Environ. Sci. Technol. 2008, 42, 7786–7792
Carbon and Hydrogen Isotope Fractionation during Anaerobic Toluene Oxidation by Geobacter metallireducens with Different Fe(III) Phases as Terminal Electron Acceptors NICOLE B. TOBLER, THOMAS B. HOFSTETTER,* AND ´ P. SCHWARZENBACH RENE Institute of Biogeochemistry and Pollutant Dynamics (IBP), ETH Zurich, 8092 Zurich, Switzerland
Received January 7, 2008. Revised manuscript received March 17, 2008. Accepted March 27, 2008.
Microbial oxidation of BTEX compounds under iron-reducing conditions is an important attenuation process for fuelcontaminated sites. We evaluated the use of compoundspecific isotope analysis for the identification and quantification of anaerobic toluene oxidation by Geobacter metallireducens. 13C and 2H enrichment of toluene was measured in laboratory batch systems and varied significantly for a solid vs a dissolved Fe(III) phase provided as terminal electron acceptor. 13C enrichment factors (C) in suspensions of a solid Fe(III) phase were between -1.0 and -1.3‰, whereas C-values were significantly higher in solutions of Fe(III) citrate (-2.9 to -3.6‰). The same trend was observed for 2H fractionation. Solid phase reduction resulted in an H-value of -34.6 ( 0.9‰, compared to -98.4 ( 3.0‰ for the reduction of dissolved Fe(III). The linear correlation of δ2H vs δ13C during toluene oxidation resulted in nearly identical slopes for both systems, confirming that the reaction mechanism, that is enzymatic methyl-group oxidation, was the same. We hypothesize that smaller 2H and 13C fractionation in suspensions is due to toluene transport limitations to cells of G. metallireducens at surfaces of solid Fe(III) phases. Enrichment factors determined in Fe(III) mineral suspensions should be more representative for anaerobic toluene degradation owing to the abundance of solid Fe(III) in soils and aquifers.
Introduction Fuel components introduced into the subsurface via leaking underground fuel tanks or pipelines, accidental spills, or landfill leachates represent a major risk for drinking water resources. Among the diverse fuel contaminants, monoaromatic hydrocarbons including benzene, toluene, ethylbenzene, and xylenes (BTEX-compounds) are of particular concern because they are relatively water-soluble, toxic, and can persist in the subsurface for several years (1, 2). Often, fuel-contaminated sites develop extended anaerobic plumes due to high carbon loads and rapid oxygen depletion (3–5). Natural attenuation of BTEX via anaerobic oxidation, a * Corresponding author phone: +41-44-632 83 28; fax: +41-44633 11 22; e-mail:
[email protected]. 7786
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 21, 2008
process which was long thought to be unfeasible (6), has been proposed as alternative to technical remediation of contaminated sites (3, 7). One of the most intensively studied BTEX transformation pathways is the anaerobic oxidation of toluene, which has been investigated with nitrate-, iron-, manganese-, and sulfate-reducing as well as methanogenic enrichment cultures and pure cultures (8–11). As shown in Scheme 1, the proposed enzymatic pathway for toluene activation and subsequent mineralization is initiated by the addition of toluene to fumarate forming benzylsuccinate via a radical mechanism catalyzed by the enzyme benzylsuccinate synthase (bss). Benzylsuccinate then reacts to benzoyl-Coenzyme A, which upon further transformations enters the citric acid cycle (8–10). The enzyme that catalyzes the transformation of toluene to benzylsuccinate was first detected and characterized in nitrate-reducing microorganisms and has since been identified in a range of phylogenetically diverse anaerobic bacteria including the dissimilatory Fe(III)-reducing microorganism Geobacter metallireducens (12). Additionally, the genes encoding for the enzyme benzylsuccinate synthase, which catalyzes the transformation of toluene to benzylsuccinate, have been found in G. metallireducens and the activity of the gene product has been confirmed (12). The use of Fe(III) minerals, which are abundant in the subsurface, as electron acceptors, makes the iron-mediated toluene oxidation an important reaction among natural BTEX attenuation processes.
SCHEME 1
Because transformation processes of BTEX in contaminated aquifers are difficult to monitor on the basis of concentration measurements of the dissolved compounds, it has been suggested to use compound-specific isotope analysis (CSIA) of C and H isotopes to assess biodegradation of BTEX qualitatively and quantitatively (13–15). CSIA allows for the monitoring of stable isotope ratios of BTEX during their microbial transformation. Owing to a normal kinetic isotope effect, such reactions lead to an enrichment of heavier isotopes in the substrate. In contrast, isotope fractionation during physical processes such as phase transfer or adsorption is usually negligible (2, 16). In addition, the observed isotope fractionation behavior can be indicative of specific reaction mechanisms (17). For the proposed anaerobic toluene oxidation pathway (Scheme 1), cleavage of one out of three C-H bonds at the methyl group in the rate-determining step gives rise to moderate 13C isotope effects and more pronounced 2H fractionation. Owing to the presence of nonreactive C and 10.1021/es800046z CCC: $40.75
2008 American Chemical Society
Published on Web 06/11/2008
H atoms and intramolecular isotopic competition (18), however, observable bulk isotope enrichment of toluene is significantly smaller. To date, 13C and 2H fractionation of anaerobic toluene oxidation by dissimilatory Fe(III)-reducing bacteria have mostly been measured with dissolved Fe(III) citrate as terminal electron acceptor (19–21). The determined bulk isotope enrichment factors in these studies were, as expected from the reaction mechanism, rather small for carbon and much more substantial for hydrogen. At nearneutral pH, however, Fe(III) is predominantly present in mineral phases due to the very low solubility of Fe(III). Fe(III) chelating humic substances may elevate dissolved Fe(III) concentrations in sedimentary environments, but it is unclear whether these concentrations are high enough to support bacterial growth and respiration (22). Dissimilatory Fe(III)reducing bacteria, especially Geobacter species, have been shown to need direct contact with a solid Fe(III) phase, whereas these bacteria are free-living in homogeneous solutions of dissolved Fe(III) phases (22–24). Therefore, isotope fractionation during toluene oxidation under ironreducing conditions could vary as a consequence of different rate-limiting reaction steps that depend on the Fe(III) bioavailability. For the quantification of BTEX transformation in anoxic environments, however, knowledge of reliable isotope enrichment factors is essential. It was the goal of this study to explore the influence of solid vs dissolved Fe(III) phases as terminal electron acceptors on the variability of 13C and 2H fractionation during anaerobic toluene oxidation. We evaluated the variability of 13C and 2H enrichment factors for toluene transformation in the presence of either a solid or a dissolved Fe(III) phase. To this end, batch experiments were set up with G. metallireducens, varying concentrations of either an amorphous Fe(III)containing mineral or Fe(III) citrate as electron acceptor, and toluene as electron donor and BTEX representative.
Experimental Section Chemicals and Minerals. Toluene (g99.9%) was purchased from Merck and Fe(III) citrate monohydrate (18-20% Fe) was purchased from Fluka. Fe(III) citrate solutions (0.5 M) were prepared by slow solubilization of the chemical suspended in Millipore water via titration with 32% NaOH (Fluka) to pH 7.0. The obtained solution was filtered (0.45 µm, cellulose acetate), deoxygenated by extended purging with argon, and autoclaved. Amorphous, biogenic Fe(III) mineral (bFe mineral) was generated by a lithotrophic, nitrate-reducing enrichment culture as described in the literature (25–27). Note that suspensions of this mineral contain structural or mineral-bound Fe(II) (between 5 and 15% of total iron) and phosphate (0.25 to 0.40 mol P/mol Fe), because microbial Fe(II) oxidation by this mixed culture is never complete and takes place in freshwater medium containing 4.4 mM phosphate. Microorganisms and Culture Conditions. Geobacter metallireducens was cultivated in freshwater medium, according to ref (28) with the following modifications: 30 mM PIPES buffer (pH 6.8), 0.068 g/L KH2PO4, and 0.35 g/L NaCl (Table S1, Supporting Information (SI)). Cysteine was supplied as reducing agent (1 mM). Cultures were incubated at 30 °C in the dark and briefly shaken every day. As electron donor, pure toluene liquid was added by microsyringe, and as electron acceptor, bFe mineral or Fe(III) citrate was added from anoxic stock solutions. Cultures were routinely grown on 10 mM bFe mineral or Fe(III) citrate and 0.5 mM toluene with 5 vol% inoculum until late exponential phase before transfer to a new culture. The concentration-time courses of biogenic Fe(II) and toluene in these cultures are shown in Figures S1 and S2 in the SI. Anaerobic Microbial Toluene Oxidation. Experiments were set up in a glovebox (100% N2) in 120 mL serum bottles,
to which was added deoxygenated freshwater medium containing no vitamins, no trace elements, and no selenitetungstate components. Pure toluene liquid was provided with a microsyringe, yielding an initial concentration of approximately 500 µM. Subsequently, the serum bottles were sealed with Viton stoppers and crimp caps. Then, batch reactors were taken out of the glovebox and autoclaved. After cooling down to room temperature, vitamins, trace elements, selenite-tungstate components, cysteine, and bFe mineral or Fe(III) citrate were added from anoxic, sterile stock solutions using N2-flushed, sterile syringes. Finally, an inoculum of 5 vol% was added from a late-exponential routinely grown culture. The total solution volume of the reactors was 80 mL. Cultures were incubated at 30 °C in the dark and briefly agitated every day and before each sampling. In one set of experiments, cultures were shaken at 150 rpm. Typically, five or six batch reactors were prepared, of which three or four were inoculated and two served as blanks with sterile, anoxic water added instead of inoculum. Samples for the determination of total Fe(II), toluene concentration, and toluene isotope signature were withdrawn using N2-flushed, sterile syringes. Analytical Methods. Fe(II) was determined by UV/vis spectrometry at 562 nm after complexation with Ferrozine (29). To measure total Fe(II), samples were extracted in 0.5 M HCl for 1 h, followed by a 0.2 µm filtration (regenerated cellulose). Biogenic Fe(II), Fe(II)bio, was calculated by subtraction of the measured total Fe(II) concentration at the start of the experiment from the measured total Fe(II) concentrations during the experiment. Toluene concentrations were measured by GC/MS and δ13C of toluene by GC/ C/IRMS according to the literature (30). To determine δ2H of toluene, the method for δ13C was adapted as follows: Toluene hydrogen atoms were converted to H2 using an untreated high-temperature reactor at 1400 °C. The reduction reactor was set to room temperature and the toluene concentrations in standards and samples were 10-fold higher than in the analysis of carbon isotopes. Measurement uncertainties of δ13C and δ2H were typically below ( 0.5‰ and ( 5.0‰ (RSD), respectively. Quantification of Isotopic Fractionation. Bulk isotopic enrichment factors, E, were obtained from linear regression of eq 1 following procedures described in refs 18 and 31, ln(δhE0 + ∆δhE + 1000) )
(
)
εE (δhE0 + 1000) (1) · ln(c) + ln 1000 c ε0E /1000
where δhE0 and ∆δhE are the initial toluene isotope signature and its change during the reaction, and c and c0 are the substrate concentration at times t and zero, respectively. To obtain a position specific enrichment factor, E, reactive position, for the reacting bond(s) eq 1 was modified as follows: ln(δhE0 + n ⁄ x · ∆δhE + 1000) )
εE, reactive position · ln(c) + 1000 h (δ E0 + 1000) ln c ε0E /1000
(
)
(2)
where n is the number of isotopic atoms of element E and x corresponds to the number of reacting sites. For calculation of C, reactive position according to the proposed reaction pathway for anaerobic toluene oxidation, n/x equals 7/1 (Scheme 1 (8–10), 6 nonreactive, aromatic C atoms), whereas H, reactive position accounts for 3 reactive methyl-group H atoms and 5 nonreactive aromatic positions (n/x ) 8/3). Apparent kinetic isotope effects (AKIEE s) were calculated from E, reactive position using eq 3. For 2H isotope effects, intramolecular isotopic competition of 3 H atoms (z ) 3) was taken into account. AKIEE )
1 1 + z · εE, reactive position ⁄ 1000
VOL. 42, NO. 21, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
(3)
9
7787
FIGURE 1. Concentration-time courses of Fe(II) and toluene in experiments with bacteria (Fe(II)bio) vs blank experiments (Fe(II)). (a) Microbial toluene oxidation with a solid Fe(III) phase as electron acceptor in assays with 20 mM Fe(III) as bFe mineral and 0.45 mM toluene. (b) Microbial toluene oxidation with a dissolved Fe(III) phase as electron acceptor in assays with 20 mM Fe(III) citrate and 0.45 mM toluene. Solid and dashed lines represent mean (1 standard deviation (σ), respectively, for stoichiometric conversion of toluene to Fe(II) by microbial respiration according to eqs 4 and 5. Error bars indicate (1σ of four reactors with bacteria and of two reactors for blank experiments (without bacteria). We calculated E, E,reactive position, and AKIEEs from combined data sets of several replicates using on the one hand linear regression of the combined data set of all replicates and on the other hand a Pitman estimator for the same data according to ref (31). All data presented in the main manuscript are from linear regression analyses. A comparison between results from linear regression analysis and from using the Pitman estimator showed only very minor discrepancies and can be found in the SI.
Results and Discussion Anaerobic Microbial Toluene Oxidation. Microbial toluene oxidation experiments with G. metallireducens were conducted with either amorphous, biogenic Fe(III) mineral (bFe mineral) or dissolved Fe(III) citrate as electron acceptor. Figure 1 illustrates the production of biogenic Fe(II) and consumption of toluene for experiments with 20 mM Fe(III) from bFe mineral or Fe(III) citrate and approximately 0.5 mM toluene as electron donor. With the solid Fe(III) phase as electron acceptor (Figure 1a), Fe(II) production was slower and stopped after 120 h compared to generation of nearly identical final biogenic Fe(II) concentrations in only 50 h with Fe(III) citrate as electron acceptor (Figure 1b). Whereas toluene was completely consumed with Fe(III) citrate, 20 µM toluene, 4% of its initial concentration, remained in bFe mineral grown cultures despite the still remaining ca. 4.5 mM Fe(III). This effect is most likely due to phase transformations of bFe mineral to less bioavailable Fe(III/II) phases (32–34), reflected in a color change of the suspension from an initial red-brown to a dark brown during microbial reduction. In blank experiments without bacteria, a very small increase of Fe(II) owing to abiotic Fe(III) reduction by cysteine and a slight decrease of toluene due to mass withdrawal by sampling, headspace partitioning, and/or sorption into Viton stoppers were observed comparable to previously published results (27, 30). According to the stoichiometry of toluene oxidation with an Fe(III) phase as terminal electron acceptor (see eqs 4 and 5), 36 mol of Fe(II) are produced from 1 mol of toluene, if 7788
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 21, 2008
toluene is exclusively used as electron donor for microbial respiration and not as carbon source. The solid and dashed lines in Figure 1 indicate mean (1 standard deviation for the stoichiometric oxidation of toluene and reduction of Fe(III) to Fe(II). Stoichiometric Fe(II) equivalents were calculated using blank-corrected toluene concentrations at the start of the stationary phase. With both electron acceptors, approximately 92% (Fe(III) citrate) to 101% (bFe mineral) of the electrons resulting from toluene oxidation were recovered as Fe(II). The missing 8% of toluene in the case of Fe(III) citrate was presumably used as carbon source for biosynthesis. bFe Mineral: 2+ + 87 H2O C7H8 + 36 Fe(OH)3 + 65 H+ f 7 HCO3 + 36 Fe (4) Fe(III) Citrate: 2+ C7H8 + 36 Fe3+ + 21 H2O f 7 HCO+ 43 H+ (5) 3 + 36 Fe
Note that increasing the electron acceptor to donor ratio, that is, the ratio of Fe(III) to toluene concentration, resulted in complete toluene oxidation and near stoichiometric formation of Fe(II) in both experimental systems (Figure S3a and b). Yields of biogenic Fe(II) from microbial toluene oxidation (between 90% and 95%) vs the use of toluene as carbon source were comparable to the experiments with 20 mM Fe(III). Carbon and Hydrogen Isotope Fractionation during Anaerobic Microbial Toluene Oxidation. δ13C and δ2H values of toluene for up to 90% toluene conversion were determined with initial 20 mM Fe(III) from bFe mineral or from Fe(III) citrate as terminal electron acceptor. As can be seen in Figure 2, we observed a significant 13C and 2H enrichment with time, which, as expected, was much smaller for carbon than for hydrogen. Figure 2 also demonstrates that 13C and 2H enrichments observed during bFe mineral reduction were substantially smaller compared to the ones found during reduction of dissolved Fe(III) citrate. As shown in Table 1
FIGURE 2. Time courses of carbon and hydrogen isotope enrichment during anaerobic microbial toluene oxidation. (a) Isotopic enrichment observed during bFe mineral reduction. Error bars indicate (1 standard deviation (σ) of triplicate isotope measurements. (b) Isotopic enrichment observed during Fe(III) citrate reduction. Error bars indicate (1σ for triplicate isotope measurements.
TABLE 1. Carbon and Hydrogen Isotope Enrichment Factors, EC and EH, Respectively, and Apparent C and H Kinetic Isotope Effects, AKIEC and AKIEH, Respectively, of Anaerobic Microbial Toluene Oxidation Experiments with Either bFe Mineral or Fe(III) Citrate As Terminal Electron Acceptor Fe(III) phase
Fe(III) [mM]
inoculum [vol%]
shaken at 150 rpm
εC a,b [‰]
bFe mineral bFe mineral bFe mineral bFe mineral bFe mineral/ goethite Fe(III) citrate Fe(III) citrate Fe(III) citrate Fe(III) citrate
20 20 20 50 8/68
5 5 5 5 5
no yes no no no
-1.3 ( 0.1 -1.3 ( 0.1 -1.0 ( (
