Article pubs.acs.org/est
Anaerobic Mercury Methylation and Demethylation by Geobacter bemidjiensis Bem Xia Lu,*,†,‡ Yurong Liu,‡,§ Alexander Johs,‡ Linduo Zhao,‡ Tieshan Wang,† Ziming Yang,‡ Hui Lin,‡ Dwayne A. Elias,∥ Eric M. Pierce,‡ Liyuan Liang,‡,⊥ Tamar Barkay,∇ and Baohua Gu*,‡ †
School of Nuclear Science and Technology, Lanzhou University, Lanzhou, China Environmental Sciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, United States § State Key Laboratory of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China ∥ Biosciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, United States ⊥ Biology and Soft Matter Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, United States ∇ Department of Biochemistry and Microbiology, Rutgers University, New Brunswick, New Jersey 08901, United States ‡
S Supporting Information *
ABSTRACT: Microbial methylation and demethylation are two competing processes controlling the net production and bioaccumulation of neurotoxic methylmercury (MeHg) in natural ecosystems. Although mercury (Hg) methylation by anaerobic microorganisms and demethylation by aerobic Hg-resistant bacteria have both been extensively studied, little attention has been given to MeHg degradation by anaerobic bacteria, particularly the iron-reducing bacterium Geobacter bemidjiensis Bem. Here we report, for the first time, that the strain G. bemidjiensis Bem can mediate a suite of Hg transformations, including Hg(II) reduction, Hg(0) oxidation, MeHg production and degradation under anoxic conditions. Results suggest that G. bemidjiensis utilizes a reductive demethylation pathway to degrade MeHg, with elemental Hg(0) as the major reaction product, possibly due to the presence of genes encoding homologues of an organomercurial lyase (MerB) and a mercuric reductase (MerA). In addition, the cells can strongly sorb Hg(II) and MeHg, reduce or oxidize Hg, resulting in both time and concentration-dependent Hg species transformations. Moderate concentrations (10−500 μM) of Hg-binding ligands such as cysteine enhance Hg(II) methylation but inhibit MeHg degradation. These findings indicate a cycle of Hg methylation and demethylation among anaerobic bacteria, thereby influencing net MeHg production in anoxic water and sediments.
■
INTRODUCTION Microbial methylation, which converts inorganic mercury (Hg) to bioaccumulative neurotoxic methylmercury (MeHg) in anoxic environments, significantly affects the fate and toxicity of Hg in natural systems.1−6 Important progress has recently been made in understanding the mechanisms and microbial communities responsible for Hg(II) methylation.2,5−7 Many species, including sulfate- and iron-reducing bacteria, methanogens, and syntrophic, acetogenic, and fermentative Firmicutes, were confirmed to possess the gene pair, hgcAB, which is essential for methylating inorganic Hg(II).2,7,8 In comparison, microbial demethylation, a process considered as a major pathway of MeHg degradation below the photolytic zone, has been much less studied in these microorganisms, particularly in anaerobic settings where MeHg is produced.9−11 Two microbial demethylation mechanisms have been described and are distinguished by their degradation products: Reductive demethylation produces elemental Hg(0) and CH4 as the major products,11,12 whereas CO2 and Hg(II) are the main products of oxidative demethylation.9 Reductive demethylation is generally favored at high Hg concentrations (e.g., at μM) © XXXX American Chemical Society
under oxic conditions, whereas oxidative demethylation can occur at low Hg concentrations (e.g., at nM) in anoxic environments.3,13−15 Much attention has been focused on reductive demethylation induced by the mer operon in aerobic organisms as a detoxification mechanism.11−13 The mer operon encodes a series of genes involved in Hg detoxification. Of particular importance is the sequential action of two enzymes: organomercurial lyase (MerB), which cleaves the C−Hg bond to yield CH4 and Hg(II), and mercuric reductase (MerA), which reduces Hg(II) to gaseous Hg(0). This process has often been attributed to Hg-resistant aerobic11−13,16 and facultative anaerobic bacteria.17,18 The discovery that slurries of anoxic sediments could also degrade MeHg by forming CO2 and Hg(II) revealed another mechanism for demethylation, possibly as a cometabolic byproduct of methylotrophic metabolism.9 Demethylation Received: January 25, 2016 Revised: March 24, 2016 Accepted: March 28, 2016
A
DOI: 10.1021/acs.est.6b00401 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology assays with metabolic inhibitors,9,15,19 as well as the limited number of pure culture experiments,20,21 have attributed this oxidative demethylation to anaerobic sulfate reducers and methanogens. However, no studies to date have reported demethylation involving the mer operon in anaerobic bacteria such as iron-reducers in anoxic environments. We examined Hg(II) methylation and MeHg degradation as potential detoxification mechanisms of the iron reducing bacterium Geobacter bemidjiensis Bem. This Geobacter strain was first enriched from subsurface sediments collected in Bemidji, Minnesota, United States.22 The genome sequence indicates that it possesses both homologues of the Hg(II) methylation gene pair hgcAB and the putative Hg-resistance genes merB and merA.2,7,23 Previous studies have also identified the G. bemidjiensis strain as a strong Hg(II) methylator in an iron-based culture medium amended with a high concentration of cysteine,7 but its ability to degrade MeHg under anoxic conditions has never been observed. Furthermore, the mechanisms and factors (such as the presence or absence of complexing ligands) that govern Hg(II) methylation or MeHg degradation by the G. bemidjiensis strain are currently unknown.
At each selected time point, 4−6 replicate sample vials were taken out of the anoxic chamber and immediately analyzed for purgeable elemental Hg(0). Half of the purged samples were filtered through 0.2-μm syringe filters (to remove cells) and analyzed for soluble MeHg (MeHgsol) and Hg(II) (Hgsol). The other half (without filtration) was analyzed for total MeHg (MeHgtot) and nonpurgeable Hg(II) (HgNP). All purged samples (with and without filtration) were then preserved in HCl [0.5% (v/v)] at 4 °C, and an aliquot (0.2−0.4 mL) was taken and analyzed for MeHg, and the remaining aliquot was oxidized overnight in BrCl [5% (v/v)] at 4 °C and analyzed for Hg(II) [including both MeHg and inorganic Hg(II)]. Total Hg (HgT) was determined by summation of the purgeable Hg(0) and nonpurgeable HgNP. The cell-associated MeHg (MeHgcell) was estimated by the difference between total and soluble MeHg, and similarly for the cell-associated Hg(II) (Hgcell). The soluble and cell-associated inorganic Hg(II) (IHgsol or IHgcell) were calculated by subtracting MeHg (MeHgsol or MeHgcell) from Hg(II) (Hgsol or Hgcell).27,28 Based on observed data, the initial rate constant (kmeth) for MeHg production was calculated using the first-order rate law: d[IHg]/dt = −d[MeHg]/dt = −kmeth[IHg], where IHg equals (THg − MeHg), and kmeth was determined by the slope of the linear regression between the natural logarithm of the IHg concentration and time.6,28 The MeHg concentration decreased linearly after reaching its maximum at ∼8 to 24 h, and the observed demethylation rate constant (kde) was thus estimated by the slope of the linear regression between MeHg concentration and time. Each Hg(II) methylation assay or MeHg degradation assay was repeated at least once to ensure data quality, and all data plotted in Figures 1−5 represent an average of all replicate samples. Potential loss of Hg(II) or MeHg was usually within ±10%, determined from measurements of the total Hg recovery. Error bars represent one standard deviation of replicate samples from different batch experiments. Hg and MeHg Analyses. The purgeable Hg(0) was analyzed by purging dissolved gaseous Hg(0) directly from cell suspensions with ultrapure N2 for 2 min into a Hg(0) analyzer (detection limit ∼2.5 × 10−4 nM) (Lumex RA-915+, Ohio Lumex).5,6,29 Nonpurgeable Hg(II) (after BrCl oxidation) was determined via SnCl2 reduction, gold-trap amalgamation, and detection with the Lumex Hg(0) analyzer. MeHg was analyzed using a modified EPA Method 1630,5,6,27 in which enriched Me200Hg was added as an internal standard. MeHg was then extracted from the sample matrix via distillation, ethylation, and trapping on a Tenax column via N2-purging. Thermal desorption and separation by gas chromatography were applied prior to the detection of Hg by an inductively coupled plasma mass spectrometer (ICP-MS) (Elan-DRCe, PerkinElmer, Inc., Shelton, CT). The detection limit was about 3 × 10−5 nM MeHg.
■
MATERIALS AND METHODS Bacterial Culture and Assay Conditions. Geobacter bemidjiensis Bem (DSM16622) was obtained from the Deutsche Sammlung von Mikroorganismen and Zellkulturen GmbH (DSMZ, Braunschweig, Germany). It was first cultured in DSMZ Medium 579 containing 55 mM ferric citrate and 30 mM acetate at pH 6.8 at 30 °C.7,22 Subcultures were stored at −80 °C. For the methylation and demethylation assays, new cultures were prepared (with ∼6% inoculum) in a modified medium containing 40 mM fumarate (replacing ferric citrate) as an electron acceptor, to avoid iron precipitation during incubation.24,25 Cells were harvested at early stationary phase (with optical density at ∼0.4 to 0.5) and washed three times by repeated centrifugation (at 1200g, 5 min, 25 °C) and resuspension in a deoxygenated 3-(N-morpholino)propanesulfonate (MOPS) buffer solution (pH 6.8), consisting of 10 mM MOPS, 0.17 mM NaCl, 1.3 mM KCl, 5 mM NaH2PO4, 0.15 mM MgSO4, 0.1 mM NH4Cl, and 1 mM sodium acetate.26 The buffer was autoclaved and then deoxygenated by boiling and purging with ultrahigh purity N2 gas, and subsequently kept in an anoxic glove chamber (Coy) (98% N2 and 2% H2) before use. All washing steps and subsequent Hg assays were conducted in the anaerobic chamber. Hg(II) Methylation and MeHg Degradation Assays. All methylation and demethylation assays were conducted in MOPS buffer in 4 mL amber glass vials (National Scientific) in the anoxic chamber. For the Hg(II) methylation assay, Hg(II) stock solution (50 μM HgCl2 in 1% HCl) was added to yield a final concentration of ∼25 nM Hg(II). Similarly, MeHg stock solution (5 μM MeHgOH in 0.5% HOAc, 0.2% HCl, Brooks Rand Laboratories) was added to yield a final concentration of 5 nM in MeHg degradation assays. For the concentrationdependent studies, various Hg(II) or MeHg concentrations from 1 to 125 nM were used. The final cell density was 5 × 108 cell mL−1. Sodium fumarate (1 mM final) was added once at time zero as the electron acceptor during all assays. All vials were immediately sealed with PTFE/silicone screw caps (Kimble/Kontes) and kept in the dark on an orbital shaker. Parallel assays with heat-killed cells (1.5 h at 80 °C) were conducted as controls.
■
RESULTS AND DISCUSSION Hg(II) Methylation and MeHg Degradation by G. bemidjiensis Bem. Results of the time-dependent Hg(II) methylation assay indicate that inorganic Hg(II) (at 25 nM) can be methylated rapidly by washed cells of G. bemidjiensis in MOPS buffer (Figure 1a). The production of MeHg increased exponentially within the first few hours and reached to a maximum of ∼1.2 nM [or ∼5% of the added Hg(II)] in 8 h. The initial methylation rate constant (within 8 h), calculated based on the first-order rate law, was 0.9 ± 0.2 h−1. The B
DOI: 10.1021/acs.est.6b00401 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
8-120 h, was attributed to microbial demethylation, as detailed below, and followed a zero-order with a calculated slope or rate constant of 5.4(±0.3) × 10−3 nM h−1. This demethylation rate is similar to those observed in pure culture studies with sulfatereducing bacteria and methanogens under anoxic conditions.20,21 To confirm the ability of G. bemidjiensis cells to degrade MeHg, additional experiments were performed, in which only MeHg [no inorganic Hg(II)] was added to the cell suspension to give a final concentration of 5 nM MeHg. MeHg concentrations decreased consistently, and approximately 60% of the added MeHg was degraded over the 5-day incubation period (Figure 2a). In comparison, no appreciable degradation
Figure 1. Hg(II) methylation assays with washed cells of G. bemidjiensis Bem (5 × 108 cell mL−1) in MOPS buffer (pH 6.8) amended with 1 mM fumarate and 1 mM acetate. (a) Net methylmercury (MeHg) production, and (b) Hg(II) reduction to elemental Hg(0) and mass balance analysis (total Hg, HgT) during the 5-day incubation. The initial added Hg(II) concentration was 25 nM. Data points represent an average of >3 independent batch experiments, and error bars represent one standard deviation from 6−9 replicate samples.
maximum amount of Hg(II) methylated is similar to that observed with washed cells of G. sulf urreducens PCA strain (5%),5,6 but much lower than that reported with the same G. bemidjiensis Bem in a previous study,7 in which methylation was performed directly in an Fe(III)-based culture media and with a high cysteine concentration (500 μM). Concurrent with Hg(II) methylation, a large fraction (38%) of the added Hg(II) was reduced to elemental Hg(0) by the cells in the first 4−8 h (Figure 1b), since G. bemidjiensis is a known iron-reducing bacterium.22,25 However, the Hg(0) concentration decreased gradually after 24 h (Figure 1b) from ∼8.7 nM at 24 h to 4.3 nM after 5 days. Removal of Hg(0) during this period was not due to the loss of Hg, as evidenced by the THg analysis, but suggests that Hg(0) was reoxidized by G. bemidjiensis cells, consistent with observations of cell thiol-induced oxidation of Hg(0) under anaerobic conditions.5,6,30 In heat-killed cells, neither Hg(II) reduction nor methylation were observed, as expected (Figure 1). Importantly, the biosynthesized MeHg also decreased linearly after reaching its maximum at 8 h; nearly 50% of the synthesized MeHg was lost or degraded after 5 days (Figure 1a). This loss of MeHg cannot be attributed to sorption or volatilization since a good mass balance was obtained by analyzing both the gaseous Hg(0) and the total Hg in the system (Figure 1b). The loss of MeHg, observed between
Figure 2. Methylmercury (MeHg) degradation assays with washed cells of G. bemidjiensis Bem (5 × 108 cell mL−1) in MOPS buffer (pH 6.8) amended with 1 mM fumarate and 1 mM acetate. (a) MeHg degradation, and (b) Hg(0) production and mass balance analysis (total Hg, HgT) during the 5-day incubation. The initial added MeHg concentration was 5 nM. Data points represent an average of >3 independent batch experiments, and error bars represent one standard deviation from 6−9 replicate samples.
of MeHg was observed with heat-killed cells (Figure 2a), suggesting that demethylation is biologically mediated. The estimated demethylation rate constant was 25.4(±0.6) × 10−3 nM h−1, which is about 5 times higher than that observed in the Hg(II) methylation assay (Figure 1a) because a higher MeHg concentration (5 nM) was added in the demethylation assay than it was formed in the methylation assay (∼1.2 nM). A short delay in MeHg degradation was also observed since the measured MeHg concentration is a net balance between MeHg production and degradation. As shown in Figure 1a, methylation can dominate initially after the addition of C
DOI: 10.1021/acs.est.6b00401 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 3. Mercury species distribution, expressed as the percentage of the measured total Hg (HgT), during (a) Hg(II) methylation and (b) methylmercury (MeHg) degradation assays with washed cells of G. bemidjiensis Bem (5 × 108 cell mL−1). Soluble inorganic Hg(II) (IHgsol) or MeHg (MeHgsol) were analyzed after removing cells with the 0.2 μm syringe filter, whereas the cell sorbed or associated inorganic Hg(II) (IHgcell) or MeHg (MeHgcell) were determined by the difference between nonpurgeable Hg (HgNP) or total MeHg and the soluble IHgsol or MeHgsol. Hg(0) = Elemental Hg (see Figures 1 and 2 for additional details). Data points represent an average of two independent batch experiments, and error bars represent one standard deviation from 4−6 replicate samples.
Figure 4. Concentration- and time-dependent Hg(II) methylation (a) and cumulative methylmercury (MeHg) degradation (b) with washed cells of G. bemidjiensis Bem (5 × 108 cell mL−1) in MOPS buffer (pH 6.8). Inset in (a) shows percentages of MeHg production with incubation time, and inset in (b) shows percentages of MeHg degradation. Data points represent an average of two independent batch experiments, and error bars represent one standard deviation from 4−6 replicate samples.
inorganic Hg(II), whereas demethylation commences after significant amounts of MeHg build up in the system. Demethylation by G. bemidjiensis cells is also evidenced by analysis of the reaction products or the production of elemental Hg(0) in the course of MeHg degradation assay (Figure 2b). At 24 and 48 h, approximately 0.4(±0.1) and 0.7(±0.4) nM of Hg(0) were produced, coinciding with the degradation of 0.4(±0.1) and 0.8(±0.2) nM of MeHg at 24 and 48 h (Figure 2a), respectively. In other words, MeHg appeared to be quantitatively converted to Hg(0) by the cells since no inorganic Hg(II) was added. However, the Hg(0) production rate decreased over time and, at day 5, the amount of Hg(0) produced accounted for only 56% of the MeHg degraded. This
imbalance between the amounts of MeHg degraded and Hg(0) produced is attributed in part to the ability of G. bemidjiensis cells to oxidize Hg(0), which is consistent with anaerobic cell thiol-induced oxidation of Hg(0), as described in detail elsewhere.5,6,30 However, we cannot rule out the possibility of oxidative demethylation,9 although the near complete conversion of MeHg to Hg(0) in the first 48 h (Figure 2) suggests that reductive demethylation may be the dominant pathway. To further elucidate the complex interactions and processes of G. bemidjiensis cells in Hg(II) methylation, reduction or oxidation, and MeHg degradation, we examined Hg species transformation and distribution in both Hg(II) methylation and MeHg degradation assays (Figure 3). When inorganic Hg(II) D
DOI: 10.1021/acs.est.6b00401 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 5. Effects of cysteine concentrations on (a) Hg(II) methylation and (b) methylmercury (MeHg) degradation with washed cells of G. bemidjiensis Bem (5 × 108 cell mL−1) at various incubation times (0, 4, 24, and 120 h). The initial added Hg(II) concentration was 25 nM (a), and the MeHg concentration was 5 nM (b). Data points represent an average of two independent batch experiments, and error bars represent one standard deviation from 4−6 replicate samples.
lation (with no inorganic Hg(II) added). Results (Figure 4b) show that the total amount of MeHg degraded increased with time and with increasing MeHg concentrations from 1 to 125 nM. This result is consistent with a mer-mediated detoxification mechanism.11,33 However, when calculated based on the ratio or percentage of MeHg degradation, we found an optimal MeHg concentration of ∼5−25 nM (Figure 4b inset), either above or below which the percentages of the degraded MeHg decreased, resulting in a bell-shaped concentration-dependent MeHg degradation. For example, at 1 nM MeHg, only about 31% of the MeHg was degraded after 5 days, whereas ∼3 nM or 60% of the MeHg was degraded at an initial MeHg concentration of 5 nM. At high MeHg concentration (125 nM), a total of 29.2 nM MeHg was degraded in 5 days, corresponding to only 23% of the total MeHg added. Again this result may be attributed in part to potential toxic effects of high MeHg concentrations and to simultaneous MeHg production by the organism as MeHg is converted to inorganic Hg. Similarly to the situation described above (Figure 2b), we observed production of elemental Hg(0) during the course of MeHg degradation, although the amount of Hg(0) produced varied with reaction time and with the concentration of added MeHg (SI Figure S2). As expected, total Hg(0) production increased with added MeHg concentration and coincided with the amount of MeHg degraded. In particular, nearly 100% of the degraded MeHg was converted to Hg(0) at 24 h when a low amount (
