentration of DMA in each tube was measured at the beginning and end of the experiments. Phytosiderophore Derivatization. Phenylisothiocyanate (PITC) derivatization of the secondary amine on the phytosiderophore was used for chromatographic separation and spectrophotometric quantification (30). The samples were placed in open Eppendorf centrifuge tubes and evaporated to dryness in a vacuum centrifuge (Savant) and then stored at -20 °C until derivatization for DMA analysis. The derivatization mixture [a 7:1:1:1 ratio of ethanol, deionized water, triethylamine (TEA), and PITC] was prepared and allowed to stand for 10 min. The dried samples (warmed to room temperature during reagent preparation) were amended with 10 µL of this mixture, vortexed, and allowed to stand for 20 min. The samples were then again dried for 30 min in a vacuum centrifuge. Then, 1 mL of 5 mM Na2HPO4 buffer adjusted to pH 7.4 with 0.1% H3PO4 was added to each sample followed by vortex mixing. The samples were analyzed by high-performance liquid chromatography (HPLC) within the next 6 h. Phytosiderophore Measurement. The following HPLC method was based on the method published by Howe et al. (30). Analyses were performed on a 3.9 × 150 mm column containing a 4-µm C-18 silica-based reverse-phase packing (Waters, Nova-pak) on a Beckman HPLC equipped with a 50-µL sample injection loop. An isocratic mobile phase was utilized consisting of 15% methanol and 85% buffer. The buffer was 0.14 M sodium acetate buffered with acetic acid to pH 6.4 with 0.2 mg of EDTA L-1 and 500 µL of TEA L-1 to block exposed silanol sites on the C-18 column. The flow rate was constant at 1 mL min-1, and the UV absorption was measured at 269 nm using a variable wavelength UV-vis detector (Shimadzu SPD-10 AV). DMA standards were generously provided by Uni-Hohenheim in Hohenheim, Germany. The isolation procedure

FIGURE 1. Calibration curves of glycine, proline, and tricine shown as nmol (in 50 µL) vs peak area. Amine groups were derivatized with PITC, and samples were run on the HPLC as described in the text. Solid lines reflect the linear best fit of the data. for these standards is described by von Wire´n et al. (27). The DMA standard was diluted, derivatized, and exhibited an HPLC retention time of approximately 6.2 min. This is slightly later than the previously reported retention time of 4.7 min under similar conditions (30). Peaks found in plant root exudates that coeluted with the DMA standard were assumed to be DMA. In follow-up experiments, we pretreated root exudates with ferrihydrite (an amorphous iron hydroxide) before PITC derivatization to test the ability of the exudates to solubilize Fe and found that the putative DMA peak’s retention time shifted significantly presumably due to the formation of an Fe complex. As the exact molar concentration of the DMA standard was unknown, it was necessary to use standards containing amines that would react similarly with PITC to estimate the concentration of DMA. Standard curves of absorbance versus concentration were prepared for PITC derivatives of three compounds: glycine containing a primary amine and tricine and proline containing secondary amines. Glycine and proline produced a similar standard curve while the tricine standard curve was significantly different, producing less peak area absorbance per nanomole injected (Figure 1). Since the structure of tricine suggests that there may be some steric hindrance in the PITC substitution reaction, the standard curves from the glycine and proline were assumed to be most appropriate for the measurement of the DMA concentrations. Reported concentrations of DMA are thus estimates. Metal Analyses. Plant root tissue was collected for determination of the trace metal concentration. The roots below the base of the foam plug were separated from each maize plant with a scalpel, dried in a 70 °C oven for 8 h, and then weighed. Root samples (∼0.1 g) were dry ashed for analysis of total Fe and Cd following the methods of Greweling (31) except that hydrogen peroxide is used as the oxidizing agent. Elemental analyses were performed on a ThermoElemental IRIS optical emission ICP with an axial torch and an yttrium internal standard. The ICP Analytical Laboratory in the Fruit and Vegetable Science Department at Cornell University performed the analyses. Cd in solutions for shortterm uptake experiments was measured by graphite furnace atomic absorption spectroscopy (GFAAS) using a PerkinElmer Analyst 100 (Norwalk, CT) equipped with a HGA 800 graphite furnace.

Results Plant Appearance. After 2 weeks of growth, the leaves of the iron-stressed plants exhibited a slight yellow color relative to the leaves of those plants with iron in their nutrient

FIGURE 2. Amount of DMA release (µmol normalized to root DW) into solution over time from plants grown under ( Fe conditions with pCd ) 7.6. “µmol” is in quotes to emphasize that this is an estimate derived from a molar concentration for DMA on the basis of a standard curve for glycine (see text). solution, indicating the effects of mild chlorosis. The root mass produced by plants was not affected by iron supply. However, the root mass was significantly lower for plants grown at the highest cadmium concentration (∼3 times less), most only having two or three lateral root stems. DMA Release Rates. The DMA release rate from control -Fe maize plants, 2.4 (( 1.5) µmol g-1 root DW d-1, was quite close to that previously reported for -Fe plants, 2 (( 0.4) µmol g-1 root DW d-1 (32); while the release from the +Fe plants, 0.84 (( 0.28) µmol g-1 root DW d-1, was somewhat higher than previously observed, 0.2 (( 0.1) µmol g-1 root DW d-1. The total amount of iron used for the +Fe plants in this experiment was 20 µM while 100 µM was used in the prior research. Since our results and those obtained by prior investigators show that plants grown in a solution containing a lower iron concentration will release phytosiderophores at a higher rate, it would be expected that maize grown in a nutrient solution with 20 µM iron would have a greater DMA release rate than the plants grown in a solution with 100 µM iron. To evaluate the effect of cadmium concentration on the DMA release rate, the concentration of DMA in solution over time was measured for both (Fe conditions with varying cadmium stress. The DMA concentration in solution increased with time for all conditions. Representative data for two different plants is shown in Figure 2 ((Fe at pCd ) 7.6). Because the rate of phytosiderophore release is not necessarily constant over time (some plants exhibit an increased production of phytosiderophores at the onset of light; 17), the DMA release rates were not calculated from the slope of the concentrations versus time but were calculated from the total DMA concentrations in solution measured at (a) 0.5 and (b) 2.3 d (Figure 3) with time zero coinciding with the start of a light cycle. The -Fe plants had, in general, higher rates of DMA release than the +Fe plants at each level of cadmium stress. The DMA release rate for maize plants in both +Fe and -Fe conditions was greater with increasing cadmium stress, although there was considerable plant to plant variability. Plant to plant variability was lower for the 2.3-d time point, and although this is a long time during which the plants have been deprived of nutrients, the rate of exudate release is not substantially different when integrated over the longer time period from that observed during the first 0.5 d. While it is possible that root leakage contributed to the total DMA release, a constant rate of leakage would be required to be consistent with the observations, which would not be expected if root integrity was deteriorating over time. The plants with the highest rate of release per gram root were the plants grown at the highest cadmium concentration. This result may reflect, in part, the lower root mass also observed in these plants. EDTA concentration is also a variable that is not readily constrained in this experiment VOL. 36, NO. 24, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5365

FIGURE 3. Rates of DMA release for ( Fe plants with varying cadmium stress at 0.5 d (a); rates for same plants after 2.3 d (b). Rates represent the total amount of DMA (in µmol) released by one plant into 25 mL of solution normalized by the root dry weight (DW) and the elapsed time in d. “µmol” is in quotes to emphasize that this is an estimate of a molar concentration for DMA on the basis of a standard curve for glycine (see text). Bars represent the average of two or three replicate plants. (Table 1), but only at the highest Cd concentration does EDTA greatly exceed concentrations typically used in nutrient medium (25-100 µM; 33), thus the significant increases observed in DMA release at pCd ) 8.6 are not likely to be caused by increased EDTA. Since bacteria can degrade phytosiderophores (34) and produce their own chelators, significant efforts were made in our experiments to minimize bacterial contamination. However, we know from plate counts that our plants were not axenic. It is possible that some of the plant to plant variability observed is due to differential degradation by bacteria; however, it is not likely that the consistent trend that we observed with respect to Cd and the presence or absence of Fe resulted from differential contamination. Furthermore, because our sterile precautions either eliminated or greatly reduced the microbial population and some period of acclimation would likely be required before an opportunistic bacterium could degrade DMA, we believe our results are likely to be free of significant artifacts caused by bacterial contamination. In addition any O2 deficiency imposed by the lack of root aeration during plant growth (eliminated to reduce microbial contamination) would have occurred in all of the plants and thus would not be expected to interfere with the experimental results. Fe and Cd Content in Roots. For +Fe plants, the concentration of iron remained fairly constant in the root tissue with increasing cadmium in the growth medium. There was a slightly higher concentration of iron in the tissue of the plants with the highest cadmium content (Figure 4b). The roots of the -Fe plants had a lower iron content that did not significantly vary with cadmium concentration, and it was presumed that the constant level of iron in the root tissue was from the iron stored in each seed. The roots of the 5366

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 24, 2002

FIGURE 4. Concentrations of (a) cadmium and (b) iron in the root tissue from ( Fe plants with varying cadmium in solution; metal in root tissue measured by ICP elemental analysis. Each bar represents the average of measurements on at least three replicate plants.

FIGURE 5. Concentration of cadmium in root tissue from plants grown in solution with no added cadmium then placed for 12 h in solution with 0.3 µM cadmium. The control contained only pH buffer and Cd. Root exudates also contained root washings from Fe-limited plants, and EDTA also contained 50 µM EDTA. Each bar is the average of three replicates. +Fe plants contained significantly more Fe than those of the -Fe plants (p < 0.05). As expected, the concentration of cadmium in the root tissue increased with increasing cadmium concentration in solution. However, with each level of cadmium stress, the +Fe plants had a greater concentration of cadmium in their tissue than the -Fe plants (Figure 4a). Cadmium Uptake Experiments. Short-term Cd accumulation by Fe-sufficient plants was examined in the presence of root exudates (including an estimated 7 µM DMA) from Fe-limited plants. Plants exposed to Cd in the presence of both root exudates and EDTA accumulated significantly less Cd (p < 0.05) than those exposed to Cd in control medium (Figure 5). Root exudates were nearly as effective as 50 µM EDTA, a strong metal chelator, at reducing the Cd bioavailability. The plants, grown in a +Fe medium, did not alter the DMA concentration in the uptake medium during the 12-h

experiment, and the amount of Cd removed from solution was commensurate with what was taken up by the plants (data not shown). While it is likely that metal ligands besides DMA were present in the root exudate treatment, back of the envelope chemical equilibrium calculations (using a conservative estimate of KML ) 1011 for the binding strength of DMA for Cd based on published stability constants for other metals such as Zn and Cu (21) and a hypothetical 1 mM citrate concentration) support the premise that Cd-DMA is the predominant species. However, in future experiments it would be preferable to use a purified DMA standard to ensure that other plant exudates are not interfering with the effects of DMA on cadmium accumulation.

Discussion This research and that of prior investigators (17, 32) has established that plants release more phytosiderophores when the availability of Fe is low. Thus, one explanation for the observed increase in phytosiderophore release with increasing cadmium stress is that cadmium may cause an iron deficiency. While total Fe concentrations in root tissues in this study do not decrease with increasing cadmium levels (Figure 4b), these plants may still be experiencing iron deficiency because of intracellular antagonisms between cadmium and iron. In a study of cadmium effects on cucumber plants, cadmium stress caused chlorosis similar to that observed in iron-deficient plants, and cadmium completely inhibited short-term iron translocation from the root to shoot (35). A study of cadmium uptake by Brassica chinensis showed that the iron concentration actually rose in plants with increasing levels of cadmium stress (36), but symptoms of chlorosis were increasingly apparent, suggesting inhibition of metabolic processes involving iron. Mn has also been shown to induce chlorosis, inhibit iron absorption and translocation from roots to shoots, and, as in our research, increase phytosiderophore production in barley (37). Shenker et al. (25) also examined the influence of Cd on phytosiderophore production in barley and wheat. Contrary to our results, they found no statistical difference between Cd-treated and untreated plants with both low and sufficient Fe in the nutrient medium. One possible explanation for the difference between our result and theirs, besides the obvious difference in plant species, is that the Cd concentration (pCd ) 8.8) used by Shenker et al. (25) was not high enough to stimulate a response. Only at pCd ) 8.6 and 7.6 do we see consistently higher rates of exudation in both +Fe and -Fe plants (Figure 3). Previous studies done on the influence of Fe limitation (or Fe-efficient vs inefficient varieties) on Cd accumulation have reported elevated Cd whereas here we report a decrease (Figure 4a). One reason is that dicots and nongrass monocots solubilize Fe from iron hydroxide by reducing Fe(III) to Fe(II), which is then taken up by an Fe(II) transporter (38). IRT1, an efficient Fe2+ transporter, is inhibited by Cd (12), and Cd transport is enhanced under conditions of Fe deficiency concurrent with increased expression of IRT1 in plants (11, 14). Thus in these plants increased Cd accumulation under Fe limitation is due to increased numbers of Fe(II) transporters. Similarly, increased Cd accumulation has been observed in Fe-limited phytosiderophore-producing plants that have been grown in soil (26, 39). While it is clear that the release of phytosiderophores mobilizes Cd from particle surfaces (24, 25), it not known whether the Cd-phytosiderophore complex is taken up directly by the plants. In our study, when maize plants are grown under conditions of plentiful Fe, Fe uptake is dominated by a reductase/Fe(II) uptake system similar to that induced in dicots and nongrass monocots under conditions of Fe deficiency (17). When maize is Fe limited, it shifts to the high-affinity Fe-phytosiderophore uptake mechanism, which

may have the combined affect of reducing the number of Fe(II) transporters and increasing the production of extracellular ligands that can bind Cd. Fewer transporters will result in lower Cd concentrations in the plant, but the question of whether Cd-phytosiderophore complexes will be taken up by the plant remains. Thus an objective of this research was to determine whether DMA plays a direct role in cadmium accumulation through membrane transport of the Cd-phytosiderophore complex. In hydroponic medium, maize plants exposed to cadmium in the presence of DMA took up less total cadmium than controls (Figure 5). While our experiments are not unequivocal because of the potential effect of other root exudates, our model calculations suggest that the DMA complex is likely the predominant complex in our experimental medium. Therefore, it is likely that the increased production of phytosiderophores in -Fe plants contributed to the lower Cd concentrations measured in these plants (Figure 4a). Together these results indicate that DMA does not provide a mechanism for direct transport of cadmium into maize plants. Experimental control through the exclusion of a soil solid phase and minimal bacterial contamination allow the mechanistic controls on DMA release and its direct role in cadmium uptake to be ascertained. Increased phytosiderophore release in soil under -Fe conditions and the resultant increase in Cd accumulation seem to be the result of delivery of dissolved Cd to the surface of the root by phytosiderophores. Further research in which a solid phase is present should be conducted to measure the influence of DMA on cadmium uptake through alteration in the soil/solution phase distribution of Cd in axenic hydroponic medium. It is likely that the release of Cd from the phytosiderophore and the ensuing Cd uptake by the plant are influenced by microorganisms present at the surface of the root. Experiments in which a defined bacterial monoculture is present are also suggested and would allow controlled evaluation of symbiotic processes (such as biodegradation of DMA and bacterial production of extracellular polymers) on cadmium uptake.

Acknowledgments Scholarship support for this project was provided to K.A.H. by the National Science Foundation, the Teresa and H. John Heinz III Foundation, and the School of Civil and Environmental Engineering at Cornell University. Funding was also provided by the USDA NY Hatch Project 123407. We thank Volker Ro¨mheld and his group at Universitaet Hohenheim in Stuttgart, Germany, for their assistance with this work. We also acknowledge Kyle Cuneo for his work on this project.

Literature Cited (1) Oehme, F. Toxicity of Heavy Metals in the Environment; Marcel Dekker: New York and Basel, 1978. (2) Nriagu, J. O., Ed. Cadmium in the Environment; WileyInterscience: New York, 1980. (3) Page, A. L.; Logan, T. J.; Ryan, J. A. Land Application of Sludge: Food Chain Implications; Lewis Publishers: Chelsea, MI, 1987. (4) Wagner, G. J. Adv. Agron. 1993, 51, 173-212. (5) Hamon, R.; Wundke, J.; McLaughlin, M.; Naidu, R. Aust. J. Soil. Res. 1997, 35, 1267-1277. (6) Li, Y.; Chaney, R. L.; Schneiter, A. A.; Miller, J. F. Crop Sci. 1995, 35, 137-141. (7) Chaney, R. L. In Metal Speciation: Theory, Analysis and Application; Kramer, J. R., Allen, H. E., Eds; Lewis Publishers: Chelsea, MI, 1988; pp 219-259. (8) Sauve´, S.; Norvell, W.; McBride, M.; Hendershot, W. Environ. Sci. Technol. 2000, 34, 291-296. (9) Rengel, Z. Plant Soil 1997, 196, 255-260. (10) Hart, J.; Welch, R.; Norvell, W.; Sullivan, L.; Kochian, L. Plant. Physiol. 1998, 116, 1413-1420. (11) Cohen, C. K.; Fox, T. C.; Garvin, D. F.; Kochian, L. V. Plant. Physiol. 1998, 116, 1063-1072. VOL. 36, NO. 24, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5367

(12) Eide, D.; Broderious, M.; Fett, J.; Guerinot, M. L. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 5624-5628. (13) Rogers, E. E.; Eide, D. J.; Guerinot, M. L. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 12356-12360. (14) Lombi, E.; Tearall, K. L.; Howarth, J. R.; Zhao, F.-J.; Hawkesford, M. J.; McGrath, S. P. Plant. Physiol. 2002, 128, 1359-1367. (15) Mench, M.; Morel, J.; Guckert, A.; Guillet, B. J. Soil Sci. 1988, 39, 521-527. (16) Krishnamurti, G.; Cieslinski, G.; Huang, P.; Van Rees, K. J. Environ. Qual. 1997, 26, 271-277. (17) Marschner, H.; Ro¨mheld, V. Plant Soil 1994, 165, 261-274. (18) Marschner, H.; Ro¨mheld, V.; Kissel, M. Physiol. Plant. 1987, 71, 157-162. (19) von Wire´n, N.; Mori, S.; Marschner, H.; Ro¨mheld, V. Plant. Physiol. 1994, 106, 71-77. (20) Curie, C.; Panaviene, Z.; Loulergue, C.; Dellaporta, S. L.; Briat, J.-F.; Walker, E. L. Nature 2001, 409, 346-349. (21) Murakami, T.; Ise, K.; Hayakara, M.; Kamei, S.; Takagi, S. Chem. Lett. 1989, 1989, 2137-2140. (22) Sugiura, Y.; Nomoto, K. In Structure and Bonding; Clarke, M. J., et. al., Eds; Springer-Verlag: Berlin, 1984; pp 107-135. (23) Treeby, M.; Marschner, H.; Ro¨mheld, V. Plant Soil 1989, 114, 217-226. (24) Awad, F.; Ro¨mheld, V. J. Plant Nutr. 2000, 23, 1847-1855. (25) Shenker, M.; Fan, T. W.-M.; Crowley, D. E. J. Environ. Qual. 2001, 30, 2091-2098. (26) Mench, M.; Fargues, S. Plant Soil 1994, 165, 227-233. (27) von Wire´n, N.; Marschner, H.; Ro¨mheld, V. Plant. Physiol. 1996, 111, 1119-1125.

5368

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 24, 2002

(28) Brown, S.; Angle, J. S.; Chaney, R. L. J. Environ. Qual. 1997, 26, 1375-1384. (29) Westall, J. C.; Zachary, J. L.; Morel, F. M. M. MINEQL: A Computer Program for the Calculation of Chemical Equilibrium Composition of Aqueous Systems; R. M. Parsons Laboratory, Massachusettes Institute of Technology: 1976. (30) Howe, J. A.; Choi, Y. H.; Loeppert, R. H.; Wei, L. C.; Senseman, S. A.; Juo, A. S. R. J. Chromatogr. 1999, 841, 155-164. (31) Greweling, T. In The Cornell University Agricultural Experiment Station Bulletin; Cornell University: Ithaca, 1976; p 4. (32) Marschner, H.; Ro¨mheld, V. Plant Soil 1990, 123, 147-153. (33) Parker, D. R.; Norvell, W. A. Adv. Agron. 1999, 65, 151-213. (34) von Wire´n, N.; Ro¨mheld, V.; Morel, J.; Guckert, A.; Marschner, H. Soil Biol. Biochem. 1993, 25, 371-376. (35) Fodor, F.; Sarvari, E.; Lang, F.; Szigeti, Z.; Cseh, E. J. Plant Phys. 1996, 148, 434-439. (36) Wong, M. K.; Chuah, G. K.; Koh, L. L.; Ang, K. P.; Hew, C. S. Environ. Exp. Bot. 1984, 24, 189-195. (37) Alam, S.; Kamei, S.; Kawai, S. J. Plant Nut. 2000, 23, 1193-1207. (38) Fox, T. C.; Shaff, J. E.; Grusak, M. A.; Norvell, W. A.; Kochian, L. V. Plant. Physiol. 1996, 111, 93-100. (39) Ro¨mheld, V.; Awad, F. J. Plant Nutr. 2000, 23, 1857-1866.

Received for review April 17, 2002. Revised manuscript received October 8, 2002. Accepted October 9, 2002. ES020695Z