Hydrogen-Driven Microbial Metabolic Network Reconstruction in Opalinus Clay Rocks


  • Applied microbiology
  • Biogeochemistry
  • Environmental microbiology
  • Metagenomics


The Opalinus Clay formation will house a repository of nuclear geological waste in Switzerland. It is expected that gas pressure will accumulate due to hydrogen production from steel corrosion, which undermines the integrity of the engineering barrier. In an in situ experiment located at the Mont Terri Underground Rock Laboratory, we showed that hydrogen is consumed by microorganisms, which trigger microbial communities. Metagenomics binning and metaproteomic analysis of these deep sub-surface communities reveal carbon cycles driven by the autotrophic hydrogen oxidizers belonging to the new genera. Necromass are then processed by fermentors, followed by total oxidation into carbon dioxide by heterotrophic sulfote reduction bacteria, which closes the cycle. This metabolic network of microbes can be integrated into geological repository designs to reduce pressure buildup. This study shows that Opalinus Clay has the potential for chemolithoautotrophic-based systems, and provides models of microbial carbon cycles in deep sub-surface environments where hydrogen and sulfate are present.

Most living microorganisms are found below the terrestrial surface 1, 2, in a zone that is hydrologically disconnected from the earth’s surface 3 . These microbial communities may not directly depend on sunlight as an energy source, but more on geogenicly reduced compounds, such as hydrogen gas (H2), when the buried organic matter is no 4.5 . Hydrogen is produced by serpentinization (anoxic oxidation) of mafik and ultramafik rocks ( 6) and water radiolysis ( 7) . Hydrogen is the main metabolic compound in many anoxic ecosystems and its oxidation has been advised to support sub-surface lithoautotrophic microbial ecosystems in 8, 9, 10, 11, 12, 13.

Due to the difficulty of access, there is a limited understanding of the microbial system below the surface in 14 . In this environment, the functional diversity of microorganisms has been assessed by several methods of sequencing, such as gene sequencing 16S rRNA 10, 15, metagenomic inscribed with the words 12, metagenomic contigs assemblies 16, 17 and metagenomic-assembled genomes (MAGs) of low levels, 18 and microbial communities with a high diversity of 13. MAG shows a significant contribution, as it allows the characterization of the suspected metabolism of individual organisms, and the reconstruction of the dislested metabolic tissue at the community level. But all of the above studies are based on the presence of genetic material, which only represents the metabolic potential given. To determine the metabolic processes that still exist, metatranscriptomic and/or metaproteomic analysis is necessary, since they can reveal the presence of active functions (by detecting messenger RNA and proteins, respectively).

Microbiological studies from beneath the surface are relevant for a variety of reasons. In particular, microbial processes affect the geochemistry of groundwater and oil reservoirs, impacting the extraction feasibility of resources. Additional motivation is related to the management of nuclear waste. Some European countries have chosen to dispose of nuclear waste in deep geological repositories, where waste will be isolated for a period of up to a million years, until a reasonable level of radioactivity is reached 19 . The design of such a system requires a detailed understanding of the main processes (physical, chemical and biological) that can occur. For example, anoxic steel corrosion will release H 2 which can increase the pressure in 20 closed repositories, which compromises the integrity of the engineering barrier. On the other hand, H 2 is a source of energy and electrons for microorganisms. Thus, microbial activity in the repository is expected to have a beneficial impact, by reducing the excessive pressure of H 2 in this limited environment of 21 .

To test this hypothesis, we used the Opalinus Clay rock formation, which is the current candidate to house a deep geological repository in Switzerland 22. These rocks are characterized by the distribution of small pore sizes (average 10-20 nm), the presence of reduced mineral species (siderite, pirit) 23, sulfate (SO 4 2− ) as dissolved substances in water (10-25 mM) 24, solid phase organic matter (1.5% b/b) 24 and dissolved organic matter (200 μM acetate) 25 . H2 injection in drill holes in Opalinus Clay, 300 m below the surface, combined with water monitoring drill holes capture the resulting chemical and biological changes as well as microbial metabolic tissue under conditions relevant to the repository.

Our study shows that Opalinus Clay has a microbial community capable of efficiently oxidizing H 2, which is beneficial for the safety of deep geological nuclear waste repositories, through the reduction of pressure buildup caused by anoxic steel corrosion. While the concentrations of H 2 in this study were unrealistic for natural sub-surface environments (because they were too high), they were relevant to repositories, and promoted the accumulation of microbial biomass, which allows metaproteomic analysis. Used in co-ornally with metagenomics sequencing and genome streaming, this technique highlights active metabolic pathways. The metabolic tissue described here, a carbon cycle based on chemolithoautotropy, is in common interest for deep sub-surface microbiology, and can serve as a model for understanding autotrophic microbial communities that rely on H 2, carbon dioxide (CO2) and sulfate.

Chemical and biological changes

To replicate the conditions relevant to the repository, we conducted an in situ experiment 300 m below the surface, at Mont-Terri Underground Rock Laboratory 26 (St-Ursanne; Switzerland; Fig. 1). We inject H2 in drill holes in Opalinus Clay (Fig. 2), and monitor chemical and biological changes for more than 500 days. Geochemical monitoring of drill well water reveals that oxygen (O 2) (resulting from drilling) is consumed quickly (within 1 month), leading to anoxic conditions (Fig. 3) after the start of amendment H 2. These events have limited consequences to long-term trials. This is because iron iron (Fe (III)), produced by exposure to O 2, is reduced back to iron iron (Fe (II)) at the beginning of the experiment after the condition returns to anocsia (Fig. 3). In addition, the customary community of anaerobic obligate, if any, continues to be sent to the drill pit through the influx of seawater (ml 20 ml per day). Within 150 days of the first H2 amendment, dissolved sulphides (especially HS – ) accumulated to concentrations of up to 600 μM, highlighting the dominance of sulfate reduction (Fig. 3). Sulfate is naturally in the water in ca. 24 mM and slowly decreases over time, due to the reduction of sulfate microbes and dilution by artificial water (APW; Figure Add-on 1). Indeed, the concentration of sulfates in apw injected up to day 317 is lower than in natural well water drill (Supplement to Table 1). H 2 is consumed together with SO 4 2 served and serves as an electron donor in its reduction (Supplement 2).

(a) Location of Mont Terri URL in Switzerland. ( b ) A cross-section of the Mont Terri highway tunnel, whose security gallery is used to dig urLs. ( c) The location of the drill holes sampled in this study. The circles indicate the position of the open drill hole and its angles indicate the orientation of the drill hole compared to the ground (negative angle means the drill hole is down). The drill hole used in the experiment was BRC-3. Modified from Swisstopo 26.

For better clarity, polyvinyl chloride screens are not displayed. Packer Neoprene isolated the water drill pit from the occult atmosphere from the Underground Stone Laboratory gallery. The polyamide line makes it possible to take drill water samples, and inject APW and H 2 (which creates head space) at different depths.

Full-size picture

Dissolved O2 concentration is indicated by a blue line (right y-axis), Fe (II) with green lines (right y-axis) and S (-II) with an orange line (left y-axis) in the water drill hole, starting 50 days before the first H2 injection that occurs on the 0th day. Initially, H2 injection is carried out through a small reservoir on the surface through a gas permeable membrane when recirculating seawater. After 50 days, H2 input is suboptimal due to membrane blockage (Additional Figure. 2). That is the reason why, starting day 134, H 2 is regularly injected directly in the drill hole, thus creating a gas phase. Shortly there was a decision to stop the circulation of seawater due to the loss of H2 in the peristaltic pump. The experimental procedure is shown at the top of the plot: the light blue line indicates when the water is recirculated, the dark red line indicates when H 2 is sent optimally from the surface, the bright red line indicates when H 2 is sent less optimally. from the surface, a dark cross indicates when the H 2 heated membrane is altered, and the red diamond indicates when H 2 is injected directly into the drill hole. Samples were also shown at the top of the plot: green boxes showed samples for gene sequencing 16S rRNA, red boxes showed samples for metagenomics sequencing, and blue box samples showed for metaproteomic analysis.

Full-size picture

Increased planktonic cell density occurs simultaneously with the reduction of O2, Fe (III), then SO4 2−, coupled with H2 oxidation (Supplemental Figure 3), indicating that this process is biological. This increase in the number of cells along with a decrease in the concentration of CO2 dissolved in water (Supplemented Figure 4), also indicates autotrophic growth. Gene sequencing of 16S rRNA region V4 DNA extracted from drill hole water containing suspended clay particles reveals a microbial community dominated by bacteria, whose composition fluctuates initially but remains stable after a sulfate reduction regime is formed (Fig. 4; Additional Data 1) . Aerob Gammaproteobacteria 27 originally represented about 60% of the microbial community, but was largely replaced by metabolically flexible Alphaproteobacteria 28, and Gram-positive and Gram-negative 29.30 sulfate-reducing bacteria (SRB) 29.30 once O2 was depleted.

This community was assessed by amplicone sequencing from the V4 region of the 16S rRNA gene. The first sample on day 0 was recovered before the first H2 injection. The colored bar at the top of the plot shows the last electron acceptor in a different redox regime (Fig. 3).


Recovery of genomes that gather with metanomes

Metagenomics binning enables the identification of 65 putative genomes (Supplemental Data 2), which is in line with other metagenomics studies of the surface in 13 . Twenty-two trays were found free of the wrong order assembled or incorrectly and completely (or almost complete; Supplement table 2) 31, and represents more than 83% of microbial communities (Fig. 5). Next, this trash can will be referred to as MAG 32. This MAG represents a subset of operational taxonomy units (OTU), defined by gene sequencing of 16S rRNA. Indeed, most MAG’s that store the 16S rRNA gene can be connected to OTU (Supplemental Table 3). Gen 16S rRNA is missing for some MAG (c0, c3, c23, c25, c36 and c42), which makes their taxonomy affiliations less precise (Supplemental Table 4). The other seven MAG (c4a, c4b, c8a, c16a, c20a, c32 and c57) represent new genera, as indicated by their taxonomy explanations (Supplement to Table 4) and the average nucleotide identity (ANI) with reference genomes (Table 5). Overall, two methods – metagenomics binning and gene sequencing 16S rRNA – describe microbial communities of the same composition (Figure 5 Supplement, Figure 4).

Completeness of selected metabolic pathways, such as respiratory processes, central carbon metabolism, H2 oxidation associated with respiratory processes, carbon pathways of fixation and fermentation, are indicated for each genome (green scale). This figure also includes, for each genome, the average proportion and proportions in each sample (blue, yellow and red scales). Seven MAG with protein expression information highlighted in gray.

Full-size picture

The metabolic pathways of the 22 MAG mentioned above are inferred from their protein explanations and produce their classification into six categories: (i) autotrophic SRB can use H2 as an energy source and electron to reduce CO2; (ii) heterotrophic BPRS is capable of oxidizing acetate into CO2; (iii) autotrophic bacteria cannot reduce sulfate, but are capable of oxidizing H 2 ; (iv) facultative autotrophic bacteria that are not able to reduce sulfate or oxidize H 2, are expected to show in situ heterotrophic growth, (v) heterotrophic bacteria are not able to reduce sulfate or oxidize H 2, and (vi) heterotrophic bacteria are not able to reduce sulfate but are able to oxidize H 2 (Fig. 5).
Metaproteomic analysis and community-level metabolic tissue

Metaproteomic analysis of the sample recovered after 483 days, when the system achieved a stable sulfate reduction condition, found enough protein information to decipher metabolic pathways for seven MAG (c4a, c8a, c12, c16a, c22, c23, c57; Table Supplement 6, Supplemental Data 3, 4 and 5), represents more than 60% of microbial communities, and allows identification of their metabolic activity. Only MAG c16a contains enough protein information to reconstruct a detailed metabolic map, as presented in Figure. 6. Protein information from other MAG is just enough to identify metabolism widely, as shown in the Figure. 7a. Analysis of 16S rRNA from metaproteomic samples showed that microbial communities were similar to those from previous metagenomics samples (Figure 4, Supplemental Data 1).


Evidence of proteins is shown with green boxes, and genome evidence with white boxes. These microorganisms reduce SO 4 2 4 to HS – using electrons from H 2 (gray compartment), resulting in a proton gradient across the cytoplasmic membrane. Intracellular proton translocation is combined into the ATP generation. ATP and electrons (the latter transferred via ferredoxin and reduced NADH/NADPH) can be used to reduce CO2 to acetyl-KoA through reductive acetyl-KoA pathways (orange compartments). Biomass biosynthesis occurs through the tricarboxylic acid cycle (green compartment), and gluconeogenesis (blue compartment). Biosynthesis of fatty acids and amino acids is not described in detail, but the proteins involved are listed in purple and red boxes respectively. Finally, these bacteria can also use acetate and propionate as a source of carbon (yellow compartment). All proteins from this number are listed in Supplement table 7.