Isolation of Key Methanogens for Global Methane Emission from Rice Paddy Fields: a Novel Isolate Affiliated with the Clone Cluster Rice Cluster I

Abstract

Despite the fact that rice paddy fields (RPFs) are contributing 10 to 25% of global methane emissions, the organisms responsible for methane production in RPFs have remained uncultivated and thus uncharacterized. Here we report the isolation of a methanogen (strain SANAE) belonging to an abundant and ubiquitous group of methanogens called rice cluster I (RC-I) previously identified as an ecologically important microbial component via culture-independent analyses. To enrich the RC-I methanogens from rice paddy samples, we attempted to mimic the in situ conditions of RC-I on the basis of the idea that methanogens in such ecosystems should thrive by receiving low concentrations of substrate (H₂) continuously provided by heterotrophic H₂-producing bacteria. For this purpose, we developed a coculture method using an indirect substrate (propionate) in defined medium and a propionate-oxidizing, H₂-producing syntroph, Syntrophobacter fumaroxidans, as the H₂ supplier. By doing so, we significantly enriched the RC-I methanogens and eventually obtained a methanogen within the RC-I group in pure culture. This is the first report on the isolation of a methanogen within RC-I.

Since their discovery in an Italian rice paddy field (RPF) in 1998 (12), the uncultivated microorganisms in the group RC-I (rice cluster I) have been a focus of attention with regard to potential ecological and biogeochemical importance. To date, the following aspects of RC-I have been elucidated. (i) Cultivation-independent molecular studies combined with stable isotope analysis strongly suggested that members of the RC-I group are most active and play a key role in methane production from RPFs (22). (ii) The members are highly likely to be H₂-utilizing methanogens because abundant populations of RC-I Archaea were observed in enrichment cultures with H₂ (8, 25, 33) and a metagenomic approach showed that an RC-I methanogen had a full set of genes involved in methanogenesis from H₂-CO₂ (9). (iii) The group was found to be a major archaeal component in natural RPFs regardless of geographical locations and seasonal changes, accounting for 20 to 50% of total methanogenic populations (20, 28, 39). (iv) In addition to the occurrence in RPFs, they have also been observed in a wide variety of anoxic environments, such as peatland (1, 33), sediment of freshwater lakes (40), and contaminated aquifer (7, 37), indicating the widespread habitat of RC-I on our planet. Given the molecular ecological results and biogeochemical data on paddy fields (6, 27), the RC-I group has been recognized as one of the most important yet-to-be isolated, and thus uncharacterized, methanogens that may significantly impact the global methane cycle. For these reasons, we set out to isolate a pure culture of an RC-I group methanogen with the hope that its characterization might contribute to the understanding of these important methane sources. Here we report the isolation and initial characterization of such a microbe.

MATERIALS AND METHODS

Environmental samples, microorganisms, and cultivation.

Rice paddy soils were obtained from four places near Nagaoka, Niigata Prefecture, Japan. Methanogenic sludge was collected from a methanogenic municipal sewage treatment plant at Nagaoka, Niigata Prefecture, Japan. Strain SANAE was isolated in this study as a novel hydrogenotrophic methanogenic archaeon. Syntrophobacter fumaroxidans strain MPOB (DSM 10017) was purchased from the Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (Braunschweig, Germany). All cultivations were carried out at 37°C with 50-ml serum vials containing 20 ml of an artificial medium (31) under an atmosphere of N₂-CO₂ (80/20 [vol/vol]) without shaking. The purity of strain SANAE was checked by 16S rRNA gene-based clone analysis with the universal archaeal primer pair Ar109f-1490R (16, 38), terminal restriction fragment length polymorphism based on a method reported by Chin et al. (2), FISH (fluorescence in situ hybridization) with 16S rRNA-targeted oligonucleotide probes for the isolates (see the FISH section below), evaluation by a failure to recover bacterial 16S rRNA gene amplification by PCR with the universal bacterial primer pair EUB338*-1490R (14, 38), and cultivation with the following media at 37 and 55°C: (i) thioglycolate medium (Difco) containing ca. 150 kPa H₂-CO₂ (in the headspace) and 10 mM sulfate; (ii) thioglycolate medium containing 20 mM lactate and 10 mM sulfate; (iii) thioglycolate medium containing 10 mM sucrose, 10 mM glucose, 10 mM cellobiose, and 10 mM xylose; and (iv) AC medium (Difco).

Analysis of 16S rRNA genes.

DNA extraction from environmental samples and enrichment cultures and PCR amplification were performed as described previously (16). PCR amplification was done with the archaeal universal primer pair Ar109f-1490R (16, 38). PCR products were purified with a MinElute purification kit (QIAGEN). Clone libraries were constructed with the TOPO TA cloning kit (Invitrogen). The nearly full-length 16S rRNA gene sequences of isolated strains were amplified with the archaeal primer pair Arch21F-1490R (5, 38). Sequences were determined with a CEQ DTC kit-quick start kit (Beckman Coulter) and a CEQ2000XL (Beckman Coulter) automated DNA sequencer. Phylogenetic analyses were performed with the ARB program (24). The base phylogenetic tree was constructed by the neighbor-joining method implemented in the ARB program with nearly full-length (>1,000 nucleotides) sequences. Partial-length sequences (<1,000 nucleotides) were inserted into the base tree with the parsimony insertion tool of the ARB program to show their approximate positions. Bootstrap resampling analysis (11) for 1,000 replicates was performed with the PAUP* 4.0 package (36) to estimate the confidence of tree topologies.

FISH.

Fixation of cells and whole-cell in situ hybridization were performed as described previously (32). The 16S rRNA-targeted oligonucleotide probe SANAE1136 (5′-GTGTACTCGCCCTCCTCG-3′) for strain SANAE was designed with the PROBE DESIGN tool of the ARB program (24). The probe was labeled with Cy3. The stringency of hybridization was adjusted by adding formamide to the hybridization buffer (25% [vol/vol] for SANAE1136).

Microscopy and chemical analyses.

An Olympus fluorescence microscope was used for studies of cell morphology and epifluorescence (Olympus BX50F). Short-chain fatty acids, methane, H₂, and carbon dioxide were measured as described previously (15-17).

Nucleotide sequence accession numbers.

Sequence data obtained in this study were deposited in the DDBJ/EMBL/GenBank databases under accession numbers AB162774, AB196288, AB236108, and AB243792 to AB243811.

RESULTS

Attempt to cultivate RC-I methanogens with hydrogen (H₂).

First, to choose the appropriate inoculum to cultivate RC-I methanogens, we began with a survey of four RPF soils via 16S rRNA gene-based clone analysis with archaeon-specific PCR primers. Fifty clones were randomly picked from each clone library. While molecular signals of the RC-I group methanogens have been detected from all RPF samples, we selected a sample from Nagaoka, Niigata Prefecture, Japan, which had the highest clonal frequency of RC-I phylotypes (Fig. 1, phylotypes NRP-O [3/50 clones] and NRP-U [1/50 clones]). On the basis of the assumption that RC-I probably is a group of hydrogenotrophic methanogens, as indicated by previous studies, the first enrichment trial involved anaerobic incubation of the soil sample with H₂ (ca. 150 kPa) as an energy source in both liquid and solid media at 37°C. Many methanogens grew in the enrichment cultures after 3 days of incubation, as judged by the methane production with cell growth and F₄₂₀ autofluorescence (data not shown). However, further analysis of these cells via 16S rRNA gene-based cloning analysis showed that not all of the archaeal clones belonged to the RC-I group but some were closely related to previously isolated methanogens such as Methanobacterium bryantii (sequence similarity, 98%; sequence accession no. AB236108).

FIG. 1.

Phylogenetic relationships among the clones obtained in this study, strain SANAE, and RC-I environmental clones inferred from 16S rRNA gene sequence comparisons. The scale bar indicates the number of nucleotide changes per sequence position. The symbols at nodes show bootstrap values obtained after 1,000 resamplings. Accession numbers are also shown after each reference sequence.

Cultivation of an RC-I methanogen by the coculture method.

Our first attempt, described above, was based on a canonical method for cultivation, resulting in the isolation of a nontargeted H₂-utilizing methanogen. One of the conceivable reasons was the difference between in vitro and in situ physicochemical conditions. In particular, we assumed that a high concentration of H₂ in the headspace (for example, 100 kPa) might be fatal because it was approximately 1,000- to 10,000-fold higher than the in situ H₂ concentrations of natural RPFs (4, 10). Most likely, this cultivation condition was not suitable for RC-I members to compete with the growth of previously isolated methanogens. In fact, it is well known that high concentrations of substrates sometimes allow minor in situ but fast-growing in vitro microbes to grow.

If the in vitro condition is similar to the in situ RC-I habitat, would it be possible to cultivate the targeted microbes in the artificial culture medium? Indeed, some of previously uncultured microbes have been isolated in pure cultures only when the substrate was provided at low concentrations close to the natural environment (3, 19, 29). The underlying idea was that RC-I methanogens in RPFs might thrive by receiving low concentrations of substrate (H₂) continuously provided by heterotrophic H₂-producing bacteria (this way of living is commonly referred to as interspecies H₂ transfer) (18, 30). We thus established a coculture method by using the syntrophic propionate-oxidizing H₂-producing bacterium Syntrophobacter fumaroxidans (13) as the supplier of H₂. Bacteria like S. fumaroxidans are called syntrophs and are fermentative bacteria that catalyze the oxidation of a variety of substrates (fatty acids, alcohols, and aromatic compounds) via reactions that are thermodynamically unfavorable unless H₂ is continuously removed (18, 30). Thus, they typically live in syntrophy with H₂-consuming microbes such as methanogens. Therefore, if cultivation is done together with a particular syntroph growing on a substrate that the syntroph can preferably metabolize, it can continuously provide low concentrations of H₂ to methanogens (ca. 10 to 30 Pa in the case of propionate) (17, 21, 34, 35). Thus, for the next challenge, we cultured the soil samples with an anaerobic propionate-oxidizing syntroph, S. fumaroxidans, as the H₂ supplier. Since the doubling growth time of S. fumaroxidans is ca. 5 days (13), as a result, hydrogen, an end product of its metabolism, will be provided to methanogens at a slow rate.

A primary enrichment culture was made from the rice paddy soil with propionate (20 mM) as the sole energy source and the pregrown cells of an S. fumaroxidans culture. After 3 months of incubation of the coculture enrichment with S. fumaroxidans, cell growth, propionate degradation, and methane production were observed (Fig. 2A). The culture was successively transferred to fresh medium every 50 to 80 days (5 to 10% [vol/vol] inoculum). During cultivation, the partial pressure of H₂ in the cultures was kept at less than 30 Pa (Fig. 2A). Microscopic observation after five transfers showed that the community consisted simply of F₄₂₀-autofluorescent methanogen-like cells and Syntrophobacter-like, oval, rod-shaped cells. To identify the methanogens thriving in the syntrophic culture, we constructed an rRNA gene clone library with an archaeal universal primer set. The clone analysis indicated that all 10 of the archaeal sequences examined were affiliated with a single phylotype within the RC-I group (Fig. 1, phylotype NRP-ProM-A). To confirm whether the clonal sequence obtained from the dominant archaeal population in the culture, a specific DNA probe (SANAE1136) was designed and applied to the culture. As a result, the syntrophic culture actually contained SANAE1136 probe-positive rods (Fig. 2D and E). These observations indicated that the microbe possessing the rRNA gene detected as phylotype NRP-ProM-A was the dominant archaeal population in the culture.

FIG. 2.

RC-I methanogens in coculture with syntrophic, propionate-oxidizing bacteria. (A) Stoichiometric conversion of propionate in the enrichment culture containing methanogenic cells of the RC-I group. (B and C) Photomicrographs of the enrichment culture after five transfers obtained by phase-contrast (B) and fluorescence (C) microscopy indicating the presence of F₄₂₀-autofluorescent methanogens in identical fields. (D and E) FISH of RC-I cells. Phase-contrast (D) and fluorescence (E) images of RC-I cells stained with Cy3-labeled 16S rRNA probe SANAE1136. Scale bars represent 10 μm.

Isolation of the RC-I methanogen in pure culture.

To obtain the RC-I methanogen in pure culture from the coculture enrichment, we used the serial-dilution method with both liquid and solid media and different substrates that are generally utilized by H₂-utilizing methanogens, such as H₂ (ca. 150 kPa) and formate (40 mM), with the propionate coculture system as the inoculum. In the H₂ cultures, we observed the growth of the RC-I methanogen that converted H₂ to methane stoichiometrically, which is explained by an equation 4H₂ + CO₂ → CH₄ + 2H₂O (Fig. 3). The growth rate, however, was found to be very slow compared with those of previously characterized methanogens; i.e., the doubling time of RC-I cells with H₂ was 4.2 days (based on the optical density at 600 nm), indicating that isolation requires a long time. After using the serial-dilution methods for over a year, we eventually succeeded in isolating the RC-I methanogen in a pure culture, designated strain SANAE (Fig. 4), by the deep-agar method with H₂ as the substrate.

FIG. 3.

Methane production from H₂ by strain SANAE at 37°C.

FIG. 4.

Phase-contrast micrograph of strain SANAE cells grown on H₂ (ca. 150 kPa) supplemented with acetate (1 mM) and yeast extract (0.01%). Scale bar represents 10 μm.

Strain SANAE had nonmotile, rod-shaped cells. They were 1.8 to 2.4 μm long and 0.3 to 0.6 μm wide. The strain formed brownish, lens-shaped colonies 1 to 1.5 mm in diameter in deep agar with H₂ (ca. 150 kPa in the headspace). In addition to H₂, the isolate can also grow with formate (40 mM). Acetate was not utilized as an energy source.

DISCUSSION

We have developed a new way to isolate methanogens, by using syntrophs that can supply H₂ at very low concentrations (coculture method). As expected, the RC-I methanogen could be cultivated by the coculture method whereas conventional enrichments from the same environmental samples with high levels of H₂ yielded fast-growing methanogens such as Methanobacterium spp. Assuming that strain SANAE is representative of the members of the RC-I group, our results suggest that the RC-I group has a higher affinity for H₂ than other, fast-growing, members and thus may be well adapted to the low-H₂ habitat of RPFs. This speculation was also suggested by two reports on stable-isotope-probing analysis (23, 26). Lu et al. reported that Methanobacteriales and Methanosarcinales incorporated ¹³C when rice root was incubated in a high-H₂ atmosphere in the presence of ¹³CO₂, while ¹³C was preferentially incorporated into RC-I under an N₂ atmosphere in which a low concentration of H₂ was produced by fermentative bacteria from rice root materials (23). According to the stable-isotope-probing analysis with ¹³C-labeled propionate for RPF soil reported by Lueders et al., ¹³C was incorporated not only into rRNA of syntrophic propionate-oxidizing bacteria but also into certain methanogenic archaeal groups, including RC-I (26). These results imply that RC-I methanogens probably live syntrophically with propionate-oxidizing bacteria in RPF environments. Consequently, given that the members of RC-I may have a high affinity for H₂ and live in syntrophy with fermentative bacteria such as syntrophic propionate oxidizers, it can be said that our coculture method is a suitable way to cultivate RC-I methanogens.

Very recently, a complete metagenomic sequence of an RC-I member was reported (9). The sequence was retrieved from a paddy sample incubated at 50°C, indicating that the organism is most likely a thermophilic methanogen. Since the temperature of natural RPFs is normally not stable, it is reasonable that a variety of sequences within the RC-I cluster, which may represent either mesophilic or thermophilic members, have been detected from RPFs. That metagenomic study revealed that the thermophile has a unique set of genes encoding antioxidant enzymes such as catalase and superoxide dismutase. Considering that RPFs receive a seasonal change in redox potential (oxic to anoxic), these unique genes may be indispensable for survival in such an environment.

The 16S rRNA gene sequence similarity between the metagenome and strain SANAE is, however, only 92.0% (Fig. 1, RC-I fosmid 1B6mer2), indicating that the two organisms are phylogenetically distinct at the genus level. Therefore, further studies cross-linking both the metagenomic information of thermophilic RC-I and our mesophilic isolate SANAE are needed to better understand how RC-I methanogens contribute to global methane emission from RPF environments. There are several important questions that have to be addressed to fully understand the role of RC-I methanogens. (i) Can thermophilic RC-I methanogens also grow slowly with a low concentration of H₂? (ii) What is the biochemical or physiological background of the low maximum specific growth rate and the assumed high affinity for hydrogen of RC-I and similar organisms? (iii) What is the difference in enzymatic characteristics, in particular, antioxidant enzymes, between the two different RC-I methanogens? To address these questions, we are now performing further biochemical (including examination of affinity for H₂ in terms of K_m) and genomic analyses of our isolate. We are also trying to isolate the thermophilic RC-I methanogen whose genomic sequence is complete.

Since the coculture method can mimic the natural ecosystem, where H₂ is provided at low concentrations, the method was supposed to have potential for the cultivation of uncharacterized methanogens residing in other ecosystems. To prove this, we also applied the cultivation technique to a methanogenic digester that contains a relatively large amount of uncultured archaeal members within the order Methanomicrobiales. By strategy described above, we successfully isolated a new methanogen, designated strain NOBI-1 (16S rRNA gene sequence accession no. AB162774), which was one of the predominant methanogens in the original ecosystem but was not cultivated so far at the genus or family level. In this case, all of the primary enrichment cultures using a high partial pressure of H₂ with the same inoculum resulted in the cultivation of well-known methanogens like those of the genus Methanobacterium. Detailed information about the enrichment, isolation, and physiological properties of the isolate will be reported in the near future. These findings strongly indicate that the coculture method should work for isolation of fastidious but ecologically important methanogens that would otherwise escape conventional isolation strategies.