Nitrous oxide inhibition of methanogenesis represents an underappreciated greenhouse gas emission feedback

Abstract

Methane (CH₄) and nitrous oxide (N₂O) are major greenhouse gases that are predominantly generated by microbial activities in anoxic environments. N₂O inhibition of methanogenesis has been reported, but comprehensive efforts to obtain kinetic information are lacking. Using the model methanogen Methanosarcina barkeri strain Fusaro and digester sludge-derived methanogenic enrichment cultures, we conducted growth yield and kinetic measurements and showed that micromolar concentrations of N₂O suppress the growth of methanogens and CH₄ production from major methanogenic substrate classes. Acetoclastic methanogenesis, estimated to account for two-thirds of the annual 1 billion metric tons of biogenic CH₄, was most sensitive to N₂O, with inhibitory constants (K_I) in the range of 18–25 μM, followed by hydrogenotrophic (K_I, 60–90 μM) and methylotrophic (K_I, 110–130 μM) methanogenesis. Dissolved N₂O concentrations exceeding these K_I values are not uncommon in managed (i.e. fertilized soils and wastewater treatment plants) and unmanaged ecosystems. Future greenhouse gas emissions remain uncertain, particularly from critical zone environments (e.g. thawing permafrost) with large amounts of stored nitrogenous and carbonaceous materials that are experiencing unprecedented warming. Incorporating relevant feedback effects, such as the significant N₂O inhibition on methanogenesis, can refine climate models and improve predictive capabilities.

Introduction

Carbon dioxide (CO₂) receives primary attention as a driver for climate change, but methane (CH₄) and nitrous oxide (N₂O) account for ~20% and 7%, respectively, of the net radiative forcing in the atmosphere [1, 2]. The central objective of the Paris Agreement, which is to hold the global average temperature increase to “well below 2°C above preindustrial levels” [3, 4], cannot be met without controlling CH₄ and N₂O emissions. Global emissions of both CH₄ and N₂O are ultimately controlled by microbial processes [2, 5]; however, human activities have massively disturbed the natural balance of microbial production and consumption of these greenhouse gases [6]. The current estimated annual net atmospheric emission increases of ~51 Tg of CH₄ [7, 8] and 2.2 Tg of N₂O [9] are both predicted to accelerate [1, 10]. The key microbial guilds and biogeochemical processes responsible for CH₄ and N₂O production and consumption are known, and their responses to climate change have been a matter of intense research [11, 12]. One area of considerable uncertainty pertains to positive and negative feedback loops affecting greenhouse gas emissions and climate [13]. An infamous scenario for a positive feedback loop in response to a warming climate is permafrost thawing, which stimulates the microbial activity and the release of CO₂, CH₄, and N₂O from newly bioavailable carbon and nitrogen pools. The massive release of greenhouse gases will further accelerate radiative forcing and in turn cause more thawing. This example illustrates why a comprehensive understanding of both positive and negative feedback loops is essential. Quantitative information is needed to meaningfully incorporate feedback effects into refined climate models.

Methanogenic archaea (methanogens) drive CH₄ production by utilizing acetate, H₂/CO₂, and methylated compounds as substrates, generating around 1 billion metric tons of CH₄ annually [14, 15]. Approximately, two-thirds of biogenic CH₄ is produced from acetoclastic methanogenesis (i.e. the conversion of acetate to CH₄ and CO₂), with the remaining one-third attributed to hydrogenotrophic CO₂ reduction (hydrogenotrophic pathway) and methylated compound utilization (methylotrophic pathways) [16–20]. While the three major methanogenesis pathways share a core set of enzymes, several mechanistically distinct enzyme systems with corrinoid prosthetic groups (i.e. vitamin B₁₂ derivatives) are involved in methyl group transfer reactions, energy conservation, and CH₄ production (Fig. 1) [18, 19, 21].

In hydrogenotrophic methanogenesis, CO₂ is sequentially reduced to a methyl group carried by tetrahydromethanopterin (H₄MPT) with H₂ as electron donor. The corrinoid-dependent N⁵-methyltetrahydromethanopterin:CoM methyltransferase (MTR) complex then transfers the methyl group onto another C₁-carrier, coenzyme M, forming CH₃–S–CoM, a step associated with energy conservation, followed by CH₄ production catalyzed by methyl coenzyme M reductase (MCR) [22]. Acetoclastic methanogens produce CH₄ by activating acetate to acetyl-CoA, which is then cleaved by the corrinoid-dependent enzyme system carbon monoxide dehydrogenase/acetyl-CoA synthase (CODH/ACS) to yield an enzyme-bound methyl group and a carbonyl group. The carbonyl moiety is oxidized to CO₂ and the methyl group is transferred to H₄MPT [23]. The MTR complex catalyzes a methyl group transfer to form CH₃–S–CoM, a step associated with energy conservation, before MCR mediates the reduction of the methyl group to CH₄ [16, 24]. Methylotrophic methanogens either directly generate CH₃–S–CoM from methylated compounds (e.g. MeOH) utilizing substrate-specific, corrinoid-dependent MTRs, or cleave the methyl group from methoxylated compounds (e.g. 2-methoxybenzoate) and form CH₃–CoM via H₄MPT and the MTR complex [25], followed by CH₄ production via MCR. All three methanogenesis pathways utilize corrinoid-dependent enzyme systems for methyl group transfers and have strict requirements for super-reduced Co(I) to generate CH₄ and conserve energy [26, 27]. These features render methanogenesis sensitive to oxidative stress (e.g. fluctuating redox conditions and oxygen intrusion) [19, 28].

N₂O is an even stronger oxidant than oxygen (E^o’(N₂O(g)/N₂(g) = +1.355 V > E^o’(O₂(g)/H₂O(l) = +0.818 V), reacts with Co(I), and has been shown to interfere with Co(I) cobamide-dependent enzyme systems [29, 30]. Demonstrated metabolic consequences include ceased corrinoid-dependent methionine biosynthesis and impaired organohalide respiration [31–33]. Based on this information, elevated N₂O in anoxic ecosystems would be expected to inhibit other corrinoid-dependent processes such as methanogenesis; however, available studies investigating the impacts of N₂O on methanogenesis have led to inconsistent N₂O inhibition patterns for axenic methanogen cultures and CH₄ producing microbial communities [34–36]. Laboratory incubations showed that Methanobacterium bryantii strain Bab1 grown with H₂/CO₂ ceased CH₄ production in the presence of 95 μM N₂O, whereas Methanosarcina barkeri strain MS maintained some methanogenic activity at 10-fold higher N₂O concentrations under the same growth conditions [35]. Different sensitivities to N₂O inhibition were also reported for mixed methanogenic cultures maintained with different substrates. For example, inhibition of CH₄ production in a mixed community bioreactor occurred only at N₂O concentrations exceeding 700 μM [36], whereas 20–28 μM N₂O completely inhibited methanogenic activity in salt marsh sediment and Amazon peatland enrichment cultures [34, 37]. These variable sensitivities to N₂O suggest that inhibition of methanogenesis by N₂O is organism- and possibly substrate-specific; however, the available data are scarce and do not allow a robust, quantitative assessment of N₂O inhibition on methanogenesis from relevant methanogenic substrates [34–36].

N₂O fluxes in soil–water systems have risen sharply due to the intensified use of synthetic N fertilizer in agriculture [6, 9, 38]. As a result, elevated N₂O concentrations occur more frequently in ecosystems with CH₄ production such as rice paddy soils [39], wastewater treatment plants [40], sediments [41], and groundwater aquifers [42]. Also, permafrost thawing accelerates N turnover, releasing large amounts of N₂O [38, 43]. More detailed knowledge about the interactions between N₂O concentrations and methanogenesis is needed to advance the predictive capabilities of climate change impacts on future greenhouse gas emission scenarios. To address the existing knowledge gaps, we assessed the inhibitory effect of N₂O on CH₄ production from major methanogenic substrates in growth experiments with axenic and mixed methanogenic cultures and in whole-cell suspension assays. Using cultures performing acetoclastic, methylotrophic, and/or hydrogenotrophic methanogenesis, we determined kinetic parameters that quantitatively describe N₂O inhibition on methanogenesis.

Materials and methods

Methanogenic cultures and growth inhibition experiments

The methanogenic archaeon M. barkeri strain Fusaro metabolizes acetate, H₂/CO₂, and MeOH while employing different, substrate-specific methanogenic pathways. To determine if M. barkeri exhibits varied sensitivities to N₂O, cultures were pregrown with MeOH, H₂/CO₂, or acetate for at least three consecutive transfers before the impact of N₂O on CH₄ production was examined. Also, we analyzed three methanogenic mixed cultures derived from anaerobic digester sludge, which is known to harbor a broad diversity of methanogenic archaea. Three different enrichment cultures from the same source material were obtained with MeOH, H₂/CO₂, or acetate as growth substrate. The mixed cultures were transferred at least six times on the respective substrate and were used to examine the impact of N₂O on CH₄ production from different methanogenic substrates (see Supplemental Information for additional information on the mixed methanogenic cultures). Experiments were performed in triplicate 60-ml glass serum bottles with 30 ml N₂/CO₂ (80/20, v/v) headspace and 30 ml bicarbonate-buffered (50 mM) mineral salt medium (pH 7.2) reduced with 0.2 mM sulfide and 0.2 mM L-cysteine [44]. Acetoclastic, methylotrophic, and hydrogenotrophic cultures received 20 mM acetate, 30 mM MeOH, and 1.24 mmol H₂, respectively. To avoid overpressure in bottles with H₂ as electron donor, the headspace of culture bottles was replaced with 30 ml of filter-sterilized H₂. For M. barkeri cultures, 0.1–1.0 ml of N₂O gas (undiluted or 10-fold diluted in N₂) was directly added to the incubation vessels to achieve final aqueous N₂O concentrations of 100 and 200 μM in cultures with MeOH, 50 and 100 μM in cultures with H₂, and 20 and 50 μM in cultures with acetate. For the methanogenic mixed cultures grown with 30 mM MeOH, 30 ml H₂ (1.24 mmol), and 20 mM acetate, 0.5–1.6 ml of 10-fold diluted N₂O gas (in N₂) was introduced to achieve final aqueous phase N₂O concentrations of 10 and 30 μM. More detailed information about N₂O additions and concentration calculations is provided in the Supplemental Methods. All cultures were incubated without agitation at 37°C in the dark with the stoppers facing up, and replicates without N₂O and without inoculum served as positive and negative controls, respectively. CH₄ and N₂O were analyzed throughout the growth experiments by injecting 100 μl headspace samples into an Agilent 3000A Micro gas chromatograph equipped with thermal conductivity detectors and a molecular sieve column and a PLOT Q column for CH₄ and N₂O measurements, respectively.

Whole-cell suspension assays to determine N₂O inhibition constants for CH₄ production

The M. barkeri and the methanogenic mixed cultures were first grown in 1.6 l medium and harvested via centrifugation when about two-thirds of the initial substrate (i.e. MeOH, H₂/CO₂, or acetate) had been converted to CH₄, as calculated based on CH₄ production according to Equations (1)–(3) (see below). The cell pellets collected from 1.6 l of medium were washed and suspended in 1.6 ml of reduced mineral salt medium in sealed 2-ml glass vials, resulting in a 1000-fold concentration of the biomass. A 0.2 ml aliquot of the concentrated cell suspension was sacrificed to measure total protein with the Bradford assay [45].

Cell suspension assays were performed at room temperature in 20-ml glass vials flushed with N₂/CO₂ (80/20, v/v) and were sealed with Teflon-lined butyl rubber stoppers held in place with aluminum crimps. The assay vials received a total of 0.9 ml reduced mineral salt medium, 0.1 ml of cell suspension, and increasing concentrations of substrates (i.e. MeOH, H_2, or acetate, with N₂O as indicated in Tables S1–S7). For assay vials receiving H₂ as electron donor, the headspace was replaced with increasing volumes of premixed H₂/CO₂ (4/1, v/v) to achieve H₂ concentrations ranging from 1.2 to 333 μM (Table S3). Small volumes (89–178 μl) of undiluted or 10-fold diluted (in N₂) N₂O were directly injected into the 20-ml assay vials, what resulted in small pressure changes with negligible impact on the distribution of N₂O between the aqueous phase and the headspace. All cell suspensions were freshly prepared following identical procedures to ensure consistency between independent experiments. Vials that received 0.1 ml of sterile mineral salt medium or 0.1 ml of heat-killed (i.e. autoclaved) cell suspension served as negative controls.

Following 10 min of equilibration, 0.1 ml of cell suspension was added to initiate the assays. Headspace samples (100 μl) were withdrawn with an air-tight syringe with a lock every 30 min over a 3-h incubation period and, for the kinetic experiments, CH₄ was analyzed using an Agilent 7890 GC series gas chromatograph equipped with a flame ionization detector and a DB-624 capillary column (60 m length × 0.32 mm diameter, 1.8 μm film thickness). For each treatment at a fixed initial substrate concentration [S], an initial CH₄ production rate v, normalized to the amount of protein per vial in the unit of nmol CH₄ min⁻¹ mg protein⁻¹, was determined. The determined rate data were fit into Michaelis–Menten competitive, noncompetitive, and uncompetitive inhibition models (Table S8) to determine the maximum CH₄ production rate V_max, the half-velocity constant K_m, and the inhibitory constant K_I of N₂O on CH₄ production from the different substrates. The best-fit inhibition model was chosen based on the highest coefficient of determination (R²) and the lowest standard deviation of the residuals (Table S9). Each datum point on the Michaelis–Menten plots (Figure 5) represents a CH₄ production rate generated from at least four time points for one substrate concentration [S].

Genomic DNA extraction and 16S ribosomal RNA gene amplicon sequencing

DNA extraction and PCR assays followed established procedures [46] and details are provided in the Supplemental Information. The purified DNA samples were processed and barcoded with primers 341F/785R targeting the V3/V4 region of the prokaryotic 16S rRNA gene [47] following established procedures [48]. The resulting sequence data were analyzed using the QIIME 2 v2021.4 environment [49]. The precise programs and settings are described in the Supplemental Information, and the QIIME 2 pipeline script and custom R file employed to parse results are available at https://github.com/rwmurdoch/methanogens_and_N2O. The raw amplicon library reads were deposited in the Sequence Read Archive (SRA) under accessions SRR19782291 to SRR19782296.

Quantitative real-time PCR and growth yield calculations

Quantitative real-time PCR (qPCR) to enumerate archaeal 16S rRNA genes in M. barkeri and methanogenic mixed cultures followed established protocols using primer set Mtgen835F/918R and probe FAM-Mtgen831 (Table S10) [50]. Samples for enumeration of cell numbers were collected at the beginning and at the end of the prolonged growth experiments. Average growth yields of methanogens were calculated from the changes in 16S rRNA gene copy numbers in triplicate culture vessels divided by the total amounts of CH₄ produced over the same time period (Table S11). Reported growth yield (i.e. cells produced per μmol of CH₄ formed) used conversion factors of 3 and 2.5 16S rRNA gene copies per methanogen cell for M. barkeri [51] and the enrichment cultures [52], respectively. For comparison with theoretical [53] and reported values from the literature (Table 2), growth yields were also calculated as μg of dry biomass per μmol of CH₄ formed making the following assumptions: an average methanogen cell has a volume of 2.5 μm³ [54] with a density equal to water [55]. The dry cell biomass is 30% of the wet cell biomass [56], and 90% of the dry biomass represents organic material [55].

Results

N₂O has distinct impact on CH₄ production from different methanogenic substrates

In the absence of N₂O, MeOH-grown M. barkeri cultures consumed 895 ± 10 μmol of MeOH within a 6-day incubation period and produced 635 ± 34 μmol of CH₄ (Fig. 2A). This stoichiometry closely matched the expected CH₄ production based on Equation (1):

Growth yields of 5.3 × 10⁶ ± 0.3 × 10⁶ cells per μmol CH₄ were measured for cultures without N₂O (Fig. 2B). By contrast, cultures that received 100 or 200 μM N₂O produced negligible amounts of CH₄ during the initial 6-day incubation period without any apparent growth after 6 days. Following an 11-day lag phase, cultures with 100 μM N₂O started consuming MeOH, and 626 ± 20 μmol of CH₄ were produced following a 38-day incubation period, indicating that 100 μM N₂O delayed, but did not prevent, CH₄ production from MeOH by M. barkeri (Fig. 2A). Although complete conversion of MeOH to CH₄ according to Equation (1) was achieved in the presence of 100 μM N₂O over a prolonged 38-day incubation period, the growth yield decreased by 63.8% ± 7.8% compared to cultures without N₂O, (i.e. 1.9 × 10⁶ ± 0.4 x 10⁶ vs. 5.3 × 10⁶ ± 0.3 × 10⁶ cells were produced per μmol of CH₄ formed) (Fig. 2B). In the presence of 200 μM N₂O, only 55 ± 11 μmol of CH₄ were produced over a 38-day incubation period, and the growth yield decreased by over 80% to 0.9 × 10⁶ ± 0.4 × 10⁶ cells per μmol CH₄, indicating a pronounced inhibitory effect of N₂O on CH₄ production and growth of M. barkeri.

More pronounced N₂O inhibition on CH₄ production and growth was observed in M. barkeri cultures that received H₂ as electron donor (i.e. hydrogenotrophic methanogenesis). In the absence of N₂O, M. barkeri cultures, that received H₂ as electron donor, produced 318 ± 21 μmol of CH₄ from 1.24 mmol of H₂ over an 11-day incubation period, consistent with Equation (2):

In the absence of N₂O, a growth yield of 0.8 × 10⁶ ± 0.2 × 10⁶ cells per μmol of CH₄ formed was measured (Fig. 2C and D). In the presence of 50 or 100 μM N₂O, CH₄ production by M. barkeri cultures commenced after prolonged lag phases ranging from 13 to 17 days. At the end of the 38-day incubation period, M. barkeri cultures with 50 μM N₂O and produced 311 ± 7 μmol of CH₄. The M. barkeri cultures that had received 100 μM N₂O only generated 17 ± 3.2 μmol of CH₄, resulting in a 95% decrease in total CH₄ production compared to cultures without N₂O. The average growth yield in cultures with 50 or 100 μM N₂O decreased by ~70% and 85% to 2.3 × 10⁵ ± 0.3 × 10⁵ and 1.2 × 10⁵ ± 0.3 × 10⁴ cells per μmol CH₄ formed, respectively, compared to cultures without N₂O.

The most pronounced N₂O inhibition was observed in acetate-fed M. barkeri cultures, and 20 μM N₂O severely diminished CH₄ production (Fig. 2E). In the absence of N₂O, M. barkeri produced 588 ± 24 μmol of CH₄ from 605 ± 17 μmol of acetate over a 38-day incubation period, consistent with Equation (3):

In the presence of 20 or 50 μM N₂O, M. barkeri cultures only produced 21.3 ± 4.1 and 20.4 ± 3.8 μmol of CH₄, respectively, a decline of over 96% in total CH₄ production compared to cultures without N₂O over the 38-day incubation period. Acetate-grown M. barkeri cultures that had received 20 or 50 μM N₂O generated 1.5 ± 0.2 × 10⁶ and 1.0 ± 0.3 × 10⁶ cells per μmol of CH₄ produced, decreases of 30.4% ± 3.3% and 52.5% ± 12.6%, respectively, compared to the average yield of 2.1 × 10⁶ ± 0.7 × 10⁶ cells per μmol CH₄ in cultures without N₂O (Fig. 2F).

In all culture vessels with observed inhibition of methanogenic activity, N₂O concentrations remained constant throughout the experiment, consistent with the absence of a “nos” operon on the genome of M. barkeri. Taken together, these results demonstrate that micromolar levels of N₂O inhibit methanogenesis, reduce growth yields of this model methanogen, and further reveal that the inhibition is most pronounced for acetoclastic methanogenesis.

N₂O adversely affects CH₄ production and methanogen growth yields in mixed methanogenic enrichment cultures

To examine the impact of N₂O on mixed methanogen communities, growth and kinetic assays were performed with enrichment cultures derived from digester sludge. Microbial community analysis revealed the presence of diverse methanogen groups known to utilize acetate, H₂/CO₂, and MeOH as substrates, with distinct methanogen taxa prevalent under the different enrichment conditions (Tables S12 and S13). In cultures that received H₂ as electron donor, sequences representing the genus Methanobacterium and other unidentified members of the family Methanobacteriaceae dominated, whereas in acetate-fed cultures, Methanosarcina was the most abundant archaeal taxon. Methanomethylovorans was the most abundant archaeal genus in the MeOH-fed cultures, but sequences representing the genus Methanomassiliicoccus were also prevalent. 16S rRNA gene amplicons representing six families and four of the known eight orders of methanogens were represented in the examined mixed cultures derived from digester sludge (Fig. 3A and B; Table S12).

Without N₂O addition, the mixed cultures produced 542.6 ± 23.3, 316.5 ± 25.4, and 589.2 ± 33.7 μmol of CH₄ from 0.9 mmol of MeOH, 1.24 mmol of H₂, or 0.6 mmol of acetate, consistent with Equations (1)–(3). When cultures were amended with 10 or 30 μM N₂O, CH₄ production in MeOH-, H₂-, and acetate-fed methanogenic mixed cultures was substantially or completely inhibited (Fig. 4A, C, and E). Even over an extended 42-day incubation period, the presence 10 μM N₂O still repressed the total CH₄ production in MeOH-, H₂-, or acetate-grown mixed cultures by ~60%, 80% and 50%, respectively, compared to incubations without N₂O. With 30 μM N₂O, only negligible amounts of CH₄ were detected in all incubations over a 42-day incubation period. Some N₂O loss was observed in the mixed culture vessels that received H₂/CO₂ or MeOH as substrates, with no more than 20% of the initial amount of N₂O consumed. Taken together, these results demonstrate that N₂O exerts a stronger inhibitory effect on methanogenesis in the mixed cultures harboring diverse methanogen populations than in axenic M. barkeri incubations.

Enumeration of total methanogens and total bacteria in the mixed cultures using qPCR revealed that N₂O diminished methanogen growth yields (Fig. 4B, D, and F). The qPCR analysis further revealed significantly decreased ratios of methanogen-to-bacterial 16S rRNA genes in all N₂O-treated cultures (Fig. 3C–E), illustrating that N₂O impacted the methanogen populations much more strongly than the bacterial populations. In the absence of N₂O, the average growth yields of methanogens in the enrichment cultures with MeOH, H₂, or acetate were 0.6 × 10⁷ ± 0.1 × 10⁷, 0.6 × 10⁵ ± 0.3 × 10⁵, and 1.0 × 10⁶ ± 0.1 × 10⁶ cells per μmol of CH₄ formed, respectively. In the presence of 10 μM N₂O, the growth yields of methanogens declined by ~90%, 60%, and 50% in enrichment cultures that received MeOH, H₂, and acetate, respectively (Fig. 4B, D, and F). Only negligible CH₄ production and methanogen growth were measured in all mixed cultures with 30 μM N₂O (Fig. 4B, D, and F). Taken together, the results of the mixed culture studies corroborate that N₂O concentrations in the low micromolar range exhibit pronounced inhibitory effects on hydrogenotrophic, methylotrophic, and acetoclastic methanogenesis in microbial communities.

Kinetic studies confirm potent N₂O inhibition of methane formation rates

Whole-cell suspension assays using M. barkeri and the methanogenic mixed cultures were performed to quantitatively assess the inhibitory effects of N₂O on CH₄ production from each methanogenic substrate (Table S1). The Michaelis–Menten single-substrate single-inhibitor model (R² > 0.95) best explained the trends of CH₄ production rates versus increasing substrate concentrations (Fig. 5). Among all assays with M. barkeri and the mixed cultures, maximum CH₄ production rates (i.e. V_max values) of acetate-fed cultures were most strongly affected by increasing N₂O concentrations, followed by the H₂- and MeOH-fed cultures (Figs 2 and 3).

In the absence of N₂O, the V_max values for CH₄ production in MeOH-, H₂-, or acetate-amended M. barkeri cell suspension assays were 440.4 ± 10.2, 138.1 ± 6.9, and 40.8 ± 4.7 nmol CH₄ min⁻¹ mg protein⁻¹, respectively (Fig. 5A, Table 1). The addition of N₂O decreased the V_max of CH₄ production in MeOH-, H₂-, and acetate-fed M. barkeri cell suspension assays to different extents. Rate data determined in M. barkeri suspensions assays with MeOH fit the Michaelis–Menten inhibition model best. The V_max values declined by ~45% and 57% to 244.1 ± 38.4 and 188.8 ± 32.5 nmol CH₄ min⁻¹ mg protein⁻¹, respectively, in the presence of 100 and 200 μM N₂O. The determined inhibitory constant, K_I, of N₂O on methylotrophic CH₄ production was 130.9 ± 4.7 μM in M. barkeri cell suspensions (Table 1), indicating that N₂O concentrations around 130 μM reduced the maximum CH₄ production rate (V_max) by 50%. More pronounced N₂O inhibition was observed in M. barkeri whole-cell suspensions assays using H₂ as the electron donor for CO₂ reduction (Fig. 5B). The addition of 50 and 100 μM N₂O reduced the V_max values in H₂-fed M. barkeri cell suspension assays by ~40% and 57% to 82.8 ± 29.5 and 59.9 ± 5.9 nmol of CH₄ min⁻¹ mg protein⁻¹, respectively. The model simulation determined a K_I value of 90.6 ± 10.8 μM N₂O (Fig. 5B, Table 1), indicating a stronger inhibition of N₂O on CH₄ production in H₂- versus MeOH-amended M. barkeri cell suspension assays. The M. barkeri assays with acetate as substrate showed the most pronounced inhibition by N₂O (Fig. 5C, Table 1). In assays amended with 20 and 40 μM N₂O, the V_max values decreased by 21% and 44% to 32.1 ± 4.2 and 22.8 ± 2.5 nmol of CH₄ min⁻¹ mg protein⁻¹, respectively. From the best-fit inhibition model, a K_I value of 24.8 ± 2.6 μM was determined for N₂O inhibition of acetoclastic methanogenesis in M. barkeri cell suspensions. Collectively, the cell suspension assays corroborate strong inhibitory effects of N₂O on methanogenesis, with the kinetics of acetoclastic methanogenesis being most impacted by N₂O.

Similar kinetic responses were observed for whole-cell suspension assays conducted with the methanogenic mixed cultures pregrown with the respective substrates (Fig. 5D–F). In the absence of N₂O, the determined V_max values for mixed culture cell suspension assays that received MeOH, H₂, or acetate were 209.0 ± 4.4, 105.8 ± 3.7, and 22.8 ± 0.7 nmol CH₄ min⁻¹ mg protein⁻¹, respectively (Table 1). Notably, the methanogenic mixed cultures were more sensitive to N₂O than axenic M. barkeri cultures irrespective of the type of methanogenic substrate provided. In cell suspension assays that received MeOH, the presence of 50 and 100 μM N₂O reduced the V_max values by 37.4 ± 10.8% and 75.4 ± 21.4%, respectively, compared to assays without N₂O. The inhibition model determined a K_I value of 109.9 ± 6.8 μM for N₂O inhibition on CH₄ production in cell suspension assays amended with MeOH (Fig. 5D, Table 1). In H₂-amended cell suspension assays, the presence of 30 and 60 μM N₂O decreased the V_max values by 32.3 ± 9.8 and 65.4 ± 12.9%, respectively, compared to assays without N₂O (Fig. 5E, Table 1). The best-fit inhibition model determined a K_I value for N₂O inhibition of 62.1 ± 6.4 μM for CH₄ production in assays that received H₂ as electron donor. Consistent with the observations made with M. barkeri, the most pronounced N₂O inhibition on CH₄ production rates was observed in mixed culture cell suspension assays that received acetate as substrate (Fig. 5F). In the presence of 10 and 30 μM N₂O, the V_max values in suspension assays with acetate decreased by 38 ± 7.8% and 74.9 ± 16.0%, respectively, compared to assays without N₂O. Using the best-fit Michaelis–Menten inhibition model, a K_I value of 17.7 ± 1.8 μM was determined for N₂O inhibition of CH₄ production in cell suspension assays with acetate (Table 1).

Taken together, the experimental data demonstrate that dissolved N₂O concentrations in the low μM range (i.e. 20–100 μM) repress CH₄ production and reduce growth yields of M. barkeri in axenic culture and of different methanogen guilds in methanogenic mixed cultures. Evaluation of the kinetic data determined in cell suspension assays revealed distinct N₂O inhibition patterns for CH₄ production from acetate, H₂, and MeOH, with the rate of acetoclastic CH₄ production being most sensitive to N₂O inhibition. The K_I values for N₂O inhibition of CH₄ production from key methanogenic substrates ranged between 18 and 130 μM, suggesting N₂O can impact CH₄ production and emissions in diverse ecosystems, including critical zone environments (e.g. thawing permafrost).

Discussion

Microorganisms drive global C and N cycling and ultimately control CH₄ and N₂O production, consumption, and thus emissions to the atmosphere [10, 11, 13]. Predictive climate models must consider the responses of microbial CH₄ and N₂O production under environmental change scenarios [10, 57]. Quantitative assessment of feedbacks that affect CH₄ production are crucial for refining greenhouse gas emission models. This study investigated the feedback effects between two important greenhouse gases, specifically the inhibitory effect of N₂O on archaeal CH₄ production, to provide quantitative data that link environmental N₂O concentrations with methanogenesis. The findings demonstrate that environmentally relevant, micromolar levels of N₂O suppress CH₄ production and the growth of methanogens. The determined K_I values reveal a concentration-dependent, progressively negative feedback by N₂O on archaeal CH₄ formation rates and the total amounts of CH₄ produced and show that the strength of the inhibition is most pronounced for acetoclastic methanogenesis.

Assuming the current day partial pressure of 335 ppb N₂O in the atmosphere, the theoretical concentration of N₂O in air-equilibrated water should be around 7 nM; however, substantially elevated levels of dissolved N₂O have been observed in groundwater and watersheds [33, 42]. In areas impacted by agricultural activities (e.g. fertilizer application), dissolved groundwater N₂O concentrations can exceed 100 μM [42]. Even in some remote, natural aquatic systems, such as ice-covered Antarctic lakes, N₂O concentrations of up to 86 μM have been reported [58, 59]. N₂O is generated during N cycling and major formation processes include microbial denitrification, ammonia oxidation, and abiotic chemodenitrification [44]. Based on the physiology of the microorganisms (e.g. ammonia oxidizers are strict aerobes) and the thermodynamics of the processes (e.g. nitrate and nitrite reduction are associated with a greater change in Gibbs free energy than methanogenesis), one might argue that N₂O formation and methanogenesis are physically separate processes, thus limiting the exposure and inhibitory effects of N₂O on methanogens. Such redox stratification does occur; however, most environmental matrices, such as soils, are highly heterogenous and characterized by dynamic spatial and temporal gradients resulting in patchy distribution of redox processes. N₂O is water-soluble and, depending on hydrology, can reach other redox zones [60]. Consequently, impacts of elevated N₂O on various biogeochemical processes, including those associated with greenhouse gas emissions, are likely. Laboratory studies have reported inhibitory effects of nitrogen oxides (NOx), including N₂O, on methanogenesis [34–37, 61, 62]; however, the available data are scarce and no uniform pattern has emerged that would support a quantitative relationship between N₂O and microbial CH₄ production. Consequently, this negative feedback of N₂O on CH₄ production has not been considered in greenhouse gas emission models.

The evaluation of N₂O inhibition on CH₄ production from methanogenic substrates (i.e. MeOH, acetate, and H₂/CO₂) with both the model methanogen M. barkeri and digester sludge-derived mixed methanogenic cultures quantitatively links N₂O concentrations with methanogen activity and growth. The growth experiments illustrate that micromolar N₂O concentrations affect CH₄ production, and the whole-cell suspension assays and kinetic model simulations provide a plausible explanation for the inconsistent literature reports about N₂O effects on methanogenesis. Specifically, acetoclastic methanogenesis was most sensitive to N₂O (K_I values of 18–25 μM), followed by hydrogenotrophic (K_I 60–90 μM) and methylotrophic methanogenesis (K_I 110–130 μM), indicating that the type of methanogenic substrate utilized affects the sensitivity of methanogens to N₂O. N₂O inhibition was significantly more pronounced in methanogenic enrichment cultures than in axenic M. barkeri cultures, and 30 μM N₂O prevented CH₄ production and methanogen growth in the mixed culture experiments regardless of the type of methanogenic substrate utilized. Previous studies support that mixed methanogenic communities are more sensitive to N₂O than commonly studied model methanogen isolates [34, 35, 62, 63]. These observations suggest that the axenic methanogen cultures used to elucidate the biochemistry and genetics of methanogenesis may not serve as general models for other features of methanogen biology (e.g. N₂O inhibition). The reasons for the reduced sensitivity of axenic versus mixed methanogen cultures to N₂O may be a result of long-term adaptation of the isolates to laboratory cultivation, or not-yet-characterized microbe–microbe interaction networks render mixed cultures more susceptible to N₂O inhibition. The enrichment cultures used to determine the K_I values for N₂O inhibition harbored diverse methanogen groups (Fig. 3A), and kinetic studies (e.g. determination of K_I values for N₂O inhibition) with representative isolates of the various lineages are warranted to capture the breadth of methanogen responses to N₂O.

Taken together, the low μM range K_I values of N₂O that impact methanogenesis suggest major consequences of rising N₂O concentrations for C cycling. Of note, the vast majority of the annual ~1 billion metric tons of biogenic CH₄ is generated from acetate- and H₂-driven CO₂ reduction [16–20, 64], the two processes with the lowest observed K_I values for N₂O inhibition. It is therefore reasonable to predict that elevated environmental N₂O will impact CH₄ production and methanogen growth in ecosystems with high bioavailable C and N loads, such as wetlands, sediments, and permafrost soils.

Key enzyme systems involved in CH₄ production and energy conservation require cobamide prosthetic groups [65] (Fig. 1). The super-reduced Co(I) form of cobamides is susceptible to oxidants such as N₂O, a plausible mechanism for the observed inhibition of methanogenesis [29, 34, 63]. The experimental efforts demonstrated that acetoclastic methanogenesis was most sensitive to N₂O, with K_I values in the range of 18–25 μM, indicating that N₂O concentrations in the range of 20 μM would reduce the V_max of CH₄ production from acetate by 50%. The K_I values for N₂O inhibition determined for hydrogenotrophic and methylotrophic methanogenesis ranged between 60–90 and 110–130 μM, respectively. The reasons for the apparently substrate-specific K_I values for N₂O inhibition likely reflect differences in the pathways leading to CH₄ production from acetate, H₂/CO₂ and MeOH (Fig. 1). The acetoclastic pathway involves two steps catalyzed by corrinoid-dependent enzyme systems, CODH/ACS and the MTR corrinoid enzyme complex [16, 65]. (Fig. 1), both of which are targets for N₂O inhibition. Distinct cobamide-dependent enzyme systems catalyze the formation of CH₃–CoM in the hydrogenotrophic and the methylotrophic pathways [22, 66]. Both the acetoclastic and hydrogenotrophic pathways depend on the MTR corrinoid enzyme complex to generate CH₃–CoM and to conserve energy [22, 27]. By contrast, methylotrophic methanogens can directly generate CH₃–CoM without the energy-conserving MTR corrinoid enzyme complex from methylated compounds (e.g. MeOH) via a substrate-specific, corrinoid-dependent methyltransferase complex (i.e. MtaA, MtaB, and MtaC in the case of MeOH) [67]. The experimental efforts consistently demonstrated that CH₄ production from acetate and H₂/CO₂ exhibited 7- and 3-fold higher sensitivities, respectively, to N₂O than CH₄ production from MeOH. These findings suggest that N₂O inhibition of methanogenesis is related to the oxidation of the super-reduced Co(I) cobamide, which is essential for the corrinoid enzyme complexes involved in the different methanogenesis pathways. While the differential susceptibilities of pathway-specific corrinoid-dependent enzymatic steps can explain the distinct inhibitory effects of N₂O on CH₄ production via the acetoclastic, hydrogenotrophic, and methylotrophic pathways, detailed enzymatic studies would be needed to assess the responses of individual enzyme systems to N₂O. Kinetic data for the enzymatic regeneration of the super-reduced Co(I) state are lacking, and the rates of reduction may differ between enzymes and pathways, which could contribute to the observed physiological response of decreased CH₄ production. Recovery from N₂O inhibition was not a focus, but the experimental data indicate partial recovery of methane formation in N₂O-treated axenic and mixed cultures; however, the methanogen growth yields remained lower in N₂O-treated cultures compared to controls without N₂O even over the extended incubation period. Further studies are needed to generate mechanistic insights into modes of recovery from N₂O inhibition.

The presence of N₂O significantly decreased growth yields or completely abolished growth in the examined methanogenic cultures (Table 2). The type of methanogenic substrate utilized determines the fraction of electrons available from electron donor oxidation directed toward cell synthesis and thus governs the growth yields of methanogens [53, 68]. N₂O affects corrinoid-dependent enzymes involved in electron transfer (e.g. the MTR enzyme complex), and it is not surprising that N₂O interferes with energy conservation in methanogens. Consistently, enumeration of methanogen 16S rRNA genes at the termination of all growth experiments illustrated that N₂O not only negatively impacted CH₄ production but also methanogen growth yields. An alternate explanation for the reduced methanogen growth yields in the mixed cultures exposed to N₂O could be competition for electron donor (e.g. H₂); however, the sequencing and the qPCR data do not support this hypothesis, and N₂O inhibition explains the decline of methanogens and the changes of methanogen-to-bacteria ratios. The measured methanogen growth yield data in the absence of N₂O were on par with reported experimental data and closely matched the theoretical values (i.e. yields calculated based on thermodynamics) (Table 2). One exception were the growth yields measured in H₂/CO₂-fed M. barkeri cultures, which were ~10-fold lower than data reported in the literature. The most pronounced growth suppression was observed in the mixed methanogenic cultures, where 30 μM N₂O was sufficient to prevent the growth of methanogenic archaea. Collectively, the data show that micromolar concentrations of N₂O decrease or abolish CH₄ production, reduce methanogen growth yields, and exhibit progressively negative feedback on microbial CH₄ production.

Microbial processes are strongly influenced by environmental factors and their responses to climate change vary both spatially and temporally [11, 69]. To improve the predictive power of climate models and potentially justify the application of biotechnological approaches for managing greenhouse gas emissions, interactions and feedbacks between relevant biotic/abiotic processes must be understood and quantitatively captured [70]. Attempts have been made to include microbial data (i.e. biomass, enzyme, and growth kinetics) to improve climate models [70], but the incorporation of multifactorial, multidirectional, and often nonlinear biotic/abiotic feedbacks underlying the global CH₄ and N₂O budgets is challenging and requires robust quantitative data. The determined K_I values for N₂O inhibition of methanogenesis reveal a relevant negative feedback effect on CH₄ emissions, and the new quantitative information generates opportunities to refine CH₄ emission models.

Conflicts of interest

The authors declare no competing interests.

Funding

This work was supported through the Dimensions of Biodiversity Program of the US National Science Foundation (award 1831599 to F.E.L.). Y.Y., Y.X., and Y.S. acknowledge the support from the China Scholarship Council.

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information files. The raw amplicon library reads are available in the SRA repository.

References

Author notes