APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Mar. 1992, p. 786-793

Vol. 58, No. 3

0099-2240/92/030786-08$02.00/0 Copyright 1992, American Society for Microbiology
Microbial Degradation of Toluene under Sulfate-Reducing Conditions and the Influence of Iron on the Process
HARRY R. BELLER,* DUNJA GRBIC-GALIC, AND MARTIN REINHARD Environmental Engineering and Science, Department of Civil Engineering, Stanford University, Stanford, California 94305-4020
Received 9 August 1991/Accepted 17 December 1991
Toluene degradation occurred concomitantly with sulfate reduction in anaerobic microcosms inoculated with contaminated subsurface soil from an aviation fuel storage facility near the Patuxent River (Md.). Similar results were obtained for enrichment cultures in which toluene was the sole carbon source. Several lines of evidence suggest that toluene degradation was directly coupled to sulfate reduction in Patuxent River microcosms and enrichment cultures: (i) the two processes were synchronous and highly correlated, (ii) the observed stoichiometric ratios of moles of sulfate consumed per mole of toluene consumed were consistent with the theoretical ratio for the oxidation of toluene to CO2 coupled with the reduction of sulfate to hydrogen sulfide, and (iii) toluene degradation ceased when sulfate was depleted, and conversely, sulfate reduction ceased when toluene was depleted. Mineralization of toluene was confirmed in experiments with [ring-U-_4Cltoluene. The addition of millimolar concentrations of amorphous Fe(OH)3 to Patuxent River microcosms and enrichment cultures either greatly facilitated the onset of toluene degradation or accelerated the rate once degradation had begun. In iron-amended microcosms and enrichment cultures, ferric iron reduction proceeded concurrently with toluene degradation and sulfate reduction. Stoichiometric data and other observations indicate that ferric iron reduction was not directly coupled to toluene oxidation but was a secondary, presumably abiotic, reaction between ferric iron and biogenic hydrogen sulfide.
In 1986, the U.S. Environmental Protection Agency (EPA) estimated that 35% of the nation's underground motor fuel storage tanks were leaking (30). Together with surface spill accidents and landfill leachate intrusion, such leaks contribute significantly to groundwater contamination by gasoline, aviation fuel, and other refined petroleum derivatives. Benzene and various alkylbenzenes (such as toluene) are relatively water-soluble constituents of gasoline and aviation fuel that present human health concerns. This article focuses on toluene, an EPA priority pollutant that is a depressant of the central nervous system (27). Toluene is less toxic than benzene (a confirmed carcinogen) and has an EPA water quality criterion for the protection of human health of 14.3 mg/liter or 0.155 mM (27). Toluene and other alkylbenzenes are readily degradable in aerobic surface water and soil systems; however, in the subsurface environment, contamination by organic compounds often results in the complete consumption of available oxygen by indigenous microorganisms and the development of anaerobic conditions. In the absence of oxygen, degradation of toluene can take place only with the use of alternative electron acceptors, such as nitrate, sulfate, or ferric iron, or fermentatively in combination with methanogenesis. To date, degradation of toluene by aquifer-, sediment-, or sewage-derived microorganisms has been found under denitrifying conditions (4, 6, 10, 11, 35, 36), fermentative and methanogenic conditions (8, 31, 33, 34), and ferric iron-reducing conditions (14, 15). Despite recent research advances in the degradation of toluene under various anaerobic conditions, we are aware of very few reports of toluene degradation under sulfate-reducing conditions. Preliminary evidence of the degradation of

Corresponding author.

alkylbenzenes under sulfate-reducing conditions has been reported in studies of fuel-contaminated aquifer material from Seal Beach, California (9), and in contaminated subsurface materials collected in Oklahoma (26). More conclusive evidence of sulfidogenic toluene degradation at the Seal Beach site is reported in this issue by Edwards et al. (5). Despite the limited study of the process, toluene degradation under sulfate-reducing conditions could be relevant in subsurface and waterlogged soil environments for the following reasons: (i) toluene oxidation coupled with sulfate reduction is an exergonic reaction that could yield energy to bacteria (although other anaerobic electron acceptors, such as ferric iron and nitrate, would yield considerably more energy than sulfate); (ii) sulfate is present at millimolar concentrations in some subsurface environments (e.g., near landfills [12, 21, 28] and adjacent to coastal and estuarine areas); and (iii) various pure cultures of sulfate-reducing bacteria are capable of degrading a number of oxygencontaining aromatic compounds (e.g., 1, 3, 25, 29, 32), some of which are proposed intermediates in anaerobic toluene degradation (e.g., p-cresol, benzoate, and 2- and 4-hydroxybenzoate [8, 11, 15]). This article presents evidence that microflora enriched from fuel-contaminated soil can degrade toluene under sulfate-reducing conditions. In addition, the stimulation of toluene degradation by amorphous ferric oxyhydroxide [amorphous Fe(OH)3] will be discussed. The microbially mediated and abiotic interactions among toluene, sulfate, and ferric iron described in this article indicate that geochemical site conditions (e.g., oxidation-reduction potential, the presence of sulfate, the presence of iron-containing minerals) are worthy of consideration when assessing the potential for in situ biological restoration of aquifers contaminated with refined petroleum products, such as gasoline and aviation fuel.
SULFIDOGENIC TOLUENE DEGRADATION VOL. 58, 1992 VOL. 58, 1992

MATERIALS AND METHODS

Construction of microcosms and enrichment cultures. Miand enrichment cultures were prepared under strictly anaerobic conditions in an anaerobic glove box (Coy Laboratory Products, Inc., Ann Arbor, Mich.). The microcosms and enrichment cultures were contained in amber glass, 250-ml, screw-cap bottles that were sealed with Mininert polytetrafluoroethylene (PTFE) valves (Alltech Associates, Inc, Deerfield, Ill.); the Mininert valves provided a tight seal for the bottles while allowing sampling of the headspace and culture medium via syringe. The combined volume of medium and 30 g of wet solids (in microcosms) or medium and culture inoculum (in enrichments) was 200 ml. The remaining volume of the bottles was headspace. Five aromatic hydrocarbons (benzene, toluene, ethylbenzene, and o- and p-xylenes) were initially spiked into bottles as pure liquids at concentrations in the range of 50 to 100 puM per compound. Xylenes were excluded from later studies. Sterile controls, which contained sediment or culture inoculum, were removed from the glove box after being sealed and were autoclaved at 121C. Replicate microcosms and controls were used routinely. After degradation had begun, regular respiking with toluene (typically 100 to 250 puM per week) and sulfate was performed as necessary. Incubation was carried out at 35C in an anaerobic glove box. (i) Growth medium. The composition of the growth medium used for microcosms and enrichment cultures was based on medium described by Lovley and Phillips (16). This medium included the following compounds at the concentrations (millimolar) specified in parentheses: NaHCO3 (30), NH4Cl (28), NaH2PO4- H20 (4.4), NaCl (1.7), KCl (1.3), CaCI2. 2H20 (0.68), MgCl2. 6H20 (0.49), MgSO4. 7H20 (0.41), MnCl2 4H20 (0.025), and Na2MoO4. 2H20 (0.004). A higher initial sulfate concentration (2 to 3 mM) was used in later experiments. The medium (excluding NaHCO3) was sterilized in an autoclave at 121C for 20 min and then purged with an oxygen-free mixture of 79% N2-21% CO2 for 45 min. The medium was amended with NaHCO3 immediately after being purged and was then flushed in the antechamber of an anaerobic glove box. The final pH of the medium was approximately 6.9. (ii) Amorphous Fe(OH)3. The initial goal of the research described in this article was to enrich ferric iron-reducing microbial communities that could degrade monoaromatic hydrocarbons. In order to enrich iron-reducing bacteria, soils were screened for hydrocarbon-degrading activity by using a basal mineral medium that either was amended with ferric iron or was not amended with significant concentrations of any potential electron acceptor. The phase of iron added to microcosms and enrichment cultures was Fe(OH)3. Amorphous Fe(OH)3 was originally chosen rather than more crystalline iron phases because the rate of biological ferric iron reduction tends to increase significantly with decreasing degree of crystallinity (13). This iron phase was prepared by neutralizing a 0.1 M ferric chloride solution with sodium hydroxide. The precipitate was aged for 4 h after neutralization; during this period, the pH was adjusted periodically with sodium hydroxide to neutralize acidification resulting from ferric iron hydrolysis. The precipitate was then rinsed with Milli-Q water in a 2-liter Buchner funnel to reduce the residual chloride concentration to a level that would correspond to less than 1 mM in a 200-ml microcosm or enrichment culture. The amorphous Fe(OH)3 was prepared with sterilized glassware and reagents prepared in sterilized Milli-Q water, but the iron phase itself could not be autocrocosms

claved because the elevated heat and pressure would facilitate crystallization, which was not desired. (iii) Soil inoculum. Patuxent River soil was collected at the Naval Air Station, Patuxent River, Md., in September 1987. The site was extensively contaminated with aviation fuel (e.g., JP-5). The sample was collected from a Quaternary stratum of an outcrop through which hydrocarbon-contaminated groundwater was seeping. Exposed soil was removed before sampling. The soil was received in water-saturated form and stored at 4C until its use in microcosms roughly 2 years after collection. The homogenized soil slurry used for inoculation contained approximately 7 ,umol of sulfate per g. Analytical methods. (i) Aromatic hydrocarbon analysis. Aromatic substrates were measured by a static headspace technique with an HP model 5890A gas chromatograph (Hewlett-Packard Company, Palo Alto, Calif.) with an HNU model PI 52-02A photoionization detector (10.2 eV lamp; HNU Systems, Inc., Newton, Mass.) and a DB-624 fused silica capillary column (30 m length by 0.53 mm inside diameter, 3.0-,um film thickness; J & W Scientific, Folsom, Calif.). Analyses were isothermal (65C) with splitless injection (the split was turned on after 0.5 min). Sampling and analysis of headspace from microcosms, enrichment cultures, and standards were performed identically: 300 ,ul of headspace was sampled through the Mininert valve of microcosms, enrichment cultures, or standards with 500-pu gas-tight syringe that included a PTFE plunger tip and a side-port needle. The aqueous concentration of aromatic compounds in microcosms and enrichment cultures was determined by comparing peak areas in samples with those of external standards. Standards for headspace analyses were prepared by spiking a methanolic stock solution of the aromatic compounds into a Mininert-sealed bottle that contained 200 ml of water. The transfer of the methanolic stock solution was made with a gas-tight 500-ptl syringe. The amount of stock solution (and correspondingly, the mass of each compound) added to the standard bottle was determined gravimetrically by weighing the syringe immediately before and after spiking. The aqueous concentrations of aromatic compounds in standards were calculated by using Henry's Law constants (25C) obtained from the literature (18), which were empirically corrected for the incubation temperature of 35C, and a mass balance expression. A potential source of systematic error in the headspace analyses was the presence of methanol in the standards. However, it was assumed that the presence of methanol at the applicable methanol concentrations would not significantly affect the gas-liquid partitioning of the compounds in the standards; this assumption was based on detailed studies of Henry's Law constants of other volatile organic compounds that entailed spiking with methanolic stock solutions (7). Another potential source of systematic error was the sorption of toluene to solids. Controlled sorption experiments were not performed to quantify this effect. However, the sorption of toluene would be of concern only for microcosms; in enrichment cultures, the solid/liquid (mass/mass) ratio was on the order of 0.03 or less and could be expected to result in a very small mass of sorbed toluene relative to the mass of dissolved toluene. Mass balances indicated that the effect of all observed losses of toluene (including sorption to solids) on aqueous concentrations in enrichment cultures would be within the range of analytical error (i.e., within 10%). Thus, quantitative data (e.g., stoichiometric ratios) in this article focus on enrichment cultures rather than microcosms.

BELLER ET AL.

APPL. ENVIRON. MICROBIOL.
(ii) Fe(II) analysis. HCl-soluble Fe(II) was measured as described by Lovley and Phillips (17). Briefly, this method involved the dissolution of 0.1 to 0.2 g of culture medium in 0.5 M HCI, the use of a spectrophotometric reagent for ferrous iron (ferrozine), and measurement of theA.62 with an HP model 8451A diode array spectrophotometer (HewlettPackard Company). An advantage of first dissolving the culture medium in HCI was that Fe(II) in poorly crystalline precipitates (such as amorphous ferrous sulfide) could be determined in addition to soluble Fe(II), which was important because precipitates were the predominant Fe(II) phases in the systems studied. Standards were prepared by dissolving a known amount of ammonium iron(II) sulfate hexahydrate (stored in the dark in an anaerobic glove box) in 0.5 M HCI. (iii) Sulfate analysis. Sulfate in filtered culture medium was determined by ion chromatography (Dionex Series 4000i with a Nelson Analytical Chromatography Software system) equipped with an HPIC-AS4A column (Dionex Corporation, Sunnyvale, Calif.), an anion micro membrane suppressor, and a conductivity detector. Analyses were isocratic, with a 0.75 mM sodium bicarbonate-2.2 mM sodium carbonate eluant flowing at a rate of 2 ml/min. Ions were identified and quantified by comparing retention times and peak areas, respectively, with those of external standards. (iv) 14C analysis. 14C-labeled toluene was fed to selected enrichment cultures to investigate the extent of toluene mineralization. For spiking into samples, [ring-U-14C]toluene (>98% purity; 10.2 mCi/mmol; Sigma Chemical Co., St. Louis, Mo.) was diluted into unlabeled toluene (>99.9% purity; Aldrich Chemical Co., Inc., Milwaukee, Wis.) to a final specific activity of 67.5 ,uCi/mmol. 14C activity in culture liquid and headspace samples was analyzed with a Tri-Carb model 4530 liquid scintillation system (Packard Instruments Co., Inc., Downers Grove, Ill.). All samples were automatically blank-corrected and were corrected for sample-specific quenching by using an external standard method (with a 226Ra gamma source) and a quenching curve developed from a series of quenched standards. Sample processing methods were similar to those described by Grbi6-Galic and Vogel (8). Three 1-ml samples of culture liquid were collected and subjected to different treatments. One 1-ml sample of culture liquid was collected with a 1-ml gas-tight syringe and was injected into a vial containing 10 ml of Universol scintillation fluid (ICN Biomedicals, Inc., Irvine, Calif.) and 0.6 ml of 1 N NaOH; this sample represented the sum of '4C activity from all possible species in the culture liquid (i.e., toluene, volatile intermediates, C02, nonvolatile intermediates, and biomass). Another 1-ml sample was injected into 0.6 ml of 1 N NaOH and was purged with N2 for 15 min, after which 10 ml of scintillation fluid was added. This sample represented the sum of 14C activity of CO2, nonvolatile intermediates, and biomass. A third 1-ml sample was injected into 0.6 ml of 1 N HCl, purged with N2 for 15 min, and mixed with 10 ml of scintillation fluid. This fraction represented the 14C activity of nonvolatile intermediates and biomass. Three 1-ml samples of headspace were treated analogously, except that 11 ml of scintillation fluid was used rather than 10 ml. 14CO2 was calculated as the difference between the 14C activities of the purged NaOH-treated and purged HCl-treated samples. Volatile 14C (other than 14C02) was calculated as the difference between the unpurged NaOH-treated and purged NaOH-treated samples. Experimental design. The two microcosm experiments performed with Patuxent River material (experiments Ml

and M2) were originally attempts to screen for activity of iron-reducing bacteria. Briefly, the experimental design consisted of setting up parallel series of microcosms, one with and one without added amorphous Fe(OH)3. The initial concentration of iron differed in the two experiments: the first experiment (Ml) involved ca. 12 mM Fe(III), and the second experiment (M2) involved ca. 20 mM Fe(III). Enrichments of certain iron-amended microcosms were prepared in an anaerobic glove box by shaking the microcosm, removing 20% (by volume) of the combined liquid and solids, adding this inoculum to a new bottle, and diluting to 200 ml with fresh medium. Series of enrichment cultures were prepared after 88 and 235 days of incubation of selected microcosms from experiment M2; these enrichment series will be referred to as experiments El and E2, respectively. Secondary enrichments (experiment E3) were prepared after 119 days of incubation of an enrichment culture from experiment E2. Iron was added over time to certain enrichment cultures at concentrations of ca. 2 to 4 mM per addition. Xylenes were excluded from experiments E2 and E3.
RESULTS AND DISCUSSION Microcosms containing Patuxent River material and enrichment cultures of these microcosms yielded analogous results. The following trends were apparent in both microcosms and enrichment cultures: (i) toluene degradation, sulfate reduction, and ferric iron reduction (in systems amended with ferric iron) proceeded concomitantly at rates that were highly correlated with each other; (ii) the presence of amorphous Fe(OH)3 enhanced toluene degradation; (iii) degradation of other aromatic hydrocarbons added to the systems (benzene and ethylbenzene) was not observed; (iv) toluene degradation, sulfate reduction, and ferric iron reduction were not observed in sterile controls. In this section, data presented for microcosms will focus on the stimulation of toluene degradation by iron, and data for enrichment cultures will be used to explore quantitative relationships between toluene degradation, sulfate reduction, and ferric iron reduction. The quantitative discussion will focus on enrichment cultures rather than microcosms for two reasons: to minimize the potentially confounding effects of any electron donors or acceptors present in the native soil, and to minimize the potentially confounding sorptive effects of solids in the system. Microcosms: toluene degradation and its apparent stimulation by iron. The results of the first microcosm experiment with Patuxent River material (experiment Ml) are shown in Fig. 1 in terms of aqueous toluene concentration versus time over the first 2 months of incubation. The data in Fig. 1 represent averages for three groups of microcosms: (i) two autoclaved controls (one with added iron and one without), (ii) three microcosms without added amorphous Fe(OH)3, and (iii) four microcosms with added amorphous Fe(OH)3 [ca. 12 mM Fe(III) on day 0]. No toluene degradation was observed in the controls. Although rates of toluene degradation for microcosms with and without added iron were very similar for the first 3 to 4 weeks of incubation, the microcosms with added iron had faster toluene degradation rates starting on day 30. These differences in rates continued throughout the next 30 days of incubation. Note that the microcosms with added iron received additional toluene on day 44, whereas those without added iron did not. The effect of the presence of iron on toluene degradation rate was reproducible in this study. As an indication of the variability among replicates, the average amount of toluene that had

VOL. 58, 1992

SULFIDOGENIC TOLUENE DEGRADATION
-*-*@** Cl (control) -*-*"-*- C2 (control)

tollluene

CS (without iron)

C6 (without iron)

e C8 (with iron) O- C9 (with iron) ,A CIO (with iron)

.--It:---~~~.

Fe(OH)3 to C9,C1O Fe(OH)3 to C8

_-----A~~~

Time (days) FIG. 2. Toluene versus time in individual Patuxent River microcosms (experiment M2). Solid arrows represent amendments of ca. 10 mM Fe(III). "With iron" represents microcosms with ca. 20 mM Fe(OH)3 added on day 0; "without iron" represents microcosms that were not amended with Fe(OH)3.

Time (days)

FIG. 1. Average toluene concentrations versus time in Patuxent River microcosms (experiment Ml). The number of replicates is indicated parenthetically, and reproducibility is discussed in the text. "With iron" represents replicates with ca. 12 mM Fe(OH)3 added on day 0; "without iron" represents replicates that were not amended with Fe(OH)3. Arrows indicate times of toluene addition.
been degraded by day 37 in microcosms with added iron (0.248 + 0.0125 mM, mean + standard deviation [SD]) was significantly greater (P < 0.001) than the average among replicates without added iron (0.164 + 0.0156 mM). Such significant differences continued after day 37. Enrichment cultures also showed stimulation of toluene degradation in the presence of iron, although the effect was more rapid (within days of iron addition) in enrichment cultures than in experiment Ml microcosms. Both ferric and ferrous iron were found to stimulate toluene degradation in enrichment cultures (data not shown). In the second microcosm experiment (experiment M2), as in the first, the presence of amorphous Fe(OH)3 had an effect on toluene degradation, but the effect was qualitatively different. Toluene concentration versus time is shown in Fig. 2 for the first 47 days of incubation of five microcosms and two autoclaved controls (one with added iron and one without). Toluene was not degraded in any microcosms during the first month of incubation. However, the addition of ca. 10 mM amorphous Fe(OH)3 to the three microcosms that initially contained added iron (microcosms C8, C9, and C10) initiated toluene degradation within a few days in each of those microcosms. Two microcosms that did not receive amorphous Fe(OH)3 over the period shown (microcosms C5 and C6) did not degrade toluene until roughly 40 days later than the iron-amended microcosms. Toluene degradation was not observed in the controls. Toluene degradation and sulfate reduction were strongly correlated in this microcosm experiment. For the three iron-amended microcosms shown in Fig. 2, regressions of cumulative sulfate reduction versus cumulative toluene degradation yielded r2 values of >0.97 (calculated over a 1.5-month period following the initiation of toluene degradation; n = 6). Enrichment cultures: interrelationships among toluene deg-

radation, sulfate reduction, and iron reduction. In Patuxent River microcosms and enrichment cultures, toluene degradation, sulfate reduction, and ferric iron reduction appeared to be strongly linked. An example of these relationships in a Patuxent River enrichment culture (experiment El) is shown in Fig. 3, in which the cumulative appearance of Fe(II) (an indication of ferric iron reduction) is shown along with cumulative sulfate reduction and cumulative toluene degradation. As shown in the figure, not only did these processes occur simultaneously, but their relative rates were approximately constant. An autoclaved control prepared identically to and incubated concurrently with the enrichment culture shown had C/Co (final/initial concentration) values for toluene, sulfate, and Fe(II) of 0.99, 0.97, and 1.08, respectively, after approximately 3 months of incubation, indicating a lack of discernible toluene degradation, sulfate reduction, and iron reduction in a sterile system. Plots similar to Fig. 3 were obtained for microcosms. The linkage between toluene deg-

0.75 -

4)0 P.

9Q> _>

FIG. 3. Cumulative Fe(I1) appearance and toluene and sulfate disappearance in an enrichment culture from experiment El. The regression equation relating toluene and sulfate consumption for this enrichment (days 0 to 84) was: sulfate consumed (mM) = [3.79 x toluene consumed (mM)] + 0.06 (r2 > 0.999; n = 11).
TABLE 1. Selected biotic and abiotic reactions involving toluene, iron, and various sulfur species

Equation

Description
(1) C7H8 + 4.5 s042- + 3 H20 = 2.25 H2S + 2.25 HS- + 7 HCO3- + 0.25 H+ (AGO' = -205 kJ/reaction)
(2) C7H8 + 4.09 s042- + 0.17 NH4+ + 2.49 H20 = 2.04 H2S + 2.04 HS- + 0.17 C5H702N (cells) + 6.16 HCO3- + 0.2 H+
Biotic toluene oxidation with sulfate (no cell growth) Biotic toluene oxidation with sulfate (cell growth) Biotic toluene oxidation with amorphous Fe(OH)3 (no cell growth; magnetite formation)
(3) C7H8 + 108 Fe(OH)3 (s) = 36 Fe3O4 (s) + 7 HC03- + 159 H20 + 7 H+ (AGO' = -3,174 kJ/reaction)
(4) 4 Fe(OH)3 (s) + 3 H2S + 3 HS- = 2 S0 (s) + 4 FeS (s) + 9 H20 + 3 OH(5) 4 Fe(OH)3 (s) + H2S + HS- = 2 S (s) + 4 Fe2+ + 3 H20 + 9 OH-
Abiotic oxidation of sulfide to S by amorphous Fe(OH)3; FeS(s) formation
Abiotic oxidation of sulfide to S" by amorphous Fe(OH)3; no FeS(s) formation Abiotic oxidation of sulfide to S0 by amorphous Fe(OH)3; limited FeS(s) formation; FeCO3(s) formation Toluene oxidation with sulfate; formation of S (s) and FeS (s) Toluene oxidation with sulfate; formation of S0 (s) but not FeS (s) Toluene oxidation with sulfate; formation of S0 (s), FeS (s), and FeCO3 (s)
(6) 4 Fe(OH)3 (s) + 1.28 H2S + 1.28 HS- + 3.48 HCO3- = 2.0 S (s) + 0.52 FeS (s) + 3.48 FeCO3 (s) + 7.28 H20 + 4.72 OH-

(7)a C7H8 + 2.72 Fe(OH)3 (s) + 4.09 S042- + 0.17 NH4+ = 1.36 S0 (s) + 2.73 FeS (s) + 6.16 HC03- + 0.17 CSH702N + 3.84 H20 + 1.84 OH(8)b C7H8 + 8.18 Fe(OH)3 (s) + 4.09 s042- + 0.17 NH4+ = 4.09 S0 (s) + 8.18 Fe2+ + 6.16 HC03- + 0.17 C5H702N + 3.85 H2O + 18.2 OH-

(9)C C7H8

+ 6.39 Fe(OH)3 (s) + 4.09 S042- + 0.17 NH4+ = 3.2 S0 (s) + 0.83 FeS (s) + 5.56 FeCO3 (s) + 0.6 HCO3- + 0.17 C5H702N + 9.31 H20 + 7.34 OH-
b Combination of equations 2 and 5. c Combination of equations 2 and 6.
a Combination of equations 2 and 4.
radation, sulfate reduction, and ferric iron reduction raises the question of which compound, sulfate or ferric oxyhydroxide, was serving as the terminal electron acceptor for toluene degradation. Several lines of evidence suggest that toluene degradation was directly linked to sulfate reduction in microcosms and enrichments: (i) the two processes were synchronous, (ii) the observed stoichiometric ratios of sulfate consumed to toluene consumed were consistent with the theoretical ratio for the complete oxidation of toluene to bicarbonate coupled with the reduction of sulfate to hydrogen sulfide, and (iii) toluene degradation ceased when sulfate was depleted, and conversely, sulfate reduction ceased when toluene was depleted. The synchronism of toluene degradation and sulfate reduction is apparent in Fig. 3 and is supported by the very strong r2 values for regressions of cumulative toluene degradation versus cumulative sulfate reduction over time (typically, r2 > 0.98 for enrichment cultures). The regression equation for the enrichment shown in Fig. 3 had an r2 value of >0.999 and a slope of 3.79 (see Fig. 3 legend). Thus, the ratio of sulfate consumed to toluene consumed for the enrichment culture shown in Fig. 3 was ca. 3.8. The ratio in enrichment cultures overall averaged 4.07 0.26 (mean SD) and ranged from 3.6 to 4.4 (calculated for 11 enrichment cultures over intervals of at least 20 days of incubation). These values observed in enrichment cultures approximate the theoretical ratios, ranging from 4.5 (toluene oxidation to bicarbonate with no bacterial cell growth; equation 1 in Table 1) to 4.1 (toluene oxidation to bicarbonate with estimated cell growth; equation 2 in Table 1); the estimation

' 3.%0

%"''%

Control

Enrichment
FIG. 4. 14C activity in a control (autoclaved medium only) and secondary Patuxent River enrichment culture spiked with ca. 37 ,umol (5.59 x 106 dpm) of [nng U-'4C]toluene. '4C02 constituted 83.6% of the 4C activity in the enrichment culture on day 8, whereas nonvolatile 4C (presumably biomass) constituted 11.4%. These results agree well with equation 2 (Table 1), which estimates that 88% of the carbon in toluene will be mineralized and the remaining 12% will be converted to biomass. The total '4C activity in the enrichment culture on day 8 was 95.6% of that on day 0. The sulfate consumed/toluene consumed (molar) ratio over this period was 3.8 (assuming complete toluene consumption, as suggested by gas chromatography analysis). Toluene was unaltered in the control.

FIG. 5. Mutual dependence of toluene degradation and sulfate reduction for a Patuxent River enrichment culture (experiment E2). Arrows indicate additions of toluene or sulfate. Datum points represent the averages of duplicate measurements.
of cell growth in equation 2 was derived by using the calculation methods described by McCarty (19, 20). The observed values of the sulfate consumed/toluene consumed ratio and the consistency of the ratio over months of monitoring provided preliminary evidence that toluene was oxidized to bicarbonate by sulfate reducers. Experiments in which [ring-U-14C]toluene was fed to enrichment cultures confirmed that oxidation to bicarbonate was occurring in accordance with equation 2 (shown for a secondary enrichment culture in Fig. 4). Further evidence of the link between toluene degradation and sulfate reduction was the dependence of toluene degradation on the presence of sulfate and, conversely, the dependence of sulfate reduction on the presence of toluene (shown for an enrichment culture from experiment E2 in Fig. 5). When sulfate was depleted (days 0 to 3 of this experiment), toluene was not appreciably degraded. After sulfate was spiked into the enrichment culture (day 3), concurrent toluene degradation and sulfate reduction occurred until day 7, at which time toluene was depleted. In the absence of toluene (days 7 to 10), sulfate reduction was not apparent. If toluene oxidation and sulfate reduction were directly linked, the concurrent process of ferric iron reduction requires explanation (e.g., identification of the electron doTABLE 2. Theoretical and observed stoichiometric ratios relating toluene degradation, ferric iron reduction, and sulfate reduction

Type and equation no.'

Fe(III) reduced/

toluene consumed

sulfate consumed

Theoretical 8 9

36 2.7 8.2 6.4

0.67 2.0 1.6

Observed"
See Table 1 for equations. ^ Observed ratios were calculated by using the enrichment culture shown in Fig. 3.
nor). Two possible explanations for ferric iron reduction are (i) direct coupling with toluene oxidation (where ferric iron could have served as a terminal electron acceptor for toluene) and (ii) coupling with sulfate reduction (where ferric iron could have served as the electron acceptor for the abiotic oxidation of biogenic hydrogen sulfide). Stoichiometric ratios were used to examine the likelihood of these two explanations, although data for other variables would be required to reach definitive conclusions. The ratio of Fe(III) reduced to toluene consumed was used to investigate the possibility that ferric iron-reducing bacteria were oxidizing toluene. The observed ratios were considerably lower than the theoretical ratio of 36 expected for toluene oxidation to bicarbonate coupled to ferric iron reduction (e.g., equation 3 in Table 1, as reported by Lovley and Lonergan [15]). For example, the observed ratio for the enrichment shown in Fig. 3 was 6.1 rather than 36 (Table 2), which suggests that not nearly enough ferric iron had been reduced to account for complete toluene oxidation. Thus, the stoichiometry suggests either that ferric iron reducers were not involved in toluene oxidation or that they carried out incomplete oxidation. Indeed, the ratio of sulfate consumed to toluene consumed (discussed earlier) further suggests that ferric iron reducers played little, if any, role in toluene degradation. Moreover, in experiments E2 and E3, rapid toluene degradation over months of incubation with no initial lag period was observed in enrichment cultures that were amended with ferrous iron rather than ferric iron (data not shown). If ferric iron reduction was not directly coupled to toluene degradation, it may have resulted from an oxidation-reduction reaction between ferric iron and biogenic hydrogen sulfide. Geochemical studies of the abiotic, anoxic or oxygen-limited reactions of goethite or amorphous Fe(OH)3 with hydrogen sulfide (2, 22, 24) have demonstrated the relatively rapid formation of ferrous sulfide and elemental sulfur as major products and, in Pyzik and Sommer (22), thiosulfate as an additional product. In accordance with these observations, possible abiotic reactions between ferric iron and hydrogen sulfide that result in elemental sulfur and ferrous sulfide formation are shown in Table 1 (equations 4, 5, and 6). Equations 4 to 6 have the same sulfide oxidation product (i.e., elemental sulfur) but different proportions of available sulfide relative to iron. In equation 4, there is sufficient sulfide present to precipitate all reduced iron as ferrous sulfide, whereas in equation 5 there is sufficient sulfide present only to reduce all available ferric iron, but not to precipitate any of the reduced iron. Equation 6 represents a case in which there is an intermediate amount of sulfide present, and a portion of the reduced iron is precipitated as ferrous sulfide. In equation 6, the remainder of the reduced iron is precipitated as siderite, in recognition of solubility equilibria that would pertain to the 30 mM bicarbonatebuffered medium used in this study. Sulfide was not measured in this study, so it was not possible to directly examine the stoichiometry of iron reduction in relation to hydrogen sulfide. However, if it is assumed that all consumed sulfate was reduced to hydrogen sulfide (a well-founded assumption based on the literature on sulfate reduction, the odor of acidified culture samples collected during this study, and the appearance of a black precipitate in culture medium that was probably iron sulfide [231), then the ratio of Fe(III) reduced to sulfate consumed can be used to examine the possibility of iron reduction via hydrogen sulfide. Equations 7, 8, and 9 in Table 1 represent complete oxidation of toluene with sulfate as the electron acceptor (with cell growth) combined with the oxidation of biogenic

sp. nov. FEMS Microbiol. Lett. 29:325-330. 4. Dolfing, J., J. Zeyer, P. Binder-Eicher, and R. P. Schwarzenbach. 1990. Isolation and characterization of a bacterium that mineralizes toluene in the absence of molecular oxygen. Arch. Microbiol. 154:336-341. 5. Edwards, E. A., L. E. Wills, M. Reinhard, and D. Grbic-Galic. 1992. Anaerobic degradation of toluene and xylene by aquifer microorganisms under sulfate-reducing conditions. Appl. Environ. Microbiol. 58:794-800. 6. Evans, P. J., D. T. Mang, K. S. Kim, and L. Y. Young. 1991. Anaerobic degradation of toluene by a denitrifying bacterium. Appl. Environ. Microbiol. 57:1139-1145. 7. Gossett, J. M. 1985. Anaerobic degradation of Cl and C2 chlorinated hydrocarbons. Final report. Publication ESL-TR85-38. Engineering and Services Laboratory, Air Force Engineering and Services Center, Tyndall Air Force Base, Fla. 8. Grbic-Galic, D., and T. M. Vogel. 1987. Transformation of toluene and benzene by mixed methanogenic cultures. Appl. Environ. Microbiol. 53:254-260. 9. Haag, F., M. Reinhard, and P. L. McCarty. 1991. Degradation of toluene and p-xylene in anaerobic microcosms: evidence for sulfate as a terminal electron acceptor. Environ. Toxicol. Chem. 10:1379-1389. 10. Hutchins, S. R., G. W. Sewell, D. A. Kovacs, and G. A. Smith. 1991. Biodegradation of aromatic hydrocarbons by aquifer microorganisms under denitrifying conditions. Environ. Sci. Technol. 25:68-76. 11. Kuhn, E. P., J. Zeyer, P. Eicher, and R. P. Schwarzenbach. 1988. Anaerobic degradation of alkylated benzenes in denitrifying laboratory aquifer columns. Appl. Environ. Microbiol. 54: 490-496. 12. Kunkle, G. R., and J. W. Shade. 1976. Monitoring ground-water quality near a sanitary landfill. Ground Water 14:11-20. 13. Lovley, D. R. 1987. Organic matter mineralization with the reduction of ferric iron: a review. Geomicrobiol. J. 5:375-399. 14. Lovley, D. R., M. J. Baedecker, D. J. Lonergan, I. M. Cozzarelli, E. J. P. Phillips, and D. I. Siegel. 1989. Oxidation of aromatic contaminants coupled to microbial iron reduction. Nature (London) 339:297-300. 15. Lovley, D. R., and D. J. Lonergan. 1990. Anaerobic oxidation of toluene, phenol, andp-cresol by the dissimilatory iron-reducing organism GS-15. Appl. Environ. Microbiol. 56:1858-1864. 16. Lovley, D. R., and E. J. P. Phillips. 1986. Organic matter mineralization with reduction of ferric iron in anaerobic sediments. Appl. Environ. Microbiol. 51:683-689. 17. Lovley, D. R., and E. J. P. Phillips. 1986. Availability of ferric iron for microbial reduction in bottom sediments of the freshwater tidal Potomac River. Appl. Environ. Microbiol. 52:751757. 18. Mackay, D., and W. Y. Shiu. 1981. A critical review of Henry's Law constants for chemicals of environmental interest. J. Phys. Chem. Ref. Data 10:1175-1199. 19. McCarty, P. L. 1971. Energetics and bacterial growth, p. 495-531. In S. D. Faust and J. V. Hunter (ed.), Organic compounds in aquatic environments. Marcel Dekker, Inc., New York. 20. McCarty, P. L. 1975. Stoichiometry of biological reactions. Prog. Water Technol. 7:157-172. 21. Nicholson, R. V., J. A. Cherry, and E. J. Reardon. 1983. Migration of contaminants in groundwater at a landfill: a case study. 6. Hydrogeochemistry. J. Hydrol. 63:131-176. 22. Pyzik, A. J., and S. E. Sommer. 1981. Sedimentary iron monosulfides: kinetics and mechanism of formation. Geochim. Cosmochim. Acta 45:687-698. 23. Rickard, D. T. 1969. The microbiological formation of iron sulphides. Stockholm Contrib. Geol. 20:49-66. 24. Rickard, D. T. 1974. Kinetics and mechanism of the sulfidation of goethite. Am. J. Sci. 274:941-952. 25. Schnell, S., F. Bak, and N. Pfennig. 1989. Anaerobic degradation of aniline and dihydroxybenzenes by newly isolated sulfatereducing bacteria and description of Desulfobacterium anilini. Arch. Microbiol. 152:556-563. 26. Sewell, G. W., H. H. Russell, and S. A. Gibson. 1991. Anaerobic

hydrogen sulfide to elemental sulfur with ferric iron as the electron acceptor. More specifically, equations 7, 8, and 9 represent the combination of equation 2 with equations 4, 5, and 6, respectively. Equation 2 was chosen because it is consistent with the results for the enrichment culture shown in Fig. 3 (i.e., the ratio of sulfate consumed to toluene consumed in this enrichment [3.8] was similar to the ratio in equation 2 [4.09]). Equation 2 is also supported by experiments with radiolabeled toluene (Fig. 4). Table 2 illustrates that, for the enrichment culture shown in Fig. 3, equation 9 is generally consistent with the observed ratios of Fe(III) reduced/sulfate consumed and Fe(III) reduced/toluene consumed. The fact that equation 9 is generally consistent with the observed stoichiometry in terms of iron, sulfate, and toluene indicates the plausibility that toluene was oxidized to carbon dioxide by a sulfate-reducing consortium and that the resulting sulfide was responsible for ferric iron reduction. The abiotic reaction proposed for hydrogen sulfide and amorphous Fe(OH)3 (equation 6) is only one of a number of possible reactions that could explain the stoichiometry. For example, a similar approach could be used in which thiosulfate, rather than elemental sulfur, was the sulfide oxidation product. The data collected in this study have been used to indicate the importance of iron in either initiating or accelerating toluene degradation and to provide support for a proposed mechanism of iron reduction; however, the data are not suitable for explaining how iron stimulated toluene degradation. Further study will be required to address this issue. Hypotheses that could explain the nature of iron's effect include the following: (i) iron may have reduced sulfide toxicity to toluene-degrading bacteria (e.g., via the oxidation and precipitation mechanisms illustrated by equation 4 in Table 1), and (ii) iron might have been a limiting nutrient that was required for toluene degradation (e.g., amendments of iron could have compensated for the indigenous ferrous iron that was removed from solution by precipitation as ferrous sulfide or other phases). Although amorphous Fe(OH)3 itself is extremely insoluble, it is possible that the reduction of millimolar amounts of ferric iron (e.g., Fig. 3) resulted in micromolar concentrations of Fe(II) that remained soluble and available to bacteria. The microbiological and geochemical complexities of these systems merit further investigation. Ongoing research is focusing on experiments to clarify the role of iron in stimulating toluene degradation, to isolate the toluene-degrading bacteria in pure culture (or, if syntrophy is required, to isolate the critical members of the community), and to elucidate the metabolic pathway of toluene degradation.

ACKNOWLEDGMENTS We thank Ron Hoeppel of the Naval Civil Engineering Laboratory (Port Hueneme, Calif.) for providing the Patuxent River materials used as the inoculum in this study. This research was supported by U.S. Environmental Protection Agency grant EPA CR-815721.
REFERENCES 1. Bak, F., and F. Widdel. 1986. Anaerobic degradation of phenol and phenol derivatives by Desulfobactenum phenolicum sp. nov. Arch. Microbiol. 146:177-180. 2. Berner, R. A. 1964. Iron sulfides formed from aqueous solution at low temperatures and atmospheric pressure. J. Geol. 72:293306. 3. Cord-Ruwisch, R., and J. L. Garcia. 1985. Isolation and characterization of an anaerobic benzoate-degrading spore-forming sulfate-reducing bacterium, Desulfotomaculum sapomandens

27. 28. 29.

biotransformation of toluene under sulfate-reducing conditions. Poster presentation at Robert S. Kerr Ground Water Research Seminar (U.S. Environmental Protection Agency), 26-28 March 1991, Oklahoma City, Okla. Sittig, M. 1985. Handbook of toxic and hazardous chemicals and carcinogens, 2nd ed., p. 868-870. Noyes Publications, Park Ridge, N.J. Smolenski, W. J., and J. M. Suflita. 1987. Biodegradation of cresol isomers in anoxic aquifers. Appl. Environ. Microbiol. 53:710-716. Szewzyk, R., and N. Pfennig. 1987. Complete oxidation of catechol by the strictly anaerobic sulfate-reducing Desulfobacterium catecholicum sp. nov. Arch. Microbiol. 147:163-168. U.S. Environmental Protection Agency. 1986. Underground motor fuel storage tanks: a national survey. Publication NTIS PB 86-216512. U.S. Environmental Protection Agency, Washington, D.C. Vogel, T. M., and D. Grbic-Galic. 1986. Incorporation of oxygen from water into toluene and benzene during anaerobic fermentative transformation. Appl. Environ. Microbiol. 52:200-202. Widdel, F., G.-W. Kohring, and F. Mayer. 1983. Studies on dissimilatory sulfate-reducing bacteria that decompose fatty
acids. III. Characterization of the filamentous gliding Desulfonema limicola gen. nov. sp. nov., and Desulfonema magnum sp. nov. Arch. Microbiol. 134:286-294. Wilson, B. H., B. Bledsoe, and D. Kampbell. 1987. Biological processes occurring at an aviation gasoline spill site, p. 125-137. In R. C. Averett and D. M. McKnight (ed.), Chemical quality of water and the hydrologic cycle. Lewis Publications, Inc., Chelsea, Mich. Wilson, B. H., G. B. Smith, and J. F. Rees. 1986. Biotransformations of selected alkylbenzenes and halogenated aliphatic hydrocarbons in methanogenic aquifer material: a microcosm study. Environ. Sci. Technol. 20:997-1002. Zeyer, J., P. Eicher, J. Dolfing, and R. P. Schwarzenbach. 1990. Anaerobic degradation of aromatic hydrocarbons, p. 33-40. In D. Kamely, A. Chakrabarty, and G. S. Omenn (ed.), Biotechnology and biodegradation (Advances in Applied Biotechnology Series, vol. 4). Portfolio Publishing Company, Woodlands, Tex. Zeyer, J., E. P. Kuhn, and R. P. Schwarzenbach. 1986. Rapid microbial mineralization of toluene and 1,3-dimethylbenzene in the absence of molecular oxygen. Appl. Environ. Microbiol. 52:944-947.