Sediments from a hydrocarbon-contaminated aquifer, where periodic shifts between sulfate reduction and methanogenesis occurred, were examined to determine whether the degradation of toluene under sulfate-reducing conditions depended on interspecies hydrogen transfer. Toluene degradation under sulfate-reducing conditions was inhibited by the addition of 5 mM sodium molybdate, but the activity was not restored upon the addition of an actively growing, hydrogen-using methanogen. Toluene degradation was not inhibited in microcosms where hydrogen levels were maintained at a level theoretically sufficient to inhibit toluene degradation if the process proceeded via interspecies hydrogen transfer. Finally, the addition of carbon monoxide, a potent inhibitor of hydrogenase activity, inhibited hydrogen but not toluene consumption in sulfate-reducing microcosms. These results suggest that toluene is degraded directly by sulfate-reducing bacteria without the involvement of interspecies hydrogen transfer. The sequence of experiments used to reach this conclusion could be applied to determine the role of interspecies hydrogen transfer in the degradation of a variety of compounds in different environments or under different terminal electron-accepting conditions.
The metabolic interactions among species involved in anaerobic hydrocarbon degradation in natural environments are not well understood. The involvement of metabolic interactions such as interspecies hydrogen transfer would make the anaerobic degradation of aromatic hydrocarbons sensitive to fluctuations in environmental conditions such as the dissolved hydrogen concentration. These fluctuations could affect the rate and extent of degradation of these compounds. Thus, a greater understanding of the involvement of interspecies interaction will provide useful insights to the operation of anaerobic bioreactors and in the design of bioremediation efforts.
In methanogenic environments, interspecies hydrogen transfer is required for the degradation of a wide range of organic compounds including alcohols, fatty acids, amino acids, organic acids such as lactate, and benzoate and other aromatic acids . Thus, it is likely that the degradation of aromatic hydrocarbons such as toluene also requires interspecies hydrogen transfer . However, it is not clear whether similar metabolic interactions are operative under other terminal electron-accepting conditions. Several pure cultures of anaerobic bacteria have been characterized that can mineralize aromatic hydrocarbons such as toluene coupled to the reduction of sulfate (equation 1, Table 1) [3,4] or other electron acceptors [5–7]. Recently, Meckenstock  found that a newly isolated sulfate reducer, strain TRM1, and the iron reducer, Geobacter metallireducens, could degrade toluene in the absence of their respective electron acceptors when grown in syntrophic association with a hydrogen-using bacterium. This study entertains the possibility that toluene degradation in sulfate- or iron-reducing environments could proceed by a syntrophic association where the hydrogen produced by the toluene-degrading organisms (equation 2, Table 1) is rapidly used by a hydrogen-using, sulfate reducer (equation 3, Table 1). The combined reaction (equation 4, Table 1) is thermodynamically favorable since hydrogen levels are maintained at a very low level by the hydrogen-using bacterium. If acetate is consumed by another sulfate-reducing bacterium (equation 5, Table 1), the overall stoichiometry for toluene degradation by a sulfate-reducing consortium (equation 1, Table 1) would be indistinguishable from that catalyzed by known pure cultures.
Table 1. Different reactions possibly involved in anaerobic toluene degradation under sulfate-reducing conditionsa
a Values obtained from [7,34] or calculated according to ΔGf values from .
In support of this concept, it is now known that some bacteria capable of growth by anaerobic respiration can grow in the absence of their electron acceptor in syntrophic associations with hydrogen-using bacteria. For example, some sulfate-reducing bacteria can grow with ethanol or lactate [8,9] in the absence of sulfate in syntrophic association with hydrogen-using methanogens. Wu et al.  suggested the involvement of sulfate-reducing bacteria in propionate and ethanol degradation in methanogenic granules treating brewery wastewater. The iron-reducing bacterium, Geobacter sulfurreducens, can metabolize acetate in the absence of iron in syntrophic association with Wollinella succinogenes, a hydrogen-using nitrate reducer . Conversely, Syntrophobacter species, which were once considered to be obligate syntrophic bacteria, can use sulfate as the electron acceptor for the oxidation of propionate and other organic acids [12,13]. These studies show that many anaerobic bacteria have diverse modes of energy metabolism and that their function in natural ecosystems cannot be inferred based only on their affiliation with a group of organisms that have been historically classified as having one mode of energy metabolism such as sulfate reduction.
We tested whether toluene degradation under sulfate-reducing conditions in a shallow aquifer contaminated with gas condensate hydrocarbons involved interspecies hydrogen transfer. Geochemical and dissolved hydrogen measurements indicated that sulfate reduction and methanogenesis were the dominant terminal electron-accepting processes and that these two processes varied temporally and spatially at the site . We hypothesized that the periodic shifts in the terminal electron-accepting processes may have established conditions where syntrophic populations involved in toluene degradation under methanogenic conditions would be able to couple toluene degradation with hydrogen-using sulfate reducers when sulfate was periodically replenished at this site. While the outcome of the experiments failed to support the hypothesis that toluene degradation under sulfate-reducing conditions involves interspecies hydrogen transfer, the protocol that we used to test this hypothesis will be useful to determine the role of interspecies hydrogen transfer in the degradation of compounds in natural ecosystems.
2. Materials and methods
2.1 Microorganisms and media
Methanospirillum hungatei strain JF1 and Desulfovibrio strain G11 were obtained from the culture collection of M.P. Bryant (Urbana, IL, USA). Syntrophus aciditrophicus strain SB, a syntrophic microorganism that utilizes benzoate in the presence of hydrogen-using bacteria, was isolated in our laboratory from a sewage treatment plant in Norman, OK, USA . A basal medium  was used for growing the above strains with the omission of rumen fluid. M. hungatei were grown in basal medium with a headspace pressurized to 140 kPa by a gas mixture of 80% H2:20% CO2. Desulfovibrio strain G11 was grown similarly, but in a basal medium containing 15 mM sodium sulfate. Both microorganisms were grown in a shaking incubator (100 rpm). S. aciditrophicus was grown in pure culture in a basal medium with crotonic acid (40 mM) as a substrate and a headspace of 80% N2:20% CO2. Cocultures of S. aciditrophicus and M. hungatei were cultured in a basal medium containing 2.5 mM benzoate as a substrate and a headspace of 80% N2:20% CO2. All microorganisms were incubated at 37°C.
2.2 Sampling and microcosms preparation
Sediment samples were obtained from a gas condensate-contaminated site near Ft. Lupton, CO, USA . Microcosms to study the factors affecting toluene degradation were prepared in an anaerobic glove box  by combining 25 g of sediment with 40 ml of basal medium containing 6 mM sodium sulfate in 120-ml serum bottles. The bottles were sealed with a composite of a butyl rubber stopper (Bellco CO, Vineland, NJ, USA) with a layer of teflon fused to the bottom of the stopper and crimped with aluminum seals. After preparation, the headspace of each microcosm was exchanged with 80% N2:20% CO2 (170 kPa). Toluene (2 μl) (99.8%, Aldrich chemical Co., Milwaukee, WI, USA) was then added as a pure liquid using a glass syringe to a final aqueous concentration of approximately 0.3–0.4 mM. All microcosms were incubated at room temperature in the dark without shaking unless otherwise stated.
2.3 Experimental design
To determine whether electron flow during toluene degradation could be shifted from sulfate reduction to methanogenesis, 15 microcosms with about 0.4 mM toluene and 6 mM sulfate were prepared as described above. The microcosms were incubated until toluene degradation and sulfate consumption commenced. After 16 days, sodium molybdate was added from a 100 mM sterile anoxic stock solution to six of the microcosms to a final concentration of 5 mM to inhibit sulfate reduction. The other three microcosms received NaCl from a 100 mM sterile anoxic stock solution to a final concentration of 5 mM and served as a positive control for ionic strength. After 70 days of incubation, 5 ml of an actively growing culture of M. hungatei strain JF1 (OD600 0.37) was added to three of the microcosms that received sodium molybdate to determine if toluene degradation could be coupled to methanogenesis. After 295 days of incubation, two of the three microcosms that received M. hungatei each received 5 ml of a coculture of S. aciditrophicus and M. hungatei and enough sodium benzoate to give a final concentration of ∼0.5 mM. The headspaces of the later three microcosms were then evacuated and flushed with 80% N2:20% CO2 and the concentrations of benzoate and methane were monitored. This treatment was done to show that syntrophic benzoate degradation could be coupled to methanogenesis in these microcosms. In addition to the above microcosms, three microcosms did not receive toluene to correct for sulfate reduction coupled to endogenous electron donors. Another three microcosms with about 0.4 mM toluene and 6 mM sodium sulfate were autoclaved (20 min, 121°C) and served as heat-killed controls.
As a positive control to show that syntrophic metabolism could be shifted from one terminal electron-accepting process (methanogenesis) to another (sulfate reduction), an experiment similar to that described above was conducted with butyrate as the electron donor. Fifteen microcosms with 5 mM sodium butyrate were prepared as described above. Three of these microcosms were autoclaved (20 min, 121°C) and served as heat-killed controls. Three other microcosms were unamended with substrate to correct for methane production from endogenous electron donors. After 14 days of incubation, bromoethanesulfonic acid (BESA) was added from a 100 mM sterile, anoxic stock solution to a final concentration of 5 mM to each of six microcosms to inhibit methanogenesis. The other three microcosms each received sodium chloride to give a final concentration of 5 mM and served as positive controls. After 44 days of incubation, three of the microcosms that received BESA each received 5 ml of an actively growing culture of Desulfovibrio strain G11 (OD600 of about 0.34) and sodium sulfate to give a final concentration of 6 mM.
2.4 Effect of hydrogen and carbon monoxide on toluene degradation
To test the effect of hydrogen on toluene degradation under sulfate-reducing conditions, nine microcosms with about 0.4 mM toluene and 5 mM ferrous sulfate heptahydrate (FeSO4.7H2O) were prepared as described above. FeSO4.7H2O was used to prevent the accumulation of potential toxic levels of sulfide  due to stimulation of sulfate reduction by repeated hydrogen additions. Three of these microcosms were autoclaved (121°C, 20 min) to serve as heat-killed controls. After 28 days of incubation, three microcosms each received hydrogen (2×10−2 atm) and the hydrogen concentration was monitored daily. Microcosms were reamended with hydrogen whenever the partial pressure of hydrogen fell below 10−2 atm. This value is approximately 100 times the hydrogen partial pressure needed to theoretically inhibit hydrogen production as calculated according to Thauer et al.  given the concentrations of reactants and products present in the microcosms and equation 1 (Table 1) for toluene degradation. The other three replicates received equal volumes of 100% N2 gas instead of hydrogen each time hydrogen was added to the above microcosms. Each gas was added to the microcosms using sterile syringes and that had been flushed with the respective gas at least 10 times . Microcosms were incubated with shaker (100 rpm) to allow free hydrogen diffusion to the liquid phase.
To test the effect of CO on toluene degradation under sulfate-reducing conditions, experiments were first conducted to determine the minimum concentration of CO that inhibits hydrogen uptake. Duplicate microcosms with about 1.5–2% headspace concentration of hydrogen and 5 mM sodium sulfate were each amended with 0–10% CO. A concentration of 10% CO in the headspace completely inhibited hydrogen uptake (Fig. 4B) and hence was used in toluene-amended microcosms. Microcosms were prepared as mentioned above with about 0.4 mM toluene and 5 mM sodium sulfate. Toluene was consumed after 68 days. At day 69, toluene (0.4 mM) and sulfate (5 mM) were amended to all microcosms. Three microcosms also received CO at a concentration of 10% while the three control microcosms received a similar volume of N2 instead of CO.
2.5 Analytical procedures
Toluene concentrations were determined by headspace analysis with a gas chromatograph (GC) equipped with a flame ionization detector and a 30-m Carbograph VOC capillary column (Alltech Inc., Deerfield, IL, USA) held isothermally at 150°C. The carrier gas was helium at a flow rate of 16 ml min−1. Hydrogen in the headspace was analyzed with a GC equipped with a mercury reduction vapor detector . Methane was also analyzed by GC  and sulfate was analyzed by ion chromatography . Benzoate concentrations were determined by high-performance liquid chromatography as described elsewhere .
3.1 Shifting of the terminal electron-accepting process
We hypothesized that if toluene degradation under sulfate-reducing conditions was mediated by a syntrophic microorganism and a hydrogen-using sulfate reducer, then toluene degradation would be restored in molybdate-inhibited, toluene-degrading enrichments by the addition of an actively growing, hydrogen-using methanogen. The result of an experiment to test this hypothesis is shown in Fig. 1. Microcosms that received toluene and sulfate but not molybdate completely degraded toluene to below the detection limit of our analysis after about 70 days of incubation. Toluene consumption occurred with concomitant consumption of nearly stoichiometric amounts of sulfate according to equation 1 of Table 1 (3.94 mol of sulfate per mol of toluene consumed, after correction for sulfate reduction in substrate-unamended microcosms, not shown). In contrast, sulfate reduction and toluene degradation were inhibited in the replicate microcosms that received 5 mM sodium molybdate at day 16. The addition of 5 ml of an actively growing culture of M. hungatei to three of the molybdate-inhibited microcosms at day 70 did not restore toluene degradation, even after prolonged incubations to 295 days (Fig. 1). The addition of 0.4 mM toluene at day 128 to microcosms that did not receive molybdate showed that these microcosms could still actively degrade toluene (Fig. 1).
Although molybdate is considered to be a specific inhibitor of sulfate reduction , the concentration used in this experiment may have been high enough to inhibit microorganisms capable of syntrophic metabolism or the methanogens in the microcosms. To rule out this possibility, 5 ml of a syntrophic, methanogenic benzoate-degrading coculture of S. aciditrophicus and M. hungatei and 0.5 mM sodium benzoate (final concentration) were added, at day 295 to two of the microcosms that received molybdate and M. hungatei. In the two microcosms that received the syntrophic coculture, benzoate was consumed and methane was produced at levels far above those observed in the microcosms that did not receive benzoate and the syntrophic coculture (Fig. 1). These latter data suggest that neither syntrophic metabolism nor methanogenesis was likely to have been inhibited by the concentrations of sodium molybdate or toluene present in the microcosms, although it is possible that the toluene oxidizer in this system is sensitive to molybdate.
The ability of sediment syntrophic consortia to shift from one electron acceptor to another was confirmed in a separate experiment where methanogenic butyrate-degrading enrichments, which are dependent on interspecies hydrogen transfer, were used. The addition of 5 mM of BESA, a potent inhibitor of methanogenesis, after an initial 14 days of incubation inhibited butyrate degradation and methanogenesis (Fig. 2). Butyrate degradation coupled to sulfate reduction and acetate production was restored by the addition of 5 mM sulfate (final concentration) and 5 ml of a culture of Desulfovibrio strain G11 at day 44 (Fig. 2).
3.2 Effect of hydrogen on toluene degradation
In order for syntrophic toluene degradation to be thermodynamically favorable, the hydrogen partial pressure must be kept at a very low level (approximately 10−5 atm). We maintained the hydrogen partial pressure at a value 100 times above the value theoretically needed to inhibit syntrophic toluene metabolism. The rate of toluene degradation in these hydrogen-amended microcosms was not different from that of control microcosms that received nitrogen (Fig. 3).
3.3 Effect of carbon monoxide on toluene degradation
Carbon monoxide is a potent inhibitor of the enzyme hydrogenase, which is required for interspecies hydrogen transfer. We compared toluene degradation in microcosms that received carbon monoxide to a final concentration of 10% to controls that received an equal volume of nitrogen gas. Fig. 4 shows that toluene degradation occurred at similar rates regardless of whether CO was present or not.
This work provides conclusive evidence that toluene degradation under sulfate-reducing conditions by bacteria derived from a hydrocarbon-contaminated aquifer does not depend on interspecies hydrogen transfer. However, we believe that the approach we used will be useful in determining the role of interspecies hydrogen transfer in the degradation of compounds in natural ecosystems. Our protocol for testing the role of syntrophic interactions in substrate degradation is based on fundamental characteristics of syntrophic metabolism such as independence of terminal electron-accepting reaction, sensitivity to high levels of hydrogen, and sensitivity to hydrogenase inhibitors.
First, by using an inhibitor of the terminal electron-accepting process, we showed that toluene degradation directly depended on sulfate reduction. If sulfate reduction was needed only for hydrogen removal (i.e. if toluene degradation under sulfate-reducing conditions required interspecies hydrogen transfer), the addition of another hydrogen-user which is not inhibited by molybdate (e.g. a methanogen) should restore toluene degradation activity. To interpret this experiment correctly, we had to show that (a) molybdate did not inhibit methanogenic or syntrophic metabolic activities and that (b) sediment-associated syntrophic populations were capable of coupling with different terminal electron-accepting processes. This was done by showing that syntrophic benzoate degradation could occur in molybdate-inhibited microcosms and that syntrophic butyrate degradation could be coupled to methanogenesis or sulfate reduction. Our protocol is also based on the assumption that maintaining high hydrogen levels will inhibit substrate turnover in systems that depend on interspecies hydrogen transfer. We maintained hydrogen concentrations at levels much higher than needed to inhibit syntrophic metabolism but low enough to avoid stimulating hydrogen-utilizing microorganisms in the enrichments. This necessitated the continuous monitoring and addition of hydrogen gas. A set of control microcosms that received only an inert gas (nitrogen) was needed to show that repeated additions of a gas did not inhibit substrate decay or stimulate aerobic metabolism by repeated addition of oxygen. It was also necessary to ensure that enough electron acceptor (sulfate in our case) was present since continuous hydrogen additions will consume sulfate and to prevent sulfide build-up, which can have inhibitory effects on microbial activities , by providing sulfate as FeSO4.7H2O. The presence of iron will remove any sulfide produced as ferrous sulfate precipitate.
It has previously been suggested that syntrophic and hydrogen-using microorganisms are closely positioned next to each other in flocs or microniches where hydrogen concentrations may not be in an equilibrium with the bulk aqueous hydrogen pool, i.e. the measured hydrogen concentrations . We used hydrogen concentrations that were roughly 100 times greater than that needed to inhibit syntrophic toluene degradation. Also, we used hydrogenase inhibitors to block hydrogen uptake and interspecies hydrogen transfer. For valid interpretation of the results, controls should be run to ensure that the concentration of the inhibitor used blocks hydrogen uptake. The techniques used in all experiments described above are common to almost all anaerobic laboratories. Thus, this approach should be useful to determine the role of syntrophic metabolism in the degradation of a variety of compounds regardless of the ecosystem.
Few reports have examined the role of syntrophic metabolism in substrate turnover in the presence of sulfate. Lovley et al.  suggested that benzene degradation in marine sediments under sulfate-reducing conditions is mediated by a single microorganism rather than a microbial consortium. This conclusion was based on the inability to detect potential extracellular intermediates of benzene degradation (including acetate) in isotope trapping experiments utilizing 14C benzene. It has been suggested that syntrophic propionate degraders can effectively compete with propionate oxidizing sulfate reducers, especially at lower concentrations of sulfate [12,26]. Schoelten and Stams  suggested that propionate and butyrate metabolism in freshwater sediments was mediated by sulfate-reducing bacteria. This conclusion was based on the observation that methane production was inhibited in propionate and butyrate-degrading enrichments by the addition of sulfate. However, it could be argued that the addition of sulfate allowed sulfate-reducing bacteria to outcompete methanogenic bacteria for hydrogen produced by syntrophic metabolism of these compounds. Balba and Nedwell  concluded that propionate and butyrate were degraded directly by sulfate-reducing bacteria in marine sediments, based on the observation that molybdate inhibited propionate and butyrate metabolism and that the addition of 80% hydrogen to the gas phase did not inhibit butyrate degradation. However, showing that molybdate inhibits substrate degradation is not sufficient in itself to conclude that syntrophic metabolism is not operative since the results will be similar. Also, the time span of such experiments must be long enough to determine if a shift of electron flow from sulfate reduction to methanogenesis occurred. The use of high initial hydrogen concentrations to inhibit syntrophic metabolism is problematic since sulfate-reducing bacteria could consume all the hydrogen initially added and lower the hydrogen concentrations to levels low enough to allow interspecies hydrogen transfer to proceed. This may also deplete sulfate pools and lead to the potentially inhibitory concentrations of sulfide.
The identification of microorganisms by molecular techniques such as phospholipid fatty acid analysis, denaturing gradient gel electrophoresis analysis, or direct cloning and sequencing of 16S RNA genes of organisms has greatly added to our understanding of the microbial community structures in many environments [29–31]. Although the benefits of these techniques are indisputable, they do not provide direct information on the types of compounds that these organisms degrade. This information is usually inferred from the known physiological traits for organisms that have culturable representatives. However for organisms with diverse modes of metabolism such as the sulfate-reducing bacteria, the assignment of a physiological function based on phylogenetic identification is extremely difficult. Also, organisms with a close phylogenetic (e.g. 16S RNA gene sequence similarity values of up to 98.7%) can have different substrate utilization patterns [32,33]. The application of our approach together with molecular analysis would help clarifying the physiological roles of different groups of microorganisms in natural ecosystems.
We thank Dr. Lisa M. Gieg for helpful editorial comments. This work was supported by Grants DE-FG03-96-ER-62287 and DE-FG03-96-ER20214/A003 from the Department of Energy.