INTRODUCTION
Methanogenesis is an important process in the carbon cycle, with a significant impact on global warming. Methane is produced exclusively by methanogenic archaea—strictly anaerobic microorganisms that occur in almost all anoxic habitats on earth, from the marine environment to freshwater sediments to soils, including hot springs and the deep subsurface, in sewage sludge, and in the digestive tracts of animals and humans (
33).
All methanogens belong to the phylum
Euryarchaeota. They presently comprise members of six orders. The basal groups are
Methanopyrales,
Methanococcales, and
Methanobacteriales (class I);
Methanomicrobiales (class II) (
3); and
Methanosarcinales (class III) (
2), with the recently recognized sister group
Methanocellales (
50). It has been hypothesized that the genes for hydrogenotrophic methanogenesis were already present in a common ancestor and were vertically inherited in a broader monophyletic unit encompassing all methanogens (
3). Consequently, it has to be postulated that methanogenesis was lost in the
Archaeoglobales (which fall among class I methanogens), the
Thermoplasmatales, and the
Halobacteriales (which fall between class I and class II) (
3).
In addition, there are many deep-branching lineages of archaea that are exclusively represented by their 16S rRNA genes (
19,
53,
60) and whose properties cannot be safely predicted for lack of any cultivated representatives. One of these lineages is a diverse clade of sequences distantly related to the
Thermoplasmatales. Originally discovered in the marine environment (
9,
17) and the deep subsurface (
59), related clones were subsequently obtained from rice field soil (
20), the water column and sediment of freshwater lakes (
27,
44), and soil and leachate of landfills (
24,
37). Other members of this clade were found in the guts of termites (
16,
42,
54), wood-feeding cockroaches (
22), and scarab beetle larvae (
14). Also, studies of the mammalian digestive tract reported sequences of uncultured archaea distantly related to the
Thermoplasmatales in cattle (
10,
26,
58,
62), sheep (
63), and wallabies (
15) and in the guts and subgingival pockets of humans (
32,
39,
40,
51). Although concrete evidence was lacking, several of these earlier reports had already suggested that such “uncultured
Thermoplasmatales” may represent a novel lineage of methanogens.
The
mcrA gene, which encodes the α-subunit of methyl coenzyme M (methyl-CoM) reductase, has been established as a molecular marker for methanogenic archaea (
36). Studies of the diversity of methanogens in landfill soil yielded several novel
mcrA gene sequences that formed a deep-branching cluster separate from those of the established orders of methanogens (
37). Related sequences were soon discovered in a eutrophic lake (
13) and in salt marsh sediments (
7). Later studies of vertebrate guts also revealed the presence of novel
mcrA genes in the cow rumen (
10); feces of pigs, chickens, and horses (
61); the guts of humans (
39,
51); and the foregut of wallabies (
15).
Kemnitz et al. (
28) observed a correlation between the abundance of rice cluster III (RC-III) archaea and the rate of methanogenesis in enrichment cultures. Mihajlovski et al. (
39) claimed that a new
mcrA phylotype and a new 16S rRNA phylotype obtained from the same stool sample belonged to the same organism and subsequently postulated that they represent a putative new order of methanogens (
40). Also, Evans et al. (
15) had speculated that the unknown
mcrA gene sequences in the foreguts of wallabies and ruminants belong to a lineage of uncultivated archaea encountered in these habitats. However, the final proof for this hypothesis is still lacking.
Previous studies have shown that 16S rRNA and
mcrA genes in the established methanogenic lineages have the same phylogeny (
36,
37). This allows the correlation of unknown
mcrA sequences with the corresponding 16S rRNA gene sequences, a strategy that has been successfully employed to predict the methanogenic nature of the uncultivated archaea in rice cluster I (
36), which eventually led to the enrichment and isolation of
Methanocella paludicola (
50).
In this study, we comprehensively analyzed the phylogeny of all
Thermoplasmatales-related 16S rRNA genes available to date and the unknown
mcrA genes from the respective habitats. To further corroborate the hypothetical congruence of the resulting trees, we obtained additional sequence sets of archaeal 16S rRNA and
mcrA genes from the hindguts of various higher termites and wood-feeding cockroaches, which are known to harbor abundant and diverse populations of uncultured
Thermoplasmatales (
5). In addition, we initiated enrichment cultures from the hindguts of termites and millipedes to isolate a potentially methanogenic member of this novel lineage.
MATERIALS AND METHODS
Termites and cockroaches.
Cubitermes ugandensis and Ophiotermes sp. were collected in Kakamega Forest Reserve (Kenya), and Macrotermes michaelseni was collected near Kajiado (Kenya). Trinervitermes sp. and Alyscotermes trestus originated from the campus of the Jomo Kenyatta University of Agriculture and Technology, Gachororo, Kenya. Only worker caste termites were used for this work. The wood-feeding cockroaches Salganea esakii and Panesthia angustipennis were collected in the vicinity of the Keta Shrine in Ishikawa Prefecture, Japan, by Kiyoto Maekawa, Toyama University. The millipede Anadenobolus sp. was obtained from a commercial breeder (b.t.b.e. Insektenzucht, Schnürpflingen, Germany). All animals were kept in plastic containers at room temperature in the dark.
DNA extraction and purification.
The hindguts of 10 to 20 termites were dissected with sterile fine-tipped forceps, pooled in 2-ml tubes containing 750 μl sodium phosphate buffer (120 mM; pH 8.0), and homogenized. Homogenates of individual cockroach hindguts were prepared in a similar manner. DNA was prepared using a bead-beating protocol combined with phenol-chloroform extraction. The homogenate was transferred to a 2-ml bead-beating vial, and 250 μl sodium dodecyl sulfate (SDS) solution (10% SDS, 0.5 M Tris-HCl, pH 8.0, 0.1 M NaCl), and 0.7 g heat-sterilized zirconia-silica beads (0.1-mm diameter; Carl Roth, Karlsruhe, Germany) was added. Cells were lysed by shaking with a cell disruptor (FastPrep-24; MP Biomedicals, Ilkirch, Germany) for 45 s at a velocity of 6.5 m/s. Cell debris was sedimented by centrifugation at 20,000 × g for 4 min. The supernatant was extracted with 1 volume of phenol-chloroform-isoamyl alcohol (24:24:1 by volume; pH 8.0). After a second centrifugation step, the supernatant was extracted with 1 volume of chloroform-isoamyl alcohol (24:1 [vol/vol]) and centrifuged again in a 2-ml phase lock gel heavy tube (Eppendorf, Hamburg, Germany). The DNA was precipitated by mixing the aqueous phase with 2 volumes of polyethylene glycol (PEG) solution (30% PEG 6000 in 1.6 M NaCl). After centrifugation for 30 min, the pellet was washed with 500 μl ice-cold ethanol (70%) and dried under vacuum. DNA was dissolved in 50 μl elution buffer (MinElute PCR Purification Kit; Qiagen, Hilden, Germany), checked photometrically for purity (Nanodrop; PeqLab, Erlangen, Germany), quantified fluorimetrically (Qubit; Invitrogen, Eugene, OR), and stored at −20°C.
PCR amplification and cloning.
16S rRNA genes were amplified using either the archaeon-specific primer pair Ar109f (5′-AMDGCTCAGTAACACGT-3′) (
25) and Ar912r (5′-CTCCCCCGCCAATTCCTTTA-3′) (
35) or the archaeon-specific primer Ar109f and the prokaryote-specific primer 1490R with the modification of Hatamoto et al. (
23) (5′-GGHTACCTTGTTACGACTT-3′), a combination that yields only archaeal 16S rRNA genes (
43). Each PCR mixture (50 μl) contained reaction buffer, 2.5 mM MgCl
2, 1 U
Taq DNA polymerase (all Invitrogen, Carlsbad, CA), 50 μM deoxynucleoside triphosphate mixture, 0.3 μM each primer, 0.8 mg/ml bovine serum albumin, and 1 μl DNA extract. The PCR program consisted of an initial denaturation step (94°C for 3 min) followed by 32 cycles of denaturation (94°C for 20 s), annealing (52°C for 20 s), and extension (72°C for 50 s) and a final extension step (72°C for 7 min). For the amplification of the
mcrA gene, the primer pair
mcrA-f (5′-GGTGGTGTMGGATTCACACARTAYGCWACAGC-3′) and
mcrA-r (5′-TTCATTGCRTAGTTWGGRTAGTT-3′;
37) was used; the reaction mixture and the PCR protocol were the same as described above, except for the annealing temperature (53.5°C) and the cycle number (35) and a decreased ramp temperature rate of 1°C/s. The PCR products were purified and cloned as described by Schauer et al. (
52).
Sequence analysis.
The 16S rRNA gene sequences obtained in this study were imported into the current Silva database (version 106) (
48;
http://www.arb-silva.de) using the ARB software package (
34). Sequences from other studies that were not included in Silva were retrieved from GenBank (
http://www.ncbi.nlm.nih.gov/). The sequences were automatically aligned, and the alignments were refined manually. A 30% consensus filter was used to exclude highly variable positions. Phylogenetic trees of almost full-length sequences (1,250 bp) were calculated using RAxML, a maximum-likelihood method (
56). Tree topology and node support (100 bootstraps) were tested using the maximum-parsimony method (DNAPARS) implemented in ARB. The
mcrA gene sequences were imported into a seed alignment complemented with sequences of unknown origin that were retrieved from the NCBI database. Trees were calculated at the amino acid level (140 amino acids) using PhyML, a maximum-likelihood method (
21) implemented in ARB. Tree topology and node support (100 bootstraps) were tested using the maximum-parsimony method (PROTPARS) implemented in ARB.
Cultivation.
Enrichment cultures were set up in anoxic, bicarbonate-buffered AM5 medium under an atmosphere of N
2-CO
2 (80:20 [vol/vol]) (
4), but dithiothreitol (DTT) was omitted. The basal medium was supplemented with Casamino Acids (2 g/liter), coenzyme M (10 mg/liter), cysteine (2 mM), and palladium on activated charcoal (10 ml/liter) and (optionally) with yeast extract (2 g/liter) or rumen fluid (10%). The medium (4.5 ml) was dispensed into 15-ml rubber-stoppered glass vials, and hydrogen gas (5 ml) was added to the headspace. Substrates were added from sterile stock solutions (final concentrations): formate (50 mM), methanol (50 mM), acetate (30 mM), or xylan (9 g/liter). Tubes were inoculated (0.5 ml) with gut homogenates of
C. ugandensis or
Anadenobolus sp. prepared in basal medium (1 gut per ml), and the tubes were incubated at 30°C in the dark. The methane content in the headspace was measured every week. The culture headspace (0.2 ml) was sampled with a gas-tight syringe, and the methane content was analyzed using a gas chromatograph with a flame ionization detector (
38).
Quantitative PCR and pyrotag sequencing.
DNA was extracted from the enrichment culture (2 ml) (see above), and the copy numbers of archaeal 16S rRNA genes were determined by quantitative real-time PCR (qPCR) as described by Kemnitz et al. (
28) using the primers A364aF (5′-CGGGGYGCASCAGGCGCGAA-3′) (
6) and A934b (5′-GTGCTCCCCCGCCAATTCCT-3′) (
20). Bacterial 16S rRNA genes were quantified as described by Stubner (
57) using the primer pair 519fc (5′-CAGCMGCCGCGGTAANWC-3′) and 907r (5′-CCGTCAATTCMTTTRAGTT-3′) (
31). In addition, the bacterial community structure of the sample was determined by 454 pyrotag sequencing as described elsewhere (
30).
Nucleotide sequence accession numbers.
The sequences obtained in this study were submitted to GenBank. The accession numbers are JX266062 to JX266091 for 16S rRNA genes and JX266092 to JX266145 for mcrA genes from hindgut homogenates. The accession numbers for the corresponding genes of strains MpT1 and MpM2 are JX266068, JX266097, JX648297, and JX648298.