Rhizosphere Bacteria Mobilize Zn for Hyperaccumulation by Thlaspi

Jun 28, 2001 - Root Fungal Endophytes Enhance Heavy-Metal Stress Tolerance of Clethra barbinervis Growing Naturally at Mining Sites via Growth Enhance...
0 downloads 12 Views 218KB Size
Download PDF
Environ. Sci. Technol. 2001, 35, 3144-3150

Rhizosphere Bacteria Mobilize Zn for Hyperaccumulation by Thlaspi caerulescens STEVEN N. WHITING,† MARK P. DE SOUZA, AND NORMAN TERRY* Department of Plant and Microbial Biology, 111 Koshland Hall, University of California, Berkeley, California 94720-3102

Thlaspi caerulescens has a remarkable ability to hyperaccumulate Zn from soils containing mostly nonlabile Zn. The present study shows that rhizosphere microbes play an important role in increasing the availability of watersoluble Zn in soil, thus enhancing Zn accumulation by T. caerulescens. The addition of bacteria to surface-sterilized seeds of T. caerulescens sown in autoclaved soil increased the Zn concentration in shoots 2-fold as compared to axenic controls; the total accumulation of Zn was enhanced 4-fold. When the same experiment was conducted with Thlaspi arvense, a nonaccumulator, bacteria had no effect on shoot Zn accumulation although they increased watersoluble Zn concentrations available to both Thlaspi species by 22-67% as compared to the axenic controls. Further evidence that bacteria increase the availability of water-soluble Zn in soil was obtained when liquid media that had supported bacterial growth mobilized 1.3-1.8-fold more Zn from soil as compared to axenic media. Other experiments with agar media showed that bacteria did not facilitate an increase in the rate of soluble Zn transport into the root nor did they enlarge the surface area of the roots of either Thlaspi species. Thus, the bacterially mediated increase in the dissolution of Zn from the nonlabile phase in soil may enhance Zn accumulation in T. caerulescens shoots.

Introduction A small number of plant species possess a unique ability to accumulate metals to exceptionally high concentrations in their tissues without symptoms of toxicity. These plants are termed metal “hyperaccumulators” (1). In the case of zinc, hyperaccumulators are characterized by concentrations greater than 10 000 µg of Zn g-1 shoot dry weight when growing in their natural habitats (2). Of the 400 taxa of metal hyperaccumulators identified to date (2), the Zn-hyperaccumulating plant Thlaspi caerulescens J. & C. Presl has received the most attention because of its extraordinary ability to extract and accumulate Zn from soils (e.g., refs 3 and 4). As a result of this ability to remove considerable quantities of metals from soils, such hyperaccumulators are attracting interest from scientists interested in phytoremediation, a cost-effective, “green” technique for decontaminating polluted environments (e.g., refs 2 and 5). * Corresponding author telephone: (510)642-3510; fax: (510)6424995; e-mail: [email protected]. † Present address: Department of Botany, University of Melbourne, Parkville, Victoria 3010, Australia. 3144

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 15, 2001

T. caerulescens has been studied extensively with respect to its plant-based characteristics for Zn hyperaccumulation when Zn is present in the growth medium in water-soluble (i.e., labile or “bioavailable”) forms. T. caerulescens has been shown, for example, to have a considerably enhanced absorption capacity as well as an increased capacity for translocation of Zn from root to shoot as compared to nonaccumulator plants (e.g., refs 6-10). Even more remarkable is the ability of T. caerulescens to accumulate Zn from soils in which Zn is present in largely nonlabile forms; research suggests that T. caerulescens has the ability to solubilize nonlabile forms of soil Zn (11, 12). The mechanisms involved in this solubilization of nonlabile Zn during the growth of T. caerulescens have not been elucidated. Knight et al. (11) proposed two possible mechanisms for the Zn solubilization they recorded during the growth of T. caerulescens: (i) the soils used had large buffering capacities to replenish Zn in the soil solution within a short time or (ii) T. caerulescens was efficient at mobilizing Zn that was not initially soluble. To date, an active metal-mobilizing mechanism operating in the roots of T. caerulescens has not been identified. In this paper, we present data indicating that rhizosphere microbes may play an important role in increasing the availability of water-soluble Zn for plant uptake by T. caerulescens. Microbes (bacteria and fungi) are ubiquitous in soils to which hyperaccumulators are native, even in those soils containing high concentrations of metals (13, 14). It is known that bacteria mobilize iron in soils in response to Fe deficiency by producing metal-chelating compounds such as siderophores (15, 16), which enhance the nutrition and Fe uptake of nonaccumulator plants (17). Rhizosphere bacteria have also been shown to enhance the accumulation of Se, Cd, and Hg in the roots of nonaccumulator plants (18-20). Rhizosphere bacteria might influence Zn accumulation by T. caerulescens in several different ways. First, the presence of bacteria in the rhizosphere of nonaccumulator plants has been shown to have beneficial effects on root growth and root hair production (21, 22); this growth promotion could increase the Zn uptake in T. caerulescens by increasing the functional surface area of its roots. Second, the presence of bacteria in the rhizosphere of T. caerulescens might increase the transport of soluble Zn into the roots, as has been found for potassium and nitrate uptake by Brassica napus (23). Third, the bacteria may facilitate the solubilization of nonlabile forms of Zn in the soil, which would also increase Zn uptake by T. caerulescens. We tested each of these hypotheses to determine the mechanism of bacterially mediated Zn accumulation by T. caerulescens and conclude that the third hypothesis is correct, i.e., the microbes increased the availability of water-soluble Zn in the soil.

Materials and Methods Preparation of Axenic Seeds. The effects of microbes on Zn uptake by plants were investigated using two species of Thlaspi, the Zn hyperaccumulator (Thlaspi caerulescens) and the related nonaccumulator (Thlaspi arvense L.). Neither T. caerulescens nor T. arvense have mycorrhizae associated with their roots. The seeds of T. caerulescens were from a population growing on Zn- and Cd-contaminated soil close to a former Zn and Cd smelter at Prayon, Belgium. Seeds of T. arvense were obtained from the Crucifer Genetics Cooperative, Madison, WI. For all experiments, seeds of both species were surface-sterilized by shaking in 70% ethanol for 30 s, followed by shaking in 10% sodium hypochlorite for 30 10.1021/es001938v CCC: $20.00

 2001 American Chemical Society Published on Web 06/28/2001

TABLE 1. Physicochemical and Microbiological Analyses of the Damp Berkeley Soil pH total Zn water-extractable Zn total Fe water-extractable Fe total S total microbial counts (×106)

6.81 ( 0.05 179 ( 2 57.46 ( 11.25† 8399 ( 164 547 ( 233† 208 ( 8 7.4 ( 1.1†

a All concentrations are given in mg kg-1 soil dry weight except for water-extractable metals (µg kg-1). Total microbial counts (bacteria plus fungi) are given as colony forming units (CFU) per gram of soil fresh weight. Values are means ( SE, n ) 4, except for the dagger (†) which indicates that n ) 3.

min, and by five 5-min washes with sterile double-distilled water (DDW). Bacterial Inoculation of Seeds for Soil and Agar Experiments. We randomly selected three strains of rhizosphere bacteria from those we isolated previously (18). The three strains (BJ1, BJ5, and BJ10) were identified as Microbacterium saperdae, Pseudomonas monteilii, and Enterobacter cancerogenes, based on >99% similarity in their 16S rDNA sequences (de Souza and Terry, unpublished results). A mixed inoculum of these three strains was prepared to add to the seeds: one loopful of each bacterial strain was added to 5 mL of sterile 0.85% saline (NaCl) solution in a sterile plastic tube and vortexed to suspend the bacteria. One milliliter of this suspension was added to 4 mL of sterile 0.5% methylcellulose solution (Sigma) prepared with 0.85% sterile saline solution to provide a bacterial suspension for inoculation of seeds. Subsequently, 2.5 mL of the bacteria/methylcellulose suspension was transferred to a clean sterile plastic tube and autoclaved to provide a suspension of dead bacteria as an axenic control. An additional tube containing 2.5 mL of sterile methylcellulose was not inoculated with bacteria to provide a second axenic control. To estimate the number of bacteria in the culture, serial dilutions of each of the three methylcellulose solutions were spread on 1.5% trypticase soy agar (TSA) plates and incubated at 25 °C for 7 days. This showed that the live bacteria suspension contained a total of 2.52 × 108 colony forming units (CFU) per milliliter. The dead bacteria and axenic methylcellulose suspensions both contained no viable bacteria, demonstrating that these controls were indeed axenic. The methylcellulose suspensions were used to prepare three seed treatments. For the principal treatment (plus bact), one-third of the surface-sterilized seeds of both T. caerulescens and T. arvense were soaked in the methylcellulose solution containing live bacteria. The second one-third of the seeds of each species were soaked in the axenic methylcellulose solution as a control (no bact). The remaining seeds were soaked in sterile 0.5% methylcellulose solution containing dead (autoclaved) bacteria to provide a second control treatment (dead bact). After 20-min soaking time, all seeds were removed from the methylcellulose solutions and dried on sterile filter papers in a laminar-flow cabinet. All subsequent seed transfers from the filter papers to the growth media described below were performed using flamed, autoclaved forceps in a laminar-flow cabinet. Effect of Bacteria on Plants Grown in Soil. This experiment tested the effect of bacteria on the Zn accumulation of T. caerulescens and T. arvense growing in soil where most of the Zn was not initially available in water-soluble forms. A Zn-contaminated agricultural soil (clay-loam) from a site adjacent to the University of California at Berkeley experimental growth facility was selected for the soil experiment and analyzed for selected physicochemical properties (Table 1). The soil was not dried after collection, and its pH was

determined in a 1:2.5 slurry of soil:DDW. The concentrations of water-extractable Zn and Fe in the damp soil were determined by extraction with DDW (WE-Zn and WE-Fe): five replicate subsamples (10 g) were shaken in 25 mL of DDW for 2 h and then filtered. The mass of soil on the filter paper was determined after drying at 80 °C. The Zn and Fe concentrations in the filtrate were determined by inductively coupled plasma atomic emission spectrometry (ICP-AES). The concentrations of total Zn, Fe, and S in five, air-dried 3-g subsamples of this soil were also determined by ICP-AES after digestion in 25 mL of a 4:1 mixture of concentrated HCl and HNO3 (24). The total microbial counts in the soil were estimated after vortexing five 0.5-g replicates of the fresh soil in 1 mL of sterile 0.85% saline in sterile plastic tubes; serial dilutions of the soil suspension using 0.85% sterile saline were spread on TSA plates and counted after 7 days incubation at 25 °C. The CFU/g of fresh soil is presented in Table 1. Fifty grams of fresh (undried) soil was placed in each of 48 Magenta boxes (Sigma) and watered with 10 mL of DDW. Twenty-four of the Magenta boxes were then autoclaved for 45 min to kill the native microbial populations in the soil (sterile soil). The soils in the other 24 Magenta boxes were not autoclaved to maintain their native populations of microbes (nonsterile soil). Twelve seeds of T. caerulescens with bacteria (plus bact) were planted into each of four replicate sterile soil Magenta boxes and four replicate nonsterile soil Magenta boxes. This was repeated for the remainder of the Magenta boxes with axenic seeds of T. caerulescens (no bact) and seeds of T. caerulescens with killed bacteria (dead bact). Similarly, four replicate Magenta boxes of each treatment were planted with seeds of T. arvense (plus bact, no bact, and dead bact). The Magenta boxes were closed and placed in a controlled-environment plant growth chamber with constant light at 25 °C, and sterile DDW was added aseptically as required. After a 35-day growth period, the plant roots and shoots were harvested from the soil in a laminar-flow cabinet. Samples of soil were collected aseptically from each of the boxes. The number of viable bacteria in 0.25-g subsamples of these damp soils after the growth of the plants was determined as described above. The pH and concentrations of water-soluble Zn in each of the soils were determined as described above after shaking duplicate 4-g soil samples with 10 mL of DDW for 2 h. The plants were washed in running DDW until free of adhering soil, and the whole plants were soaked for 20 min in 1 mM Ca(NO3)2‚4H2O at 4 °C to remove adsorbed Zn. The number of bacteria per gram of root was determined in subsamples of the roots by serial dilution and TSA plate counts after thoroughly grinding the roots in 0.25 mL of sterile 0.85% saline using autoclaved micropestles. The plants were separated into roots and shoots, dried at 80 °C, and weighed. The concentrations of Zn in plant tissues were determined by ICP-AES following digestion in 2 mL of concentrated HNO3 at 85 °C for 4 h. Effect of Bacteria on Plants Grown in Agar Containing Zn. This experiment tested the effect of bacteria on the Zn accumulation of T. caerulescens and T. arvense growing in media where the Zn was soluble and hence predominantly available for plant uptake. Two sterile growth media, one with a low concentration of Zn and the other with a high Zn concentration, were prepared using 0.4% Phytagar (Gibco) and one-quarter strength Hoagland’s nutrient solution (25). The low Zn treatment had 0.1 µmol of Zn L-1, and the high Zn treatment had 76.5 µmol of Zn L-1 (5 mg of Zn L-1) added to the one-quarter strength Hoagland’s nutrient solution as ZnSO4‚7H2O. Fifty milliliters of each kind of agar was pipetted aseptically into each of two sets of 24 autoclaved Magenta boxes and allowed to set. Twelve seeds of T. caerulescens (plus bact; see above) were added to each of four replicate VOL. 35, NO. 15, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3145

Magenta boxes of the low Zn agar and also to four replicate boxes of the high Zn agar in a laminar-flow cabinet. This procedure was repeated with the other sets of seeds for T. caerulescens (no bact) and (dead bact) and for T. arvense (plus bact, no bact, and dead bact). The boxes were closed, placed in a controlled-environment growth room, and harvested after 35 days. At harvest, the plants were removed from the boxes, counted, rinsed thoroughly in DDW to remove adhering agar, and desorbed for 20 min in 1 mM Ca(NO3)2‚4H2O at 4 °C. The roots and shoots were separated, dried at 80 °C, and weighed. The concentrations of metals in the shoots were determined by ICP-AES following digestion in HNO3 as described above. Soil Extractions Using the Products of Bacterial Growth. This experiment tested the ability of the products of bacterial growth to extract Zn from the soil. Two different media were used to culture the bacterial strains: 1.5% trypticase soy broth (TSB; Becton Dickinson) and a rhizosphere medium (RSM) designed by Buyer et al. (26) to resemble a rhizosphere habitat, except that we omitted the pH buffer, ACES. Four sterile 500-mL Erlenmeyer flasks were prepared, two with 250 mL of TSB and two with 250 mL of RSM. The three bacterial strains were added to one flask of each medium as a 0.5-mL inoculum of the bacterial saline suspension described previously. The second flask of each medium was not inoculated as a sterile (axenic) control. The bacterial cultures and the axenic control flasks were placed on an orbital shaker at 100 rpm for 24 h at 25 °C. The medium in each flask was centrifuged at 8000g for 15 min; the supernatant was decanted and vacuum filtered through sterile filters (0.22 µm pore size). The pH of the filtrate was recorded. The ability of the filtrate to extract Zn from the soil was determined by shaking five replicate 2-g samples of the soil with 10 mL of each of the bacterial filtrates or axenic filtrates for 2 h. The soil suspensions were centrifuged at 4000g for 15 min and filtered, the concentration of metals in the filtrate was determined by ICP-AES, and the mass of soil was determined after drying at 80 °C. Effects of Bacteria on Root Hair Production. This experiment examined the effects of the bacteria on root hair growth. A technique for morphological observations of root hairs growing on the surface of vertical agar plates was modified from that described by Dubrovsky et al. (22). This technique allows observations to be made without any mechanical interference and without risk of contamination, which allowed observations to be made on more than one occasion. Petri dishes were prepared with 50 mL of sterile 1.5% Phytagar agar made with one-quarter strength Hoagland’s solution. Three seeds of T. caerulescens plus bact (see above) were placed aseptically in a row toward one edge of the Petri dish. This was repeated for T. caerulescens no bact, T. arvense plus bact, and T. arvense no bact. Five replicate Petri dishes were prepared for each treatment. The Petri dishes were placed vertically in a rack so that the roots would grow on the surface of the agar. At regular intervals, the roots of both species were examined through the Petri dish lid. Twelve days after planting, when the tap root of each species was well-developed, the length and morphology of the root hairs 1 cm behind the root apex were examined. Statistical Analyses. Analysis of variance procedures and Tukey multiple comparison tests were performed to test for statistically significant differences between means.

Results and Discussion Effect of Bacteria on Growth and Zn Accumulation of Plants Grown in Sterile Soil. The Zn-hyperaccumulator, Thlaspi caerulescens, and the nonaccumulator, Thlaspi arvense, were grown on an autoclaved soil containing 179 mg of Zn kg-1 DW, of which only 0.06 mg of Zn kg-1 (0.03%) was water 3146

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 15, 2001

FIGURE 1. In autoclaved soil, the inoculation of bacteria (+ bact) onto the seeds of the Zn hyperaccumulator T. caerulescens (black bars) resulted in a higher concentration of Zn in the shoots (A), a higher shoot biomass (B), and a greater mass of Zn accumulated (C) than the two axenic treatments (no bact and dead bact). These effects were not seen for the nonaccumulator T. arvense (gray bars). Data are shown as means plus standard errors, n ) 4. Letter codes are shown for significant ANOVA; bars showing the same letter code are not significantly different (Tukey multiple comparisons test, P > 0.05). soluble (Table 1). Here we show that bacteria facilitated increases in the shoot Zn concentration and the biomass of T. caerulescens growing in sterile soil; there were no bacterial effects on the nonaccumulator, T. arvense. When bacteria were supplied to axenic seeds of T. caerulescens (plus bact), the Zn concentrations in the shoot of T. caerulescens reached 1000 mg kg-1, 2-fold higher than the Zn concentrations in shoots of T. caerulescens plants grown in axenic control treatments (no bact and dead bact) (Figure 1A). The bacteria also caused a 1.7-fold increase in the shoot mass of T. caerulescens as compared to the controls (Figure 1B), but this was not significant (P > 0.05). Thus, the shoots accumulated 4 times as much Zn as the axenically grown plants (4 µg as compared to 1 µg), which was highly significant (Figure 1C; P < 0.05). With T. arvense, there was no significant effect of bacterial inoculation on Zn accumulation in the shoots, which had a Zn concentration of 100 mg kg-1 with all treatments (Figure 1A). The addition of bacteria to the seeds of T. arvense also had no effects on the biomass and total Zn accumulated by the T. arvense shoots (Figure 1B,C). Effect of Bacteria on Root Biomass and Root Hair Production. The masses of the roots of the two species of Thlaspi had the same patterns as their shoots. In the presence of bacteria the mass of T. caerulescens roots was higher than in the axenic controls but not significantly so (P > 0.05; data not shown); there was no significant effects of the bacteria on the mass of roots of T. arvense. The addition of bacteria to the seeds also had positive effects on the concentration of Zn in the roots of both Thlaspi species as compared to the

TABLE 2. Analyses of Soil at the End of Sterile Soil Experimenta species

T. caerulescens bacterial counts WE-Zn (µg kg-1) pH T. arvense bacterial counts WE-Zn (µg kg-1) pH

+ bact

no bact

dead bact

10.0 × 107 73.85a 6.56

0 50.68c 6.63

0 60.44b 6.91

7.5 × 107 70.82a 6.74

0 58.64ab 7.01

0 42.32b 6.73

a Bacterial counts given as CFU g-1 soil DW; zero counts indicates that the soil remained sterile. The concentrations of water extractableZn (WE-Zn; µg kg-1) and the pH of the bulk soil are also shown. Letter codes are shown for significant ANOVA; values showing the same letter are not significantly different (Tukey multiple comparisons test, P > 0.05). Values are means, n ) 2.

axenic controls, but none of these differences were statistically significant (P > 0.05; data not shown). It is not surprising that the significant effects of bacteria on Zn concentrations were found in the shoots (but not the roots) because T. caerulescens is well-known to rapidly translocate most of the Zn it takes up to its shoots (6, 27). The 2-fold increase in Zn concentration in the shoots of T. caerulescens plants supplied with bacteria as compared to axenic plants (P < 0.05) confirms that there was an increase in the uptake of Zn that was independent of positive effects of bacteria on the plant biomass. There was also no evidence that the bacterially mediated increase in Zn uptake in T. caerulescens was due to an increase in root surface area. In the sterile soil experiment, the root/shoot biomass ratio was calculated to determine if there was a disproportionate increase in the mass of the roots as compared to the shoots. No such increases were found; the root/shoot biomass ratio of T. caerulescens with bacteria was 0.076 as compared to 0.074 for axenic plants (data not shown; P > 0.05). For T. arvense, the root:shoot biomass ratio with bacteria was 0.165 as compared to 0.179 for axenic plants, which were not significantly different (P > 0.05). In a separate experiment, the root hairs were examined on roots of both species growing down the surface of vertical agar plates. The addition of bacteria to the seeds had no effects on root hair production of T. caerulescens or T. arvense. After 12 days growth, the taproots of axenic and plus bact T. caerulescens were approximately 2.5 cm long. The root hairs 1 cm behind the root apexes in both treatments were approximately 1.9 mm, indicating that the bacteria had no effect on root hair length. For the nonaccumulator, T. arvense, the taproot was 4 cm long; the root hairs with bacteria were slightly shorter than the axenic plants (1 mm as compared to 1.2 mm). No visible differences in root hair density or morphology were observed between the plus and minus bacterial treatments of either species. Effect of Bacteria on Mobility of Zn in Sterile Soil. The addition of bacteria to the seeds of both plant species facilitated increases in the solubility of Zn in the soil during the growth period. When bacteria were present in the rhizosphere of either plant species, the soils had greater concentrations of water-extractable Zn as compared to the axenic soils (22 to 67%, P < 0.05; Table 2). Moreover, the concentration of water-extractable Zn in the presence of bacteria (Table 2; >71 µg kg-1) was higher than that at the beginning of the experiment (Table 1; 57.46 µg kg-1). Because of its superior plant-based abilities to absorb soluble Zn from the growth medium (6, 9), T. caerulescens was able to benefit from the increased Zn solubility, resulting in the higher Zn accumulation than the axenic plants. T. arvense does not accumulate Zn and, hence, did not accumulate more Zn

FIGURE 2. Filtered media that had supported bacterial growth for 24 h extracted higher concentrations of Zn from soil than the sterile media. Data are shown for the two media used, trypticase soy broth (TSB) and simulated rhizosphere medium (RSM) plus standard errors, n ) 5. The pH values of the media are also shown as numbers above the bars. The concentration of Zn extracted by sterile DDW is given for comparison. despite the increased Zn solubility in the soil. The soils from both the T. caerulescens and T. arvense treatments where bacteria were added to their seeds (plus bact) contained bacterial populations of the same order of magnitude (Table 2; 107). In contrast, there were no bacteria present in the other treatments (no bact and dead bact), confirming that these treatments had remained axenic throughout the experiment. Mobilization of Zn in Sterile Soil with Products of Bacterial Growth. Further evidence to support the conclusion that the bacteria facilitated an increase in the solubility of nonlabile Zn in the soil was obtained by measuring the concentration of Zn extracted from soil by bacteria cultured in two liquid media, TSB and RSM. The concentrations of Zn solubilized from the soil by both filtered media (RSM and TSB) after 24 h of bacterial growth were higher than those extracted by the sterile growth media (Figure 2), indicating that the products of bacterial growth could mobilize Zn in the soil. The TSB medium that had supported bacterial growth mobilized 34% more Zn than the sterile TSB (P < 0.05). For the RSM that had supported bacterial growth, 77% more Zn was mobilized (P < 0.05). Both media mobilized many times more Zn than water alone (Figure 2); this might be expected as both TSB and RSM contained amino acids from the hydrolyzed soy and milk proteins, respectively. Extracting the soil with the filtrates of the media that had supported bacteria also mobilized significantly greater quantities of Fe but not S (data not shown). In the case of RSM, the plus bacteria medium mobilized 6 times more Fe than the sterile medium (P < 0.05). For TSB, the plus bacteria medium mobilized 1.5 times more Fe than the sterile medium. These results indicate that the activity of the bacteria in the soil would very likely have had a significant effect on increasing the mobility of metals in the rhizosphere of the plants in the sterile soil experiment. Effect of Bacteria on Growth and Zn Accumulation of Plants Grown in Agar. We investigated two alternate mechanisms by which the bacteria might have increased the concentration of Zn in the shoots of T. caerulescens in the sterile soil experiment but found no evidence to support these. First, we tested whether the effect of bacteria on Zn uptake was via increasing the functional surface area of the root. This hypothesis was not supported by the measurements of root hair length and the root/shoot biomass ratio, which were not significantly different between bacterial and axenic treatments for either Thlaspi species (see above). Second, the presence of bacteria did not facilitate an increase in the VOL. 35, NO. 15, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3147

FIGURE 3. In sterile agar containing 0.1 µM Zn L-1, the inoculation of bacteria (+ bact) onto the seeds of T. caerulescens (black bars) or T. arvense (gray bars) did not have any effects on the concentration of Zn in the shoots (A), shoot biomass (B), or mass of Zn accumulated (C) as compared to the two axenic treatments (no bact and dead bact). Data are shown as means plus standard errors, n ) 4. Letter codes are shown for significant ANOVA; bars showing the same letter code are not significantly different (Tukey multiple comparisons test, P > 0.05). rate of soluble Zn transport from the rhizosphere into the root. This conclusion was made from an experiment using sterile agar media (instead of sterile soil) to determine the effects of bacteria on Thlaspi when the availability of watersoluble Zn was not limited. When grown in agar media containing 0.1 or 76.5 µmol of Zn L-1, the positive effects of bacteria on the growth and Zn accumulation of T. caerulescens were not reproduced. The concentrations of Zn in the shoots and the biomass of T. caerulescens were the same in all three of the bacterial treatments (Figures 3A and 4A; P > 0.05). Similarly, the addition of bacteria to the seeds of T. arvense had no significant effects on the growth or Zn accumulation in the shoot (Figures 3 and 4). These results demonstrate that the bacteria did not influence the rate of transport of Zn into the hyperaccumulator plant, since more water-soluble Zn was available in this medium. This suggests that the positive effects of bacteria on T. caerulescens seen in the sterile soil experiment were due to the acquisition of nonlabile Zn from the soil and not due to an increase in the rate of Zn transport into the plant. The concentration of Zn in the shoots of the hyperaccumulator, T. caerulescens, grown in the high-Zn agar (Figure 4A; 1500 mg kg-1) was significantly greater than those for the low-Zn agar (Figure 3A; ∼200 mg kg-1). However, there were no differences in the shoot biomass of T. caerulescens, which was 3 mg in both the low-Zn and high-Zn agars (Figures 3B and 4B); this indicates that T. caerulescens was not Zn deficient in the low-Zn agar, nor was it suffering from Zn toxicity in the high-Zn agar. By contrast, the nonaccumulator, T. arvense, 3148

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 15, 2001

FIGURE 4. In sterile agar containing 76.5 µM Zn L-1, the inoculation of bacteria (+ bact) onto the seeds of T. caerulescens (black bars) or T. arvense (gray bars) did not have any effects on the concentration of Zn in the shoots (A), shoot biomass (B), or mass of Zn accumulated (C) as compared to the two axenic treatments (no bact and dead bact). Data are shown as means plus standard errors, n ) 4. Letter codes are shown for significant ANOVA; bars showing the same letter code are not significantly different (Tukey multiple comparisons test, P > 0.05). was not tolerant to the 76.5 µmol of Zn L-1 in the high-Zn agar. The leaves of T. arvense were extremely chlorotic, and the shoot biomass (Figure 4B; 2 mg) was approximately 40% lower than the plants in the low-Zn agar (Figure 3B). Remarkably, the concentration of Zn in the shoots of the nonaccumulator reached 2000 mg of Zn kg-1 (Figure 4A), which was higher than that in the hyperaccumulator. Since the T. arvense plants were very unhealthy, this indicates that the plants were no longer able to control the uptake and translocation of Zn from this high-Zn medium. Lasat et al. (6) found similar extremely high concentrations of Zn in the shoots of T. arvense growing in hydroponic solutions with >50 µM Zn. Possible Mechanisms for Increased Availability of WaterSoluble Zn in Soil. Clearly, bacteria in the soil are important for increasing the availability of water-soluble Zn for hyperaccumulation from soils with a high proportion of nonlabile Zn. This study indicates that the bacteria facilitated the release of Zn from the nonlabile phase in the soil, thus enhancing the availability of water-soluble Zn to T. caerulescens. The increase in the solubility of Zn in the soil was not due to changes in pH. Evidence for this came from two experiments: (i) In the sterile soil experiment there were no significant differences in the pH of the soil between bacteriatreated and axenic soils at harvest (Table 2). Furthermore, the soil pH at harvest was not significantly lower than that at the beginning of the experiment (Table 1). These findings agree with other studies of T. caerulescens that concluded that the rhizosphere pH of T. caerulescens did not decrease significantly during plant growth (11, 12, 28, 29). (ii) In the

experiment where the bacteria were grown in vitro in the TSB and RSM media, the increased solubilization of Zn and Fe from the soil by the bacterial extracts was also not caused by lower pH; the presence of bacteria in RSM actually raised its pH by 0.8 unit, and yet this medium extracted 2-fold more Zn from the soil than the sterile RSM (Figure 2). The increase in the solubility of Zn in the soil may therefore have been due to compounds such as Zn-chelating metallophores produced by the bacteria. Metallophores are commonly produced by strains of Pseudomonas and Enterobacter, which were used here (16). The production of metallophores may account for the fact that both bacterial media increased the concentrations of Zn and Fe extracted from the soil but not S. Indeed, the RSM medium that had supported bacterial growth extracted 2-fold more Zn (Figure 2) and 6-fold more Fe than the sterile RSM (data not shown). These increases in the water-soluble Zn concentration with bacteria in both the in vivo (sterile soil) and the in vitro experiments support our conclusion that bacteria enhanced Zn accumulation in T. caerulescens shoots by increasing dissolution of Zn from the nonlabile phase. Effect of Bacteria on Growth and Zn Accumulation in Plants Grown in Nonsterile Soil. There were no significant effects of bacterial inoculation on the Zn concentration, biomass, or Zn accumulation of T. caerulescens or T. arvense grown in soil that had not been autoclaved (nonsterile) (data not shown; all P > 0.05). In all three treatments (plus bact, no bact, and dead bact), the shoots of the hyperaccumulator, T. caerulescens, had Zn concentrations of approximately 1500 mg kg-1, which was much higher than T. arvense, which had only 150 mg kg-1 (data not shown). If the results from the sterile soil and nonsterile soil experiments are compared, the native populations of microbes in the nonsterile soil significantly increased Zn accumulation by T. caerulescens (with or without added bacteria) as compared to T. caerulescens grown axenically in sterile soils. In the nonsterile soil experiment, the concentration of Zn in the shoots of all three treatments of T. caerulescens (1500 mg kg-1; data not shown) was 3-fold higher than that in the axenic treatments in the sterile soil (Figure 1; 500 mg kg-1) and similar to that in the bacteria-treated T. caerulescens plants grown in sterile soil (Figure 1; 1100 mg kg-1). Since the addition of bacteria to the seeds did not have any significant effects on the Zn concentration in the shoots of T. caerulescens in the nonsterile soil experiment (data not shown), the native microbial populations in that soil (7.4 × 106 CFU g-1 soil DW) must therefore have increased the availability of Zn to T. caerulescens, masking the effects of the three species of bacteria we added to its seeds. This suggests that the positive effects on T. caerulescens in the sterile soil (Figure 1) were not specific to the strains of bacteria we added in that experiment; the native populations of microbes in soil increased growth and Zn accumulation by T. caerulescens. Microbes are ubiquitous in soil. Hence, the nonspecific microbially mediated solubilization of Zn will contribute to the ability of T. caerulescens to acquire Zn from field soils or from nonsterile soils used in pot studies, which contain low concentrations of labile metals. Such microbial mobilization may therefore account for the solubilization of Zn described in other studies of the rhizosphere of T. caerulescens (e.g., refs 11, 12, and 29). Furthermore, the contribution of microbes to Zn accumulation by T. caerulescens may be further enhanced by the larger numbers of microbes on its roots as compared to T. arvense. At the end of the sterile soil experiment presented here, the roots of T. caerulescens supported 6.1 × 108 bacteria/g of root, which was significantly higher (P < 0.05) than the 3.6 × 108 seen for T. arvense; this is probably a result of the higher specific root length of T. caerulescens as compared to T. arvense (10).

To conclude, this study provides a new insight into the mechanisms by which hyperaccumulator plants access nonlabile metals in soils. The addition of bacteria to the seeds of the Zn-hyperaccumulator, T. caerulescens, greatly enhanced its Zn uptake when growing in a soil where a majority of the Zn was not initially available in water-soluble forms. The results demonstrated that the increased Zn uptake by T. caerulescens with bacteria was not a result of increased root growth, nor was it a function of increased root surface area (root hairs). Thus, the bacteria enhanced the availability of water-soluble Zn in the soil, which overcame a major ratelimiting step for Zn acquisition by T. caerulescens growing on soils with low concentrations of labile Zn. The constitutively high Zn accumulation in the nonsterile soils indicates that the effects were not specific to the three strains of bacteria added. Many microbial strains may therefore contribute to Zn acquisition by T. caerulescens and other hyperaccumulators growing on soils with low concentrations of labile Zn. This microbial phenomenon could potentially be exploited (e.g., by bioaugmentation of soil) for the phytoremediation and phytomining of metals from soils, thus decontaminating metal-polluted soils that pose worldwide environmental health hazards.

Acknowledgments We thank Alan Baker for kindly donating the seed of T. caerulescens. We also thank Marina Ma for assistance in setting up the experiments and Peter Neumann for his helpful comments. This work was funded by Electric Power Research Institute Grants W08021-30 and W04163 and by the Agriculture Experiment Station.

Literature Cited (1) Brooks, R. R.; Lee, J.; Reeves, R. D.; Jaffre T. J. Geochem. Explor. 1977, 7, 49. (2) Baker, A. J. M.; McGrath, S. P.; Reeves, R, D.; Smith, J. A. C. In Phytoremediation of Contaminated Soil and Water; Terry, N., Ban ˜ uelos, G. S., Eds.; Lewis Publishers: Boca Raton, FL, 2000; pp 85-107. (3) Baker, A. J. M.; McGrath, S. P.; Sidoli, C. M. D.; Reeves, R. D. Resour. Conserv. Recycl. 1994, 11, 41. (4) McGrath, S. P.; Dunham, S. J.; Correl, R. L. In Phytoremediation of Contaminated Soil and Water; Terry, N., Ban ˜ uelos, G. S., Eds.; Lewis Publishers: Boca Raton, FL, 2000; pp 109-128. (5) Salt, D. E.; Smith, R. D.; Raskin, I. Annu. Rev. Plant. Physiol. Plant Mol. Biol. 2000, 49, 643. (6) Lasat, M. M.; Baker, A. J. M.; Kochian, L. V. Plant Physiol. 1998, 118, 875. (7) Salt, D. E.; Prince, R. C.; Baker, A. J. M.; Raskin, I.; Pickering, I. J. Environ. Sci. Technol. 1999, 33, 713. (8) Ku ¨ pper, H.; Zhao, F. J.; McGrath, S. P. Plant Physiol. 1999, 119, 305. (9) Pence, N. S.; Larsen, P. B.; Ebbs, S. D.; Letham, D. L. D.; Lasat, M. M.; Gravin, D. F.; Eide, D.; Kochian, L. V. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 4956. (10) Whiting, S. N.; Leake, J. R.; McGrath, S. P.; Baker, A. J. M. New Phytol. 2000, 145, 199. (11) Knight, B.; Zhao, F. J.; McGrath, S. P.; Shen, Z. G. Plant Soil. 1997, 197, 71. (12) McGrath, S. P.; Shen, Z. G.; Zhao, F. J. Plant Soil. 1997, 188, 153. (13) Schlegel, H. G.; Cosson, J. P.; Baker, A. J. M. Botan. Acta 1991, 104, 18. (14) Ghaderian, Y. S. M.; Lyon, A. J. E.; Baker, A. J. M. New Phytol. 2000, 146, 219. (15) Kloepper, J. W.; Leong, J.; Teintze, M.; Schroth, M. N. Nature 1980, 286, 885. (16) Neilands, J. B.; Leong, S. A. Annu. Rev. Biochem. 1986, 37, 187. (17) Kapulnik, Y. In Plant Roots: The Hidden Half; Waisel, Y., Eshel, A., Kafkazi, U., Eds.; Marcel Dekker: New York, 1996; pp 769781. (18) de Souza, M. P.; Chu, D.; Zhao, M.; Zayed, A. M.; Ruzin, S. E.; Schichnes, D.; Terry, N. Plant Physiol. 1999, 119, 565. VOL. 35, NO. 15, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3149

(19) de Souza, M. P.; Huang, C. H. A.; Chee, N.; Terry, N. Planta 1999, 209, 259. (20) Salt, D. E.; Benhmanou, N.; Leszczynieka, M.; Raskin, I.; Chet, I. Int. J. Phytorem. 1999, 1, 67. (21) Haahtela, K.; Ro¨nkko¨, R.; Laakso, T.; Williams, P. H.; Korhonen, T. K. Mol. Plant-Microbe Interact. 1990, 3, 358. (22) Dubrovsky, J. G.; Esther Puente, M.; Bashan, Y. Soil. Biol. Biochem. 1994, 26, 1657. (23) Bertrand, H.; Plassard, C.; Pinochet, X.; Touraine, B.; Normand, P.; Cleyet-Marel, J. C. Can. J. Microbiol. 2000, 46, 229. (24) Allen, S. E. Chemcal Analysis of Ecological Materials, 2nd ed.; Blackwell Scientific: Oxford, 1989. (25) Hoagland, D. R.; Arnon, D. Calif. Agric. Exp. Stn., Berkeley, Circ. 1938, 347.

3150

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 15, 2001

(26) Buyer, J. S.; Sikora, L. J.; Chaney, R. L. Biol. Fert. Soil 1989, 8, 97. (27) Shen, Z. G.; Zhao, F. J.; McGrath SP. Plant Cell Environ. 1997, 20, 898. (28) Hutchinson, J. J.; Young, S. D.; McGrath, S. P.; West, H. M.; Black, C. R.; Baker, A. J. M. New Phytol. 2000, 146, 453. (29) Luo, Y. M.; Christie, P.; Baker, A. J. M. Chemosphere 2000, 41, 161.

Received for review December 5, 2000. Revised manuscript received May 3, 2001. Accepted May 8, 2001. ES001938V