Metabolic Potential for Reductive Acetogenesis and a Novel Energy-Converting [NiFe] Hydrogenase in Bathyarchaeia From Termite Guts - A Genome-Centric Analysis

Abstract

Symbiotic digestion of lignocellulose in the hindgut of higher termites is mediated by a diverse assemblage of bacteria and archaea. During a large-scale metagenomic study, we reconstructed 15 metagenome-assembled genomes of Bathyarchaeia that represent two distinct lineages in subgroup 6 (formerly MCG-6) unique to termite guts. One lineage (TB2; Candidatus Termitimicrobium) encodes all enzymes required for reductive acetogenesis from CO₂ via an archaeal variant of the Wood-Ljungdahl pathway, involving tetrahydromethanopterin as C₁ carrier and an (ADP-forming) acetyl-CoA synthase. This includes a novel 11-subunit hydrogenase, which possesses the genomic architecture of the respiratory Fpo-complex of other archaea but whose catalytic subunit is phylogenetically related to and shares the conserved [NiFe] cofactor-binding motif with [NiFe] hydrogenases of subgroup 4 g. We propose that this novel Fpo-like hydrogenase provides part of the reduced ferredoxin required for CO₂ reduction and is driven by the electrochemical membrane potential generated from the ATP conserved by substrate-level phosphorylation; the other part may require the oxidation of organic electron donors, which would make members of TB2 mixotrophic acetogens. Members of the other lineage (TB1; Candidatus Termiticorpusculum) are definitely organotrophic because they consistently lack hydrogenases and/or methylene-tetrahydromethanopterin reductase, a key enzyme of the archaeal Wood-Ljungdahl pathway. Both lineages have the genomic capacity to reduce ferredoxin by oxidizing amino acids and might conduct methylotrophic acetogenesis using unidentified methylated compound(s). Our results indicate that Bathyarchaeia of subgroup 6 contribute to acetate formation in the guts of higher termites and substantiate the genomic evidence for reductive acetogenesis from organic substrates, possibly including methylated compounds, in other uncultured representatives of the phylum.

Keywords: Bathyarchaeota; Wood-Ljungdahl pathway; acetogens; comparative genomics; gut microbiota; metagenome-assembled genomes; termites.

FIGURE 1

Genome-based phylogeny of termite gut Bathyarchaeia, illustrating the relationship of lineages TB1 and TB2 to other MAGs in the Bathy-6 subgroup (f__UBA233 in the GTDB taxonomy). MAGs of other subgroups that are mentioned in the text are marked in bold. The maximum-likelihood tree was inferred from a concatenated alignment of 43 marker genes using the LG+F+I+G4 model and rooted with selected Crenarchaeota and Euryarchaeota as outgroup. A fully expanded tree with the accession numbers for all genomes is shown in the Supplementary Material (Supplementary Figure S2). The scale bar indicates 10-amino-acid 10% sequence divergence. Highly supported nodes (SH-aLRT, • ≥ 95%, 1,000 replications) are indicated.

FIGURE 2

16S rRNA-based phylogeny of subgroup Bathy-6, indicating the placement of the termite clade among Bathyarchaeia from other environments. The maximum-likelihood tree is based on a curated alignment (1,424 positions) of all sequences in the SILVA database and their homologs retrieved from the bathyarchaeal MAGs and the low-quality bins obtained from the termite gut metagenomes (Hervé et al., 2020). The tree was rooted with members of Bathy-5 as outgroup. The scale bars indicate 0.05 nucleotide substitutions per site. SH-aLRT values (• ≥ 95%; ∘ ≥ 80%, 1,000 replications) indicate node support. Branches marked with dashed lines indicate shorter sequences that were added using the parsimony tool. A fully expanded tree with the accession numbers of all sequences is shown in the Supplementary Material (Supplementary Figure S3).

FIGURE 3

Gene functions encoded by termite gut bathyarchaea (TB1 and TB2) and other representatives of the Bathy-6 subgroup. All phylotypes with sufficiently complete genomes were included; their phylogenetic relationship was taken from Figure 1 (for strain designations, see Table 1). Colored circles indicate presence, and open circles indicate absence of the respective function; light blue indicates that a gene set is incomplete. The asterisks (*) in MtrABCDEFGH and FrhABG indicate that only MtrH or FrhB, respectively, is present. The number of aminotransferases encoded by each phylotype is indicated in the circle. If a phylotype is represented by more than one MAG, the annotation results were combined; details can be found in the Supplementary Material (Supplementary Table S2). H₄MPT, tetrahydromethanopterin; MFR, methanofuran; Fpo, F₄₂₀:methanophenazine oxidoreductase.

FIGURE 4

Catabolic pathways encoded by the MAGs of termite gut Bathyarchaeia. The circles next to each enzyme indicate the presence of the coding genes in all, some, or none of the phylotypes of TB1 and TB2 (more details in Figure 3). Gray shading indicates pathways that are absent from all MAGs. The directionality of Fpo-like hydrogenase (Hfo) and ATP (synth)ase is discussed in the text. Dashed lines with question marks indicate hypothetical interactions. MFR, methanofuran; H₄MPT, tetrahydromethanopterin (archaeal pathway); THF, tetrahydrofolate (bacterial pathway). A detailed list of genes present/absent in the respective MAGs is provided as Supplementary Material (Supplementary Table S2).

FIGURE 5

Organization of the gene clusters encoding the respiratory complexes (blue background) and the ancestral [NiFe] hydrogenases in group 4 (red background). Identical colors indicate homologous genes; a phylogenetic analysis of the catalytic subunit of [NiFe] hydrogenases and its homologs (red) is shown in Figure 6. The font style of the gene labels indicates differences in the subunit nomenclature of Nuo/Fpo (uppercase), Mbh (lowercase), and Ech (italics).

FIGURE 6

Phylogeny of the catalytic subunit of selected group 4 [NiFe] hydrogenases and their homologs in the respiratory complexes. The maximum-likelihood tree is based on a curated alignment of the deduced amino acid sequences; the scale bar indicates 0.1-amino-acid substitutions per site. SH-aLRT values (• ≥ 95%; ∘ ≥ 80%, 1,000 replications) indicate node support. The genomic context of the highlighted genes is shown in Figure 5. Gene numbers indicate IMG/Mer gene IDs.

FIGURE 7

Comparison of the [NiFe]-binding motifs (L1 and L2) in the large subunits of selected group 4 [NiFe] hydrogenases with the corresponding amino acid residues (IUPAC code) of their homologs in the Nuo and Fpo complexes. The shading indicates the typical motifs of [NiFe] hydrogenases (L1 motif: C[GS][ILV]C[AGNS]xxH; L2 motif: [DE][PL]Cx[AGST]Cx[DE][RL]; Vignais and Billoud, 2007). The four cysteine residues that coordinate the [NiFe] cluster are marked in red; other conserved residues are marked in blue.