
==== Front
mSystems
mSystems
msystems
mSystems
2379-5077
American Society for Microbiology 1752 N St., N.W., Washington, DC

39189956
msystems00999-24
10.1128/msystems.00999-24
msystems.00999-24
Research Article
environmental-microbiologyEnvironmental MicrobiologyDiverse non-canonical electron bifurcating [FeFe]-hydrogenases of separate evolutionary origins in Hydrogenedentota
https://orcid.org/0000-0001-9731-4476
Zheng Xiaowei 1 zhengxw@im.ac.cn

https://orcid.org/0000-0003-2421-8162
Huang Li 1 2 huangl@im.ac.cn

1 State Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences , Beijing, China
2 College of Life Sciences, University of Chinese Academy of Sciences , Beijing, China
Editor Greening Chris Monash University , Melbourne, Victoria, Australia

Address correspondence to Xiaowei Zheng, zhengxw@im.ac.cn
Address correspondence to Li Huang, huangl@im.ac.cn
The authors declare no conflict of interest.

9 2024
27 8 2024
27 8 2024
9 9 e00999-2423 7 2024
25 7 2024
Copyright © 2024 Zheng and Huang.
2024
Zheng and Huang.
https://creativecommons.org/licenses/by/4.0/ This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license.

ABSTRACT

Hydrogenedentota, a globally distributed bacterial phylum-level lineage, is poorly understood. Here, we established a comprehensive genomic catalog of Hydrogenedentota, including a total of seven clades (or families) with 179 genomes, and explored the metabolic potential and evolutionary history of these organisms. We show that a single genome, especially those belonging to Clade 6, often encodes multiple hydrogenases with genomes in Clade 2, which rarely encode hydrogenases being the exception. Notably, most members of Hydrogenedentota contain a group A3 [FeFe]-hydrogenase (BfuABC) with a non-canonical electron bifurcation mechanism, in addition to substrate-level phosphorylation and electron transport-linked phosphorylation pathways, in energy conservation. Furthermore, we show that BfuABC from Hydrogenedentota fall into five sub-types. Phylogenetic analysis reveals five independent routes for the evolution of BfuABC homologs in Hydrogenedentota. We speculate that the five sub-types of BfuABC might be acquired from Bacillota (synonym Firmicutes) through separate horizontal gene transfer events. These data shed light on the diversity and evolution of bifurcating [FeFe]-hydrogenases and provide insight into the strategy of Hydrogenedentota to adapt to survival in various habitats.

IMPORTANCE

The phylum Hydrogenedentota is widely distributed in various environments. However, their physiology, ecology, and evolutionary history remain unknown, primarily due to the limited availability of the genomes and the lack of cultured representatives of the phylum. Our results have increased the knowledge of the genetic and metabolic diversity of these organisms and shed light on their diverse energy conservation strategies, especially those involving electron bifurcation with a non-canonical mechanism, which are likely responsible for their wide distribution. Besides, the organization and phylogenetic relationships of gene clusters coding for BfuABC in Hydrogenedentota provide valuable clues to the evolutionary history of group A3 electron bifurcating [FeFe]-hydrogenases.

KEYWORDS

Hydrogenedentota
electron bifurcation
[FeFe]-hydrogenase
gene transfer and loss
Ministry of Science and Technology of the People's Republic of China (MOST) SQ2021QZKK0302 Zheng Xiaowei CAS | Bureau of Development and Planning, Chinese Academy of Sciences (Bureau of Development and Planning, CAS) XDB42000000 Huang Li MOST | National Natural Science Foundation of China (NSFC) 42376238 Zheng Xiaowei MOST | National Natural Science Foundation of China (NSFC) 32393970 Huang Li MOST | National Natural Science Foundation of China (NSFC) 91751000 Huang Li cover-dateSeptember 2024
==== Body
pmcINTRODUCTION

Hydrogenedentota bacteria were first found in a clone library of 16S rRNA genes prepared from sediment samples collected in Nankai Trough at a depth of 3,843 m and identified as candidate division NKB19 in 1999 (1). After its genomes were sequenced by using single-cell technology, NKB19 was named Hydrogenedentes (2), probably due to its suspected ability to consume or produce hydrogen through [Ni-Fe] hydrogenase, and this name was later amended to Hydrogenedentota according to Genome Taxonomy Database (GTDB) release86 for orthography (3, 4). Members of this phylum have been detected in various habitats, such as wastewater, agricultural soil, hot spring, desert, sea water, and polar ice shelf (5–10). However, very little is currently known about this phylum, since no cultured isolates are available, and none of the public genomes of this phylum have been analyzed in detail until now.

Hydrogenases are highly complex metalloenzymes that catalyze the reversible interconversion of H2 to 2H+ + 2e−, enhancing the host’s survival in diverse habitats by utilizing H2 as a source of low potential electrons or dispensing excess reducing equivalents to control the cellular redox balance (11, 12). Three evolutionarily distinct types of hydrogenases are classified on the basis of both sequence similarities and compositions of metal-containing catalytic centers and are termed [FeFe]-, [NiFe]-, and [Fe]-hydrogenases (13–17). Whereas [Fe]-hydrogenase is only found in archaea (16), the other two types of hydrogenases are more widely distributed in nature (13, 15, 18, 19). [NiFe]-hydrogenase exists in both bacteria and archaea. [FeFe]-hydrogenase was once believed to be only present in bacteria and eukaryotes. Recently, Greening et al. revealed the presence of [FeFe]-hydrogenase in archaea (20), challenging the previous view that [FeFe]-hydrogenase evolved later in bacteria than [NiFe]-hydrogenase in the last universal common ancestor (15, 21). Previously, [NiFe]-hydrogenase has been annotated in four single-amplified genomes (SAGs) of Hydrogenedentota (2), but [FeFe]-hydrogenase has never been reported before in this phylum.

In this study, we first analyzed the global distribution of Hydrogenedentota based on 16S rRNA amplicon data from the Earth Microbiome Project (EMP) and then retrieved published metagenome-assembled genomes (MAGs) and SAGs for the genetic and metabolic investigation of the phylum. We show that members of this phylum are globally distributed, employ diverse strategies in energy conservation, and are capable of using a non-canonical catalytic mechanism involving diverse electron bifurcating [FeFe]-hydrogenase (BfuABC) for energy conservation or disposing of excess reducing equivalents. We also reveal multiple evolutionary origins of BfuABC homologs in Hydrogenedentota. Our data shed light on the evolution of this non-canonical electron bifurcating [FeFe]-hydrogenases and the strategy of energy conservation of these widespread low-abundance microorganisms in adaptation to surviving in various habitats.

RESULTS AND DISCUSSION

The phylum of Hydrogenedentota is a globally distributed low-abundance lineage

We retrieved a total of 23,323 qualified samples from the EMP (22). About 35% (8,009) of the samples were found to contain species belonging to Hydrogenedentota. These species are distributed around the globe and generally in low abundance in various habitats. They have been found in terrestrial soil, freshwater sediment, marine sediment, fresh water, marine water, and host-associated and other habitats, with the median relative abundances of only 0.013%, 0.054%, 0.067%, 0.006%, 0.002%, 0.019%, and 0.015%, respectively (Fig. S1 and S2; Data Set S1, Sheet 1).

Hydrogenedentota are facultative anaerobic bacteria with a heterotrophic lifestyle

A total of 179 medium- to high-quality Hydrogenedentota genomes (≥50% completeness and ≤10% contamination) were retrieved from NCBI, GTDB, Genomes from Earth’s Microbiomes (GEM), Figshare, and European Nucleotide Archive (ENA) databases (Data Set S1, Sheet 2). These genomes range from 2.68 to 12.10 Mb in size with a median size of 4.68 Mb, and their GC contents were between 36.63% and 69.02% with a median of 59.34% (Fig. 1). The large variation in estimated genome size and GC content points to considerable genome diversity within the phylum. These genomes are grouped into 100 potential strains and 90 candidate species-level taxa, respectively. As revealed by phylogenetic analysis based on 120 concatenated single-copy marker proteins, the 179 genomes are grouped into seven clades (Clades 1–7, Fig. 1; Fig. S3), which may represent seven families in the order of Hydrogenedentiales according to the GTDB classification.

Fig 1 Phylogenomic tree of Hydrogenedentota inferred from 120 concatenated marker proteins. Tree annotations from inside to outside circles are (1) completeness (%), (2) contamination (%), (3) GC content (%), (4) environmental origin of the 179 Hydrogenedentota genomes, and (5) number of hydrogenases detected in each genome. One hundred non-redundant genomes are indicated by red stars. Branches in yellow, green, black, red, brown, pink, and blue represent seven family-level clades (Clades 1–7, respectively) of Hydrogenedentota. The bootstraps are labelled with black dots in the middle of branches when ≥90%. The tree scale bar indicates the mean number of substitutions per site. The bar plot (top right) shows the number of genomes retrieved per environmental category. The box plot (top center) indicates the statistics of completeness, contamination, and GC content estimates of the 179 Hydrogenedentota genomes. Shaded gradients are consistent with the values in the heatmap boxes in the tree. Details of the 179 Hydrogenedentota genomes are shown in Data Set S1 (Sheet 2).

Circular phylogenetic tree depicts Hydrogenedentota clades. Outer rings depict habitat, GC content, contamination, completeness, and hydrogenase count. Inset bar chart shows genome counts from different habitats.

As illustrated in Fig. 2 and Data Set S1 (Sheet 3), members of Hydrogenedentota are predicted to have a Gram-negative cell wall, since the Raetz pathway for lipopolysaccharide biosynthesis was identified in over 94% of genomes, and UDP-N-acetylmuramoyl-L-alanyl-D-glutamate:meso-2,6-diaminopimelate ligase, which is involved in peptidoglycan biosynthesis by adding meso-diaminopimelic acid (mDAP), the hallmark amino acid residue of Gram-negative bacteria (23), to the pentapeptide, were detected in over 85% genomes. The presence of genes encoding proteins involved in cell division and a rod-shaped cell determination, such as rodA, mreB, mreC, and mreD (24, 25), in 97 of the 100 non-redundant genomes suggests that the cells of Hydrogenedentota may be rod-shaped.

Fig 2 Structure, motility, and metabolic potential of Hydrogenedentota. The phylogenomic tree of 100 non-redundant genomes was constructed and annotated as described in Fig. 1. A metabolic pathway or protein complex is indicated with a solid symbol if complete or intact, open symbol if incomplete or partial, and no symbol if absent. A tick indicates the presence of a specific protein or subunit. Those hydrogenases with defined and uncertain electron bifurcating activities are marked by red and purple ticks, respectively. Details of the structure, motility, and metabolic potential of Hydrogenedentota can be found in Text S1.

Phylogenetic tree with annotations depicts clades 1 to 7 and the presence of various gene families. Each clade is represented by different colors, and colored symbols mark specific gene family presence. Scale bar indicates evolutionary distance.

Further function annotation and interpretation suggest members of Hydrogenedentota are motile bacteria with a heterotrophic lifestyle (Fig. 2; Data Set S1, Sheets 3–5; see Text S1 for details). Hydrogenedentota appear to be capable of generating ATP via substrate-level phosphorylation (SLP) since genes encoding pyruvate kinase and phosphoglycerate kinase, two key enzymes involved in the phosphor group transfer from phosphoenolpyruvate or 1,3-diphosphoglycerate to ADP during glycolysis, were annotated in over 68% (123) and 83% (150), respectively, of the 179 genomes (Data Set S1, Sheet 4).

To gain insights into the energy conservation of Hydrogenedentota, we also characterized the electron transport chain (ETC) of the organisms. As shown in Fig. 2 and Data Set S1 (Sheet 4), complex I (NADH-quinone oxidoreductase) appears to be patchily distributed across the seven clades of Hydrogenedentota and found in only 36% of the genomes. However, electrons may be fed into ETC in genomes lacking complex I by complex II (succinate dehydrogenase) or Na+ translocating NADH:quinone oxidoreductase (Na+-NQR). Other quinone reductases, such as glycerol-3-phosphate dehydrogenase and quinoprotein glucose dehydrogenase, which are involved in glycerophospholipid and carbohydrate metabolism, respectively, are found in almost half of the members of Hydrogenedentota. These observations suggest the presence of diverse pathways for electrons entering the respiratory chain in Hydrogenedentota. While both complex IV, caa3-type heme–copper oxygen reductase and cytochrome bd oxidase, were detected, nearly 70% of the genomes have at least one of them, indicating that Hydrogenedentota mainly use O2 as the electron acceptor. However, oxygen reductase may also serve to protect oxygen-sensitive enzymes in strictly anaerobic organisms when they are exposed to oxic conditions, as shown in Desulfovibrio desulfuricans ATCC 27774 (26). It is worth noting that ubiquinone (UQ or coenzyme Q), usually used by aerobic organisms, is speculated to have appeared about 2.5 billion years ago as a strategy of life to adapt to rising oxygen levels (27, 28), whereas menaquinone (MK or vitamin K2) exists in microorganisms living under low O2 or anoxic conditions (29). Hydrogenedentota possess a complete set of men-like genes for the synthesis of MK, while genes encoding UQ synthesis are incomplete in this phylum (Data Set S1, Sheets 4 and 5). In addition, genes encoding nitrite reductase, nitrate reductase and nitric oxide reductases were detected, suggesting that the organisms are capable of using NO2−, NO3−, or NO as electron acceptors. Further, over half (95) of the 179 Hydrogenedentota genomes are derived from anaerobic environments, such as wastewater, sludge, and bioreactor (Data Set S1, Sheet 2). These results are consistent with the notion that Hydrogenedentota are facultative anaerobes. We also found that FoF1-ATPase and V/A-ATPase are encoded in 61% and 71% of the Hydrogenedentota genomes, respectively (Fig. 2; Data Set S1, Sheet 4). The co-existence of the two types of ATP synthases may offer fitness advantages for survival in various habitats (30–33). Taken together, our data indicate that Hydrogenedentota primarily employ electron transport-linked phosphorylation (ETP) in energy conservation.

Strikingly, these organisms have evolved an electron bifurcation pathway, in addition to SLP and ETP. Specifically, group A3 [FeFe]-hydrogenases (Fig. 2; Fig. S4; Data Set S1, Sheet 6), which use either H2 as a source of reducing power or H+ as an oxidant to dispose of excess reducing equivalents (13, 34), are more widely distributed than [NiFe]-hydrogenases in Hydrogenedentota. This contrasts sharply with the previous belief (2). We identified six types of [NiFe]-hydrogenases (groups 1a, 1c, 3b, 3c, 3d, and 4g) in Hydrogenedentota (Fig. 1 and 2; Fig. S4e; see Text S1 for details), among which only group 3c [NiFe] methyl viologen-reducing (Mvh) hydrogenases are believed to have electron bifurcating activity (35). Due to the lack of two FeS clusters in the hoxF/HydB subunit, a homolog of diaphorase (36) or the BfuB subunit in group A3 [FeFe]-hydrogenase (35, 37) (see below), group 3d [NiFe]-hydrogenases from Hydrogenedentota are probably incapable of electron bifurcation (Fig. S4f). The reversible process catalyzed by the hydrogenases is referred to as electron bifurcation in this report for convenience. Notably, members of Hydrogenedentota are often found to encode either [FeFe]- or [NiFe]-hydrogenase, and the two types of hydrogenases are rarely encoded in the same genome (Fig. 1 and 2).

Electron bifurcating [FeFe]-hydrogenases have diversified into distinct sub-types in Hydrogenedentota

Four types (492 sequences) of putative [FeFe]-hydrogenases, i.e., groups C1 (53), C2 (1), C3 (156), and A3 (282), were identified in the 179 Hydrogenedentota genomes (Fig. 2; Fig. S4a; Data Set S1, Sheet 6). Group C [FeFe]-hydrogenases are believed to be involved in hydrogen sensing (15). However, we were hardly able to find the H-cluster (i.e., the hydrogenase active site consisting of a [4Fe4S] cluster bound via a cysteine to a 2Fe sub-cluster) binding motifs P1, P2, and P3 in group C1 [FeFe]-hydrogenases from Hydrogenedentota by using the reported signature sequences (14, 17, 38). Therefore, it is unclear if they are capable of sensing hydrogen. Twenty-one of the 53 genes encoding group C1 [FeFe]-hydrogenases and nearly all the genes encoding group C3 [FeFe]-hydrogenases are located next to those encoding group A3 [FeFe]-hydrogenases (Fig. S4c and S5). This finding indicates that group C [FeFe]-hydrogenases from Hydrogenedentota are mainly involved in the regulation of the production and consumption of hydrogen. It is difficult to speculate the physiological processes in which the other 32 group C1 [FeFe]-hydrogenases are involved, based on the analysis of genes adjacent to their coding genes due to the limitation incurred by the length of contigs or scaffolds (Data Set S1, Sheet 11). Besides, a histidine kinase/HSP90-like ATPase (HATPase_C) domain, often found in ATP-binding proteins, e.g., histidine kinase (39) and topoisomerases (40), was detected in group C1 [FeFe]-hydrogenases from Hydrogenedentota (Fig. S4d). However, this domain is absent in the best-matched group C1 [FeFe]-hydrogenase in the HydDB database (Fig. S4d) (41). Unlike group C1, two types of C3 [FeFe]-hydrogenases (corresponding to the T subunit in sub-types IV and V of BfuABC, respectively, see below) contain a PAS domain (42) and a histidine kinase domain (His_kinase_dom), respectively, which play a regulatory role. These two domains are also found in the best hits of two types of C3 [FeFe]-hydrogenases in the HydDB database (WP_005996231.1 and WP_006928132.1) (41) (Fig. S4d and S3f).

Since group C2 [FeFe]-hydrogenase was found only in a single genome, and genes encoding groups C3 and A3 [FeFe]-hydrogenases are usually located together in the same gene cluster (Fig. S5), we will focus our analysis on group A3 [FeFe]-hydrogenase and its encoding gene cluster. Three phylogenetically distinct groups of flavin-based electron bifurcating (FBEB) enzymes, i.e., EtfAB, NfnAB, and Hdr-Mvh complexes, are known to use FAD to bifurcate electrons (43–45). However, BfuABC is recently proposed to use a different electron bifurcation mechanism involving the combination of a single FMN and nearby iron sulfur clusters (35, 37). This non-canonical electron bifurcation mechanism suggests that BfuABC might not be FBEB enzymes and represent a novel type of electron bifurcating enzymes, which, unlike flavin (only)- and quinone-based electron bifurcating enzymes, use a unique “FMN-FeS” center to bifurcate electrons (35).

To understand the diversity of BfuABC in Hydrogenedentota and predict the roles of the three subunits in hydrogen production/consumption (BfuA) and bifurcation (BfuB and BfuC) (13, 14, 34, 37, 46), we looked into their organizational features. Gene clusters encoding no or an incomplete BfuB or BfuC subunit in the neighborhood (i.e., within five genes up- or downstream) of the bfuA gene were not retained for further analysis. A total of 195 gene clusters (Fig. S5) were predicted to encode an intact trimeric BfuABC or tetrameric BfuABCD complex. As shown in Fig. 3a and Fig. S6, both the individual and concatenated phylogeny analyses of BfuA, BfuB, and BfuC reveal five independent clusters (termed sub-types I to V). Interestingly, the five BfuABC sub-types nearly perfectly correspond to five distinct patterns of gene cluster organization (Fig. 3b; Fig. S5). To the best of our knowledge, the pattern of BfuABC encoded in the “BAC” orientation, as shown in sub-type III, has never been reported.

Fig 3 Phylogenetic characteristics of group A3 [FeFe]-hydrogenases (BfuABC) in Hydrogenedentota. (a) Unrooted radiation tree based on the concatenate of three subunits “ABC” from 195 nearly complete group A3 BfuABC [FeFe]-hydrogenases. Numbers in parentheses represent the number of hydrogenases in corresponding sub-types. (b) Five sub-types of gene clusters for group A3 [FeFe]-hydrogenases in Hydrogenedentota. To facilitate comparison, the gene orders are displayed according to the gene clusters located on the negative (−) DNA strand. (c) Schematic diagram indicating the extent of conservation in residues comprising the H-cluster motifs (P1, P2, and P3) in Hydrogenedentota BfuA. Residues with higher bit scores (i.e., larger font sizes) represent a greater degree of conservation at a given aligned position. The property of a given amino acid is indicated by a color. (d) Unrooted phylogenetic tree of the concatenate of three subunits “ABC” and domains of BfuB subunit of group A3 [FeFe]-hydrogenase in CSSed162cmA_463 (GCA_007118765.1). (e) Ribbon diagram showing the predicted structures of sub-type IV and V BfuA subunits. (f) Ribbon diagram showing the predicted structures of sub-type IV and V BfuT subunits (group C3 [FeFe]-hydrogenase). “PAS” represents a signal sensor domain, named after three proteins that it occurs in: Per-period circadian protein, Arnt-Ah receptor nuclear translocator protein, and Sim-single-minded protein. “His” represents a histidine kinase domain (His_kinase_dom). (g) Ribbon diagram showing the predicted structures of five sub-types BfuB. The two dotted coiled regions in sub-type II represent the missing binding domains of the “B5” and “B3B4” (the cause of the lack of electron bifurcating) FeS clusters, respectively. In panels (d and e), “A1,” “A5,” “B2,” “B5,” and “C1” represent [2Fe-2S] clusters; “A2,” “A3,” “A4,” “B1,” “B3,” and “B4” represent [4Fe-4S] clusters; “2Fe” represents the active center for hydrogen production or consumption; H-cluster should contain a “2Fe” center and a [4Fe-4S] cluster. “FMN,” flavin mononucleotide binding domain; “Nqo1,” soluble ligand binding β-grasp (SLBB) domain within a region of Nqo1 middle domain-like superfamily; “DPD_II,” dihydroprymidine dehydrogenase domain (DPD_II) carrying two [4Fe-4S] clusters; “FAD,” flavin adenine dinucleotide (FAD) or NAD(P) binding domain. The detailed characteristics of all of the 195 gene clusters and BfuB subunit domains are shown in Fig. 6 and S5. The genome of CSSed162cmA_463 (GCA_007118765.1), which contains all five types of gene clusters, is shown here as an example to illustrate the features of gene clusters and BfuA, BfuB, and BfuT sub-units’ domains of group A3 [FeFe]-hydrogenases in Hydrogenedentota. A gene, denoted T (in sub-type IV gene cluster), in the genome of CSSed162cmA_463 (GCA_007118765.1) is a partial gene residing on the edge of the contig (SKQZ01000441.1; Fig. S5d). SKQZ01000441.1 is nearly 100% identical to SLMA01000044.1 in the genome of CSSed165cm_333 (GCA_007133815.1). The average nucleotide identity value of the two genomes is close to 100%, indicating that they belong to the same species (Fig. S3). Therefore, we speculate that the partial T gene may have resulted from incomplete assembly, and the T gene is thus shown in CSSed162cmA_463 (GCA_007118765.1).

Phylogenetic tree with different groups marked, gene clusters with color-coded gene annotations, motif sequence alignment, gene cluster alignment with gene types marked by colors, protein structures in different conformations, and colors.

BfuA, which is involved in ligating the H-cluster, contains three motifs, i.e., P1 (xTSCxPxWx), P2 (xxMPCxAKK), and P3 (ExMxCPGGCxxGxGQ). Although these motifs are in general conserved, several positions in each of the three motifs vary significantly among the five sub-types (Fig. 3c; Fig. S7). For example, Cys at the fifth position in P1 is often present in sub-type IV while Ser is common in the other four sub-types. Only BfuA in sub-type IV exhibits the “M3c” modular structure (14), which is predicted to contain three [4Fe-4S] clusters (“A2”, “A3,” and “A4”) and two [2Fe-2S] clusters (“A1” and “A5”), of which “A5” is located at the C-terminal end of the protein (Fig. 3d and e). A “M3a” modular structure is typical of sub-types I, II, III, and V. The variations at the motif and domain levels suggest the evolutionary independence of BfuAs of the five sub-types. However, because of the lack of biochemical validation, it remains uncertain whether, or to what extent, these changes affect the oxidation or production of hydrogen.

The presence and location of FeS clusters in BfuB homologs have been verified in HydABC (BfuABC) [FeFe]-hydrogenase from Thermotoga maritima and a related HydABC hydrogenase (NiFe-HydABCSL) from Acetomicrobium mobile DSM 13181 by cryo-electron microscopy (35, 37, 46). A complete set of “B1” to “B5” FeS clusters appears to be necessary for the activity of bifurcating hydrogenases. In support of this notion, this set is incomplete in non-bifurcating homologous Fdh (formate dehydrogenase) and Nqo (NADH quinone oxidoreductase, complex I) enzymes (35, 47). In addition, a HydABC (BfuABC) [FeFe]-hydrogenase from the Syntrophomonas wolfei DSM 2245B, whose HydB (BfuB) subunit lacks “B3” and “B4” FeS clusters, was shown to lack the bifurcating ability in in vitro assays (48, 49). Interestingly, the non-bifurcating hydrogenase (HydABC) was still able to produce hydrogen from NADH without the need for Fd at low hydrogen partial pressure (48, 49). Thus, it is suggested that “B1” and “B2” FeS clusters are capable of reversible electron transfer between NADH and protons, and “B3” and “B4” FeS clusters play a role in interacting with Fd. Similarly, we propose the following complete and reversible electron transport pathways for [FeFe]-hydrogenases from Hydrogenedentota. Two electrons generated from H2 by the H-cluster domain in BfuA are transferred to FMN via the “B1” FeS cluster in the electron bifurcating BfuBC core, followed by bifurcation to the first electron acceptor NAD (high potential) directly and the second electron acceptor Fd (low potential) through FeS clusters in the order of “C1-B2-B3-B4”.

Some suggest that “B5” FeS clusters may display the bifurcating ability since the “B5” and “C1” FeS clusters can act as a gate or switch separating the low potential electron pathway (“B2-B3-B4”) from the mid potential electron pathway (“FMN-B1”) (35), while others argue that the “B5” FeS cluster is at a dead end and unlikely involved in electron transfer (37). Therefore, the function of double “B5” FeS clusters in sub-type III BfuB is still unclear. Interestingly, the domain containing the “B3” and “B4” FeS clusters is replaced by a dihydropyrimidine dehydrogenase domain (DPD_II) with an extra FAD/NAD(P)-binding domain located at the C-terminus of the protein in sub-type I BfuB (Fig. 3d and g). This potentially creates an alternative bifurcation center of “FAD” in addition to the “FMN-FeS” center. As aforementioned, FAD is able to bifurcate electrons alone without the need for FeS clusters in a process known as a canonical bifurcating pathway, as found in FBEB enzymes (43–45). However, the co-existence of “FAD” and “FMN-FeS” bifurcating centers in a single BfuABC hydrogenase has never been reported. The extra FAD binding domain would induce conformational changes in the enzyme, possibly altering both flavin redox properties and electron transport. However, more studies are required to determine if both, either, or neither of the two potential bifurcation centers bifurcate electrons. We postulate that sub-type I BfuABC may represent a novel bifurcating mechanism as it possesses a possible bifurcating “FAD” center in addition to the “FMN-FeS” center.

The gene cluster coding for BfuABC is highly mobile among bacterial phyla

To interrogate the evolutional relationship of the five sub-types of BfuABC [FeFe]-hydrogenases from Hydrogenedentota, we aligned the 195 BfuABC concatenated sequences against a local reference genome database and selected a total of 2,415 BfuABC homologs from 2,144 reference genomes from species belonging to 26 bacterial phyla (Fig. 4; Data Set S1, Sheet 7). The phylogenetic tree of the 2,415 BfuABC homologs from reference genomes, together with 195 BfuABC homologs from Hydrogenedentota genomes, point to complex evolution pathways for group A3 [FeFe]-hydrogenases (Fig. 4). As shown in Fig. 4, there are substantially more BfuABC [FeFe]-hydrogenases than just the five sub-types identified in Hydrogenedentota. Phylum Bacillota (synonym Firmicutes) contains the most diverse group A3 [FeFe]-hydrogenases. Intriguingly, group A3 [FeFe]-hydrogenases do not appear to be strictly vertically inherited, as revealed by phylogenetic and taxonomic conflicts both at and below the phylum level (Fig. 4). For example, the BfuABC homologs from Bacillota are not all clustered together but are scattered, forming clusters with those from other 26 bacterial phyla. These results suggest that inter-phylum horizontal gene transfer has been a major driving force in the evolution of group A3 [FeFe]-hydrogenases.

Fig 4 Phylogenetic analysis of Hydrogenedentota BfuABC proteins in relation to those in other bacterial phyla. The strip with different colors in the outer ring indicates different bacterial phyla. Sub-types I to V of BfuABC are marked in blue, purple, red, pink, and green, respectively. The bootstraps of over 90% are shown in brown branches. The genome id, gene cluster types of BfuABC homologs, and domains of BfuB subunit are displayed sequentially at the end of the branches. “X” represents genes that encode subunits or proteins other than BfuABCDT. The gene orders and orientations and BfuB domains are marked as in Fig. 3.

Circular phylogenetic tree of BfuABC homologs depicts taxonomy at the phylum level with various phyla color-coded, including Acidobacteriota, Actinomycetota, Aquificota, Armatimonadota, Bacillota, Bacteroidota, Chloroflexota, Cyanobacteria, and others.

Moreover, the five sub-types appear to be distributed patchily among Hydrogenedentota. Although this may be attributed in part to the incompleteness of the genome sequences, it more likely reflects the results of gene transfer and loss, two key drivers of gene changes in the evolution of Hydrogenedentota, as revealed by amalgamated likelihood estimation (ALE) (Fig. S8; Data Set S1, Sheet 10). To shed more light on the evolution of electron bifurcating [FeFe]-hydrogenases from Hydrogenedentota, we then examined the individual phylogenetic clusters containing the five sub-types of BfuABC homologs and their nearest neighbors. As shown in Fig. 5, the five BfuABC sub-types appear to have been initially acquired from Bacillota through separate horizontal gene transfer events, and thus, each of them exhibits unique evolutionary features.

Fig 5 Phylogenetic analysis of five sub-types of BfuABC proteins in Hydrogenedentota and other bacterial phyla. The radiation tree was constructed based on the concatenate of the three subunits “ABC” of 195 and 2,144 BfuABC homologs from the genomes of Hydrogenedentota and the other 26 bacterial phyla, respectively. Sub-types I to V of BfuABC are marked in blue, purple, red, pink, and green, respectively, and numbers in parentheses represent the number of hydrogenases in corresponding sub-types in the radiation tree. More information, shown in a circular form of this tree, is available in Fig. 4. Panels (a–e) are enlarged regions I to V of the radiation tree, respectively, and show the detailed relationships between the sub-types I to V of BfuABC in Hydrogenedentota and those in other closely related phyla. “1,” phylum affiliation; “2,” genome id; “3,” contig id or strain name in other phyla; “4,” clade in Hydrogenedentota. Clades 6-1 and 6-2 are two sub-clades of Clade 6, as illustrated in Fig. 2; Fig. S3. Nodes “ND” and “YD” in panel (a) represent homologs in the corresponding clusters with and without a BfuD subunit, respectively, and “H” in “ND_H” means “Hydrogenedentota.” Nodes “UT” and “DT” in panel (e) represent the location of bfuT either upstream or downstream of bfuA in the gene clusters. The non-redundant Hydrogenedentota genomes at strain levels are indicated by red stars. The bootstraps are labeled with black dots when ≥90%.

Phylogenetic tree of BfuABC homologs depicts taxonomy at the phylum level, with color-coded branches representing different phyla such as Actinomycetota, Bacteroidota, Cyanobacteria, Firmicutes, and others.

Sub-type I BfuABC homologs from Spirochaetota, Pseudomonadota, Planctomycetota, Verrucomicrobiota, and Kiritimatiellota, which are most closely related to that from Hydrogenedentota, generally lack BfuD (node ND in Fig. 5a). Unexpectedly, several homologs from Hydrogenedentota in Clade 6 also lack BfuD (node ND_H in Fig. 5a; Fig. S5). Interestingly, BfuD is present in a distantly related homolog from Syntrophus gentianae DSM 8423 and absent in a homolog from Desulfotignum balticum DSM 7044. Both of the strains belong to phylum Thermodesulfobacteriota, suggesting that BfuD might be susceptible to loss during evolution. Similarly, as shown in Fig. 4 and 5c, the ancestral sub-type III BfuABC homolog might contain a BfuD subunit and subsequently lost it in Hydrogenedentota and some lineages in Acidobacteriota and Planctomycetota (Thermogutta terrifontis R1) since this subunit is encoded by both Planctomycetota (Anaerohalosphaera lusitana ST-NAGAB-D1) and Ca. Omnitrophota (Velamenicoccus archaeovorus LiM) (Fig. 4 and 5c). All closely related sub-type II BfuABC homologs from Acidobacteriota, Actinomycetota, and Planctomycetota lack the “B3,” “B4,” and “B5” FeS clusters (Fig. 4 and 5b). Both the least and most related species are from the phylum of Acidobacteriota (Thermotomaculum hydrothermale AC55 and Holophaga foetida DSM 6591), suggesting that sub-type II BfuABC homologs in Hydrogenedentota and Acidobacteriota were inherited from a common ancestor. Horizontal gene transfer of sub-type II BfuABC within Hydrogenedentota may occur frequently, especially in Clade 1, where the BfuABC homologs are not clustered together (Fig. 5b).

As shown in Fig. 4, sub-types VI and V BfuABC homologs reside in the same large cluster, indicating their close evolutionary relationship. These two sub-types are featured by a bfuT, an extra gene encoding a group C3 hydrogen-sensing [FeFe]-hydrogenase. bfuT is often located upstream of bfuC in a sub-type IV gene cluster, which mainly exists in the genomes of Clade 6 (Fig. 5d). However, two exceptions were observed. First, bfuT is downstream of and adjacent to bfuA in the two genomes of Clade 1 (GCA_019695095.1 and GCA_003557825.1) (Fig. S5d). All of the most closely related BfuABC homologs of the two genomes are from species in phylum Thermodesulfobacteriota. No bfuT was found downstream of bfuA in these thermodesulfobacteria, except for Desulfobulbus elongatus DSM 2908 and Desulfobulbus propionicus DSM 2032 (Fig. 4 and 5d). Second, the bfuT and bfuA genes in the D. elongatus DSM 2908 and D. propionicus DSM 2032 genomes are separated by three genes (Fig. 4). These observations suggest that BfuABC homologs from Thermodesulfobacteriota and Hydrogenedentota share a common ancestor, which was acquired through horizontal gene transfer before the divergence of the two phyla, and exhibit separate evolution trajectories, resulting in variation in the location of bfuT in relation to bfuA. Puzzlingly, one of the two evolutionary paths (node DT in Fig. 5d) appears to have flourished in descendants (Clade 6), while the other (node UT in Fig. 5d) does not in Hydrogenedentota. In contrast, sub-type V gene clusters are inherited vertically in both Clades 1 and 6. The organization of a sub-type V gene cluster from Hydrogenedentota is consistent with that from phylum Calditrichota (Fig. 4 and 5e), represented by Caldithrix abyssi DSM 13497. However, three distantly related BfuABC homologs from phylum Bacteroidota (Perlabentimonas gracilis M08_MB, Tenuifilum thalassicum 38H-str, and Williamwhitmania taraxaci A7P-90m) have an additional bfuT gene upstream of bfuC in addition to one downstream of bfuA. The physiological and ecological implications of the presence of two bfuT genes in a single gene cluster are unclear but may concern habitat adaptation. We speculate that both Hydrogenedentota and Calditrichota have lost the second bfuT after initially obtaining it through ancient horizontal gene transfer from Bacteroidota. Clearly, BfuT is also susceptible to gain and loss during the evolution of both sub-type IV and V BfuABC homologs.

Electron bifurcating [FeFe]-hydrogenases increase the versatility of Hydrogenedentota in energy conservation

Genes encoding bifurcating [FeFe]-hydrogenases of multiple sub-types often co-exist in a single genome. For example, CSSed162cmA_463 (Hydrogenedentota, GCA_007118765.1) has all the five sub-types (Fig. 3 and 6), and Anoxybacter fermentans DY22613 (Bacillota, GCF_003991135.1) has three (Fig. 4). Interestingly, nearly all of the strains containing sub-type II BfuABC complexes, which lack the electron bifurcating ability, have at least one of the other four sub-types of putative bifurcating [FeFe]-hydrogenases (Fig. 6). Moreover, Clade 6-1 and Clade 6-2 genomes are capable of translocating protons or Na+ ions across the cytoplasmic membrane using different mechanisms for energy generation through ATP synthase. Clade 6-1 genomes usually possess complexes I (NADH-quinone oxidoreductase), II (succinate dehydrogenase), and IV (cytochrome bd oxidase or cyt-bd oxygen reductase) for transporting electrons successively while translocating protons across the membrane at the same time (Fig. 2). On the other hand, Clade 6-2 genomes encode a Rnf complex, a Na+-dependent ferredoxin:NAD+ oxidoreductase, instead of the above three respiratory chain complexes. Rnf functions to generate a transmembrane electrochemical Na+ gradient by coupling the reduction of NAD+ with reduced ferredoxin. Interestingly, all of the genomes lacking the gene encoding BfuD of sub-type I [FeFe]-hydrogenase belong to Clade 6-1 (node ND_H in Fig. 5a), and those having the BfuD-encoding gene belong to Clade 6-2 (node YD; Fig. 5a). Therefore, unlike Clade 6-1, which uses respiratory chain complexes, Clade 6-2 may couple reactions catalyzed by Rnf with those by BfuABCD homologs in generating the transmembrane electrochemical gradient and maintaining intracellular redox balance. And the tetrameric complex of sub-type I BfuABCD may facilitate the oxidation of NADH generated by Rnf more efficiently than the trimeric one due to the presence of BfuD functioning in rapid regulatory response to fluctuating environmental conditions (50). In other words, the interaction between reactions catalyzed by BfuABCD and those by Rnf is beneficial or even vital for Hydrogenedentota to obtain an adequate energy supply in adapting to changing environments.

Fig 6 Phylogenetic tree of the concatenate of the three subunits “ABC” from the 195 nearly complete group A3 [FeFe]-hydrogenases in Hydrogenedentota. Five sub-types (I to V) of group A3 [FeFe]-hydrogenases, i.e., BfuABDC, BfuABC, BfuCAB, BfuABCT, and BfuTABC, are shown by blue, purple, red, pink, and green strips in genome id, respectively. Domain structures of BfuA, BfuB, and BfuC are shown outside sequentially. For a genome containing the gene cluster coding for BfuABC as well as any of those coding for BfuABDC, BfuCAB, BfuABCT, and BfuTABC, the coexistence of the gene clusters is indicated by dotted lines in color.

Circular phylogenetic tree of BfuABC homologs with color-coded branches and concentric rings representing various taxonomic levels. Inner rings contain detailed connections between specific groups. Tree scale bar indicates branch length.

Conclusions

Hydrogenedentota is a group of globally distributed and low-abundance bacteria living in various habitats. Based on the analysis of currently available genomes, members of this phylum are predicted to be Gram-negative, rod-shaped, motile, facultative anaerobes with a heterotrophic lifestyle. Most of these organisms have evolved a non-canonical electron bifurcation pathway, in addition to SLP and ETP, as a strategy for energy conservation. Specifically, they employ the electron bifurcating [FeFe]-hydrogenase (group A3) in hydrogen metabolism, a process presumably vital for their survival in diverse habitats. Five independent pathways have been proposed for the evolution of the gene clusters coding for BfuABC in Hydrogenedentota, and these gene clusters are predicted to have been subjected to extensive intra- and interphyla horizontal gene transfers and undergone gene losses and gains.

MATERIALS AND METHODS

16S rRNA gene amplicon data set

Sample processing, sequencing, and core amplicon data analysis were performed by the EMP (www.earthmicrobiome.org), and all amplicon sequence data and metadata are publicly available through the EMP data portal (http://qiita.microbio.me/emp) (22). The global distribution of Hydrogenedentota was analyzed as described previously (51).

Hydrogenedentota genome collection

Published MAGs or SAGs of Hydrogenedentota (2, 9, 52–54), verified by GTDB-Tk (v2.0.0) (55), were retrieved from the NCBI Assembly database (ftp://ftp.ncbi.nlm.nih.gov/genomes/genbank/bacteria/, July 2023), GTDB (https://data.gtdb.ecogenomic.org/releases/release207/), GEM catalog (https://portal.nersc.gov/GEM) (56), Figshare (https://doi.org/10.6084/m9.figshare.c.5564844.v1) (54), and ENA (https://www.ebi.ac.uk/ena/browser/view/PRJEB45951 and https://www.ebi.ac.uk/ena/browser/view/PRJEB35712) (9, 53). All these MAGs and SAGs were collectively referred to as “genomes” in this article for simplicity.

Genome completeness assessment, dereplication, and phylogeny

The completeness, potential contamination, and strain heterogeneity of these genomes were evaluated via CheckM (v1.1.9) with lineage-specific marker genes (57). Genome size was estimated as described by Chen et al. (58). One hundred seventy-nine medium- to high-quality (completeness ≥50% and contamination ≤10%) genomes were retained in downstream analysis (Data Set S1, Sheet 2). To reduce redundancy, the 179 genomes were dereplicated at 99% average nucleotide identity using dRep (v2.3.2; option: -comp 50 -con 5 -sa 0.99) (59), resulting in a total of 100 strain-level genomes. Phylogenetic analysis of the above genomes based on 120 bacterial marker proteins were conducted using the “identify” and “align” steps in GTDB-Tk (v2.0.0) (55). Maximum likelihood phylogeny was inferred with FastTree (v2.1.11) in the WAG+GAMMA model (60). The phylogenetic tree was then visualized and annotated on the iTOL (https://itol.embl.de/) (61).

Function annotation

Open reading frames (ORFs) in the above genomes were predicted using Prokka (v1.14.6) (62) with default parameters. All ORFs were annotated using the Kyoto Encyclopedia of Genes and Genomes database with GhostKOALA (63) and against the Pfam (release 35.0) (64), NCBIfam (release 11.0) (65), and SUPERFAMILY (release 1.75) (66) HMM models using Interproscan (v5.62-94.0) (67). All genomes were also submitted to the METABOLIC v2.0 annotation pipeline for functional annotation (68). Detailed descriptions of the cell structure, motility, metabolic reconstruction, hydrogenase identification and classification, and ALE can be found in Text S1.

ACKNOWLEDGMENTS

This work was supported by the Second Tibetan Plateau Scientific Expedition and Research Program (SQ2021QZKK0302), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB42000000), the National Natural Science Foundation of China (42376238 and 32393970), and the Major Research Plan of the National Natural Science Foundation of China (91751000).

DATA AVAILABILITY

All data generated or analyzed during this study are included in this published article and its supplementary files.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/msystems.00999-24.

10.1128/msystems.00999-24.SuF1 Data Set S1 msystems.00999-24-s0001.xlsx

Metagenomic analysis results.

10.1128/msystems.00999-24.SuF2 Text S1 msystems.00999-24-s0002.pdf

Supplemental methods, notes, and legends.

10.1128/msystems.00999-24.SuF3 Fig. S1 msystems.00999-24-s0003.pdf

Distribution of Hydrogenedentota around the globe.

10.1128/msystems.00999-24.SuF4 Fig. S2 msystems.00999-24-s0004.pdf

Relative abundance of Hydrogenedentota in different habits.

10.1128/msystems.00999-24.SuF5 Fig. S3 msystems.00999-24-s0005.pdf

Phylogenomic tree of Hydrogenedentota.

10.1128/msystems.00999-24.SuF6 Fig. S4 msystems.00999-24-s0006.pdf

Classification of [FeFe]-hydrogenases and [NiFe]-hydrogenases.

10.1128/msystems.00999-24.SuF7 Fig. S5 msystems.00999-24-s0007.pdf

Organization of 195 gene clusters coding for BfuABC homologs.

10.1128/msystems.00999-24.SuF8 Fig. S6 msystems.00999-24-s0008.pdf

Unrooted radiation tree of the subunits BfuA, BfuB, and BfuC.

10.1128/msystems.00999-24.SuF9 Fig. S7 msystems.00999-24-s0009.pdf

Schematic diagram of H-cluster motifs.

10.1128/msystems.00999-24.SuF10 Fig. S8 msystems.00999-24-s0010.pdf

Events of inferred duplications, intra-phylum transfers, losses and gene originations.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.
==== Refs
REFERENCES

1 Li L, Guenzennec J, Nichols P, Henry P, Yanagibayashi M, Kato C. 1999. Microbial diversity in Nankai trough sediments at a depth of 3,843 m. J Oceanogr 55 :635–642. doi:10.1023/A:1007897020042
2 Rinke C, Schwientek P, Sczyrba A, Ivanova NN, Anderson IJ, Cheng JF, Darling A, Malfatti S, Swan BK, Gies EA, Dodsworth JA, Hedlund BP, Tsiamis G, Sievert SM, Liu WT, Eisen JA, Hallam SJ, Kyrpides NC, Stepanauskas R, Rubin EM, Hugenholtz P, Woyke T. 2013. Insights into the phylogeny and coding potential of microbial dark matter. Nature 499 :431–437. doi:10.1038/nature12352 23851394
3 Trüper HG, Euzéby JP. 2009. International code of nomenclature of prokaryotes. Appendix 9: orthography. Int J Syst Evol Microbiol 59 :2107–2113. doi:10.1099/ijs.0.016741-0 19567558
4 Parks DH, Chuvochina M, Chaumeil PA, Rinke C, Mussig AJ, Hugenholtz P. 2020. A complete domain-to-species taxonomy for Bacteria and Archaea. Nat Biotechnol 38 :1079–1086. doi:10.1038/s41587-020-0501-8 32341564
5 Din GMU, Du Z, Zhang H, Zhao S, Liu T, Chen W, Gao L. 2021. Effects of Tilletia foetida on microbial communities in the rhizosphere soil of wheat seeds coated with different concentrations of Jianzhuang. Microb Ecol 82 :736–745. doi:10.1007/s00248-021-01696-w 33527233
6 Rasuk MC, Fernández AB, Kurth D, Contreras M, Novoa F, Poiré D, Farías ME. 2016. Bacterial diversity in microbial mats and sediments from the Atacama Desert. Microb Ecol 71 :44–56. doi:10.1007/s00248-015-0649-9 26224164
7 Pereira AD, Leal CD, Dias MF, Etchebehere C, Chernicharo CAL, de Araújo JC. 2014. Effect of phenol on the nitrogen removal performance and microbial community structure and composition of an anammox reactor. Bioresour Technol 166 :103–111. doi:10.1016/j.biortech.2014.05.043 24907569
8 Zhang Y, Wu G, Jiang H, Yang J, She W, Khan I, Li W. 2018. Abundant and rare microbial biospheres respond differently to environmental and spatial factors in Tibetan hot springs. Front Microbiol 9 :2096. doi:10.3389/fmicb.2018.02096 30283408
9 Martínez-Pérez C, Greening C, Bay SK, Lappan RJ, Zhao Z, De Corte D, Hulbe C, Ohneiser C, Stevens C, Thomson B, Stepanauskas R, González JM, Logares R, Herndl GJ, Morales SE, Baltar F. 2022. Phylogenetically and functionally diverse microorganisms reside under the Ross Ice Shelf. Nat Commun 13 :117. doi:10.1038/s41467-021-27769-5 35013291
10 Yakimov MM, Cappello S, Crisafi E, Tursi A, Savini A, Corselli C, Scarfi S, Giuliano L. 2006. Phylogenetic survey of metabolically active microbial communities associated with the deep-sea coral Lophelia pertusa from the Apulian plateau, Central Mediterranean Sea. Deep Sea Res Part I 53 :62–75. doi:10.1016/j.dsr.2005.07.005
11 Leung PM, Daebeler A, Chiri E, Hanchapola I, Gillett DL, Schittenhelm RB, Daims H, Greening C. 2022. A nitrite-oxidising bacterium constitutively consumes atmospheric hydrogen. ISME J 16 :2213–2219. doi:10.1038/s41396-022-01265-0 35752717
12 Greening C, Grinter R. 2022. Microbial oxidation of atmospheric trace gases. Nat Rev Microbiol 20 :513–528. doi:10.1038/s41579-022-00724-x 35414013
13 Morra S. 2022. Fantastic [FeFe]-hydrogenases and where to find them. Front Microbiol 13 :853626. doi:10.3389/fmicb.2022.853626 35308355
14 Poudel S, Tokmina-Lukaszewska M, Colman DR, Refai M, Schut GJ, King PW, Maness P-C, Adams MWW, Peters JW, Bothner B, Boyd ES. 2016. Unification of [FeFe]-hydrogenases into three structural and functional groups. Biochim Biophys Acta 1860 :1910–1921. doi:10.1016/j.bbagen.2016.05.034 27241847
15 Greening C, Biswas A, Carere CR, Jackson CJ, Taylor MC, Stott MB, Cook GM, Morales SE. 2016. Genomic and metagenomic surveys of hydrogenase distribution indicate H2 is a widely utilised energy source for microbial growth and survival. ISME J 10 :761–777. doi:10.1038/ismej.2015.153 26405831
16 Shima S, Pilak O, Vogt S, Schick M, Stagni MS, Meyer-Klaucke W, Warkentin E, Thauer RK, Ermler U. 2008. The crystal structure of [Fe]-hydrogenase reveals the geometry of the active site. Science 321 :572–575. doi:10.1126/science.1158978 18653896
17 Vignais PM, Billoud B. 2007. Occurrence, classification, and biological function of hydrogenases: an overview. Chem Rev 107 :4206–4272. doi:10.1021/cr050196r 17927159
18 Leung PM, Grinter R, Tudor-Matthew E, Lingford JP, Jimenez L, Lee HC, Milton M, Hanchapola I, Tanuwidjaya E, Kropp A, Peach HA, Carere CR, Stott MB, Schittenhelm RB, Greening C. 2024. Trace gas oxidation sustains energy needs of a thermophilic archaeon at suboptimal temperatures. Nat Commun 15 :3219. doi:10.1038/s41467-024-47324-2 38622143
19 Peters JW, Schut GJ, Boyd ES, Mulder DW, Shepard EM, Broderick JB, King PW, Adams MWW. 2015. [FeFe]- and [NiFe]-hydrogenase diversity, mechanism, and maturation. Biochim Biophys Acta 1853 :1350–1369. doi:10.1016/j.bbamcr.2014.11.021 25461840
20 Greening C, Cabotaje PR, Valentin Alvarado LE, Leung PM, Land H, Rodrigues-Oliveira T, Ponce-Toledo RI, Senger M, Klamke MA, Milton M, Lappan R, Mullen S, West-Roberts J, Mao J, Song J, Schoelmerich M, Stairs CW, Schleper C, Grinter R, Spang A, Banfield JF, Berggren G. 2024. Minimal and hybrid hydrogenases are active from archaea. Cell 187 :3357–3372. doi:10.1016/j.cell.2024.05.032 38866018
21 Weiss MC, Sousa FL, Mrnjavac N, Neukirchen S, Roettger M, Nelson-Sathi S, Martin WF. 2016. The physiology and habitat of the last universal common ancestor. Nat Microbiol 1 :16116. doi:10.1038/nmicrobiol.2016.116 27562259
22 Thompson LR, Sanders JG, McDonald D, Amir A, Ladau J, Locey KJ, Prill RJ, Tripathi A, Gibbons SM, Ackermann G, et al. . 2017. A communal catalogue reveals Earth's multiscale microbial diversity. Nature 551 :457–463. doi:10.1038/nature24621 29088705
23 Irazoki O, Hernandez SB, Cava F. 2019. Peptidoglycan muropeptides: release, perception, and functions as signaling molecules. Front Microbiol 10 :500. doi:10.3389/fmicb.2019.00500 30984120
24 Henriques AO, Glaser P, Piggot PJ, Moran Jr CP. 1998. Control of cell shape and elongation by the rodA gene in Bacillus subtilis. Mol Microbiol 28 :235–247. doi:10.1046/j.1365-2958.1998.00766.x 9622350
25 Wang S, Furchtgott L, Huang KC, Shaevitz JW. 2012. Helical insertion of peptidoglycan produces chiral ordering of the bacterial cell wall. Proc Natl Acad Sci U S A 109 :E595–E604. doi:10.1073/pnas.1117132109 22343529
26 Lobo SAL, Melo AMP, Carita JN, Teixeira M, Saraiva LM. 2007. The anaerobe Desulfovibrio desulfuricans ATCC 27774 grows at nearly atmospheric oxygen levels. FEBS Lett 581 :433–436. doi:10.1016/j.febslet.2006.12.053 17239374
27 Bergdoll L, Ten Brink F, Nitschke W, Picot D, Baymann F. 2016. From low- to high-potential bioenergetic chains: thermodynamic constraints of Q-cycle function. Biochim Biophys Acta 1857 :1569–1579. doi:10.1016/j.bbabio.2016.06.006 27328272
28 Schoepp-Cothenet B, Lieutaud C, Baymann F, Verméglio A, Friedrich T, Kramer DM, Nitschke W. 2009. Menaquinone as pool quinone in a purple bacterium. Proc Natl Acad Sci U S A 106 :8549–8554. doi:10.1073/pnas.0813173106 19429705
29 Melin F, Hellwig P. 2020. Redox properties of the membrane proteins from the respiratory chain. Chem Rev 120 :10244–10297. doi:10.1021/acs.chemrev.0c00249 32820893
30 Mahendrarajah TA, Moody ERR, Schrempf D, Szánthó LL, Dombrowski N, Davín AA, Pisani D, Donoghue PCJ, Szöllősi GJ, Williams TA, Spang A. 2023. ATP synthase evolution on a cross-braced dated tree of life. Nat Commun 14 :7456. doi:10.1038/s41467-023-42924-w 37978174
31 Hwang Y, Schulze-Makuch D, Arens FL, Saenz JS, Adam PS, Sager C, Bornemann TLV, Zhao W, Zhang Y, Airo A, Schloter M, Probst AJ. 2021. Leave no stone unturned: individually adapted xerotolerant Thaumarchaeota sheltered below the boulders of the Atacama Desert hyperarid core. Microbiome 9 :234. doi:10.1186/s40168-021-01177-9 34836555
32 Zhong H, Lehtovirta-Morley L, Liu J, Zheng Y, Lin H, Song D, Todd JD, Tian J, Zhang XH. 2020. Novel insights into the Thaumarchaeota in the deepest oceans: their metabolism and potential adaptation mechanisms. Microbiome 8 :78. doi:10.1186/s40168-020-00849-2 32482169
33 Wang B, Qin W, Ren Y, Zhou X, Jung MY, Han P, Eloe-Fadrosh EA, Li M, Zheng Y, Lu L, Yan X, Ji J, Liu Y, Liu L, Heiner C, Hall R, Martens-Habbena W, Herbold CW, Rhee SK, Bartlett DH, Huang L, Ingalls AE, Wagner M, Stahl DA, Jia Z. 2019. Expansion of Thaumarchaeota habitat range is correlated with horizontal transfer of ATPase operons. ISME J 13 :3067–3079. doi:10.1038/s41396-019-0493-x 31462715
34 Birrell JA, Rodríguez-Maciá P, Reijerse EJ, Martini MA, Lubitz W. 2021. The catalytic cycle of [FeFe] hydrogenase: a tale of two sites. Coord Chem Rev 449 :214191. doi:10.1016/j.ccr.2021.214191
35 Schut GJ, Haja DK, Feng X, Poole FL, Li H, Adams MWW. 2022. An abundant and diverse new family of electron bifurcating enzymes with a non-canonical catalytic mechanism. Front Microbiol 13 :946711. doi:10.3389/fmicb.2022.946711 35875533
36 Di Leonardo PF, Antonicelli G, Agostino V, Re A. 2022. Genome-scale mining of acetogens of the genus Clostridium unveils distinctive traits in [FeFe]- and [NiFe]-hydrogenase content and maturation. Microbiol Spectr 10 :e0101922. doi:10.1128/spectrum.01019-22 35735976
37 Feng X, Schut GJ, Haja DK, Adams MWW, Li HL. 2022. Structure and electron transfer pathways of an electron-bifurcating NiFe-hydrogenase. Sci Adv 8 :eabm7546. doi:10.1126/sciadv.abm7546 35213221
38 Vignais PM, Billoud B, Meyer J. 2001. Classification and phylogeny of hydrogenases. FEMS Microbiol Rev 25 :455–501. doi:10.1111/j.1574-6976.2001.tb00587.x 11524134
39 Song Y, Peisach D, Pioszak AA, Xu Z, Ninfa AJ. 2004. Crystal structure of the C-terminal domain of the two-component system transmitter protein nitrogen regulator II (NRII; NtrB), regulator of nitrogen assimilation in Escherichia coli. Biochemistry 43 :6670–6678. doi:10.1021/bi049474r 15157101
40 Bellon S, Parsons JD, Wei Y, Hayakawa K, Swenson LL, Charifson PS, Lippke JA, Aldape R, Gross CH. 2004. Crystal structures of Escherichia coli topoisomerase IV ParE subunit (24 and 43 kilodaltons): a single residue dictates differences in novobiocin potency against topoisomerase IV and DNA gyrase. Antimicrob Agents Chemother 48 :1856–1864. doi:10.1128/AAC.48.5.1856-1864.2004 15105144
41 Søndergaard D, Pedersen CNS, Greening C. 2016. HydDB: a web tool for hydrogenase classification and analysis. Sci Rep 6 :34212. doi:10.1038/srep34212 27670643
42 Hefti MH, Françoijs K-J, de Vries SC, Dixon R, Vervoort J. 2004. The PAS fold. A redefinition of the PAS domain based upon structural prediction. Eur J Biochem 271 :1198–1208. doi:10.1111/j.1432-1033.2004.04023.x 15009198
43 Buckel W, Thauer RK. 2018. Flavin-based electron bifurcation, a new mechanism of biological energy coupling. Chem Rev 118 :3862–3886. doi:10.1021/acs.chemrev.7b00707 29561602
44 Feng X, Schut GJ, Lipscomb GL, Li HL, Adams MWW. 2021. Cryoelectron microscopy structure and mechanism of the membrane-associated electron-bifurcating flavoprotein Fix/EtfABCX. Proc Natl Acad Sci U S A 118 :e2016978118. doi:10.1073/pnas.2016978118 33372143
45 Appel L, Willistein M, Dahl C, Ermler U, Boll M. 2021. Functional diversity of prokaryotic HdrA(BC) modules: role in flavin-based electron bifurcation processes and beyond. Biochim Biophys Acta Bioenerg 1862 :148379. doi:10.1016/j.bbabio.2021.148379 33460586
46 Furlan C, Chongdar N, Gupta P, Lubitz W, Ogata H, Blaza JN, Birrell JA. 2022. Structural insight on the mechanism of an electron-bifurcating [FeFe] hydrogenase. Elife 11 :e79361. doi:10.7554/eLife.79361 36018003
47 Yu H, Schut GJ, Haja DK, Adams MWW, Li H. 2021. Evolution of complex I-like respiratory complexes. J Biol Chem 296 :100740. doi:10.1016/j.jbc.2021.100740 33957129
48 Losey NA, Poudel S, Boyd ES, McInerney MJ. 2020. The beta subunit of non-bifurcating NADH-dependent [FeFe]-hydrogenases differs from those of multimeric electron-bifurcating [FeFe]-hydrogenases. Front Microbiol 11 :1109. doi:10.3389/fmicb.2020.01109 32625172
49 Losey NA, Mus F, Peters JW, Le HM, McInerney MJ. 2017. Syntrophomonas wolfei uses an NADH-dependent, ferredoxin-independent [FeFe]-hydrogenase to reoxidize NADH. Appl Environ Microbiol 83 :e01335-17. doi:10.1128/AEM.01335-17 28802265
50 Jacq-Bailly A, Benvenuti M, Payne N, Kpebe A, Felbek C, Fourmond V, Léger C, Brugna M, Baffert C. 2020. Electrochemical characterization of a complex FeFe hydrogenase, the electron-bifurcating Hnd from Desulfovibrio fructosovorans. Front Chem 8 :573305. doi:10.3389/fchem.2020.573305 33490032
51 Zheng X, Dai X, Zhu Y, Yang J, Jiang H, Dong H, Huang L. 2022. (Meta)genomic analysis reveals diverse energy conservation strategies employed by globally distributed Gemmatimonadota. mSystems 7 :e0022822. doi:10.1128/msystems.00228-22 35913193
52 Zhang H, Wang M, Wang H, Chen H, Cao L, Zhong Z, Lian C, Zhou L, Li C. 2022. Metagenome sequencing and 768 microbial genomes from cold seep in South China Sea. Sci Data 9 :480. doi:10.1038/s41597-022-01586-x 35933411
53 Paoli L, Ruscheweyh H-J, Forneris CC, Hubrich F, Kautsar S, Bhushan A, Lotti A, Clayssen Q, Salazar G, Milanese A, et al. . 2022. Biosynthetic potential of the global ocean microbiome. Nature 607 :111–118. doi:10.1038/s41586-022-04862-3 35732736
54 Nishimura Y, Yoshizawa S. 2022. The OceanDNA MAG catalog contains over 50,000 prokaryotic genomes originated from various marine environments. Sci Data 9 :305. doi:10.1038/s41597-022-01392-5 35715423
55 Chaumeil PA, Mussig AJ, Hugenholtz P, Parks DH. 2022. GTDB-Tk v2: memory friendly classification with the genome taxonomy database. Bioinformatics 38 :5315–5316. doi:10.1093/bioinformatics/btac672 36218463
56 Nayfach S, Roux S, Seshadri R, Udwary D, Varghese N, Schulz F, Wu D, Paez-Espino D, Chen I-M, Huntemann M, et al. . 2021. A genomic catalog of Earth's microbiomes. Nat Biotechnol 39 :499–509. doi:10.1038/s41587-020-0718-6 33169036
57 Parks DH, Imelfort M, Skennerton CT, Hugenholtz P, Tyson GW. 2015. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res 25 :1043–1055. doi:10.1101/gr.186072.114 25977477
58 Chen M-Y, Teng W-K, Zhao L, Hu C-X, Zhou Y-K, Han B-P, Song L-R, Shu W-S. 2021. Comparative genomics reveals insights into cyanobacterial evolution and habitat adaptation. ISME J 15 :211–227. doi:10.1038/s41396-020-00775-z 32943748
59 Olm MR, Brown CT, Brooks B, Banfield JF. 2017. dRep: a tool for fast and accurate genomic comparisons that enables improved genome recovery from metagenomes through de-replication. ISME J 11 :2864–2868. doi:10.1038/ismej.2017.126 28742071
60 Price MN, Dehal PS, Arkin AP. 2010. FastTree 2--approximately maximum-likelihood trees for large alignments. PLoS One 5 :e9490. doi:10.1371/journal.pone.0009490 20224823
61 Letunic I, Bork P. 2021. Interactive tree of life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res 49 :W293–W296. doi:10.1093/nar/gkab301 33885785
62 Seemann T. 2014. Prokka: rapid prokaryotic genome annotation. Bioinformatics 30 :2068–2069. doi:10.1093/bioinformatics/btu153 24642063
63 Kanehisa M, Sato Y, Morishima K. 2016. BlastKOALA and GhostKOALA: KEGG tools for functional characterization of genome and metagenome sequences. J Mol Biol 428 :726–731. doi:10.1016/j.jmb.2015.11.006 26585406
64 Mistry J, Chuguransky S, Williams L, Qureshi M, Salazar GA, Sonnhammer ELL, Tosatto SCE, Paladin L, Raj S, Richardson LJ, Finn RD, Bateman A. 2021. Pfam: the protein families database in 2021. Nucleic Acids Res 49 :D412–D419. doi:10.1093/nar/gkaa913 33125078
65 Li W, O’Neill KR, Haft DH, DiCuccio M, Chetvernin V, Badretdin A, Coulouris G, Chitsaz F, Derbyshire MK, Durkin AS, Gonzales NR, Gwadz M, Lanczycki CJ, Song JS, Thanki N, Wang J, Yamashita RA, Yang M, Zheng C, Marchler-Bauer A, Thibaud-Nissen F. 2021. RefSeq: expanding the prokaryotic genome annotation pipeline reach with protein family model curation. Nucleic Acids Res 49 :D1020–D1028. doi:10.1093/nar/gkaa1105 33270901
66 Gough J, Karplus K, Hughey R, Chothia C. 2001. Assignment of homology to genome sequences using a library of hidden Markov models that represent all proteins of known structure. J Mol Biol 313 :903–919. doi:10.1006/jmbi.2001.5080 11697912
67 Jones P, Binns D, Chang HY, Fraser M, Li W, McAnulla C, McWilliam H, Maslen J, Mitchell A, Nuka G, Pesseat S, Quinn AF, Sangrador-Vegas A, Scheremetjew M, Yong SY, Lopez R, Hunter S. 2014. InterProScan 5: genome-scale protein function classification. Bioinformatics 30 :1236–1240. doi:10.1093/bioinformatics/btu031 24451626
68 Zhou Z, Tran PQ, Breister AM, Liu Y, Kieft K, Cowley ES, Karaoz U, Anantharaman K. 2022. METABOLIC: high-throughput profiling of microbial genomes for functional traits, metabolism, biogeochemistry, and community-scale functional networks. Microbiome 10 :33. doi:10.1186/s40168-021-01213-8 35172890
