Environ. Sci. Technol. 2008, 42, 9171–9177
Mercury Stable Isotope Fractionation during Reduction of Hg(II) by Different Microbial Pathways K . K R I T E E , * ,† J O E L D . B L U M , ‡ A N D TAMAR BARKAY† Rutgers University, 76 Lipman Drive, New Brunswick, New Jersey 08901, and University of Michigan, 1100 N. University Avenue, Ann Arbor, Michigan 48109
Received June 10, 2008. Revised manuscript received September 25, 2008. Accepted October 2, 2008.
Mercury (Hg) stable isotope fractionation has recently been developed as a tool in biogeochemistry. In this study, the extent of Hg stable isotope fractionation during reduction of ionic mercury [Hg(II)] by two Hg(II)-resistant strains, Bacillus cereus 5 and the thermophile Anoxybacillus sp. FB9 [which actively detoxify Hg(II) by the mer system] and a Hg(II)-sensitive metalreducing anaerobe, Shewanella oneidensis MR-1 [which reduces Hg(II) at low concentrations], was investigated. In all cases, barring suppression of fractionation that is likely due to lower Hg(II) bioavailability, the Hg(II) remaining in the reactor became progressively enriched with heavy isotopes with time and underwent mass-dependent Rayleigh fractionation with R202/198 values of 1.0016 ( 0.0004 (1 SD). Based on a multistep framework for the Hg(II) reduction pathways in the three strains, we constrain the processes that could contribute toward fractionation and suggest that for Hg(II)-resistant strains, reduction by mercuric reductase is the primary step causing fractionation. The proposed framework helps explain the variation in the extent of Hg stable isotope fractionation during microbial reduction of Hg(II), furthering the promise of Hg isotope ratios as a tool in determining the role of microbial Hg transformations in the environment.
Introduction Reduction of ionic mercury [Hg(II)] to elemental mercury [Hg(0)] is one of the most important transformations in the Hg biogeochemical cycle because it competes for the substrate of methylation, i.e., Hg(II), and results in the transport of Hg(0) from aquatic ecosystems to the atmosphere. Both photochemical processes and specific (i.e., mer-mediated) and nonspecific microbiological activities reduce Hg(II) to Hg(0), but it is not clear which mechanisms lead to the production of Hg(0) in a given aquatic environment. Because the atmospheric distribution of Hg(0) remains a major source of Hg contamination, the determination of sources and pathways for the formation of Hg(0) emissions is central to regulatory efforts that are aimed at controlling this problem (1). The two primary approaches used to assess sources of * To whom correspondence should be addressed. Present address: Department of Geosciences, Washington Road, Guyot Hall, Princeton, NJ 08544. E-mail:
[email protected]. Tel.: 1-608-258-0896. Fax: 1-608-258-1274. † Rutgers University. ‡ University of Michigan. 10.1021/es801591k CCC: $40.75
Published on Web 11/11/2008
2008 American Chemical Society
Hg(0) emissions, direct measurements and modeling, are subject to high levels of uncertainty (1). Given the relatively large range of Hg fractionation seen in environmental samples and experimental evidence suggesting fractionation of Hg during biological, physical, and photochemical processes (2-4), Hg isotope ratios hold great promise for elucidating transformations and determining the relative contributions of multiple sources to Hg deposition (see supporting text S1.1 for references). The ability to use stable isotope ratios of Hg as an indicator of Hg biogeochemical processes in environmental management, however, depends on wellfounded constraints on the extent of fractionation (i.e., values of fractionation factors) during all the transformation processes known to be a part of the global and regional Hg biogeochemical cycles. A variety of biological and abiotic processes can cause Hg(II) reduction in aquatic environments. The Hg(II) biological reduction mechanism that is the most specific and extensively studied is mediated by the Hg resistance (mer) system that is found in a broad range of mostly aerobic Hgresistant bacteria and archaea from diverse environments (5). Transport of Hg(II) into the cell and its reduction by the cytoplasmic mercuric reductase (MerA) are the steps central to the mer pathway in bacteria. Specific, high-affinity, and dedicated uptake of Hg(II) is initiated in the bacterial periplasm (in Gram-negative bacteria) where MerP acts as an extracytoplasmic Hg(II) “sponge” that transfers the Hg(II) ion to transmembrane R-helical transporters such as MerT, which then transports it across the inner membrane. It has been suggested that in Gram-positive bacteria, which lack both the outer membrane (OM) and a “proper” periplasmic space, MerP is present outside the cell but attached to the cytoplasmic/inner membrane (6). There has been no systematic exploration of the Hg(II) reduction mechanism in anaerobes, but this might involve respiratory electron-transport chains in the cell wall that probably reduce Hg(II) before it enters the cytoplasm and becomes complexed to cytoplasmic thiols (Sue Miller, personal communication). This is supported by Wiatrowski et al. (7), who showed that Hg(II) reduction in anaerobes is dependent on the presence of both an electron donor and an electron acceptor. In the presence of organic compounds, a variety of dark abiotic processes can also cause Hg(II) reduction (see references in ref 5). In a previous study (2), we showed that reduction of Hg(II) by the mer-mediated pathway in a Hg(II)-resistant Gramnegative bacterium, Escherichia coli JM109/pPB117, systematically reduced the lighter isotopes of Hg(II), with a fractionation factor, (R202/198)reactant/product, ranging from 1.0014 to 1.0020 irrespective of incubation temperature. However, it is likely that even though a narrow range of overlapping R values was found for one Gram-negative bacterium when grown at different temperatures (22-37 °C), other mercarrying bacteria might not lead to a similar extent of Hg fractionation because of differences in the nature of cell walls and Hg(II) transport mechanisms or the number and nature of the rate-limiting step(s) in the reduction pathway(s). In the case of selenium isotopes, experiments with a Grampositive microbe and a Gram-negative microbe each suggested that cell wall structure might significantly influence the extent of selenium isotope fractionation during selenate, but not during selenite reduction to elemental selenium (8). Therefore, an examination of several strains that employ a variety of reduction pathways (see also supporting text S1.2) is vital for beginning to constrain the range of fractionation possible during Hg(II) reduction. VOL. 42, NO. 24, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
9171
Here, we have measured the extent of Hg isotope fractionation by three Hg(II)-reducing bacterial strains. We used two Hg(II)-resistant Gram-positive bacterial strains, a mesophilic soil bacterium, Bacillus cereus Strain 5 (9), and a thermophile, Anoxybacillus sp. FB9 (10). We also used a Hg(II)-sensitive Gram-negative facultative metal-reducing bacterium, Shewanella oneidensis MR-1, that has been shown to reduce Hg(II) at low concentrations by a non-mermediated process likely associated with respiratory electrontransport chains (7). We discuss the results in light of the steps known or proposed for the multi-step processes of Hg(II) reduction by these three strains.
Materials and Methods Bacterial Strains and Growth Conditions. The overall strategy for the determination of fractionation factors during microbial Hg(II) reduction was described previously (2). All bacterial strains were grown in the dark at their respective optimal growth temperatures in defined media, except as noted. (1) Strain 5. B. cereus strain 5, a low-GC Gram-positive soil organism (9) for which Hg(II) resistance is inducible and is encoded by an uncharacterized plasmid (see supporting text S2.1), was grown in M9-based defined medium (2) with the addition of 0.1% yeast extract. Cells were grown overnight at 30 °C with shaking to an optical density at 660 nm (A660) of 0.6. The overnight culture was diluted 1:10 in fresh medium containing ∼600 ng/g Hg(II) in the form of HgCl2. After 3.5 h (when an A660 value of 0.35 was reached), another addition of 600 ng/g Hg(II) was made to reinduce the culture, and 15 min later, 300 µL of this starter culture was added to 300 mL of fresh M9 medium containing 600 ng/g Hg(II) in the form of SRM NIST 3133 in a 1-L reactor. An uninoculated reactor containing the same medium and Hg(II) dose was used as an abiotic control. The starting cell density in the reactor as determined by spread plating on LB plates was 5 × 105 CFU/ mL. Samples were collected from the reactor at regular time intervals to determine the concentrations and isotopic composition of Hg(II) that remained in the reactor. (2) Anoxybacillus sp. FB9. Strain FB9 is a low-GC Grampositive thermophilic facultative chemolithoautotroph that was isolated from a Hg-rich geothermal spring (10). The strain was grown in a defined medium with 10 mM acetate at its optimum growth temperature of 60 °C (10). The cells were grown overnight to a cell count of 1.5 × 109 CFU/mL as determined by direct microscopic counts following acridine orange staining (10). The overnight culture was induced by the addition of 200 ng/g Hg(II), incubated for an additional 30 min, and diluted 1:10 into a 1-L reactor that contained ∼350 mL of fresh medium with 500 ng/g Hg(II) (see also supporting text S2.1 for isotopic composition of starting material). Two control incubations employing the same medium and conditions were included: a control consisting of heat-killed cells, which had been placed at 90 °C for 40 min prior to inoculation of the reactor, and an abiotic control that was performed on a different date, which contained a higher starting Hg(II) concentration (∼875 ng/g). Heat killing ensures that the cells become metabolically inactive without damaging the cell wall structure and without releasing the cytoplasmic contents that could impact Hg(II) chemistry. Sample collection was performed as for B. cereus strain 5. (3) Shewanella oneidensis MR-1. Strain MR-1 is a facultative anaerobe that can grow with multiple electron acceptors and reduces Hg(II) by a non-inducible non-mer-dependent pathway (7). This strain was grown in a defined medium under fumarate reducing conditions (7). Briefly, autoclaved medium components (except as noted below) were assembled aerobically in a pre-sterilized 1-L reactor, and the reactor was bubbled with Hg and oxygen-free nitrogen for 45 min to purge any dissolved oxygen. Anaerobic and filter 9172
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 24, 2008
FIGURE 1. Reactor isotope data for B. cereus Strain 5, Anoxybacillus sp. FB9, and S. oneidensis MR-1 grown at their optimum growth temperatures in 1-L reactors plotted as ln(R/R0) vs ln(f). The two bold dotted lines have slopes of -0.0018 and -0.0012 and represent the maximum (1.0018) and the minimum (1.0012) extent of Rayleigh-type fractionation observed during early phases of the experiments. sterilized stock solutions of sodium fumarate (electron acceptor), sodium lactate (electron donor), amino acids, and sodium bicarbonate were added toward the end of this bubbling period. NIST SRM 3133 was added to the reactor to a final concentration of 50 ng/g Hg(II), before the addition of the starter culture. A fumarate reducing starter culture was grown to an A660 value of 0.07, corresponding to ∼108 cells/mL (7), and diluted 1:1000 in fresh medium to start the experiment. A heat-killed control consisted of the same medium, Hg(II) concentration, and incubation conditions as above, but contained the same number of heat-killed (at 80 °C for 30 min) cells. Mercury Concentration and Isotopic Determination. The concentration of total Hg and the isotopic composition of the reactant Hg(II) remaining in the growth medium [i.e., strictly “residual” Hg(II) plus cell-associated Hg(II)] were measured using cold vapor generation atomic fluorescence and multiple-collector inductively coupled plasma mass spectrometry, respectively, as described previously (2, 11). Isotope ratios for all experiments are reported in delta notation, in units of per mil (‰), referenced to a standard (NIST SRM 3133) (12). δxxxHg is calculated as δxxxHg )
[
(xxxHg/198Hg)sample (xxxHg/198Hg)NIST
]
- 1 × 1000
Fractionation factors (R202/198 ) [202Hg/198Hg]instantaneous reactant/ [202Hg/198Hg]instantaneous product) and statistical uncertainties associated with the fractionation factors [one standard deviation (1SD) or two standard errors (2SE)] were calculated using the Rayleigh equations and York regression method on reactor isotopic data [ln(R/R0) vs ln(f)] (2). Theorectical R202/198 values for equilibrium fractionation during exchange of a ligand A to a ligand B by a Hg(II) ion can be estimated using β factors: (R202/198)A-B ) βA/βB. The β factors are theoretically calculated values of fractionation between a Hg compound (A or B) and Hg(0) (13). Hence, net fractionation during an exchange process occurring at equilibrium can be estimated by using the relationships δ202A - δ202B (in ‰) ) 1000 ln[(R202/198)A-B] ) 1000[ln(β202/198)A - 1000 ln(β202/198)B] ) (∆202/198)A-B.
Results and Discussion Reduction of Hg(II) to Hg(0) by all Hg(II)-reducing strains used in this study led to enrichment of heavier isotopes in the reactor, indicating preferential reduction of lighter isotopes of Hg. Initially, this isotope fractionation caused mass-dependent “Rayleigh-type” fractionation (Figures 1 and
TABLE 1. Summary of Characteristics of Strains Used for Hg(II) Reduction Experiments and the Corresponding r202/198 Values
strain
optimum growth temp (°C)
Escherichia coli JM109/pPB117
37
JM109 replicate
-
common habitat
cell wall
reduction pathway
used electron acceptor
used electron donor
soil/waterb
Gram-neg mer
oxygen
pyruvate
-
-
-
-
-
added Hg(II) (ng/g) na
fractionation factor (r202/198)
600
5
(1.0014 ( 0.0001)c
-
5
(1.0016 ( 0.0005)c
Bacillus cereus Strain 5
30
soil
Gram-pos mer
oxygen
pyruvate
600
4
1.0012 ( 0.0001
Anoxybacillius sp. Strain FB9
60
geothermal areas
Gram-pos mer
oxygen
acetate
500
5
1.0014 ( 0.0001
Shewanella oneidensis MR-1
30
sediment/ subsurface Gram-neg unknown fumarate lactate
50
7
1.0018 ( 0.0003
a
Number of data points used for regression analysis. b A mer system cloned from the soil bacterium Pseudomonas stutzeri OX and described in ref 2. c Best estimate of fractionation factor based on reactor data (2).
S1-S3; also see below) of Hg(II) in the reactor with mean values of fractionation factors (R202/198) ranging from 1.0012 to 1.0018. However, as the reaction progressed, a progressive suppression in Hg isotope fractionation was observed during each of the experiments. The Hg(II) concentrations and isotopic compositions of the samples (Tables S1-S3) are provided as a part of the Supporting Information. All R202/198 values calculated based on the isotopic data for the early phases of the reduction process (Tables S1-S3) overlapped those previously reported for E. coli JM109/pPB117 (2) (Table 1). B. cereus Strain 5. For the experiment with Strain 5 (Figures 1 and S1, Table S1), the R202/198 value was 1.0012 ( 0.0001 (2SD). As compared to the live control, where 65% of the added Hg(II) was lost from the reactor in the first 6.5 h (Table S1), only 2.5% was lost during that period from the dark abiotic control, with the value of δ202Hg for the reactor staying close to 0.07 ( 0.08‰ (2SD of repeat analyses of the in-house standard (2)). After ∼3 h of incubation, when the cell density had increased to ∼107 CFU/mL and the Hg(II) concentration dropped to
