Widespread soil bacterium that oxidizes atmospheric methane

Abstract

The global atmospheric level of methane (CH₄), the second most important greenhouse gas, is currently increasing by ∼10 million tons per year. Microbial oxidation in unsaturated soils is the only known biological process that removes CH₄ from the atmosphere, but so far, bacteria that can grow on atmospheric CH₄ have eluded all cultivation efforts. In this study, we have isolated a pure culture of a bacterium, strain MG08 that grows on air at atmospheric concentrations of CH₄ [1.86 parts per million volume (p.p.m.v.)]. This organism, named Methylocapsa gorgona, is globally distributed in soils and closely related to uncultured members of the upland soil cluster α. CH₄ oxidation experiments and ¹³C-single cell isotope analyses demonstrated that it oxidizes atmospheric CH₄ aerobically and assimilates carbon from both CH₄ and CO₂ Its estimated specific affinity for CH₄ (a⁰_s) is the highest for any cultivated methanotroph. However, growth on ambient air was also confirmed for Methylocapsa acidiphila and Methylocapsa aurea, close relatives with a lower specific affinity for CH₄, suggesting that the ability to utilize atmospheric CH₄ for growth is more widespread than previously believed. The closed genome of M. gorgona MG08 encodes a single particulate methane monooxygenase, the serine cycle for assimilation of carbon from CH₄ and CO₂, and CO₂ fixation via the recently postulated reductive glycine pathway. It also fixes dinitrogen and expresses the genes for a high-affinity hydrogenase and carbon monoxide dehydrogenase, suggesting that atmospheric CH₄ oxidizers harvest additional energy from oxidation of the atmospheric trace gases carbon monoxide (0.2 p.p.m.v.) and hydrogen (0.5 p.p.m.v.).

Keywords: USC alpha; filter cultivation; isolate; methane; trace gases.

Fig. 1.

Phylogenetic relationship of the PmoA and electron micrographs of M. gorgona MG08. (A) The unrooted maximum-likelihood tree based on 155-aa positions was computed using the Jones Taylor Thornton matrix-based model of amino acid substitution. Env. Seq., environmental sequence (clone or DGGE sequence). The PmoA of M. gorgona MG08 clusters with other uncultured MOB within the USCα JR1/cluster 5 (Methylocella species does not carry the genes that encode pMMO and is therefore not represented in the tree). Bootstrap values are presented at branch points (1,000 data resamplings). Bar, 0.05-aa substitutions per site. (B) Scanning electron micrograph of M. gorgona MG08 cells grown at 21 °C in liquid culture with 20% CH₄ headspace without shaking. (C) Transmission electron micrograph of cells grown under the same conditions showing intracytoplasmic membranes of type III and inclusions that resemble PHB granules.

Fig. 2.

Average nucleotide and amino acid identities, phylogenetic relationship, and central metabolism comparison between M. gorgona MG08 and its genome sequenced relatives. (A) Symmetrical matrix of pairwise gANI and AAIs between all strains and MAGs and ordered as in B. ANI is presented in the Lower Left triangle and values ≥74 are provided. AAI is presented in the Upper Right triangle and values ≥60 are provided. M. gorgona MG08 and Ca. M. lahnbergensis (AAI, 71.3; ANI, 78.1), MAG USC1 (AAI, 70.1; ANI, 78.0), MAG USC2 (AAI, 70.0; ANI, 77.3), M. aurea KYG T (AAI, 69.3; ANI, 72.5), M. acidiphila B2 (AAI, 67.4; ANI, 74.75), M. palsarum NE2 (AAI, 66.1; ANI, 67.9) and M. silvestris BL2 (AAI, 62.5; ANI, 69.9) are below the species threshold of 96.6 ANI (3) and 95 AAI (4). (B) The phylogenomic tree was calculated with 10 independent chains of 11,000 generations under the LG model with four rate categories, using an input alignment of 34 concatenated marker genes (Materials and Methods). A total of 6,000 generations of each chain were discarded as burn-in, the remainder were subsampled every five trees (bpcomp -x 6000 5 11000) and pooled together for calculation of the reported 50% consensus tree and bipartition posterior probabilities (maxdiff = 0.814, meandiff = 0.010076). The model and number of rate categories was identified using ModelFinder (Materials and Methods). (C) Distribution of functional complexes presented in Fig. 4 and SI Appendix, Table S1 were determined using blast (5), OrthoFinder (6), and manual examination of trees. Presence of a complete complex is indicated by a solid square. Complexes that are incomplete are indicated with an embedded diamond. Abbreviations for functional complexes: aca, carbonic anhydrase; acc, acetyl-CoA carboxylase; atp, ATP synthase; cox, carbon monoxide dehydrogenase; cyd, terminal cytochrome oxidase; eno, enolase; fae, formaldehyde activating enzyme; fdh, formate dehydrogenase; fdx, ferredoxin, 2Fe-2S; fhc, formyltransferase/hydrolase complex; fhs, formate–tetrahydrofolateligase; FNR, ferredoxin-NADP+ oxidoreductase; fol, bifunctional 5,10-methylene-tetrahydrofolatedehydrogenase, and 5,10-methylene-tetrahydrofolatecyclohydrolase; gck, 2-glycerate kinase; gcv, glycine cleavage complex; gly, serine hydroxymethyltransferase; hhy, [NiFe] hydrogenase; hpr, hydroxypyruvate reductase; mch, methenyl tetrahydromethanopterin cyclohydrolase; mcl, malyl-CoA lyase; mdh, malate dehydrogenase; mtd, NAD(P)-dependent methylene tetrahydromethanopterin dehydrogenase; mtk, malate thiokinase; mxa, methanol dehydrogenase; nif, nitrogenase; nuo, NADH-quinone oxidoreductase; pet, ubiquinol-cytochrome c reductase; pmo, particulate methane monooxygenase; ppc, phosphoenolpyruvate carboxylase; and sga, serine–glyoxylate aminotransferase.

Fig. 3.

Global distribution of the M. gorgona lineage. The map indicates sampling locations of public SRA datasets containing members of the M. gorgona lineage (Materials and Methods). The M. gorgona lineage was identified in 1,537 SRA datasets, of which 1,240 are shown. Latitude and longitude were unavailable for the remaining 297 datasets.

Fig. 4.

The central carbon and energy metabolism of M. gorgona MG08 as predicted from its genome and confirmed by proteomics. H₄MPT, tetrahydromethanopterin. Dashed black arrows indicate passive diffusion across the cell membrane. Numbers for the metabolic steps in the figure refer to the following enzyme names: (1) particulate methane monooxygenase, (2) methanol dehydrogenase and corresponding cytochrome c, (3) formaldehyde activating enzyme, (4) NAD(P)-dependent methylene tetrahydromethanopterin dehydrogenase, (5) methenyl tetrahydromethanopterin cyclohydrolase, (6) formyltransferase/hydrolase, (7A) NAD-dependent formate dehydrogenase, (7B) molybdopterin binding reversible formate dehydrogenase/CO₂ reductase, (8) formate-tetrahydrofolate ligase, (9) bifunctional 5,10-methylene-tetrahydrofolate dehydrogenase and 5,10-methylene-tetrahydrofolate cyclohydrolase, (10) glycine cleavage system, (11) serine hydroxymethyltransferase, (12) serine–glyoxylate aminotransferase, (13) hydroxypyruvate reductase, (14) 2-glycerate kinase, (15) enolase, (16) phosphoenolpyruvate carboxylase, (17) malate dehydrogenase, (18) malate thiokinase, (19) malyl-CoA lyase, (20) carbonic anhydrase, (21) acetyl-CoA carboxylase, (22) [MuCo] class I carbon monoxide dehydrogenase, (23) cytochrome c reductase and corresponding cytochrome c, (24) cytochrome c/d/o terminal oxidases (cytochrome d oxidase—cydAB, cytochrome o ubiquinol oxidase—cyoABCD, heme-copper cytochrome c oxidase type A1 – ctaAEGBC, heme-copper cytochrome type C cbb₃ oxidase—ccoNOQP and heme-copper cytochrome c oxidase type A1—coxCAB). (25) [NiFe] group 1h hydrogenase, (26) NADH dehydrogenase, (27) ATP synthase, (28) nitrogenase, (29) ferredoxin, 2Fe-2S, and (30) ferredoxin-NADP⁺ oxidoreductase. All enzymes and electron carrier proteins were also detected in the proteome, with the exception of the nitrogenase and the cbb₃ oxidase. A full list of the proteins, corresponding genes, and genome entries are found in SI Appendix, Table S1.

Fig. 5.

Microcolonies of M. gorgona MG08 cultivated at different CH₄ concentrations. Microcolonies were grown on polycarbonate filters floating on liquid nitrate mineral salt medium for the number of days specified on each picture, either in closed jars with air amended with different concentrations of CH₄, or exposed to unamended air. For fixation, the filters were transferred to fresh-made 2% paraformaldehyde in 1× PBS in the refrigerator overnight. For staining, filters were transferred (side with bacteria up) on top of 200 µL droplets of 1,000× SYBRgreen and incubated for 10 min, washed, and air dried.

Fig. 6.

CH₄ oxidation by M. gorgona MG08 microcolonies incubated on floating filters under atmospheric air (A). Microcolony formation under atmospheric air (B). (A) Five 170-mL bottles with floating polycarbonate filters were incubated on 35 mL 1/10 diluted nitrate mineral salt media (Materials and Methods) under atmospheric air (135 mL 1 atm headspace, sealed with rubber stopper) for 120 d. An additional set of five bottles was incubated without filters as a negative control for CH₄ oxidation. A two-sample t test assuming equal variances confirmed that headspace CH₄ concentrations were significantly different in bottles with filters containing cells (P value < 0.001), compared with those without cells. (B) Filters from A were manually inspected to identify colony formation. (Top) Showing single cells before incubation. (Bottom) Colonies formed after 120-d incubation in one of the bottles from A. For staining in B, filters were transferred (side with bacteria up) on top of 200 µL droplets of 1,000× SYBRgreen and incubated for 10 min, washed, and air dried.

Fig. 7.

Carbon fixation in M. gorgona MG08. NanoSIMS visualization of the ¹³C isotope label incorporation in M. gorgona MG08 cells grown on polycarbonate filters in atmospheres containing 20 p.p.m.v. CH₄ and 1,000 p.p.m.v. CO₂. Two incubations are compared: ¹³CH₄ in combination with ¹²CO₂ (A) and ¹²CH₄ in combination with ¹³CO₂ (B). ¹³C/(¹²C + ¹³C) isotope fraction values, given in at%, are displayed on a false color scale ranging from 0.9 at% (dark blue) to 2.5 at% (red). Cells grown in isotopically unlabeled methane and carbon dioxide showed a ¹³C content of 1.09 ± 0,01 at% (SD, n = 60).