A nitrite-oxidising bacterium constitutively consumes atmospheric hydrogen

Abstract

Chemolithoautotrophic nitrite-oxidising bacteria (NOB) of the genus Nitrospira contribute to nitrification in diverse natural environments and engineered systems. Nitrospira are thought to be well-adapted to substrate limitation owing to their high affinity for nitrite and capacity to use alternative energy sources. Here, we demonstrate that the canonical nitrite oxidiser Nitrospira moscoviensis oxidises hydrogen (H₂) below atmospheric levels using a high-affinity group 2a nickel-iron hydrogenase [K_m(app) = 32 nM]. Atmospheric H₂ oxidation occurred under both nitrite-replete and nitrite-deplete conditions, suggesting low-potential electrons derived from H₂ oxidation promote nitrite-dependent growth and enable survival during nitrite limitation. Proteomic analyses confirmed the hydrogenase was abundant under both conditions and indicated extensive metabolic changes occur to reduce energy expenditure and growth under nitrite-deplete conditions. Thermodynamic modelling revealed that H₂ oxidation theoretically generates higher power yield than nitrite oxidation at low substrate concentrations and significantly contributes to growth at elevated nitrite concentrations. Collectively, this study suggests atmospheric H₂ oxidation enhances the growth and survival of NOB amid variability of nitrite supply, extends the phenomenon of atmospheric H₂ oxidation to an eighth phylum (Nitrospirota), and reveals unexpected new links between the global hydrogen and nitrogen cycles. Long classified as obligate nitrite oxidisers, our findings suggest H₂ may primarily support growth and survival of certain NOB in natural environments.

Fig. 1. Hydrogen oxidising activities of Nitrospira moscoviensis during nitrite-replete and nitrite-deplete conditions.

a Oxidation of molecular hydrogen (H₂) to sub-atmospheric levels by N. moscoviensis cultures. Error bars show standard errors of three biological replicates, with heat-killed cells and medium-only incubations as negative controls. The mixing ratio of H₂ is on a logarithmic scale, and the grey dotted line indicates the average atmospheric mixing ratio (0.53 ppmv). Black arrows indicate time points at which nitrite had been completely oxidised and was replenished in the nitrite-replete cultures. b Kinetics of H₂ oxidation by N. moscoviensis cells. The curve was fitted and kinetic parameters were calculated based on a Michaelis–Menten non-linear regression model. Note that dissolved H₂ levels in the culture medium were 0.39 nM H₂ at atmospheric conditions and at 37 °C.

Fig. 2. Heatmap of selected N. moscoviensis proteins that were differentially abundant under nitrite-deplete versus nitrite-replete conditions.

Notably, atmospheric H₂ was oxidised under both conditions (see Fig. 1a). Normalised protein abundance values were log₂-transformed, grey colour indicates absence from the proteome. The complete set of untransformed values is listed in Table S1. ND1 to ND3 and NR1 to NR3 are three biological replicates under nitrite-deplete and nitrite-replete conditions, respectively. Fold changes under nitrite-deplete conditions and the corresponding significance values (adj. p value ≤ 0.001, ***; ≤0.01, **; ≤0.05, *) are shown in the two boxes next to each protein. Yellow-green and blue colour indicate significant positive and negative fold changes, respectively, and grey asterisks indicate that abundance values were imputed. Select low-abundance proteins of interest were included despite being not statistically significant. Depicted proteins are sorted based on their functional category as displayed on the left side. UniProtKB accession numbers are indicated in parentheses. CI_1, 2M-type NADH-quinone oxidoreductase 1; CI_2, NADH-quinone oxidoreductase 2; nomenclature of nitrite oxidoreductase subunits according to Mundinger et al. [8].

Fig. 3. Genome-based model of energy and CO₂ fixation metabolism in N. moscoviensis, showing significant fold changes (min. twofold) in the relative abundance levels of selected proteins under nitrite-deplete conditions compared to nitrite-replete conditions.

The significant fold change level is indicated for each protein by colour, with grey indicating no significant fold change of at least 2-fold. Quinone reduction by the hydrogenase (Huc) is shown in the context of respiration, but the resulting quinol may also fuel reverse electron flow as outlined in the main text. Dashed lines indicate features and electron flow pathways awaiting experimental confirmation in Nitrospira or other organisms. Respiratory complexes are labelled with roman numerals. Cyt. cytochrome; Fd ferredoxin; OGOR 2-Oxoglutarate-ferredoxin oxidoreductase; POR pyruvate-ferredoxin oxidoreductase; Q quinone. For further details, refer to the main text, Fig. 2, and Supplementary Table S1.

Fig. 4. Thermodynamic modelling of the power yield from H₂ and nitrite oxidation at various substrate concentrations based on the Michaelis–Menten kinetics of the reactions by N. moscoviensis cultures.

The power yields for nitrite oxidation at nitrate:nitrite molar ratios 1, 100, and 10,000 are shown. The dotted vertical line indicates the atmospheric H₂ concentration.