M. fumariolicum strain SolV used in this study was initially isolated from the volcanic region Campi Flegrei, near Naples, Italy (Pol et al., 2007). In this study the medium to obtain an OD₆₀₀ of 1.0 was composed of 0.2 mM MgCl₂.6H₂O; 0.2 mM CaCl₂.2H₂O; 1 mM Na₂SO₄; 2 mM K₂SO₄; 2 mM (NH₄)₂SO₄ (or 5 mM KNO₃) and 1 mM NaH₂PO₄.H₂O. A trace element solution containing 1 μM NiCl₂, CoCl₂, MoO₄Na₂, ZnSO₄ and CeCl₃; 5 μM MnCl₂ and FeSO₄; 10 μM CuSO₄ and 40–50 μM nitrilotriacetic acid (NTA). The pH of medium was adjusted to 2.7 using 1 M H₂SO₄ (1 ml H₂SO₄ per 1 L medium). To avoid precipitation, CaCl₂.2H₂O and the rest of medium were autoclaved separately and mixed after cooling. This medium composition was used in batch and continuous cultures, unless otherwise stated.

The continuous culture with CH₄ as an electron donor and nitrate (NO3-) as N-source (CH₄/NO3-), liquid volume 500 ml, was operated at 55°C with stirring at 900 rpm with a stirrer bar. The chemostat was supplied with medium at a flow rate of 14.5 ml.h⁻¹ (D = 0.026 h⁻¹), using a peristaltic pump. The cell-containing medium was removed automatically from the chemostat by a peristaltic pump when the liquid level reached the 500 ml level sensor in the reactor. A supply of 10% CH₄ (v/v), 8% O₂ (v/v), and 68% CO₂ (v/v) took place by mass flow controllers through a sterile filter and was sparged into the medium just above the stirrer bar (total gas flow rate ≈20 ml.min⁻¹). The initial pH was 3.4 and was regulated with 1 M carbonate connected to the vessel by a peristaltic pump. The pH was gradually increased to 6 and after obtaining a steady state, all experiments were performed at this pH. In the continuous culture with H₂ as an electron donor and NH4+ as N-source (H₂/NH4+), liquid volume was 1.2 L and this culture was operated at 55°C with stirring at 1,000 rpm. The chemostat was supplied with medium at a flow rate of 29.9 ml.h⁻¹ (D = 0.023 h⁻¹). A gas supply of 12% H₂ (v/v), 10% air (v/v), and 5% CO₂ (v/v) was provided by mass flow controllers through a sterile filter and sparged into the medium (total gas flow rate ≈16.5 ml.min⁻¹). The initial pH was 2.9 and the pH was regulated by 1 M NaOH. A pH range from 3 to 5.5 was investigated in the steady state. In the continuous culture with CH₄ as an electron donor and NH4+ as N-source (CH₄/NH4+), the liquid volume was 0.3 L and the culture was operated at 55°C with stirring at 700 rpm at pH 2.7. The chemostat was supplied with medium at a flow rate of 0.35 ml.h⁻¹ (D = 0.0012 h⁻¹). A gas supply of 0.16% CH₄ (v/v), 0.6% O₂ (v/v), and 5% CO₂ (v/v) was directed by mass flow controllers through a sterile filter and sparged into the medium (total gas flow rate ≈10 ml.min⁻¹). An O₂ sensor in the liquid was coupled to a Biocontroller (Applikon) regulating the O₂ mass controller in each reactor.

In order to obtain cells growing at maximum growth rate (μ_max), cells were grown without any limitation in 250-ml serum bottles containing 40 ml medium (4 mM NH4+; pH 2.7), and sealed with red butyl rubber stoppers. The headspace contained air with (v/v) 10% CH₄, 5% CO₂ at 55°C with shaking at 250 rpm. Incubations were performed in duplicate.

Nitric oxide and nitrous oxide (NO and N₂O) were analyzed on an Agilent series 6890 gas chromatograph (Agilent, USA) equipped with a Porapak Q and a Molecular sieve column, coupled to a thermal conductivity detector and a mass spectrometer (MS; Agilent 5975 Cinert MSD; Agilent, USA) as described before (Ettwig et al., 2008). For all gas analyses, 100 μl gas samples were injected into the gas chromatograph. Furthermore, nitric oxide production was monitored directly from the gas outlet of the reactors using a nitric oxide analyzer (NOA 280i, GE) with a suction rate of 11.6 ml.min⁻¹.

To determine the dry weight, samples of 8–10 ml from the culture suspension were filtered through pre-weighed 0.45 μm filters and dried to constant weight in a vacuum oven at 70°C (n = 3). In order to determine the total content of carbon and nitrogen, 10 ml of the culture suspension (duplicate) was centrifuged at 4,500 g for 30 min and the clear supernatant was used for the analysis. The nitrogen and carbon content in the supernatant was compared with the corresponding values in the whole cell suspension. The total carbon and nitrogen contents were measured using TOC-L and TNM-1 analyzers (Shimadzu).

To determine nitrite (NO2-) concentrations, 50 μl of sample, and 450 μl of MilliQ water were added to a cuvette. Then, 500 μl of reagent A [1% (w/v) sulfanilic acid in 1M HCl; kept in the dark] and 500 μl of reagent B [0.1% (w/v) naphtylethylene diaminedihydrochloride (NED) in water; kept at 4°C in the dark] were added to the same cuvette and mixed well. After incubation for 10 min at room temperature, the absorbance at 540 nm was measured and the values were compared with a calibration curve using known concentrations of nitrite in a range of 0–0.5 mM. If necessary, the sensitivity of this assay could be increased 10-fold using 500 μl samples without addition of water. NH4+ concentrations were measured using the ortho phthaldialdehyde (OPA) method (Taylor et al., 1974). In order to determine hydroxylamine concentrations, 200 μl reagent A (50 mM potassium phosphate buffer pH 7), 160 μl demineralized water, 200 μl sample, 40 μl reagent B [12% (w/v) trichloroacetic acid in water, kept in the dark], 200 μl reagent C (1% w/v 8-hydroxyquinoline (quinolinol) in 100% ethanol, kept in the dark) and 200 μl reagent D (1 M Na₂CO₃) were mixed and incubated at 100°C for 1 min. The absorption was measured at 705 nm and the values were compared to a calibration curve using hydroxylamine concentrations 0.02–0.1 mM.

To determine the affinity constant of pMMO for NH₃ of each sample, a volume of 5 ml of cells from the CH₄/NO3- continuous culture were washed and resuspended in the same medium at pH 6 (The pH of the medium was adjusted to 6 using MES buffer at a final concentration of 25 mM), transferred to a 60-ml serum bottle and capped. After a pre-incubation for 30 min, CH₄ was added to each bottle at final concentrations of 0.5, 1, 2, 3, 4, and 8% (v/v). To each incubation, with a certain concentration of CH₄, NH4+ was added in a range of 0.5–16 mM. The initial production of NO2- was measured, and the values were normalized to the total protein content of the cells. Incubations were performed at 55°C and shaking at 380 rpm. Each condition was performed in duplicate and values did not deviate more than 5%.

The complete genome sequence of strain SolV (Anvar et al., 2014), which is also available at the MicroScope annotation platform (https://www.genoscope.cns.fr/agc/microscope/home/), was used as the template for the transcriptome analysis (RNA-seq). A 4-ml volume of cells (OD₆₀₀ = 1) was sampled from the continuous cultures (H₂ and CH₄ grown cells under O₂ limitation) and from a batch culture (cells at μ_max grown on CH₄ without limitation) and harvested by centrifugation. The pellet was further used for mRNA isolation using the RiboPure™-Bacteria Kit according to the manufacturer's protocol (ThermoFisher, Waltham, USA). Briefly, cells were disrupted by cold Zirconia beads and after centrifugation, 0.2 volumes of chloroform was added to the supernatant for initial RNA purification. Next, 0.5 volumes of 100% ethanol was added to the aqueous phase obtained after chloroform addition and the whole sample was transferred to a filter cartridge. After washing, the RNA was eluted from the filter cartridge. Afterwards, using MICROBexpress™ kit (ThermoFisher, Waltham, USA) the ribosomal RNAs were removed from the total RNA. The rRNA removal efficiency was checked using the Agilent 2100 Bioanalyzer (Agilent, Santa Clara, USA). Next, Ion Total RNA-Seq Kit v2 (ThermoFisher, Waltham, USA) was used to construct the cDNA libraries from rRNA-depleted total RNA. Briefly, the rRNA-depleted total RNA was fragmented using RNase III and then, reverse transcription was performed on the fragmented RNAs. The obtained cDNAs were amplified and further purified to prepare barcoded libraries. To prepare the template for the Ion Personal Genome Machine® (PGM™) System, a volume of 15 μl from two sample libraries with a concentration of 14 pM were mixed. This mixture of two libraries was used to prepare the template-positive Ion Sphere™ particles (ISPs) using the Ion OneTouch™ 2 instrument. Afterwards, the template-positive ISPs were enriched using the Ion OneTouch™ ES instrument. Both template preparation and enrichment were performed using the Ion PGM™ Template OT2 200 Kit (Ion Torrent, Life technologies). Enriched templates were sequenced on an Ion 318™ Chip v2 using the Ion PGM™ sequencing 200 Kit v2. Expression analysis was performed with the RNA-seq Analysis tool from the CLC Genomic Work bench software (version 7.0.4, CLC-Bio, Aarhus, Denmark). The sequencing reads were first mapped to the ribosomal RNA operon and all tRNA and ncRNA genes, and mapped reads were discarded. The remaining reads were mapped to the CDS sequences extracted from the genome sequence of strain SolV (Anvar et al., 2014). Expression values are given as RPKM (Reads per Kilo base of exon model per Million mapped reads; Mortazavi et al., 2008). The total number of reads obtained and mapped on the coding sequences of the genome for each sample together with the calculated expression levels (RPKM) is provided in the Supplementary Material (Table S1).

To study the effect and conversion of nitrogenous compounds, three different continuous cultures were used which are referred to as CH₄/NH4+, H₂/NH4+, and CH₄/NO3-. In the second and third cultures, oxygen was limiting. Using a NOx analyzer and GC-MS, we demonstrated that in the CH₄/NH4+ culture with low actual CH₄ concentrations in the liquid (0.3 μM) and with NH4+ (4 mM), NO2- was not detected, and N₂O production rate was only 0.015 nmol N₂O.h⁻¹.mg DW⁻¹ (Table 1) which was 12,000-fold less than the CH₄ conversion rate (180 nmol.h⁻¹.mg DW⁻¹). To increase NO2- concentrations and study potential toxic effects of this compound, we used the H₂/NH4+ continuous culture applying different conditions. Initially, the production of NO2-, NO, and N₂O were measured under steady state conditions at a pH range of 3–5.5 under O₂ limitation (Figure 1). We showed that the NO2-, NO and N₂O concentrations were elevated by increasing the pH from 3 to 5.5 in the presence of 4 mM NH4+. Changing pH from 3 to 5.5 introduces more NH₃ in the medium. The NH₃ concentration in a range of 12 nM to 5 μM was calculated using the Henderson–Hasselbalch equation (Hütter, 1992), considering the temperature of 55°C at pH 3 to 5.5, respectively. At pH 5.5, we measured a NO2- concentration at steady state of about 420 μM in the reactor (Figure 2) resulting from a production rate of ≈ 48 nmol NO2-.h⁻¹.mg DW⁻¹, while nitrite production at pH 3 was very limited. Based on the clear effect of increasing pH on the production of NO2-, one could speculate that the real substrate for pMMO to produce NO2- is NH₃ (not NH4+). Furthermore, the NO2- reduction activities (NO and N₂O production) were measured at 0.81 nmol NO2-.h⁻¹.mg DW⁻¹ (1.7% of NH4+ oxidation rate) which is 53-fold higher than that in the CH₄/NH4+ culture (Table 1). A rapid NO2- consumption (≈83 nmol NO2-.h⁻¹.mg DW⁻¹) was observed when O₂ supply was switched off completely (Figure 2), and the NO2- reduction rate (as NO and N₂O) increased about 100-fold (74.4 nmol NO2-.h⁻¹.mg DW⁻¹). A rapid initial increase of NO suggests that conversion to N₂O is the rate limiting step. The decrease of N₂O levels was due to the continuous dilution of the gas present in the reactor headspace (total gas flow rate in the outlet ≈ 15 ml.min⁻¹). Concentrations of 1–5 μM NH₂OH were measured in data points before and after switching off O₂ supply.

We further tested the effect of different concentrations of NH4+ (4–20 mM) on the NO2-, NO and N₂O production at pH 4 under oxygen limitation in the H₂/NH4+ continuous culture (Figure 3). We showed that the concentrations of NO2-, NO, and N₂O slightly increased once the NH4+ concentration was gradually elevated. This observation indicates that at pH 4, even a 4-fold increase in the NH4+ concentration did not result in a high production of NO2- similar to what we observed at pH 5.5 supporting our assumption that pH plays an important role regarding the availability of NH₃ molecules. Furthermore, we showed that the cells in the CH₄/NO3- continuous culture were able to perform NO2- reduction at a rate of 120 nmol NO2-.h⁻¹.mg DW⁻¹ by converting the added NO2- (50 μM) to NO and further to N₂O in the absence of oxygen (Figure 4). Table 1 shows an overview of rates of ammonium oxidation to nitrite and nitrite reduction to NO/N₂O) in the different continuous culture.

The affinity constants (K_s) for NH4+ and NH₃ were determined using SolV cells from the CH₄/NO3- continuous culture. From the initial production rates of nitrite the best fitting curves to Michaelis–Menten kinetics were predicted (Figure S1). Since part of the NH4+ is present as NH₃ at pH 6 (1 M NH4+ is about 3 mM NH₃ at pH 6), the Michaelis–Menten curves were also produced based on the NH₃ concentrations (Figure S1). Therefore, we calculated apparent affinity constants (K_s) for both NH4+ and NH₃ in strain SolV (Table 2).

To identify the type of inhibition, the Michaelis–Menten curves were transformed to Lineweaver-Burk plots. Figure S2 shows a set of double reciprocal plots, obtained with different NH4+ concentrations in the presence of CH₄ at a range of 2, 4, and 8% (v/v). Increasing the CH₄ concentration resulted in a group of lines with a common intercept on the 1/V₀ axis but with different slopes. The intercept is 1/V_max and V_max is constant regardless of increasing CH₄ concentration (V_max = 1.61 ± 0.05 μmol.h⁻¹.mg protein⁻¹). The constant intercept of all lines suggests a competitive inhibition between CH₄ and NH₃. The NO2- production rate in the absence of CH₄ was about 3- to 4-fold lower compared to the rate in the presence of 0.5% (v/v) CH₄ suggesting that traces of CH₄ are essential for the pMMO activation. Table 2 shows an overview of affinity constants calculated for NH₃ (NH4+) obtained in the incubations with different CH₄ concentrations. Affinity constants for NH₃ were calculated based on the Henderson–Hasselbalch equation considering a temperature of 55°C (Hütter, 1992). These results showed that increasing CH₄ concentration limits the affinity of pMMO for NH₃ significantly, which correlates with the observed competitive inhibition between CH₄ and NH₃.

Expression levels of housekeeping genes and genes involved in metabolism of nitrogenous compounds were determined for H₂- and CH₄-grown cells (both under O₂ limited conditions). These values were compared to the expression values in cells growing at μ_max on CH₄ (without limitation). To compare baseline expression levels, we selected a group of 384 housekeeping genes (in total 437.9 kbp) involved in energy generation, ribosome assembly, carbon fixation (CBB cycle), C1 metabolism (except for pmo), amino acid synthesis, cell wall synthesis, translation, transcription, DNA replication, and tRNA synthesis (Khadem et al., 2012a,b). All ratios of expression levels of the housekeeping genes under these conditions were between 0.5 and 2 (Table S1). The robustness of the transcriptome data were tested using the method of Chaudhuri et al. (2011). In this method, the logarithmic value of RPKM + 1 of each condition (in duplicates) was calculated and the values were plotted against each other. This resulted in correlation coefficients of 0.80, 0.82, and 0.87 (Figure S3), showing the high robustness of the transcriptome data.

The transcriptome data showed that genes encoding the enzymes involved in NH4+ assimilation in strain SolV including glutamine synthase (GlnA)/glutamate synthase (GltB) and the alanine and glutamate dehydrogenases (Ald, Gdh) were equally expressed under all conditions (Table 3). Among these genes, only glnA was about 2.5-fold less expressed in the continuous cultures compared to the cells grown at μ_max (Table 3). We also found that the carAB operons (encoding the glutamine hydrolyzing carbamoyl-phosphate synthase) were constitutively expressed. The conversion of glutamine and carbon dioxide into glutamate and carbamoyl phosphate is performed by this enzyme (Khadem et al., 2012a). Similarly, the argDHFG operons (encoding enzymes from the urea cycle) were expressed under all conditions. Interestingly, we detected the ammonium/ammonia transporter (amtB) was at least 3-fold up-regulated in the CH₄/NO3- continuous culture compared to the other conditions reflecting that cells may have a preference for NH4+ as N-source. In addition, the genes encoding the NO3-/NO2- transport (nasA) and the assimilatory nitrite and nitrate reductases were 9- to 45-fold up-regulated in the CH₄/NO3- continuous culture compared to the cells at μ_max (Table 3). Both latter observations correlate with the fact that nitrate was used as N-source under this condition. Interestingly, the transcriptome analysis showed that the nirK and norC genes were up-regulated in the chemostat continuous culture compared to those at μ_max, while results for norB (encoding the catalytic subunit) were less clear. This may imply that other NO reductases were active. We also found that the haoA gene was about 2-fold down-regulated in the CH₄/NO3- continuous culture compared to the H₂/NH4+ and μ_max cultures, likely due to the absence of NH4+ in this condition. The haoA gene showed comparable high expression levels in the H₂/NH4+ continuous and batch μ_max culture (Table 3).

The transcriptome data showed different expression levels of two of the three different pmo operons in strain SolV (Table 4). We found that the pmoCAB2 operon including the mfumv2_1793, mfumv2_1792 and mfumv2_1791 subunits was significantly expressed (RPKM values 14,899–37,218) in the cells growing at μ_max with no limitation and the pmoCAB1 operon showed very low expression. In contrast, cells in the continuous cultures on H₂/NH4+ and CH₄/NO3- under O₂ limitation showed a significantly different expression pattern of the pmoCAB operons. We found that the pmoCAB1 operon including mfumv2_1796, mfumv2_1795 and mfumv2_1794 subunits was very highly expressed under these conditions (RPKM values 5,003–47,785), whereas the expression levels of the pmoCAB2 operon was found to be 2- to 19-fold lower in comparison to the cells growing at μ_max. The pmoCAB3 operon including the mfumv2_1606, mfumv2_1605 and mfumv2_1604 subunits showed low expressed under all conditions although expression in H₂/NH4+ grown cells seems to be slightly up-regulated. The conversion of methanol to formaldehyde is the second step in CH₄ oxidation pathway. Interestingly, it has been shown that strain SolV contains a XoxF-type methanol dehydrogenase (MDH) that can convert methanol directly to formate (Pol et al., 2014). We found that the xoxFGJ operon encoding the methanol dehydrogenase and pqqABCDEF operon encoding the proteins involved in biosynthesis of the methanol dehydrogenase cofactor pyrroloquinoline quinone were expressed more or less similar under all conditions tested. The last step of the CH₄ oxidation pathway is conversion of formate to CO₂ catalyzed by NAD-dependent formate dehydrogenase and a membrane-bound formate dehydrogenase. The genes encoding these enzymes were expressed under all conditions, although the expression levels of these enzymes (except for fdsD and fdh) in continuous cultures under O₂ limitation was 2- to 2.5-fold lower compared to cells grown at μ_max (Table 4).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.