Copper control of bacterial nitrous oxide emission and its impact on vitamin B₁₂-dependent metabolism

Significance

Global atmospheric loading of nitrous oxide (N₂O) is on the increase. This stable, long-lived greenhouse gas is a major contributor to radiative forcing by Earth’s atmosphere. Here we describe the genetic regulation of N₂O reductase nosZ, encoding the only known N₂O-removing enzyme that limits the release of this denitrification intermediate during the bacterial usage of nitrogenous fertilizers. Expression of nosZ is down-regulated in copper-limited environments, leading to net emission of N₂O. This cytotoxic N₂O emission subsequently modulates expression of genes controlled by vitamin B₁₂ riboswitches, because N₂O binds to and inactivates vitamin B₁₂. Cytotoxicity of N₂O can be relieved by the addition of vitamin B₁₂. This interaction provides a role for NosZ in N₂O-detoxification in nondenitrifying bacteria.

Abstract

Global agricultural emissions of the greenhouse gas nitrous oxide (N₂O) have increased by around 20% over the last 100 y, but regulation of these emissions and their impact on bacterial cellular metabolism are poorly understood. Denitrifying bacteria convert nitrate in soils to inert di-nitrogen gas (N₂) via N₂O and the biochemistry of this process has been studied extensively in Paracoccus denitrificans. Here we demonstrate that expression of the gene encoding the nitrous oxide reductase (NosZ), which converts N₂O to N₂, is regulated in response to the extracellular copper concentration. We show that elevated levels of N₂O released as a consequence of decreased cellular NosZ activity lead to the bacterium switching from vitamin B₁₂-dependent to vitamin B₁₂-independent biosynthetic pathways, through the transcriptional modulation of genes controlled by vitamin B₁₂ riboswitches. This inhibitory effect of N₂O can be rescued by addition of exogenous vitamin B₁₂.

Global atmospheric loading of the ozone-depleting greenhouse gas, nitrous oxide (N₂O), is on the increase (1). Molecule for molecule, its radiative potential is ∼300-fold higher than carbon dioxide (2, 3), comprising ∼9% of global radiative forcing by greenhouse gases (4). In addition, atmospheric N₂O is stable for ∼120 y. Approximately 70% of anthropogenic N₂O loading arises from agriculture, mainly from the use of nitrogen-containing fertilizers by soil microbes for dissimilatory purposes. Taken together, these features make N₂O an important target for mitigation strategies (5).

N₂O is an intermediate in the sequential reduction of nitrate (NO₃⁻) to di-nitrogen (N₂), via nitrite (NO₂⁻), nitric oxide (NO), and N₂O, a process known as denitrification (6). Under certain conditions, the final step in denitrification is dispensed with and N₂O is released into the atmosphere. One limiting factor in this process is copper (Cu) availability, the metal cofactor required by the N₂O reductase (NosZ) that destroys N₂O (5, 7, 8). During Cu-limitation the catalytic capacity of the Nos system may be exceeded by the rate of the preceding reactions that generate N₂O (i.e., NO₃⁻, NO₂⁻, and NO reduction) and thus, N₂O is emitted by denitrifying bacteria (7, 9, 10).

Much attention has been given to the cytotoxic properties of NO as a free-radical and oxidant, but N₂O is often described as a relatively inert intermediate of the nitrogen cycle. However, N₂O exhibits cytotoxicity, as it is known to bind to and inactivate vitamin B₁₂ (B₁₂), an essential cellular cofactor in B₁₂-dependent enzymes involved in methionine and DNA synthesis (11, 12). B₁₂ also acts as a ligand for B₁₂ riboswitches that modulate gene expression in the absence of this cofactor (13, 14). The possible impact of environmental N₂O emissions on B₁₂ metabolism in microbiological communities has largely been ignored. As levels of N₂O increase in the environment, there is a compelling argument for better understanding the impact of this gaseous intermediate on the regulation of cellular metabolism in denitrifying bacteria, particularly in Cu-limited culture conditions that are associated with N₂O emissions (7, 9). Here we link these two important issues in bacterial N₂O research and show that the nos genes (for N₂O reduction) are strongly regulated by Cu and that the accumulation of only micromole amounts of N₂O in Cu-deficient cultures leads to up-regulation of B₁₂ independent anabolism, a response indicative of destruction of the B₁₂ pool by N₂O. This cytotoxicity of N₂O is relieved by the addition of exogenous B₁₂.

Results and Discussion

Impact of Cu-Limitation on Paracoccus denitrificans Transcription Under Denitrifying Conditions.

Paracoccus denitrificans was grown under anaerobic batch culture conditions with NO₃⁻ as electron acceptor in medium containing 13 µmol/L (Cu-H) and 0.5 µmol/L (Cu-L) copper. Under both culture conditions NO₃⁻ was consumed in a growth-linked fashion, decreasing from ∼8 mmol to 0 mmol as the culture density increased (Fig. 1). The growth rate and final yield was identical in both Cu-regimes and NO₂⁻ was observed at a maximum of 1 mmol and not detectable once cultures had reached stationary phase. The key difference between the two was a transient accumulation of N₂O in the Cu-L culture that was not observed in the Cu-H culture. This accumulation reached a maximum of around 2 mmol N⋅N₂O (Fig. 1). These results suggest that in the Cu-L cultures the catalytic capacity of the Cu-dependent Nos system is transiently exceeded by the rate of the reactions that generate nitrous oxide (i.e., NO₃⁻, NO₂⁻, and NO reduction) and is consistent with other observations that Cu limitation can lead to nitrous oxide release by denitrifying bacteria (7, 9, 10).

Fig. 1.

Growth characteristics of P. denitrificans PD1222, nosC⁻ PD2301, nosR⁻ PD2302, and nosZ⁻ PD2303 strains in anaerobic batch culture conditions in Cu-H media (♦) or Cu-L media (■). (Left) Optical density (OD_600nm), (Center) N⋅NO₃⁻ (millimole N in the form of NO₃⁻), and (Right) N⋅N₂O (millimole N in the form of N₂O) for each respective strain. Bars represent SE between triplicates, and where not visible, these were smaller than the symbols.

P. denitrificans has four distinct multidomain metalloproteins for the reduction of NO₃⁻, NO₂⁻, NO, and N₂O (Fig. 2), which are encoded by narGHJI (for nitrate reductase), nirSECFDGHJN (for nitrite reductase), norCBQDEF (for nitric oxide reductase) and nosCRZDFYLX (for nitrous oxide reductase) loci, respectively. NosZ is the catalytic subunit of the N₂O reductase and binds 12 Cu ions per functional homodimer in the copper-sulfide redox centers Cu_A and Cu_Z (8). To examine if Cu limitation has an impact on the denitrification genes at the level of transcription, global gene-expression analyses were performed using total RNA isolated from anaerobically grown cells under either Cu-H or Cu-L conditions. Cultures were compared at midexponential growth, where in Cu-L media, N₂O levels had exceeded 100 µmoles per flask (Fig. 1). Under these conditions, only 41 genes were differentially expressed (Fig. 3) and the expression levels of a selection of these were verified by quantitative RT-PCR (qRT-PCR) (Fig. 4 and Tables S1–S3). There was no significant change in any of the nar, nir, or nor genes encoding the enzymes for NO₃⁻ reduction to N₂O, all of which were highly expressed under both culture conditions (Fig. S1). In contrast, Cu-limitation had a major effect on the nosRZDFYLX genes, required for the functional N₂O reductase system, the expression of which decreased by 8- to 25-fold in the Cu-limited cultures (Fig. 3 and Table S3). This result demonstrates that the nos genes are subject to regulation by Cu.

Fig. 2.

Reactions of denitrification in which NO₃⁻ is sequentially reduced to N₂. Above each arrow is the metallo-enzyme complex and below is the metal cofactor required for each reaction in P. denitrificans. N₂O reduction is carried out by Nos (reaction 4), the only denitrification enzyme dependent on Cu in this model denitrifier.

Fig. 3.

Heat map representing the expression level of genes of P. denitrificans PD1222, grown anaerobically with NO₃⁻ and either 13 µmol/L (Cu-H) or 0.5 µmol/L (Cu-L) Cu in the media. (A) Genes regulated by B₁₂ riboswiches that are modulated by N₂O. (B) The Cu-responsive genes for N₂O reduction and Cu-metabolism. Colors indicate average log₂ normalized expression values between three biological replicates. Black arrows indicate gene orientation. Bold text denotes those genes ratified using qRT-PCR. ID represents unique locus tag of each gene in the P. denitrificans genome (Accession: NC_008686–008688).

Fig. 4.

Expression of selected genes from P. denitrificans, as monitored by qRT-PCR. (A) The effects of either Cu-limitation (Cu-L/Cu-H) or N₂O (Cu-H+N₂O/Cu-H) exposure on the expression of a selection of genes identified from transcriptomic studies. (B) Expression of nosZ in wild-type P. denitrificans PD1222, nosC⁻ PD2301, or the nosR⁻ PD2302-deficient strains during anaerobic growth and Cu-limitation. (C) The expression of the three-gene cluster involved in Cu-metabolism, during Cu-limitation in aerobic and anaerobic conditions. Bars show SE of three biological replicates.

NosC and NosR Are both Required for Cu-Mediated Regulation of nosRZDFYLX.

Interestingly and in contrast to the adjacent nosRZDFYLX, the proximal gene in the nos cluster, nosC was up-regulated in response to Cu-limitation. The gene’s product, NosC, is a hypothetical protein with unknown function and close (>50% identical) homologs appear to be only distributed among other Paracoccus species. Notably, all known homologs of NosC contain a CXXCXXC motif that may bind a redox active cofactor, the significance of which is unknown. NosR is a transmembrane iron-sulfur cluster containing flavoprotein required for reduction of N₂O that also contains two putative metal binding CXXXCP motifs (15–17), noted for their ability to bind Cu in some proteins, as discussed below. NosDFYLX are all thought to be important for NosZ assembly and activation, but a clear function for these polypeptides is yet to be demonstrated (16). Similarly, previous studies have not formally identified a function for the NosR protein (15–17) and its involvement in the Cu-dependency of the Nos system has never been reported. Nevertheless, nosR occurs adjacent to nosZ in the vast majority of denitrifier genomes, underpinning its involvement in N₂O reduction. Given that nosC was differentially expressed in Cu-limited conditions compared with the rest of the nos genes, and NosR contains the necessary residues for binding Cu, we reasoned that these proteins might be involved the Cu-dependency of N₂O reduction.

To test this hypothesis, we deleted nosC and nosR independently and examined the phenotypes of these strains during anaerobic growth with nitrate as the electron acceptor. In the nosR⁻ mutant (PD2302), we noted that N₂O reduction was vastly reduced, but not completely abolished in both Cu-H and Cu-L conditions, because 50–70% (∼4.5–7 mmol) of the NO₃⁻ consumed was detected as N₂O (Fig. 1). In the nosC⁻ mutant (PD2301), we noted that N₂O reduction was altered in Cu-H media, because up to 2 mmols of N₂O were detected (Fig. 1). As a control, we generated a nosZ-deficient strain (PD2303) and saw that ∼90–100% of NO₃⁻ was detectable as N₂O following anaerobic growth using NO₃⁻. To further investigate the role of the nosC and nosR genes, we measured the transcription of nosZ under both Cu regimes in both mutant strains. Strikingly, we found that nosZ transcript levels in both the nosC⁻ strain (PD2301) and the nosR⁻ strain (PD2302) were similar in Cu-H and Cu-L conditions (Fig. 4), providing strong evidence that repression of nosZ during Cu-limitation had become deregulated. This result identifies a role for both NosC and NosR in the Cu-regulation of nosZ, although the mechanisms by which these proteins mediate this Cu-response remains to be established and warrants further investigation.

Cu-Responsive senC2 and pcuC Are Involved in Cu-Metabolism in P. denitrificans.

The remaining genes highlighted by our transcriptomic studies were all up-regulated during Cu limitation (Fig. 3). Of these genes, we found a three-gene cluster that plays a pivotal role in anaerobic Cu-metabolism, encoded by Pden_4445, 4444 (here we term pcuC), and 4443 (here we term senC2). Pden_4445 encodes a hypothetical protein with unknown function. The product of pcuC is predicted to be secreted to the periplasm by the Sec system and is homologous to the bacterial Cu(I) protein PCu_AC (periplasmic Cu_A chaperone), including the conserved Cu(I)-binding H(M)X₁₀MX₂₁HXM motif. In other bacterial species, PCu_AC directs the insertion of Cu(I) into the Cu_A site of cytochrome ba₃ oxidase (18). The product of senC2 is a member of the bacterial homologs of Sco1, involved in the assembly of the Cu_A center of cytochrome oxidases in mitochondria. The sensor of C_ox (SenC)/Sco1 family, including SenC2, have a CXXXCP Cu-binding motif that provides for cysteinyl coordination of Cu, completed by an axial histidinyl ligand (19–21).

Deletions in each of these genes significantly decreased N₂O reduction in Cu-L cultures, reduced the final biomass yield, and in the case of the pcuC⁻ strain PD2305, severely delayed the transition into anaerobic metabolism using NO₃⁻ (Fig. 5). Interestingly, mutations in Pden_4445 and senC2 had no effect in Cu-H conditions, suggesting that these mutants can be complemented by micromolar Cu levels. However, the pcuC⁻ strain PD2305 accumulated up to 1 mmol N⋅N₂O in Cu-H media, indicating that even in Cu-H media the catalytic activity of NosZ is compromised by the loss of pcuC. Given the observed phenotypes of the respective mutants (PD2304 – 2306), we predict that the proteins encoded by Pden_4445, pcuC, and senC2 are involved in either insertion or maintenance of the Cu-centers of NosZ. Furthermore, these proteins may have a role in aerobic Cu metabolism, given that they are also up-regulated in aerobic Cu-limited conditions (Fig. 4). However, the functions and activities of the putative Cu-binding proteins PcuC and SenC2 and the role of the product of Pden_4445 are yet to be fully established.

Fig. 5.

Growth characteristics of Pden_4445⁻ PD2304, pcuC⁻ PD2305, and senC2⁻ PD2306 strains in anaerobic batch culture conditions in Cu-H media (♦) or Cu-L media (■). (Left) Optical density (OD_600nm), (Center) N⋅NO₃⁻ (millimole N in the form of NO₃⁻), and (Right) N⋅N₂O (millimole N in the form of N₂O) for each respective strain. Bars represent SE between triplicates, and where not visible, these were smaller than the symbols.

N₂O Modulates Expression of Vitamin B₁₂-Independent Metabolic Pathways.

Upon interrogation of the remaining up-regulated genes, we observed that many have an association with B₁₂-dependent metabolism and may be under the control of B₁₂ riboswitches, known for their ability to modulate transcription of downstream genes dependent on the presence of the B₁₂ ligand (13, 14, 22–24). Nine putative gene clusters downstream of B₁₂ riboswitches had elevated expression in Cu-limited conditions (Fig. 3, Fig. S2, and Table S3). These clusters included genes for the B₁₂-independent versions of methionine synthase (metEF) and both the class Ib and class III ribonucleotide reductases, nrdHIEF and nrdDG respectively. Other genes encoded products associated with B₁₂-scavenging apparatus, including a cobalt transporter (cbtAB), several iron/B₁₂ ABC-type transporter systems (e.g., btuBCDF), and several hypothetical proteins of unknown function.

Because Cu-limited cultures accumulate significant amounts of N₂O, a molecule that directly interacts with B₁₂ or B₁₂-dependent enzymes (11, 12, 25, 26), we hypothesized that the up-regulation of these riboswitch-controlled genes was being modulated not by Cu, but by the consequential N₂O emitted during Cu-limitation. To test this theory, we cultured P. denitrificans under anaerobic conditions in Cu-H media, in the presence and absence of excess N₂O, and monitored the response of the nos and other genes by qRT-PCR. Indeed, all of the gene clusters downstream of B₁₂ riboswitches were up-regulated upon exposure to N₂O. Conversely, expression of the nos genes and the senC2/pcuC gene clusters did not change (Fig. 4 and Table S3). This finding suggested that up-regulation of B₁₂-linked gene clusters is linked to the presence of N₂O, consistent with this gaseous intermediate accumulating during exponential growth in the Cu-limited media. Such an interaction between B₁₂ and N₂O would diminish the active B₁₂ pool and lead to the synthesis of B₁₂-independent enzymes to counteract this loss of cofactor.

Cytotoxicity of N₂O Exposure Is Alleviated by the Addition of Vitamin B₁₂.

To confirm the phenomenon that N₂O exhibited a cytotoxic effect mediated by the destruction of B₁₂, the atmospheres of sealed flasks containing Cu-H media was replaced with N₂O, such that each vessel had media and atmosphere completely saturated with N₂O. These vessels were then inoculated with P. denitrificans and incubated for 30 h at 30 °C. During this initial incubation cultures grew poorly in an N₂O atmosphere. After 30 h additions of B₁₂, l-methionine (l-met), or unsupplemented media was made to the vessels, and growth was monitored for a further 30 h. Notably, following addition of either B₁₂ or l-met, the growth of P. denitrificans rapidly entered exponential phase and reached high cell densities, but the culture without either of these compounds was unable to grow (Fig. 6). This restoration of growth confirms that the cytotoxicity of N₂O is inextricably linked to the status of l-met/B₁₂ cellular pools, likely mediated through the inactivation of the B₁₂-requiring methionine synthase MetH by this potent greenhouse gas, as previously reported (11, 12).

Fig. 6.

Cytotoxicity of N₂O can be alleviated by the addition of vitamin B₁₂ or l-met. (A) Growth of P. denitrificans in anaerobic conditions with NO₃⁻ in Cu-H conditions with N₂O-saturated atmospheres. After 30 h (black arrow) additions of either vitamin B₁₂ (▲), l-met (■), or unsupplemented media (♦) was made to the cultures, and growth monitored for a further 30 h. B and C compare the growth of P. denitrificans wild-type (PD1222; ♦), metE⁻ (PD2307; ▲), or metE⁻ nosZ⁻ (PD2308; ■) cultures, grown either in Cu-H conditions (B) or the N₂O-genic Cu-L conditions (C). Bars represent SE between triplicates, and where not visible, these were smaller than the symbols.

Vitamin B₁₂-Independent Methionine Synthase MetE Counteracts the Toxicity of N₂O.

Given that the B₁₂-independent methionine synthase MetE was up-regulated in the presence of N₂O, we postulated that this enzyme compensates for the loss of functional MetH, which is inactivated by N₂O. In other bacteria MetE is regulated by a B₁₂ riboswitch (27). To examine the role of MetE, we generated a metE⁻ strain (PD2307) and assessed the impact of anaerobic growth in N₂O-genic, Cu-L media on the survival of this mutant. The metE⁻-deficient strain was hampered by growth in Cu-L media, but not in Cu-H media (Fig. 6 and Fig. S3), suggesting that in conditions in which N₂O accumulated, growth was significantly perturbed. To further demonstrate this phenomenon, we analyzed the growth of a nosZ⁻ metE⁻ double-mutant (PD2308), which would accumulate N₂O during anaerobic metabolism, irrespective of Cu levels. Growth of PD2308 was impaired compared with the wild-type, regardless of Cu-content, with an extended lag phase and lower final biomass. This finding further demonstrates the influence of N₂O on aspects of vitamin B₁₂-linked metabolism (Fig. 6 and Fig. S4).

In support of our findings, B₁₂-deficiency and hyperhomocysteinemia have been previously reported in humans as a consequence of N₂O-based anesthesia (28). Under these circumstances both patients and clinicians are exposed to N₂O. B₁₂ is essential for a variety of mammalian enzymes, which in humans is obtained primarily through diet and from the gastrointestinal microflora. Hyperhomocysteinemia and methylmalonic acidemia arise from the malfunction of B₁₂-dependent enzymes required for methionine synthesis and fatty acid catabolism (28, 29), and can be treated with dietary supplements of B₁₂.

Concluding Remarks

Bacterial anaerobic denitrification is well characterized at the structural and biochemical level. Expression of the NO₃⁻, NO₂⁻, NO, and N₂O reductases are subject to regulation by gas-sensitive regulatory proteins (30), which sense O₂, NO, or the anion NO₃⁻ (31–33), and regulate a variety of genes involved in the transition to anaerobic respiration (34). For nearly three decades it has been known that Cu concentrations can drastically affect the activity of N₂O reductase and the subsequent yield of N₂O from denitrifying cells (7, 9, 10). Emerging from this study are two additional features in the biology of this potent green-house gas.

Here, we describe evidence for transcriptional modulation of nos expression by Cu-status, mediated through the NosC and NosR proteins. In addition, we reveal a remarkable interplay between N₂O emission and the expression of B₁₂ riboswitch-regulated gene clusters that has far-reaching implications. In P. denitrificans, N₂O modulates the expression of genes for anaerobic anabolism that are controlled by B₁₂ riboswitches. This result in turn impacts on growth under denitrifying conditions. In other bacteria, such as the human pathogens Enterococcus and Listeria, B₁₂ riboswitches control a variety of responses including pathogenesis and virulence (22, 23). Our data suggest that N₂O accumulation is toxic to cellular metabolism at extracellular concentrations as low as ∼0.1 mmol/L. Thus, NosZ may play a previously unrecognized role in N₂O resistance in microbial communities and this provides an insight into the function of NosZ in nondenitrifier bacterial genomes (35). The implications are that organisms that live in microbial communities and do not have B₁₂-independent mechanisms, or NosZ, could be compromised by the release of N₂O following incomplete denitrification by partial denitrifiers or during Cu-limitation. This result may give rise to a selection pressure in which nosZ is conserved for reasons other than for the generation of proton motive force during denitrification.

Materials and Methods

For detailed methods, refer to the SI Materials and Methods. Batch cultures of P. denitrificans were grown anaerobically in minimal media supplemented with 30 mmol/L succinate, 10 mmol/L NH₄Cl, and 20 mmol/L NO₃⁻, containing either 13 µmol/L (Cu-H) or <0.5 µmol/L Cu (Cu-L) in the form of CuSO₄, determined by inductively coupled plasma optical emission spectrometry. N-oxyanions were quantified as previously described (7) with minor changes. Mutant strains were generated using pK18mobsacB and selected using a modified Luria-Bertani recipe containing sucrose. For N₂O exposure, vessels containing Cu-H media were saturated by sparging with N₂ and subsequently with N₂O for 15 min. Additions of l-met, AdoCbl, or blank media were added anaerobically after 30 h of N₂O exposure. Total RNA was harvested from midexponential phase batch cultures. Minimum Information About a Microarray Experiment-compliant microarrays were carried out using RNA from three independent replicates using custom designed tiled 44 K microarray slides (Agilent), and transcripts were quantified by qRT-PCR as described previously (36), conforming to the Minimum Information of Quantitative Real-Time PCR Experiment guidelines.