Key Physiology of a Nitrite-Dependent Methane-Oxidizing Enrichment Culture

Methane is an important greenhouse gas with a radiative forcing 28 times that of carbon dioxide over a 100-year time scale. The emission of methane to the atmosphere is controlled by aerobic and anaerobic methanotrophs, which are microorganisms that are able to oxidize methane to conserve energy. While aerobic methanotrophs have been studied for over a century, knowledge on the physiological characteristics of anaerobic methanotrophs is scarce. Here, we describe kinetic properties of “Candidatus Methylomirabilis lanthanidiphila,” a nitrite-dependent methane-oxidizing microorganism, which is ecologically important and can be applied in wastewater treatment.

ABSTRACT

Nitrite-dependent methane-oxidizing bacteria couple the reduction of nitrite to the oxidation of methane via a unique oxygen-producing pathway. This process is carried out by members of the genus Methylomirabilis that belong to the NC10 phylum. Contrary to other known anaerobic methane oxidizers, they do not employ the reverse methanogenesis pathway for methane activation but instead a canonical particulate methane monooxygenase similar to those used by aerobic methanotrophs. Methylomirabilis-like bacteria are detected in many natural and manmade ecosystems, but their physiology is not well understood. Here, using continuous cultivation techniques, batch activity assays, and state-of-the-art membrane-inlet mass spectrometry, we determined growth rate, doubling time, and methane and nitrite affinities of the nitrite-dependent methane-oxidizing bacterium “Candidatus Methylomirabilis lanthanidiphila.” Our results provide insight into understanding the interactions of these microorganisms with methanotrophs and other nitrite-reducing microorganisms, such as anaerobic ammonium-oxidizing bacteria. Furthermore, our data can be used in modeling studies as well as wastewater treatment plant design.

IMPORTANCE Methane is an important greenhouse gas with a radiative forcing 28 times that of carbon dioxide over a 100-year time scale. The emission of methane to the atmosphere is controlled by aerobic and anaerobic methanotrophs, which are microorganisms that are able to oxidize methane to conserve energy. While aerobic methanotrophs have been studied for over a century, knowledge on the physiological characteristics of anaerobic methanotrophs is scarce. Here, we describe kinetic properties of “Candidatus Methylomirabilis lanthanidiphila,” a nitrite-dependent methane-oxidizing microorganism, which is ecologically important and can be applied in wastewater treatment.

INTRODUCTION

For a long time, microbial methane oxidation was thought to be carried out only by aerobic bacteria (1, 2). However, in 2000, sulfate-dependent anaerobic oxidation of methane (AOM) was demonstrated as a microbially mediated process (3). This finding was followed by the discovery of nitrate- and nitrite-dependent anaerobic oxidation of methane (nitrate- and nitrite-AOM) (4–7). Several studies in the last decade established that nitrite-AOM is performed by bacteria belonging to the NC10 phylum, represented by the cultured species “Candidatus Methylomirabilis oxyfera,” “Candidatus Methylomirabilis lanthanidiphila,” and “Candidatus Methylomirabilis sinica” (4, 7, 8). Even though these microorganisms have an anaerobic metabolism, they oxidize methane via an intra-aerobic pathway and activate methane using oxygen through a particulate methane monooxygenase (pMMO), while reducing nitrite to N₂ (4). In contrast, nitrate-AOM is catalyzed by archaea from the ANME-2d clade, such as “Candidatus Methanoperedens nitroreducens,” which oxidize methane via the reverse methanogenesis pathway and reduce nitrate preferentially to nitrite and further to ammonium (5, 9, 10).

Since nitrite-AOM was first discovered and the important role of Methylomirabilis-like bacteria in anaerobic oxidation of methane became apparent, molecular tools based on 16S rRNA and pmoA gene amplification were developed to survey this group of bacteria (11, 12). Consequently, ribosomal 16S rRNA and pmoA gene sequences related to this clade have been reported from diverse environments, such as coastal aquifers (13), wastewater treatment plants (14, 15), sediments from freshwater and saline lakes (16–19), peat soil from minerotrophic peatlands (20), agricultural lands and paddy field soils (21–24), marine oxygen minimum zones, sediments (25, 26), and rumen fluid samples (27). Methylomirabilis-like bacteria appear to be present in a wide variety of environments; however, their contribution to global nitrogen and methane fluxes is not yet fully understood (28). Archaea performing nitrate-AOM are usually detected in similar environments where NC10 bacteria are present, such as paddy fields, agricultural sediments, and wastewater treatment plants (6, 24, 29, 30). Furthermore, sequences related to ANME-2d clade archaea have been reported in marine sediments (31), methane seeps (32), and intertidal zones (33). However, their environmental distribution and relevance are less known than those of Methylomirabilis-like bacteria and even fewer studies are specifically focused on their physiology (10, 34, 35).

There have been many attempts to enrich bacteria performing nitrite-AOM to further study their physiology under controlled conditions (8, 29, 36). So far, published enrichment cultures were obtained from limited types of inocula, and the overall enrichment conditions did not change significantly since the first publication of a culture that contained both Methylomirabilis- and Methanoperedens-like microorganisms (6). When nitrate and nitrite were supplied together to anoxic sediments collected from agricultural ditch canals, both nitrate- and nitrite-dependent methane-oxidizing microorganisms were enriched (6, 36). Enrichments solely fed with nitrate resulted in similar cocultures where nitrate-AOM supports nitrite-AOM by producing nitrite (10, 37, 38). On the other hand, when ammonium, methane, and nitrite were supplied as substrates, cocultures of Methylomirabilis species and anaerobic ammonium-oxidizing (anammox) bacteria were achieved (14, 39–41). It was demonstrated that a coculture of Methylomirabilis-like and anammox bacteria could simultaneously oxidize ammonia and methane using nitrite as the common electron acceptor (40). Clearly, there is an intricate interplay between nitrite- and nitrate-dependent anaerobic methane- and ammonium-oxidizing microorganisms in natural ecosystems. These interactions can be advantageous when trying to apply these processes for ammonium and methane removal from wastewater (28, 42, 43). However, key physiological information that is necessary to understand how these processes function in nature or how to apply them for wastewater treatment is still lacking.

Here, we report a combined strategy to selectively enrich a culture that solely contained nitrite-dependent anaerobic methane-oxidizing bacteria as the dominant microorganism, in the absence of nitrate-dependent methane-oxidizing archaea. After revision of the mineral medium composition, we supplemented cerium as a trace metal due to its relevance in methanotrophic metabolism (44). Our efforts resulted in a culture dominated by “Candidatus Methylomirabilis lanthanidiphila” (8). We determined the affinity of this enrichment culture for dissolved methane and nitrite. Furthermore, we used constant biomass removal to achieve a stable growth rate and were, thus, able to calculate values for apparent yield and doubling time. These results contribute to the necessary knowledge to understand the ecophysiology of the nitrite-AOM process and enable progress toward its sustainable application in wastewater treatment.

RESULTS AND DISCUSSION

Growth medium composition, including macronutrients and micronutrients as well as the supplied substrates and applied physical parameters, control which groups of microorganisms will be dominant in a microbial culture. In previous studies, the mineral medium described by Raghoebarsing et al. in 2006 has been used to enrich nitrite- and nitrate-dependent anaerobic methane-oxidizing microorganisms without significant changes (6). Recent studies have revealed that trace elements, such as Cu, Fe, and cerium (Ce), might be limiting in this medium (21, 36, 44–46). While Cu and Fe are typical metals involved in many enzymes, Ce has only been recently recognized as an important metal for the XoxF-type methanol dehydrogenase (MDH) of methanotrophic and methylotrophic microorganisms, which employ Ce instead of Ca (44, 47). Therefore, the inclusion of this trace metal in the growth medium could influence the type of Methylomirabilis bacteria based on the MDH it carries and expresses. Indeed, when cerium was supplied to an existing enrichment culture, a new species named “Candidatus Methylomirabilis lanthanidiphila” was enriched (8). Genome analysis of this new species and of “Candidatus Methylomirabilis limnetica” showed that both microorganisms only carry the XoxF-type MDH and no MxaF-type Ca-dependent MDH (18).

In the present study, biomass from the enrichment culture of “Candidatus Methylomirabilis lanthanidiphila” was used to seed a new bioreactor. The seeding culture contained both Methanoperedens-like archaea and “Candidatus Methylomirabilis lanthanidiphila” (8). The addition of nitrate to bioreactors as a substrate, next to nitrite and methane, determines whether a coculture of Methanoperedens-like archaea and Methylomirabilis-like bacteria (6, 10, 37, 38) or a monoculture of Methylomirabilis-like bacteria (29, 45) will be obtained. Therefore, to achieve a monoculture of “Candidatus Methylomirabilis lanthanidiphila,” we supplied our bioreactor solely with nitrite as an electron acceptor and methane as an electron donor. The population of Methanoperedens-like archaea decreased (initially approximately 10%) and after 300 days was below the detection limit of fluorescence in situ hybridization (FISH) microscopy (<1%) (Fig. 1).

FIG 1.

Fluorescence in situ hybridization of biomass from the “Candidatus Methylomirabilis lanthanidiphila” enrichment culture. (A) Inoculum, (B) after 9 months of enrichment, and (C) after 17 months of enrichment. In cyan, “Candidatus Methylomirabilis lanthanidiphila” (simultaneous hybridization with Dbact193 in FLUOS; EubMix in CY5); in orange, Methanoperedens-like archaea (Darch641 in CY3). Bar = 20 μm.

We used this culture to determine the key physiological properties of nitrite-dependent anaerobic methane-oxidizing bacteria. In order to have an estimate of their growth rate and yield, biomass was removed from the bioreactor at a flow rate of 250 ml d⁻¹ (Fig. 2). When the biomass removal rate was increased to above 300 ml d⁻¹, the culture showed nitrite accumulation and biomass removal was stopped immediately to counteract potential inhibition of the culture due to high nitrite concentrations. Under the imposed growth conditions, during two distinct periods of protein measurements (208 to 238 days and 430 to 460 days), an apparent μ_max of 0.14 d⁻¹ was estimated, which corresponds to a doubling time of 5 days. This value was lower than the previous estimates that were between 2 to 8 weeks, which were calculated based on an increase of copy numbers of 16S rRNA genes or nitrogen loading rate, without constant biomass washout (11, 36). A doubling time of 5 days could be advantageous for Methylomirabilis-like bacteria when coping with dynamic conditions in natural habitats or humanmade ecosystems, such as wastewater treatment plants. Nevertheless, it should be noted that the shortest doubling time of Methylomirabilis-like bacteria could be even lower, as demonstrated for anaerobic ammonium-oxidizing (anammox) bacteria. Initially, anammox bacteria were estimated to have a doubling time of around 11 days based on experiments with flocculent biomass (48), whereas experiments with planktonic cells showed that the doubling time of anammox bacteria could be as short as 3 days (49).

FIG 2.

Nitrite removal and protein content of the “Candidatus Methylomirabilis lanthanidiphila” enrichment culture. Solid and dashed lines represent nitrite load in the medium and in the effluent of the enrichment culture, respectively. Filled circles represent protein measurements. Cerium was added to the reactor starting with inoculation, and biomass removal was started on day 27.

The two distinct periods used for growth rate estimations were also used to determine the yield of “Candidatus Methylomirabilis lanthanidiphila.” In these periods, the average protein content of the bioreactor was 155 ± 57 μg_protein ml⁻¹ (day 208 to 238) and 133 ± 39 μg_protein ml⁻¹ (day 430 to 460). From these measurements and the mass balance determined in these time periods in our enrichment culture, yields of 0.073 ± 0.027 and 0.085 ± 0.025 C mol_biomass mol_methane⁻¹ were estimated, respectively. A stoichiometry of 8:3 (nitrite:methane) was used to calculate these values (6). The measured yield of “Candidatus Methylomirabilis lanthanidiphila” was slightly higher than the yield of anammox bacteria (0.066 C mol_biomass mol_ammonium⁻¹ [50]), which co-occur with Methylomirabilis species in natural and manmade ecosystems where they compete for the same electron acceptor, nitrite (14, 40). At first glance, this might suggest that a higher yield would give a competitive advantage to Methylomirabilis bacteria; however, it has been demonstrated that anammox bacteria were better competitors under electron acceptor limitation, suggesting that their affinity for nitrite was better than that of Methylomirabilis bacteria (40).

Here, we also determined the affinity of “Candidatus Methylomirabilis lanthanidiphila” to nitrite and methane. Using in situ batch incubations in the bioreactor, an apparent affinity constant of 7 μM for nitrite was determined (Fig. 3), which was the same as what was previously measured for Methylomirabilis oxyfera (51). This value was higher than that of anammox bacteria, which was estimated to be below 5 μM (52, 53). This result was in line with earlier studies, which reported that under nitrite limitation, anammox bacteria outcompete microorganisms that perform nitrite-AOM (40).

FIG 3.

Methane-dependent nitrite reduction by the “Candidatus Methylomirabilis lanthanidiphila” enrichment culture.

Methylomirabilis bacteria appear to have a high affinity for their electron donor, methane. Previous experiments were all performed by measuring methane in the gas phase of bioreactors, and values as high as 87 to 97 μM were reported (54). Using ex situ batch incubations and online dissolved gas measurements with membrane-inlet mass spectrometry (MIMS), a V_max of approximately 5.5 nmol methane min⁻¹ mg protein⁻¹ was determined (Fig. 4). The estimated apparent affinity constant of “Candidatus Methylomirabilis lanthanidiphila” for methane was 2.6 ± 0.7 μM, which was comparable to aerobic methane oxidizers that have an affinity constant in the range of 0.2 to 6 μM (55–57). Affinity to methane could be an important controlling factor determining which methane-oxidizing clades would prevail under methane-limited conditions in natural and humanmade ecosystems. It should be noted that it is most likely that Methylomirabilis bacteria have higher intrinsic affinities (i.e., lower apparent affinity constants) for both methane and nitrite than the estimated values, as experiments with Methylomirabilis bacteria were conducted using flocculent biomass in the absence of available planktonic cells.

FIG 4.

(A) Methane oxidation by the “Candidatus Methylomirabilis lanthanidiphila” enrichment culture measured by membrane-inlet mass spectrometry. Four replicates are displayed. Data are normalized to an initial concentration of 8.5 μM methane. (B) Average methane oxidation calculated from the four distinct measurements displayed in panel A. (C) Michaelis-Menten kinetics calculated from values in panel B.

Recently, the methane affinity of nitrate-dependent methane-oxidizing Methanoperedens-like archaea was estimated at 0.5 mM (35). This value is similar to the affinities of microorganisms (ranging from 1.1 mM to 37 mM) that perform sulfate-dependent methane oxidation employing the same reverse methanogenesis pathway (58–61) but considerably higher than that of bacterial methane oxidizers. The difference of affinities for methane could be due to the fact that Methylomirabilis bacteria use canonical particulate methane monooxygenases (pMMO), whereas Methanoperedens-like archaea employ the reverse methanogenesis pathway and activate methane using methyl-coenzyme M reductase (MCR), the enzyme that is responsible for methane production in methanogenesis (62).

Taken together, these results shed light on the ecophysiology of Methylomirabilis bacteria and help understand which groups of microorganisms could prevail under different regimes, such as methane, nitrite, and/or nitrate fluxes, availability of trace elements such as Ce that can be used as a cofactor, and limitation of different substrates or washout rates. Furthermore, the basic kinetic information is also important when designing wastewater treatment plants aiming to remove ammonium and methane simultaneously by making use of anaerobic ammonium and methane oxidation processes. Currently, the anammox process is applied in oxygen-limited systems where aerobic ammonium-oxidizing bacteria supply nitrite to anammox bacteria. The lower affinity of Methylomirabilis bacteria toward nitrite than that of anammox bacteria suggests that the availability of ammonium would be an important control parameter for the competition between these two clades of microorganisms in natural and humanmade ecosystems. This means that the wastewater treatment plants would have to be run under ammonium limitation to prevent the loss of Methylomirabilis bacteria from the system (40). Furthermore, the presence of oxygen in an environment where methane is also present would most likely induce the growth and activity of aerobic methane-oxidizing bacteria (34, 63). In such a situation, aerobic ammonia-oxidizing microorganisms would compete for oxygen with aerobic methane-oxidizing microorganisms, while Methylomirabilis bacteria would compete for nitrite with anammox bacteria and for methane with aerobic methane-oxidizing microorganisms (28, 42). Studying the fascinating interplay of these four clades of microorganisms under ammonium, oxygen, or methane limitation in controlled laboratory-scale bioreactors is required to elucidate their interactions in nature as well as to determine the feasibility of the application of a bioreactor containing these microorganisms for ammonium and methane removal.

MATERIALS AND METHODS

Enrichment conditions.

From an established enrichment culture (8), biomass (4 liters) was taken to inoculate a new bioreactor, which was made anoxic by flushing with a mixture of Ar/CO₂ (95%/5%, vol/vol) for more than 72 h. This enrichment culture was operated as a continuous sequencing batch reactor (nominal volume, 6 liters; Applicon Biotechnology B.V., Applisens, Schiedam, the Netherlands). The reactor was monitored with an oxygen and a pH probe (Applikon Instruments B.V.). Temperature and pH were maintained at 30°C and 7.5, respectively. The enrichment culture was supplied with a mixture of CH₄/CO₂ (95%/5%, vol/vol), where methane was the sole electron donor. The medium was constantly flushed with a gas mixture of Ar/CO₂ (95%/5%, vol/vol). The volume was kept at 4 liters through a level-sensor-controlled pump, the operation occurred in sequential feed (11 h), with an average medium flow rate of 22.7 ml h⁻¹, and rest (1 h) cycles, and stirred at a constant speed of 200 rpm during feed periods. The hydraulic retention time was 4 days. Biomass removal started on day 27 and was performed throughout the reactor operation at a rate of 250 ml d⁻¹ (solids retention time [SRT], 16 days). Biomass removal was stopped only temporarily in periods of reactor instability, such as hardware malfunction or transient nitrite accumulation.

Medium composition.

The composition of the mineral medium was adapted from the initial medium described previously (6). The importance of lanthanides for methanotrophic bacteria described by Pol et al. prompted us to include Ce (44). The mineral medium contained the following per liter: 94 mg MgSO₄, 217 mg CaCl₂, 100 mg KH₂PO₄, 800 μg FeSO₄, 70 g ZnSO₄, 33 μg CoCl₂, 750 μg CuSO₄, 52 μg NiCl₂, 7 μg H₃BO₃, 64 μg MnCl₂, 22 μg Na₂WO₄, 53 μg Na₂MoO₄, 25 μg SeO₂, and 100 μg CeCl₂. The nitrite concentration in the medium was adjusted according to the consumption rate and varied from 1 to 40 mM. Nitrite concentration in the reactor was mostly below the detection limit (<40 μM), seldom increasing up to approximately 250 μM.

Dissolved methane measurements.

Methane consumption was determined through membrane-inlet mass spectrometry (MIMS), in a custom-made glass cylinder cell (volume, 32 ml). The biomass sample was collected from the seeding bioreactor culture (8), while sparging with Ar/CO₂ (95%/5%, vol/vol). Biomass was anaerobically washed 3 times with 10 mM MOPS buffered medium (pH 7.3), and MIMS cell was kept anaerobic through sparging with argon before the experiment. Methane was supplied with sequential injections (100 μl) of methane-saturated demineralized water at 1 bar of overpressure (approximate methane concentration of 2.6 mM).

Determination of kinetic parameters.

The results from the MIMS, in the form of counts per second, were converted to concentration during the measurement events. The raw data were subsampled in intervals of 10 seconds and the kinetic determinations were expressed per minute. To determine the approximate values for V_max and the apparent methane affinity constant (K_{methane app}), linearization through Lineweaver Burk plotting was performed to a set of 4 replicates of methane oxidation curves (Fig. 4). The numerical values obtained for all curves were averaged, and from the mathematical model, the kinetic parameters were determined.

The data were converted from counts per second to concentration, by considering an average of 300,000 cps as a zero baseline, based on 3 hours of constant equilibration to determine the baseline from the system. An average of 390,000 cps was consistently the signal intensity corresponding to the starting point of repetitive injections of 100 μl of Milli-Q water saturated with methane at 1 bar of overpressure, which corresponded to 8.5 μM methane at every injection. The reproducibility was experimentally observed with repetitive injections of increasing volumes, resulting in a consistent increase in counts per second.

Protein determination.

Biomass (10 ml) was taken two times per week when the reactor volume was at 4 liters. Three independent samples were taken and homogenized using a glass homogenizer. Each sample was mixed with a 3-M NaOH solution (1:1) and incubated for 30 minutes at 90°C. After being incubated, the sample was brought to room temperature, and CuSO₄ (4%, wt/vol) was added (2:1, sample:CuSO₄). Samples were vortexed for 30 seconds and centrifuged (13,000 × g, 5 min). The supernatant was measured at 540 nm in a spectrophotometer (Spectronic 200; Thermo Scientific, USA). Bovine serum albumin was used as the protein standard.

Fluorescence in situ hybridization.

Biomass (1.5 ml) was disrupted by passing the sample through a series of needles (inner diameter, 0.8, 0.6, and 0.4 mm). Once disrupted, the sample was washed twice with 1 ml phosphate-buffered saline (PBS; 130 mM NaCl and 10 mM phosphate buffer, pH 7.4). After the washing steps, the sample was centrifuged (13,000 × g, 5 min), and the pellet was resuspended in 0.3 ml of PBS and 900 μl paraformaldehyde (4%) and was incubated overnight at 4°C for chemical fixation. FISH was performed as previously described using 35% formamide (45). Fluorescent probes targeting Methylomirabilis-like bacteria (Dbact193 labeled with FLUOS, [6]), most bacteria (EUBI-III labeled with CY5 [64]), and Methanoperedens-like archaea (Darch641 labeled with CY5 [65]) were used. Images were obtained with a Zeiss Axioplan 2 epifluorescence microscope, together with the Axiovision software package (Zeiss, Germany).