Genomic organization, gene expression and activity profile of Marinobacter hydrocarbonoclasticus denitrification enzymes

Genome analysis

The genome sequence of M. hydrocarbonoclasticus SP17 (Grimaud et al., 2012), with the accession number NC_017067 (NCBI) was analyzed using bioinformatic tools to search for homologous genes encoding the catalytic subunits of the four enzymes that catalyze each step of the denitrification pathway, as well as accessory factors (see Supplementary Information S2). The virtual Footprint of PRODORIC database was used to detect the putative binding sites for FNR and IHF in the promoter regions of the denitrification genes (Table S3). As a remark, the genomic organization of denitrification genes was analyzed using the deposited genome of M. hydrocarbonoclasticus SP17, while the strain that was cultured was M. hydrocarbonoclasticus 617. In fact, nos gene cluster had been sequenced in our lab (GenBank: dQ504302.1) and showed to have an identical organization including dnr gene upstream nos gene cluster.

Marinobacter hydrocarbonoclasticus 617 growth conditions

M. hydrocarbonoclasticus 617 was grown in artificial seawater (ASW) liquid medium containing 1.17% (w/v) NaCl, 0.075 % (w/v) KCl, 0.3% (w/v) NH₄Cl, 1.23% (w/v) MgSO₄.7H₂O, 0.6% (w/v) Tris, 0.1% (w/v) yeast extract and 0.12% (w/w) sodium lactate as energy and carbon source (Baumann & Baumann, 1981). The pH was adjusted to 7.5 with 37% (w/w) HCl. The medium was supplemented with 0.33 mM K₂HPO₄, 10 mM CaCl₂.2H₂O, 7.2 µM FeSO₄.7H₂O and a Starkey oligoelement solution (Starkey, 1938) (1 mL/L of culture), with the composition listed in Supplementary Information S4.

Nucleic acid extraction and cDNA generation

Samples of bioreactor cultures taken at different time-points were immediately frozen in liquid nitrogen, until further use (1 mL). Total RNA, from the samples, was isolated using the Isolate II RNA mini kit (Bioline, Memphis, TN, USA), according to manufacturer’s instructions. Genomic DNA contamination was removed by on-column digestion with DNase I, supplied with the kit. RNA yields were determined at 260 nm and purity was estimated by determining the A_260 nm/A_280 nm and A_260 nm/A_230 nm ratios. To generate cDNA, 500 ng of each RNA sample (after its extraction) was reversely transcribed using the SensiFAST cDNA Synthesis Kit (Bioline), using the following conditions: 10 min at 25 °C, followed by 15 min at 42 °C and 5 min at 85 °C (for enzyme inactivation). cDNAs were diluted (1:100) and stored at −80 °C until further use. DNA used in standard curves was extracted from a sample of the bioreactor culture taken during the exponential phase, using the Isolate II Genomic DNA Kit (Bioline, Memphis, TN, USA), according to the manufacturer’s instructions and stored and −80 °C until further use.

Quantitative Real-Time PCR

The expression of M. hydrocarbonoclasticus genes involved in the denitrification pathway—narG, nirS, c-norB (MARHYR3054), q-norB (MARHY3014) and nosZ—was analyzed by quantitative real-time PCR (qPCR). Reactions were performed using SensiFAST™ SYBR No-ROX Kit (Bioline) with primers specific for each gene at 250 nM and 3 µL of cDNA. Gene expression levels were determined using the relative standard curve method, which uses standard curves to interpolate unknown quantity values from a sample. This was performed using DNA standard curves, as previously described (Gomes et al., 2005). Briefly, for each target, as well as for the reference (16S rRNA), a standard curve was generated using serial dilutions of genomic DNA. The amount of cDNA of each target and control gene, present in every sample, was determined from their respective DNA standard curve, through conversion of the obtained threshold cycle (C_t) value (Applied Biosystems, 2008; Gomes et al., 2005). Relative normalized expression values were obtained by dividing the values obtained for each target gene by the values obtained for the reference gene (16S rRNA), as its expression is stable, during the different growth phases. Primers were designed using Primer3 software (Untergasser et al., 2012) (Table 1). “No template” and “RT minus” (a Reverse Transcription reaction containing all reagents except the reverse transcriptase enzyme) controls were also included in every PCR assay. Reactions were run in a Corbett Rotor-Gene instrument (Qiagen, Hilden, Germany) using the following thermal cycling conditions: 95 °C for 5 min followed by 40 cycles of 95 °C for 15 s and 60 °C for 30 s. Analysis of melting curves generated by the stepwise increase of the temperature from 60 °C to 95 °C, were used to verify the specificity of the amplified products. Three biological replicates were analyzed but not at the exact same time points, thus a single representative experiment is shown (similar results were obtained for the other replicates). Two to three technical replicates were performed depending on the biological replicates.

Co-transcriptional analysis of nosRZDFYL

To determine if nosRZDFYL is composed of a single transcriptional unit, a sample collected during the exponential phase of M. hydrocarbonoclasticus growing in a bioreactor performed at pH 7.5 was used. The cDNA was generated similarly to those described above but using 2 µg of total RNA. PCR reactions were performed using 6 µL of the resulting cDNA and NZYTaq DNA polymerase (Nzytech) with primer pairs that would amplify regions between nosR and nosZ (nosR_F/nosZ_R), nos Z and nosD (nosZ_F/nosD_R) and nosD to nosL (nosD_F/nosL_R) (Table 1). Genomic DNA and a “RT minus” reactions were included as controls. The amplified PCR products were analyzed by 1% agarose gel electrophoresis, run at 100 V, during 20 min, and visualized after incubation in a SYBR Safe solution (Invitrogen) under UV light.

Nitrate/nitrite quantifications and nitric oxide and nitrous oxide reduction assays

Nitrate and nitrite quantifications along the growth curve were performed using the Nitrate/Nitrite Assay kit (Sigma), according to the manufacturer’s instructions, in triplicate for each time-point. The absorbance was measured at 540 nm in a VersaMax ELISA Microplate Reader using the SoftMax Pro 6.4 software.

Whole-cell activity assays for nitric oxide and nitrous oxide reduction were performed spectrophotometrically using a TIDAS diode array spectrophotometer, at room temperature inside an anaerobic chamber (MBraun), by following the oxidation of reduced methyl viologen (non-physiologic electron donor) at 600 nm, as previously described (Kristjansson & Hollocher, 1980) (data analysis is described in Supplementary Information S6). Nitric oxide (5% NO/95% He, Air liquid) and nitrous oxide (>95% N₂O, Air Liquid) solutions were freshly prepared before the assays. Briefly, 5 mL of Milli-Q water were degassed for 30 min with argon in sealed serum flasks (sealed with a rubber septum and aluminium caps) and then flushed with N₂O for 1 h. Taking into account the solubility of N₂O in water, at 25 °C and 1 atm, this solution is 25 mM in N₂O (Kristjansson & Hollocher, 1980). The NO solution was prepared similarly but a 10% (per mass) KOH solution was used between the NO bottle and the Milli-Q solution to remove other nitrogen oxides from NO gas. A concentration of 1.91 mM in NO is reported at 20° C for a 98.5% NO stock, thus this solution is 9.7 µM (Timoteo et al., 2011). The assays were performed with constant stirring by adding 40 µL cell suspension to a quartz cuvette already containing 120 µM methyl viologen and 60 µM sodium dithionite in 100 mM Tris–HCl, pH 7.6 (in a final volume of 1 mL). The nitric oxide reduction assay was initiated by the addition of NO-saturated water to a final concentration of 9.6 µM, while the nitrous oxide reduction assay was initiated (without any further incubation) by the addition of N₂O-saturated water to obtain a 1.25 mM concentration in the assay. The reduction rate (micromoles of NO or N₂O reduced per minute per optical density) was obtained by analyzing the slope of the reaction after nitric oxide or nitrous oxide addition. Controls of the reactions were performed in which nitric oxide or nitrous oxide were added to the assay in the absence of cell suspension. In these controls no oxidation of MV was observed.

The dependence of N₂O in the reduction rate by whole-cells was performed in a 1 mL quartz cuvette, containing 100 µM methyl viologen and 50 µM sodium dithionite in 100 mM Tris–HCl pH 7.6, to which 3 µL of cells was added (containing 0.09 mg of total protein), collected by centrifugation at the end of the growth performed at pH 7.5 in the bioreactor, under microoxic conditions. The assay was initiated by the addition of 20, 40, 80, 120, 250 and 752 µM of N₂O-saturated water. Triplicates for each substrate concentration were performed. Activities were calculated as described above but reported as µmol_N2O min⁻¹ mg⁻¹ of total protein (determined using the Pierce™ 660 nm protein assay with bovine serum albumin as standard). The parameters K_m and V_max were calculated by fitting the curve with the Michaelis–Menten equation.

Genomic organization of denitrification genes in M. hydrocarbonoclasticus

The genome of M. hydrocarbonoclasticus SP17 has several genes encoding proteins involved in the different steps of the denitrification pathway clustered together (see Fig. S3). The nosZ, encoding the catalytic domain of N₂OR, is included in the nosRZDFYL gene cluster. Upstream the nos cluster is the NaR gene cluster, narLXKGHJV (see Fig. S3), with narG encoding the catalytic subunit of NaR, and narXL, the two-component system, NarXL, involved in transcriptional regulation of these genes in response to nitrate and/or nitrite (Philippot, 2002; Spiro, 2012).

Several nitrate/nitrite transporters have been identified in the genome of M. hydrocarbonoclasticus, one being associated with the nar cluster, other two upstream the nir operon, and a narU-type gene is also annotated in this genome. Multiple NarK-like transporters has also been identified in P. aeruginosa and P. denitrificans (Jia et al., 2009; Moir & Wood, 2001), but their biological function is still a matter of debate, with different transporters being proposed to be required under specific growth conditions. The identification of the mechanism of action of these membrane proteins can only be accomplished through the study of the physiology of selected knock-out strains and site-directed mutagenesis.

Further upstream in the genome is norBC, encoding the two subunits of the short-chain membrane-bound c-NOR (see Fig. S3). Neither norE nor norF homologues were annotated in the genome of M. hydrocarbonoclasticus. However, the translated sequence of MARHY3057 shares 46% sequence identity with P. stutzeri NorE (designated in that microorganism as NirQ). A BLAST search was performed using the translated sequence of MARHY3058, showing 53% identity with P. stutzeri NirP (the NorF homologue in Pseudomonas species with 82 residues), though in a very restricted region, covering only 49% of the primary sequence. Indeed, this putative NorF (with 61 a.a.) is 40% shorter than its homologues. However, the analysis of the entire intergenic region between MARHY3057 and MARHY3058 identifies a longer ORF encoding a protein with 79 residues, that has 39% primary sequence identity with P. stutzeri NirP. Therefore, we propose that NorF is misannotated in M. hydrocarbonoclasticus SP17 genome, being encoded in the 3141724–3141963 genomic region rather than in the 3141778–3141963 region.

A second norB gene (MARHY3014) also encoding a NOR is annotated in M. hydrocarbonoclasticus genome, downstream the nos cluster and transcribed in the opposite direction (see Fig. S3). This NOR is composed of a single long-chain subunit being a q-NOR homologue (primary sequence identity with Geobacillus stearothermophilus q-NOR is 37%, with a 95% coverage, and 25% with P. aeruginosa large subunit NorB, but with only 60% coverage) (Table S2). While c-NOR uses a c-type cytochrome as electron donor, q-NOR accepts electrons from the quinone pool. The presence of two types of NOR in the same bacteria does not seem to be frequent (Heylen et al., 2007; Jones et al., 2008) and its biological relevance is unknown.

After the norBC genes, a large cluster of 10 genes involved in the nitrite reduction step of the denitrification pathway was found linked together, nirFCSDLGHJEN. Different genetic arrangements of nir genes have been identified in bacteria. P. aeruginosa presents the most similar genetic organization of the nitrite reductase gene cluster in comparison to the one of M. hydrocarbonoclasticus (Bali et al., 2014; Philippot, 2002). The only differences are in the nirM gene, encoding cytochrome c ₅₅₁, which is absent in M. hydrocarbonoclasticus and in nirFC, which are located upstream nirS and are transcribed in the opposite direction. NirM is the electron donor of NiR in P. aeruginosa, while the cytochrome c ₅₅₂, encoded by MARHY3556, was identified as the physiological electron donor of NiR in M. hydrocarbonoclasticus (Lopes et al., 2001) and is distantly located in the genome.

The denitrification pathway has been proposed to be regulated mainly by transcription regulators of the CRP/FNR family, FNR and DNR, that in M. hydrocarbonoclasticus are encoded by MARHY0862 (FNR homologue) and MARHY3023 (DNR homologue), with this later ORF being located upstream the nos gene cluster. FNR, that activates genes during anaerobiosis, is a homodimer containing a [4Fe-4S] center per protomer, which is sensitive to oxygen and to nitrosylation (Crack et al., 2014). On the other hand, DNR senses N-oxides, in particular NO (Arai, Kodama & Igarashi, 1999), through a heme (Castiglione et al., 2009). These two transcription regulators bind specifically to consensus sequences of the type TTGATN₄ATCAA (named FNR binding box). Thus, the promoter regions of the four gene clusters were analyzed to identify the presence of these binding boxes, up to 200 bp upstream their ATG site. Homologous consensus sequences were identified based on its score and distance to the putative transcription site (Fig. 2). For most of the gene clusters, the putative binding regions are located around −10 to −35 bp, except for dnr (−88 bp). These distances are in agreement with the average location of these sites identified in the whole genome of P. aeruginosa (Trunk et al., 2010) and also with the more common spacing locations observed for FNR in E. coli, which are usually centered at approximately −41, −61, −71, −82 and −92 bp from the transcription initiation point (Wing, Williams & Busby, 1995). Although these binding sites still need to be experimentally confirmed, the involvement of MARHY0862 (FNR homologue) and MARHY3023 (DNR homologue) in the regulation of the denitrification pathway will be discussed.

The other regulatory system of denitrification is NarXL that responds to nitrate and is proposed to be involved in the regulation of the nar genes (Schreiber et al., 2007). NarL, the transcription regulator, recognizes specific heptameric motifs (TACYYMT) with a 7 − 2 − 7 arrangement located in the promoter regions of nar gene cluster (Schreiber et al., 2007). The promoter region of narK and narGHJV was analyzed to identify the −10 and −35 boxes, which were only found upstream narK, which lead us to propose that narKGHJV is transcribed into a single polycistronic unit. Moreover, upstream narK a potential NarL binding site (around −21 bp, TACCCCA, with a score of 4.13 for NarL of E. coli) and a FNR/DNR binding site were identified but none on narGHJV (Fig. 2). Relative to narXL, no binding sites for FNR/DNR nor NarL were identified upstream the −10/−35 boxes, which can indicate that these genes are constitutive. Moreover, NarL is not involved in regulation of dnr as no consensus sequences were identified upstream of FNR box, as opposite to what was described for P. stutzeri (Härtig et al., 1999) (dnrE) and P. aeruginosa (dnr) (Schreiber et al., 2007) that carried the motifs and were shown to have a NarL-dependent transcription.

Analysis of gene expression during M. hydrocarbonoclasticus 617 growth in the presence of nitrate - microoxic batch growth

The expression of denitrification genes of M. hydrocarbonoclasticus 617 grown in the presence of nitrate, in the 2 L-bioreactor, was analyzed by quantitative real-time PCR. The results obtained for both norB genes (MARHY3054 and MARHY3014) indicates that norB_MARHY3054 (c-norB) gene is the one transcribing an active nitric oxide reductase under the denitrifying conditions used here, as no significant expression levels are observed for the norB_MARHY3014 (q-norB) gene in these conditions (Fig. 3). Thus, q-type NOR (encoded by norB _MARHY3014) does not contribute to the M. hydrocarbonoclasticus denitrification pathway, under these conditions, and its role in the physiology of this microorganism remains unknown.

Expression profiles show that the genes encoding the catalytic subunits of the denitrifying enzymes (narG, nirS, norB_MARHY3054 and nosZ) are expressed simultaneously after 1 h, and the maximum level of transcription occurs at time-point 5 h (end of the first diauxic phase), for these four genes (Fig. 3).

At 5 h of growth, narG expression level is slightly lower than the other genes (Fig. 3), reflecting the existence of additional regulatory mechanisms in nar expression or a lower strength of the fnr promoter in the nar gene cluster.

Transcriptional analysis of nos gene cluster

In M. hydrocarbonoclasticus nos cluster is composed of nosRZDFYL (see Fig. S2). The nosR and nosZ genes are separated by 75 bp, while 68 bp separate nosZ from nosD. The nosD, nosF and nosY genes overlap each other by 2 bp and finally, nosY and nosL are separated by 22 bp. This organization suggests that these genes might be transcribed into a single transcriptional unit. To investigate this hypothesis reverse transcription of total RNA, from a sample of the bioreactor, was performed and the intergenic regions of the corresponding cDNAs were amplified by PCR. Amplicons with the expected sizes were obtained from nosR and nosZ, nosZ and nosD, and nosDFYL intergenic regions (Fig. 4A and 4B), showing that nosRZDFYL is transcribed into a polycistronic unit, similarly to P. aeruginosa nos gene cluster (Arai, 2003).

Marinobacter hydrocarbonoclasticus growth under denitrifying conditions

M. hydrocarbonoclasticus is a ubiquitous marine bacterium that populates the subtropical area and the Mediterranean Sea. In the present work, this bacterium was grown under two different conditions, microoxic and anoxic, as a model for the metabolism occurring at two different sea depths. The growth in the bioreactor was performed under microoxic conditions as a model for the layers close to the surface, while the growth in the sealed serum flasks are a model for deeper sea layers.

Growth in a bioreactor under microoxic conditions

M. hydrocarbonoclasticus cultures were performed in a 2-L bioreactor under microoxic conditions (low aeration rate) and in the presence of nitrate (10 mM), to promote denitrification (Moura et al., 2017; Eady, Antonyuk & Hasnain, 2016; Zumft, 1997). The analysis of the growth curve shows the presence of a diauxic growth. The first growth phase occurs without any lag phase during the first 5 h with a growth rate of 0.173 ± 0.008 h⁻¹, slowing down (growth rate of 0.011 ± 0.001 h⁻¹) until approximately 35 h, when it reaches the stationary phase (Fig. 5A). The lag phase was not observed in any of the replicated assays, indicating that adaptation-period was not required, or it was very small; and the death phase was not reached even after 48 h of growth, suggesting that the carbon source, lactate, had not been completely consumed.

The oxygen level was monitored during the growth, showing a rapid decrease in the first 2 h, and remaining negligible until the end of the growth (see Fig. S2). This apparent absence of oxygen does not mean that there is no oxygen, but that the amount dissolved is being completely consumed: the low aeration rate combined with the low agitation at 50 rpm after 5 h of growth will lower the amount of dissolved oxygen in the growth media. This foresees that the microorganisms will grow in microenvironments experiencing anoxic or near anoxic conditions.

A diauxic growth was reported for P. denitrificans during anoxic growth in the presence of nitrate (60 mM) (Hahnke et al., 2014), with the first corresponding to the consumption of nitrate and the second with the start of nitrite consumption. However, in the present study the two exponential phases seem to correspond to the simultaneous use of oxygen and nitrate as terminal electron acceptors by M. hydrocarbonoclasticus, in the first phase, and in the sole use of the low amount of oxygen dissolved in the growth media in the second exponential phase, with the first having an expected energetic advantage.

Activity profile of denitrifying enzymes

Nitrate and nitrite were quantified during M. hydrocarbonoclasticus growth, to monitor the action of NaR and cd ₁NiR, respectively, which catalyze the first and second steps of the denitrification pathway (Fig. 1). In addition, the reduction rate of nitric oxide and nitrous oxide by whole-cells was also determined at different time-points during growth, to identify when c-NOR and N₂OR were active in the cells (Fig. 5).

The data obtained indicate that nitrate was being consumed in the second hour of the growth and was completely consumed after 3 h (first mid-exponential phase of the diauxic growth and the time-point at which the oxygen level is negligible, see Fig. S1). This is an indication that nitrate transport to the cytoplasm and NaR were active before time-point 2 h in the bioreactor. Indeed, at time-point 1 h, narG gene expression was already occurring at low levels (Fig. 3), and thus it is likely that NaR was already active in the cells. Moreover, the decrease in nitrate concentration was concomitant with the formation of nitrite, which reaches a maximum after 3 h of growth with a value close to the initial nitrate concentration (8.5 mM), and four hours later (during the initial stages of the second diauxic growth phase) nitrite was completely consumed (Fig. 5A). This decrease in nitrite concentration is attributed to the activity of cd ₁NiR. In fact, a correlation between nirS expression and nitrite consumption is observed (Fig. 3 and 5A). Moreover, similar nitrate and nitrite profiles were observed for the growth in the sealed serum flasks, indicating that both enzymes are also active under these conditions (Fig. 5B).

The last two steps of the denitrification pathway are catalyzed by c-NOR and N₂OR, and to determine if these enzymes are active in M. hydrocarbonoclasticus the reduction rate of NO and N₂O by the whole-cells was monitored (Fig. 5). The results obtained from the growth carried in the bioreactor indicate that both enzymes started to be active in the cells after 5 h of growth. Moreover, it was observed that these enzymes are active in the cells until the end of the growth (Fig. 5), even if NO and N₂O are expected to have been completely consumed earlier on (Bergaust et al., 2010).

On the other hand, the profile of the reduction rate of NO and N₂O by the whole-cells grown in the sealed serum flasks showed maximum NO reduction rate at 3 h, when high levels of nitrite are still present in the medium, while the N₂O reduction was only observed after approximately 6 h of growth, indicating that both enzymes are active under these conditions at pH 7.5 (Fig. 5B). Moreover, contrary to what was observed for the cells cultured in the bioreactor under microoxic conditions, the ability to reduce these metabolites decreases towards the end of the growth.

Kinetic parameters for N₂O reduction by whole-cells

The N₂O reduction by the whole-cells of M. hydrocarbonoclasticus as a function of substrate, N₂O, has a hyperbolic behavior (Fig. 4C). This data was fitted to the Michaelis–Menten equation to estimate the kinetic parameters, V_max of 2.2 ± 0.1 µmol_N2O min⁻¹ mg⁻¹ and K_m of 18 ± 5 µM. This K_m is of the same order of magnitude as the K_m of Clade I P. stutzeri strain DCP-Os1 N₂OR (36 ± 9 µM) determined by N₂O consumption measurements, which are lower than those reported for whole-cells containing Clade II N₂OR (1.0 ± 0.2 µM and 1.3 ± 0.4 µM, from Dechloromonas denitrificans strain ED-1 and Anaeromyxobacter dehalogenans strain 2CP-C, respectively) (Yoon et al., 2016).

Gene expression and regulation of denitrification in M. hydrocarbonoclasticus

In M. hydrocarbonoclasticus there is a 70 kb region that encodes all the genes involved in the denitrification pathway (see Fig. S3), with the promoter regions of these gene clusters presenting a FNR/DNR consensus sequences, while NarL binding sequences were only identified upstream the nar operon (Fig. 2).

The genome of this bacterium has genes that encode two different types of NORs, q-NOR and c-NOR, but under the denitrifying conditions used here only the later catalytic gene subunit is expressed at significant levels (Fig. 3), and in fact this was the enzyme isolated from M. hydrocarbonoclasticus membrane extracts (Timoteo et al., 2011).

In M. hydrocarbonoclasticus, the onset and maximum of nosZ, nirS, narG and norB transcription occurs simultaneously, which suggests that a common signal(s) triggers and regulates the expression of these genes. The triggering signal can be the oxygen, which levels are very low during the growth in the bioreactor (Fig. S1). This behavior is different from the one in P. denitrificans, in which it was observed that nosZ is expressed earlier than nirS and norB (Bergaust et al., 2010). Those authors interpreted this pattern of transcription by nosZ transcription being triggered by oxygen, while the onset of nirS and norB transcription was associated with the increase in nitric oxide concentration that occurs afterwards (Bergaust et al., 2010) (vide infra).

Relative to the regulation of M. hydrocarbonoclasticus nar genes, similarly to what was observed in P. stutzeri (Härtig et al., 1999) and P. aeruginosa (Schreiber et al., 2007), these might be under the regulation of narXL, a nitrate-responsive two-component system. This was supported by the identification of NarL DNA binding sites in the promoter region of nar gene cluster (Fig. 2), as mentioned.

The data obtained here does not univocally identified the signals involved in the regulation of M. hydrocarbonoclasticus denitrification genes. However, it is expected that the common effectors are oxygen and nitric oxide, as identified in other denitrifying bacteria (Spiro, 2012), through the action of FNR and DNR, respectively. In fact, given the pattern of transcription observed for the genes encoding the catalytic denitrifying subunit, it can be postulated that oxygen and nitric oxide are the main effectors in this organism. Upon decrease of oxygen tension in the cultures, FNR monomers with its [4Fe-4S] centers dimerize, becoming activate, and bind specific DNA regions, activating the transcription of several genes required for the anoxic metabolism, which include activation of dnr and narGHJ (Crack et al., 2013; Trunk et al., 2010). In turn, upon the initial activity of cd ₁NiR, originated from the initial low levels of nirS transcription, NO concentration will increase by the reduction of nitrite, and DNR will activate nir, nor and nos genes (Fig. 6). On the other hand, nar genes would be regulated by oxygen and nitrate that activates their expression through narXL.

The existence of a cascade type regulation for these genes will prevent the accumulation of intermediate toxic products. Although, FNR and DNR recognize the same promoter regions (as mentioned before), it has been shown that DNR is the specific regulator of denitrification (Ebert et al., 2017; Trunk et al., 2010). The mechanism for the discrimination of these binding sites is still unknown, though nitrosylation of FNR could play a role, as upon nitrosylation (which might occur when nitric oxide concentration increases), FNR reduces its binding affinity to the specific DNA regions (Crack et al., 2013). This double regulatory mechanism might be important to avoid increasing release of toxic nitric oxide by NiR, together with down-regulation of nir, nor and nos gene clusters, as with a second regulator, as DNR, denitrification can proceed by activating those gene clusters.

Metabolites and NO/N₂O reduction rates by whole-cells

Nitrate and nitrite concentration during the growth of M. hydrocarbonoclasticus was monitored, being observed a rapid consumption of nitrate with concomitant formation of nitrite, which is completely consumed after 2 h, in both types of growth conditions tested. Considering that the decrease in nitrite concentration is due to cd ₁NiR activity, as the denitrification pathway is active, the concomitant formation of nitric oxide is expected, followed by formation of N₂O. The activity of the two last enzymes of the denitrification pathway was proven by the observation of the NO and N₂O reduction rate by the whole-cells, which occurs until the end of the growth (Fig. 5).

There was a delay between maximum gene expression (nirS, norB and nosZ) and enzymatic activity or complete nitrite consumption: (i) nirS maximum expression occurs at 5 h and complete consumption of nitrite occurs only after time-point 8 h; and (ii) nosZ and norB maximum expression occurs at 5 h, while ability of the cells to reduce NO and N₂O is maximum at 7 h. This displacement can be attributed to the maturation process of these three enzymes, as c-NOR is a membrane-bound enzyme and, cd ₁NiR and N₂OR catalytic center assembly involves several accessory factors.

Comparison between the activity profiles obtained for the two types of growth showed that lower reduction rates, especial of N₂O, are obtained in the sealed serum flasks. Two major explanations can be hypothesized, either there is reduced transcription of nosZ genes under this condition or the amount of N₂OR produced is lower. In fact, the yield of N₂OR isolated from a batch bioreactor operated under microoxic conditions is 2.6 ± 0.4 mg/L medium, while from a similar growth under anoxic conditions is 0.5 ± 0.3 mg/L medium.

Nevertheless, since cells isolated from both growth conditions were able to reduce nitrous oxide, it is plausible to argue that M. hydrocarbonoclasticus can carry out the four steps of the denitrification pathway under these two growth conditions, and this organism has a denitrifying chain as the one schematically presented in Fig. 1.

N₂OR Clade and nitrous oxide affinity

M. hydrocarbonoclasticus has an affinity for nitrous oxide consistent with the presence of a Clade I N₂OR, which also agrees with the arrangement and gene constitution of nos operon in this organism, nosRZDFYL (Fig. 4B). Furthermore, the estimated K_m for N₂O reduction by the whole-cells (18 ± 5 µM) (Fig. 4C) is similar to the value reported for the N₂O reduction by the in vitro activated fully-reduced N₂OR (K_m of 14 ± 3 µM) (Dell’acqua et al., 2008).

Considering that 40 µL of growth under microoxic conditions has a reduction rate of 3.4 × 10⁻² µmol_N2Omin⁻¹ (before being collected), and 1.49 × 10⁻⁴ mg of N₂OR (considering a 70% recovery for the isolation of N₂OR from this growth), this gives an estimate of 228 µmol_N2O min⁻¹ mg_N2OR⁻¹ for the cells in vivo. This value is in fact similar to the specific activity of N₂OR in the activated fully reduced form (200 µmol_N2O min⁻¹ mg_N2OR⁻¹) (Johnston et al., 2014). In the case of the anoxic growth, an activity of the same order of magnitude as that one was estimated at the end of the growth (340 µmol_N2O min⁻¹ mg_N2OR⁻¹).