ABSTRACT
The transformation of ferrihydrite to stable iron oxides over time has important consequences for biogeochemical cycling of many metals and nutrients. The response of methanogenic activity to the presence of iron oxides depends on the type of iron mineral, but the effects of changes in iron mineralogy on methanogenesis have not been characterized. To address these issues, we constructed methanogenic cocultures of Geobacter and Methanosarcina strains with different ferrihydrite mineralization pathways. In this system, secondary mineralization products from ferrihydrite are regulated by the presence or absence of phosphate. In cultures producing magnetite as the secondary mineralization product, the rates of methanogenesis from acetate and ethanol increased by 30.2% and 135.3%, respectively, compared with a control lacking ferrihydrite. Biogenic magnetite was proposed to promote direct interspecies electron transfer between Geobacter and Methanosarcina in a manner similar to that of c-type cytochrome and thus facilitate methanogenesis. Vivianite biomineralization from ferrihydrite in the presence of phosphate did not significantly influence the methanogenesis processes. The correlation between magnetite occurrence and facilitated methanogenesis was supported by increased rates of methane production from acetate and ethanol with magnetite supplementation in the defined cocultures. Our data provide a new perspective on the important role of iron biomineralization in biogeochemical cycling of carbon in diverse anaerobic environments.
IMPORTANCE It has been found that microbial methanogenesis is affected by the presence of iron minerals, and their influences on methanogenesis are associated with the mineralogical properties of the iron minerals. However, how changes in iron mineralogy affect microbial methanogenesis has not been characterized. To address this issue, we constructed methanogenic cocultures of Geobacter and Methanosarcina strains with different ferrihydrite mineralization pathways. The experimental results led to two contributions, i.e., (i) the transformation of iron minerals might exert an important influence on methanogenesis under anaerobic conditions and (ii) both biogenic and chemical magnetite can accelerate syntrophic ethanol oxidization between Geobacter metallireducens and Methanosarcina barkeri. This study sheds new light on the important role of iron biomineralization in the biogeochemical cycling of carbon in diverse anaerobic environments, particularly in iron-rich natural and agricultural wetland soils.
INTRODUCTION
Ferrihydrite is a common iron (hydr)oxide in soils and sediments that is thermodynamically unstable and is a precursor for other iron (hydr)oxides that are more stable (1–4). Ferrihydrite reduction by dissimilatory iron-reducing bacteria (DIRB), which is an important biotic process, produces Fe(II) and, consequently, secondary Fe(II) or Fe(II)/Fe(III) mixed-valence mineral phases, such as magnetite, green rust, vivianite, or siderite (1–4). Many environmental factors influence secondary mineralization of ferrihydrite, including inorganic oxyanions, natural organic matter, and the types and amounts of bacterial cells (5). In particular, phosphate has a profound impact on the extent of iron reduction and the mineralization pathway (6–8). For example, magnetite is formed as the predominant biomineralization product of ferrihydrite by Shewanella putrefaciens CN32 in the absence of phosphate, whereas this pathway leads to carbonate green rust and vivianite formation in the presence of phosphate (9).
As the most bioavailable iron oxide, ferrihydrite can serve as a terminal electron acceptor for microbial respiration (10), thus influencing carbon cycling in anaerobic soils and sediments (11). For example, microbial reduction of ferrihydrite competes with microbial methanogenesis for organic matter under anaerobic conditions. The competition between DIRB and methanogenic archaea for common electron donors such as acetate and H2 has been postulated as the primary mechanism by which ferrihydrite inhibits methanogenesis (12–15). However, (semi)conductive iron oxides, such as hematite or magnetite, can accelerate methanogenesis from acetate, ethanol (16, 17), propionate (17, 18), and butyrate (19). It has been proposed that the mechanism for the stimulation of methanogenesis involves direct interspecies electron transfer (DIET) between syntrophic oxidizers and methanogens via the conductive iron oxide minerals (16, 18, 19). These findings imply that the methanogenic pathway may be correlated with the mineralogical properties of iron oxide, particularly its crystallinity and conductivity. This raises the following question: if microbial methanogenesis is influenced by iron mineralogy, how do changes in iron mineralogy affect microbial methanogenesis?
To address this question, we previously conducted long-term anaerobic incubation of paddy soil and ferrihydrite-supplemented soil cultures to investigate how secondary mineral products of ferrihydrite reduction affect methanogenesis; the results showed that microbial magnetite formation was correlated with facilitated methanogenesis, in terms of the average methane production rate and the acetate degradation rate (20). However, the association between the promotion of methanogenesis and magnetite formation is very speculative, due to the complexity of the soil system. In this study, using cocultures of Geobacter sulfurreducens and Methanosarcina barkeri and of Geobacter metallireducens and Methanosarcina barkeri undergoing secondary mineralization of ferrihydrite, we investigated whether changes in the mineralogical properties of iron-containing minerals influence the methanogenic pathway; the different mineralization pathways of ferrihydrite were regulated by the modulation of phosphate levels. The results observed here have implications for our understanding of the biogeochemical cycling of organic carbon, particularly in iron-rich subsurface environments.
MATERIALS AND METHODS
Microorganisms, culture media, and growth conditions.Geobacter sulfurreducens strain PCA (DSM 12127), Geobacter metallireducens strain GS-15 (DSM 7210), and Methanosarcina barkeri strain MS (DSM 800) were purchased from DSMZ (Braunschweig, Germany). All culturing and sampling were performed under strictly anaerobic conditions, with an atmosphere of 80% N2/20% CO2 (vol/vol). G. sulfurreducens inocula were grown in NBAF medium (21), using acetate (10 mM) as the electron donor and fumarate (40 mM) as the electron acceptor. G. metallireducens strain GS-15 was grown in FC medium (22), using ethanol (10 mM) as the electron donor and ferric citrate (55 mM) as the electron acceptor. M. barkeri was cultured anaerobically in DSMZ methanogenic medium 120 (Tables 1 to 3), using methanol as the methanogenic substrate. The culturing temperatures for the Geobacter strains and M. barkeri were 30°C and 37°C, respectively. Before inoculation into the coculture system, the pure culture solution was centrifuged and resuspended twice in sterilized phosphate-free medium 120 without methanol.
Composition of DSMZ medium 120
Composition of vitamin solution
Composition of trace element solution SL-10
Coculture experiments.Secondary mineralization of ferrihydrite was performed in 125-ml serum bottles containing 25 ml of sterilized medium (autoclaved at 121°C for 30 min) supplemented with 4 mM ethanol (or 5 mM acetate) and 25 mM ferrihydrite as Fe atoms. The stock solution of ferrihydrite was prepared by neutralizing 0.4 M ferric chloride solution with 1 M NaOH to a pH of approximately 7.0 (23). To control the mineralization products from ferrihydrite reduction, one treatment was conducted with phosphate-free medium 120 and the other was conducted using medium 120, which has a phosphate concentration of 3.7 mM. Microbial reduction of ferrihydrite was initiated by injecting each vial with 1.25 ml of Geobacter strain (G. sulfurreducens for acetate or G. metallireducens for ethanol; cell density, ∼6.0 × 105 cells/ml). After secondary mineralization, the concentration of ethanol or acetate in each treatment was adjusted to an initial concentration of 4 or 5 mM, respectively. Methanogenesis in the cocultures was initiated by the addition of 1.25 ml of M. barkeri (cell density, ∼1.0 × 105 cells/ml); at that point, the initial phosphate-free treatment was supplemented with 3.7 mM phosphate. The yeast extract and Casitone present in medium 120 were removed in the incubation experiments, to eliminate the possibility of methane production from the yeast extract and Casitone. All experiments were conducted in triplicate. The anaerobic bottles for secondary mineralization of ferrihydrite and coculture methanogenesis were incubated at a constant temperature of 37°C in the dark. For regular sampling, triplicate vials were used for DNA extraction and for measurement of the concentrations of Fe(II), ethanol, acetate, and CH4.
Magnetite experiments.To further investigate the relationship between accelerated methanogenesis and magnetite biomineralization from ferrihydrite, defined methanogenic cocultures were conducted in the presence and absence of magnetite. Magnetite nanoparticles were synthesized by slowly adding a Fe(II)/Fe(III) acidic solution (0.8 M FeCl3 and 0.4 M FeCl2 in 0.4 M HCl) into a vigorously mixed NaOH solution (1.5 M). The nanoparticles were then purified by centrifugation and suspended in deoxygenated water (24). Cocultures were initiated with 1.25 ml of the Geobacter strain and 1.25 ml of M. barkeri in 125-ml serum bottles containing 25 ml of sterilized medium 120 (without yeast extract and Casitone), 4 mM ethanol or 5 mM acetate as the carbon source, and 25 mM magnetite (as Fe atoms) when required. M. barkeri incubated alone in the same medium was used as a control for syntrophic methanogenic cocultures.
Analytical analyses.A 200-μl sample of gas was extracted from the headspace of the anaerobic bottles using a sterile syringe, and the concentrations of CH4 were determined using a GC9700 gas chromatograph (Techcomp Instruments, Shanghai, China) equipped with a flame ionization detector (FID). The concentrations of ethanol were also analyzed by gas chromatography. The concentrations of acetate were measured by high-performance liquid chromatography (HPLC) (Shimadzu LC-15C; Japan) with a Wondasil C18 reverse-phase column (250 mm by 4.6 mm; 5-μm pore size), and the lower detection limit was 0.1 mg/liter. The HCl-extractable Fe(II) concentrations were determined via the ferrozine technique, as described previously (11). To identify secondary minerals by X-ray diffraction (XRD) analysis, the samples were centrifuged at 4,000 rpm for 5 min, washed twice with deionized water (5 min each time), and then dried overnight in a vacuum desiccator. All XRD analyses were performed with a PANalytical (Netherlands) X'pert Pro multipurpose diffractometer (MPD) at 40 kV and 40 mA using CuKα1,2 radiation (λ = 0.15406 and 0.15444 nm), with a scan rate of 0.1°θ/min.
RESULTS AND DISCUSSION
Effects of phosphate on secondary mineralization of ferrihydrite.Microbial reduction of ferrihydrite by Geobacter strains occurred without a lag phase regardless of the presence of phosphate (see Fig. S1A and S2A in the supplemental material). In the G. sulfurreducens cultures, the total HCl-extractable Fe(II) production accumulated to maximum levels of 9.8 mM in incubations both with and without phosphate. In the G. metallireducens cultures, the maximum Fe(II) production levels were very similar between the phosphate-free cultures (11.6 mM) and the phosphate-supplemented cultures (11.2 mM). Ferrihydrite reduction was coupled to acetate/ethanol utilization; approximately 2.4 mM acetate and 1.5 mM ethanol were consumed in the G. sulfurreducens cultures and the G. metallireducens cultures, respectively (see Fig. S1B and S2B). According to the stoichiometric equations, microbial Fe(III) reduction coupled to acetate/ethanol oxidation produces 8 or 12 mol of Fe(II) at the expense of every 1 mol of acetate (CH3COO− + 8 Fe3+ + 4 H2O → 8 Fe2+ + 2 HCO3− + 9 H+) or ethanol (CH3CH2OH + 12 Fe3+ + 3 H2O → 12 Fe2++ 2 CO2 + 12 H+), respectively. Thus, approximately 50% of the electrons from acetate oxidization were recovered by dissimilatory Fe(III) reduction by G. sulfurreducens, and 63 to 66% of ethanol was consumed by dissimilatory Fe(III) reduction by G. metallireducens. Without considering substrate conversion to biomass, the electron balances were estimated to be 50 to 66% (see Tables S1 and S2 in the supplemental material), similar to values observed previously during dissimilatory selenate reduction (25). A portion of the carbon was assimilated by the microbial cells. For example, the cell concentrations of G. sulfurreducens increased 7.1- and 7.6-fold due to the microbial reduction of ferrihydrite in the presence and absence of phosphate, respectively (see Fig. S3 in the supplemental material).
Secondary minerals produced following ferrihydrite reduction were characterized by XRD (Fig. 1 and 2). In the absence of phosphate, the nonmagnetic reddish brown ferrihydrite was gradually transformed to black solids that were strongly attracted to a magnet (see Fig. S4 in the supplemental material). The formation of magnetite was confirmed by the characteristic diffraction pattern of lines 220, 311, 400, 422, 511, and 440 (Fig. 1A and 2A) (10). At the end of the ferrihydrite reduction in the absence of phosphate, the extent of Fe(III) reduction in the G. sulfurreducens cultures and the G. metallireducens cultures resulted in nearly 84% and 67% conversion of ferrihydrite to magnetite, respectively, given the FeII/FeIII ratios in stoichiometric magnetite (FeIIFeIII2O4). During the same incubation period, vivianite was identified as the secondary mineralization product from ferrihydrite reduction in the presence of phosphate (Fig. 1B and 2B). Based on the stoichiometry of vivianite [Fe3(PO4)2·8H2O], a total phosphate concentration of 3.7 mM allowed the precipitation of 49% and 38% of the total Fe(II) as vivianite in the G. sulfurreducens cultures and G. metallireducens cultures, respectively.
X-ray diffraction (XRD) spectra of ferrihydrite incubated with Geobacter sulfurreducens in the absence (A) and presence (B) of phosphate (P). Acetate served as the electron donor. The iron oxide samples were taken after 20, 50, and 100 hours of anaerobic incubation.
XRD spectra of ferrihydrite incubated with Geobacter metallireducens in the absence (A) and presence (B) of phosphate (P). Ethanol served as the electron donor. The iron oxides samples were taken after 24, 72, and 168 hours of anaerobic incubation.
The impact of phosphate on the mineralization pathway of ferrihydrite has been studied extensively, particularly in Shewanella putrefaciens CN32 (7, 9). As observed consistently in many studies of CN32, magnetite was the dominant mineralization product in the absence of phosphate, whereas vivianite and green rust were generally formed in the reduction system in the presence of high concentrations of phosphate. Magnetite formation in this study was consistent with an earlier report of ferrihydrite reduction using G. metallireducens GS-15 (10). These results imply that magnetite is a very common terminal product during dissimilatory ferrihydrite reduction. Although the mineralization pathway or ferrihydrite is proposed to be highly governed by Fe(II) concentration (2, 4, 26), the mineralization pathways in this study did not appear to be controlled by Fe(II) concentration. The concentration of HCl-extractable Fe(II) in this study was similar to those obtained by ferrihydrite reduction in the presence and absence of 4 mM phosphate (1). Although the high concentrations of Fe(II) in both systems are conducive to magnetite precipitation, the strong surface complexation of phosphate with ferrihydrite might suppress the transformation of ferrihydrite to magnetite in the systems containing phosphate (27).
Response of methanogenic activity to mineralogical transformations of ferrihydrite.After ferrihydrite reduction, M. barkeri was added to initiate methanogenesis from acetate and ethanol. The coculture of G. sulfurreducens and M. barkeri initiated methane production from acetate without an apparent lag phase (Fig. 3A). In the cocultures with magnetite as the secondary mineralization product, the complete conversion of acetate to methane was achieved within 25 days, generating a total methane amount of 138.8 ± 9.6 μmol. With a longer incubation period (31 days), the cocultures with vivianite formation approached a steady methane level of 128.8 ± 6.9 μmol. The average methane production rates were calculated to be 4.1 and 5.6 μmol · day−1 in the cocultures undergoing ferrihydrite reduction in the presence and absence of phosphate, respectively. Compared with the control cocultures lacking ferrihydrite (4.3 μmol · day−1), ferrihydrite reduction to magnetite increased the methanogenesis rate by 30.2%, whereas ferrihydrite reduction to vivianite altered the methane production rate to a much smaller extent. Regardless of the methanogenic pathway, the stoichiometric conversion of acetate to CH4 is 1:1. In all methanogenic incubations, acetate was converted to methane at levels of 91.7 to 94.5%, indicating that acetate was completely converted to CH4 and that methanogenesis was the only terminal electron-accepting process in these cocultures. Acetate oxidation followed a pattern similar to that of methane production, and the higher acetate degradation rate was mirrored by a higher methane production rate (Fig. 3B).
Temporal changes in CH4 production (A) and acetate concentrations (B) (mean ± standard deviation [SD], n = 3) in G. sulfurreducens-M. barkeri cocultures metabolizing acetate to methane with secondary mineralization of ferrihydrite (Fh), in the absence and presence of phosphate (P).
Despite a long lag phase (29 days), the G. metallireducens-M. barkeri cocultures metabolized ethanol to CH4 (Fig. 4). In the presence of vivianite as the secondary mineral product, ethanol was slowly metabolized to CH4, with an average methane production rate of 2.0 μmol · day−1, comparable to that of the control cocultures without ferrihydrite (1.7 μmol · day−1). In the presence of magnetite that was produced as the main mineral product of ferrihydrite reduction, ethanol was converted to CH4 at a much faster average rate of 4.0 μmol · day−1, approaching a total methane amount of 170.2 ± 1.5 μmol. During ethanol degradation, acetate transiently accumulated to a maximum level of 1.2 ± 0.06 mM and was further metabolized with methane production (Fig. 4B). The total CH4 production was comparable to the 181.5 μmol of CH4 expected from the complete oxidation of ethanol to CH4 (2 CH3CH2OH → 3 CH4+ CO2), with 4.4 mM ethanol in 27.5 ml, suggesting that 93.8% of the electrons from ethanol metabolism were recovered in methane. Throughout the incubation period, there was no ethanol metabolism or CH4 production in the incubations with M. barkeri alone, which was expected because M. barkeri cannot utilize ethanol. The levels of HCl-extractable Fe(II) were relatively steady throughout the period of methanogenesis for both treatments (see Fig. S5 in the supplemental material), indicating the stability of the secondary minerals in the methanogenic cocultures. Similar results were observed when G. metallireducens and M. barkeri were incubated together at the beginning of the experiments with and without phosphate, and the average methane production rate in the phosphate-free cocultures was 84% higher than that in the phosphate-supplemented cocultures (see Fig. S6 in the supplemental material).
Temporal changes in CH4 production (A) and ethanol and acetate concentrations (B) (mean ± SD, n = 3) in G. metallireducens-M. barkeri cocultures metabolizing ethanol to methane with secondary mineralization of ferrihydrite (Fh), in the absence and presence of phosphate (P).
Our results suggested that methanogenic activities in the defined cocultures were linked to the secondary mineral products from microbial ferrihydrite reduction. The reaction of Fe(II) with ferrihydrite in the absence of phosphate yielded magnetite and facilitated methanogenesis from both acetate and ethanol, whereas vivianite biomineralization in the presence of phosphate had an insignificant influence on the methanogenic activities of the cocultured strains. It was noted that secondary mineralization of ferrihydrite had a larger effect on the conversion of ethanol to CH4 in the G. metallireducens-M. barkeri cocultures than on acetate metabolism to CH4 in the defined cocultures (see Fig. S7 in the supplemental material).
Proposed mechanism for facilitated methanogenesis.Syntrophic interactions are required to metabolize ethanol to CH4, because methanogens are restricted to utilizing only acetate, H2/CO2, or methanol for CH4 production (28). The best-known strategy for electron exchange in syntrophy is hydrogen interspecies electron transfer (29) or formate interspecies electron transfer (30, 31). Direct interspecies electron transfer (DIET) is a newly described alternative in which electrons are transferred from the electron-donating partner to the electron-accepting partner via biological electrical connections (32–37) or conductive materials (16–19, 22, 38–40). DIET has been considered more effective than hydrogen/formate interspecies electron transfer due to the elimination of multiple enzymatic steps involved in hydrogen production and consumption (41–43). Methanosaeta harundinacea (36) and M. barkeri (37) are the only methanogens known to be capable of DIET.
G. metallireducens does not produce H2 or formate when metabolizing ethanol to acetate, which eliminates the possibility of hydrogen/formate interspecies electron transfer in the coculture syntrophic metabolism (35). Alternatively, the syntrophy in the cocultures of G. metallireducens and M. barkeri metabolizing ethanol to CH4 might occur via DIET. The results suggested that magnetite biomineralization from ferrihydrite accelerated DIET-mediated methanogenesis, whereas vivianite biomineralization from ferrihydrite had an insignificant effect on syntrophic methanogenesis in the cocultures.
Magnetite-facilitated methanogenesis via DIET has been observed in anaerobic paddy soils and anaerobic digesters with acetate, ethanol (16, 17), propionate (17, 18), and butyrate (19). Kato et al. (16) first proposed that Geobacter species donate electrons directly to methanogens of the genus Methanosarcina and that they utilize magnetite as electron conduits, similar to the conductive pili of Geobacter species. This hypothesis was recently revised by the finding that magnetite can compensate for the electron transfer function of OmcS in promoting electrical contacts with pili (44). In the defined cocultures of G. metallireducens and G. sulfurreducens, magnetite compensated for the loss of OmcS in the OmcS-deficient strains but was not very effective in compensating for the loss of pili in the PilA-deficient strains. In contrast, conductive granular activated carbon (GAC) can promote DIET and compensate for the loss of both pili and OmcS involved in biological electrical connections (22). Although magnetite and GAC are both conductive and effective in promoting DIET, their functions in electron transfer between syntrophic partners might differ due to their particle sizes (GAC diameter, 1 to 2 mm; magnetite diameter, 20 to 50 nm).
Although syntrophy is not necessary in the cocultures of G. sulfurreducens and M. barkeri to metabolize acetate to CH4, magnetite-accelerated methanogenesis from acetate has been observed under mesophilic and thermophilic conditions (16, 17, 20, 40). Based on the advantage of Geobacter species utilizing acetate over acetoclastic methanogens (15), it was proposed that Geobacter oxidized acetate to CO2 and that electrons were transferred via magnetite-mediated DIET to methanogens to reduce CO2 to CH4 (16, 20, 40). The growth of Geobacter in the presence of magnetite was dependent on methanogenesis, which demonstrated the syntrophic association between Geobacter and methanogens (16, 40).
Here, we propose pilus-mediated DIET for interspecies electron transfer between G. metallireducens and M. barkeri metabolizing ethanol to CH4 (Fig. 5). Scanning electron microscopy (SEM) results showed close physical contact between the strains, as required for syntrophy via biological connections (Fig. 6A and B); magnetite was densely formed on the cell surface (Fig. 6A). For the G. sulfurreducens-M. barkeri cocultures, the cells were in close physical contact in the presence of magnetite (Fig. 6C), whereas the cells were dispersed on the surface of vivianite (Fig. 6D). Conductive nanoscale magnetite might promote DIET in a manner similar to that of c-type cytochrome and thus facilitate methanogenesis in syntrophic cocultures, whereas nonconductive vivianite does not contribute to stimulate DIET. Further isotope experiments and studies with the H2 interference method (18) are necessary to demonstrate DIET-based syntrophic acetate oxidation.
Proposed syntrophic mechanisms for the conversion of ethanol to methane in cocultures with different ferrihydrite mineralization pathways in the absence and presence of phosphate.
SEM images of cocultures of G. metallireducens and M. barkeri plus ferrihydrite (A), G. metallireducens and M. barkeri plus ferrihydrite and phosphate (B), G. sulfurreducens and M. barkeri plus ferrihydrite (C), and G. sulfurreducens and M. barkeri plus ferrihydrite and phosphate (D). Rod-shaped cells, Geobacter species; spherical cells, M. barkeri. Arrows, magnetite.
Magnetite facilitation of microbial methanogenesis in cocultures.Magnetite-mediated DIET has been directly demonstrated only in cocultures of Geobacter sulfurreducens and Thiobacillus denitrificans (45) and of G. metallireducens and G. sulfurreducens (44). To date, magnetite-mediated DIET has not been observed in defined syntrophic methanogenic cocultures. Here, we compared the methanogenic degradation of acetate and ethanol in cocultures of G. sulfurreducens and M. barkeri and of G. metallireducens and M. barkeri, respectively, in the presence and absence of magnetite. After a long lag phase (40 days), the cocultures of G. metallireducens and M. barkeri started to slowly metabolize ethanol to CH4 in the magnetite-free cocultures; however, the presence of magnetite significantly facilitated methanogenesis, in terms of ethanol degradation and methane production rates (Fig. 7). By the end of the 66-day incubation period, approximately 5.9 ± 0.57 and 92.1 ± 4.4 μmol CH4 accumulated in the magnetite-free and magnetite-supplemented cocultures, respectively. The cultures containing M. barkeri alone did not produce methane during the course of the entire incubation (data not shown). These results are consistent with previous findings that conductive GAC (22) and biochar (38) substantially reduced the adaptation period for the conversion of ethanol to CH4 by G. metallireducens-M. barkeri cocultures via DIET.
Temporal changes in CH4 production and ethanol concentrations (mean ± SD, n = 3) in G. metallireducens-M. barkeri cocultures metabolizing ethanol to methane in the presence and absence of magnetite.
For the acetate-fed incubations, no lag phase was observed in the magnetite-free cocultures, whereas methanogenesis in the presence of magnetite was inhibited for 16 days (Fig. 8A). The suppressive effect of magnetite on methanogenesis might be due to (i) outcompetition of M. barkeri by G. sulfurreducens for acetate, as indicated by Fe(II) production (see Fig. S8 in the supplemental material) and acetate consumption during the lag phase (Fig. 8B), or (ii) direct inhibition of M. barkeri by magnetite, which might increase the redox potential and thus lead to unfavorable conditions for M. barkeri growth. Although M. barkeri can itself reduce Fe(III), such as amorphous Fe(OH)3 (46, 47) and structural Fe(III) in nontronite (48), the ability of M. barkeri to reduce magnetite has not been reported. Thus, the suppression of methanogenesis is likely not due to the diversion of electron flow from CO2 reduction (methanogenesis) to Fe(III) reduction by M. barkeri. As the suppressive effect in the magnetite-supplemented cocultures was mitigated, CH4 accumulated at an average production rate of 8.5 μmol · day−1, 93.2% greater than that in the magnetite-free cocultures (4.4 μmol · day−1) (Fig. 8A). It seemed that Fe(III) reduction by G. sulfurreducens decreased the redox potential to a level that would trigger the initiation of methanogenesis (49). Although M. barkeri can reduce ferricyanide to decrease the redox potential (50), M. barkeri did not reduce magnetite to decrease the redox potential in this study. Thus, conversion of acetate to CH4 by M. barkeri was not observed in the presence of magnetite.
Temporal changes in CH4 production (A) and acetate concentrations (B) (mean ± SD, n = 3) in G. sulfurreducens-M. barkeri cocultures metabolizing acetate to methane in the presence and absence of magnetite.
In summary, using cocultures of Geobacter and Methanosarcina strains, this study demonstrated that (i) the transformation of iron minerals might exert an important influence on methanogenic activity under anaerobic conditions and (ii) both biogenic and chemical magnetite can accelerate DIET-based syntrophic ethanol oxidization between G. metallireducens and M. barkeri. Whether acetate is being metabolized via syntrophy in the cocultures of G. sulfurreducens and M. barkeri will be further examined using isotopically labeled acetate.
Environmental implications.Microbial reduction of Fe(III) oxide and iron mineralization play important roles in the fate and transport of toxic metals (51, 52) and organic contaminants (53) in anaerobic soils and sediments; however, the effect of iron biomineralization on microbial methanogenesis has not been considered. Using cocultures of Geobacter and Methanosarcina strains, which are representative dissimilatory Fe(III)-reducing bacteria and methanogenic archaea, respectively, this study demonstrated that the secondary iron minerals formed from microbial ferrihydrite reduction might play an important role in natural methane emission. When magnetite is formed as the predominant secondary iron mineral, syntrophic methanogenesis is accelerated by the establishment of DIET. Understanding the biogeochemical controls on microbial methanogenesis is important for predicting spatial and temporal patterns of methane emission into the atmosphere. This study sheds new light on the important role of iron biomineralization in biogeochemical cycling of carbon in diverse anaerobic environments, particularly in iron-rich natural and agricultural wetland soils.
ACKNOWLEDGMENTS
This study was funded by the National Natural Science Foundation of China (grants 31470561 and 41301256), Guangdong Natural Science Funds for Distinguished Young Scholars (grant S20120011151), and the Guangdong Provincial Science-Technology Project (grant 2015A020209089).
FOOTNOTES
- Received 18 May 2016.
- Accepted 20 July 2016.
- Accepted manuscript posted online 22 July 2016.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/AEM.01517-16.
- Copyright © 2016, American Society for Microbiology. All Rights Reserved.
