Pathways of N₂O production by marine ammonia-oxidizing archaea determined from dual-isotope labeling

Significance

Ammonia-oxidizing archaea (AOA) are the major source of marine nitrous oxide (N₂O); however, the cellular kinetics and pathways of archaeal N₂O production remain unclear. We characterize N₂O production kinetics of a model marine AOA species Nitrosopumilus maritimus at low ammonia concentrations and quantify the relative contributions of multiple N₂O production pathways using dual ¹⁵N-¹⁸O isotope labeling. We provide direct evidence for the enzymatic regulation of N₂O production by AOA. We found hydroxylamine oxidation contributes substantially to N₂O production, which had not been previously recognized, and that nitrite reduction is not a significant source of N₂O. These findings are important for the interpretation of pathways and regulation of N₂O production in the ocean.

Abstract

The ocean is a net source of the greenhouse gas and ozone-depleting substance, nitrous oxide (N₂O), to the atmosphere. Most of that N₂O is produced as a trace side product during ammonia oxidation, primarily by ammonia-oxidizing archaea (AOA), which numerically dominate the ammonia-oxidizing community in most marine environments. The pathways to N₂O production and their kinetics, however, are not completely understood. Here, we use ¹⁵N and ¹⁸O isotopes to determine the kinetics of N₂O production and trace the source of nitrogen (N) and oxygen (O) atoms in N₂O produced by a model marine AOA species, Nitrosopumilus maritimus. We find that during ammonia oxidation, the apparent half saturation constants of nitrite and N₂O production are comparable, suggesting that both processes are enzymatically controlled and tightly coupled at low ammonia concentrations. The constituent atoms in N₂O are derived from ammonia, nitrite, O₂, and H₂O via multiple pathways. Ammonia is the primary source of N atoms in N₂O, but its contribution varies with ammonia to nitrite ratio. The ratio of ⁴⁵N₂O to ⁴⁶N₂O (i.e., single or double labeled N) varies with substrate ratio, leading to widely varying isotopic signatures in the N₂O pool. O₂ is the primary source for O atoms. In addition to the previously demonstrated hybrid formation pathway, we found a substantial contribution by hydroxylamine oxidation, while nitrite reduction is an insignificant source of N₂O. Our study highlights the power of dual ¹⁵N-¹⁸O isotope labeling to disentangle N₂O production pathways in microbes, with implications for interpretation of pathways and regulation of marine N₂O sources.

Ammonia-oxidizing archaea (AOA) are ubiquitous and abundant members of the marine plankton; they are almost exclusively responsible for ammonia oxidation in the world ocean (1, 2). Globally, over 80% of marine nitrous oxide (N₂O) is estimated to be produced as a side product of ammonia oxidation (3–5), indicating a dominant role of marine AOA in determining N₂O distribution and its flux from ocean to atmosphere. Compared to their bacterial counterparts, ammonia-oxidizing bacteria (AOB), marine AOA lack the genes encoding the known bacterial machineries for N₂O production (6) and exhibit lower N₂O yield (7–9), implying distinct mechanisms of N₂O production between AOA and AOB (10–12). The marine AOA demonstrate significantly higher affinity toward total ammonia (NH₃ plus NH₄⁺, hereafter referred to as NH₄⁺) for ammonia oxidation during nitrite (NO₂⁻) production than AOB (2), but the cellular kinetics of N₂O production have not been explored. Hydroxylamine (NH₂OH) and nitric oxide (NO) have been identified as key intermediates in AOA metabolism (13–15), implying multiple potential pathways for N₂O production as both NH₂OH and NO are likely precursors of N₂O. However, the explicit pathways of archaeal N₂O production are still incompletely known and remain controversial (10–12). The hybrid N₂O formation pathway (in which the two N atoms in N₂O are derived from different sources via the abiotic reaction between NH₂OH and NO) has been experimentally demonstrated and is considered the dominant pathway for N₂O production in AOA cultures and natural environments (Fig. 1A) (10–12). Other AOA pathways may include N₂O production via NO₂⁻ reduction at low pH (16) and a novel NO dismutation pathway in marine AOA under anaerobic conditions (17). In contrast, no experimental evidence that AOA can directly convert NH₂OH to N₂O via NH₂OH oxidation, either enzymatically or abiotically, has been reported. It is important to determine which pathways occur during archaeal ammonia oxidation and which are relevant in various environmental conditions because 1) the pathway and rate of N₂O production might vary with NH₄⁺ and NO₂⁻ availability, as both are involved in N₂O formation; and 2) different sources of N and O used by AOA to produce N₂O may impart distinct isotope signatures, which are used to deduce the sources of N₂O in natural and man-made systems.

Fig. 1.

Summary of N₂O production pathways during marine archaeal ammonia oxidation. (A) The hybrid pathway, which combines one N atom from NH₂OH and one from nitrite (either ambient or newly produced inside the cell), has been observed previously (10–12). Black arrows represent known pathways for N atoms; dashed pathways represent hypothesized pathways involving NO. (B) Potential N₂O pathways and the N and O atom sources identified in the current study. Five N₂O pathways are identified: 1) NH₂OH oxidation; 2) hybrid pathway a: NH₂OH reaction with NO that was sourced from the reduction of newly produced NO₂⁻; 3) hybrid pathway b: NH₂OH reaction with NO that was sourced from ambient NO₂⁻; 4) reduction of newly produced NO₂⁻; and 5) reduction of ambient NO₂⁻. Color of the N and O atoms depicts the sources: red and black denote N atoms from NH₄⁺ and ambient NO₂⁻, respectively; blue, purple, and gray denote O atoms from O₂, H₂O, and ambient NO₂⁻, respectively; green represents O atoms of newly produced NO₂⁻, which is a mixture of H₂O and O₂. The gray square denotes the membrane and periplasmic space of the AOA cell. The black arrows represent the ammonia oxidation pathway, the red arrows show the potential N₂O production processes, and the dashed arrows indicate potential pathways that are unresolved in our study. (C) Summary of the N, O atom sources of N₂O and the fractional contribution of each pathway during marine archaeal ammonia oxidation with initial NH₄⁺: NO₂⁻ ratio of 1:1.

Elucidation of all potential sources of N₂O, however, remains a challenging task. To date, most investigations on AOA N₂O production pathways have focused on the source of the nitrogen (N) atoms (13, 16–18). However, using N isotopes alone is insufficient to distinguish multiple N₂O production pathways that occur simultaneously and to quantitatively estimate the relative contribution of each pathway. For instance, the ¹⁵N-NH₄⁺ isotope labeling approach cannot completely distinguish N₂O production through NH₂OH oxidation, hybrid formation, or NO₂⁻ reduction, because all these precursors of N₂O could ultimately be sourced from NH₄⁺ during ammonia oxidation. The ¹⁸O-labeling approach provides an independent avenue to identify the source of N₂O by tracking the oxygen (O) atom in N₂O. A few studies have used ¹⁸O-H₂O to show that the O atom in N₂O can be partly derived from H₂O in lab culture and field studies (9, 19, 20). However, the alternative sources of the O atom in N₂O are still unknown. Importantly, because the O atom source differs among the various N-containing precursors, i.e., NH₂OH, NO₂⁻, and NO, dual ¹⁵N and ¹⁸O isotope labeling is a powerful method to disentangle the complex and interconnected N₂O production pathways in AOA. In this study, using a model marine AOA species Nitrosopumilus maritimus strain SCM1 (hereafter refer to as SCM1), we conducted a comprehensive set of dual-isotope labeling incubation experiments to systematically investigate N₂O production kinetics and the associated pathways during archaeal ammonia oxidation under various substrate conditions (SI Appendix, Table S1).

Results and Discussion

Kinetics of N₂O Production during Archaeal Ammonia Oxidation.

Marine AOA have a remarkably high affinity for NH₄⁺, which is consistent with their dominant role in ammonia oxidation to NO₂⁻ in oligotrophic marine environments (2). However, the effect of NH₄⁺ concentrations on N₂O production by AOA remains unknown. If N₂O is primarily generated via abiotic hybrid reactions between intermediates of archaeal ammonia oxidation, N₂O production by AOA may not follow normal enzyme kinetics. We found that both NO₂⁻ and N₂O production rates varied with NH₄⁺ concentration (Experiment 1) and both followed Michaelis-Menten-type kinetics (Fig. 2 A and B). The apparent half saturation constants (K_m(app)) for NO₂⁻ production (220 ± 50 nmol L⁻¹ NH₄⁺) were comparable to those that were previously determined for ammonia (132 nmol L⁻¹ NH₄⁺) and oxygen uptake (133 nmol L⁻¹ NH₄⁺), suggesting all essential enzymatic steps for ammonia oxidation and respiration are highly efficient and tightly coupled at low ammonia concentrations in marine AOA. Likewise, although the maximum rate (V_max) for N₂O production (0.97 ± 0.02 amol N cell⁻¹ d⁻¹) was more than three orders of magnitude lower than that for NO₂⁻ production (7.14 ± 0.25 fmol N cell⁻¹ d⁻¹), K_m(app) values for the two rates were comparably low. The comparable K_m(app) values for ammonia during NO₂⁻ and N₂O production imply that both are controlled by enzyme activity in SCM1 at low NH₄⁺ concentrations. This implies that enzyme activity provides the key intermediates NH₂OH and NO for both NO₂⁻ and N₂O production, i.e., there is no separate completely abiotic reaction that is responsible for N₂O production. The K_m(app) for N₂O production is four orders of magnitude lower than the K_m reported for NO production by SCM1 measured during ammonia oxidation (measured at >2 mol L⁻¹ NH₄⁺) (14). Therefore, the supply of NO is unlikely to be a limiting factor in determining the kinetics of N₂O production, implying a critical role for NH₂OH in determining the observed kinetics.

Fig. 2.

Experiment 1. Kinetics of ammonia oxidation and N₂O production by SCM1. (A) Michaelis-Menten-type plot of substrate-dependent rate of ammonia oxidation to NO₂⁻ (normalized to per cell per day). (B) Michaelis-Menten plot of substrate-dependent rate of N₂O production. Error bars represent SD from triplicate samples. The black lines and gray shadows show the Michaelis-Menten type regressions and the 95% CIs, respectively.

NH₂OH is enzymatically produced by ammonia monooxygenase and rapidly converted to NO₂⁻ (15). The tight coupling of its production and consumption in AOA species at low NH₄⁺ concentrations results in its limiting conversion to N₂O as a side product. However, a small fraction of NH₂OH can escape from being oxidized by the enzymatic reaction, providing the key precursor for N₂O production. For example, only 0.46% of NH₂OH was released during ammonia oxidation by the AOA Nitrososphaera gargensis even at very high NH₄⁺ concentrations [2 mmol L⁻¹ (21)], which is one to two orders of magnitude lower than the reported NO accumulation ratio during ammonia oxidation by SCM1 (14). The similarly high affinities (low K_m) for NO₂⁻ and N₂O production further suggest that both kinetics were determined by the NH₂OH supply; if the conversion of NH₂OH to NO₂⁻ was the rate-limiting step, the production of N₂O should increase continuously with the accumulation of NH₂OH. These results suggest that the SCM1-like marine AOA effectively use the trace level NH₄⁺ in the vast N-depleted ocean for both NO₂⁻ and N₂O generation, providing direct evidence for the capability of marine AOA to dominate N₂O production in the ocean and its subsequent release to the atmosphere.

Impact of NH₄⁺: NO₂⁻ Ratio on Pathways of N₂O Production.

Both NH₄⁺ and NO₂⁻ are involved in N₂O production by AOA (9, 18), but the rates and pathways might vary depending on relative substrate availability. Therefore, we investigated N₂O production under a wide range of ¹⁵NH₄⁺: ¹⁴NO₂⁻ ratios (from 0.05 to 10) (Experiments 1 and 2). The ratio of single labeled N₂O to double labeled N₂O (⁴⁵N₂O: ⁴⁶N₂O) decreased from 5.82 ± 2.58 to 0.37 ± 0.05 as the ¹⁵NH₄⁺: ¹⁴NO₂⁻ ratio increased from 0.05 to 10 (Fig. 3A and SI Appendix, Fig. S1). This dependence on substrate ratio indicates that ⁴⁵N₂O: ⁴⁶N₂O ratio is not a constant but is highly variable and implies that more than one pathway contributes to N₂O production in AOA. The strong and significant correlation (R² = 0.97, P < 0.0001) between the ⁴⁵N₂O: ⁴⁶N₂O ratio and substrate ¹⁵NH₄⁺: ¹⁴NO₂⁻ ratio implies that the relative contributions of different pathways to N₂O production and the source of the N atoms in N₂O should vary with NH₄⁺: NO₂⁻ ratio in the environment (Fig. 3B).

Fig. 3.

Experiment 2. ¹⁵N-N₂O production and isotope composition under different ¹⁵NH₄^{+ 14}NO₂⁻ ratios. (A) ⁴⁵N₂O (single labeled) and ⁴⁶N₂O (double labeled) production rate under different ¹⁵NH₄⁺ and ¹⁴NO₂⁻ concentrations (normalized to per cell per day). All of the ⁴⁵N₂O represents hybrid formation. (B) Regression between ⁴⁵N₂O: ⁴⁶N₂O production rate against ¹⁵NH₄^{+ 14}NO₂⁻ concentration ratio. Error bars represent SD from triplicate samples.

NH₄⁺ and NO₂⁻ are highly dynamic nitrogen cycle components that rarely accumulate in the global ocean. However, NH₄⁺ and NO₂⁻ can accumulate in specific regions (e.g., oxygen minimum zones and eutrophic waters) and depths (e.g., NH₄⁺ maximum, primary NO₂⁻ maximum), leading to substantial variation of NH₄⁺: NO₂⁻ ratio in these biogeochemically active marine environments (22). Our data indicate that N₂O can be produced via distinct pathways by marine AOA and sourced from different N atoms in waters with different ratios of NH₄⁺: NO₂⁻, even though the main source process is always archaeal ammonia oxidation. For example, at the primary NO₂⁻ maximum where NO₂⁻ accumulates, the low NH₄⁺: NO₂⁻ ratio might lead to the higher contribution of NO₂⁻ to N₂O production via the hybrid pathway. In contrast, at the NH₄⁺ maximum and certain hotspots of NH₄⁺ supply such as zooplankton excretion or decay of phytoplankton blooms (23), NH₄⁺ would dominate N₂O formation under the elevated NH₄⁺: NO₂⁻ ratio. These findings provide new insights in interpreting the natural abundance isotope signature of N₂O in the water column of the global oligotrophic oceans, where a subsurface dual-isotope minimum is consistently observed and has been widely interpreted as resulting from N₂O production via nitrifier-denitrification (24–27). These new data would suggest, however, that low NH₄⁺: NO₂⁻ ratio (i.e., relatively higher NO₂⁻ concentration) would also lead to the observed dual-isotope minimum by the incorporation of more isotopically depleted NO₂⁻ by the hybrid pathway during the ammonia oxidation process.

Our results are also important for the interpretation of isotope labeling patterns observed in isotope tracer experiments in the ocean, where production of ⁴⁵N₂O has been generally attributed to the hybrid N₂O formation pathway (one N from NH₄⁺, one N from NO₂⁻). We suggest that ⁴⁶N₂O could also be partially hybrid, from the combination of ¹⁵NH₄⁺ and newly produced ¹⁵NO₂⁻. The contribution of the hybrid pathway would be underestimated by ignoring ⁴⁶N₂O production. However, our results cannot fully explain the high fraction of ⁴⁵N₂O production (i.e., >70%) in the ocean (24, 28–30), nor the finding that the %⁴⁶N₂O is insensitive to short-term experimental NO₂⁻ enrichment (31) (comparison to field observations; SI Appendix, Text 1). Nevertheless, the variable N₂O atom composition by SCM1 under different NH₄⁺: NO₂⁻ ratio provides new insights into marine N₂O pathways and interpretation of its isotope composition. These pathways and the relative contributions of N and O from multiple sources are explored in experiments described below.

Contribwutions of NH₄⁺ and NO₂⁻ to N₂O Production.

The potential pathways and relative contributions of the two substrates, NH₄⁺ and NO₂⁻, via various pathways of N₂O formation in AOA were explored using multiple tracer combinations (Experiment 3). NH₄⁺ and NO₂⁻ concentrations were controlled by adding the substrates to cells that had first been washed and resuspended in substrate-free fresh medium. When ¹⁵NH₄⁺ was added without ¹⁴NO₂⁻, double labeled ⁴⁶N₂O was the main product (93.4 ± 10.1% of the total labeled N₂O production rate) (Fig. 4A). ¹⁵NH₄⁺ was the only N source in the experiment, so the small production of ⁴⁵N₂O (6.6% of the total labeled N₂O production) can be attributed to trace amounts of intracellular ¹⁴NH₄⁺ and/ or ¹⁴NO₂⁻, or to carry over from the inoculum. When equimolar amounts of ¹⁵NH₄⁺ and ¹⁴NO₂⁻ were provided, the fractional contribution of ⁴⁵N₂O increased to 28.9%, indicating that ambient NO₂⁻ is involved in ⁴⁵N₂O production, although the labeled N₂O pool was still primarily ⁴⁶N₂O. This hybrid N₂O formation indicates involvement of ambient NO₂⁻ in N₂O production, but the process varies among AOA strains (18) and with substrate ratio (Fig. 3B). Newly produced (presumably intracellular, or at least in the pseudo-periplasmic space) and ambient NO₂⁻ might both be involved in N₂O hybrid formation (pathways 2, 3 in Fig. 1B). Total labeled N₂O (combined ⁴⁵N₂O and ⁴⁶N₂O) production rate in the ¹⁵NH₄⁺ tracer incubation (8.7 ± 0.7 nmol N L⁻¹ d⁻¹, Fig. 4A) was comparable to the rate in the ¹⁵NH₄⁺ + ¹⁴NO₂⁻ incubation (9.6 ± 1.5 nmol N L⁻¹ d⁻¹, Fig. 4B), indicating no discernible difference between the effects of ambient and newly produced NO₂⁻ on N₂O production rate by SCM1.

Fig. 4.

Experiment 3. ¹⁵N-N₂O production during ¹⁵NH₄⁺ and ¹⁵NO₂⁻ labeling incubations using viable cells. (A–D) ^l5N labeled N₂O production rate from ¹⁵NH₄⁺, ¹⁵NH₄⁺ + ¹⁴NO₂⁻, ¹⁵NO₂⁻, and ¹⁵NO₂⁻ + ¹⁴NH₄⁺ labeling incubations, respectively. Error bars represent SD from triplicate incubations. Total N₂O refers to total labeled N₂O (⁴⁵N₂O + ⁴⁶N₂O). Approximately, 6 nmol N L⁻¹ d^{−1 44}N₂O must have been produced in the ¹⁵NO₂⁻ + ¹⁴NH₄⁺ incubation (D), but the amount of ⁴⁴N₂O could not be determined in these experiments due to lack of sensitivity in small volume incubations. ⁴⁴N₂O would not have been present in the other three experiments because in A all the NH₄⁺ was labeled; in B we have shown that N₂O cannot be formed from NO₂⁻ alone, and C all the NO₂⁻ was labeled.

¹⁵N-labeled N₂O production from ¹⁵NO₂⁻ tracer was negligible (0.3 ± 0.2 nmol N L⁻¹ d⁻¹) in the absence of NH₄⁺ (Fig. 4C). By comparison, when ¹⁴NH₄⁺ was added with ¹⁵NO₂⁻, the ¹⁵N-labeled N₂O production rate increased significantly to 2.6 ± 0.3 nmol N L⁻¹ d⁻¹ (P < 0.001), which is strong evidence for a hybrid N₂O formation mechanism that involves both intermediates from ammonia oxidation and NO₂⁻ reduction (Fig. 4D). It is not surprising that N₂O production rate decreased greatly in the absence of NH₄⁺, as NH₄⁺ is the substrate for energy generation and the source of the key N₂O precursor NH₂OH. Thus, N₂O can be produced from NH₄⁺ alone in the absence of NO₂⁻, but N₂O cannot be produced from NO₂⁻ alone. When NH₄⁺ is present, however, NO₂⁻ contributes to N₂O formation via a hybrid pathway. In contrast to the ¹⁵NH₄⁺ tracer experiment, ⁴⁵N₂O dominated the ¹⁵N-N₂O pool in the ¹⁵NO₂⁻ + ¹⁴NH₄⁺ incubation (92.1 ± 16.8%). The small fraction (7.9 ± 3.4%) of ⁴⁶N₂O detected in this treatment indicated a minor contribution of the chemo-denitrification-like pathway (i.e., both N atoms from NO₂⁻, pathways 4, 5 in Fig. 1B) to N₂O production (Fig. 4D) (13). However, the ⁴⁶N₂O production rate from ¹⁵NO₂⁻ measured here was two orders of magnitude lower than rates measured when external oxygen was exhausted (5 to 22 nmol L⁻¹ h⁻¹) (17), indicating production of N₂O in AOA by the proposed NO₂⁻ reduction-NO dismutation pathway is restricted to anoxic conditions.

The Role of NH₂OH as Key Precursor for N₂O Production.

Although NH₂OH is not an important N source in the marine environment (its reactivity guarantees a very low ambient concentration) (32), it is a critical intracellular intermediate in archaeal ammonia oxidation and N₂O production. Experiments using ¹⁵NH₂OH and ¹⁵NO₂⁻ + ¹⁴NH₂OH were used to explore the pathways by which NH₂OH participates in N₂O production during ammonia oxidation (Experiment 4). Notably, here we used 1 μmol L⁻¹ of NH₂OH concentration, which was two orders of magnitude lower than previous studies that explored N₂O pathways by AOA (i.e., 200 μmol L⁻¹) (13), thus closer to environmental concentrations but also ensuring the sensitivity of the assays to probe the potential mechanisms. Moreover, no discernible difference was observed for NO₂⁻ production with or without NH₂OH amendment, suggesting no detectable inhibition of 1 μmol L⁻¹ of NH₂OH on archaeal ammonia oxidation (P > 0.05) (SI Appendix, Fig. S2).

⁴⁵N₂O increased slowly but continuously over 12 h from tracer ¹⁵NO₂⁻ in viable cells, while no discernible ⁴⁶N₂O accumulation was detected (Fig. 5A). ¹⁴NH₂OH added with ¹⁵NO₂⁻ stimulated ⁴⁵N₂O production from ¹⁵NO₂⁻ in viable cells, indicating ¹⁴NH₂OH was directly involved in the reaction with ¹⁵NO₂⁻ to produce hybrid ⁴⁵N₂O (Fig. 5B). In contrast, ⁴⁵N₂O production nearly stopped after removing the cells by filtration. The fact that the viable cells produced more ⁴⁵N₂O from the ¹⁵NO₂⁻ tracer than the filtrate indicates that cellular metabolism facilitated N₂O production from the hybrid pathway, and that abiotic formation rate of N₂O by NO₂⁻ and NH₂OH was low (Fig. 5C).

Fig. 5.

Experiment 4. ¹⁵N-N₂O production from NH₂OH. ^l5N labeled N₂O production rate from (A) viable cells with ¹⁵NO₂⁻ tracer (10 μmol L⁻¹); (B) viable cells with ¹⁵NO₂⁻ (10 μmol L⁻¹) + ¹⁴NH₂OH (1 μmol L⁻¹); (C) filtrate with ¹⁵NO₂⁻ (10 μmol L⁻¹) + ¹⁴NH₂OH (1 μmol L⁻¹); (D) viable cells with ¹⁵NH₂OH (1 μmol L⁻¹) + ¹⁴NO₂⁻ (50 μmol L⁻¹); (E) filtrate with ¹⁵NH₂OH (1 μmol L⁻¹) + ¹⁴NO₂⁻ (50 μmol L⁻¹). Error bars represent SD from triplicate samples.

Much higher labeled N₂O production rates occurred in incubations supplemented with ¹⁵NH₂OH (Fig. 5 D and E) than in incubations supplemented with either ¹⁵NO₂⁻ alone or ¹⁵NO₂⁻ + ¹⁴NH₂OH, despite the concentration of ¹⁵NO₂⁻ (10 μmol L⁻¹) being tenfold higher than ¹⁵NH₂OH (1 μmol L⁻¹), demonstrating active involvement of NH₂OH in N₂O production. In the presence of viable cells, both ⁴⁵N₂O and ⁴⁶N₂O were produced (Fig. 5D), while for the filtrate, only ⁴⁶N₂O production was observed (⁴⁵N₂O accounted for <5% of the total labeled N₂O production) (Fig. 5E). The comparable ⁴⁶N₂O production between viable cell and filtrate groups indicated that ⁴⁶N₂O was mainly produced via abiotic NH₂OH oxidation, and NO₂⁻ is not involved in that reaction. In contrast, viable cells are needed for the production of ⁴⁵N₂O from ¹⁵NH₂OH and ¹⁴NO₂⁻, suggesting the hybrid reaction may require enzymatic activity to produce NO from NO₂⁻. Therefore, it appears that the ambient NO₂⁻ can enter the periplasmic space of the cell for the production of NO that is most likely catalyzed by a putative periplasmic copper-containing nitrite reductase (10, 11). These results indicate the central role of NH₂OH as a precursor of N₂O (pathways 1, 2, 3, and 4 in Fig. 1B). Moreover, the co-production of ⁴⁵N₂O and ⁴⁶N₂O from ¹⁵NH₂OH shows that hybrid N₂O formation and oxidation of NH₂OH both contributed to N₂O production in viable cells. Note that the high affinity for NH₄⁺ and the typical dependence of N₂O production rate on NH₄⁺ concentration (Experiment 1) implicate enzymatic control of N₂O production, even though the last step in both pathways (hybrid and NH₂OH oxidation) is abiotic.

Taking all the results with ¹⁵N-labeled substrates together, these findings demonstrate that N derived from both NH₄⁺ and NO₂⁻ is involved in N₂O production by AOA, but that NH₄⁺ is the major N source. Under the initial NH₄⁺: NO₂⁻ ratio of 1:1, NH₄⁺ accounts for ~85% of N atoms to N₂O. Paired analysis of ⁴⁵N₂O: ⁴⁶N₂O further implied that an additional N₂O formation pathway solely sourced from NH₄⁺ was required to explain the dominance of NH₄⁺ as the source of N atoms in N₂O. Thus, apart from the hybrid formation pathway, NH₂OH oxidation might be an important and previously overlooked pathway that contributes to archaeal N₂O production. However, it is difficult to discriminate all the associated pathways and to quantify the contributions of each process using N isotopes alone, because all N atoms in N₂O precursors could be directly or indirectly sourced from NH₄⁺.

Tracing the O in N₂O: Contributions of H₂O, O₂, and NO₂⁻ as the Source of O in N₂O.

¹⁸O leaves footprints in N₂O and NO₂⁻ that are independent of ¹⁵N and can thus provide further information on the pathways to N₂O. We developed a comprehensive set of ¹⁸O-labeling experiments to determine the source of O atoms in both NO₂⁻ and N₂O, and to quantify potential N₂O production pathways from H₂O, NO₂⁻, and O₂ (Experiment 5). The isotopic enrichment of δ¹⁸O-NO₂⁻ showed an approximately linear increase over time in the 24 h abiotic O atom exchange experiment, i.e., from ¹⁸O-H₂O in the absence of viable cells (Fig. 6A). The measured values of δ¹⁸O-NO₂⁻ agreed well with the amount of δ¹⁸O-NO₂⁻ predicted using an exchange rate constant of 0.117 (33) under the experimental conditions (pH: 7.8; temperature: 30°C).

Fig. 6.

Experiment 5. δ¹⁸O of the produced NO₂⁻ and N₂O during the 24 h ¹⁸O-labeling incubations. (A) Change of δ¹⁸O-NO₂⁻ due to abiotic O atom exchange. (B and C) δ¹⁸O of the produced NO₂⁻ in ¹⁸O-H₂O and ¹⁸O-O₂ labeling experiments. (D–F) δ¹⁸O of the produced N₂O in ¹⁸O-H₂O, ¹⁸O-O₂, and ¹⁸O-NO₂⁻ labeling experiments, respectively. Dashed lines denote the best linear regression, and gray shadow represents 95% CIs. Error bars represent SD from triplicate samples.

This correction for abiotic O atom exchange was applied to determine the δ¹⁸O of the produced NO₂⁻ in incubations with viable cells. O atoms from both H₂O and O₂ were incorporated into NO₂⁻ during archaeal ammonia oxidation (and the amount of δ¹⁸O-NO₂⁻ was proportional to the amount of labeled substrate in both ¹⁸O-H₂O and ¹⁸O-O₂ labeling incubations) (Fig. 6 B and C). The slope of δ¹⁸O-NO₂⁻ vs. δ¹⁸O-H₂O (63 ± 3%) was significantly greater than the slope of δ¹⁸O-NO₂⁻ vs. δ¹⁸O-O₂ (26 ± 2%) (P < 0.001). The significantly higher contribution of H₂O than O₂ to the O atoms in NO₂⁻ is consistent with the hypothesis that NH₂OH and NO act as co-substrates to produce two molecules of NO₂⁻. Then one NO₂⁻ molecule is reduced back to NO and another O atom from H₂O is incorporated into NO₂⁻ (13). Alternatively, an intracellular O atom exchange could occur during NO₂⁻ production by AOA (SI Appendix, Text 2).

δ¹⁸O of the produced N₂O increased with increasing δ¹⁸O-H₂O, δ¹⁸O-O₂ and δ¹⁸O-NO₂⁻, indicating that O atoms from all three potential sources were incorporated into N₂O (Fig. 6 D–F) with different contributions. Interestingly, in contrast to the O atom source structure in NO₂⁻, O₂ contributed the largest fraction of O atoms to N₂O (44 ± 2%), followed by NO₂⁻ (30 ± 3%) and H₂O (14 ± 1%). Because the O atoms in NH₂OH are sourced from O₂ and no further exchange occurs between NH₂OH and H₂O (34), the fact that O₂ (via NH₂OH) contributed most to O atoms in N₂O supports our finding of a substantial role for NH₂OH oxidation in producing N₂O (pathway 1 in Fig. 1B). The incorporation of O atoms from NO₂⁻ and H₂O into N₂O indicated internally produced and externally added (ambient) NO₂⁻ can both be involved in N₂O production. In the absence of known nitric oxide reductase catalyzing NO reduction to N₂O through nitrifier-denitrification, potential N₂O pathways associated with NO₂⁻ in marine AOA include abiotic NO₂⁻ reduction (chemo-denitrification-like) and hybrid formation. However, our ¹⁵NO₂⁻ labeling incubations showed that NO₂⁻ was involved in N₂O production only in the presence of NH₄⁺, and NO₂⁻ alone did not contribute substantially to N₂O production (Fig. 4 C and D). Therefore, hybrid formation is the dominant pathway by which NO₂⁻ contributes to N₂O production. Moreover, the incorporation of O atoms from H₂O and ambient NO₂⁻ further revealed that during hybrid formation, O atoms in NO₂⁻ or NO, rather than NH₂OH, were retained in the N₂O molecule. If the O atom was sourced from NH₂OH during the hybrid process, all the O atoms should be contributed by O₂. The incorporation of the O atom from ambient NO₂⁻ into N₂O also shows that at least some NO was produced via NO₂⁻ reduction, as previously hypothesized (11, 14).

Quantifying Multiple N₂O Sources during Archaeal Ammonia Oxidation.

The dual-isotope labeling method enabled us to fully resolve multiple N₂O production pathways in AOA. Combining results from ¹⁵N and ¹⁸O-labeling incubations (Experiments 3 to 5), we can quantitatively estimate the fractional contribution of the five potential N₂O production pathways (Fig. 1B). NO₂⁻ reduction (pathways 4 and 5) is an insignificant N₂O source under aerobic growth conditions when both NH₄⁺ and NO₂⁻ are present in equimolar amounts (Fig. 4). NH₂OH is revealed as the main contributor to N₂O production (Fig. 5) via both hybrid pathway and NH₂OH oxidation. In NH₂OH oxidation (pathway 1), 100% of O atoms in N₂O were sourced from O₂, while in hybrid formation (pathways 2 and 3), the O atom was derived from NO₂⁻ via reduction to NO. The NH₂OH involved in hybrid N₂O formation contributed an N atom but not the O atom. There were two sources of NO₂⁻: the original ambient NO₂⁻ (100% of O atoms were sourced from ambient NO₂⁻) and the NO₂⁻ newly produced from ammonia oxidation (63% of O atoms from H₂O and 26% from O₂) (Fig. 6). Under these conditions, we calculated the following contributions to O atoms in N₂O: 38.2% from NH₂OH oxidation (pathway 1), 59.8% from the hybrid source (22.2% by pathway 2 and 37.6% by pathway 3), and 2.1% from NO₂⁻ reduction (pathways 4 and 5) (Fig. 1C). This combination best fits the observed results that 14% of O atoms in N₂O were from H₂O and 44% of O atoms were from O₂ under our experimental conditions. Although the fractional contribution of the pathways might vary with various substrate concentrations (i.e., different NH₄⁺: NO₂⁻ ratios), the comprehensive ¹⁵N-¹⁸O dual-isotope labeling technique developed here provides a novel avenue to disentangle and quantify the relative contribution of multiple pathways to N₂O production in both lab and field studies.

Archaeal ammonia oxidation is the primary source of marine N₂O, yet the mechanistic understanding of archaeal N₂O production remains elusive. We present the first study determining the kinetics of archaeal N₂O production and provide strong evidence of the capability of marine AOA in producing N₂O at trace levels of NH₄⁺, supporting their dominant role in contributing to N₂O production in the ocean. We further show a direct control of NH₄⁺ and NO₂⁻ concentrations on the sources of N₂O, providing new insights into understanding the varying isotope composition of N₂O in the ocean. The increased incorporation of N and O atoms from NO₂⁻ into N₂O by marine AOA at low NH₄⁺: NO₂⁻ ratios suggests a new mechanism for interpreting the ubiquitous N₂O isotope minimum without the need to invoke nitrifier-denitrification by AOB, which are rarely detected in the oligotrophic open ocean. Our comprehensive dual ¹⁵N-¹⁸O-labeling techniques identify a substantial contribution of NH₂OH oxidation to archaeal N₂O production that was previously not recognized. These explicit descriptions of the N₂O production pathways and kinetics in AOA should improve our understanding of marine N₂O production, and the multiple N and O atom sources of N₂O identified here should inform biogeochemical models that aim to resolve the marine nitrogen cycle and constrain the air-sea N₂O flux. Moreover, our dual-isotope labeling technique could be applied in combination with manipulative experiments, such as temperature, pH, and dissolved oxygen (DO), to explore rates and pathways of archaeal N₂O production in response to ocean warming, acidification and deoxygenation.

Materials and Methods

SCM1 Cultivation and Isotope Labeling Incubation.

N. maritimus strain SCM1 was cultured in 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffered Synthetic Crenarchaeota Medium (SCM) (pH: ~7.8) at 30 °C in the dark following Qin et al. (2014) (35). Six sets of incubations were carried out (SI Appendix, Table S1). The SCM1 cells were grown and maintained in 2-L bottles containing 600 mL SCM for the preservative test and experiments 3 to 5 and were grown in 100-mL bottles containing 40 mL SCM for experiments 1 to 2. The medium contains FeNaEDTA and other trace metals including, Cu, Ni, Zn, and Co. The approximate concentrations of the trace metals can be found in Amin et al. (2013) (36). The medium was supplemented with NH₄Cl at different initial concentrations for each of the experiments: Initial concentrations were 1,000 μmol L⁻¹ in the preservative test experiment, 500 μmol L⁻¹ in experiments 3 and 5, 200 μmol L⁻¹in experiment 1 and 2, and 100 μmol L⁻¹ in experiments 4. Growth was monitored both by measuring NO₂⁻ concentration and NH₄⁺ consumption and by performing cell counts using flow cytometry (Accuri C6, BD Biosciences). After determining the best strategy for terminating the incubation to preserve the concentration and isotopic content of analytes (Preservative Test), the five experimental incubations were carried out to 1) test the kinetic response of N₂O production during SCM1 ammonia oxidation; 2) examine the impact of NH₄⁺: NO₂⁻ substrate ratio on the pathways and composition of N atoms in N₂O; 3) track N atom sources through ¹⁵N labeling experiments; 4) test the contribution of ambient NH₂OH to N₂O production; and 5) track O atom sources through multiple ¹⁸O-labeling experiments. All labeling incubation experiments were performed using mid- to late-exponential phase cultures.

Preservative test.

A total of ~1.0 L of culture was collected and aliquoted into two groups in the mid-exponential phase: 1) viable cells and 2) killed control (autoclaved at 120°C for 30 min and cooled overnight). In each group, each ¹⁵N tracer (¹⁵NH₄⁺ and ¹⁵NO₂⁻, 99% ¹⁵N, Cambridge, United States) was added to separate bottles to a concentration of 50 μmol L⁻¹. For the viable cells, one additional treatment (100 μmol L⁻¹ of ¹⁵NO₃⁻, 98% ¹⁵N, Cambridge, United States) was performed. After tracer addition, 10 mL of sample was dispensed into triplicate 20-mL serum bottles and sealed with 20-mm butyl stoppers and aluminum crimp seals (Wheaton, United States). Two preservatives (20 μL of saturated HgCl₂ and 500 μL of 10 mol L⁻¹ of NaOH) were used to compare the effect of preservatives on terminating biological activity and archaeal N₂O production. For the viable cells, the incubations were performed at 30°C in the dark and terminated at 0 and 24 h. For the autoclaved samples, the preservatives were added only at 0 h. All incubations were performed in triplicate. Our results showed that HgCl₂ induces artifacts of N₂O production from pathways involving NO₂⁻. Such artifacts are negligible when using NaOH as a preservative (SI Appendix, Fig. S3 and Text 3). Thus, NaOH was chosen as the preservative for all further experiments.

Experiment 1: Kinetic test.

When all amended NH₄⁺ (200 μmol L⁻¹) was completely consumed (i.e., below the substrate threshold of SCM1, ~10 nmol L⁻¹ NH₄⁺), 1% inoculum was transferred into NH₄⁺ free fresh medium supplemented with labeled tracers. The initial cell abundance was around 1.39 × 10⁵ cells ml⁻¹, which is comparable to AOA cell abundance in the ocean, and initial carry over ¹⁴NO₂⁻ concentration was ~2 μmol L⁻¹. A total of eight ¹⁵NH₄⁺ concentrations (0.1, 0.5, 1, 1.5, 2, 5, 8, 10 μmol L⁻¹) were used for the kinetic test. Immediately after tracer amendment, 50 mL aliquots of sample were dispensed into 60-mL serum bottles and sealed with 20-mm butyl stoppers and aluminum crimp seals (Wheaton, United States). The incubation was performed at 30°C in the dark. Depending on the initial ¹⁵NH₄⁺ concentration, the incubation was terminated at 0 h, ~2 h (0.1 μmol L⁻¹), ~6 h (0.5 μmol L⁻¹), and ~12 h (>0.5 μmol L⁻¹) by adding 2.5 mL of 10 mol L⁻¹ of NaOH. A third time point (~24 h) was also applied for all treatments. However, the third time point was only used for rate calculation in the high NH₄⁺ treatments (>1.5 μmol L⁻¹) because the substrate was nearly completely consumed or exhausted before 24 h in those low substrate treatments. The incubations were carried out in triplicates at each time point.

Experiment 2: Substrate ratio experiment.

When 100 μmol L⁻¹ of NH₄⁺ was consumed, 1% inoculum was transferred into NH₄⁺ free medium. The initial cell abundance was around 6.2 × 10⁴ cells mL⁻¹. Three NH₄⁺: NO₂⁻ ratios were achieved by adding different amounts of ¹⁵NH₄⁺ and ¹⁴NO₂⁻ (10 and 1 μmol L⁻¹ of NH₄⁺ and NO₂⁻, 5 and 5 μmol L⁻¹ of NH₄⁺ and NO₂⁻, and 1.5 and 15 μmol L⁻¹ of NH₄⁺ and NO₂⁻, respectively). Immediately after tracer amendment, 40 mL aliquots of sample were dispensed into 60-mL serum bottles and sealed with 20-mm butyl stoppers and aluminum crimp seals (Wheaton, United States). The incubation was performed at 30°C in the dark and terminated by adding 2 mL of 10 mol L⁻¹ of NaOH at 0, 6 and 24 h with triplicates at each time point.

Experiment 3: Source of N atoms in N₂O.

A total of ~4 L of culture was harvested by gentle filtration onto two 0.2 μm pore size Sterivex filters (Millipore). Immediately after the filtration, the filters were flushed using 2-L fresh substrate-free medium to collect the cells and to remove the high background NH₄⁺ (~210 μmol L⁻¹) and NO₂⁻ (~330 μmol L⁻¹). The cell densities before (4 L original culture) and after the filtration (resuspended in 2 L medium) were 1.15 × 10⁷ and 2.10 × 10⁷ cells mL⁻¹, respectively, demonstrating a good recovery efficiency (~60%) of the pre-concentration process. The measured ammonia oxidation rate of the washed cells (~16 μmol N L⁻¹ d⁻¹) was lower than the rate of unwashed cells measured on the same day (~84 μmol N L⁻¹ d⁻¹). This reduced oxidation rate indicates that the manipulation process caused physiological stress on the SCM1 cells and resulted in decreased cellular activity. Nevertheless, the activity of the washed cells was high enough to allow precise measurement of rates of ammonia oxidation and N₂O production in the experiments. The recovered cells were aliquoted into eight acid washed 250-mL PC bottles (Nalgene) and for four groups of tracers (¹⁵NH₄⁺, ¹⁵NH₄⁺ + ¹⁴NO₂⁻, ¹⁵NO₂⁻, ¹⁵NO₂⁻ + ¹⁴NH₄⁺). Immediately after tracer amendment, 10 mL aliquots of sample were dispensed into 20-mL serum bottles and sealed with 20-mm butyl stopper and aluminum crimp seals (Wheaton, United States). The incubation was performed at 30°C in the dark and terminated by adding 500 μL of 10 mol L⁻¹ of NaOH at 0 and 24 h with triplicates at each time point.

Experiment 4: Role of NH₂OH in N₂O production.

Around 1 L of culture was aliquoted into five groups when 50 μmol L⁻¹ of NH₄⁺ had been oxidized: 1) viable cells amended with ¹⁵NO₂⁻ (10 μmol L⁻¹); 2) viable cells amended with ¹⁵NO₂⁻ (10 μmol L⁻¹) + ¹⁴NH₂OH (1 μmol L⁻¹); 3) filtrate (through 0.2 μm PES filter) amended with ¹⁵NO₂⁻ (10 μmol L⁻¹) + ¹⁴NH₂OH (1 μmol L⁻¹); 4) viable cells amended with ¹⁵NH₂OH (1 μmol L⁻¹); and 5) filtrate (through 0.2 μm PES filter) amended with ¹⁵NH₂OH (1 μmol L⁻¹). After tracer amendment, 10 mL aliquots of sample were dispensed into 20-mL serum bottles and sealed with 20-mm butyl stoppers and aluminum crimp seals (Wheaton, United States). Time-course incubation (0, 1, 3, 6, 12 h) was carried out for all the groups, and the incubation was performed at 30°C in the dark with triplicates and terminated by adding 500 μL of 10 mol L⁻¹ of NaOH at each time point.

Experiment 5: Source of O atoms in N₂O.

A total of ~7.2 L of culture was harvested by gentle filtration onto two 0.2-μm pore size Sterivex filters (Millipore). Immediately after the filtration, the filters were back flushed using 2.5 L fresh substrate-free medium to collect the cells and to remove the high background NH₄⁺ and NO₂⁻. Three groups of tracers (H₂¹⁸O, ¹⁸O₂ and N¹⁸O₂⁻) were used to track the source of the O atom. For the ¹⁸O-H₂O labeling experiment, a range of δ¹⁸O-H₂O tracer amendments (−13 to 1003‰) were made by adding 0.2 mL of ¹⁸O-H₂O stocks with different ¹⁸O enrichment into the samples (¹⁸O-H₂O stocks were made by mixing the H₂¹⁸O (99% ¹⁸O, Sigma-Aldrich, United States) with distilled deionized H₂O). Similarly, six levels of δ¹⁸O-O₂ (24 to 714‰) were made by adding 0.2 mL of ¹⁸O-O₂ stocks with different ¹⁸O enrichment (¹⁸O-O₂ stocks were made by mixing the ¹⁸O₂ (99% ¹⁸O, Sigma-Aldrich, United States) with He) into the samples. For the ¹⁸O-NO₂⁻ labeling experiment, five levels of ¹⁸O-NO₂⁻ (4 to 539‰) were made by using the O atom exchange between NO₂⁻ and H₂¹⁸O. In each treatment, 20 mL of sample was dispensed into 60 -mL serum bottles and sealed with 20-mm butyl stoppers and aluminum crimp seals (Wheaton, United States). The ¹⁸O-labeled substrates were injected into the serum bottles. Both the NH₄⁺ and NO₂⁻ were set at 20 μmol L⁻¹ in each incubation. For ¹⁸O₂ labeled incubations, after ¹⁸O₂ injection, the bottles were shaken at ~120 rpm for 15 min to equilibrate the ¹⁸O₂ with the dissolved oxygen (DO) in water. The incubation was performed at 30°C in the dark and terminated by adding 1 mL of 10 mol L⁻¹ NaOH. A time-course (0, 6, 12, 24 h) incubation was performed in selected tracer treatments (δ¹⁸O-H₂O of 129‰; δ¹⁸O-O₂ of 110‰; δ¹⁸O-NO₂⁻ of 276‰), and the remaining treatments were terminated at 0 and 24 h; all experiments were performed in triplicates at each time point. An additional set of experiments was performed to examine the rate of abiotic O atom exchange between NO₂⁻ and H₂O. Briefly, NH₄⁺ and NO₂⁻ were added into ~10 mL of fresh medium to a concentration of 20 μmol L⁻¹, and ~0.1 mL of ¹⁸O-H₂O stock was added to get δ¹⁸O-H₂O of ~76‰. The incubation was performed at 30°C in the dark and terminated by adding 500 μL of 10 mol L⁻¹ of NaOH at 0, 6, 12, and 24 h with duplicates at each time point.

Sample Analysis.

The samples for NH₄⁺ and NO₂⁻ concentration measurement were stored at −20°C until analysis. The concentration of NH₄⁺ and NO₂⁻ was measured by colorimetric methods with an AA3 nutrient analyzer or a spectrophotometer. The detection limit for NH₄⁺ and NO₂⁻ was 0.5 and 0.03 μmol L⁻¹, and the analytical precision was better than ±3% and ±1%, respectively (37).

The N₂O samples were stored at 4°C after incubation. For the preservative test and experiments 3 and 4, before measurements, 1.5 nmol of N₂O of known isotope composition (δ¹⁵N = −3.2 ± 0.1‰ relative to air N₂, δ¹⁸O = 36.6 ± 0.1‰ relative to Vienna Standard Mean Ocean Water) was introduced into each serum bottle to provide enough mass for isotopic analysis. For experiments 1, 2, and 5, the samples were measured directly without N₂O carrier addition. Concentration and isotopes of N₂O were measured using a modified Gas Chromatograph-Isotope Ratio Mass Spectrometer (GC-IRMS) (38). Briefly, two needles were used for He pressurization and N₂O purging. For the 20-mL bottles, sample was purged for 6.7 min at a flow rate of 40 mL min⁻¹, and for 60-mL bottles, the purge time was 30 min. The extracted gases were passed through an ethanol trap with dry ice and a chemical trap filled with magnesium perchlorate and Ascarite to remove H₂O and CO₂. N₂O was trapped by liquid nitrogen twice for purification and concentration and then injected into the GC-IRMS with He as carrier gas. N₂O mass was determined by ion peak area [m/z of 44, 45, 46] with standard gases of 199.6, 501.0, and 1,000.2 ppmv N₂O/He, which were run at ten sample intervals. The precision of this method for N₂O mass measurement was estimated to be better than ±3%. δ¹⁵N and δ¹⁸O were calibrated against two reference tanks (R1: 199.6 ppmv N₂O/He, δ¹⁵N = −3.2 ± 0.1‰, δ¹⁸O = 36.6 ± 0.1‰; R2: 501.0 ppmv N₂O/He, δ¹⁵N = −1.6 ± 0.1‰, δ¹⁸O = 36.6 ± 0.3‰). The precision of δ¹⁵N and δ¹⁸O measurements with 2 nmol N₂O reference gas was better than 0.3‰ and 0.4‰, respectively (n = 20) (30). All the samples were measured within 2 wk after the incubations.

After N₂O measurement, the samples were stored at 4°C before further analysis. δ¹⁵N and δ¹⁸O of NO₂⁻ were determined using the bacterial denitrifier method (39, 40) using a Thermo Finnigan Gasbench system with cryogenic extraction and purification system interfaced to a Delta V^PLUS isotopic ratio mass spectrometer. Briefly, ~5 to10 nmol of NO₂⁻ was quantitatively converted to N₂O using the bacterial strain Pseudomonas aureofaciens. The produced N₂O was then introduced to the GC-IRMS through an online N₂O cryogenic extraction and purification system. δ¹⁵N of NO₂⁻ values were calibrated against NO₃⁻ isotope standards USGS 34, IAEA N3, and USGS 32; δ¹⁸O of NO₂⁻ values were calibrated against NO₃⁻ isotope standards USGS 34, IAEA N3 and USGS 35. The standards were run before, after, and at ten sample intervals. Because of the different branching effect during NO₃⁻ and NO₂⁻ reduction by P. aureofaciens (i.e., 38‰ vs. 12‰), the δ¹⁸O of NO₂⁻ was further calibrated by taking account of the branching effect between NO₃⁻ and NO₂⁻ (26‰) (41). Accuracy (pooled SD) was better than ±0.2‰ for δ¹⁵N and ±0.4‰ for δ¹⁸O according to analyses of these standards with an injection of a similar amount of NO₃⁻. Quality control was also conducted by analyzing laboratory working reference material (3,000 m deep sea water from the South China Sea).

δ¹⁸O of H₂O was measured following McIlvin and Casciotti (2006) (42) using the full exchange of O atom between H₂O and NO₂⁻ under acidic conditions (pH: 6) at room temperature (~25°C) for 2 wk. δ¹⁸O of NO₂⁻ was measured as described above, and δ¹⁸O of H₂O was calculated based on the isotope effect of 13‰ between the O atom exchange at room temperature (33).

δ¹⁸O of O₂ was not measured directly. The δ¹⁸O of the ¹⁸O₂ tracer was calculated from the mixing ratio of air (assuming δ¹⁸O air O₂ is 24‰) and ¹⁸O₂ tracer. Briefly, during our incubation, the DO concentration in the medium was near equilibration with air (~244 μmol L⁻¹); thus, a total of ~348 μmol of O₂ was present in the bottle (20 mL of medium and 40 mL of air in the headspace). During the ¹⁸O-O₂ labeling incubation, 0 to 12 μL of ¹⁸O₂ gas was introduced into the headspace and was then fully equilibrated with the water to attain different enrichments of ¹⁸O in the incubation, and the δ¹⁸O was then calculated from the ¹⁸O/¹⁶O after tracer addition.

Calculations.

Rate of labeled N₂O production from the ¹⁵N-labeled substrate was calculated based on the accumulation of ⁴⁵N₂O (single labeled) and ⁴⁶N₂O (double labeled) during the incubation. Total labeled N₂O production rate was defined as the ¹⁵N from both ⁴⁵N₂O and ⁴⁶N₂O (Eq. 1). ¹⁵NH₄⁺ oxidation rate was calculated from the increase of ¹⁵NO₂⁻.

$${\hspace{0.17em}}^{15}N-N_{2}O={\hspace{0.17em}}^{45}{N}_{2}O+2\times {\hspace{0.17em}}^{46}{N}_{2}O,$$ [1]

where the total ¹⁵N-N₂O includes ¹⁵N atom from ⁴⁵N₂O (one ¹⁵N atom) and ⁴⁶N₂O (two ¹⁵N atoms in each molecule).

δ¹⁸O of the produced N₂O during the incubation was calculated using a two-endmember mixing model (Eq. 2) (43).

$${\mathrm{\delta }}^{18}O-N_2{O}_P=\frac{M_{t24}\times {\mathrm{\delta }}^{18}{O}_{N_2{O}_{t24}}-M_{t0}\times {\mathrm{\delta }}^{18}{O}_{N_2{O}_{t0}}}{M_{t24}-M_{t0}},$$ [2]

where δ¹⁸O-N₂O_P, δ¹⁸O-N₂O_t24, and δ¹⁸O-N₂O_t0 denote the δ¹⁸O-N₂O value of the net produced N₂O during the ¹⁸O-labeling incubation, δ¹⁸O-N₂O at the end and beginning of incubation, respectively. M_t24 and M_t0 denote the measured N₂O mass at the end and beginning of incubation, respectively.

δ¹⁸O of the produced NO₂⁻ during the ¹⁸O-H₂O and ¹⁸O-O₂ incubations was calculated using Eq. 2 after calibrating the abiotic O atom exchange between NO₂⁻ and H₂O (9, 33). Briefly, the abiotic O atom exchange rate was derived from the time-course experiment using cell-free medium, which was then used to calibrate the contribution of abiotic O atom exchange during our incubation (Eq. 3). The difference between the measured ¹⁸O-NO₂⁻ and the predicted ¹⁸O-NO₂⁻ by abiotic exchange was then used to calculate the δ¹⁸O of newly produced NO₂⁻ during the incubation. The slopes of the newly produced δ¹⁸O-NO₂⁻ and δ¹⁸O-N₂O against δ¹⁸O of different substrates (H₂O, O₂, NO₂⁻) were identified as the fraction contribution of O atom from various substrates to the NO₂⁻ and N₂O (9, 33).

$${\mathrm{\delta }}^{18}{O}_{N{O}_{2abio}^{-}}=({\mathrm{\delta }}^{18}{O}_{N{O}_{2t0}^{-}}-{\mathrm{\delta }}^{18}{O}_{N{O}_{2eq}^{-}})\times \exp (-k\times t)+{\mathrm{\delta }}^{18}{O}_{N{O}_{2eq}^{-}},$$ [3]

where δ¹⁸O-NO₂⁻_abio, δ¹⁸O-NO₂⁻_t0, and δ¹⁸O-NO₂⁻_eq are δ¹⁸O value of NO₂⁻ at the end, beginning, and the equilibrated NO₂⁻ with H₂O due to abiotic O exchange. t is the incubation length in hours and k is the rate constant.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All data needed to evaluate the conclusions in the paper are deposited in Zenodo database that can be accessed through (https://doi.org/10.5281/zenodo.7378577) (44).