Microbial Isotopic Fractionation of Perchlorate Chlorine

Abstract

Perchlorate contamination can be microbially respired to innocuous chloride and thus can be treated effectively. However, monitoring a bioremediative strategy is often difficult due to the complexities of environmental samples. Here we demonstrate that microbial respiration of perchlorate results in a significant fractionation (∼−15‰) of the chlorine stable isotope composition of perchlorate. This can be used to quantify the extent of biotic degradation and to separate biotic from abiotic attenuation of this contaminant.

Environmental contamination with perchlorate (ClO₄⁻) has recently been recognized as posing a significant health threat (30, 31). In general, this contaminant is anthropogenically sourced, resulting from its extensive use by the munitions industry as a major component of explosives and rocket fuels (29, 30). Prior to 1997, perchlorate was an unregulated compound; however, the recent discovery of perchlorate contamination in drinking water resources throughout the United States, especially those in the southwestern states of Nevada, Utah, and California (25), prompted the emplacement of new regulatory guidelines within the United States, with a final maximum concentration limit to be set pending the outcome of ongoing toxicological studies. In January 2002, the current provisional action level of 1 μg liter⁻¹ was recommended for drinking water following the publication of the first draft of the U.S. Environmental Protection Agency review on toxicological and risk characterization data associated with perchlorate contamination (31).

Because of perchlorate's unique chemical stability under environmental conditions and its high solubility (30), microbial reduction of perchlorate to innocuous chloride is considered the most practical method for remediation (30). Recent studies have demonstrated that specialized microorganisms have evolved that can couple growth to the anaerobic reduction of perchlorate and completely reduce it to chloride (1, 7, 10, 11, 19, 21, 26-28, 32). Phenotypic characterization studies have demonstrated that the known perchlorate-reducing bacteria exhibit a broad range of metabolic capabilities and can thrive in adverse environments (10; J. Pollock, L. A. Achenbach, and J. D. Coates, unpublished data). Significant advances have similarly been made in the biochemistry and genetic systems involved in microbial perchlorate reduction and on the environmental factors which affect their activity (5, 6, 8, 10). As such, the applicability of this metabolism offers great potential for the bioremediation of perchlorate-contaminated environments, and recent field studies have successfully demonstrated this potential.

Although several molecular and immunological tools based on unique signature molecules are now available to monitor the microbial populations associated with perchlorate reduction in the environment (5, 6, 24), monitoring of the effectiveness of a bioremediative strategy in field environments is often difficult due to the complex nature of environmental samples. Results can often be tainted by many abiotic factors, including adsorption or chemical reactivity of the target contaminant. One potential strategy for overcoming these shortcomings with many compounds is to follow the changes in stable isotope composition of the molecule of interest. Variations of the stable isotope ratios of many elements have been used for a long time to give valuable information about biogeochemical processes occurring in the environment (4, 18, 22). Many atoms can exist in two or more forms, chemically identical but differing in mass. The relative abundances of the stable (nonradioactive) isotopes are effectively constant for each element; however, microbial processes are known to make small but significant changes to isotopic compositions (3, 9, 14, 15, 17, 20). In the case of chlorine (Cl), there are relatively few examples of major fractionating processes operating naturally: probably the largest effect is attributed to aqueous diffusion of dissolved chloride in marine pore waters in low-permeability rocks (originally ∼0‰). As a result of the faster diffusion of ³⁵Cl, dilute brines at the diffusion front have values of ∼−0.9‰, showing relative depletion of ³⁷Cl, while the residual brine is ∼+1.9‰ (12). However, even larger changes result from chemical manufacturing processes, where, for example, chlorinated hydrocarbon solvents, products from the common feedstock natural sodium chloride (∼0‰), can show a range of values from −3‰ to + 4‰ (16). Because of chlorine's inherently conservative nature, Cl isotope methods have been used and proposed for use as a natural tracer of sources and mixing processes of solutions. Recently, however, studies demonstrated that the chlorine isotope composition of the chlorinated solvents perchloroethene and trichloroethene (TCE) can be altered during microbial reductive dechlorination, and the lighter isotope (³⁵Cl) is preferentially utilized (23; M. L. Coleman, T. J. McGenity, and M. C. P. Isaacs, Abstr. 9th Annu. V. M. Goldschmidt Conf., abstr. 7253, 1999). If perchlorate-reducing bacteria would similarly fractionate the chlorine isotope composition of perchlorate, this would offer a simple yet effective mechanism to monitor the success of a bioremediative strategy for perchlorate-contaminated environments.

As part of a study on the metabolic diversity of organisms capable of growth by the anaerobic respiration of perchlorate, we isolated a novel organism, Dechlorosoma suillum strain PS, from a swine waste lagoon (1, 11, 21). Physiological characterization revealed that D. suillum rapidly reduced perchlorate with acetate as the electron donor and that reduction was dependent on molybdate and was negatively regulated by the presence of either oxygen or nitrate (6, 8, 24). The ability of D. suillum to fractionate chlorine isotopes while growing on perchlorate was determined by culturing the organism at 37°C in modified basal medium (7) from which the major chloride salts were removed and replaced with equivalent sulfate salts. Culturing was performed with acetate (10 mM) as the sole electron donor and perchlorate (10 mM) as the sole electron acceptor in a 6-liter automatically controlled (pH, temperature, and dissolved oxygen) batch fermentor (Bioflow 2000, New Brunswick Scientific Co., Inc., Edison, N.J.) as previously described (8). A value of pH 7.0 was maintained through the automatic dispensation of anoxic sterile solutions of either 0.5 M H₂SO₄ or 1.0 M NaOH, as appropriate. The inoculum culture of D. suillum was prepared in an identical chloride-free medium except that nitrate (10 mM) replaced perchlorate as the sole electron acceptor. Growth was monitored by microscopic observation and optical density measurements at 600 nm. Perchlorate and chloride concentrations were determined on subsamples collected at regular intervals throughout the growth cycle by ion chromatography analyses as previously described (8). Ion chromatographic analysis of the basal medium components prior to inoculation indicated the presence of a minor chloride contamination (0.2 mM) resulting from the vitamin and mineral stock solutions, which was expected, as some chloride salts were used in the preparation (7).

Growth and perchlorate reduction by D. suillum were rapid and reproducible in this medium (Fig. 1a). The initial lag phase was due to the transfer of the nitrate-grown culture into medium with perchlorate as the sole electron acceptor, as previously observed (8). Ion chromatographic analysis (8) of the culture broth throughout the growth phase revealed that the perchlorate was quantitatively reduced to chloride by D. suillum, and no potential intermediates, such as chlorate or chlorite, were observed (Fig. 1b). Subsamples were collected throughout the growth phase of the culture for determination of the chlorine isotope composition of the perchlorate. Cells were removed by filtration through sterile 0.2-μm-pore-size filters under an N₂-CO₂-H₂ (75/20/5, vol/vol/vol) atmosphere in an anaerobic glove bag (Coy Laboratory Products, Inc.) and dispensed into previously prepared, clean, sterile, sealed 10-ml serum vials filled with N₂. Isotopic analysis of the perchlorate and residual chloride contents of the samples was performed as previously described (2). Briefly, chloride was separated from the perchlorate by precipitation as AgCl₂, and the perchlorate content of the samples was then quantitatively reduced to chloride by alkaline fusion-decomposition as previously described (2). As before, stoichiometric conversion of the perchlorate to chloride (>98% yield) was achieved (data not shown). Chlorine isotope analyses were performed in duplicate on each sample. The precision of a single isotopic analysis was <0.05‰, and the average difference between duplicate analyses was 0.1‰. The results indicated a systematic variation in isotopic composition as the extent of perchlorate reduction to chloride by D. suillum proceeded (Fig. 2). The relationship between the isotope composition of the residual product, the fraction reacted (f), and the isotopic difference between reagent and product (α) was a function of a variant of the Rayleigh distillation formula, namely, R/R₀ = f ^{(α − 1)}, where R₀ is the ³⁷Cl/³⁵Cl ratio of initially present Cl⁻ in ClO₄⁻; R is the ³⁷Cl⁻/³⁵Cl⁻ ratio of residual ClO₄⁻; and α is the isotopic fractionation factor for the reaction, R_perchlorate/R_chloride. The equation operates similarly but independently with the relevant chloride isotope data. Perchlorate isotope data from an initial experiment were consistent with a very significant isotopic fractionation factor, α = 0.9842 ± 0.0004, equivalent to an isotopic difference such that Δ_{perchlorate-chloride} = 15.8‰ ± 0.4‰ (mean ± standard deviation; n = 3) (Fig. 2). Because of the low-level chloride contamination in the basal medium associated with the use of chloride salts in the trace metal and vitamin stock solutions, chloride data were not of sufficient quality to calculate an independent estimate of α. However, two more-extensive repeat experiments gave a Δ_{perchlorate-chloride} value of 14.8 ± 1.3 from perchlorate chlorine isotopes, while chloride gave an independent determination of 14.8‰ ± 0.7‰ (both values are means ± standard deviation; n = 10). Although the isotopic differences from all the experiments are not identical, they are very consistent and are indicative of an exceptional level of fractionation and the robustness of the data.

FIG. 1.

Cell density increase (a) and perchlorate reduction with corresponding chloride formation (b) by D. suillum during growth with acetate as the sole electron donor. The results depicted are the averages of triplicate determinations.

FIG. 2.

Chlorine isotope compositions of chloride and residual perchlorate during microbial reduction by D. suillum. Instant Cl⁻ indicates the calculated isotopic composition of chloride produced from the remaining perchlorate at that point in the process, as though unmixed with previously produced chloride.

The extent of this fractionation is much greater than for chlorine isotopic fractionation (<−2.2 ‰) previously observed during microbial reductive dechlorination of TCE and tetrachlroethene (to cis-dichloroethene) (23). Interestingly, although previous studies (13) have demonstrated that microbial fractionation of some stable isotopes, such as sulfur, is a function of the rate of microbial reduction, the results presented here indicate that the microbial fractionation of chlorine by D. suillum was constant throughout growth and show no indication of any rate-dependent effects (Fig. 2). Similar reduction rate-independent chlorine isotope fractionation was previously observed during the reductive metabolism of TCE by halorespiring organisms utilizing either pyruvate or hydrogen as alternative electron donors (23; Coleman et al., 9th Annu. V. M. Goldschmidt Conf.).

Intrinsic remediation is the preferred option for remediation of perchlorate-polluted sites (30). The results presented here represent the first demonstration of the microbial fractionation of the chlorine-stable isotopes of perchlorate. The discovery of the large isotopic fractionation factors associated with microbial perchlorate reduction allows the measurement of the rate and extent of natural or engineered bioremediation at any site and the discrimination between biotic removal of perchlorate from any possible abiotic mechanisms. Presently, little is known about abiotic mechanisms which may result in the chemical reduction or attenuation of perchlorate in the natural environment. As such, the distinction between biological and abiotic mechanisms, such as dilution, dispersion, or adsorption, during the treatment of a perchlorate-contaminated site is an essential part of monitoring the success of a bioremediation strategy. The use of chlorine isotopic signature determination throughout a bioremediative process may also be of use for the development and application of appropriate predictive models to identify future movement of the pollutant and its potential impact on sensitive target environments. Equally important is the possibility of using the mass and bulk isotopic balances to determine an independent validation of the approach.