Environ. Sci. Technol. 2007, 41, 8301–8307
Enantioselective Microbial Transformation of the Phenylpyrazole Insecticide Fipronil in Anoxic Sediments W. JACK JONES,* CHRISTOPHER S. MAZUR, JOHN F. KENNEKE, AND A. WAYNE GARRISON Ecosystems Research Division, National Exposure Research Laboratory, U.S. Environmental Protection Agency, 960 College Station Rd., Athens, Georgia 30605
Received June 12, 2007. Revised manuscript received October 2, 2007. Accepted October 19, 2007.
Fipronil, a chiral insecticide, was biotransformed initially to fipronil sulfide in anoxic sediment slurries following a short lag period. Sulfidogenic or methanogenic sediments transformed fipronil with half-lives of approximately 35 and 40 days, respectively. In all microbially active sediment slurries tested, the transformation of fipronil to fipronil sulfide was enantioselective. In the sulfidogenic sediment slurry, the enantiomeric fraction (EF) of fipronil decreased from an initial racemic EF value of 0.46 to a value of 0.22 during the incubation period of active fipronil transformation, indicating preferential transformation of theS-(+)-enantiomer.Apreviouslyunidentifiedproduct,5-amino1-[2,6-dichloro-4-(trifluoromethyl)-phenyl]-4-(trifluoromethylthio)-1-H-pyrazole-3-carboxyamide, or fipronil sulfide-amide, was detected in the sulfidogenic slurries and coincided with the loss of fipronil sulfide. Biota from methanogenic freshwater sediment slurries also transformed fipronil enantioselectively but with a preference for the R-(-)-enantiomer. In all microbially inhibited (autoclaved) sediment slurries tested, no changes in the enantiomeric fractions of fipronil were observed and only low levels (95%) S-(+)- and R-(-)-enantiomers of fipronil were provided by Chiral Technologies (West Chester, PA) and were prepared by HPLC using a CHIRALPAK AS-H preparative column. FeS was obtained from Alfa Aesar (Ward Hill, MA), and FeS2 was obtained from Wards Scientific (Rochester, NY). All other chemicals were purchased from Aldrich Chemical Co. (Milwaukee, WI) and were of ACS reagent grade or higher. Organic solvents were purchased from Fisher Scientific Co. (Atlanta, GA) and were Optima grade or equivalent. Sediment Collection. Two sampling sites, representing a sulfate-reducing sediment and a methanogenic sediment, were selected to assess the microbial transformation of fipronil. Sediment samples from the sulfate-reducing site (35) were collected from a tidal wetland (approximately 20–25 cm depth from the West Branch Canal Creek) located at Aberdeen Proving Ground, Maryland. The sediment (from site WB-26) was characterized as anoxic and sulfidogenic due to the presence of abundant sulfide minerals and reduced iron. Sediments characterized as methanogenic were collected from the upper 5–20 cm zone of a freshwater pond located at a USDA field research site near Watkinsville, Georgia. The USDA sediments were actively methanogenic and contained low sulfide levels relative to the Aberdeen wetland site. Immediately following collection, sediments were placed in glass jars, sealed, and stored at 4 °C. 8302
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 24, 2007
TABLE 1. Physiochemical Characteristics of USDA Watkinsville Pond Sediment and Aberdeen Wetland Sediment site
pHa
%TOCb
%Nc
redox statusd
USDA Aberdeen
6.7 7.8
4.0 2.2
0.30 0.22
methanogenic sulfidogenic
a pH of 1:1 slurry of sediment plus deionized water. Weight percent. c Total nitrogen (wt %). d Determined by headspace gas analysis and modified Cline assay (38) for sulfide. b
Specific physicochemical characteristics of the Aberdeen and USDA sediments are presented in Table 1. Kinetic Studies (Sediments). Sediment slurries from the freshwater pond and Aberdeen wetland were prepared inside an anaerobic chamber (Coy Laboratory Products, Grass Lake, MI) containing an atmosphere of 99% N2 and 1% H2. The collected sediment was passed through a 1-mm sieve and mixed with either anoxic (N2-sparged) site water (USDA sediment) or half-strength artificial seawater (Aberdeen wetland sediment) to achieve a sediment solids concentration of approximately 100 g/L of slurry. A stock solution of fipronil (3.5 g/L), dissolved in acetone, was aseptically added to sterile, empty, amber serum bottles, and the acetone was evaporated to dryness prior to the addition of sediment slurry (10–20 mL). The serum bottles were subsequently sealed with Teflon-lined, butyl-rubber septa and crimped with aluminum caps. The initial target concentration of fipronil was approximately 3.3 mg/L. Microcosm bottles were incubated at 25 °C on a rotary shaker, and slurry samples were collected by syringe to assess fipronil transformation. Control microcosms were prepared by autoclaving sediment slurries at 120 °C for 30 min over three consecutive days and aseptically adding slurries to sterile serum bottles containing fipronil, as described above. Triplicate samples were collected at selected time points to assess fipronil transformation. Kinetic Studies (FeS and FeS2). To assess the possible contribution of abiotic fipronil transformation, batch experiments were performed using, as reactants, sulfide minerals that are commonly present in anoxic sediments. Reduced mineral solutions (approximate pH of 6.5) were prepared in an anaerobic chamber. Approximately 2.5 g of either FeS or FeS2 was weighed into a series of 60-mL serum bottles containing 50 mL of argon-sparged, deionized water and sealed with Teflon-faced butyl-rubber septa. Bottles containing the mineral solution were crimp-sealed and spiked with 100 µL of a 1540 mg/L stock solution of fipronil prepared in acetone. Bottles were incubated on a rotor drum in the dark at 25 °C. Periodically, a 1 mL sample was removed from the experimental bottles using an argon-purged, gastight syringe and transferred to a sample vial containing 200 µL of hexane. A second sample was removed and added to 500 µL of methyl tertiary butyl ether (MTBE). Samples were vigorously mixed on a vortex shaker for 10 s. The hexane and MTBE extracts were analyzed for fipronil and fipronil transformation products, respectively. Analytical Techniques. Gas chromatography (GC) techniques were used to quantify fipronil, products of fipronil transformation, and headspace methane concentrations. For the analysis of fipronil, fipronil sulfide, and other products, samples of the experimental sediment slurry were extracted overnight by vigorous shaking using an equal volume of acetonitrile. An aliquot of the acetonitrile extract was then extracted with hexane for GC analysis. For the analysis of polar transformation products, a portion of the aqueous phase of the centrifuged (5000g for 10 min) sediment slurry was extracted with an equal volume of MTBE.
FIGURE 1. GC profile of S-(+)- and R-(-)-enantiomers of technical-grade fipronil showing near-baseline separation using a chiral selective capillary column (* indicates chiral center of the fipronil structure). Fipronil and fipronil transformation products were analyzed by gas chromatography and electron capture detection (ECD). Separation was achieved with a 30-m, 0.32-mm-i.d., 0.25-µm-film-thickness DB-5 capillary column (J&W Scientific, Folsom, CA). The detector and injection-port temperatures were 300 and 250 °C, respectively. The temperature program was as follows: 60 °C, hold 5 min; 10 °C min-1 to 90 °C, no hold; 20 °C min-1 to 220 °C, 5 min hold. To confirm product identities, sample extracts were also analyzed by splitless injection GC/MS (Agilent 6890/5872) operated in electron ionization mode. A 30-m, 0.25-mm-i.d., 0.25-µmfilm-thickness DB-5 MS fused silica capillary column (J & W Scientific) was used for analyte separation. Electron energy was 70 eV. Fipronil sulfide and fipronil amide were confirmed by comparison of mass spectra and retention times with authentic standards. A standard of fipronil sulfide-amide was prepared by hydrolyzing fipronil sulfide at pH 12 and was confirmed by gas chromatography/mass spectrometry (GC/ MS). An enantiomer-specific chiral analysis of fipronil was also achieved using a gas chromatograph equipped with ECD. Enantiomers were separated using a 30-m, 0.25-mm-i.d., 0.2µm-film-thickness BGB-172 chiral column (BGB Analytic AG, Laufrainweg, Germany). The column consisted of 20% tertbutyldimethylsilylated-β-cyclodextrin in BGB-15 (15% phenyl-, 85% methylpolysiloxane). The injection and detector temperatures were 260 and 325 °C, respectively. The temperature program was as follows: 150 °C, hold 2 min; 1 °C min-1 to 220 °C, hold for 20 min. Methane was analyzed using a gas chromatograph equipped with a flame ionization detector and separated on a Poropak N column (6′ × 1/8′′ o.d., 80/100 mesh; Alltech, Deerfield, IL) at 50 °C with a N2 carrier flow of 15 mL min-1. Ferric iron (Fe3+) and ferrous iron (Fe2+) species were analyzed in both the pore water and the solid phase of experimental sediment slurries using the ferrozine assay described in Viollier et al. (36, 37). Dissolved sulfide concentrations were measured in the aqueous phase of experimental sediment slurry samples using a modified Cline assay (38) and direct filtration (0.22 µm syringe filter) into an acidified 2% (w/v) zinc acetate trapping solution. Sulfide standards were prepared using thoroughly washed and dried sodium sulfide crystals.
Results and Discussion Chromatographic Separation of Fipronil Enantiomers. Fipronil enantiomers were separated by chiral GC (Figure 1) using a commercially available 20% tert-butyldimethylsilylated-β-cyclodextrin column. The first eluting peak was identified as the S-(+)-enantiomer by comparing its retention time with an enantiomerically pure standard. The enantiomeric fraction (EF ) area of (+)-enantiomer/[area of (+)enantiomer + area of (-)-enantiomer]) of the fipronil racemate used in all experimental studies was calculated to
be 0.46 ((0.04). Near-baseline separation of the fipronil enantiomers was achieved using the chiral cyclodextrin column. Fipronil Transformation and Enantioselective Degradation in Anoxic Sediments. Fipronil transformation in environmental samples was investigated using anoxic sediments collected from two geographically distinct locations representing different physicochemical characteristics (Table 1). Differences in sediment pH, the percentage of organic carbon, and redox conditions were apparent. While lower in total organic carbon (TOC) than the USDA site, sediment samples from the Aberdeen site consisted of high levels of sulfide minerals typical of sulfidogenic wetland environments. Fipronil transformation was observed in the sulfidogenic Aberdeen wetland sediment slurries following a short lag phase of approximately 5 days (Figure 2A). A greater than 98% loss of the initial fipronil amendment (3.3 mg/L concentration) was noted after a total incubation period of 88 days. On the basis of the available data, the half-life of fipronil in the sulfidogenic wetland sediment slurry was determined to be approximately 35 days. In this nonacclimated, microbially active sediment, fipronil loss was associated with a near stoichiometric production of fipronil sulfide, indicating that the reductive transformation of fipronil was the predominant, initial biotransformation process in the anoxic sediment microcosms. Similar rates of fipronil transformation (to fipronil sulfide) were observed in studies by Ngim and Crosby (21) in flooded rice soils, and Walse et al. (23) recently reported rapid fipronil transformation to the sulfide metabolite in the sediment fraction of anoxic estuarine mesocosms. The highest concentration of fipronil sulfide detected in the Aberdeen study (approximately 3.4 mg/L) occurred on day 62 in the microbially active sediment slurries; after continued incubation, an appreciable loss of the fipronil sulfide was evident (Figure 2A). Subsequent product analysis (GC-MS) of the experimental sediment slurry following the loss of fipronil sulfide revealed the presence of a previously unidentified metabolite, 5-amino-1-[2,6-dichloro-4-(trifluoromethyl)-phenyl]-4-(trifluoromethylthio)-1-H-pyrazole-3carboxyamide, or fipronil sulfide-amide (Figure 3), hypothesized to be the biohydrolysis product of fipronil sulfide (Figure 4). Identification of this novel product was confirmed from mass spectral data and by the comparison of GC retention time with a synthesized standard. The sulfide-amide metabolite exhibited major ions at m/z 438 (M+), 255, and 369 (M+ - 69), the latter representing a loss of the trifluoromethyl group. An aqueous fipronil sulfide standard treated at pH 12 yielded a transformation product with similar mass spectra and GC retention times as the postulated fipronilsulfide-amide metabolite. Fipronil amide, or 5-amino-1-[2,6dichloro-4-(trifluoromethyl)-phenyl]-4-(trifluoromethylsulfinyl)-1-H-pyrazole-3-carboxyamide, primarily a base-catalyzed hydrolysis product of fipronil (Figure 4), was detected at low VOL. 41, NO. 24, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8303
FIGURE 2. (A) Kinetics of fipronil biotransformation to fipronil-sulfide in sulfidogenic Aberdeen wetland sediment slurries. Panels B and C illustrate the mean enantiomer fraction values of fipronil biotransformation in sterile and microbially active (live) sediment slurries from the sulfidogenic Aberdeen wetland site (B) and the methanogenic USDA site (C), respectively. Note that the methanogenic USDA sediment (C) exhibited the opposite enantioselective preference of the Aberdeen sediment.
FIGURE 3. Mass spectrum of the fipronil biotransformation product (from the Aberdeen wetland site) identified 5-amino-1-[2,6-dichloro-4-(trifluoromethyl)-phenyl]-4-(trifluoromethylthio)-1H-pyrazole-3-carboxyamide, or fipronil sulfide-amide. levels in other fipronil transformation experiments (21, 22) but was detected only at trace levels in the Aberdeen sediment slurries. In microbially inhibited sediment slurries (sterilized by autoclaving), minor loss (
