The potential biofertilization effect of H₂ is accompanied by a modest impact on the composition of microbial communities in the rhizosphere of common vetch

Introduction

N₂-fixing nodules of legume plants release H₂ as a by-product of the N₂ fixation reaction. In the case of symbiosis involving rhizobia without the [NiFe]-hydrogenase (hup- genotype), the release of H₂ creates a concentration gradient from approximately 10,000 ppmv at the soil-nodule interface to sub-atmospheric concentration (<0.5 ppmv) five cm away from the source (Dong & Layzell, 2001; Piché-Choquette, Khdhiri & Constant, 2018). At least 40 phyla of H₂-oxidizing bacteria (HOB) exploit the energy potential of H₂ (Greening et al., 2016). In carbon limitation conditions, HOB can survive by relying on H₂ as the sole energy source or by a mixotrophic strategy combining H₂ with organic carbon (Bay et al., 2021; Greening et al., 2015; Islam et al., 2020; Liot & Constant, 2016). According to metagenomic surveys, approximatively 12% of soil bacteria use H₂-derived electrons to fix CO₂ (Xu et al., 2021).

An elevated H₂ concentration in the rhizosphere of legume plants promotes the enrichment of HOB (La Favre & Focht, 1983). Incubating soil samples in modified atmospheres comprising different H₂ concentrations isolates the impact of H₂ on the composition and activity of soil microbial communities. Soil exposure to 250 nmol H₂ g(soil)⁻¹ h⁻¹ mimics H₂ fluxes from N₂-fixing nodules and enriches CO₂ fixation activity from HOB (Stein et al., 2005). The number of cells affiliated to Cytophaga-Flavobacterium-Bacteroides was also impacted by H₂, as shown by phylotype enumeration by fluorescence in situ hybridization analyses (Stein et al., 2005). Soil exposure to elevated H₂ concentrations promotes the activity of HOB degrading persistent organic pollutants in soil (Xu et al., 2023). The activation of HOB in soil exposed to H₂ for a few weeks led to variable impacts on the composition of soil microbial communities (Khdhiri et al., 2017; Khdhiri et al., 2018), calling into question the impact of H₂ on soil microbiota.

Soil fumigation by H₂ enhances plant growth. For instance, wheat biomass increased by 48% in soil samples exposed to a H₂-enriched atmosphere compared to control soil samples (Dong et al., 2003). This phenomenon is known as the H₂ fertilization effect (Golding & Dong, 2010). The exact mechanism is currently unknown, but some evidence suggests that it involves HOB activated by H₂, including Variovorax spp., displaying plant growth-promotion activity (Maimaiti et al., 2007). The current study was conducted to determine whether a high concentration of H₂ in soil activates and increases the population of HOB and if HOB activation modifies the composition of the soil microbial community and enhances the growth of V. sativa. These hypotheses were tested by setting up a two-stage experimental design. The first stage consisted of conditioning two contrasting HOB populations by exposing a sandy loam soil to either (i) sub-atmospheric (<0.5 ppmv) or (ii) supra-atmospheric (5,000 ppmv) H₂ concentrations. Soil exposure to H₂ lasted two months to favour the activation and enrichment of HOB in soils exposed to high H₂ concentration. The second stage consisted of growing V. sativa using the enriched soils as a substrate, without further exposure to the modified atmosphere. The plant biomass and microbial communities of the plant rhizosphere were analyzed to examine the legacy effect of H₂ soil exposure. The legacy effect is defined here as a modification of the composition or activity of the microbial community during the activation stage that persists in the second stage of the experiment, without an external supply of H₂. The legacy effect includes stimulation of soil biological processes through plant-soil feedback caused by the plant growth-promotion effect of activated HOB. The activation of HOB did not support apparent growth of the HOB population and caused little impact on the composition of the soil microbial community. Elevated H₂ exposure reduced the below-ground biomass of V. sativa and no difference was noticed for the above-ground biomass. This manuscript was previously published as a preprint (bioRxiv 2025.02.12.637852; doi: https://doi.org/10.1101/2025.02.12.637852).

Material and Methods

Results

HOB activation stage

The H₂ oxidation rate measured in the control treatment was relatively stable, at 29 ± 3 nmol g_(soil−dw)⁻¹ h⁻¹ on average, without significant temporal pattern (Fig. 1). H₂ oxidation rate of the control soil decreased in weeks 6 and 7 prior to increasing again at weeks 9 and 10. Repetition of the experiment would be required to verify if the transient loss of activity was caused by technical issues or biological processes. No variations in soil water content or incubation conditions were noticed. Activation of HOB in microcosms exposed to elevated H₂ concentration was proved to be efficient with the H₂ oxidation rate varying from 47 ± 8 to 80 ± 11 nmol g_(soil−dw)⁻¹ h⁻¹ after 30 days of exposure to 63 ± 5 nmol g_(soil−dw)⁻¹ h⁻¹ (Fig. 1). The plateau of the H₂ oxidation rate indicates the energy potential of H₂ and nutrient content in soil were not sufficient to support proliferation of HOB. This conclusion was supported by indistinguishable hhyL gene copy numbers among control, with 2.2 ± 1.7 ×10⁶ copies per g, and eH₂ treatments, with 2.9 ± 2.2 ×10⁶ copies per g (t-test, p-value = 0.56; Table S5).

Soil microbial communities

Most bacterial 16S rRNA gene sequences were affiliated with the phyla Actinobacteria (28%), Proteobacteria (27%), Thaumarchaeota (16%), and Acidobacteria (10%). The impact of H₂ exposure during the activation stage on soil bacterial communities was negligible, as variations of alpha diversity were not significantly influenced by the eH₂ treatment (Tables S2 and S3). Aggregation of the ASV at the genus level led to the identification of three genera displaying contrasting differential abundances between the two treatments. A higher log₂ fold-change (LFC) was observed in the eH₂ treatment compared to the control, and two groups were identified as being favored by H₂ exposure, namely Aeromicrobium spp. (LFC = 1.02, p-value = 0.002) from the Actinobacteria phylum and Ramlibacter spp. (LFC = 0.74, p-value = 0.00005) from the Proteobacteria phylum (Fig. 2). A single group of ASVs aggregated to the genus Pseudoduganella spp. appeared more favored by the control treatment than by the eH₂ treatment, but the difference was not significant (LFC = −1.25, p-value = 0.08). No legacy effect on the eH₂treatment was observed on soil microbial communities in the rhizosphere of vetch as H₂ exposure neither influenced the alpha diversity nor the distribution of ASVs aggregated at the genus level (Tables S1 and S3).

The legacy of HOB activation of vetch

The eH₂ treatment did not influence the aerial biomass of V. sativa (Fig. 3A), whereas it did reduce root biomass (p-value= 0.034; Fig. 3B). The root/shoot ratio, usually used as an observation of the plant response to the environment, was reduced in eH₂ exposure treatment (p-value = 0.028; Fig. 3C).

Discussion

The production of H₂ in N₂-fixing nodules comprising rhizobia not possessing [NiFe]-hydrogenase creates steep H₂ concentration gradients in the rhizosphere of legume plants. These concentration gradients expand at long distances, up to several centimeters away from nodules. In addition to redefining the spatial boundaries of the rhizosphere effect, these gradients are expected to play a role in modifying the microbial community structure and function (De La Porte et al., 2020). This notion was first supported by the enrichment of HOB activity in soil surrounding N₂-fixing nodules in pioneering investigations performed in pot experiments (La Favre & Focht, 1983). Subsequent efforts were invested in artificial soil microcosms exposed to elevated H₂ concentrations. H₂ exposure enhanced net CO₂ fixation in soil, which is due to the facultative chemolithoautotrophic life strategy of numerous HOB in soil (Stein et al., 2005). The application of terminal restriction fragment length polymorphism (T-RFLP) led to the conclusion that only a few members of the microbial community were favored by H₂ exposure (Osborne, Peoples & Janssen, 2010). This observation was contrasted in subsequent investigations by applying high-throughput PCR amplicon and metagenomic sequencing techniques coupled with loose differential abundance analyses, suggesting H₂ exposure induced significant alteration of microbial communities, mainly in the rare biosphere (Khdhiri et al., 2017; Piché-Choquette et al., 2016). Revisiting these experiments with more stringent statistical tools tailored to the sparse microbiome data in the current study and in a recent work (De la Porte et al., 2024) confirmed conclusions supported by earlier T-RFLP techniques that H₂ exerts negligible impact on the soil microbial community. The exact mechanisms responsible for variations in biogeochemical processes upon soil exposure to H₂ are therefore decoupled from HOB proliferation and compositional change of microbial communities.

Only two ASV groups, Aeromicrobium and Ramlibacter, were positively influenced by the eH₂ treatment during the activation stage (Fig. 2). Pumphrey, Ranchou-Peyruse & Spain (2011) reported the enrichment of labelled DNA from a phylotype affiliated with Aeromicrobium ginsengisoli from hairy vetch rhizospheric soil incubated with 100 ppm of H₂. There is, therefore, evidence for chemolithoautotrophic metabolism in Aeromicrobium spp., but demonstration of the activity in isolates has not been reported. In addition to the negligible alteration of microbial communities by H₂, there is also no evidence supporting the proliferation of HOB encoding for group 1H/5 [NiFe]-hydrogenase in soil exposed to eH₂ as quantification of hhyL was indistinguishable between treatments (Table S5). The enrichment of HOB encoding [NiFe]-hydrogenase from other groups cannot be ruled out as they were not detected in the quantitative PCR assay. For instance, HOB encoding for [NiFe]-hydrogenase encompassing group 2a (Bay et al., 2021; Cordero et al., 2019; Islam et al., 2020), group 1f (Hyo type; Myers & King, 2016), group 1l (Hyl type; Ortiz et al., 2021), and group 1d (6C type; Pandelia, Lubitz & Nitschke, 2012) were not detected by hhyL PCR assays and are potential contributors of the H₂ oxidation activity measured in the current study. The stabilization of the H₂ oxidation activity in soil exposed to eH₂ was suggested in a previous microcosms experiment where the response of high-affinity H₂ oxidation activity plateaued after a one-week H₂ exposure period (Piché-Choquette et al., 2016). Extended incubation under a headspace enriched with H₂ was insufficient to induce a secondary enrichment after the enrichment plateau (Fig. 1). The level of CO₂ in the static headspace of soil microcosms during the activation stage as well as organic carbon and other nutrients were likely insufficient to support the proliferation of HOB.

Another aspect of H₂ exposure in agroecosystems is the fertilization effect of this gas. This was first reported in pot experiments where soil substrate was fumigated with H₂ prior to planting (Dong et al., 2003). The plant biomass increased by 15–48% in fumigated soil compared to controls. The fertilization effect was attributed to HOB displaying the plant growth-promotion effect (Maimaiti et al., 2007), such as 1-aminocyclopropane-1-carboxylate (ACC) deaminase activity reducing the accumulation of ethylene, which acts as a stress phytohormone (Yang & Hoffman, 1984). This is indirect evidence suggesting that recruitment of symbiotic N₂-fixing rhizobia releasing H₂ (Hup⁻) is an evolutionary advantage for legume plants. Rhizobia possessing [NiFe]-hydrogenase (Hup⁺) to recycle H₂ produced by nitrogenase exist, but they appeared less represented by their Hup⁻ counterparts in nature (Annan et al., 2012). Consideration of the fertilization effect of H₂ and the selection of Hup- symbioses in nature led to the suggestion that H₂ could be a missing link in clarifying why there are multiple benefits of crop rotation that have not been explained by nitrogen fertilization alone (Golding & Dong, 2010).

The benefits of H₂ on plants are likely multifactorial. The fertilization effect was not observed for wheat grown in soil subjected to fumigation by H₂, CO, or CH₄ (De la Porte et al., 2024). The aerial biomass of vetch in the present study was not enhanced in soil conditioned with eH₂ exposure treatment (Fig. 3). The reduction in root biomass, as well as in root/shoot ratio, is thought to be related to the plant growth-promotion activity of HOB, including ACC deaminase reducing root proliferation for nutrient prospection. If this mechanism holds true, it was realized in the absence of evident HOB proliferation. These results must be considered in trials integrating irrigation treatments with H₂-saturated water, enhancing stress resistance in plants. The antioxidant potential of H₂, microbial activation of immune system in plants, and HOB are potential contributors to plant benefits (Wang et al., 2024). A more holistic analysis, examining both plant and microorganism interactions in soil exposed to different H₂ levels or in contrasting Hup+/Hup- symbioses, is recommended to disentangle the contributions of microorganisms and plant metabolism in explaining the benefit of H₂.

Conclusions

In conclusion, soil exposure to eH₂ led to an activation of HOB without supporting their growth. H₂ was therefore sufficient to support the persistence energy requirement of HOB, probably by rewiring their energy metabolism toward an inorganic energy source and a repression of carbon metabolism. Examination of the potential synergetic effect of H₂ with nutrients will decipher the limiting factors for HOB proliferation and activation in soil.

Supplemental Information