Biotransformation of 8:2 Fluorotelomer Alcohol in Soil and by Soil

In the presence of 1,4-dioxane, which is not easily biodegraded, 8:2 FTOH .... to all bottles regardless of the initial condition (0 or 1 mM octanol a...
1 downloads 0 Views 267KB Size
Environ. Sci. Technol. 2007, 41, 8024–8030

Biotransformation of 8:2 Fluorotelomer Alcohol in Soil and by Soil Bacteria Isolates J I N X I A L I U , † L I N D A S . L E E , * ,† LORING F. NIES,‡ CINDY H. NAKATSU,† AND RONALD F. TURCO† Department of Agronomy, and School of Civil Engineering, Purdue University, West Lafayette Indiana 47907-2054

Received April 13, 2007. Revised manuscript received September 18, 2007. Accepted September 20, 2007.

The microbial transformation of 8:2 fluorotelomer alcohol (FTOH) to perfluorocarboxylic acids, including the globally detected perfluorooctanoic acid (PFOA), has recently been confirmed to occur in mixed bacteria cultures, activated sludge, and soil. However, little is known to date about the microbes involved in the transformation. In the present study, the effect of three carrier solvents (ethanol, octanol, and 1,4-dioxane), which may serve as carbon sources, on the aerobic degradation rate of 8:2 FTOH and metabolite distribution was evaluated both in a clay loam soil and in two pure soil bacterial cultures. Biodegradation pathways appeared similar regardless of the solvent; however, significant differences in 8:2 FTOH degradation rates were observed: 1,4-dioxane > ethanol > octanol. In the presence of 1,4-dioxane, which is not easily biodegraded, 8:2 FTOH degradation was the fastest. With octanol, which is a structural analogue of 8:2 FTOH, the transformation was inhibited, but upon depletion of octanol, 8:2 FTOH was biodegraded. In the pure culture study, two soil bacterial strains, Pseudomonas species OCY4 and OCW4, enriched from soil using octanol as a sole carbon source, also transformed 8:2 FTOH without prior exposure or acclimation to 8:2 FTOH. Increased biomass resulting from octanol metabolism did increase 8:2 FTOH transformation rates; however, 8:2 FTOH could not support bacterial growth, indicating the transformation by pure cultures was via cometabolic processes.

Introduction The confirmed global presence of long-chain alkyl perfluorocarboxylates (PFCAs) and their potential precursors in various environmental compartments (1–3), as well as in wildlife and humans (4), has drawn the attention of scientists and the public. Recent studies have shown that the major contributors to environmental loads are direct sources from the use of perfluorooctanate (PFOA) and perfluorononanate (PFNA) ammonia salts in fluoropolymer manufacturing (5). However, scientific data for assessment of the relative contributions from indirect sources such as precursors from fluorotelomer alcohols (FTOHs) and their surfactant or polymer derivatives is still lacking. FTOHs [F(CF2CF2)nCH2CH2OH], with even-numbered fluorocarbon chains and an ethanol moiety, are important * Corresponding author phone: 765-494-8612; fax: 765-496-2926; e-mail: [email protected]. † Department of Agronomy. ‡ School of Civil Engineering. 8024

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 23, 2007

industrial intermediates used in the synthesis of fluorotelomer intermediates, surfactants, and polymeric materials (6) and, in some cases, may be present as residuals in consumer products (7). Cleavage of covalent bonds in these fluorotelomer-based substances is suspected to release FTOHs into the environment over time. Therefore, gleaning information on their environmental stability and the contribution of PFCAs from FTOHs is pivotal in the estimation of the load of potential PFCAs from fluorotelomer-based consumer products. Although the volatile nature of FTOHs has been the focus of numerous studies (8, 9), high volatility does not necessarily mean that FTOHs would not be present in soils at significant levels. Studies that monitor the levels of FTOHs in soil environments are still sparse; however, 8:2 FTOH and longer FTOHs are known to exhibit strong sorption to soils (10, 11). Ultimately, the competition of several parallel processes, including sorption and transformation in soil and partitioning into the air, collectively determine the fate of FTOHs. The ethanol moiety that facilitates FTOH derivatization makes FTOHs susceptible to degradation. Past studies with an acclimated chlorinated alkanes/alcohols degrading culture (12), activated sludge (13), mixed bacterial culture (14), soils (15), and rat hepatocytes (16) have all verified the production of PFOA from 8:2 FTOH oxidation and subsequent removal of the ethanol moiety. The breakdown is mainly limited to nonfluorine functionality, but a higher degree of defluorination has been observed with the formation of a six-carbon perfluorocarboxylate (PFHA), and several parallel pathways were proposed to be functioning simultaneously (14). In previous microbial degradation studies (12–14), 8:2 FTOH, which has a low aqueous solubility (10), was dissolved in ethanol to introduce it into microcosms for degradation. Wang et al. (14) observed that ethanol significantly enhanced production of PFHA and the release of 14CO2 from the β-carbon, indicting that ethanol was likely used as a carbon substrate for microbial growth. Therefore, questions arise as to how FTOH biotransformation is affected by carrier solvents or other carbon substrates and if biotransformation of 8:2 FTOH is growth-sustaining or through cometabolic reactions. The latter, if true, indicates that the degradation rates of FTOHs and likely other telomer compounds in the natural environment may be highly dependent on the availability of other carbon sources. Elucidation of these questions will help to quantify the degradation kinetics of fluorotelomer surfactants or polymers, whose long-term environmental stability and potential contribution to PFCAs is of great concern. In the present study, the effect of three carrier solvents (ethanol, octanol, and 1,4-dioxane) on the aerobic degradation rate of 8:2 FTOH and metabolites distribution was evaluated in both a clay loam soil and pure soil bacterial isolates. Ethanol is a common carrier solvent that degrades easily and can be used as a direct carbon source for growth by several microbial populations (17). Octanol, a structural analogue of 8:2 FTOH, can undergo β-oxidation (18), which is a speculated, yet debatable, transformation mechanism for 8:2 FTOH (14, 16). We hypothesize that octanol and 8:2 FTOH are likely subject to similar enzymatic reactions, which may be carried out by similar microflora. 1,4-Dioxane is not easily degraded in soil (19), and if degraded, the biochemical pathways and enzymes involved would be different than those active for alcohols, thus serves as a reasonable control. 10.1021/es0708722 CCC: $37.00

 2007 American Chemical Society

Published on Web 10/26/2007

Materials and Methods Chemicals and Soil. The following perfluorinated compound standards were obtained as detailed in the Supporting Information: 8:2 fluorotelomer alcohol (8:2 FTOH), 8:2 fluorotelomer carboxylic acid (8:2 FTCA), 8:2 fluorotelomer R,β-unsaturated carboxylic acid (8:2 FTUCA), perfluorohexanoic acid (PFHA), perfluoroheptanoic acid (PFHeA), perfluorooctanoic acid (PFOA), perfluorononanoic acid (PFNA), 2H,2H,3H,3H-pentadecafluorodecanoic acid (7:3 FTCA), 2H,2H,3H,3H- pentadecafluorodecenoic acid (7:3 FTUCA), and 2H-pentadecafluoro-2-nonanol (7:2 sFTOH) and the internal standards [1D,1D,2D,2D,313C]8:2 FTOH and [1,213 C]PFOA. A Chalmers soil obtained from the Purdue Agronomy Farm (West Lafayette, IN) was used in the degradation studies. Soil characteristics and all other chemicals used, including the xenobiotic basal medium (XBM) (20), are provided in the Supporting Information. Biotransformation in Closed-Bottle Moist Soil Microcosms. Amber 125 mL serum bottles with crimped seals and silicone/aluminum septa were used for the biodegradation studies. Eight treatments in triplicate included soil amended with 8:2 FTOH dissolved in one of three carrier solvents (ethanol, octanol, and 1,4-dioxane), carrier solvent controls (no FTOH), abiotic controls (sterile soil amended with 8:2 FTOH dissolved in octanol), and soil controls (no FTOH and no carrier solvent). All treatments were conducted at 22.6 °C ( 0.3 in the dark with 10 g of soil (oven-dry mass) wetted to field capacity with 5 µM CaCl2 and preincubated for 7 days. After preincubation, concentrated 8:2 FTOH solutions prepared in ethanol, octanol, or 1,4-dioxane were added into different microcosms with a microliter syringe to give initial soil concentrations of 100 µg of 8:2 FTOH/g and 500 µg of solvent/g, and manually mixed with a sterile spatula. The carrier solvent controls received the same 500 µg of solvent/g but no FTOH. The soil in the abiotic controls was γ-irradiated intermittently three times for a 3.2 Mrad accumulative dose with 1–2 day incubation at field capacity in between irradiations, and then it was amended aseptically with 8:2 FTOH dissolved in octanol. The soil matrix controls were used to establish baseline levels of CO2 evolution and background perfluorinated chemicals. Isolation and Identification of Bacterial Strains. Several attempts were made to isolate 8:2 FTOH-utilizing bacteria from the Chalmers soil using 8:2 FTOH as the sole added carbon substrate; however no such strains were obtained over a 6 month enrichment period. Alternatively, two octanolutilizing bacteria, Pseudomonas spp. OCY4 and OCW4, were isolated with octanol as the sole added carbon source and were identified using 16S rRNA gene sequencing (details provided in the Supporting Information). The BLASTn comparison of the nucleotide sequences with the National Center for Biotechnology Information (21) gave identities of 99.9% between Pseudomonas sp. OCY4 (GenBank accession no. EU143652; 1429 bp sequenced; orange colonies on trypic soy agar) and Pseudomonas aurantiaca (GenBank accession no. DQ682655, ref 22) and identities of 99.8% between Pseudomonas sp. OCW4 (accession no. EU143651;1434 bp sequenced; white colonies on trypic soy agar) and Pseudomonas chlororaphis (accession no. Z76673, ref 23). Biotransformation in Pure Bacterial Cultures. Both open-bottle and closed-bottle setups were employed for the pure culture biotransformation test. Open bottles (500 mL screw-cap bottles containing 200 mL of XBM medium) were used for the examination of the carbon utilization between octanol and 8:2 FTOH with repetitive sampling. Closed bottles (25 mL serum bottle with crimped seal containing 7.9 mL of medium) were used for examination of the mass balance and metabolite distributions by destructive sampling. 8:2 FTOH was aseptically precoated onto the inner walls of sterile bottles (11.2 µmol FTOH in open bottles and 0.42 µmol in

closed bottles) by addition of 8:2 FTOH dissolved in acetone and evaporating off the acetone in a sterile hood. Washed cell suspensions of the two octanol-utilizing bacteria (preparation detailed in the Supporting Information) were used to inoculate all the bottles (except the noninoculated controls) to achieve an initial optical density of 0.01 at 600 nm (OD600) as measured with a cell density meter (Biowave CO8000, WPA, Cambridge, UK). Each setup included a set of bottles that had only basal salts medium XBM (no octanol) and a set in which octanol was added to achieve 1 mM (open bottle) or 0.5 mM (closed bottle) octanol in the liquid XBM. In the open bottle setup, 1 mM octanol was added after 48 h of incubation to all bottles regardless of the initial condition (0 or 1 mM octanol at time zero). In both open and closed setups, five treatments in triplicate were created, including pure culture with only 8:2 FTOH, only octanol, both octanol and FTOH, neither octanol nor FTOH, and a noninoculated control (autoclaved medium plus 8:2 FTOH). Extra replicates in the closed bottle test were prepared for the treatments containing 8:2 FTOH for conducting the whole-bottle solvent extraction (three replicates) and fluoride and biomass measurements (three replicates). The 1:2 liquid to gas volume ratio in the closed bottles ensured aerobic conditions over the 2-month incubation period as confirmed by oxygen sensor measurements (OxyMicro System, World Precision Instruments, Inc., Sarasota, FL). The details of culture sampling and sample preparation are listed in the Supporting Information. All incubations were at 30 °C in the dark for one week (open bottles at 225 rpm) or two months (closed bottles at 150 rpm). LC/MS/MS and Fluoride Analysis. The fluorinated compounds in the soil headspace eluents, soil extracts, and cell culture extracts were analyzed via liquid chromatography tandem mass spectrometry (LC/MS/MS) using the methods described previously (11). Matrix effects in soil extracts were completely removed using a dispersive graphitized carbon adsorbent similar to that described by Powley et al. (24). [1,2-13C]PFOA was used as an internal standard for all the anionic metabolites, and [1D,1D,2D,2D,313C]8:2 FTOH was used as the standard for 8:2 FTOH and 7:2 sFTOH. Fluoride concentrations in the pure culture incubations were assayed with an Orion 96-09BNWP ion-selective electrode (Thermo Electron, Beverly, MA). See Supporting Information for further details on LC/MS/MS analysis, method performance, and fluoride analysis.

Results and Discussion 8:2 FTOH Transformation in Soil Microcosms. O2 and CO2 Levels in Soil Microcosms. All soil microcosms remained aerobic during the 7 d incubation with headspace O2 remaining above 60% atmospheric saturation (Figure 1A). Octanol-treated soils resulted in the greatest O2 reduction, followed by the ethanol-treated soils. The 1,4-dioxane treatment had profiles similar to those for the soil and abiotic controls, implying that oxidation of 1,4-dioxane did not occur, which was verified by 1,4-dioxane analysis in aqueous soil extracts using gas chromatography with flame ionization detection. Headspace CO2 concentrations were consistent with the observed O2 demands (Figure 2B) with greater CO2 evolution corresponding well to lower O2 levels. For octanol and ethanol treatments, CO2 evolution was initially large relative to soil controls, followed by slower increases, similar to those of the soil controls, indicating essentially complete consumption of ethanol by day 2 and of octanol by day 4. Although substantial degradation of 8:2 FTOH did occur (Figure 1C), no significant differences were observed in O2 loss or CO2 evolution (p > 0.05 for both ethanol and octanol treatments at day 7) between the 8:2 FTOH treatments and the no-8:2 FTOH carrier solvent controls (Figure 1A and B). Wang et al. (14) indicated that 14CO2 was produced from VOL. 41, NO. 23, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8025

FIGURE 2. Metabolite distribution at day 7 from the biodegradation of 8:2 FTOH (n ) 3) in soil with ethanol, octanol, or 1,4-dioxane as a carbon source. In the abiotic control (not shown), the metabolites recovered totaled less than 0.02 mol %.

FIGURE 1. Biotransformation of 8:2 FTOH in the presence of octanol, ethanol, or 1,4-dioxane carrier solvents in soil microcosms as exemplified by profiles of (A) O2, (B) CO2, and (C) 8:2 FTOH remaining (including that from soil, bottle septum, and headspace) over time. Error bars are the standard deviation of three replicates in all figures. mineralization of β-14C using 14C-labeled 8:2 FTOH. However, in our nonradiolabeled study, the CO2 evolved from partial FTOH mineralization could not be differentiated from the CO2 evolved from carrier solvent oxidation. CO2 evolution from the abiotic controls (γ-irradiated soils) and the nonsterile soil controls were also similar. The CO2 generated in the abiotic controls may have been a result of radiolytic decarboxlation of soil organic matter or possibly respiration of radioresistant microbes or microbial enzyme residuals (25). If the latter were present, they showed little to no ability to degrade 8:2 FTOH (Figure 1C). Biotransformation of 8:2 FTOH. The fastest biotransformation of 8:2 FTOH occurred when the carrier solvent was 8026

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 23, 2007

1,4-dioxane, followed by ethanol (Figure 1C). With octanol, no observable 8:2 FTOH degradation occurred during the first 2 day, followed by degradation at rates similar in magnitude to what was observed with the other two carriers. Assuming a first-order reaction, estimated 8:2 FTOH transformation rates were 0.28 day-1 with 1,4-dioxane, 0.18 day-1 with ethanol, and 0.13 day-1 with octanol after the 2 day lag period. Differences in rates were all significant at the 95% confidence level (ANOVA tests at R ) 0.05). The onset of 8:2 FTOH degradation between day 2 and 4 coincided with the apparent complete utilization of octanol as evidenced by the tapering off of CO2 evolution within the same time frame (Figure 1B). Loss of 8:2 FTOH in the abiotic control was 12.1% with the metabolites recovered totaling less than 0.02 mol %, relative to the 8:2 FTOH applied. In addition to degradation, loss of 8:2 FTOH over time can include irreversible sorption (10), which may account for the majority of the 8:2 FTOH not recovered in the abiotic control and contribute to an overestimation in the biotransformation rates in the biotic treatments. All metabolites observed in previous studies with activated sludge and mixed bacterial cultures (13, 14) were detected in this study and quantified when authentic standards were available. Metabolites quantified included 7:2 sFTOH, PFOA, PFHeA, PFHA, 8:2 FTCA, 8:2 FTUCA, 7:3 FTCA, and 7:3 FTUCA. PFNA was only observed at or near the method detection limit (MDL, 0.21 ng mL-1). The total moles of polyfluorinated compounds (PFCs) recovered (Table 1) relative to the moles of 8:2 FTOH applied included 8:2 FTOH and all known quantifiable organic metabolites that were extracted from the soil, bottle, and alum septa, and recovered from the headspace via SPE cartridge trapping. At day 7 (Figure 2), 8:2 FTUCA was the dominant metabolite (6.14–8.39 mol %, relative to applied 8:2 FTOH), followed by 8:2 FTCA (2.32–2.88%), and 7:2 sFTOH (2.29–3.46%). All other metabolites monitored were below 1 mol %, including PFOA (0.58–0.78%). All metabolites monitored were observed in all nonsterile soil microcosms, indicating similar biodegradation pathways regardless of the carrier solvent (carbon source). Despite the late onset of 8:2 FTOH degradation with octanol as the carrier, the PFOA and PFHA concentrations in the octanol treatments were not significantly different from the 1,4-dioxane or ethanol treatments (p > 0.05) by day 7. The moles of PFCs recovered decreased with time regardless of the carrier solvent. A strong overall inverse linear relationship between the moles of PFCs recovered and 8:2 FTOH remaining was observed across all times and treatments (Figure 3). Extrapolation of the linear regression to zero 8:2 FTOH remaining yields a PFC recovery of 18 mol %. Given that complete mineralization of 8:2 FTOH is unlikely, the PFCs not recovered likely includes parent compound or

TABLE 1. Mole-Based Recovery of 8:2 FTOH and Metabolites in Closed-Bottle Soil Microcosmsa total mol % ( SEb of polyfluorinated compounds recovered relative to mol of 8:2 FTOH applied day 2 treatment abiotic control ethanol 1,4-dioxane octanol

headspace 11.2 ( 0.5 0.6 ( 0.1 0.6 ( 0.1 0.8 ( 0.1

day 4

soil

total

79.4 ( 2.9 78.6 ( 2.0 59.4 ( 2.5 92.8 ( 3.2

92.4 ( 2.5 79.2 ( 2.1 60.2 ( 2.5 93.7 ( 3.1

headspace 11.5 ( 0.2 0.5 ( 0.0 0.5 ( 0.1 0.6 ( 0.1

day 7

soil

total

77.1 ( 1.5 53.2 ( 4.9 44.6 ( 4.8 75.0 ( 0.3

90.7 ( 1.6 53.7 ( 4.9 45.1 ( 4.7 75.5 ( 2.4

headspace 10.0 ( 0.3 0.7 ( 0.0 0.4 ( 0.0 0.5 ( 0.0

soil

total

75.7 ( 1.1 41.5 ( 1.9 29.2 ( 0.5 59.5 ( 2.2

87.9 ( 1.3 42.3 ( 1.8 29.7 ( 0.5 60.1 ( 2.2

a The total moles recovered included the parent compound and all the metabolites extracted from the soil, alum septum, and recovered from the headspace relative to the moles of 8:2 FTOH applied at day 0. b Mean ( SE (standard error), n ) 3.

FIGURE 3. Relationship between remaining 8:2 FTOH and the total mass recovered, relative to mass of 8:2 FTOH applied at time zero. The total mass recovered included 8:2 FTOH and all the metabolites shown in Figure 2. The solid line presents a linear regression (R 2 ) 0.996) of all the data points. metabolites that have been irreversibly sorbed as well as unmonitored metabolites, such as transient fluorotelomer aldehydes (14). The moles of PFCs recovered from the headspace and septa were also inversely related to microbial activity. The abiotic control had 10 ( 0.3 mol % recovered in the headspace and 2.3 mol % from the septa (all as 8:2 FTOH), whereas the biotic treatments had much lower amounts recovered from the headspace (