Detection, Isolation, and Characterization of Acidophilic Methanotrophs from Sphagnum Mosses

Abstract

Sphagnum peatlands are important ecosystems in the methane cycle. Methane-oxidizing bacteria in these ecosystems serve as a methane filter and limit methane emissions. Yet little is known about the diversity and identity of the methanotrophs present in and on Sphagnum mosses of peatlands, and only a few isolates are known. The methanotrophic community in Sphagnum mosses, originating from a Dutch peat bog, was investigated using a pmoA microarray. A high biodiversity of both gamma- and alphaproteobacterial methanotrophs was found. With Sphagnum mosses as the inoculum, alpha- and gammaproteobacterial acidophilic methanotrophs were isolated using established and newly designed media. The 16S rRNA, pmoA, pxmA, and mmoX gene sequences showed that the alphaproteobacterial isolates belonged to the Methylocystis and Methylosinus genera. The Methylosinus species isolated are the first acid-tolerant members of this genus. Of the acidophilic gammaproteobacterial strains isolated, strain M5 was affiliated with the Methylomonas genus, and the other strain, M200, may represent a novel genus, most closely related to the genera Methylosoma and Methylovulum. So far, no acidophilic or acid-tolerant methanotrophs in the Gammaproteobacteria class are known. All strains showed the typical features of either type I or II methanotrophs and are, to the best of our knowledge, the first isolated (acidophilic or acid-tolerant) methanotrophs from Sphagnum mosses.

INTRODUCTION

Methane is an important greenhouse gas, and its concentration has been rising rapidly since industrial times (31). Methanotrophs are a sink for methane and occur in many different ecosystems like rice paddies, soils, volcanic areas, and peat bogs (17, 33, 55). Acidic peat bogs are the most extensive type of wetland, occupying about 3% of the total land area and storing an enormous amount of carbon (32). Methanotrophs present in these peatlands can act as a filter for methane, thereby reducing its emissions from these wetlands (56, 57). The biodiversity of methanotrophic communities can be investigated using PCR with primers targeting the 16S rRNA gene or functional genes like the methane monooxygenase genes pmoA, pxmA, and mmoX (18, 47, 68). To quickly screen the methanotrophic community of an ecosystem, a microarray technique based on the pmoA gene has been developed (8). The primers used in this technique are based on pmoA sequences of currently available cultivated methanotrophs and some uncultivated methanotrophs, and therefore, culture-dependent studies, e.g., isolation of new strains, remain important to expand our knowledge on the microbial methanotrophic communities. Molecular analysis of the endophytic methanotrophic community of Sphagnum mosses is sparse. Recently the abundance of both alpha- and gammaproteobacterial methanotrophs within this community was described (43, 44). Other molecular surveys focused on peat lands (soil, mires, etc.) rather than mosses (3, 16, 22, 81).

Aerobic methanotrophs occur within the Alpha- and Gammaproteobacteria classes and Verrucomicrobia (17, 52, 55). The gammaproteobacterial (type I) methanotrophs use the ribulose monophosphate pathway for carbon fixation and possess disc-shaped intracellular membranes (ICMs) throughout the cell. The alphaproteobacterial (type II) methanotrophs use the serine pathway for carbon fixation and have ICMs parallel to the cytoplasmic membrane (33, 62, 72). The recently described extremely acidophilic methanotrophic members of the Verrucomicrobia phylum isolated from volcanic areas do not contain intracellular membrane structures, and their biochemistry and physiology still needs to be resolved in detail (37, 41, 55). Currently used primers for the functional gene pmoA are not able to detect these new methanotrophs, which again demonstrates the need for culture-dependent studies to update our understanding of the methanotrophic world.

Isolation of methanotrophs from peat ecosystems is a challenge. Peat bogs are a harsh environment for microbes to live in because of the low pH, around 4.5 or lower, and the low nutrient content. Several methanotrophs have been isolated from peatlands using strongly diluted oligotrophic media (21, 24–26), but molecular analysis of peat soils and Sphagnum mosses indicate that many methane-oxidizing bacteria remain uncultured (19, 43, 44).

The acidophilic or acid-tolerant methanotrophs isolated from peat so far all belong to the Alphaproteobacteria and include Methylocella species (21, 25, 28), Methylocapsa species (24), and Methylocystis species (20). No methanotrophic Gammaproteobacteria have been isolated from peat ecosystems yet, and none of the known gammaproteobacterial methanotrophic isolates are acidophilic, i.e., capable of growth below pH 5. It has been shown that methanotrophs are present on and can live inside Sphagnum mosses (43, 44, 57). Until now, no methanotrophs have been isolated directly from the Sphagnum peat mosses, since the acidophilic Alphaproteobacteria were all isolated from peat soils.

The present study describes a pmoA-based microarray to analyze methanotrophic diversity in Sphagnum mosses from a Dutch peat bog and the enrichment, isolation, and characterization of methanotrophs from Sphagnum mosses.

MATERIALS AND METHODS

Sampling site and methane oxidation tests.

Samples were collected from an acidic Sphagnum peat bog at the Mariapeel nature reserve (Netherlands; 51°24′28.4′′N, 5°55′8′′E) (61, 65, 70), the Hatertse Vennen (Netherlands; 51°47′4′′N, 5°48′2′′E), and the nature reserve Parc Naturel Haute Fagnes (Belgium; 50°27′0.6′′N, 5°55′39′′E). The pH of the peat water at the time of sampling was between 3.8 and 4.3. The moss species from the Dutch peat bogs was identified as Sphagnum cuspidatum and that from Belgium as Sphagnum denticulatum. Samples were taken from floating Sphagnum rafts at the surface of the water. Only alive Sphagnum mosses were sampled, showing chlorophyll and no degradation at the bottom of the plant. For all experiments, complete Sphagnum mosses were used. Sphagnum mosses were thoroughly washed and incubated in 120-ml bottles with 1 ml of methane to determine methane oxidation rates. Incubations were performed in triplicate.

Media.

Two different enrichments media, medium M2 (27) and medium N, were used. Medium N is a newly developed medium based on the peat water composition analysis of the Mariapeel (Netherlands). The medium contained, per liter, 1.75 mg KH₂PO₄, 1.01 mg KNO₃, 8.02 mg NH₄Cl, 2.92 mg NaCl, 17.6 mg CaCl₂ 0.2H₂0, 9.86 mg MgSO₄ 0.7H₂0, 2.44 mg Na₂SiO₃, 1.33 mg AlCl₃, and 0.2% of trace element solution as described previously (27), which contained 0.12 g CuSO₄·5H₂O rather than 0.1 g CuCl₂·5H₂O.

Culture conditions and isolations.

After incubation, the mosses from the Mariapeel were transferred to a Fernbach flask with 100 ml filter sterile (0.2 μm) peat water from the Mariapeel. The flasks were closed with red butyl rubber septa and screw caps. Incubations were performed with 5% (vol/vol) methane at room temperature on a table shaker at 90 rpm, while methane concentration and turbidity were monitored daily. Sphagnum from Haute Fagnes was incubated in a 250-ml bottle with 100 ml medium N, 5% methane, and 5% carbon dioxide to obtain a pH of 4.5. Upon methane consumption of both enrichments, dilution series were made after crushing the Sphagnum moss with a mortar. Dilutions (10¹- to 10⁹-fold) were made in 120-ml or 60-ml serum bottles containing 6 or 3 ml of medium N or M2 and 5% CO₂-1% CH₄ in the headspace. The bottles were sealed with gray butyl rubber septa and alumina crimp seals. Methane and carbon dioxide were added before autoclaving. Bottles were incubated at room temperature, and methane concentrations were measured when the cultures became turbid. Growth and methane consumption were found in up to the 10⁷-fold dilution, and aliquots of this dilution were transferred to agarose plates (1% [wt/vol] Agarose MP; Roche). After colonies appeared, these were restreaked to obtain single colonies and pure cultures, which were used for colony PCR and further culturing.

Isolations and enrichments on medium M2 resulted in seven pure cultures, designated strains 29, M162, M167, M169, M175, M212, and M242. Strain M242 originated from a Belgian peat ecosystem, strain M169 from the Hatertse Vennen, and the others from the Mariapeel.

From the Mariapeel Sphagnum enrichments on medium N, one methanotroph was obtained in pure culture and showed slimy, mucous-like colonies on plate. This culture will be referred to as strain M5. From the Belgian Sphagnum enrichments on medium N, one methanotroph was also obtained in pure culture and showed pink, slimy growth on agarose plates and flocs in liquid culture. This culture will be referred to as strain M200. The flocs could be destroyed by passing the culture through a small needle, using a syringe, and this was repeated several times before measuring the optical density at 600 nm (OD₆₀₀). All isolates can be provided upon request.

Genomic DNA extraction and microarray.

The genomic DNA from the thoroughly washed Sphagnum mosses was isolated as described previously (44). The pmoA-based microarray was performed as described previously (8), except for the pmoA-based PCR. Genomic DNA extracted from Sphagnum mosses was used as template in a touchdown PCR with the pmoA primer set A189/T7-A682 using the following PCR program: 94°C for 5 min, 11 cycles for 1 min at 94°C, 1 min at 62°C with decrease of 1°C at every cycle, 1 min at 72°C, followed by 25 cycles of 1 min at 94°C, 1 min at 52°C, and 1 min at 72°C, followed by 10 min at 72°C. The PCR product obtained was diluted 100 times and used in a nested pmoA touchdown PCR with the primer set A189/T7-mb661 and the following program: 94°C for 5 min, 11 cycles of 1 min at 94°C, 1 min at 62°C with a decrease of 1°C at every cycle, and 1 min at 72°C, followed by 14 cycles of 1 min at 94°C, 1 min at 52°C, and 1 min at 72°C, followed by 10 min at 72°C. To perform the microarray, 300 ng of PCR product was needed.

PCR amplification of the 16S rRNA gene and the pmoA, pxmA, and mmoX genes.

Single colonies were picked with a sterile toothpick and transferred to 20 μl sterile Milli-Q (MQ). The inoculated Milli-Q water was used as a template for PCR. The 16S rRNA gene was amplified with general bacterial primers 616F (39) and 1492R (45). Both the pmoA and the mmoX genes were amplified with general primers using the recommended annealing temperatures (2, 36, 48, 50, 63). The pxmA genes were amplified using the primers pxmA230F and pxmA732R (68) at a gradient from 50 to 60°C.

All PCR products were purified using the QIAquick PCR purification kit (Qiagen Benelux B.V., Venlo, Netherlands). DNA sequencing was performed with the primers used in the PCR, and for the 16S rRNA gene, the universal bacterial primer 612R (29) was also used. DNA sequencing was performed at the sequencing facility of the UMC Sint Radboud, Nijmegen, Netherlands. Phylogenetic analyses were conducted with the MEGA4 package (67). Phylogenetic dendrograms were constructed using the neighbor-joining and maximum parsimony algorithms, as implemented in the MEGA software. Evolutionary distances between sequences were calculated using the DNA and deduced amino acid sequences.

Physiological tests.

The pH and temperature optimum tests were performed with M2 medium (27). The medium was adjusted to pH 3.5 to 7 with 0.1 M solutions of H₃PO₄, KH₂PO₄, and K₂HPO₄. Serum bottles (120 ml) containing 12 ml of medium were inoculated with an active preculture of the isolate. Bottles contained 1 to 5% (vol/vol) methane and were sealed with butyl rubber septa and aluminum caps. After 5 to 8 days of incubation at room temperature (20°C), culture turbidity and methane consumption were analyzed. When no growth occurred within 8 days, the culture was incubated for up to 40 days to check for growth. A temperature optimum test was performed at pH 4.7 at 4, 10, 15, 20, 25, 30, and 37°C. Maximum OD₆₀₀ was between 0.5 and 0.6 for all tested strains in the pH and temperature optima, but no additional methane was added during incubation. Cultures of different incubation periods were tested for resistance to freeze-drying, desiccation, and storage in glycerol at −80°C (79). Viability of the cultures was tested by growth on solid and liquid M2 medium. Cultures were centrifuged for 1 min at maximum speed in an Eppendorf centrifuge. Desiccation resistance was tested by spreading the culture on sterile glass slides and letting them dry, and cultures were stored in 20% glycerol at −80°C. All stored cultures were tested for viability of the cells at different time intervals of growth (1, 2, and 4 weeks of growth) and storage (1, 2, 8, and 24 weeks of storage).

Carbon sources.

Isolates were tested for growth on M2 medium with methanol (0.5%, vol/vol), acetate (0.1%, wt/vol), and formate (0.1%, wt/vol) both in liquid medium and on agarose plates. Cultures were incubated for 1 month and monitored daily. Methane consumption and growth by optical density (OD₆₀₀) were monitored. Final OD₆₀₀ values were between 0.15 and 0.3.

Nitrogen fixation.

Cultures were tested for growth and methane consumption on nitrogen-free M2 medium with various headspace oxygen concentrations (20, 10, 5, 2, and 1%) and a methane concentration of 1%. Upon growth on nitrogen-free medium with methane as the energy source, the cultures were transferred to nitrogen-free medium with 0.1% methanol, and the nitrogenase activity assay was performed by adding 2% of acetylene and measuring ethylene production as described previously (41).

Analytical techniques.

Methane was measured on a Hewlett-Packard model 5890 gas chromatograph equipped with a flame ionization detector (FID) and a Porapak Q column (80/100 mesh). Turbidity as a measure of growth was analyzed on an Ultraspect K spectrophotometer at 600 nm.

Phospholipid fatty acid determination.

The isolates were grown in 500-ml bottles with 200 to 250 ml medium M2, initial pH 4.5, and 5 to 10% methane in the headspace. Bottles were incubated at room temperature and were shaken at 75 to 90 rpm. Cells were collected by centrifugation for 10 min at 16,264 × g.

Total lipids were obtained from 0.5 to 5 mg freeze-dried cell material using a modified Bligh and Dyer extraction procedure (5, 9), after which it was separated into three increasingly polar fractions over an activated silicic acid column using chloroform, acetone, and methanol, respectively. The methanol fraction, containing the phospholipids, was subjected to mild alkaline methanolysis, after which an authentic C_19:0 fatty acid standard was added. The fatty acids, released from the phospholipids, were methylated using diazomethane. An aliquot of this fraction was used for dimethyl disulfide (DMDS) adduction, as described by Nichols et al. (53), to determine double-bond positions of the monounsaturated fatty acids. Quantification was performed on a gas chromatograph (HP 6890) equipped with a flame ionization detector (FID) set at constant pressure (100 kPa) using relative retention times. A fused silica column (inside diameter [i.d.], 50 m by 0.32 mm; film thickness, 0.1 μm) coated with CP Sil-5CB was used, with helium as a carrier gas. Extracts were injected on-column at 70°C. The temperature increased at 20°C/min to 130°C and 4°C/min to 320 °C, followed by an isothermal hold for 20 min. Components were identified using a gas chromatograph-mass spectrometer (Thermo Trace GC Ultra).

Sample preparation for transmission electron microscopy (TEM).

For strains M5, M175, M242, and M200, actively growing cells were cryofixed by high-pressure freezing, freeze substituted in anhydrous acetone containing 2% osmium tetroxide, 0.2% uranyl acetate, and 1% H₂O or anhydrous acetone containing 2% osmium tetroxide, embedded in Epon resin, sectioned, poststained, and imaged as described previously (73). Additionally, strain M5 was imaged by negative staining. For negative staining, cells were first adsorbed to carbon-Formvar-coated grids (copper, 100 mesh, hexagonal) for 10 min. After incubation, excess liquid was removed with filter paper, and cells were negatively stained by incubation for 5 min on a drop of 1.8% methyl cellulose containing 0.4% aqueous uranyl acetate on ice, after which they were air dried and imaged with a transmission electron microscope (Tecnai 12; FEI Company, Eindhoven, Netherlands). Images were recorded using a charge-coupled-device (CCD) camera (MegaView II; AnalySis).

For strain 29, actively growing cells were collected by centrifugation, embedded in 1% water agar, and fixed in 1% OsO₄-2% glutaraldehyde in 50 mM cacodylate buffer (pH 6.5) for 1 h at 4°C. After dehydration in an ethanol series, the samples were embedded in Spurr epoxy resin. Thin sections were cut on a Sorvall MT-5000 Ultra microtome, stained with 2% (wt/vol) uranyl acetate in water, and then poststained with lead citrate (38). The specimen samples were examined with a JEOL JEM 100 CX-II transmission electron microscope. Cryoscanning electron microscopy (cryo-SEM) was performed on active batch cultures to examine cell morphology. A stub with a droplet of culture was frozen in liquid nitrogen. The sample was transferred in a transfer holder under vacuum at liquid nitrogen temperature to the cold stage at −95°C into a cryopreparation chamber, CT 1500 HF (Oxford Instruments, High Wycomb, United Kingdom). The specimen was sputter coated with 5 nm Pt, conveyed under high vacuum to the cold stage of a scanning electron microscope equipped with a cold-field emission electron gun (JSM 6300F; JEOL, Tokyo, Japan), analyzed, and recorded at −180°C using a 5-kV accelerating voltage (74).

Nucleotide sequence accession numbers.

The sequences of the 16S rRNA genes, the pmoA genes, the pxmA genes, and the mmoX gene of strain M5 and strain M200 have been deposited in the GenBank database under accession numbers HM564015 to HM564021. The 16S rRNA genes, the pmoA genes, and the mmoX gene of strain 29 have been deposited in GenBank under accession numbers DQ076754 to DQ076756, and those of strains M162, M167, M169, M175, M212, and M242 have been deposited in GenBank under accession numbers JN036511 to JN036528.

RESULTS AND DISCUSSION

Methane oxidation rates and methanotrophic community analysis.

Sphagnum mosses from the Mariapeel and the Hatertse Vennen showed high initial methane oxidation rates of around 40 to 60 μmol CH₄ per day per g (dry weight), and the mosses of the Haute Fagnes showed relatively low initial methane oxidation rates of 2 to 3 μmol CH₄ per day per g (dry weight), which are both within the range reported previously (44, 57). Peat water controls showed no activity confirming the presence of methanotrophic bacteria inside and/or attached to the Sphagnum mosses. DNA was isolated from the mosses from both peat bogs. A pmoA-based PCR was performed using a combination of general primers, and these products were used for microarray hybridization (8). Since the Haute Fagnes sample did not yield sufficient DNA, only the Mariapeel sample was used for the pmoA microarray analyses. The pmoA gene, encoding the 27-kDa subunit of the membrane-bound methane monooxygenase (pMMO), is generally used as a genetic marker for methanotrophs. All known methanotrophs contain this gene, except Methylocella and Methyloferula spp., which possess only a cytoplasmic soluble methane monooxygenase (sMMO) (21, 25, 28, 69, 77). The pmoA diversity within DNA extracted from the Sphagnum mosses was very high (Fig. 1) compared to that found in other studies using the same microarray on peat soils (15), alpine meadow soils (1), upland soils (14), and rice fields (76) but was comparable to what was found in Sphagnum mosses from Sweden, Canada, or Patagonia (44).

Fig. 1.

Results for the pmoA-based microbial methanotrophic community analysis microarray of two Sphagnum moss samples from the Mariapeel. The color-coding bar on the left represents the achievable signal for an individual probe (1 indicates maximum signal obtained, 0.1 indicates 10% signal, i.e., only 10% hybridization to that probe, and 0 indicates no signal). Mp = Mariapeel; NL = The Netherlands. Mp1.1A and Mp1.2 are two samples taken in the same peat bog. A list of probes in their respective order is given in Table S2 in the supplemental material.

The microarray results revealed the presence of both type I and type II methanotrophs. The type II probes showed that both Methylocystis spp. and Methylosinus spp. were present in the Sphagnum mosses. Furthermore, a probe targeting a group of uncultivated type II methanotrophs (I.R. McDonald and J. C. Murrell, unpublished data) from a peatland, Peat264, hybridized with the PCR product. Surprisingly, the microarray also showed a strong signal with type Ia probes, which target the gammaproteobacterial genera Methylomonas, Methylobacter, and Methylomicrobium. A broad diversity of Methylomonas and Methylobacter species was detected. Probes based on clone sequences obtained from the soil-water interface of rice fields (SWI1-375, SWI1-377) (30) also showed a strong signal. The type Ib-specific probes, which target the thermotolerant and thermophilic gammaproteobacterial genera Methylococcus, Methylothermus, and Methylocaldum and related, uncultivated clades, showed a weaker signal but still indicated a diverse type Ib community, especially the Methylocaldum probes. Type II methanotrophs are commonly found in peat and other ecosystems, and several acidophilic methanotrophs have been isolated (19). Comparable to our results, other studies using the same microarray (15, 16) showed the abundance of Methylocystis species in peat soils. The presence of type I methanotrophs in peatlands has been reported previously (15, 22, 49), but it is new that such a predominance of type I methanotrophs is observed in these ecosystems. Methylocella spp., which are commonly found in northern peatlands (21, 25, 59), and the recently discovered Methyloferula species (77), are the only known methanotrophs that do not have a pmoA gene and are therefore were not detected in this microarray. In addition, the PCR primers used are not able to amplify verrucomicrobial pmoA genes. Nevertheless, the microarray results indicated that many different methanotrophic genera are present and might contribute to Sphagnum-associated methane oxidation, although thus far, no isolates from these mosses have been reported. Therefore, Sphagnum mosses were enriched with methane in order to isolate methanotrophs.

Enrichment and isolation of methanotrophs.

The microarray indicated the presence of both Alpha- and Gammaproteobacteria in Sphagnum mosses. M2 medium is a strongly diluted medium which was previously used to successfully isolate alphaproteobacterial methanotrophs from peat (26, 27) but seems to be less suited to isolate gammaproteobacterial methanotrophs. Several attempts have been described in the literature to enrich methanotrophs by modifying the medium composition (13, 38, 82). Another widely used and modified medium is Whittenbury's medium (80), but acidophilic methanotrophs are not able to grow on this medium (26, 27). To increase the chance of successful isolation of both type I and type II methanotrophs, a new medium was designed (medium N) based on the water composition of the Mariapeel and the M2 medium (Table 1). Thoroughly washed Sphagnum mosses were incubated in both medium M2 and medium N supplemented with methane to enrich the endophytic and epiphytic methanotrophs. The enrichments resulted in increasing methane-oxidizing activity, but the liquid medium always stayed clear and transparent, indicating an enrichment of methanotrophs inside the mosses. Thereafter, Sphagnum mosses were crushed with a mortar and used for dilution series. The highest dilutions consuming methane were transferred to medium M2- and medium N-agarose plates and subsequently streaked until single colonies and pure cultures were obtained. The colonies growing on medium M2-agarose were transferred to liquid culture to test methane-oxidizing activity and were used for colony PCR. Some of the colonies did not show methane-oxidizing activity and were identified as Burkholderiales and Pseudomonaceae spp. (see Fig. S1 in the supplemental material), known contaminants of methanotrophic enrichments and common in peat ecosystems (4, 60). Seven pure methanotrophic cultures, showing high methane-oxidizing activity in liquid medium M2, were obtained and designated strains 29, M162, M167, M169, M175, M212, and M242. Strain M242 originated from a Belgian peat ecosystem, strain M169 from the Hatertse Vennen, and the others from the Mariapeel. All cultures showed white or creamy colonies on agarose plates.

Table 1.

Comparison of the concentrations of the different elements in peat water, medium N, and medium M2

Element	Concn (μmol/liter)
	Peat watera	Medium N	Medium M2
PO43−	1.24	10	293.92
NO3−	11.60	10	1,978.04
NH4+	51.88	50	0
Na+	237.02	220	0.02
K+	40.45	20	2,271.97
Cl−	343.39	426.09	54.96
Ca2+	124.60	120	27.21
Mg2+	39.75	40	81.14
Mn2+	0.60	0.30	0.03
Fe2+	14.55	14.39	1.44
Si2+	20.96	20	0
Zn2+	1.01	0.7	0.07
Al3+	6.47	10	0
Cu2+	ND	0.96	0.09
Co2+	ND	1.68	0.17
Ni2+	ND	0.17	0.02
SO42−	55.06	56.04	82.65
Mo2+	ND	0.25	0.02

^aND = not determined.

Mosses were also incubated with methane in peat water at different pH values (from 2.5 to 5.0). Methane-oxidizing activity was measured in all enrichments except at pH 2.5. The enrichment at pH 3.3 was further analyzed and purified on peat water-Gelrite plates. The methanotrophic culture obtained in this way showed 99% homology to Methylocystis sp. strain H2s (3), based on 16S rRNA gene analysis. This microorganism was previously isolated from peat and was reported to have a pH optimum between 6 and 6.5 (3). Upon further testing, our methanotrophic culture did result in better growth at higher pH and showed the same optimum, but the culture could also be maintained at initial-medium pH 3.3.

The enrichment and isolation from Sphagnum mosses on newly designed medium N resulted in different methane-oxidizing cultures. Since the methane-oxidizing culture enriched from the Mariapeel Sphagnum moss consisted of at least two different species, a dilution series was made on medium N-agarose plates. 16S rRNA gene analysis revealed the enrichment culture to consist of a methanotrophic Gammaproteobacterium species showing highest homology to Methylomonas species and a nonmethanotrophic Betaproteobacterium species that showed highest homology to Burkholderia species, which are commonly found in Sphagnum bogs (4). The methanotrophic culture was designated strain M5. Apparently gammaproteobacterial methanotrophs were able to outcompete the alphaproteobacterial methanotrophs on medium N. For further cultivation of strain M5, medium M2 was used since it resulted in better growth. Medium M2 contains more nitrate, phosphate, and potassium and less trace elements, sodium, chloride, and calcium than medium N (Table 1). On agarose plates, 1-week-old colonies of strain M5 had a mucous-like appearance. The methanotrophic culture originating from the Haute Fagnes Sphagnum enrichment on medium N showed pink slimy colonies on agarose plates and methane-oxidizing pink flocs in liquid culture. This culture was designated M200, and phylogenetic analysis showed this isolate also belonged to the Gammaproteobacteria of the Methylococcaceae family.

Strain purity of all pure isolates was confirmed by repetitive plating on selective media and (transmission electron) microscopy. No growth occurred on LB, 10-fold diluted LB, or M2 supplemented with 0.1% glucose agar plates.

Phylogenetic analysis.

All isolates were shown to contain a pmoA gene, strain M5 and M200 also contained the putative pmoA gene, pxmA (68), and all strains except strain M200 also possessed a mmoX gene. The mmoX genes of these strains were obtained with the primer set mmoXA-mmoXB (2). No mmoX gene could be amplified with DNA from strain M200 as a template, with all possible combinations of the general primer sets tested.

Phylogenetic affiliation based on the 16S rRNA genes, pmoA genes, pxmA genes, and mmoX genes showed that strains 29, M242, M162, M167, M169, M175, and M212 all belonged to the Alphaproteobacteria and that both strains M5 and M200 belonged to the Gammaproteobacteria (Fig. 2, 3, and 4). Strains 29 and M242 were most related to Methylosinus spp., and the 16S rRNA gene, pmoA, and mmoX sequences of strains M162, M167, M175, M169, and M212 were 100% identical to those of Methylocystis sp. strain H2s (GenBank accession no. FN422003) and Methylocystis sp. strain F10V2a (GenBank accession no. AJ458504), and only the mmoX sequence of strain M169 was 98 to 99% identical to those of the two above-mentioned strains. This indicated that these strains were very similar and genetically identical to Methylocystis sp. strain F10V2a and to Methylocystis strain H2s isolated from a peat bog, which was shown to be able to use acetate for survival (3).

Fig. 2.

16S rRNA gene neighbor-joining tree showing the relationship of all the isolated strains to selected methanotrophs (>500 bootstrap replicates). Bootstrap values of >60% are indicated at the nodes of the branches with black dots. The scale bar represents the number of base substitutions per site.

Fig. 3.

pmoA and pxmA gene neighbor-joining tree based on DNA showing the relationship of all the isolated strains to selected methanotrophs (>500 bootstrap replicates). Bootstrap values of >60% are indicated at the nodes of the branches with black dots. The scale bar represents the number of base substitutions per site. *, GenBank accession no. EF591086, EF591085, FJ462788, FJ462789, FJ462791, and CP000975; **, GenBank accession no. EF591087, FJ462790, and CP000975.

Fig. 4.

mmoX gene neighbor-joining tree based on DNA showing the relationship of all the isolated strains (>500 bootstrap replicates). Bootstrap values of >60% are indicated at the nodes of the branches with black dots. The scale bar represents the number of base substitutions per site.

Compared to each other, the 16S rRNA and pmoA genes of strains 29 and M242 were 97 to 98% identical. Based on all the genes, strain 29 showed the highest similarity (96 to 99%) to Methylosinus sp. strain 8 (GenBank accession no. AJ458486) and Methylosinus sp. strain NCIMB 11126 (GenBank accession no. Y18946). Strain M242 showed the highest similarity to Methylosinus sp. strain SE2 (GenBank accession no. AJ458478) and Methylosinus sporium 8 (GenBank accession no. AJ459018). The mmoX sequences of strains 29 and M242 are 99.6% identical, and they both showed highest similarities to Methylosinus sp. strain LW3 (GenBank accession no. AY007287) and Methylosinus sp. strain PW1 (GenBank accession no. AY007292), both 96 to 97% on the DNA level.

The 16S rRNA gene sequence of strain M5 showed 95% similarity to Methylomonas strains LW21 (GenBank accession no. AF150800), LW19 (GenBank accession no. AF150798), and LC1 (GenBank accession no. DQ119049) and to Methylomonas scandinavica (GenBank accession no. AJ131369). In the neighbor-joining tree (Fig. 2), strain M5 clustered with the genus Methylomonas, and this topology was confirmed with minimum evolution and maximum parsimony trees. Based on pmoA, pxmA, and mmoX genes, strain M5 was also most similar to Methylomonas species (Fig. 3 and 4). These genes showed 80 to 87% sequence homology to different Methylomonas species. The pmoA and mmoX genes were 84 to 87% homologous to Methylomonas sp. strain LC1 (GenBank accession no. DQ119046) and other Methylomonas strains. On the amino acid level, the similarities were 91 to 98% to those strains. Interestingly, the pmoA sequence showed homology to the Methylomonas-related clade MHPSr2, representing six clones originating from Sphagnum-covered soils in the Moor House peat reserve, United Kingdom (15). The pmoA sequence of strain M5 was 88 and 92% identical to the MHPSr2 clones on the DNA and protein levels, respectively, and together formed a separate clade, with high bootstrap value, within the Methylomonas genus. This shows that similar Methylomonas spp. have been detected in other Sphagnum-dominated peatlands. The pxmA gene of strain M5 is related to the pxmA gene of Methylobacter marinus (81%; GenBank accession no. EU722430), Methylomonas methanica strain S1 (81%; GenBank accession no. EU722433) and Methylomonas sp. strain LW13 (79%; GenBank accession no. EU722432). On an amino acid level, this homology was 94 to 96%. However, so far the pxmA gene data set is not extensive (68).

Strain M200 showed, based on the 16S rRNA gene, highest homology to many uncultured bacteria from different ecosystems among many wetlands. The 16S rRNA of strain M200 showed 92 to 93% homology to Methylovulum miyakonense HT12 (GenBank accession no. AB501287), Methylosoma difficile LC2 (GenBank accession no. DQ119050), Methylobacter tundripaludum (GenBank accession no. AJ414655), Methylobacter psychrophilus (GenBank accession no. NR025016), and Methylosarcina quisquiliarum (GenBank accession no. NR025040). Phylogenetic analysis based on neighbor-joining (Fig. 2) and maximum parsimony revealed that strain M200 had homology to the Methylobacter, Methylovulum, and Methylosoma genera. The pmoA gene sequence of strain M200 was clustering between Methylosoma difficile, Methylovulum miyakonense, and Methylobacter spp. (Fig. 3). The closest relatives of the pmoA of strain M200 were Methylovulum miyakonense (89%), Methylosoma difficile (86%; GenBank accession no. DQ119047), and Methylobacter spp. BB5.1 and sp5FB (both 85%; GenBank accession no. AF016982 and AJ868410, respectively). Amino acid sequences revealed the same results, and all above-mentioned relatives showed a 90 to 96% homology. Based on the pxmA gene and PxmA protein sequences of M200, the closest relatives were Methylobacter marinus (GenBank accession no. EU722430) and Methylomicrobium album (GenBank accession no. EU722431), having 71 to 72% homology on the DNA level and 79 to 82% on the amino acid level (Fig. 3).

Physiological properties.

Physiological properties of the different isolates in relation to their original habitat were determined and are compiled in Table 2. Sphagnum-dominated peatlands are low in pH and nutrients and vary in concentrations of methane and oxygen. All isolates showed growth at an acidic pH and at a range of different temperatures (Table 2; see also Fig. S2 in the supplemental material). Sphagnum-dominated peat bogs are limited in all nutrients and also in fixed nitrogen, and the ability to fix nitrogen can be a great advantage in this ecosystem. All strains were able to grow on nitrogen-free medium, but strain 29 showed growth and ethylene production on nitrogen-free M2 medium at 10% oxygen or lower, while strains M5 and M200 showed growth and ethylene production only at 5% oxygen or lower. This higher sensitivity to oxygen on nitrogen-free medium for Gammaproteobacteria than for Alphaproteobacteria has been found previously (51).

Table 2.

Physiological properties of the different isolates

Parameter	Result for indicated straina
	M5	M200	29	M242	M169	M212
Phylogenetic affiliation	Methylomonas	Methylovulum-Methylosoma	Methylosinus	Methylosinus	Methylocystis	Methylocystis
pH range	3.5–6.5 (5.0)	4.1–7.0 (5.5)	3.1–7.0 (5.0–5.5)	ND	ND	ND
Temp range (°C)	10–30 (20)	4–30	10–30 (20–25)	ND	ND	ND
Carbon source
Methanol	+	+	+	+	+	+
Acetate	−	−	−	−	−	−
Formate	−	−	−	−	−	−
Formaldehyde	ND	ND	−	−	−	−
Glucose	−	−	−	−	−	−
Nitrogen fixation
Growth on N-free medium	+	+	+	+	+	+
Ethylene production (%)b	1, 2, 5	1, 2, 5	1, 2, 5, 10	ND	ND	ND
Storage
Spore formation	−	−	+	+	−	−
Glycerol	−	−	+	+	+	+
Heat shock	−	−	+	+	+	+
Freeze-drying	−	−	+	+	−	−
Desiccation	−	−	+	+	−	−

^apH and temperature optima are given in parentheses. ND = not determined.

^bThe oxygen concentration (%) at which ethylene production from acetylene takes place are given.

Transmission electron microscopy (TEM) was used to analyze some of the cultures and showed that all cultures consisted of single cells with cell walls typical of Gram-negative bacteria, and it showed the typical intracytoplasmic membranes (Fig. 5, 6, and 7). Both TEM and phase-contrast microscopy showed all strains to be pure cultures. For all strains, indications of internal cellular storage particles were observed, possibly consisting of polyhydroxy-alkanoates or -butyrates, which are commonly found in methanotrophs.

Fig. 5.

(A) Scanning electron micrograph of actively growing cells of strain 29. Bar, 5 μm. (B) Transmission electron micrograph of a cell of strain 29. Bar, 5 μm. (C) Transmission electron micrograph of a cell of strain M175. (D) Transmission electron micrograph of a cell of strain M242.

Fig. 6.

Transmission electron micrographs of high-pressure frozen, freeze-substituted, and Epon-embedded cells (A, B) and negatively stained cells (C, D) of strain M5. (A) Whole cell with typical intracytoplasmic membranes. (B) Whole cell with a putative storage particle (white). (C) Negative staining of several single cells of strain M5. (D) Negative staining of one/two cells of strain M5. The putative flagellum is visible at the right side.

Fig. 7.

Transmission electron micrographs of high-pressure frozen, freeze-substituted, and Epon-embedded cells of strain M200. (A) Whole cell with typical intracytoplasmic membranes. (B) Whole cell with a putative storage particle (white). (C) Cell periphery with the Gram-negative cell wall and the slime layer around the cell. (D) Several single cells of strain M200.

Cells from both strains 29 and M242 showed the typical sinus shape and were motile. Methylosinus-related strain M242 showed a 70-nm-wide layer outside the cell wall, which could represent a polysaccharide layer or fimbriae. Strains 29, M242, and M175 showed intracytoplasmic membranes in bundles distributed along the periphery of the cell, which is a typical feature of type II Methylocystis/Methylosinus-related methanotrophs (Fig. 5). Both strains 29 and M242 showed the production of spores, visualized with light microscopy, and cell cultures remained viable after all the different storage methods tested (Table 2). Strains M162, M169, and M175 were not resistant to desiccation or freeze-drying, and no cysts were observed during microscopic examination. However, the strains remained viable after heat treatment and storage in glycerol at −80°C.

Cells of both strains M5 and M200 showed intracytoplasmic membranes in bundles distributed throughout the cells, which is a typical feature of type I methanotrophs (Fig. 6 and 7), and both strains showed a polysaccharide-like matrix around them. Negative staining of strain M5 where uranyl acetate had bound to the cells instead of to the grid showed that some cells contained a singular flagellum (Fig. 6D). Neither strain survived any known storage method, indicating there is no production of spores and most probably no cysts production either (64). No spores or cysts were observed with TEM either.

Phospholipid-derived fatty acids.

The proteobacterial methanotrophs possess distinct patterns of phospholipid ester-linked fatty acids (PLFA). Type I methanotrophs contain mainly C₁₆ PLFA, and type II methanotrophs mainly C₁₈ PLFA. Methanotrophic biomarkers include C_16:1ω8 and C_16:1ω5 for type I methanotrophs and C_18:1ω8 for type II methanotrophs. The major fatty acids of strains 29 and M242 were C_16:1ω7, C_18:1ω7, and some C_16:0 and C_18:00, and for strain M5 and strain M200, C_16:1ω8, C_16:1ω7, C_16:1ω6, C_16:1ω5, C_14:0, and C_16:0 were found to be abundant (see Table S1 in the supplemental material).

The abundance of C₁₈ PLFA and absence of most of the C₁₆ PLFA in strains 29 and M242 are in accordance with the type II methanotrophic features. The PFLA distribution of strains 29 and M242 are quite similar to the general PLFA distribution of methanotrophs within the Alphaproteobacteria. Strains 29 and M242 showed the highest similarity to the PLFA distribution of Methylosinus sporium sp. KS17 (5). All Methylosinus sporium strains, as well as strains 29 and M242, lack 18:1ω8c, while Methylosinus trichosporium strains show an abundance of this lipid. This is consistent with the fact that both isolates phylogenetically cluster together with Methylosinus sporium strains (Fig. 2, 3, and 4).

The abundance of C₁₆ PLFA and absence of C₁₈ PLFA in M5 and M200 is in accordance with their other type I features. Both strains M5 and M200 showed highest abundance of C_16:1ω8 and C_16:1ω5, and also C_16:1ω7 and C_16:0 were relatively high. The PLFA distributions of both M5 and M200 were quite distinct from PLFA patterns of other isolated methanotrophs but showed the highest similarity to some Methylobacter, Methylomonas, and Methylomicrobium species. Strain M5 fits within the PLFA pattern of all the different Methylomonas spp., and strain M200 fits best with Methylobacter tundripaludum, Methylomicrobium album, and also Methylosoma difficile, although this genus misses C_16:1ω8 and C_16:1ω5, which are highly abundant in strain M200. Phylogenetic analysis of M200 revealed homology to Methylovulum miyakonense, but based on PFLA, no similarity was found since this genus possesses only C_14:0, C_15:0, and C_16:0 PFLA (38).

Detection in the environment.

As described above, the methanotrophic community of the Mariapeel was investigated using a pmoA microarray. There are no probes on this microarray that specifically detect Methylosoma or Methylovulum spp. Based on its pmoA sequence, strain M5 has a 100% sequence match with the general type Ia probe Ia575, and strain M200 has a 100% sequence match with Ia575 and the Methylobacter-related Mb271 probe. The general type Ia probe Ia193 has one mismatch with strain M5 and three with strain M200. To confirm this, the pmoA PCR products of both strains were used for microarray hybridization, and both indeed showed hybridization with the general methanotrophic probes and the general type Ia probe Ia575, but only strain M5 hybridized with Ia193 (see Fig. S3 in the supplemental material).

Both strains show a hybridization pattern in accordance with the phylogenetic analysis. Strain M5 hybridized with Methylomonas-related probes, and strain M200 hybridized with Methylobacter-related probes. The hybridization with the Mha-500 probes of both isolates is an unspecific cross-hybridization. Almost all of these probes showed hybridization in the microarray of the Mariapeel, indicating that both isolates could have been detected in the Mariapeel Sphagnum.

Final conclusion.

In conclusion, this study showed that only alphaproteobacterial methanotrophs could be isolated using M2 medium (26, 27), while use of the new medium N resulted in pure cultures of gammaproteobacterial methanotrophs. Despite several different isolation techniques resulting in a number of pure cultures, unfortunately, many of the alphaproteobacterial isolates were almost identical. Two new Methylosinus sporium strains were isolated from Sphagnum mosses, strains 29 and M242, which were both able to grow at pH 4.5. Strains 29 and M242 follow the description of the Methylosinus genus as published in Bowman et al. (10, 11) and are most similar to Methylosinus sporium. Several studies showed the presence of Methylosinus spp. in peat ecosystems (15, 23), but so far, only a neutrophilic Methylosinus strain has been isolated from an acidic peat lake (35), and this is the first report of an acidophilic Methylosinus strain and the first Methylosinus strains isolated from Sphagnum mosses.

Strains M172, M169, and M175 have properties similar to those of the genus Methylocystis, as described by Bowman et al. (10, 11), and the Methylocystis strain H2s isolated from peat (1). Methylocystis strain H2s was reported to be moderately acid tolerant, but here we showed it is able to grow down to pH 3. Methylocystis strain H2s was shown to represent 17 to 58% of the microbial population in peat soils of different geographical regions (3), and it is able to use acetate for survival over long periods of time in the absence of methane (18). Surprisingly, our isolates were not able to grow on acetate. In general, Methylocystis spp. can be found in many different ecosystems (18) and have been shown to be abundantly present in peat ecosystems (19, 78). This could explain the large amount of Methylocystis isolates obtained in this study.

The use of the oligotrophic medium N enabled the isolation of two strains that belong to the Gammaproteobacteria. Strain M5 is a new strain belonging to the genus Methylomonas. All 16S rRNA, pmoA, pxmA, and mmoX genes of strain M5 indicated that it is most closely related to the Methylomonas genus, and it follows the genus description as published by Bowman et al (11, 12). Some Methylomonas spp. are able to fix nitrogen, and most have a single polar flagellum. Inclusions of poly-β-hydroxybutyrate have been found in Methylomonas spp., and they produce capsules and surface pellicle and desiccation-sensitive cysts. Methylomonas species have been isolated from various ecosystems like marine ecosystems (46, 64), groundwater (40), soil, mud, sludge, and wastewater (11) but never from peat ecosystems or Sphagnum mosses. Strain M5 was shown to be abundantly present in pmoA amplicons from Sphagnum mosses from the Mariapeel (43).

The 16S rRNA, pmoA, and pxm genes of strain M200 in different tree topologies did not produce a consistent affiliation to a particular genus and will therefore need further investigation. Highest homology to Methylobacter species, Methylovulum miyakonense, and Methylosoma species was found. Some properties of strain M200 are similar to those of the genera Methylovulum (38) and Methylosoma (58), but others resemble the Methylobacter genus as described by Bowman et al (11). Strain M200 is able to fix nitrogen, as are Methylosoma spp., but so far, no Methylobacter and/or Methylovulum spp. are reported to fix nitrogen. The pink strain M200 resembles the pink Methylosoma spp., but none of the known Methylobacter and Methylovulum spp. are reported to produce carotenoids. Both Methylobacter and Methylosoma species (12, 58) are able to grow between pH 5.5 and 9.0, while the pH range of strain M200 is pH 4.1 to 6.5, and Methylovulum miyakonense grows between pH 6.0 and 7.5 (38). No reports in which any of the M200-related species have been detected in peat ecosystems or Sphagnum mosses have been found. Several Methylobacter species have been isolated from environments like polar tundra (54, 71), soda lakes (42), and Artic wetland soil (78) but are also commonly found in soil, sewage, and mud (11). Methylosoma difficile strain LC2 was isolated from littoral sediment of Lake Constance (58), and Methylovulum miyakonense HT12 was recently isolated from forest soil in Japan (38). Strain M200 most probably represents a new strain belonging to the genus Methylosoma or may even represent a novel genus.

Gammaproteobacterial methanotrophs can be found in a broad range of environments and comprise a large variety of methanotrophs, including the moderately thermophilic Methylococcus spp. and Methylocaldum spp (6), thermophilic Methylothermus (7), halophilic Methylohalobius (34), and filamentous bacteria Crenothrix and Clonothrix (66, 75), but no acidophilic gammaproteobacterial methanotrophs have been described, and to the best of our knowledge, strains M5 and M200 are the first acidophilic members of this class.

The isolation and characterization of these new strains expand our knowledge on methanotrophs and can help us to update the microarray and future molecular surveys.