Environ. Sci. Technol. 2008, 42, 9437–9442
Is Perchlorate Metabolized or Re-Translocated within Lettuce Leaves? A Stable-Isotope Approach ANGELIA L. SEYFFERTH,† NEIL C. STURCHIO,‡ AND D A V I D R . P A R K E R * ,§ Department of Environmental Sciences, University of California, Riverside, California 92521, and Department of Earth and Environmental Sciences, University of Illinois, Chicago, Illinois 60607
Received July 18, 2008. Revised manuscript received September 10, 2008. Accepted October 2, 2008.
Perchlorate is an environmental contaminant that is found in drinking water and a variety of foodstuffs, but many questions regarding its uptake, transport, and persistence in higher plants remain unanswered. In a series of hydroponic experiments, a stable-isotope tracer of perchlorate (95% 37ClO4-) was utilized to determine the extent of in vivo metabolism and of phloem re-translocation of perchlorate at low (i.e., nmol/kg of fresh weight (FW)) concentrations in lettuce. Chlorate and chlorite metabolites were not detected in lettuce leaves at detection limits of 13.1 and 291 nmol/kg FW, respectively, and chloride isotopic signatures were not substantially different from natural chloride. Perchlorate exhibited no significant movement from older leaves into new leaves, nor to roots. Stable isotopes proved useful in assessing perchlorate metabolism and re-translocation within lettuce at nmol/kg levels. The absence of any metabolism or re-translocation indicates that perchlorate is relatively persistent within leafy produce, and that the primary mode of transport of perchlorate is through the xylem of higher plants.
Introduction Perchlorate is an inorganic contaminant found at generally low concentrations in surface and groundwater across the United States (1, 2), but the health implications of low-level contamination is a topic of ongoing debate (3, 4). Perchlorate may interfere with iodide uptake at the sodium-iodide symporter of the human thyroid, and may thus lead to a lower production of key thyroid hormones, especially in sensitive groups such as developing children, individuals with pre-existing thyroid disorders, and individuals with less than sufficient iodide intake (5, 6). Although much of the concern over perchlorate ingestion has considered drinking-water contamination, perchlorate is also found at concentrations of concern in fresh produce (7-11) dairy milk, and breast milk (6, 12, 13), indicating that food-chain transfer is likely. Recent evidence suggests that the importance of human exposure through contaminated produce may be equal to or * Corresponding author phone: (951) 827-5126; fax: (951) 8273993; e-mail:
[email protected]. † Current address: Department of Environmental Earth Systems Science, Stanford University, Stanford, California 94305. ‡ University of Illinois. § University of California. 10.1021/es802006e CCC: $40.75
Published on Web 11/08/2008
2008 American Chemical Society
greater than that from drinking water and should be considered in risk assessment calculations (7, 14). Previous work in our laboratory demonstrated that different varieties of winter lettuce (i.e., green leaf, butter head, crisphead) accumulate varying concentrations of perchlorate in their edible tissue (15). However, the reason(s) for these findings is not yet clear, and there are at least three possible explanations. First, the root morphologies of the lettuce varieties may vary and lead to different rates of uptake due to differences in root surface area, or to differences in the density of ion transporters that have affinity for perchlorate. Second, each variety may actually accumulate the same amount of perchlorate, but the varieties differ in their abilities to metabolize perchlorate through chlorate and chlorite intermediates to chloride. Third, the different leaf concentrations seen at harvest may be due to varying rates of phloem re-translocation to the roots and, perhaps, back into the solution. Metabolism (phytodegradation) of perchlorate has been demonstrated in higher plants and appears to follow the stepwise reduction to chloride through chlorate and chlorite intermediates, as is the case with certain anaerobic bacteria (16-18): ClO4 f ClO3 f ClO2 f Cl
(1)
There is evidence that certain plants are able to metabolize perchlorate to chloride, but all of the studies were conducted at concentrations of at least 2 orders of magnitude higher than those found in typical irrigation water (e.g., Colorado River water) (16-20). Moreover, there are very limited data on the presence of chlorate and chlorite intermediates within plant tissues after perchlorate metabolism. Susarla et al. (20) measured chlorate, chlorite, and chloride within all 11 plant species that took up perchlorate, but there is no mention of how these metabolites were analyzed or of their leaf concentrations. Van Aken and Schnoor (18) have provided the most unequivocal evidence for perchlorate phytodegradation to date using radio-labeled perchlorate (36ClO4-) at concentrations of 25 mg/L in a four-week uptake experiment with hybrid poplar trees (Populus deltoides x nigra). Radiolabeled chlorate, chlorite, and chloride were detected in the poplar leaf extracts, but only labeled chloride was detected in the solution after 30 days (18). It remains inconclusive whether other plant species are capable of perchlorate metabolism, and whether metabolism takes place in produce that is exposed to nanomolar perchlorate concentrations in the irrigation water. To our knowledge, there have been no studies on phloem re-translocation of perchlorate in higher plants. Site assessments and market surveys have shown that perchlorate is found in greater concentrations in leaves than in fruits and stems (10, 21, 22). It may be that because transpiration is a key factor in perchlorate accumulation in plants (15), and since fruits transpire less water than leaves, fruits accumulate less perchlorate. However, it is also possible that this observation is merely a reflection of a greater extent of in vivo perchlorate metabolism in fruits than in leaves. Nzengung and McCutcheon (23) observed that old-leaf perchlorate concentrations steadily decreased while new-leaf concentrations increased, and the authors suggested that perchlorate may move from old leaves to new leaves in some plant species. However, it is not clear whether this was due to phloem re-translocation of perchlorate, or to metabolism of perchlorate in older leaves with concurrent uptake and translocation of perchlorate into new leaves. VOL. 42, NO. 24, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
9437
To date, there have been no studies on the in vivo metabolism of nanomolar concentrations of perchlorate, or of the extent of phloem re-translocation of perchlorate in fresh produce. In this study, we used nanomolar concentrations of 37Cl-enriched perchlorate (95% 37ClO4-) as a tracer to assess the extent, if any, of (1) in vivo metabolism of perchlorate and (2) phloem re-translocation of perchlorate in lettuce. To our knowledge, this is the first study to use stable isotopes to quantify both metabolism and phloem re-translocation of perchlorate in higher plants.
Experimental Section Reagents. All solutions were prepared using 18 MΩ · cm water or better, and all salts used for nutrient solutions were ACS certified 99% purity or higher. Perchlorate standards ranging from 2.5 to 200 nM were prepared by dilution of a liquid 10 mM ClO4- standard solution (SPEX CertiPrep, Metuchen, NJ). Natural (nonlabeled) perchlorate treatments were administered using the same standard stock solution. Labeled perchlorate treatments were administered using a stock solution of 461 mM ClO4- in which chloride atoms were labeled as 37Cl with 95% purity (Icon Isotopes, Summit, NJ). Metabolism Experiment. Green leaf, butter head, and crisphead lettuce seeds were germinated (day 0) and grown hydroponically in accordance with our previous work (15) with the following modifications. The presence of natural chloride was minimized by the substitution of nitrate or sulfate salts for chloride salts, and by the use of sulfuric acid for pH adjustments. Basal nutrients were supplied as (µM): NO3-, 4900; NH4+, 100; P, 80; K, 1080; Ca, 1900; Mg, 500; S, 558; Fe, 20; Mn, 0.6; Zn, 8; Cu, 2; B, 10; Mo, 0.1; Ni, 0.1; Si (as SiO2 gel), 187; Cl, 5.4. Germinated seedlings were transferred to a 22-L perchlorate-free nursery (day 10) and remained in the nursery for 10 days until seedlings were transferred to individual buckets (day 20) that contained either 100 or 500 nM of labeled perchlorate (95% 37ClO4-), with three replicates. Two control plants per plant type were also transplanted to individual buckets, which consisted of perchlorate-free nutrient solution; these plants were used to determine the ClOx MDLs (see below), and to compare natural chloride signatures in plant extracts with those of treatment plants. The nursery and individual buckets were continuously aerated using spaghetti tubing that was fitted to an aquarium pump; dissolved oxygen was maintained at approximately 5.8 mg/L throughout the experiments. To buffer the nutrient solution at pH 6, the solution contained 1.0 mM 2-(Nmorpholino)ethanesulfonic acid (MES) and 0.5 mM NaOH. Each day, pH, water level, and phosphate concentration were monitored and adjusted to appropriate levels (6.0, 4 L, and 80 µM, respectively). Every 2-5 days, exhausted buckets were subsampled and replaced with clean, acid-washed buckets of fresh nutrient solution. Solution changes became more frequent as plants grew larger. All plants were grown to approximately marketable size for a total of 42 days from seed to harvest, and the treatment plants were exposed to 95% 37Cl-labeled perchlorate for a total of 21 days. Upon harvesting, shoots were separated from roots and were placed into zip-locked bags and frozen for at least 2 days in a -22 °C freezer. The entire head of the frozen shoots was then processed according to the procedure described later in this section. Phloem Re-Translocation Experiment. Plants were grown in a similar fashion to those in the metabolism experiment with the following modifications. Here, there was no need to minimize natural chloride, and thus, nutrient solutions were prepared according to our previous work (15) with basal nutrients supplied exactly as above with the exception of the following (µM): S, 500; Cl, 191. Treatments were imposed at the nursery stage, prior to transferring to individual buckets, 9438
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 24, 2008
and only green leaf and butter head lettuces (not crisphead) were used to assess the extent of perchlorate re-translocation. After 10 days of germination, plants were transferred (day 10) to a 22-L nursery that contained nutrient solution with 100 nM of labeled perchlorate (as 95% 37ClO4-). At day 20, seven plants each of green leaf and butter head lettuce were transferred to individual 4-L solutions that also contained 100 nM of labeled perchlorate (as 95% 37ClO4-). At day 30, all plants were transferred to perchlorate-free solutions and remained there for two days to flush the xylem of any remaining labeled perchlorate. To ensure complete xylem flushing of labeled perchlorate, three plants of each lettuce type were harvested at day 32, separated into roots and shoots, and stored frozen in plastic zip-locked bags until roots were processed. Three of the remaining four plants of each type were transferred to 4-L nutrient solutions that contained 100 nM of natural (nonlabeled) perchlorate, and the remaining plant of each type was transferred to a perchlorate-free 4-L nutrient solution; plants remained so exposed for the duration of the experiment. Also at day 32, all of the leaves on the remaining plants were marked with different colored nail polish to distinguish between leaves that were mature (MAT), midsized (MID), or juvenile (JUV) at the time of switch from labeled to natural (nonlabeled) perchlorate (or from labeled to perchloratefree solution for one plant of each type). At day 34, each plant was transferred to larger, 8-L solutions and remained in the larger reservoirs for the duration of the experiment. Plants were thus grown for a total of 43 days, and three plants of each type were exposed to labeled perchlorate (95% 37ClO4-) for 20 days at the beginning of the experiment and then to natural (nonlabeled) perchlorate (24.22% 37ClO4-) for the last 11 days. One plant of each type was exposed to labeled perchlorate (95% 37ClO4-) for 20 days at the beginning of the experiment and then to a perchlorate-free solution for the last 11 days. Upon harvesting, roots were separated from shoots, and shoots were separated into four portions: leaves that were juvenile (JUV), midrange (MID), and mature (MAT) at the time of switch from labeled to natural (nonlabeled) perchlorate, as well as new leaves that emerged (NEW) since the switch from labeled to natural (nonlabeled) perchlorate (or from labeled to perchlorate-free solution for one plant of each type). Each portion was placed in a zip-locked bag and frozen until processed. Tissue Processing. Lettuce leaves and roots were processed separately using a recently developed extraction procedure for ClO4- in plant matrices and analyzed by ion chromatography-electrospray ionization-mass spectrometry (IC-ESI-MS) (24). Briefly, leaves were sealed in a plastic bag and placed in a -22 °C freezer for at least 2 days to help rupture cell membranes. Frozen leaves were weighed, diluted at least 3:1 (w/w) with water, and macerated in a blender. Extracts were then shaken for 4 h, centrifuged, and filtered, and the aqueous extracts were rendered water-clear using a one-step solid phase extraction (SPE) method to remove organic pigments. Total time for extraction and sample cleanup was 6 h. For the metabolism experiment, the entire lettuce head was processed according to the aforementioned paragraph, but an aliquot (ca. 750 mL) of each extract was further processed for stable-isotope analysis of chloride. These aliquots were not passed through SPE cartridges; instead, ca. 40 g of activated carbon powder (Mallinckrodt Baker, Phillipsburg, NJ) was added to the extracts to further remove organic species present in the matrix. Extracts were shaken for 2 h, centrifuged, and filtered rendering a water-clear extract. Spike additions of chloride into lettuce extracts resulted in over 90% recovery, indicating that minimal chloride is lost (or added) due to the addition of activated
carbon. Extracts were kept at 4 °C until the chloride was purified and precipitated as AgCl (see Supporting Information S2 for complete details). The δ37Cl values were obtained by isotope-ratio mass spectrometry (IRMS) from the AgCl precipitates following a procedure previously described (25). Analysis of Perchlorate, Chlorate, and Chlorite. Aqueous lettuce extracts were analyzed for chlorite, chlorate, and perchlorate using IC-ESI-MS (Dionex, Sunnyvale, CA) (see Supporting Information S3 for complete details). Flow from the IC was directed into a Thermo Finnigan Surveyor MSQ Plus single quadropole MS (Thermo Electron Corporation, Waltham, MA). In order to correct for ion suppression during perchlorate analysis, Cl18O4- was used as an internal standard at a concentration of 10 nM in all standards and samples, and was added just prior to analysis (24). The MS was equipped with an AXP-MS auxiliary pump, pumping 50/50 acetonitrile/water at a flow rate of 0.3 mL/min. Matrix diversion was used to divert IC flow to waste when chlorite, chlorate, and perchlorate were not eluting. For the metabolism experiment, negative ion monitoring of m/z 69((0.5), 85((0.5), and 101((0.5) was utilized to identify and quantify chlorite (37Cl16O2-), chlorate (37Cl16O3-), and perchlorate (37Cl16O4-), respectively, and m/z 107((0.5) was monitored to quantify the internal standard. Chromeleon Version 6.6 (Dionex, Sunnyvale, CA) was used to control the instrumentation and to quantify ClOx. Measured Isotope Ratios in the Re-Translocation Experiment. To attain measured isotope ratios (MIRs) of 37ClO4to 35ClO4- in the re-translocation experiment, mass chromatograms were generated using the Thermo Finnigan MSQ Plus single quadropole mass spectrometer described in the previous section, and area counts of m/z 101 were divided by area counts of m/z 99. This procedure was verified by comparing the MIR and the known isotopic ratios (IRs) of standards (see Results section and Supporting Information Figure S1), which were prepared by mixing various amounts of tracer and natural (nonlabeled) perchlorate. Method Detection Limit (MDL). A widely used method for calculating the detection limit of analytical methods as described in EPA Method 314.0 (26) was utilized here. In short, seven replicate injections of 5 nM perchlorate, 24 nM chlorate, and 297 nM chlorite in initially ClOx-free green leaf lettuce extract were analyzed by IC-ESI-MS, and an MDL was calculated based on the Student’s t-value at the 99% confidence interval. The MDL is given by: MDL ) t · s
(2)
where t ) 3.14 for six degrees of freedom and s is the standard deviation of the quantified perchlorate concentration in the seven replicate injections.
Results Metabolism Experiment. Using the AS 19 analytical column, MDLs of chlorite, chlorate, and perchlorate in initially ClOxfree lettuce leaves were 291, 13.1, and 4.0 nmol/kg FW, respectively (Table 1). While we were able to detect low levels of chlorite and chlorate in these spiked plant matrices, neither chlorate nor chlorite intermediates were found in the extracts of perchlorate-exposed lettuce, even with the relatively high treatment level of 500 nM labeled perchlorate. In contrast, perchlorate concentrations were always g1600 nmol/kg FW when lettuce was grown at this higher perchlorate level (Figure 1). When grown at 100 nM labeled perchlorate, all three genotypes contained from 265 to 1600 nmol/kg FW, consistent with prior results (15). If labeled perchlorate was significantly reduced within lettuce leaves to chloride, the δ37Cl values would be substantially different from natural chloride (Figure 2). However, the chloride isotopic signatures in lettuce extracts from both the 100 and 500 nM perchlorate treatment were not
TABLE 1. MDL Determination for Chlorite, Chlorate, and Perchlorate by 7 Replicate Injections of 297 nM ClO2-, 24 nM ClO3-, and 5 nM ClO4- in Initially ClOx-Free Green Leaf Lettuce Extract Using the AS 19 Analytical Column and IC-ESI-MS Detection nmol/L -
replicate number
ClO2
ClO3-
ClO4-
1 2 3 4 5 6 7
254 219 283 259 256 291 276
22.7 22.1 24.2 24.2 24.3 21.8 23.1
4.8 4.1 4.8 4.8 4.1 4.3 4.4
average SDa MDLb
262 23.8 74.8
23.2 1.1 3.4
4.5 0.3 1.0
equivalent MDL
291
nmol/kg FW a
Standard deviation.
b
13.1
4.0
MDL ) 3.14 × SD.
FIGURE 1. Metabolism experiment results. Average whole-head perchlorate concentrations in three types of lettuce grown with either 100 or 500 nM 37ClO4- using IC-ESI-MS (note that neither chlorate nor chlorite metabolites were detected). Error bars represent standard errors of the mean where n ) 3. substantially different from natural chloride (Figure 2). The δ37Cl values of treated lettuce leaves ranged from -1.27 to 0.49, -0.23 to 0.34, and -0.83 to 0.02 ((0.2) ‰ (relative to the seawater chloride isotopic reference) for butter head, crisphead, and green leaf lettuce, respectively. These values are similar to the δ37Cl values obtained for perchlorate-free control plants, which were 0.19, -0.09, and 0.04 ((0.2) ‰ for butter head, crisphead, and green leaf lettuce, respectively. The δ37Cl value of the 37Cl-enriched perchlorate tracer (95% 37Cl) is about +1980‰ thus, contributions from metabolism of this perchlorate would have caused significant increases in the δ37Cl values of chloride in the lettuce extracts relative to their initial values (Figure 2). However, observed changes in δ37Cl were limited to slight decreases, indicating no apparent contribution of leaf Cl from perchlorate metabolism. Thus, within the sensitivity limits of this method, perchlorate does not appear to be substantially metabolized within lettuce leaves when plants are grown at nanomolar perchlorate concentrations. VOL. 42, NO. 24, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
9439
Discussion
FIGURE 2. Metabolism experiment results. Expected δ37Cl values of chloride based on percentages of the highest total leaf ClO4- (2.46 µmol) being metabolized to Cl- in the presence of the minimal chloride needed for the isotopic measurement (141 µmol natural Cl-) (0). The measured δ37Cl values for green leaf, butter head, and crisphead lettuce extracts are similar to each other and are also represented (×). Re-Translocation Experiment. Root extracts of plants that were harvested after xylem flushing had minimal (
