INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion References

Biomethylation of Hg²⁺ by sulfidogens in anoxic aquatic sediments (6) produces highly toxic methylmercury pollutants that have a tendency to accumulate in fish (8, 26). Somewhat paradoxically, in high-sulfate estuarine sediments, the rate of Hg²⁺ methylation is far lower than the rate of Hg²⁺ methylation in low-sulfate freshwater sediments (4, 7). The reason for this appears to be the generation of H₂S by the Hg²⁺-methylating sulfidogens, although the exact mechanism of the H₂S inhibition is uncertain at this time. Most oligotrophic freshwater lakes show no outward evidence of sulfidogenic activity, such as blackening or an H₂S odor in their sediments. We measured high rates of Hg²⁺ methylation in oligotrophic lake sediments that were free of detectable H₂S and evolved methane vigorously (19). As inhibition studies have consistently excluded methanogens and implicated sulfidogens in Hg²⁺ methylation (6, 7), one is led to question how sulfidogens stay active and methylate Hg²⁺ in environments that, because of likely sulfate limitation, seem to be inhospitable to them. Based on these considerations and on several reports of interspecies hydrogen transfers between sulfidogens and methanogens (1, 5, 20), we decided to explore the potential effect of such transfers on the methylation of Hg²⁺ under sulfate-limited conditions. We modeled the proposed interaction in vitro by using pure cultures of sulfidogens and a methanogen.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion References

Routine anaerobic Hungate techniques were used. Nitrogen gas was passed through a reduced heated copper column to remove traces of oxygen. Desulfovibrio desulfuricans LS was isolated in our laboratory (6). It was repurified by repeatedly picking single colonies from shake tubes containing diagnostic media B and E of Postgate (21) solidified with 1.5% agar. D. desulfuricans ND 132 was kindly donated by C. Gilmour (Academy of Natural Sciences of Philadelphia). It was repurified similarly. The cultures were maintained on medium C of Postgate (21) but were pregrown for experiments in medium D (21) modified by omitting FeSO₄. Incubation was at 37°C under oxygen-free 100% N₂ with slow shaking. After 24 h, 20-ml suspensions in 50-ml serum bottles were centrifuged at 3,200 × g to pellet the cells. The modified medium D was removed from the inverted bottles with a syringe and replaced by 10 ml of a specially formulated coculture medium. This medium contained (per liter) 0.33 g of KCl, 2.75 g of MgCl₂ · 7H₂O, 0.25 g of NH₄Cl, 0.14 g of CaCl₂ · 2H₂O, 0.14 g of K₂HPO₄, 18 g of NaCl, 5 g of NaHCO₃, 0.25 g of yeast extract, 0.25 g of cysteine, 1.0 g of resazurin, 0.1 g of sodium ascorbate, 0.1 g of thioglycolate, 6.0 g of sodium lactate, and the trace elements in ATCC medium 1043 (2) modified by replacing the ZnSO₄ · 7H₂O with 0.1 g of ZnCl₂ per liter and adding 0.5 g of ferric ammonium citrate per liter. This coculture medium always contained a 1-µg/ml spike of HgCl₂ and was routinely incubated at 37°C under 100% N₂ with slow shaking. The D. desulfuricans cell pellets were suspended in 10 ml of this coculture medium with shaking, and 1-ml portions were used to inoculate replicate 50-ml bottles containing 19 ml of coculture medium.

Methanococcus maripaludis ATCC 43000 (2) was kindly provided by W. B. Whitman, University of Georgia, Athens. It was maintained and pregrown for experiments in ATCC medium 1439 (2) under 80% H₂-20% CO₂ with slow shaking at 37°C. After 24 h, the cells were pelleted and ATCC medium 1439 was replaced with 10 ml of coculture medium, as described above for D. desulfuricans LS. One-milliliter portions of the resuspended cells were used to inoculate a series of 50-ml bottles containing 19 ml of coculture medium. In addition, a series of bottles containing 18 ml of coculture medium were each inoculated with 1 ml of D. desulfuricans and 1 ml of M. maripaludis. All of the bottles were incubated at 37°C under 100% N₂ with slow shaking. At appropriate times, bottles were removed from the incubator, 2.0 ml of a 1.0 M CuSO₄ solution was added to each bottle to stop the reaction, and the vials were kept at 20°C until analysis.

For experiments performed with resting cells, the D. desulfuricans strains were pregrown in modified (sulfate-free) medium D (21) as described above for the coculture experiments. The cells were sedimented in the serum bottles by centrifugation, washed once with coculture medium, and suspended to the original volume in coculture medium containing 1.0 µg of HgCl₂ per ml.

Prior to analysis, the bottles were thawed and mixed by shaking. Five milliliters was removed from each of the bottles, and the cells were centrifuged and used for protein determination (17). The remaining 15 ml was analyzed for methylmercury by the procedure of Longbottom et al. (13) by using a Hewlett-Packard model 5890 gas chromatograph equipped with a macrobore capillary column (inside diameter, 0.53 mm; length, 15 m: type AT-35; Alltech, Deerfield, Ill.). The operating conditions were as follows: the carrier gas was Ar-CH₄ (95:5, vol/vol; ultrapure grade; Matheson Gas Products, East Rutherford, N.J.) at a flow rate of 35 ml/min, the injector temperature was 150°C, and the electron capture detector temperature was 250°C. Monomethylmercury peak (retention time, 1.25 min) areas were recorded with a Hewlett-Packard model 3392A integrator that was calibrated by using monomethylmercury (CH₃HgI) standards (American Tokyo Kasei, Inc., Portland, Oreg.). The detection limit for monomethylmercury was 1 ng/ml. The procedure of Longbottom et al. (13) analyzes all methylmercury as monomethylmercury, and the synthesis of CH₃Hg²⁺ was plotted over time. In order to make comparisons between experiments easier, the rates of methylmercury synthesis were normalized to the initial protein concentrations of the inocula. The data for time points given below represent single determinations.

Methane determinations in the headspaces of bottles were made with a model 1200 gas partitioner (Fisher Co., Springfield, N.J.) operated at 50°C with helium as the carrier gas at a flow rate of 30 ml/min. The sample volume analyzed was 250 µl, and the instrument was calibrated with methane standards (Fisher). Microscopic observations were made with a Reichert Zetopan microscope at a magnification of ×1,000.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion References

As shown in Fig. 1A, neither D. desulfuricans LS nor M. maripaludis was able to grow individually in the coculture medium, since the organic substrate (lactate) could be used only by the former bacterium and the electron acceptor (carbonate) could be used only by the latter bacterium. However, in the coculture of the two bacteria, the protein level increased 15-fold in 8 days, from 3 to 45 µg/ml.

View larger version (24K): [in this window] [in a new window]   FIG. 1.   (A and B) Protein contents of D. desulfuricans LS and M. maripaludis cultures and a coculture in lactate-carbonate coculture medium (A) and synthesis of CH3Hg+ (per microgram of initial protein) from a 1-µg/ml HgCl2 spike in the same experiment (B). (C) Protein contents of D. desulfuricans ND 132 and M. maripaludis cultures and a coculture in lactate-carbonate coculture medium (C) and synthesis of CH3Hg+ (per microgram of initial protein) from a 1-µg/ml HgCl2 spike in the same experiment (D). Symbols: , D. desulfuricans culture; , M. maripaludis culture; , coculture.

The morphologies of the two bacteria are quite distinct. M. maripaludis cells are large cocci that are of 2 to 2.5 µm in diameter, while D. desulfuricans cells are motile curved rods that are 1.0 to 1.5 µm long. Microscopic observations of the coculture showed that the two types of cells multiplied proportionally and roughly equal numbers of the two types were produced (results not shown).

The methylmercury synthesis pattern was similar to the protein production pattern (Fig. 1B). M. maripaludis alone failed to form any methylmercury, and D. desulfuricans LS alone formed only a trace amount, but the coculture synthesized 22 ng of methylmercury per µg of initial protein in 8 days, methylating 2.6% of the available Hg²⁺. This rate was very similar to the rate of synthesis of methylmercury by D. desulfuricans LS in a sulfate-free pyruvate medium in which this sulfidogen can grow fermentatively without an electron acceptor (6, 19, 21). In this medium the typical D. desulfuricans LS protein yield was 30 µg/ml. M. maripaludis was unable to synthesize any methylmercury even in a complete methanogen medium (2) in which the typical M. maripaludis protein yield is 38 µg/ml (19).

The coculture experiment was also performed with D. desulfuricans ND 132, and the results were only slightly different. This strain was selected for its exceptional ability to methylate mercury. It also appeared to have a limited ability to use lactate fermentatively in the absence of sulfate or to grow on the amino acids in yeast extract, since the protein concentration of the D. desulfuricans ND 132 culture approximately doubled (Fig. 1C). As in the experiment described above, the protein concentration in the coculture increased to about 50 µg/ml. The coculture methylated large amounts of mercury, but, in contrast to D. desulfuricans LS, D. desulfuricans ND 132 alone methylated as much as two-thirds of the mercury methylated by the coculture (Fig. 1D). These experiments were repeated several times to ascertain that the difference observed was strain specific and reproducible.

D. desulfuricans LS and ND-132 belong to the "incomplete oxidizer" subgroup of sulfidogens (21, 25). In the presence of sulfate, they oxidize lactate to pyruvate, and the latter yields CO₂, acetate, and reducing equivalents for SO₄² reduction. D. desulfuricans strains are unable to use acetate (6), but M. maripaludis, like many other methanogens, can use acetate as a substrate for methanogenesis and growth (2). Thus, D. desulfuricans strains, in addition to transferring lactate hydrogens to M. maripaludis for CO₂ reduction, also benefited the methanogen by producing acetate as a methanogenic substrate. In turn, the removal of H₂ and acetate by the methanogen allowed D. desulfuricans strains to utilize lactate even in the absence of SO₄².

The experiments described above were performed with a 5% inoculum, and mercury methylation occurred during growth. To determine whether preformed D. desulfuricans biomass could methylate mercury under nongrowth conditions, we repeated the control experiments with the two D. desulfuricans strains in coculture medium but with a 100% inoculum. As shown in Fig. 2, D. desulfuricans LS failed to grow or to form any methylmercury under these conditions. The protein content of D. desulfuricans ND 132, which showed a limited ability to grow in the absence of sulfate, doubled, but this organism synthesized only a trace amount of methylmercury. It may be concluded that growth conditions are necessary for mercury methylation.

View larger version (15K): [in this window] [in a new window]   FIG. 2.   (A) Protein contents of D. desulfuricans LS and D. desulfuricans ND 132 cultures in lactate-carbonate coculture medium without M. maripaludis. (B) Synthesis of CH3Hg+ (per microgram of initial protein) from a 1-µg/ml HgCl2 spike in the same experiment. Note the large inoculum and essentially nongrowth conditions. Symbols: , D. desulfuricans LS culture; , D. desulfuricans ND 132 culture.

The effects of selective inhibitors of methanogenesis and sulfate reduction (18) on the coculture of D. desulfuricans LS and M. maripaludis are shown in Fig. 3. Bromoethanesulfonic acid (BESA) at a concentration of 0.5 mM allowed only marginal protein synthesis by the coculture (Fig. 3A), but it allowed almost normal mercury methylation by the coculture (Fig. 3B). Molybdate (2 mM) prevented both growth of and mercury methylation by the coculture (Fig. 3A and B). The methylation of mercury observed in the BESA-inhibited coculture was unexpected, since M. maripaludis was completely inhibited by 0.2 mM BESA when it was grown alone in methanogen medium (results not shown). A possible explanation for this apparent paradox is that under sulfate-limited conditions, some sulfidogens are capable of using BESA as alternate electron acceptor (18), making cooperation with the methanogen unnecessary. That M. maripaludis was in fact inhibited by BESA in the coculture was evident from the effect of BESA on methanogenesis (Fig. 3C). Methane evolution was completely suppressed by the inhibitor, while in 7 days the control coculture produced 1,200 µl of methane per µg of initial protein in the inoculum. The results of an experiment to determine whether BESA in fact serves as an electron acceptor for D. desulfuricans LS are shown in Fig. 4. Inoculated alone into coculture medium, this strain failed to grow or to methylate mercury. In the presence of BESA, however, it methylated one-half as much mercury as it methylated when it was grown fermentatively on pyruvate without an electron acceptor. However, the rate of protein synthesis was low when BESA served as an electron acceptor compared to the rate when the organism was grown in medium D. D. desulfuricans ND 132 behaved in a similar manner (results not shown).

View larger version (15K): [in this window] [in a new window]   FIG. 3.   (A) Protein contents of a D. desulfuricans LS-M. maripaludis coculture in lactate-carbonate coculture medium, in coculture medium containing 0.5 mM BESA, and in coculture medium containing 2.0 mM molybdate. (B) Synthesis of CH3Hg+ (per microgram of initial protein) in the same experiment. (C) Evolution of CH4 (per microgram of initial protein) in the same experiment. Symbols: , coculture medium; , coculture medium containing 0.5 mM BESA; , coculture medium containing 2.0 mM molybdate.

View larger version (19K): [in this window] [in a new window]   FIG. 4.   (A) Protein contents of D. desulfuricans LS cultures inoculated into lactate-carbonate coculture medium with and without 0.5 mM BESA as the electron acceptor. The protein content of a culture in low-sulfate pyruvate medium D served as a positive control. (B) Synthesis of CH3Hg+ (per microgram of initial protein) in the same experiment. Symbols: , coculture medium containing 0.5 mM BESA; , coculture medium not containing BESA; , medium D.

The utilization of BESA as an alternate electron acceptor by sulfidogens presents a future methodological challenge for demonstrating interspecies electron transfer-mediated mercury methylation in sulfate-limited sediments. A straightforward approach would be to inhibit methanogens in such sediments by using BESA, expecting this to inhibit mercury methylation. However, since BESA serves as an alternate electron acceptor for sulfidogens, mercury methylation is likely to remain unaffected, even when the methanogens are inhibited.

In anoxic sediments, the microbial community preferentially uses the electron acceptor that maximizes the energy yield from the available electron donors (3). Thus, there is usually little methanogenesis from substrates utilizable by both groups until sulfate becomes limiting to sulfidogenic activity. There is no agreement in the literature concerning the sulfate concentration at which sulfate becomes limiting for the activity of sulfate reducers. On the high end, Ingvorsen et al. (10) described 300 µM as the limiting concentration, while other workers (16, 22, 24) have claimed that levels as low as 20 to 30 µM are limiting. Several authors have reported values between these extremes (23, 27). The way that sulfate reduction is measured (i.e., by spiking sediments with ³⁵SO₄²) is relevant to this controversy. In sulfate-rich marine and estuarine sediments, the spike adds an insignificant amount to the total sulfate concentration, but when the sulfate concentration is near the limiting concentration, the spike can significantly alter the sulfate pool (11). If this is the case, the measurement does not reflect the in situ rate accurately but instead gives a value between the actual in situ rate and the in situ potential. In addition to this methodological difficulty, obviously the quantity and quality of the available organic substrates, the temperature, the pH, and other environmental factors may influence the sulfate concentration at which sulfate becomes limiting.

Regardless of the exact inhibitory limit, it may be assumed that at least partial sulfate limitation occurs when anoxic sediments evolve methane vigorously, indicating that much of the energy flow passes through the methanogens. This is typical of many oligotrophic freshwater lakes. Most methanogens can use CO₂ as an electron acceptor, but their substrate ranges are extremely limited and they have higher K_s values for H₂ and acetate than sulfidogens have (12, 14-16, 27). In sulfate-limited situations, sulfidogens may still have primary access to the organic substrates but can gain energy from them only by passing on hydrogen and acetate to methanogens and in this manner utilize the only remaining electron sink (CO₂). Such interspecies hydrogen transfer has been demonstrated previously in several sulfidogen-methanogen cocultures (1, 5, 20). Our contribution explores the potential significance of this phenomenon with respect to mercury methylation.

Hg²⁺ is methylated principally by sulfidogens (6), yet at sulfate levels above 200 µM the product of the normal metabolism of these organisms (H₂S) inhibits mercury methylation (4, 7, 9). Therefore, situations that allow sulfidogens to be active without sulfate reduction are very favorable for Hg²⁺ methylation. One such situation is the growth of sulfidogens on a fermentable substrate like pyruvate (6, 21). Another one, which has potentially broader ecological and environmental significance, may be the interspecies transfer of hydrogen and acetate from sulfidogens to methanogens.

We do not wish to imply that interspecies hydrogen transfer is an exclusive mechanism for Hg²⁺ methylation; it merely represents a very favorable opportunity. It has been demonstrated that at concentrations just above the limiting concentration, sulfate additions to a lake sediment increase Hg²⁺ methylation (9). However, at SO₄² concentrations of more than 200 µM, accumulating H₂S apparently starts to interfere with CH₃Hg⁺ synthesis. As the window for sulfate-reducing activity without inhibition of Hg²⁺ methylation by H₂S seems to be narrow, there is reason to believe that Hg²⁺ methylation via interspecies hydrogen and acetate transfer has environmental significance.