Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

9.1 Introduction

Methane is an important molecule in the environment. From a societal perspective, it is a clean alternative fuel as well as a greenhouse gas contributing 18 % of total atmospheric radiative forcing (IPCC). From an ecological point of view it is a key intermediate in the global carbon cycle. In the absence of primary electron acceptors such as O2, SO4 2−, and NO3 −, methanogenic archaea couple the oxidation of small molecules like acetate and H2 to the reduction of CO2 or a methyl group to form methane. An estimated 2 Gt of methane is produced each year through anaerobic degradation of organic matter (Thauer et al. 2008). Methane can also be formed by nonmicrobial processes in the deep subsurface and efflux to the atmosphere from seeps and geothermal zones.

Methanogenic environments are therefore usually subsurface environments, either shallow flooded sediments for archaeal methanogenesis or deeper thermal zones for geochemical processes. As methane diffuses upwards from its subsurface origins towards surface oxic zones, it becomes an energy source for methanotrophs. These microbes couple the oxidation of methane to the reduction of O2, SO4 2−, or NO2 −. The SO4 2−-driven methanotrophic consortia are particularly important in ocean seeps and methane hydrate deposits (Knittel and Boetius 2009). The only known NO2 −-using methanotroph belongs to the bacterial candidate phylum NC10 and has been tentatively named “Candidatus Methylomirabilis oxyfera” (Ettwig et al. 2009, 2010). It was demonstrated that this bacterium has the unique ability to produce intracellular O2 through an alternative denitrification pathway (Ettwig et al. 2010). The importance of this process in the environment is currently being investigated. However, of the three physiological types of methanotrophs, the O2-using or aerobic methanotrophs are the best-studied group.

The taxonomic and ecological diversity of aerobic methanotrophs was once thought to be extremely limited. Until recently all known species belonged to the bacterial phylum Proteobacteria, specifically to the classes Gammaproteobacteria and Alphaproteobacteria. It was considered that gammaproteobacterial and alpha-proteobacterial methanotrophs could be easily distinguished on the basis of a few physiological and biochemical characteristics (Hanson and Hanson 1996), and they were therefore often treated as two coherent groups called type I and type II methanotrophs, respectively. However with ongoing study the known physiological and ecological diversity within each group has expanded. There are now 18 described genera of proteobacterial methanotrophs (Table 9.1), and many do not conform to the simple characteristics once used to delineate the two types (Op den Camp et al. 2009; Stein et al. 2012). Even more interesting from a biodiversity perspective is that non-proteobacterial methanotrophs have now also been found. These include some members of the phylum Verrucomicrobia (Dunfield et al. 2007; Pol et al. 2007; Islam et al. 2008) and a member of the candidate division NC10 (Ettwig et al. 2009, 2010).

Table 9.1 Comparison of the three main phylogenetic groups of aerobic methanotrophic bacteria
Full size table

9.2 Methanotrophic Verrucomicrobia

Methanotrophic Verrucomicrobia have to date only been found in geothermal environments. Geothermal gas typically contains 0.1–1 % CH4, with a δ13C signature of −20 ‰ to −30 ‰, strongly indicating a nonmicrobial source (Etiope and Klusman 2002; Giggenbach 1994). This nonmicrobial source may be thermogenic, in which methane is formed from the thermal cracking of buried organic matter, or abiotic, in which H2 and CO2 react to form methane in Fischer-Tropsch-type reactions (Xu 2010). Methane is emitted from geothermal systems via degassing of spring water and diffuse gas seepage from underground reservoirs through soil cover. As in non-geothermal systems this release may be constrained by the activity of methanotrophic bacteria. In 2005, aerobic methane oxidation activity was reported in laboratory incubations of soils from the Solfatara volcano (Italy) with pH levels as low as 1.8 and temperatures as high as 70 °C (Castaldi and Tedesco 2005). A later study of steaming soil in Hell’s Gate (Tikitere), New Zealand (pH 3–5, 60 °C), suggested that methanotrophic bacteria at this site were also actively consuming methane in geothermal gas (Dunfield et al. 2007). The acidic conditions in these environments were too harsh to allow the survival of any known methanotroph.

The microorganisms responsible for methane oxidation in extremely acidic geothermal environments were identified in three independent publications in 2007–2008 (Dunfield et al. 2007; Pol et al. 2007; Islam et al. 2008). Remarkably, they were not alpha- or gammaproteobacteria but rather members of a distinct phylum within the domain Bacteria, the Verrucomicrobia. The three isolates described probably represent a single genus based on the high level of similarity among their 16S ribosomal RNA (rRNA) gene sequences (>98.4 %). There have been difficulties in getting these bacteria preserved in a viable state in public culture collections, and therefore, they are still not taxonomically validated. However, Op den Camp et al. (2009) suggested that they be all considered members of a single genus called “Methylacidiphilum,” with three proposed species: “M. kamchatkense” Kam1, “M. infernorum” V4, and “M. fumariolicum” SolV. The three isolates of “Methylacidiphilum” are remarkably similar (see comparison table in Op den Camp et al. 2009), and they will be treated as a single entity in this chapter.

Hedlund (2010) classifies all 16S rRNA sequences from cultured and uncultured Verrucomicrobia into 7 subphyla or classes, of which only three are named: Verrucomicrobiae, Opitutae, and “Spartobacteria.” “Methylacidiphilum” is the only cultured representative of subphylum 6 (Fig. 9.1). Most sequences in this group have been recovered from geothermal habitats, although a few are from soils or invertebrate guts (Hedlund 2010). It is unknown whether other bacteria in subphylum 6 are methanotrophic.

Fig. 9.1
figure 00091

Neighbor-joining 16S rRNA gene-based phylogeny showing 6 of the 7 main subdivisions of Verrucomicrobia, including “Methylacidiphilum” in subdivision 6, in relation to other phyla of the domain Bacteria. The other aerobic methanotrophs within the phylum Proteobacteria are included for reference. Node support values are based on 500 bootstrap replicates. Phylogenetic analyses were conducted in MEGA4 (Tamura et al. 2007). The scale bar represents 0.05 changes per nucleotide position

Full size image

Published physiological studies of “Methylacidiphilum” are still quite few. However, a complete genome is available for strain V4 (Hou et al. 2008) and draft genomes for SolV (Khadem et al. 2012a) and Kam 1 (unpublished). These have allowed detailed comparisons of metabolic potential and evolutionary histories to be made to model proteobacterial methanotrophs, for which several genomes are now available. An interesting element of the genome analysis of strain V4 by Hou et al. (2008) was a correspondence analysis of its gene content compared to 58 other bacteria. Orthologous groups were defined based on the eggNOG database. This analysis suggested that the gene content of V4 was more similar to species of Proteobacteria than to other Verrucomicrobia. This may indicate a history of promiscuous lateral gene transfer (LGT). However, given that the analysis was more a comparison of metabolic lifestyle than a phylogenetic analysis, this conclusion is not certain. Based on the closest BLAST hits, the JGI IMG site for strain V4 predicts 970 genes (38.5 %) as putatively laterally transferred. Again, this is likely to some extent an artifact of the few verrucomicrobial genomes in the database (9), but it is also suggestive of a fluid genome. Alien Hunter, a program that uses nucleotide dimer to octamer frequencies to identify large islands of LGT (Vernikos and Parkhill 2006), identifies 29 regions of the V4 genome as putatively foreign DNA, summing to a total of 270,000 nt (11.80 % of the genome). The largest region is 32,500 nt. This program is conservative in that it can only identify large regions of LGT (5,000 nt windows), as only these provide enough data to estimate octamer frequencies.

A close phylogenetic examination of key methane metabolism genes shows some evidence for lateral transfer. Formate dehydrogenase and formate hydrogen lyase genes associate most closely with genes from Proteobacteria and Aquificae. Several genes involved in nitrogen metabolism, which are intimately linked to methane oxidation (Stein et al. 2012), show phylogenetic evidence of lateral transfer as well. These include the nitric oxide reductase-encoding genes norB and norC, which form a phylogenetic cluster together with sequences from Aquificae and Thermus/Deinococcus, nested within a large group of genes from Proteobacteria (Fig. 9.2). The hydroxylamine oxidoreductase-encoding haoA was possibly obtained via LGT from Proteobacteria (see the phylogeny in Klotz et al. 2008). Nitrogen fixation genes such as nifHDK are most closely related to genes from Proteobacteria as well (Khadem et al. 2010), and the entire gene set encoding nitrogen fixation is likely to have been acquired via lateral transfer. Methylacidiphilum is also highly unusual in possessing two distinct H+ pumping F1F0 ATPases, one of which appears to be ancestral to Verrucomicrobia and the other obtained from a gammaproteobacterium (Hou et al. 2008). We present more evidence of LGT later.

Fig. 9.2
figure 00092

Neighbor-joining phylogenetic tree of 166 derived amino acid sequences (length 474) of nitric oxide reductase subunit B (NorB). The tree was constructed using the ARB software environment (Ludwig et al. 2004). Node support values are based on 5,000 bootstrap replicates. The scale bar represents 0.1 changes per amino acid position

Full size image

A unique acidic and thermostable bacterial hemoglobin has been characterized from “Methylacidiphilum infernorum,” named Hell’s Gate globin I (HGbl). HGbl is unusual due to an extremely high resistance to oxidation at low pH and high temperature. It also has high structural resemblance to the vertebrate neuroglobins, providing a strong evolutionary link between the bacterial flavohemoglobins and eukaryotic neuroglobins (Teh et al. 2011).

9.3 Physiological Comparison with Proteobacterial Methanotrophs

Species of proteobacterial methanotrophs differ from each other in their dominant phospholipid fatty acids, type of methane monooxygenase (MMO) present, nitrogen fixation ability, geometric arrangement of intracellular membranes, ecological tolerances (to temperature, salt, and acid stresses), genetic systems available to deal with nitrification stress, and carbon assimilation pathways. C-assimilation pathways include the ribulose monophosphate pathway (RuMP), the serine cycle, and the Calvin-Benson-Bassham cycle (Hanson and Hanson 1996). The characteristics that most distinguish the verrucomicrobial genus “Methylacidiphilum” from its proteobacterial counterparts are an extremely acidophilic phenotype, a prevalence of saturated fatty acids, a general lack of intracytoplasmic membrane stacks but presence of other inclusion bodies, at least as deduced from thin sections of chemically fixed cells, and an inability to fix C at the level of formaldehyde. Most of the other properties of “Methylacidiphilum,” particularly its central enzymatic machinery for processing methane, are common to several proteobacterial methanotrophs.

Most proteobacterial methanotrophs are neutrophilic (pH range 5.5–8.5) and mesophilic (temperature range 20–40 °C), although some moderately acidophilic and some thermophilic isolates have been identified (Dunfield 2009) (Table 9.1). The alphaproteobacteria Methylocapsa and Methylocella are mild acidophiles growing between pH 4.2 and 7.2. These genera, along with the alphaproteobacterium Methylocystis and occasionally some gammaproteobacterial methanotrophs like Methylomonas, are commonly detected via cultivation-independent methods in acidic peats at pH 3.5–5.0 (Dunfield and Dedysh 2010; Kip et al. 2011). However, the lower limit of pH 0.8 for “Methylacidiphilum” is well below this range. The upper temperature limit for growth of “Methylacidiphilum” falls just short of that reported (67 °C) for the most thermophilic proteobacterial methanotroph, Methylothermus thermalis (Tsubota et al. 2005).

Proteobacterial methanotrophs contain mainly 14C, 16C, and 18C phospholipid fatty acids (PLFAs), usually monounsaturated. Some are signature compounds found exclusively in proteobacterial methanotrophs and not in any other known organism: C16:1ω8c and C16:1ω5t (usually in the gammaproteobacterial methanotrophs) and C18:1ω8c (usually in the alphaproteobacterial methanotrophs) (Bodelier et al. 2009). The PLFA profiles of “Methylacidiphilum” are very different and show distinctive characteristics of extreme acidophiles. They contain predominantly (>96 %) saturated fatty acids i14:0, a15:0, and 18:0, consistent with the minimization of membrane permeability to protons (Op den Camp et al. 2009). This is in contrast to even the moderately acidophilic proteobacterial methanotrophs, which contain predominantly unsaturated PLFAs. However, the bacteriohopanepolyol compounds in “Methylacidiphilum” are not unique compared to other methanotrophs. The main bacteriohopanepolyol is aminotriol, which is common to many other bacteria, including Methylocella palustris (van Winden et al. 2012).

Most proteobacterial methanotrophs oxidize methane via a membrane-bound particulate methane monooxygenase (pMMO). In order to increase the membrane surface area to hold this enzyme, they produce internal cytoplasmic membrane stacks. Although “Methylacidiphilum” also possesses pMMO, it does not typically display internal membranes. Internal membrane systems were sometimes observed in strain V4 but were not reported in SolV or Kam1. Instead, multiple electron dense inclusion bodies are seen (Fig. 9.3). Although these internal compartments were originally proposed to be carboxysomes, Khadem et al. (2011) determined that the ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO) in the SolV genome is a non-carboxysomal type and that carboxysome shell proteins are missing. The exact structure and function of these inclusion bodies has been difficult to resolve as chemical fixation of these bacteria lacks resolution, a phenomenon also seen with members of the Planctomycetes. Whether the cells of these strains possess a cell plan like many non-methanotrophic verrucomicrobia and planctomycetes including a major intracytoplasmic membrane also will depend on future application of cryotechniques alternative to chemical fixation.

Fig. 9.3
figure 00093

Transmission electron micrograph of a thin-sectioned chemically fixed cell of “Methylacidiphilum infernorum” V4 showing a large putative periplasmic space and internal inclusion bodies. Scale bar, 100 nm

Full size image

9.4 Methane Metabolism

Most methanotrophs are highly specialized. They are incapable of growing on substrates with C-C bonds but rather survive solely on methane, methanol, and a few other 1-carbon compounds (Stein et al. 2012). There are exceptions, notably the facultative methanotroph Methylocella (Dedysh et al. 2005). When strain V4 was tested on alternative growth substrates used by Methylocella, none supported growth, suggesting it is an obligate methylotroph like most known species (Dunfield et al. 2007). However, genome data provides interesting clues to a possible facultative metabolism in strain SolV. There is a hydrogenase gene cluster in the draft genome (Khadem et al. 2012a) and the strain was shown to oxidize H2, although it did not support growth when used as a sole substrate (Pol et al. 2007). Interestingly, the draft genome of SolV also contains acetate kinase and acetyl-coenzyme A synthase (Khadem et al. 2012a), which in theory might confer the ability to grow on acetate. The genome of V4 also contains acetate kinase, but no acetyl-coenzyme A synthase or other phosphotransacetylase could be identified (Hou et al. 2008).

Methanotrophs convert methane to carbon dioxide in the overall reaction: CH4 + 2O2 → CO2 + 2H2O. The first step is catalyzed by the enzyme methane monooxygenase (MMO). In this step, one atom of oxygen is incorporated into methane to generate methanol (CH3OH). There are two forms of the methane monooxygenase enzyme, a soluble form (sMMO) and particulate form (pMMO). The pMMO form is an enzyme of the copper-containing membrane-bound monooxygenase superfamily that also includes ammonia monooxygenase (AMO). pMMO is universally found in all methanotrophic genera with the exception of Methylocella, while the sMMO is found in a few genera. Multiple copies of pMMO genes were detected in each “Methylacidiphilum” strain, but no genes for sMMO. Each gene pmoC, pmoA, and pmoB is a distant homologue of its counterpart in proteobacteria, and a similar pmoCAB operon structure as in most proteobacterial operons was observed. All “Methylacidiphilum” isolates contain three complete pmoCAB operons, with V4 possessing an orphan fourth pmoC copy and Kam1 an additional pmoCA operon (Op den Camp et al. 2009). Interestingly, within each strain the three orthologous operons differ by as much as 50 % at the amino acid level. The most divergent of the three operons is that dubbed pmoCAB3 (as opposed to pmoCAB1 and pmoCAB2). It has been speculated that each ortholog may have evolved different kinetic properties and thus is differentially expressed under distinct environmental conditions. Baani and Liesack (2008) demonstrated a similar phenomenon in Methylocystis sp. strain SC2. This isolate has two copies of the pmoCAB operon that exhibit a low degree of amino acid sequence similarity (59.3–65.6 %) between the respective pmoA, pmoB, and pmoC genes of each operon. The two operons have different kinetic affinities to methane and are differentially expressed in response to methane concentration (Baani and Liesack 2008).

Recently, RNA-seq transcriptome analysis of “M. fumariolicum” SolV showed that the pmoCAB2 operon was highly expressed in cells growing at μmax with excess oxygen and ammonium (Khadem et al. 2012b). The other two pmoCAB operons were hardly expressed under these culture conditions. The expression pattern of cells from CH4-limited, N2-fixing, and O2-limited chemostat cultures (dO2 of 0.5 % and 0.03 % oxygen saturation, respectively) was remarkably different. Under these conditions, the pmoCAB2 operon was downregulated 40 times, while the pmoCAB1 operon was highly upregulated. The pmoCAB3 operon was hardly expressed in these cells. The results point to a regulation of the pmoCAB1/pmoCAB2 expression by oxygen concentration. In addition, qPCR analysis of the four pmoA genes of “Methylacidiphilum kamchatkense” Kam1 showed that the pmoA2 gene was 35-fold more strongly expressed than the other copies during batch cultivation (Erikstad et al. 2012). Suboptimal pH and temperature during growth did not change this pattern, but growth on methanol resulted in a tenfold decreased expression of pmoA2.

9.5 Evolutionary History of pmo

Phylogenetic analysis of all three pmoA gene copies relative to pmoA/amoA genes of other prokaryotes revealed that all pmoA genes from Verrucomicrobia fell into a monophyletic cluster, suggesting that the three orthologues are derived via lineage-specific duplication rather than lateral gene transfer (Dunfield et al. 2007) (Fig. 9.4). Phylogenetic analysis of pmoB and pmoC genes revealed similar topologies (Op den Camp et al. 2009). However, the placement of the pmoCAB3 genes was often quite distant and sensitive to the particular sequence set and tree-building algorithm used. One phylogenetic conclusion at least is obvious: proteobacterial and verrucomicrobial pmo operons form distinct lineages without recent lateral gene transfer between the two groups. However, the relationship of the pmoCAB3 genes to the other two operons in “Methylacidiphilum” is not clear.

Fig. 9.4
figure 00094

Phylogenetic tree of partial (495 nucleotides) pmoA/amoA gene sequences of major groups of ammonia oxidizers and methanotrophs. The tree was constructed with TREE_PUZZLE, a quartet maximum-likelihood method, using a Schoeniger-von Hasseler distance calculation (Schmidt et al. 2002) and 50,000 iterations. Support values for the major nodes are given. The scale bar represents 0.1 changes per nucleotide position

Full size image

A Bayesian phylogenetic analysis done by Tavormina et al. (2010) led to some interesting speculation about the deepest nodes of the trees. The topologies of amoA/pmoA and amoB/pmoB in their Bayesian trees support the hypothesis that “Methylacidiphilum” pmoA3 genes are related to the amoA genes of ammonia-oxidizing Crenarchaeota. Nodes connecting verrucomicrobial pmoA3/pmoB3 with amoA/amoB from Crenarchaeota were highly supported. Unfortunately, Bayesian phylogenies are notorious for delivering misleadingly high support values for deep nodes (e.g., Yang and Rannala 2005). In addition, the extremely deep branch to the Crenarchaeota amoA cluster may cause long-branch-attraction artifacts, a known hazard in any phylogenetic construction. Quartet-puzzling phylogenies constructed by Dunfield et al. (2007) and Stein et al. (2012) also include crenarchaeal amoA. These sometimes support a monophyletic Verrucomicrobia clade, but sometimes also support an affiliation of the verrucomicrobial pmoA3 with crenarchaeal amoA. As is often the case, the conclusions are sensitive to different phylogenetic construction methods and datasets. A phylogeny constructed with a quartet-puzzling method and a modified dataset is shown in Fig. 9.4. The support for the pmoA3 and crenarchaeal amoA node is present but weak at 57 %. Although the speculations of Tavormina et al. (2010) on the evolutionary connection of the pmoA3 with amoA from Crenarchaeota are intriguing, they will require more data for confirmation.

Tavormina et al. (2010) also noted that the pmoA1 and pmoA2 genes in V4 also have a similar GC content as the overall genome (47 % vs. 45 %), but the pmoA3 genes consist of only 37 % GC. They concluded that the pmoA3 gene may be foreign to the genome. The Alien Hunter analysis noted earlier predicts the same, supporting the speculation of Tavormina et al. (2010) about the separate origin of the pmoCAB3 operon compared to the other operons.

Conserved residues in derived PmoA and PmoC amino acid sequences of “Methylacidiphilum” were identical to those in the sequences from Proteobacteria. However, most of the conserved residues in proteobacterial PmoB were different in the Verrucomicrobia. These residues are involved in copper binding in proteobacterial pMMO, and their variability suggests a different metal coordination in the verrucomicrobial patterns (Op den Camp et al. 2009).

9.6 Pathways for C1 Metabolism and CO2 Fixation

Aerobic methanotrophs first convert methane to methanol, which is oxidized to formaldehyde, formate, and CO2. Methanotrophs employ a modular design for these metabolic steps (see Chistoserdova 2011) and Chap. 8 in this volume), with different species combining different modules for each step. The systems in “Methylacidiphilum” are not unique, although the particular combination of them is not seen in any known proteobacterial methanotroph. “Methylacidiphilum” has a gene encoding the large subunit (mxaF/xoxF) of a pyrroloquinoline quinone-dependent alcohol dehydrogenase of the methanol/ethanol family. No genes encoding the small subunit (mxaI) of methanol dehydrogenase or genes for several accessory proteins were found (Hou et al. 2008). Formaldehyde is probably oxidized via an enzymatic pathway using the C1-carrying cofactor tetrahydrofolate. “Methylacidiphilum” does not have the methylene-H4F dehydrogenase and methenyl-H4F cyclohydrolase enzymes that have been found in other methylotrophs, but does have a bifunctional folD gene product that may perform these functions. A more common pathway in proteobacterial methanotrophs uses the C1-carrier tetrahydromethanopterin, but this is absent in “Methylacidiphilum.” Formate is oxidized by a simple formate dehydrogenase.

Formaldehyde is a key intermediate in methane oxidation. It can be oxidized to CO2 for energy generation or directly assimilated into cell material. Methanotrophs assimilate formaldehyde via either the serine cycle, in which formaldehyde and CO2 are utilized in a one-to-one ratio to produce acetyl-CoA for biosynthesis, or via the RuMP pathway, in which formaldehyde is added to ribulose-5-monophosphate, which via sugar rearrangements produces glyceraldehyde-3-phosphate as a biosynthetic building block. Therefore, although there is some anapleurotic CO2 assimilation, particularly in methanotrophs employing the serine cycle, these bacteria are considered heterotrophs that assimilate organic carbon. An estimated 85–95 % of biomass C in gammaproteobacterial methanotrophs and 50 % of the biomass C in alphaproteobacterial methanotrophs is derived from methane rather than CO2 (Trotsenko and Murrell 2008).

Key enzymes of the RuMP pathway are absent in “Methylacidiphilum.” Two key enzymes of the serine cycle, malyl-CoA lyase (mcl) and glycerate kinase (glc), could not be identified in the V4 genome, and other putative serine cycle enzymes were not arranged in genomic islands as they are in other methanotrophs, suggesting that their function may not be in the serine cycle. Hou et al. (2008) suggested that a modified serine cycle may operate, but further evidence would be required. The only clear pathway for C fixation identified in the genome of V4 was the Calvin-Benson-Bassham cycle.

RuBisCO, the key enzyme in the Calvin-Benson-Bassham (CBB) cycle, is present in both V4 and SolV. A number of methanotrophs possess the enzymes of the CBB, most in addition to other C1 assimilation pathways (Chistoserdova 2011). The CBB cycle is associated with a large energy use per mole of CO2 fixed and was not considered a likely way to support growth on CH4. However, Khadem et al. (2011) applied both 13CH4 and 13CO2 during growth experiments to verify the genome prediction that CO2 is in fact the only carbon source for strain SolV. Genome and transcriptome analyses verified that all the genes necessary for the CBB cycle are present and expressed, most prominently the two genes encoding RuBisCO. Carboxysomes are thought to enhance the concentration of CO2 for RuBisCO as it has a low affinity for CO2, but strain SolV uses a non-carboxysome-associated form of RuBisCO, in agreement with the high concentrations of CO2 required for growth (>0.3 %). The verrucomicrobial methanotrophs have been isolated from volcanic regions with high concentrations of CO2, suggesting that there is no need to sequester CO2 for growth (Khadem et al. 2011). Phylogenetic analysis of the rbcL gene suggests an LGT event from the Actinobacteria; however, this divergence must be ancient given that the Verrucomicrobia rbcL gene clade is clearly distinct from the Actinobacteria.

The autotrophic nature of the methanotrophic verrucomicrobia has recently been confirmed in strain V4 by Sharp et al. (2012) by performing a modified stable-isotope probing (SIP) technique labeling with both 13CH4 and 13CO2, individually and in combination. Strain V4 only assimilated 13CO2, not 13CH4, confirming the source of carbon for the verrucomicrobial methanotrophs is CO2. In combination with quantitative PCR (qPCR) of verrucomicrobial-pmoA genes, this technique was then applied to geothermal soils from New Zealand, providing the first demonstration of the autotrophic nature of verrucomicrobial methanotrophs in situ. This study also demonstrated that there are several diverse clades of Verrucomicrobia methanotrophs present at the study site. Further application of the technique should allow an assessment of the overall diversity of Verrucomicrobia methanotrophs in nature.

9.7 Nitrogen Metabolism

The ability to fix gaseous nitrogen confers an advantage in environments where nitrogen is limiting. Genomic analyses of strains V4 and SolV suggest that these strains should be able to fix gaseous nitrogen, as they possess complete sets of genes for nitrogen fixation. Most of these genes have their organization in putative operons resembling those of Methylococcus capsulatus Bath (Ward et al. 2004), a proteobacterial methanotroph that has been shown to fix nitrogen. Khadem et al. (2010) demonstrated that the verrucomicrobial methanotroph strain SolV was able to fix nitrogen under low O2 concentrations. Based on acetylene reduction assays and growth experiments, the nitrogenase enzyme of strain SolV appeared to be extremely O2 sensitive compared to that of proteobacterial methanotrophs. Phylogenetic analysis of a concatenated set of the derived amino acid sequences of the nifH, nifD, and nifK genes showed that the verrucomicrobial nitrogenases group with those of the Proteobacteria and acidophilic Leptospirillum species that inhabit acid mine drainage. The tree is supportive of an LGT scenario with Proteobacteria (Khadem et al. 2010).

Ammonia is a strong competitive inhibitor of pMMO. Many methanotrophs oxidize ammonia to nitrite in two steps. In the first step, ammonia is oxidized to hydroxylamine by pMMO. In the second step, hydroxylamine is oxidized to nitrite. Many methanotrophs possess hydroxylamine cytochrome c oxidoreductase (HAO) activity to prevent accumulation of toxic hydroxylamine (Ward et al. 2004; Klotz et al. 2008). Recently, ammonia-responsive transcription of the haoAB genes encoding HAO have been reported for the methanotroph Methylococcus capsulatus Bath (Poret-Peterson et al. 2008). The finding of the haoAB gene on a plasmid in Silicibacter pomeroyi suggests a high mobility of these genes (Klotz et al. 2008). Both strains V4 and SolV of Methylacidiphilum possess an haoA gene encoding the large subunit of HAO. Phylogenetically, these groups with HAOs from methanotrophic Proteo bacteria form a separate clade from the nitrifying bacteria Nitrosococcus, Nitrosospira, and Nitrosomonas. This may indicate lateral transfer of haoA from Proteobacteria to the Verrucomicrobia (Klotz et al. 2008), although in theory it may instead indicate transfer in the opposite direction.

The oxidation of hydroxylamine by HAO is accompanied by the production of small amounts of nitric oxide (NO). In “Methylacidiphilum,” NO is then reduced to N2O via a norCB-encoded nitric oxide reductase. This is found in some but not all proteobacterial methanotrophs (Stein and Klotz 2011). Phylogenetic analysis of norC and norB suggests that these are other strong candidates for lateral transfer. A small cluster of norB from Aquificae, Thermus/Deinococcus, and Verrucomicrobia cluster within a much larger clade of genes from Proteobacteria (Fig. 9.2). This phylogenetic clade may represent a thermotolerant form of the enzyme that was originally obtained via transfer from Proteobacteria and then shared among several thermophiles.

9.8 Conclusions

We are only beginning to understand the physiology of the “Methylacidiphilum”-type methanotrophs. They share many enzymatic modules with proteobacterial methanotrophs, but have a unique combination of these. Particularly special is their reliance on autotrophic C-assimilation, which seems counterintuitive for a bacterium growing on a reduced carbon substrate. Methanotrophs have often been defined in the literature as bacteria that can use methane as their sole energy and carbon source, but in fact “carbon” should be removed from this definition. Studies into the ecology and physiology of these bacteria are ongoing.