Climate change is driven by increased emissions of CO2, CH4, and N2O. Unlike CO2, both production and removal of CH4 and N2O are largely driven by microbial processes [1, 2]. CH4 is produced in highly reduced anoxic environments by methanogenic microorganisms [3]. N2O is produced in more oxidized environments, primarily as a byproduct of nitrification or denitrification [2]. Most CH4 produced in subsurface anoxic environments is oxidized at the oxic–anoxic interface by the aerobic methanotrophs [4]. The sole biological sink of N2O is via nitrous oxide reductase (NosZ) that converts N2O to N2, and significant correlations have been observed between N2O emissions and the abundance and/or diversity of nosZ [2, 5, 6].

Aerobic methanotrophy is highly dependent on copper, with expression and activity of methane monooxygenases controlled by copper availability [7, 8]. Many of these microbes synthesize and utilize a copper binding compound termed methanobactin for copper acquisition [9]. Methanobactin binds copper ions with 1:1 stoichiometry and has an unprecedentedly high affinity for copper ions, with binding constants ranging from 1018 to 1058 M−1 (values vary depending on measurement technique used [10]). Interestingly, biological removal of N2O is also strongly dependent on copper as NosZ has a high copper requirement [11]. Under copper stress induced with synthetic chelators, denitrification led to transient N2O accumulation [12,13,14]. Given the importance of copper availability for both methanotrophy and N2O reduction, we hypothesized that methanotrophs may be able to exert a “monopoly” over copper through production of methanobactin, leading to increased rates and amounts of N2O production due to inactivation of NosZ.

N2O production was first monitored upon incubations of the denitrifier Pseudomonas stutzeri DCP-Ps1 with NO3− as the terminal electron acceptor alone, in a coculture with Methylosinus trichosporium OB3b wildtype, or in a coculture with a mutant defective in methanobactin synthesis (ΔmbnAN [15]). In the axenic cultures of P. stutzeri, transient N2O production was observed (a maximum of 76 ± 8 nmol N2O-N) but N2O was eventually consumed (Fig. 1a). In the coculture with M. trichosporium OB3b wildtype, however, 430 ± 30 µmol N2O-N permanently accumulated (Fig. 1b). In the presence of the ΔmbnAN strain, the N2O production profile was indistinguishable from the axenic cultivation (Fig. 1c). Incubation of P. stutzeri with 1 μM purified methanobactin from M. trichosporium OB3b resulted in permanent production of N2O (85.4% of the added NO3−; Fig. 1d), confirming that N2O accumulation was due to the presence of methanobactin. When methanobactin was pre-loaded with the stoichiometric concentration of copper (1:1 molar ratio) and then added to cultures of P. stutzeri, no substantial accumulation of N2O was observed, indicating that NosZ activity was not affected (Fig. 1e).

Fig. 1
figure 1

N2O production from denitrification of P. stutzeri DCP-Ps1 in a axenic cultures or b in cocultures with either M. trichosporium OB3b wildtype or c M. trichosporium OB3b ΔmbnAN mutant. The axenic cultures of P. stutzeri DCP-Ps1 with d 1 μM purified methanobactin and e 1 μM purified methanobactin pre-incubated with 1 μM Cu2+ were also examined. The initial measurement (t=0) was made immediately after inoculation of P. stutzeri DCP-Ps1. For each experiment, the amounts of NO3− (□), NO2− (○), and N2O (●) in the culture vessel were monitored. The data points are the averages of triplicate experiments, with the error bars representing the standard deviations. The inserts are magnifications of N2O measurements

Full size image

Copper uptake by P. stutzeri DCP-Ps1 in the absence and the presence of 1 μM methanobactin was then assayed by examining the partitioning of copper after incubation with 5 mM NO3− (Supplementary Table S2). In the absence of methanobactin, 0.14 ± 0.03, 0.28 ± 0.06, and 0.38 ± 0.04 µg of copper were recovered from the cell biomass, the wash solution, and the spent medium, respectively, out of 0.77 ± 0.10 µg (0.24 ± 0.03 μM) initially added to the medium. In the presence of methanobactin, the partitioning of copper changed significantly (p < 0.01), with 0.02 ± 0.01, 0.06 ± 0.00, and 0.61 ± 0.02 µg recovered from the cell biomass, the wash solution, and the spent medium, respectively, out of 0.67 ± 0.06 µg (0.21 ± 0.02 μM) initially added to the medium. The total mass of cellular copper was 0.69 ± 0.14 μg Cu∙mg protein−1 upon incubation in the absence of methanobactin, while this value decreased significantly to 0.10 ± 0.03 μg Cu∙mg protein−1 in the presence of 1 μM methanobactin (p < 0.01). The presence of methanobactin inhibited adsorption of copper onto cellular surfaces, as well, suggesting a lower affinity of methanobactin–Cu complex to the outer membrane than free Cu2+ or Cu-EDTA. Such inhibition of copper uptake reduced nosZ transcription. In the absence of methanobactin, nosZ was expressed at 4.42 ± 1.13 nosZ transcript/recA transcript, while in the presence of methanobactin, nosZ transcription decreased significantly (p < 0.05) to 1.59 ± 0.55 nosZ transcript/recA transcript (a ∆∆Ct value of 1.669). Furthermore, the addition of methanobactin pre-loaded with copper (such that it could not bind any additional copper) had no significant effect on the expression of nosZ by P. stutzeri DCP-Ps1 (Supplementary Table S3). Thus, the effect of methanobactin on nosZ expression was in the reduction of bioavailable copper.

The effect of methanobactin was also examined with three other denitrifers: Dechloromonas aromatica RCB, a denitrifier with a clade II NosZ structurally and kinetically distinct from the clade I NosZ in P. stutzeri DCP-Ps1 [16, 17]; Paracoccus denitrificans ATCC17741, previously found to synthesize copropophryin III suggested to be used as a copper chelator [18]; and Shewanella loihica PV-4, a denitrifier utilizing NirK (copper-dependent) for NO2− reduction to NO [19]. In the absence of methanobactin, no significant accumulation of N2O was observed in the three denitrifier cultures (Fig. 2a, b, c). When these denitrifiers were cultured with 1 μM methanobactin, near stoichiometric accumulations of N2O were observed from denitrification (Fig. 2d, e, f). NirK-mediated NO2− reduction was also partially inhibited in S. loihica (Fig. 2f) ; nevertheless, eventual permanent production of 43.8 ± 2.3 µmol N2O-N was observed before NO2− reduction came to a complete stop with 50.6 ± 3.1 μmol NO2− remaining. This suggests that NirK is less sensitive to copper deprivation than NosZ, possibly due to its lower copper requirement [20].

Fig. 2
figure 2

N2O production from denitrification in axenic cultures of Dechloromonas aromatica RCB (a, d), Paracoccus denitirificans ATCC17741 (b, e), and Shewanella loihica PV-4 (c, f) without (a–c) and with (d-f) 1 μM methanobactin. For each experiment, the amounts of NO3− (□), NO2− (○), and N2O (●) in the culture vessel were monitored until the reaction was completed. The data points are the averages of triplicate experiments, with the error bars representing the standard deviations

Full size image

The threshold concentration of bioavailable copper for unhindered NosZ activity has been estimated to be on the order of 10−15–10−24 M [12, 21]. Previously, such low copper concentrations could only be attained in water stripped of metals with help of artificially synthesized ligands, e.g., triethylenetriamine and tetrathiolmolybdate [12, 21]. Thus, enhanced release of N2O due to the inhibition of N2O reduction caused by the lack of bioavailable copper was previously deemed environmentally unlikely [14]. Contradicting these previous findings, we show here that a biogenic copper chelator produced by methanotrophs blocks N2O reduction, eliciting N2O emissions from incomplete denitrification. These findings have significant relevance for greenhouse gas emissions in situ as co-occurrence of methanotrophy and denitrification is not rare. As the main substrates for aerobic methanotrophs are CH4 and O2, which diffuse towards the oxic–anoxic interface from opposite ends of the oxygen gradient, the largest abundance of methanotrophs is often found at the oxic–anoxic interfaces in subsurface environments [22]. The oxic–anoxic interface is also often a “hotspot” for soil denitrification in agricultural soils, as the major source of the electron acceptors for denitrifiers is NO3−/NO2− leaching from the oxic upper layers of soil where nitrifiers oxidize NH4+-N [23]. In such ecological niches where methanotrophs and denitrifiers coexist, these findings suggest that hitherto unrecognized microbial competition for copper may be a significant mechanism controlling N2O emissions.

An open question is how much methanobactin may be produced in situ. Levels of methanobactin have yet to be reported in environmental samples, but genomic analyses, as well as physiological and biochemical investigations, indicate that methanobactin synthesis is a widely distributed phenotype among methanotrophs at environmentally relevant copper concentrations [10]. Given that methanotrophs constitute a substantial portion of the active microbial community in soils, it seems likely that denitrifier activity can be affected in situ by production of methanobactin. Methanotroph–denitrifier interactions must be more fully explored to better understand how best to predict, and possibly minimize, net greenhouse gas emissions.