Nitrous Oxide Is a Potent Inhibitor of Bacterial Reductive Dechlorination

Abstract

Organohalide-respiring bacteria are key players for the turnover of organohalogens. At sites impacted with chlorinated ethenes, bioremediation promotes reductive dechlorination; however, stoichiometric conversion to environmentally benign ethene is not always achieved. We demonstrate that nitrous oxide (N₂O), a compound commonly present in groundwater, inhibits organohalide respiration. N₂O concentrations in the low micromolar range decreased dechlorination rates and resulted in incomplete dechlorination of tetrachloroethene (PCE) in Geobacter lovleyi strain SZ and of cis-1,2-dichloroethene (cDCE) and vinyl chloride (VC) in Dehalococcoides mccartyi strain BAV1 axenic cultures. Presumably, N₂O interferes with reductive dechlorination by reacting with super-reduced Co(I)− corrinoids of reductive dehalogenases, which is supported by the finding that N₂O did not inhibit corrinoid-independent fumarate-to-succinate reduction in strain SZ. Kinetic analyses revealed a best fit to the noncompetitive Michaelis−Menten inhibition model and determined N₂O inhibitory constants, K_I, for PCE and cDCE dechlorination of 40.8 ± 3.8 and 21.2 ± 3.5 μM in strain SZ and strain BAV1, respectively. The lowest K_I value of 9.6 ± 0.4 μM was determined for VC to ethene reductive dechlorination in strain BAV1, suggesting that this crucial dechlorination step for achieving detoxification is most susceptible to N₂O inhibition. Groundwater N₂O concentrations exceeding 100 μM are not uncommon, especially in watersheds impacted by nitrate runoff from agricultural sources. Thus, dissolved N₂O measurements can inform about cDCE and VC stalls at sites impacted with chlorinated ethenes.

Graphical Abstract

INTRODUCTION

Chlorinated solvents such as chlorinated ethenes are widespread groundwater contaminants of concern due to their adverse impact on human and ecosystem health.¹ The discovery of organohalide-respiring bacteria (OHRB) triggered the development of enhanced bioremediation treatment at sites impacted with chlorinated ethenes.^1,2 Geobacter lovleyi (Geo) strain SZ³ reductively dechlorinates tetrachloroethene (PCE) via trichloroethene (TCE) to cis-1,2-dichloroethene (cDCE), and several strains of the species Dehalococcoides mccartyi (Dhc)⁴ and Candidatus Dehalogenimonas etheniformans (Dhgm)⁵ reduce cDCE and vinyl chloride (VC) to environmentally benign ethene. Reductive dechlorination is an electron-consuming process, and biostimulation with fermentable substrates to increase the flux of hydrogen is a commonly applied approach to enhance in situ contaminant detoxification.^2,6 Biostimulation alone or combined with bioaugmentation has been applied at many sites impacted with chlorinated ethenes, and substantial reductions in total organic chlorine are generally achieved; however, declining PCE and TCE concentrations are often associated with increasing trends in cDCE and VC concentrations.^7–10 Stalled reductive dechlorination may be due to a lack of electron donor (i.e., hydrogen) or unfavorable geochemical conditions such as the presence of competing electron acceptors (e.g., sulfate),^11,12 the presence of oxygen,¹³ low pH conditions,¹⁴ or the absence of organohalide-respiring Dehalococcoidia (i.e., Dhc and Dhgm).^2,15

Interestingly, a data-mining prediction model applied to geochemical and microbial groundwater monitoring data sets collected from sites impacted with chlorinated ethenes ranked nitrate and nitrite concentrations as the most relevant predictors for in situ reductive dechlorination and detoxification potential.¹⁶ Microcosm studies have attributed nitrate inhibition of PCE-to-cDCE reductive dechlorination to elevated redox potential;^7,11 however, nitrate had no inhibitory effect on PCE or TCE dechlorination in axenic cultures of Geo strain SZ and Dhc strain FL2.^3,17 These contrasting observations can be reconciled if a nitrate transformation product exerts an inhibitory effect on reductive dechlorination. Nitrous oxide (N₂O) is a known metabolite of microbial nitrogen metabolism including nitrate reduction or ammonium oxidation,^18–21 and inhibitory effects of N₂O on methanogenesis,^22–24 and in one case on PCE reductive dechlorination,¹¹ have been reported. Although detailed mechanistic studies are lacking, a likely explanation is the reaction of N O with cob(I)amides^25,26 and the ensuing inhibition of corrinoid-dependent enzyme systems involved in e.g., methanogenesis.²⁷ The reductive dehalogenase (RDase) enzyme systems catalyzing carbon−chlorine bond cleavage require a cobamide prosthetic group for function;^28–30 however, the effects of N₂O on reductive dechlorination by OHRB are largely unclear.

As global fertilizer usage continues to increase and elevated concentrations N₂O occur in groundwater,³¹ detailed understanding of the impact of N₂O on relevant microbial processes, including organohalide respiration, is needed. This study explored the impact of N₂O on microbial reductive dechlorination and determined inhibitory constants (K_I values) of N₂O on reductive dechlorination of PCE in Geo strain SZ and of cDCE and VC in Dhc strain BAV1.

MATERIALS AND METHODS

Chemicals.

PCE, cDCE (both ≥99.5%), VC (≥99.5%), ethene (≥99.9%), Vitamin B₁₂ (≥98%), sodium fumarate (≥98%), and N₂O (≥99%) were purchased from Sigma-Aldrich (St Louis, MO, USA). All other chemicals used in this study were reagent grade or of higher purity.

Bacterial Strains and Growth Conditions.

The impact of N₂O on reductive dechlorination was investigated using two organohalide-respiring isolates: Geo strain SZ, a corrinoid-prototroph capable of dechlorinating PCE and TCE to cDCE,^3,32 and Dhc strain BAV1, a corrinoid-auxotroph capable of dechlorinating cDCE and VC to nontoxic ethene.^4,33 Both cultures were grown in 160 mL serum bottles containing a CO₂/N₂ (20/80, vol/vol) headspace and 100 mL of synthetic, bicarbonate-buffered (30 mM, pH 7.3) and reduced (0.2 mM Na₂S and 0.2 mM l-cysteine) mineral salt medium as described.^30,34 Strain SZ cultures received 5 mM acetate as electron donor, 0.28 mM PCE (38 μmol/bottle), 10 mM fumarate, or both PCE and fumarate as electron acceptors, and a vitamin B₁₂-free Wolin vitamin solution.³⁵ Dhc strain BAV1 culture vessels received 10 mL of hydrogen, 0.8 mM (90 μmol/bottle) cDCE or 0.25 mM (40 μmol/bottle) VC, and 5 mM acetate as electron donor, electron acceptor, and carbon source, respectively. Dhc strain BAV1 culture vessels received the Wolin vitamin stock solution containing vitamin B₁₂ to achieve a final concentration of 25 μg L⁻¹. For inoculation, culture vessels received 3% (v/v) inocula from actively dechlorinating cultures maintained under the same conditions. All experiments used triplicate culture vessels, and culture bottles without N₂O and without inoculum served as positive and negative controls, respectively. Culture vessels were incubated without agitation at 30 °C in the dark with the stoppers facing up.

Inhibition Experiments.

For Geo strain SZ cultures, undiluted N₂O gas was directly added to incubation vessels using plastic syringes (BD, Franklin Lakes, NJ, USA) to achieve final aqueous N₂O concentrations of 9.5, 19.1, and 57.3 μM in PCE-amended cultures, of 191.4 μM and 10 mM in fumarate-amended cultures, and of 191.4 μM in cultures grown with both PCE and fumarate. For Dhc strain BAV1 cultures, undiluted or 10-fold-diluted (with N₂) N₂O gas was added with plastic syringes to achieve final aqueous phase N₂O concentrations of 9.5, 19.1, and 57.3 μM in cDCE-amended vessels and of 2.9, 5.7, and 19.1 μM in VC-amended vessels. The aqueous phase concentrations (in μM) of N₂O in all medium bottles were calculated from the headspace concentrations using a dimensionless Henry’s constant of 1.94 for N₂O at 30 °C³⁶ according to C_aq = C_g/H_cc, where C_aq, C_g, and H_cc are the aqueous concentration (in μM), the headspace concentration (in μmol L⁻¹), and the dimensionless Henry’s constant, respectively. Average dechlorination rates were calculated based on the continuous accumulation of dechlorination products (i.e., before stable product concentrations were observed). Since each reductive dechlorination step is associated with the release of one chloride ion (Cl⁻), the average dechlorination rates determined in growth experiments are reported as the total amount of Cl⁻ released per volume per unit time (i.e., μmol Cl⁻ L⁻¹ d⁻¹).

Whole Cell Suspension Dechlorination Assays.

Biomass for whole cell suspension assays was harvested by centrifugation at 10 000g at 4 °C for 30 min from 1.6 L Geo strain SZ and Dhc strain BAV1 cultures that had dechlorinated three feedings of PCE and cDCE, respectively. Inside an anoxic chamber (Coy Laboratory Products Inc., MI, USA), the supernatants were decanted, and the pellets were suspended in 8−10 mL of reduced mineral salts medium. For protein quantification, triplicate 0.2 (strain SZ) and 1.0 mL (strain BAV1) of concentrated cell suspensions were transferred to 2 mL screw cap tubes containing 20 mg of 0.1 mm diameter glass beads, and cells were broken at room temperature using a Bead Ruptor (OMNI, GA, USA) at a speed of 6.0 m s⁻¹ for three 10 min cycles with 2 min breaks. After centrifugation at 13 000g for 2 min to remove cell debris, protein content in the supernatants was estimated using the Bradford assay³⁷ on a plate reader (BioTek Instruments, VT, USA). To ensure consistency between replicate experiments, cell suspensions of both isolates were freshly prepared following identical procedures.

Dechlorination assays for rate determinations were performed in 8 mL glass vials sealed with Teflon-lined butyl rubber septa held in place with aluminum crimps. Each vial contained 3.80−4.86 mL of reduced mineral salts medium with 5 mM acetate, a N₂/CO₂ (80/20, vol/vol) headspace for Geo strain SZ, and a H₂/CO₂ (80/20, vol/vol) headspace for Dhc strain BAV1. Electron donor was provided in at least 40-fold excess of the theoretical demand to ensure that reductive dechlorination was not electron-donor limited. For Geo strain SZ cell suspensions, 36−1100 μL of an aqueous 1 mM PCE stock solution was added directly to the assay vials to achieve final aqueous PCE concentrations ranging from 5 to 150 μM. For Dhc strain BAV1 cell suspensions, 2−550 μL of cDCE stock (from a 5.0 mM aqueous stock solution) or 12−630 μL of VC stock (from a 2.0 mM aqueous stock solution) was added to achieve final aqueous cDCE and VC concentrations ranging from 1 to 500 μM and 3 to 150 μM, respectively. The total aqueous volume in all vials was 4.9 mL before introducing 0.1 mL of cell suspension. Replicate vials at each initial PCE and cDCE concentration received N₂O to achieve 0, 10, and 60 μM, and VC-amended vials received N₂O to achieve 0, 15, and 50 μM dissolved phase N₂O concentrations. The added N₂O concentrations were quantified in all assay vials prior to and at the end of the incubation period and determined constant N₂O concentrations. Following equilibration, each assay vial received 0.1 mL of the respective cell suspension (corresponding to 56.3 ± 2.2 μg of protein per vial for Geo strain SZ and 19.7 ± 1.4 μg of protein per vial for Dhc strain BAV1) using plastic syringes to start dechlorination activity. Vials that received 0.1 mL of sterile mineral salt medium instead of cell suspension and vials that received 0.1 mL of heat-killed cell suspension served as negative controls. During the 6 h assay incubation, liquid samples (1 mL) were collected from sacrificial assay vials every 60 min and transferred to 20 mL glass autosampler vials containing 0.1 mL of 25 mM H₂SO₄ to terminate dechlorination activity. The cell titers and substrate concentrations were chosen such that the dechlorination rates could be determined within the 6 h incubation period and no more than 80% of the initial chlorinated ethene concentration had been consumed at the end of the incubation period.

Analytical Procedures.

Chlorinated ethenes and ethene were analyzed with an Agilent G1888 headspace sampler connected to an Agilent 7890 gas chromatograph (GC) equipped with a flame ionization detector (method detection limit ≈ 0.2 μM) and a DB-624 capillary column (60 m length × 0.32 mm diameter, 1.8 μm film thickness).³⁴ Fumarate and succinate were quantified using an Agilent 1200 series high-performance liquid chromatography system equipped with an Aminex HPX-87H column and a dual-wavelength absorbance detector set to 210 nm.³⁸ N₂O was analyzed by injecting 100 μL headspace samples into an Agilent 7890A GC equipped with an HP-PLOT Q column (30 m length × 0.320 mm diameter, 20 μm film thickness) and a microelectron capture detector. The OD measurements were conducted with a PerkinElmer Lambda 35 UV−vis spectrophotometer by transferring 1 mL cell suspension into a cuvette and recording readings at 600 nm.

Dechlorination Kinetics and Inhibition Models.

The shortest doubling times reported for Geo strain SZ and Dhc strain BAV1 are 6 h³ and 2.2 days,^4,33 respectively. Therefore, additional cell growth was considered negligible over the assay period (<6 h) and confirmed by constant OD₆₀₀ values. For this reason, the Michaelis−Menten model (Table S1), rather than the Monod model for systems involving cell growth, was used to analyze the cell suspension dechlorination data. Therefore, the half-velocity constant K_m, rather than the Monod half-saturation constant K_s, was applied in the analyses of cell suspension kinetic parameters. For each treatment at a different initial substrate concentration [S], an initial dechlorination rate v, normalized to the amount of protein per vial, in units of nmol Cl⁻ released min⁻¹ mg protein⁻¹, was calculated from the sum of all dechlorination products measured with the GC. In brief, the amended PCE, cDCE, or VC concentrations in the respective assay vials served as the initial substrate concentrations, and the corresponding dechlorination rates were determined from the slope of progression curves representing total Cl⁻ released. The linear regression analysis included five data points and at least three for the assays with a low initial chlorinated ethene concentration. Thus, each datum point in the Michaelis− Menten plots represents a dechlorination rate extracted from one initial substrate concentration (Tables S2−S4).

The maximum dechlorination rate V_max and the half-velocity constant K_m for each treatment were calculated using the Michaelis−Menten nonlinear regression method in the Enzyme Kinetics Module for SigmaPlot 13 (Systat Software Inc., Chicago, IL, USA). This software module evaluated competitive, noncompetitive, and uncompetitive inhibition models for best fit to the rate data based on the highest coefficient of determination (R²), the lowest corrected Akaike’s Information Criterion (AICc) values, and the lowest standard deviation of the residuals (Sy.x.). The best-fit model (i.e., noncompetitive inhibition) was used to determine the inhibitory constant, K_I, for N₂O on reductive dechlorination. For data visualization, Michaelis−Menten (V over [S]), Lineweaver−Burk (1/V over 1/[S]), and Dixon (1/V over [I]) plots were generated using the SigmaPlot Enzyme Kinetics Module for each inhibition model and the different electron acceptors (i.e., PCE, cDCE, and VC).

RESULTS

N₂O Affects Reductive Dechlorination Performance in Geobacter lovleyi Strain SZ Cultures.

In the absence of N₂O, Geo strain SZ cultures completely dechlorinated 38.1 ± 3.1 μmol of PCE to stoichiometric amounts of cDCE over a 5 day incubation period, and an average PCE-to-cDCE dechlorination rate of 155.6 ± 27.2 μmol Cl⁻ L⁻¹ d⁻¹ was measured (Figure 1A). Cultures amended with 9.5 μM or higher concentrations of N₂O exhibited decreased dechlorination rates and incomplete PCE-to-cDCE dechlorination. In the presence of 9.5 μM N₂O, Geo strain SZ cultures dechlorinated the initial amount of PCE (38.0 ± 3.9 μmol) at a rate of 90.0 ± 21.6 μmol Cl⁻ L⁻¹ d⁻¹, leading to an extended time period of at least 7 days to achieve complete consumption of PCE. Although PCE was completely consumed in cultures that received 9.5 μM N₂O, small amounts of TCE (7.6 ± 0.3 μmol) remained, even after an extended incubation period of 180 days (data not shown). With the addition of 19.1 μM N₂O, the average PCE dechlorination rate in Geo strain SZ cultures decreased to 64.2 ± 4.6 μmol Cl⁻ L⁻¹ d⁻¹, and no more than 78 ± 3.4% of the initial amount of PCE (38.5 ± 1.1 μmol) was dechlorinated. Following an extended incubation period of 180 days, PCE (9.9 ± 0.2 μmol), TCE (6.6 ± 1.6 μmol), and cDCE (25.0 ± 3.2 μmol) were measured in strain SZ cultures that had received an initial dose of 38.5 ± 1.1 μmol of PCE and 19.1 μM N₂O. A further decline in dechlorination rate to 12.0 ± 2.1 μmol Cl⁻ L⁻¹ d⁻¹ was observed in cultures that received 57.3 μM N₂O, and 63.4 ± 4.5% of the initial amount of PCE (38.3 ± 3.7 μmol) remained in culture bottles at the termination of the experiments (Figure 1A). In all culture bottles with observed inhibition of dechlorination activity, electron donor (i.e., 5 mM acetate) was not limiting electron-acceptor reduction. Furthermore, consistent with the absence of N₂O reductase (nos) operons in the genome of Geo strain SZ,³² the amended N₂O remained constant throughout the experiment.

Figure 1.

Effect of N₂O on the consumption of PCE (A), fumarate (B), or both PCE and fumarate (C) as electron acceptors in cultures of the corrinoid−prototrophic bacterium Geo strain SZ. (A) PCE-to-cDCE reductive dechlorination rates and extents in Geo strain SZ cultures without N₂O (top) and in the presence of 9.5, 19.1, and 57.3 μM N₂O. (B) Fumarate-to-succinate reduction in Geo strain SZ cultures without N₂O and in the presence of 191.4 μM and 10 mM N₂O. (C) PCE-to-cDCE reductive dechlorination and fumarate-to-succinate reduction in Geo strain SZ cultures in the absence of N₂O (top) and in the presence of 191.4 μM N₂O (bottom). Solid red circles, PCE; open squares, TCE; solid blue triangles, cDCE; open diamonds, fumarate; solid green squares, succinate. Error bars represent the standard deviation (sd) of triplicate cultures.

In addition to catalyzing PCE-to-cDCE dechlorination, Geo strain SZ also derives energy from acetate oxidation coupled to fumarate to succinate reduction.³ In contrast to organohalide respiration catalyzed by corrinoid-dependent reductive dehalogenases, fumarate respiration via fumarate reductase does not involve a corrinoid-dependent enzyme system.³⁹ Therefore, investigating the impact of N₂O on fumarate reduction by Geo strain SZ cultures served as a control experiment to illustrate the selective effect of N₂O on the PCE RDase in Geo strain SZ. As shown in Figure 1B, the presence of 191.4 μM and 10 mM N₂O did not impact fumarate to succinate reduction rates and extents in Geo strain SZ cultures. In the absence of N₂O, Geo strain SZ reduced PCE and fumarate concomitantly (Figure 1C, top); however, in cultures amended with 191.4 μM N₂O, only fumarate was reduced to succinate and no PCE reductive dechlorination occurred (Figure 1C, bottom).

N₂O Affects cDCE and VC Reductive Dechlorination Performance in Dhc strain BAV1 Cultures.

Without N₂O, Dhc strain BAV1 dechlorinated cDCE (97.2 ± 3.1 μmol) to stoichiometric amounts of ethene (102.3 ± 6.1 μmol) within a 14-day incubation period at an average cDCE-to-ethene dechlorination rate of 146.1 ± 21.6 μmol Cl⁻ L⁻¹ d⁻¹ (Figure 2A). In contrast, cultures amended with 9.5 μM or higher N₂O concentrations all exhibited incomplete transformation of cDCE and VC and required longer incubation periods (up to 28 days) before stable dechlorination product patterns were observed. Cultures that received 9.5 μM N₂O showed a significantly lower dechlorination rate of 66.8 ± 21.6 μmol Cl⁻ L⁻¹ d⁻¹, and the initial amount of cDCE (97.2 ± 1.2 μmol) was dechlorinated to a mixture of VC (74.1 ± 1.2 μmol) and ethene (23.1 ± 0.7 μmol). Cultures that received 29.0 μM N₂O dechlorinated only about one-half (50.7 ± 3.3%) of the initial amount of cDCE (96.3 ± 1.1 μmol) to predominantly VC (48.8 ± 3.2 μmol) at a rate of 24.4 ± 1.1 μmol Cl⁻ L⁻¹ d⁻¹, and only small amounts of ethene (4.7 ± 0.2%) were produced. At a higher N₂O concentration of 57.3 μM, strain BAV1 dechlorinated cDCE to VC at a dechlorination rate of 18.9 ± 0.7 μmol Cl⁻ L⁻¹ d⁻¹, no ethene was formed, and about one-third (26.3 ± 0.4 μmol) of the initial amount of cDCE remained. Extended incubation periods of up to 6 months did not result in further dechlorination in all tested strain BAV1 culture vessels with N₂O. Notably, in all N₂O-amended Dhc strain BAV1 cultures, the VC-to-ethene dechlorination step occurred at such low rates resulting in VC rather than ethene formation as the major product.

Figure 2.

Effect of N₂O on reductive dechlorination of cDCE (A) and VC (B) in corrinoid auxotrophic Dhc strain BAV1. (A) cDCE reductive dechlorination rates and extents in Dhc strain BAV1 without N₂O (top panel) and in the presence of 9.5, 29.0, and 57.3 μM N₂O. (B) VC reductive dechlorination rates and extents in Dhc strain BAV1 without N₂O (top) and in the presence of 2.9, 5.7, and 19.1 μM N₂O. Solid blue triangles, cDCE; open red squares, VC; inverted, solid green triangles, ethene. Error bars represent the sd of triplicate cultures.

To further investigate the impact of N₂O on the reductive dechlorination of VC, Dhc strain BAV1 cultures amended with VC as electron acceptor received N₂O at concentrations of 2.9, 5.7, and 19.1 μM. Cultures without N₂O completely dechlorinated the initial amount of 41.2 ± 0.4 μmol of VC to ethene within 11 days at an average VC dechlorination rate of 37.2 ± 2.7 μmol Cl⁻ L⁻¹ d⁻¹ (Figure 2B). N₂O had a profound impact on VC dechlorination in Dhc strain BAV1 cultures, and the VC dechlorination rates decreased to 18.3 ± 2.1 and 9.8 ± 0.4 μmol Cl⁻ L⁻¹ d⁻¹ in the presence of 2.9 and 5.7 μM N₂O, respectively. Compared to the complete VC to ethene conversion in control incubations without N₂O, the amount of VC dechlorinated to ethene was diminished by 37.0 ± 1.3% and 76.2 ± 4.1%, respectively, in cultures amended with 2.9 and 5.7 μM N₂O. The most pronounced inhibition was observed in Dhc strain BAV1 cultures that received 19.1 μM of N₂O (Figure 2B), and only negligible amounts (<1.6 ± 0.6 μmol) of ethene were formed even after extended incubation periods of 180 days, indicating that the VC to ethene step was particularly susceptible to N₂O inhibition.

Quantification of N₂O Inhibition in Whole Cell Suspension Dechlorination Assays.

To further investigate the inhibitory effects of N₂O on reductive dechlorination, whole cell suspension assays were performed. Plots of dechlorination rates versus initial substrate concentrations with increasing N₂O concentrations are presented in Figures 3 and 4. In all cases, the maximum dechlorination rates decreased with the addition of increasing N₂O concentrations. In Geo strain SZ cell suspensions without N₂O addition, a maximum PCE-to-cDCE dechlorination rate (V_max,PCE) of 76.3 ± 2.6 nmol Cl⁻ released min⁻¹ mg protein⁻¹ was calculated using nonlinear regression in the Michaelis−Menten model (Figure 3A and Table 1). With the addition of 10 and 60 μM N₂O, V_max,PCE in Geo strain SZ cell suspensions declined to 61.3 ± 4.1 and 30.9 ± 4.2 nmol Cl⁻ released min⁻¹ mg protein⁻¹, respectively. Even more pronounced inhibitory effects of N₂O were observed for cDCE and VC dechlorination in Dhc strain BAV1 assays. In the presence of 10 and 60 μM N₂O, the maximum cDCE-to-VC dechlorination rate decreased from a V_max,cDCE of 119.4 ± 2.1 nmol Cl⁻ released min⁻¹ mg protein⁻¹ in assays without N₂O to 81.1 ± 5.1 and 31.2 ± 2.4 nmol Cl⁻ released min⁻¹ mg protein⁻¹, respectively (Figure 4A and Table 1). The strongest inhibitory effect of N₂O was observed in whole cell suspension assays of strain BAV1 with VC as electron acceptor. Compared to the maximum VC-to-ethene dechlorination rate V_max,VC of 123.3 ± 2.2 nmol Cl⁻ released min⁻¹ mg protein⁻¹ without N₂O, the addition of 15 and 50 μM N₂O reduced V_max,VC to 78.9 ± 2.1 and 19.9 ± 1.3 nmol Cl⁻ released min⁻¹ mg protein⁻¹, respectively (Figure 4C and Table 1). No dechlorination was detected in control incubations (data not shown), confirming that suspended cells, rather than any abiotic reactions, were responsible for the observed dechlorination activity.

Figure 3.

Kinetics of PCE-to-cDCE reductive dechlorination in cell suspensions of Geo strain SZ in the presence of increasing concentrations of N₂O. (A) Michaelis−Menten plot of initial PCE-to-cDCE dechlorination rates versus PCE concentrations without and in the presence of 10 and 60 μM N₂O. (B) Lineweaver−Burk plot with inserted Dixon plot illustrating N₂O inhibition on PCE-to-cDCE reductive dechlorination. Solid lines represent the best fit to each data set based on nonlinear regression using the noncompetitive inhibition model. Solid green circles represent rate data measured in the absence of N₂O, solid blue triangles show rate data measured in the presence of 10 μM N₂O, and solid red squares show rate data measured in the presence of 60 μM N₂O. Solid and open red circles depict the graphical determination of −K_I and −1/K_m, respectively.

Figure 4.

Kinetics of cDCE-to-VC and VC-to-ethene reductive dechlorination in cell suspensions of Dhc strain BAV1 in the presence of increasing concentrations of N₂O. (A) Michaelis−Menten plot of initial cDCE-to-VC dechlorination in cell suspensions of Dhc strain BAV1 without and in the presence of 10 and 60 μM N₂O. (B) Lineweaver−Burk plot with inserted Dixon plot illustrating N₂O inhibition of cDCE-to-VC reductive dechlorination. (C) Michaelis−Menten plot of initial VC-to-ethene dechlorination in cell suspensions of Dhc strain BAV1 without and in the presence of 15 and 50 μM N₂O. (D) Lineweaver−Burk plot with inserted Dixon plot illustrating N₂O inhibition of VC-to-ethene reductive dechlorination. Solid lines represent the best fit to each data set based on nonlinear regression using the noncompetitive inhibition model. Solid green circles represent rate data measured in the absence of N₂O; solid blue triangles and solid red squares show dechlorination rate data measured in the presence of N₂O (panels A and B, cDCE; panels C and D, VC). Solid and open red circles shown in B and D depict the graphical determination of −K_I and −1/K_m, respectively.

Table 1.

Kinetic (V_max, K_m) and Inhibition (K_I) Parameters for PCE, cDCE, and VC Reductive Dechlorination in Cell Suspensions of Geo Strain SZ and Dhc Strain BAV1 in Response to Increasing N₂O Concentrations^a

e− acceptor	culture	N2O (μM)	Vmax (nmol Cl− min−1 mg protein−1)	Km (μM)	KI (μM)
PCE	strain SZ	0	76.3 (±2.6)	25.1 (±2.9)	40.8 (±3.8)
		10	61.3 (±4.1)
		60	30.9 (±4.2)
cDCE	strain BAV1	0	119.4 (±4.1)	19.9 (±2.5)	21.2 (±3.5)
		10	81.1 (±5.1)
		60	31.2 (±2.4)
VC	strain BAV1	0	123.3 (±3.2)	18.9 (±1.3)	9.6 (±0.4)
		15	78.9 (±2.1)
		50	19.9 (±1.3)

^aThe best fit data were achieved with the Michaelis−Menten noncompetitive inhibition model. Error values represent 95% confidence intervals.

Kinetic Parameters Reveal Pronounced N₂O Inhibition.

The experimental data generated in Geo strain SZ cell suspension assays (Table S2) fit the Michaelis−Menten model and the corresponding Lineweaver−Burk plot (R² > 0.95) (Figure 3, Figure S1). A maximum PCE dechlorination rate V_max,PCE of 76.3 ± 2.6 nmol Cl⁻ released min⁻¹ mg protein⁻¹ and a half-velocity constant K_m,PCE of 25.1 ± 2.9 μM characterized PCE-to-cDCE dechlorination kinetics for strain SZ in the absence of N₂O (Table 1, Figure 3B). Without N₂O and in the presence of 10 and 60 μM N₂O, the PCE-to-cDCE dechlorination data fit the competitive, uncompetitive, and noncompetitive inhibition models (R² > 0.90); however, the noncompetitive inhibition model exhibited the highest R² and the lowest AICc and Sy.x. values (Figure 3 and Table S5). Using the noncompetitive inhibition model, an inhibitory constant, K_I, of N₂O for PCE dechlorination in Geo strain SZ whole cell suspensions of 40.8 ± 3.8 μM was determined (inserted Dixon plot, Figure 3B; Table 1).

The experimental data generated in Dhc strain BAV1 whole cell suspension assays (Tables S3 and S4) also fit the Michaelis− Menten model simulations and the corresponding Lineweaver− Burk plots (R² > 0.90) (Figure 4, Figures S2 and S3). In the presence of increasing N₂O concentrations, both cDCE-to-VC and VC-to-ethene reductive dechlorination data showed the best fit to the noncompetitive inhibition model based on the highest R² and the lowest AICc and Sy.x. values (Figure 4 and Table S5). While Dhc strain BAV1 assays produced comparable V_max,cDCE and V_max,VC values of 119.4 ± 2.1 and 123.3 ± 2.2 nmol Cl⁻ released min⁻¹ mg protein⁻¹, respectively, the K_I of N₂O for cDCE dechlorination (21.2 ± 3.5 μM) was approximately 2-fold greater than that for VC dechlorination (9.6 ± 0.4 μM) (inserted Dixon plots, Figures 4B and 4D; Table 1), consistent with the greater inhibitory effect of N₂O observed for the VC dechlorination step. To investigate if the differences in susceptibility of the cDCE and VC dechlorination steps to N₂O resulted from different affinities of the BvcA RDase for cDCE and VC, the K_m values for these two electron acceptors were compared. On the basis of the Dhc strain BAV1 assays, similar K_m,cDCE (19.9 ± 2.5 μM) and K_m,VC (18.9 ± 1.3 μM) values were determined (Figure 4B and 4D; Table 1), indicating that the substrate affinity of the BvcA RDase does not explain the more potent inhibition of the VC-to-ethene dechlorination step by N₂O.

DISCUSSION

Effects of N₂O on Corrinoid-Dependent Processes.

Toxic effects of N₂O were first observed in the mid-1950s after two patients died following prolonged N₂O inhalation.⁴⁰ This incident triggered studies to elucidate the mechanism underlying N₂O toxicity, and detailed investigations revealed that N₂O reacts with cobalt-containing corrinoids (e.g., cobalamin) when the coordinated cobalt atom is in the reduced Co(I) state.^25,26 Due to the vital roles of enzyme systems with corrinoid prosthetic groups,^30,41 the N₂O-mediated oxidation of the Co(I) supernucleophile interferes with key metabolic functions in all domains of life.^42,43 For example, N₂O inhibits the corrinoid-dependent methionine synthase (MetH) required for the biosynthesis of the essential amino acid methionine.⁴⁴ When used as inhalation anesthetic for human patients, N₂O can lead to a malfunction of MetH, resulting in elevated levels of homocysteine in plasma (i.e., hyperhomocysteinemia).⁴⁵ Patients generally recover from N₂O exposure after 3−4 days,⁴⁴ presumably through de novo MetH synthesis or replenishment of Co(I) corrinoid.⁴⁶

The lowest reported N₂O concentration that affected MetH activity in animals and humans was around 4.1 mM (~1000 ppm);⁴⁶ however, much lower N₂O concentrations inhibited members of the Bacteria and Archaea, presumably due to the more diverse roles of corrinoid-dependent enzyme systems in their metabolisms.⁴¹ For example, in the denitrifying bacterium Paracoccus denitrificans strain PD1222, 0.1 mM N₂O not only repressed MetH but also modulated the expression of anabolic genes under the control of vitamin B₁₂ riboswitches.⁴⁷ To compensate for the loss of MetH function, organisms such as P. denitrificans and Escherichia coli activate a corrinoid-independent methionine synthase, MetE.^41,47 Geo strain SZ possesses both the metE and the metH genes,³² which may explain the observation that up to 10 mM N₂O did not inhibit the bacterium’s growth with fumarate in defined minimal medium. In contrast, N₂O at micromolar concentrations inhibited the growth of Geo strain SZ cultures when PCE served as the sole electron acceptor. These observations corroborate that the inhibitory effect of N₂O to organisms varies markedly based on whether corrinoid-dependent enzyme systems are essential for key metabolic steps, including respiratory energy conservation.

OHRB as a Model To Study the Effects of N₂O Inhibition.

A key feature of OHRB is the involvement of corrinoid-dependent RDases in electron transfer to the chlorinated organohalogen electron acceptor.^1,28,29,48 Functional and structural analyses demonstrated that RDases represent a distinct subfamily of corrinoid-dependent enzymes, and a Co(I) supernucleophile is crucial for RDases to initiate the cleavage of carbon−chlorine bonds.^28,29,49 Since corrinoid-dependent RDases are essential in the energy metabolism of OHRB, N₂O inhibition on Co(I) corrinoid-dependent RDases can be readily observed by quantitative measurement of dechlorination activity and growth when alternate electron acceptors are absent or the energy metabolism of the OHRB is restricted to chlorinated electron acceptors, as is the case for Dhc.⁴ This effect was convincingly demonstrated in both Geo strain SZ and Dhc strain BAV1 cultures amended with chlorinated electron acceptors and N₂O. Thus, OHRB are excellent model organisms to assess the inhibitory effects of N₂O on enzymes involving the Co(I) supernucleophile in catalysis. Consistent with this assumption, experimental results showed that N₂O at micromolar concentrations exhibited strong inhibitory effects on reductive dechlorination performance in both Geo strain SZ and Dhc strain BAV1 cultures. Theoretically, a shortage of methionine caused by N₂O inhibition on MetH could also have affected dechlorination activity; however, growth of Geo strain SZ with fumarate was not impaired at much higher N₂O concentrations of up to 10 mM, indicating that N₂O inhibition of RDases diminished dechlorination performance.

The PCE-to-cDCE dechlorinating Geo strain SZ tolerated at least 3-fold higher N₂O concentrations than Dhc strain BAV1 before cessation of dechlorination activity was observed, which possibly relates to strain SZ’s ability for de novo cobamide biosynthesis.³² Active cobamide biosynthesis may allow strain SZ to maintain some level of catalytically active PCE RDase. In contrast, the characterized obligate organohalide-respiring Dhc strains^4,50 and Dehalogenimonas spp., including the VC-respiring Candidatus Dehalogenimonas etheniformans,⁵ cannot de novo synthesize corrinoid, although exceptions may exist.⁵¹ All characterized cDCE and VC dechlorinators strictly require exogenous corrinoid, which renders these bacteria more susceptible to N₂O inhibition, and low micromolar concentrations of N₂O (i.e., 3−10 μM) repressed the growth of the corrinoid-auxotroph Dhc strain BAV1. Unlike corrinoid-auxotrophic Dehalococcoidia, the majority of PCE-to-TCE- and TCE-to-cDCE-dechlorinating OHRB are corrinoid prototrophs,⁵² a feature that may enable these organisms to tolerate higher N₂O concentrations. The difference in the ability for de novo corrinoid biosynthesis is one possible explanation why PCE and TCE dechlorination to cDCE is generally achieved at sites with nitrate,^16,53 but ethene is not produced.^7–10

Corrinoid-dependent enzyme systems fulfill essential metabolic functions for organisms in all branches of life, but only a subset of the bacteria and the archaea have the machinery for de novo corrinoid biosynthesis.^41,54,55 Therefore, N₂O effects on microbial processes that hinge on the activity of corrinoid-dependent enzyme systems may expand beyond organohalide respiration.

Elevated Groundwater N₂O and Kinetic Parameters.

Based the current day atmospheric N₂O concentration of 330 ppb, the theoretical equilibrium concentration of N₂O in groundwater should be around 7 nM, assuming no mass transfer limitations; however, much higher groundwater N₂O concentrations were reported, indicating other sources exist.^31,56 For example, the increased usage of synthetic nitrogen fertilizer for agricultural production causes substantial nitrate runoff and elevated N₂O concentrations in groundwater.^19,31 Indeed, correlations between fertilizer application and associated nitrate runoff with elevated groundwater N₂O concentrations have been established.^19,56 Nitrate is not conservative, and processes including denitrification,^18,19 respiratory ammonification (i.e., dissimilatory nitrate reduction to ammonium, DNRA)⁵⁷ and ensuing nitrification,²¹ as well as chemodenitrification³⁸ contribute to the formation of N₂O. Such biogeochemical processes are likely responsible for elevated N₂O concentrations, and up to 140 μM N₂O was observed in watersheds impacted by agricultural activities.³¹ Thus, the N₂O concentrations measured in many groundwater aquifers exceed the theoretical equilibrium with atmospheric N₂O by up to 5 orders in magnitude, and intensified agriculture will exacerbate this issue.

A K_I value indicates the inhibitor (i.e., N₂O) concentration at which the maximum reaction rate V_max (i.e., the reductive dechlorination rates of chlorinated ethenes) is reduced by 50%. The determined K_I values for N₂O on reductive dechlorination of PCE, cDCE, and VC are in the range of 40, 20, and 10 μM, respectively, well below reported N₂O concentrations encountered in many groundwater aquifers, particularly at sites impacted by agricultural runoff.³¹ Consequently, N₂O inhibition could be a major cause for incomplete reductive dechlorination and cDCE and VC stalls observed at field sites.^7–10,53 Of note, a 50% V_max,cDCE and V_max,_VC inhibition occurred in strain BAV1 at 2- and 4-fold lower N₂O concentrations, respectively, compared to the same level of inhibition of V_max,_PCE in strain SZ. The higher K_I values for N₂O determined for Geo strain SZ compared to strain BAV1 may be related to the ability of strain SZ for de novo corrinoid biosynthesis (see above) or to mechanistic differences in the dechlorination steps catalyzed by the PceA versus the BvcA RDases. Similar K_m values for cDCE and VC were determined in strain BAV1, indicating the organism exhibits similar affinities for cDCE and VC; however, the 2-fold higher K_I value for the cDCE versus the VC dechlorination step cannot be easily explained considering the same BvcA RDase is involved in both dechlorination steps,^52,58 and detailed mechanistic studies are warranted.

Predicting the fate of chlorinated ethenes at bioremediation sites relies on accurate estimates of the intrinsic kinetic parameters of OHRB;⁵⁹ however, kinetic constants determinations using various dechlorinating cultures at different cell densities reported highly variable V_max and K_m values (or K_S values when Monod kinetics were applied).^12,60,61 Likely explanations for these discrepancies are that different, potentially competing types of dechlorinators with distinct RDases and present in varied abundances contributed to reductive dechlorination.^12,62,63 The current study accomplished kinetic measurements in axenic cultures under defined conditions and over short incubation periods (<6 h, no growth occurred), which facilitates the determination of intrinsic kinetic parameters.⁵⁹ The Michaelis−Menten model simulations predicted the behaviors of Geo strain SZ and Dhc strain BAV1 (R² > 0.90), and both organisms fit the noncompetitive inhibition model (R² > 0.96) with micromolar levels of N₂O as the inhibitor. These findings imply N₂O as a noncompetitive inhibitor that oxidizes the Co(I) corrinoid cofactor of RDases, thereby decreasing reductive dechlorination rates.

Implications for in Situ Bioremediation.

Electron donor (i.e., hydrogen) limitations,⁶ nutrient availability (e.g., fixed nitrogen),¹⁵ unfavorable redox potential,^7,11 low pH conditions,¹⁴ or toxic and/or inhibitory effects of cocontaminants (e.g., sulfide, chloroform, 1,1,1-trichloroethane)^12,62 can impact the microbial reductive dechlorination process. The findings of the current study indicate that decreased reductive dechlorination performance can be the result of N₂O inhibition. A common strategy to improve in situ degradation of chlorinated ethenes involves the injection of nutrients (i.e., biostimulation), typically fermentable substrates aimed at increasing the flux of hydrogen.^2,64 The formulations can include fertilizer (nitrate, ammonium, urea, phosphorus) to proactively address possible nutrient limitations.^2,13,15,65 Biogeochemical transformations of fixed nitrogen will generate N O^21,38,57 and exert inhibitory effects on microbial reductive dechlorination, which can result in undesirable cDCE or VC stalls. Thus, practitioners should carefully evaluate the need for fixed nitrogen additions to avoid possible N₂O inhibition.

The inhibitory constants, K_I, for N₂O inhibition of PCE, cDCE, and VC dechlorination were within the N₂O concentration ranges observed in groundwater aquifers (i.e., up to 143 μM),³¹ suggesting that N₂O should be of concern at contaminated sites where practitioners seek to rely on microbial reductive dechlorination as a remedial strategy. Considering the relevance of the microbial reductive dechlorination process for achieving cleanup goals and the commonality of elevated N₂O concentrations in aquifers, groundwater monitoring regimes should include nitrate/nitrite and N₂O measurements, so that potential inhibition and cDCE and VC stalls can be explained and predicted.