Molecular Hydrogen, a Neglected Key Driver of Soil Biogeochemical Processes

Abstract

The atmosphere of the early Earth is hypothesized to have been rich in reducing gases such as hydrogen (H₂). H₂ has been proposed as the first electron donor leading to ATP synthesis due to its ubiquity throughout the biosphere as well as its ability to easily diffuse through microbial cells and its low activation energy requirement. Even today, hydrogenase enzymes enabling the production and oxidation of H₂ are found in thousands of genomes spanning the three domains of life across aquatic, terrestrial, and even host-associated ecosystems. Even though H₂ has already been proposed as a universal growth and maintenance energy source, its potential contribution as a driver of biogeochemical cycles has received little attention. Here, we bridge this knowledge gap by providing an overview of the classification, distribution, and physiological role of hydrogenases. Distribution of these enzymes in various microbial functional groups and recent experimental evidence are finally integrated to support the hypothesis that H₂-oxidizing microbes are keystone species driving C cycling along O₂ concentration gradients found in H₂-rich soil ecosystems. In conclusion, we suggest focusing on the metabolic flexibility of H₂-oxidizing microbes by combining community-level and individual-level approaches aiming to decipher the impact of H₂ on C cycling and the C-cycling potential of H₂-oxidizing microbes, via both culture-dependent and culture-independent methods, to give us more insight into the role of H₂ as a driver of biogeochemical processes.

Keywords: H2 oxidation; anaerobic processes; biogeochemical processes; carbon cycle; environmental microbiology; hydrogen; soil.

FIG 1

Consensus tree of 25 [Fe]-hydrogenase sequences, with an alignment length of 389 amino acids. Protein sequences were imported from a study by Greening et al. (10), which included nonredundant putative hydrogenase catalytic subunits from cultured and environmental metagenomes sourced from NCBI RefSeq and Joint Genome Institute, Microbial Dark Matter (JGI, MDM) databases. More details are found in the article by Greening et al. (10). See Fig. S1 for complete taxa names and a more detailed phylogenetic tree. Hydrogenase amino acid sequences were aligned with MUSCLE (198) and clustered into phylogenetic trees using the following algorithms: maximum likelihood (Jones-Taylor-Thornton substitution model) using RAxML version 8.2.10 (199), maximum parsimony using PAUP 4.0 (200), and neighbor-joining (Jones-Taylor-Thornton substitution model) using BIONJ (201). CIPRES Science Gateway version 3.3 (202) servers were used for phylogenetic trees construction. Tree branches supported by over 50% of the 1,000 bootstrap replications were represented in the final consensus tree. The consensus tree was built using the “ape” package (203) in R (204). Branch colors represent taxonomic classification at the class level, since all 25 sequences are part of the Euryarchaeota phylum, while pie charts show the relative abundance of each class within Euryarchaeota.

FIG 2

Consensus tree of 1,217 [FeFe]-hydrogenase sequences, with an alignment length of 3,525 amino acids, computed as described in Fig. 1. See Fig. S2 for complete names of taxa and a more detailed phylogenetic tree. Branch colors represent taxonomic classification at the phylum level, while pie charts show the relative abundance of each phylum within hydrogenase subgroups (except for Eukaryota being a superkingdom). Dashed brackets show which tree branches belong to a specific hydrogenases group. Numbers between parentheses depict the amount of sequences within that group.

FIG 3

Consensus tree of 1,988 [NiFe]-hydrogenases sequences, with an alignment length of 1,850 amino acids, computed as described in Fig. 1. See Fig. S3 for complete names of taxa and a more detailed phylogenetic tree. Branch colors represent taxonomic classification at the phylum level, while pie charts show the relative abundance of each phylum within hydrogenase subgroups (except for Eukaryota being a superkingdom). Dashed brackets show which tree branches belong to a specific hydrogenases group. Numbers between parentheses depict the amount of sequences within that group.

FIG 4

Juxtaposition of the carbon cycle and main H₂-oxidizing functional groups. HOM contribute to all key steps of the C cycle. All numbers below microbial functional groups consist of reactions performed by these groups. While the oxidation of CH₄ can be performed in anoxic ecosystems, the anaerobic oxidation of CH₄ (AOM) is not performed by HOM, only by their syntrophic bacterial partners (i.e., sulfate- or nitrate-reducing bacteria). H₂-utilizing processes occur in a thermodynamically favored fashion according to available substrates, where O₂ is used first, followed by nitrate, iron oxides, sulfate, and carbon dioxide, as shown in the gradient at the bottom of the figure. The only key C-cycling process missing in oxic ecosystems is methanogenesis, yet residual CH₄ diffuses from anoxic to oxic layers in various ecosystems. An exception to this is that nonhydrogenotrophic methanogens are also active in oxic layers (138), thus providing CH₄ directly to oxic ecosystems.