Environ. Sci. Technol. 2008, 42, 289–293
Degradation of Low Molecular Weight Volatile Organic Compounds by Plants Genetically Modified with Mammalian Cytochrome P450 2E1 C. ANDREW JAMES,† GANG XIN,† SHARON L. DOTY,‡ AND S T U A R T E . S T R A N D * ,† Department of Civil and Environmental Engineering, and College of Forest Resources, University of Washington, Seattle, Washington
Received May 21, 2007. Revised manuscript received October 12, 2007. Accepted October 22, 2007.
Cytochrome P450 2E1 (CYP2E1) is a key enzyme in the mammalian metabolism of several low molecular weight volatile organic compounds (VOCs), such as trichloroethylene (TCE), vinyl chloride (VC), carbon tetrachloride (CT), benzene, chloroform, and bromodichloromethane (BDCM), which are all common environmental pollutants that pose risks to human health. We have developed a transgenic tobacco (Nicotiana tabacum cv. Xanthii) that expresses CYP2E1 with increased activity toward TCE and ethylene dibromide. In experiments with tobacco plant cuttings exposed to VOCs in small hydroponic vessels, the transgenic tobacco had greatly increased rates of removal of TCE, VC, CT, benzene, toluene, chloroform, and BDCM, compared to wild-type or vector control tobacco, but not of perchloroethylene or 1,1,1-trichloroethane.
Introduction Low molecular weight volatile organic compounds (VOCs) such as trichloroethylene (TCE), carbon tetrachloride (CT), benzene, and chloroform are common pollutants in the air, soil, and groundwater. Exposure can lead to adverse health effects; all are listed either as known or possible human carcinogens. Trichloroethylene is a widespread organic pollutant, having been used for years in industrial and commercial settings as a cleaning solvent. It is commonly found at United States Environmental Protection Agency Superfund sites and is one of the pollutants most frequently detected in groundwater (1, 2). Carbon tetrachloride was widely used as a dry cleaning solvent and precursor for the production of chlorofluorocarbons during the last century. Although use and production have been phased out through a series of regulations, CT is still a common groundwater pollutant (3). Benzene is associated with petroleum contamination and is found in soil and groundwater at numerous leaking underground storage tank sites. Benzene is a leading gaseous pollutant of indoor air due to fuel storage in attached garages and environmental tobacco smoke (4). In addition to being a frequent contaminant of groundwater, chloroform is a disinfection byproduct formed during the chlorine * Corresponding author phone: 206-543-5350; e-mail: sstrand@ u.washington.edu; address: College of Forest Resources, P.O. Box 352100, University of Washington, Seattle, WA 98195. † Department of Civil and Environmental Engineering. ‡ College of Forest Resources. 10.1021/es071197z CCC: $40.75
Published on Web 12/04/2007
2008 American Chemical Society
treatment of drinking water and wastewater. Chloroform can be volatilized into indoor air during household water use, especially in showers or other hot water applications. Fiss et al. (5) found that use of household products containing triclosan, a common antibacterial agent, could result in the production of sufficient chloroform to present a significant exposure risk through the air. There are many traditional treatment technologies, such as pump-and-treat and excavation, available to address sites contaminated with VOCs, but they may be cumbersome, expensive, and time-consuming to apply. Phytoremediation is an innovative treatment that uses plants and plant-related processes (e.g., rhizosphere activity, transpiration) to remediate contaminated sites; it is potentially less intrusive, cheaper, and more widely accepted. However, there are limiting factors such as extensive land requirements for planting and a limited growing season, which may restrict its employment. To help optimize phytoremediation applications, plants have been genetically modified to increase metabolism of specific contaminants (6, 7). Transgenic tobacco plants expressing the nitroreductase nsfI from Enterobacter cloacae showed a marked improvement in the ability to tolerate and transform TNT (8). An Arabidopsis thaliana transformed to express the bacterial gene xplA, which encodes for a cyclotrimethylenetrinitramine (RDX)degrading cytochrome P450, has been shown to degrade RDX whereas vector control plants only sequester small amounts of RDX (9). Transgenic rice plants expressing human cytochrome P450 isoforms 1A1, 2B6, and 2C19 were found to improve the phytoremediation potential of the herbicides atrazine and metolachlor in soils (10). Doty et al. (11) have shown that tobacco transformed to express mammalian cytochrome P450 2E1 (CYP2E1) rapidly metabolizes TCE and ethylene dibromide. These plants are the subject of this study. Cytochrome P450 2E1 is an enzyme that is significant in the degradation of xenobiotics in mammalian systems, including TCE, CT, 1,1,1-trichloroethane (TCA), chloroform, and benzene (12). The goal of this work was to evaluate the activity of transgenic tobacco against various organic pollutants. Sterile apical cuttings from either modified or unmodified tobacco plants were placed in sealed vials with liquid growth media and dosed with VOCs. The headspace concentration was measured over time and the normalized loss rates were calculated and compared.
Materials and Methods Plant Material. Transgenic, vector control, and wild type tobacco plants (Nicotiana tabacum cv. Xanthii) were used in this experiment. Plant transformations were as described in Doty (11). Briefly, plasmid pSLD3 containing cDNA for human CYP2E1 under control of the Mac promoter was introduced into tobacco leaf discs via Agrobacterium mediated transformations. Transformations were verified by polymerase chain reaction and the plants regenerated. The CYP2E1 expression and TCE transformation capability of several lines were evaluated and the best-performing line was self-crossed and progeny, designated 3–1–2, were used for further study (11). Vector control plants were transformed with plasmid pCGN1578, which did not contain cDNA for CYP2E1. Wild type plants were not transformed. Sterile plant material was obtained by propagating apical cuttings from whole plants in sterile vessels with amended growth media consisting of Murashige and Skoog (MS) + vitamins (Caissons MSP002), 30 g L-1 sucrose, 1 g L-1 phytogot gellan gum, 3.5 g L-1 phytagar, 1 mL L-1 Fungigone (Bioworld, Dublin OH), and 1 mM indole-3-butyric acid. Apical cuttings VOL. 42, NO. 1, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
289
TABLE 1. Initial Mass Dosing (µg) and Equilibrium Liquid Concentration (µg/mL) of the Compounds Tested in This Study
compound perchloroethylene trichloroethylene vinyl chloride chloroform carbon tetrachloride bromodichloromethane 1,1,1 - trichloroethane toluene benzene
initial mass (µg)
equilibrium liquid concentration (µg/mL)
54 125 1000 0.25 10 0.6 16 15 85
1.64 5.6 22 0.0025 0.2 0.05 0.5 0.9 1
were placed in water in 50 mL plastic centrifuge tubes (Sarstedt, Newton, NC) for 24 h prior to sterilization. Sterilization was performed by submersing the cuttings for 10 min in 10% bleach solution (final concentration 0.35% NaClO), rinsing with sterile, deionized water, 5 min in 1% Iodophor solution, rinsing three times with sterile, deionized water, and transferring into a sterile growth vessel. All work was performed in a sterile, laminar-flow hood. Plants grew in 6-8 weeks and served as sterile source material for the vial experiments. Cuttings for the vial experiments were obtained from the apical ends of the sterile plants. Each apical cutting had 3-5 leaves and wet mass from 1.5 to 2 g. Experimental Setup. Sterile plant cuttings were placed in clear glass volatile organic analysis (VOA) vials, 40 mL nominal volume, in 10 mL of half-strength MS basal salt mixture with 30 g L-1 sucrose (pH 5.7). Vials were capped with 24 mm Mininert valves with silicone septa (VICI Valco Instruments, Houston, TX) and held for 24 h prior to chemical exposure. Cuttings were incubated at approximately 23 °C under fluorescent lights with a 16-h photoperiod. Several chemicals were investigated in this study including perchloroethylene (PCE), TCE, vinyl chloride (VC), CT, TCA, chloroform, bromodichloromethane (BDCM), benzene, and toluene. For each experiment, a set of at least three transgenic and three vector control plant cuttings was set up, with one cutting per vial, and injected with known amounts of chemical using gastight glass syringes (Hamilton Co., Reno, NV) either in the form of a stock solution in water, or as a gas (Table 1). Chemical mass from the stock solutions was calculated based on volume, solubility in water, and temperature. Mass from gas injections was calculated based on volume, temperature, and pressure. Sampling and Analysis. Initial sampling occurred at least two hours following the introduction of chemicals into the VOA vials to allow air–liquid equilibrium. Subsequent sampling frequency was either 24 or 48 h, according to the uptake rate of the chemical of interest. Samples were obtained by withdrawing a known volume of vial headspace with a gastight glass syringe and injecting manually into a gas chromatograph (GC) for analysis. Headspace samples from vials containing PCE, TCE, CT, TCA, and chloroform were analyzed with a GC-electron capture detector (ECD). Two GC-ECDs were used during the course of the experiments: Perkin-Elmer AutoSystem XL GC-ECD with a Supelco PTE-5 column (30 m, 0.32 mm i.d., 0.32 µm film thickness); and, SRI 8610C with a Supelco SPB-624 column (60 m, 0.53 mm i.d., 3.0 µm film thickness). Headspace of vials containing VC, benzene, and toluene was analyzed with a SRI 8610C GC-flame ionization detector (FID) with a Supelco Alumina Sulfate Plot column (50 m, 0.53 mm i.d.). The GC oven temperatures were isothermal for all experiments. Sample injections were performed in triplicate. 290
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 1, 2008
FIGURE 1. Degradation of chloroethenes TCE (A) and VC (B) by CYP2E1-expressing and vector control or wild type tobacco plant cuttings over time. Chemical dosing occurred at time zero. Sample collection began 2–12 h after initial dosing to allow air–liquid equilibrium. Points are the average mass of individual vial tests (n ) 4 TCE; n ) 3 VC); error bars indicate ( 1 standard deviation. Analysis for trichloroethanol (TCOH) was performed in the TCA degradation experiments. Plants were removed from vials, flash frozen in liquid nitrogen, and ground in a mortar and pestle. Approximately 1.5 g of frozen tissue was placed in a chilled centrifuge tube with 2 mL of 1 N H2SO4/10% NaCl and shaken vigorously for 1 min. Methyl tert-butyl ether (MTBE; 10 mL) was added and the mixture was shaken for 1 min, followed by centrifugation at 8000 rpm for 10 min. The supernatant (7 mL) was transferred to a septa-capped, amber glass vial containing 2 g of Na2SO4 and stored for 1 h. An aliquot was transferred to an autosampler vial for analysis on a GC-ECD. The GC oven temperature program was 40 °C for 1 min, ramped at 10 °C minute-1 for 19 min, and held at 230 °C for 4 min. Analysis for BDCM was performed by solvent extraction due to difficulties in optimizing the peak response with manual injection in the SRI GC-ECD. Initially, cuttings were placed in 15 mL of liquid media. Following chemical dosing and equilibration, 5 mL of media was extracted through the septa with a gastight syringe and mixed with 5 mL of 1 N H2SO4/10% NaCl in a 40 mL amber glass vial and shaken vigorously for 1 min. Ten mL of MTBE was added, the mixture was shaken for 1 min, and allowed to stand for 10 min. Seven mL of MTBE was transferred, held with Na2SO4, and analyzed by GC-ECD as described above for TCOH. Analysis of the uptake kinetics was performed by a regression analysis of the natural-log transformed concentration data over time. The significance of the regression was verified through analysis of variance testing for each regression at a significance level of 0.05. First-order chemical loss
FIGURE 2. Degradation of haloalkanes by CYP2E1-expressing and vector control or wild type tobacco plant cuttings over time. Data for chloroform (A), BDCM (B), CT (C), and TCA (D) are shown. Chemical dosing occurred at time zero. Sample collection began 2-12 h after initial dosing to allow air–liquid equilibrium. Points are the average mass of individual vial tests (n ) 3 BDCM; n ) 4 chloroform, CT, and TCA); error bars indicate ( 1 standard deviation. rates were normalized for plant mass and compared using Student’s t statistic at a significance level of 0.05.
Results Chloroethenes. Tobacco plant cuttings genetically modified with CYP2E1 displayed significant degradation of TCE, while the vector control plant cuttings did not display activity significantly different from that of the no-plant controls (Figure 1A). Similarly, tobacco plant cuttings expressing CYP2E1 degraded VC, while the wild type did not demonstrate statistically significant degradation (Figure 1B). There was no removal of PCE by any of the plants tested (data not shown). Analysis of the log-transformed data indicated that transformation kinetics were first-order for TCE (r 2 ) 0.93, P < 0.05) and VC (r 2 ) 0.99, P < 0.05). Haloalkanes. Chloroform (Figure 2A) and BDCM (Figure 2B) were degraded in vials containing the CYP2E1-transformed plant cuttings, while there was no significant removal in vials with either the wild type or vector control plant cuttings. The rate of removal of CT (Figure 2C) by plant cuttings expressing CYP2E1 was significantly greater than that of either the no-plant or vector controls assuming firstorder kinetics (P < 0.05). The normalized degradation rate of CT was 3.3 times greater for the plant cuttings expressing CYP2E1 compared to the vector controls. Analysis of the logtransformed data indicated that chloroform (r 2 ) 0.88, P < 0.05) and CT (r 2 ) 0.98, P < 0.05) degradation followed firstorder kinetics. The experimental method used to measure BDCM precluded analysis of uptake kinetics. Neither plant cuttings expressing CYP2E1 nor vector control cuttings significantly degraded TCA in any of the experimental runs (Figure 2D). To determine whether any TCA was degraded to TCOH, free TCOH was measured in
plant tissues following exposure of CYP2E1-transformed plant cuttings to TCA or TCE for 7 days. TCE transformation was used as a positive control. Plant cuttings expressing CYP2E1 exposed to TCE had an average TCOH concentration of 0.226 µg per g wet plant while the TCA-exposed plants had no detectable levels of TCOH. Aromatic Hydrocarbons. Benzene and toluene were almost completely removed within 4 days from vials containing CYP2E1-transformed plants (Figure 3A and B). The control plant cuttings did not show any activity toward benzene. Toluene was degraded by the vector control plant cuttings, though at a significantly lower rate than the CYP2E1transformed plant cuttings (P < 0.05). The first-order removal rate was 10.4 times higher for the CYP2E1-transformed plant cuttings compared to the vector control plant cuttings. Analysis of the log-transformed data indicated that benzene (r 2 ) 0.97, P < 0.05) and toluene (r 2 ) 0.98, P < 0.05) degradation followed first-order kinetics.
Discussion These results demonstrate the increased degradation of a wide variety of VOCs by genetically modified plants that express mammalian CYP2E1. For most of the chemicals investigated, the transgenic plant cuttings displayed either increased activity, or were able to degrade pollutants that were otherwise recalcitrant to the wild type or vector control tobacco plant cuttings. The greater activity conferred by CYP2E1 expression has the potential to improve the effectiveness of phytoremediation in the cleanup of soil, groundwater, and indoor air. The enhanced metabolism of TCE by plants expressing CYP2E1 has been previously reported (11). The replication VOL. 42, NO. 1, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
291
FIGURE 3. Degradation of aromatic hydrocarbons toluene (A) and benzene (B) by CYP2E1-expressing and vector control tobacco plant cuttings over time. Chemical dosing occurred at time zero. Sample collection began 2 h after initial dosing to allow air–liquid equilibrium. Points are the average mass of individual vial tests (n ) 4); error bars indicate ( 1 standard deviation. of these results herein suggests that utilizing plant cuttings is an appropriate method for comparing metabolic activity. Wolfe and Hoehamer (13) reported the degradation of TCE and CT by cell-free extracts of Canadian Waterweed, suggesting that certain plants express native enzymes capable of degrading these compounds. Here, we demonstrated a marked increase in the metabolism of both TCE and CT by transgenic plant cuttings indicating the potential for accelerated remediation. To our knowledge, no other reports have shown the metabolism of VC by plants. The ability of these transgenic plants to degrade both TCE and VC may be valuable in areas of mixed contamination. 1,1,1-Trichoroethane was not significantly degraded by the transgenic tobacco plants despite the fact that it is a known substrate for mammalian CYP2E1 (12). To confirm that TCA was not metabolized by transformed plants, we looked for the production of TCOH in CYP2E1-transformed plants exposed to TCA. Trichloroethanol has been shown to be a metabolite of CYP2E1 in mammalian tissues and hepatic microsomes exposed to TCA (12). However, no TCOH was detected, which was surprising considering the wide range of activity displayed by CYP2E1 in this plant system. Cytochrome P450 enzymes function with the accessory proteins NADPH-P450 reductase (NPR) and Cytochrome b5 (b5). Several of the P450 enzyme isoforms, including 2E1, require b5 to carry out their full range of activities, though the mechanism appears to differ depending on the P450 (14). While other P450s can be stimulated by either holo-b5 or apo-b5, or even through protein–protein interactions with other P450s, CYP2E1 may only be enhanced by holo-b5, suggesting a more stringent requirement for the enhance292
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 1, 2008
ment interactions (15). The pSLD3 plasmid used to transform the tobacco contained only genes for human CYP2E1, but not NPR or b5. The results of this work indicate that the CYP2E1 still catalyzes a wide range of activity against xenobiotics, likely utilizing native plant homologues to NPR and b5 to facilitate electron transport. However, differences between the homologous versions found in plants compared to those in humans possibly causes conformational changes that inhibit catalytic activity toward TCA. Benzene and toluene can be taken up by plants and released to the atmosphere with limited transformation (16). Toluene volatilization through leaves can be reduced by inoculating plants with endophytes expressing the toluenedegrading pTOL plasmid, thereby increasing in planta degradation (17). Similarly, the improved degradation of benzene and toluene by CYP2E1-transformed plants shown here could also reduce transpiration and increase phytoremediation effectiveness. The trihalomethanes (THMs) chloroform and BDCM were degraded by the transgenic plant cuttings. Trihalomethanes, which may be present as disinfection byproduct in chlorinated drinking water, can volatilize into indoor air. In preliminary work, we found a 9-fold increase in average daily chloroform concentration in the air with shower use (data not shown; BDCM was not measured). Studies have looked at the potential of wild type household plants for the remediation of contaminants such as formaldehyde, benzene, and TCE, from indoor air (18, 19). Assuming that plant degradation was limited by the endogenous enzymatic activity, transgenic plants similar to those described in this paper may have application in the remediation of indoor air. Overall, our findings suggest potential applications for genetically modified plants in the field. The CYP2E1transformed tobacco plants have demonstrated activity against broad classes of environmental pollutants such as chlorinated solvents (both alkenes and alkanes) and aromatic hydrocarbons suggesting that CYP2E1-transformed plants could be used to remove pollutants faster in environments contaminated by single or mixed contaminants, particularly when cleanup is limited by planting area or time. It is important, however, to acknowledge the limitations of this study, which was designed specifically to investigate differences in plant metabolism. Field applications are complex; plant metabolism alone may not be the determining measure of success. Further, field applications may require plants with higher biomass production and deeper root systems than the tobacco to effectively reach contaminated areas. The spread of transgenic material may be a concern, particularly with plants with an annual reproductive cycle, such as the tobacco. Transgenic trees that express CYP2E1 could combine the higher metabolic activity with deeper root systems, while having a more controlled reproductive cycle. This is potentially quite desirable. This study and others, which have reported on various applications of genetically modified plants for detoxification of polluted environments (9, 10), demonstrate the potential of transgenic modifications to increase the range and utility of phytoremediation.
Acknowledgments This work was funded by NIEHS grant 2P42-ES004696-19, DOE grant DE-FG02-03ER63663, and the Valle Fellowship and Exchange Program at the University of Washington. C.A.J. and G.X. should be considered co-first authors on this manuscript.
Literature Cited (1) U.S. Environmental Protection Agency. Common Chemicals Found at Superfund Sites. Available at http://www.epa.gov/ superfund/resources/chemicals.htm.
(2) Moran, M. J.; Zogorski, J. S.; Squillace, P. J. Chlorinated solvents in groundwater of the United States. Environ. Sci. Technol. 2007, 41, 74–81. (3) Doherty, R. E. A history of the production and use of carbon tetrachloride, tetrachloroethylene, trichloroethylene and 1,1,1trichloroethane in the United States: Part 1 - Historical background; Carbon tetrachloride and tetrachloroethylene. Environ. Forens. 2000, 1, 69–81. (4) Isbell, M. A.; Stolzberg, R. J.; Duffy, L. K. Indoor climate in interior AK: simultaneous measurement of ventilation, benzene and toluene in residential indoor air of two homes. Sci. Total Environ. 2005, 345, 31–40. (5) Fiss, E. M.; Rule, K. L.; Vikesland, P. J. Formation of chloroform and other chlorinated byproducts by chlorination of triclosancontaining antibacterial products. Environ. Sci. Technol. 2007, 41, 2387–2394. (6) Pilon-Smits, E. A. H.; Freeman, J. L. Environmental cleanup using plants: Biotechnological advances and ecological considerations. Front. Ecol. Environ. 2006, 4, 203–210. (7) Cherian, S.; Oliveira, M. M. Transgenic plants in phytoremediation: Recent advances and new possibilities. Environ. Sci. Technol. 2005, 39, 9377–9390. (8) Hannink, N.; Rosser, S. J.; French, C. E.; Basran, A.; Murray, J. A. H.; Nicklin, S.; Bruce, N. C. Phytodetoxification of TNT by transgenic plants expressing a bacterial nitroreductase. Nat. Biotechnol. 2001, 19, 1168–1172. (9) Rylott, E. L.; Jackson, R. G.; Edwards, J.; Womack, G. L.; SethSmith, H. M. B.; Rathbone, D. A.; Strand, S. E.; Bruce, N. C. An explosive-degrading cytochrome P450 activity and its targeted application for the phytoremediation of RDX. Nat. Biotechnol. 2006, 24, 216–219. (10) Kawahigashi, H.; Hirose, S.; Ohkawa, H.; Ohkawa, Y. Phytoremediation of the herbicides atrazine and metolachlor by transgenic rice plants expressing human CYP1A1, CYP2B6, and CYP2C19. J. Agric. Food Chem. 2006, 54, 2985–2991. (11) Doty, S. L.; Shang, T. Q.; Wilson, A. M.; Tangen, J.; Westergreen, A. D.; Newman, L. A.; Strand, S. E.; Gordon, M. P. Enhanced metabolism of halogenated hydrocarbons in transgenic plants
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
containing mammalian cytochrome P450 2E1. Proc. Natl. Acad. Sci., U.S.A. 2000, 97, 6287–6291. Guengerich, F. P.; Kim, D. H.; Iwasaki, M. Role of human cytochrome P450 2E1 in the oxidation of many low-molecularweight cancer suspects. FASEB J. 1991, 5, A1515–A1515. Wolfe, N. L.; Hoehamer, C. F. Enzymes used by plants and microorganisms to detoxify organic compounds. In Phytoremediation: Transformation and Control of Contaminants; McCutcheon, S. L., Schnoor, J. L., Eds.; Wiley-Interscience: Hoboken, NJ, 2003; pp 159–187. Shimada, T.; Mernaugh, R. L.; Guengerich, F. P. Interactions of mammalian cytochrome P450, NADPH-cytochrome P450 reductase, and cytochrome b(5) enzymes. Arch. Biochem. Biophys. 2005, 435, 207–216. Yamazaki, H.; Nakamura, M.; Komatsu, T.; Ohyama, K.; Hatanaka, N.; Asahi, S.; Shimada, N.; Guengerich, F. P.; Shimada, T.; Nakajima, M.; Yokoi, T. Roles of NADPH-P450 reductase and apo- and holo-cytochrome b(5) on xenobiotic oxidations catalyzed by 12 recombinant human cytochrome P450s expressed in membranes of Escherichia coli. Protein Expression Purif. 2002, 24, 329–337. Burken, J.; Schnoor, J. Distribution and volatilization of organic compounds following uptake by hybrid Poplar trees. Int. J. Phytorem. 1999, 1, 139–151. Barac, T.; Taghavi, S.; Borremans, B.; Provoost, A.; Oeyen, L.; Colpaert, J. V.; Vangronsveld, J.; van der Lelie, D. Engineered endophytic bacteria improve phytoremediation of watersoluble, volatile, organic pollutants. Nat. Biotechnol. 2004, 22, 583–588. Wolverton, B. C.; McDonald, R. C.; Watkins, E. A. Foliage plants for removing indoor air-pollutants from energy-efficient homes. Econ. Bot. 1984, 38, 224–228. Cornejo, J. J.; Munoz, F. G.; Ma, C. Y.; Stewart, A. J. Studies on the decontamination of air by plants. Ecotoxicology 1999, 8, 311–320.
ES071197Z
VOL. 42, NO. 1, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
293
