Enceladus-like geochemistry fuels methanogenesis under extreme CO₂ limitation

Abstract

Saturn’s icy moon Enceladus features chemical signatures consistent with an alkaline soda ocean with H₂ production via hydrothermal water-rock reactions. Chemolithoautotrophic methanogenesis is thermodynamically favorable under these conditions, but it is unknown whether CO₂ scarcity in the alkaline Enceladus soda ocean limits autotrophy. Here, we show that simulated Enceladus’ ocean chemistry containing high concentrations of dissolved inorganic carbon and H₂ from mineral-water reactions enables growth of the chemolithoautotroph H₂-oxidizing methanogen Methanothermococcus okinawensis until pH 11, far exceeding its previously known pH limit. Transcriptomics revealed that growing cells overcome CO₂ scarcity in the simulated Enceladus soda ocean by overexpressing the reductive acetyl-CoA pathway for CO₂ fixation, which enabled efficient scavenging of CO₂ at extremely low concentrations. The carbon and energy metabolism of M. okinawensis was entirely fueled by the abiotic H₂ derived from the simulated Enceladus mineral-water reactions.

INTRODUCTION

Enceladus, a moon of Saturn, is covered in an ice shell with a liquid subsurface ocean beneath (1–4). Plumes of ice particles and water vapor erupting from fractures in the South Polar region (2, 5) contain signatures of molecular H₂ (6) and CH₄ (7). Chemical signatures from the plumes (2) indicate that Enceladus has an alkaline soda ocean with high concentrations of dissolved inorganic carbon (DIC) primarily in the form of carbonate (CO₃⁻²) (3, 8). The pH estimates range from 8 (9, 10) to 11.6 (11). The rocky core of Enceladus likely has a composition similar to carbonaceous chondrites and is rich in primitive Fe-Mg silicates (12). Water-rock reactions, such as serpentinization, are predicted to form secondary minerals and abiotic H₂ on Enceladus (6, 12–15). Abiotic H₂ from serpentinization (16) fuels chemolithoautotrophic life on Earth (17), and therefore, serpentinization on Enceladus makes it a promising extraterrestrial environment for potential life in our solar system (13, 18).

Thermodynamic models predict that the concentrations of H₂ on Enceladus are sufficient to fuel hydrogenotrophic methanogenesis as an energetically favorable metabolism (6, 12, 18–20). Methane (CH₄) was detected in the plumes of Enceladus, but the source of the CH₄ (whether biological or abiotic) remains unknown (11, 19, 21, 22). Biological methanogenesis is performed by archaea, which are versatile microorganisms that can tolerate alkaline pH up to 10.2 (23, 24), possibly higher, and temperatures from ∼0° to 110°C (25, 26). However, at pH values above 9, chemolithoautotrophic methanogens face an additional challenge: CO₂ scarcity. At pH >9, most of the DIC exists in the form of CO₃²⁻ (and to some extent HCO₃⁻). The dominance of CO₃²⁻ in the DIC of Enceladus available to autotrophs may hinder CO₂ fixation due to CO₂ scarcity. Chemolithoautotrophic methanogens depend on CO₂ as a carbon source and are generally optimized to live at circumneutral or slightly acidic pH (27). Living at pH <7 is beneficial to cells because most DIC is available as CO₂, which can freely diffuse across the cell membrane (28). Laboratory experiments with autotrophic methanogens in Enceladus-like conditions [but at acidic pH ∼5; (29, 30)] showed successful CH₄ production. However, it is currently unknown whether chemolithoautotrophic methanogens would be able to grow under the extreme CO₂ scarcity at the alkaline pH of Enceladus’ ocean where DIC is dominated by CO₃⁻²/HCO₃⁻.

Here, we experimentally investigated whether cellular growth via chemolithoautotrophic methanogenesis is possible under the high pH of simulated Enceladus ocean-floor conditions. As a model organism, we used the chemolithoautotrophic, thermophilic, H₂-oxidizing methanogen Methanothermococcus okinawensis that was isolated from a deep-sea hydrothermal vent (Iheya Ridge, Okinawa Trough, Japan) and was reported to grow at pH up to 8.5 (31). M. okinawensis belongs to the group of methanogens that lack cytochromes and, as such, has a relatively lower H₂ threshold for growth compared to methanogens that contain cytochromes (32). Moreover, methanogens that lack cytochromes usually use sodium gradients for ATP synthesis (32). These traits might represent an advantage for M. okinawensis in high pH environments, such as the Enceladus soda ocean.

RESULTS AND DISCUSSION

Mineral precipitation in an alkaline Enceladus simulant

We simulated an alkaline Enceladus seafloor environment, here referred to as “Enceladus simulant,” which is a mixture of an Enceladus ocean simulant (EOS) and a simulated Enceladus rock powder (SERP) (Fig. 1A and fig. S1). The EOS was prepared over a range of alkaline pH (10–12) and contained DIC at a concentration of 100 mmol kg⁻¹, which represents the upper end of estimated DIC concentrations (20 to 100 mmol kg⁻¹) in the Enceladus ocean (3). This pH range was proposed as thermodynamically favorable for hydrogenotrophic methanogenesis in the Enceladus ocean (6, 11). The SERP simulated the proposed chondritic composition of the core of Enceladus (11, 12, 33), containing SiO₂, MgO, FeO, and CaO in powderized form (34) (Materials and Methods). This setup is meant to simulate hydrothermal sediments that would accumulate at the base of a hydrothermal vent.

Fig. 1. Mineral precipitates and viable cells in Enceladus simulant experiments.

(A) Raman spectroscopy of mineral precipitates in the Enceladus simulant after 48 hours. Bands show the presence of ferroan brucite, iowaite, and aragonite. The picture shows the mineral precipitates with a blue-green color. (B) Raman spectra from the 0 to 1400 range showing plausible carbonate (aragonite and calcite) and iowaite mineralogy. (C) High-magnification photograph of mineral precipitate from the Enceladus simulant. (D) Fluorescence microscopy image of intact M. okinawensis cells (blue) in the Enceladus simulant after 48 hours; the fluorescence indicates the presence of coenzyme F420 and cellular viability. All results displayed are from pH 9.

The Enceladus simulants were incubated at 65°C for 48 hours under strictly anoxic conditions (<1 ppm O₂), forming a blue-green precipitate (Fig. 1A). The pH of the Enceladus simulant was measured at the beginning and the end of the experiments, which showed that the pH of the Enceladus simulants did not change more than 0.5 units over the course of the incubation. Raman spectroscopy, scanning electron microscopy, and differential scanning calorimetry–thermogravimetric analysis (DSC-TGA) showed that the precipitate consisted of ferroan brucite (Mg_xFe²⁺_1−x)(OH)₂), iowaite [Mg₆Fe³⁺₂(OH)₁₆Cl₂ * 4H₂O], calcite (CaCO₃), and aragonite (CaCO₃) (Fig. 1 and fig. S2). Ferroan brucite forms as a secondary mineral in serpentinizing systems (35) from the hydration of olivine (Fig. 1A), indicated by Raman bands at 276, 442, and 3645 cm⁻¹. The prominent Raman band at 3645 cm⁻¹, which shifted down from 3650 cm⁻¹ for pure magnesian brucite, shows the presence of Fe in brucite (35, 36). DSC-TGA of the Enceladus simulant (fig. S2) revealed two main weight loss events below 200°C and between 345° and 360°C pointing to the dehydroxylation of ferroan brucite (36). Energy dispersive x-ray analysis (fig. S2) suggested that the precipitate is chiefly composed of Mg and Fe, namely, 89 mol% Mg and 11 mol% Fe [x in Mg_xFe²⁺_1−x(OH)₂ = 0.097 to 0.128], which is in the observed range of x = 0.05 to 0.35 for ferroan brucite (37). Iowaite, a layered double hydroxide (38), can form at the expense of ferroan brucite (35). Natural iowaite is characterized by a prominent Raman band around 526 cm⁻¹ (38).

In addition to ferroan brucite and iowaite, calcite and aragonite are present among the mineral precipitates. The Raman band at 1085 cm⁻¹ corresponds to calcium carbonates (39) while Raman bands at 712 and 702 cm⁻¹ are indicative of calcite and aragonite, respectively. Intact M. okinawensis cells were observed in Enceladus simulants (Fig. 1D) that were autofluorescent under UV, indicating the presence of coenzyme F420 and cellular viability. Because M. okinawensis requires H₂ for growth, we investigated the formation of abiotic H₂ through hydrothermal mineral-water reactions in the Enceladus simulants.

Mineral-water reactions produce abiotic H₂ and fuel chemolithoautotrophic methanogenesis in the Enceladus simulant

In all Enceladus simulants, we detected ample concentrations of abiotic H₂ ranging from 281 ± 5 μM to 166 ± 8 μM (Fig. 2A and table S1). This is well above the minimal levels of H₂ that support growth of cytochrome lacking methanogens, such as M. okinawensis (lower limit of 1 to 10 Pa H₂) (32). No abiotic H₂ was detectable in controls that did not contain FeCl₂, demonstrating that production of abiotic H₂ in the Enceladus simulant was dependent on the presence of FeCl₂. This also shows that organic contamination from the reagents or laboratory environment cannot explain the copious amounts of H₂ production observed and that dissolved Fe(II) from the FeCl₂ was directly linked to the abiotic H₂ formed over our 48-hour timescale. Our use of 60 mM Fe(II) is higher than predicted Fe(II) concentrations in Enceladus hydrothermal fluids that were found to be in the nanomolar to micromolar range (34). Accordingly, the abiotic H₂ formation rates measured in our experiments may exceed those occurring on Enceladus, as our experimental conditions were designed to produce detectable reaction progress on laboratory timescales. The addition of FeCl₂ enhanced H₂ production by accelerating thermodynamically favorable reactions between Fe-bearing phases and water that may proceed over geologic timescales. Nevertheless, the measured μM H₂ concentrations in our Enceladus simulants match the predicted H₂ concentration range in Enceladus’ soda ocean (6), and also fall within the nanomolar to millimolar range predicted for Enceladus hydrothermal fluids (19). Similar to previous estimates (6, 30), we measured higher abiotic H₂ formation at relatively lower pH in the Enceladus simulant (Fig. 2A). Abiotic H₂ concentrations in the Enceladus simulants were roughly two to three times lower compared to the MMC medium where H₂ concentrations ranged from 323 ± 30 μM to 605 ± 20 μM.

Fig. 2. Mineral-water H₂ production and methanogenesis by M. okinawensis in Enceladus simulant compared to controls at pH 9 to 11.

(A) Concentrations of abiotic H₂ produced via mineral-water reactions in the Enceladus simulant in the absence (unfilled blue bars) and presence (filled blue bars) of M. okinawensis. Control experiments with MMC medium (green color) with overpressured headspace H₂/CO₂ gas mixtures are displayed as a comparison. Error bars correspond to standard error of the mean across three biological replicates. (B) Stable isotope labeling experiments with DI¹³C show H₂-dependent ¹³CH₄ production in Enceladus simulants (blue) compared to MMC medium controls (green) across the same range of pH (9 to 11) from (A). The highlighted region at the bottom of the plot shows the natural abundance of ¹³C. ¹³C-methane at % values above this indicate biological methanogenesis and CO₂ reduction from the added DI¹³C. Total CH₄ production is displayed in fig. S3 confirming a CH₄ increase with decreasing pH.

Ferroan brucite forming in the Enceladus simulant is noteworthy, because it is among the common secondary mineral products of low-temperature serpentinization and linked to H₂ generation (16, 35, 36, 40–42). Iron in ferroan brucite is in the ferrous state, which is one of the most important factors controlling H₂ concentrations in hydrothermal systems. In hydrothermal and serpentinizing systems, water oxidizes the Fe(II) in ferroan brucite, which increases H₂ flux (35, 36). The dissolution of ferroan brucite results in iowaite formation. Our detection of ferroan brucite, iowaite, and copious amounts of H₂ compared to controls shows that this process was occurring in our setup. When the generated H₂ is consumed, e.g., by microbes, the brucite reaction is pulled forward, generating even more H₂ (35). Considering the mineral phases, a possible H₂ generating reaction for Fe-rich brucite to iowaite is shown in Eq. 1

$$\begin{array}{c}8\left({\text{Mg}}_{0.75}{\text{Fe}}_{0.25}^{2+}\right){\left(\text{OH}\right)}_2+6{\mathrm{H}}_2\mathrm{O}+2{\text{Cl}}^{-}\to \hfill \\ {\text{Mg}}_6{\text{Fe}}_2^{3+}{\left(\text{OH}\right)}_{16}{\text{Cl}}_2\ast 4{\mathrm{H}}_2\mathrm{O}+{\mathrm{H}}_2+2{\text{OH}}^{-}\hfill \end{array}$$ (1)

When M. okinawensis was added to the Enceladus simulants, the concentrations of H₂ were 38 to 31% lower compared to abiotic controls, indicating that M. okinawensis consumed some of the abiotic H₂ (Fig. 2A). Stable isotope labeling experiments (table S1, experiment 3) with ¹³C-labeled NaHCO₃ (Materials and Methods) allowed tracking of ¹³C-labeled CH₄ produced from the added DI¹³C. Compared to MMC medium, stable isotope labeling experiments showed that M. okinawensis exhibited higher rates of methanogenesis correlated to the presence of abiotic H₂ (R² = 0.81) in the Enceladus simulants at alkaline pH (Fig. 2B). ¹³C-labeling of CH₄ in the Enceladus simulant varied from 10.9 to 2.1 atomic % (at %) ¹³C with more ¹³CH₄ produced at lower pH. In contrast, M. okinawensis produced less ¹³CH₄ at alkaline pH in MMC medium, namely, 4.2 to 1.7 at % ¹³CH₄ until pH 10.5. At pH 11, no ¹³CH₄ production was detectable in the MMC medium (Fig. 2B). In contrast, M. okinawensis still produced ¹³CH₄ even at pH 11 in the Enceladus simulant to a level of 2.1 at % ¹³C (Fig. 2B). In addition to the stable isotope labeling, the total amount of methane produced shows the same trends of increased CH₄ production in the Enceladus simulants relative to the MMC medium at alkaline pH, with a positive correlation with H₂ availability (fig. S3).

Despite the MMC medium having higher concentrations of H₂, lower CH₄ production was observed at alkaline pH compared to the Enceladus simulants (fig. S3 and Fig. 2B). This indicates that hydrogenotrophic methanogenesis was limited in the MMC medium at alkaline pH, potentially due to lower CO₂ concentrations. Specifically, at a pH ranging from 9 to 11, the concentration of CO_2(aq) in the MMC medium ranges from 10⁻⁶ to 10⁻¹⁰ mol kg⁻¹, whereas in the Enceladus simulants, CO_2(aq) was present at concentrations one to two orders of magnitude higher (10⁻⁵ to 10⁻⁸ mol kg⁻¹) due to the higher DIC levels.

Abiotic H₂ production fuels autotrophic growth

In line with prior studies (23, 31), M. okinawensis grew best at a circumneutral pH of ∼6.5 in the MMC medium (table S1) and exhibited an exponential growth phase (growth rate μ = 0.12 hour⁻¹) (fig. S4; Materials and Methods) yielding >10⁷ cells ml⁻¹ after 48 hours. In contrast, in the MMC medium at pH 9, linear growth was observed (Fig. 3A), resulting in 10⁵ cells ml⁻¹, a 100-fold reduction compared to growth at pH 6.5 in the MMC medium (fig. S4). At pH 10 in the MMC medium, there was a 10,000-fold reduction in growth compared to pH 6.5. At pH 11, no growth was observed in the MMC medium (Fig. 3, B and C).

Fig. 3. Increased growth and differential gene expression of M. okinawensis in Enceladus simulants compared to MMC medium at pH 9, 10, and 11.

(A to C) Growth curves of M. okinawensis at pH 9 (A), pH 10 (B), and pH 11 (C) compared to MMC medium and sterile water controls show an increased growth in the Enceladus simulant at all pH values tested. The highest growth was observed at pH 9 in the Enceladus simulant. Error bars indicate standard error of the mean across three biological replicates (table S1). (D to F) Volcano plots showing differential gene expression of M. okinawensis in the Enceladus simulant compared to MMC medium. Dark blue points highlight genes encoding proteins involved in the reductive acetyl-CoA pathway for methanogenesis as well as H₂ oxidation. The red point in (F) is a gene (Metok_0190) encoding the substrate binding domain of a predicted ABC-type nitrate/bicarbonate transporter. Gene acronyms are listed in table S2.

In the Enceladus simulant, however, M. okinawensis showed increased growth compared to the MMC medium at alkaline pH reaching up to 5.3 × 10⁵ cells ml⁻¹ after 48 hours at pH 9 (Fig. 3A), 2.5-fold higher compared to MMC medium controls (Fig. 3A). At pH 10 in the Enceladus simulants, cells grew up to slightly lower concentrations compared to pH 9 (2.4 × 10⁵ cells ml⁻¹), but still 100 times higher compared to the MMC controls (Fig. 3B). At pH 11 in the Enceladus simulants, growth was still detectable up to 1.4 × 10⁵ cells ml⁻¹ (Fig. 3C) compared to no growth in the MMC medium. To our knowledge, this extends the known pH limit [pH 8.5 (31)] of M. okinawensis to pH 11. Our findings are consistent with prior studies, showing that autotrophic methanogenesis is possible under alkaline conditions ranging from pH 7 to 12 (43–45).

Overexpression of reductive acetyl-CoA pathway scavenges CO₂ from Enceladus simulant

Chemolithoautotrophic methanogens, such as M. okinawensis, use the reductive acetyl-CoA pathway to fix CO₂ via H₂ reduction, which serves to provide CO₂ as a carbon source for cell growth and produces CH₄ as a metabolic by-product (46). We used differential gene expression to investigate the expression of this pathway under the CO₂-limited conditions at alkaline pH in the Enceladus simulant compared to MMC controls. The number of overexpressed genes increased as a function of pH in the Enceladus simulants (fig. S5), whereas in MMC controls, the number of overexpressed genes decreased with increasing pH. A general trend was observed, that under alkaline growth conditions, the number of overexpressed genes increased compared to the growth conditions in MMC at pH 6.5 (fig. S6). At pH 9, 10, and 11, a total of 170, 82, and 255 genes were significantly overexpressed (P < 0.05, FDR < 0.05, |log₂ fold change| ≥1) in the Enceladus simulant compared to the MMC medium (Fig. 3, D to F). In MMC medium, 101, 35, and 3 genes were significantly overexpressed relative to the Enceladus simulant at pH 9, 10, and 11, respectively. These results show that a distinct physiological response of M. okinawensis was induced by the geochemical conditions of the Enceladus simulant at alkaline pH. This physiological response is mirrored in the overexpression of arCOGs [Archaeal Clusters of Orthologous Genes database (47)] in the Enceladus simulant, where energy production and conversion (7 to 11%) was a large portion of the overexpressed genes (33 to 39%) at alkaline pH (fig. S7).

The fixation of CO₂ by M. okinawensis at alkaline pH may involve carbonic anhydrase, an enzyme many autotrophic microbes use to convert HCO₃⁻ to CO₂ (48). The genome of M. okinawensis encodes several carbonic anhydrases, which were expressed in some alkaline treatments albeit not significantly. However, several genes encoding proteins of the acetyl-CoA pathway for CO₂ fixation were significantly overexpressed at pH 9, 10, and 11 in the Enceladus simulants (Fig. 3, D to F, and table S2). At pH 10, genes encoding proteins of the acetyl-CoA pathway were also overexpressed in the MMC medium control (Fig. 3E).

The highest number of overexpressed genes involved in methanogenesis (n = 18 genes) was observed at pH 11 in the Enceladus simulant, relative to the MMC medium control (Fig. 3F). This included overexpression of genes encoding CO dehydrogenase/acetyl CoA synthase subunits (cdhA, cdhB, and cdhD) producing acetyl-CoA for cell growth (49). Furthermore, overexpression of genes encoding formylmethanofuran dehydrogenase subunits (fwdA, fwdC, and fwdE) (50), formate dehydrogenase subunits (fdhA and fdhD) (50), and methyl-coenzyme M reductase system component A2 (atwA) (51) was detected (Fig. 3). The marked increase in overexpression of genes encoding proteins of the acetyl-CoA pathway at pH 11 in the Enceladus simulants relative to the MMC controls (Fig. 3F) indicates a role for this metabolism in M. okinawensis overcoming CO₂ limitation at pH 11 in the Enceladus simulant. At pH 11, M. okinawensis appears to overexpress its CO₂-fixation machinery in response to high DIC concentrations in the Enceladus simulant to scavenge extremely low concentrations of CO₂ for energy conservation and growth.

At high pH, energy limitation due to H⁺ scarcity makes it difficult for cells to maintain a proton motive force driving ATP synthesis (52). In cytochrome-lacking methanogens such as M. okinawensis, however, sodium ion gradients are used for ATP synthesis (32), which might help to overcome H⁺ scarcity under high pH conditions. Several overexpressed genes were found involving the creation of a Na⁺ gradient (Na⁺/H⁺ antiporters, sodium-extruding oxaloacetate decarboxylase subunit alpha, table S2), and the expression of archaeal ATP synthases was also observed, which can pump both Na⁺ and H⁺ to generate ATP (52). The gene expression data point toward the use of a Na⁺ motive force to synthesize ATP as a means to overcome energy limitation and H⁺ scarcity at alkaline pH.

High DIC in Enceladus simulants alleviates CO₂ limitation for M. okinawensis

Despite the MMC medium containing higher concentrations of H₂, nutrients, and trace metals, M. okinawensis growth was limited in the MMC medium at alkaline pH compared to the Enceladus simulants (Fig. 3, A to C). Because the Enceladus simulant has a 10-fold higher DIC concentration compared to the MMC controls, DIC (the carbon source) may be the limiting factor for growth of M. okinawensis in MMC controls. In the Enceladus simulant, CO_2(aq) ranges from 7.5 × 10⁻⁵ to 3.6 × 10⁻⁸ mol kg⁻¹, compared to orders of magnitude lower CO_2(aq) concentrations in the MMC medium (4.7 × 10⁻⁶ to 8.4 × 10⁻¹⁰ mol kg⁻¹) at pH from 9 to 11. The orders of magnitude higher CO_2(aq) concentrations in the Enceladus simulants corresponded to higher observed numbers of new cells that formed in the Enceladus simulant compared to the MMC medium (Fig. 4). In the Enceladus simulants at pH 10 and 11, one to two orders of magnitude more cells were observed compared to MMC controls (Fig. 4).

Fig. 4. Increased CO₂ supply in Enceladus simulants promotes chemolithoautotrophic growth of M. okinawensis at pH > 9.

The number of new cells forming at various pH levels was dependent on the aqueous CO₂ concentrations (molality) in the Enceladus simulant and MMC medium (calculated with SpecE8 in Geochemist’s Workbench). The order of magnitude of CO₂ concentration at each pH is displayed underneath each treatment. Note that the CO₂ concentration generally decreases with increasing pH and that the CO₂ concentration in the Enceladus simulant is one to two orders of magnitude higher compared to MMC medium controls at pH 9 to 11.

Although the concentration of CO_2(aq) is orders of magnitude higher in the Enceladus simulants compared to the MMC medium at pH >9, alkaline conditions cause the majority of DIC to be in the form of either HCO₃⁻ and CO₃⁻². The high pH causes extreme CO₂ limitation for M. okinawensis, which requires CO₂ for biomass production, because they cannot directly access HCO₃⁻ and CO₃⁻² for carbon fixation. Enzymatic conversion of HCO₃⁻ to CO₂ via carbonic anhydrase is one mechanism used by autotrophs that live at neutral or alkaline pH to access CO₂. M. okinawensis encodes several carbonic anhydrases, which were expressed at pH 11 in the Enceladus simulants, albeit not significantly overexpressed relative to MMC controls. However, a gene encoding a substrate binding domain of ABC-type nitrate/bicarbonate transporters was significantly overexpressed in the Enceladus simulants at pH 11 relative to MMC controls (Fig. 3F and table S2) and also in the Enceladus simulants at pH 9 and 11 relative to MMC 6.5 (fig. S6). This gene might have been involved in increased HCO₃⁻ transport in Enceladus simulants, which could be related to the increased growth in Enceladus simulants relative to MMC controls at pH 11.

An alternative mechanism that does not require carbonic anhydrase could be Le Chatelier’s Principle, namely, if a system at equilibrium experiences a change (such as removal of CO₂ from an alkaline solution), the equilibrium will shift in the direction that counteracts that change. If M. okinawensis cells continuously consume CO₂, more HCO₃⁻ will convert to CO₂ from the DIC pool to maintain equilibrium. Therefore, at alkaline pH, M. okinawensis could still maintain a constant supply of newly formed CO₂, because removal of CO₂ via autotrophy pulls the overall equilibrium to regenerate more CO₂ from HCO₃⁻. It is also possible that H₂CO₃ freely diffuses into the cell and is used by the organism (53). The overexpression of genes encoding the acetyl-CoA pathway at pH 11 where growth of M. okinawensis still occurred (Fig. 3) indicates that the autotrophic metabolism and CO₂ consumption by M. okinawensis maintains a constant replenishing of the CO₂ pool from DIC due to Le Chatelier’s Principle.

Growth of methanogenic archaea requires numerous trace metals and nutrients that were not added to the Enceladus simulant. The source of trace metals and nutrients must have then been from the small volume of medium (1 ml) that was transferred together with the methanogens when inoculating the Enceladus simulant experiments. This diluted the concentration of trace nutrients by a factor of 20. Despite this, the methanogens still grew better in the Enceladus simulants in comparison to MMC medium showing that trace nutrients were not limiting the growth of M. okinawensis in the Enceladus simulants.

The key factor enabling cellular growth under alkaline conditions in the Enceladus simulants is likely the 10-fold higher concentrations of DIC relative to MMC medium. Despite having higher concentrations of available H₂ in the MMC medium (Fig. 2A), reduced cell growth at alkaline pH was observed (Figs. 3 and 4). This can be explained by carbon-limited conditions in the MMC medium at alkaline pH, whereas in the Enceladus simulants, DIC concentrations are high enough to supply CO₂ at levels enabling one to two orders of magnitude higher biomass production.

In summary, our findings potentially widen the habitability window of Enceladus’ subsurface soda ocean by demonstrating that the geochemical environment enables the growth of a chemolithoautotrophic methanogen at alkaline pH 9 to 11, far beyond its natural growth range on Earth, by providing two key substrates: abiotic H₂ and DIC at high concentrations. The ability of M. okinawensis to grow via the reductive acetyl-CoA pathway under these extreme conditions is consistent with the presence of methanogens (40, 44, 45, 54, 55) and acetogens (56, 57) in hyperalkaline environments generated by serpentinization, potential terrestrial model systems for Enceladus-like geochemistry.

The high carbon abundance in Enceladus’ ocean (3, 4) has already been found to be a prerequisite for the observed high phosphate availability (33) and thus a high DIC appears to be a positive factor for habitability in more than one sense. This high DIC is a consequence of the high carbonate availability for planetary bodies in the solar system that formed beyond the CO₂ snowline, that is, the distance from the Sun beyond which CO₂ is in a solid (icy) state and is available as a planetary building material. Therefore, our findings enhance the habitability potential of subsurface oceans in the outer solar system in a more general sense.

MATERIALS AND METHODS

The Enceladus simulant is an anoxic Enceladus-like microenvironment in the lab that was created to simulate key features of the seafloor of the icy moon. To perform the laboratory experiments, two key features of Enceladus had to be simulated, the rocky core of the icy moon similar to a chondritic composition (11, 12, 33, 34) and the overlying soda ocean (3, 11). Because the experiment targets habitability near seafloor hydrothermal environments, the ice crust above the ocean was not taken into account. At the interface of rocky core and ocean, hydrothermal activity is suspected (12, 14). All experiments were conducted in an anoxic chamber (MBraun) with a N₂ atmosphere and O₂ values below 1 ppm. The Enceladus simulant consisted of a SERP and an Enceladus ocean solution (EOS), both mimicking key features of Enceladus’ rocky core and soda ocean (Supplementary Methods). Experiments were conducted at an alkaline pH ranging from 9 to 11 (58) and at 65°C over 48 hours. In addition, the thermophilic methanogen M. okinawensis (DSM strain designation IH1, German Collection of Microorganisms and Cell Cultures GmbH) was added to the Enceladus simulant to test for habitable conditions. All M. okinawensis cultures were prepared at the Institute of Microbiology and German Archaea Centre at the University of Regensburg and grown in MMC medium, a common growth medium for methanogens (59, 60). It was previously used to cultivate methanogenic archaea for an early Earth experiment with an iron-sulfide chemical garden (61).

RNA extraction was done with the Direct-zol RNA Microprep kit (ZYMO Research) and the protocol from Grünberger et al. (60) and Helmbrecht et al. (61). Transcriptomes were sequenced to a median depth of 71.5 (±7.2) million paired end reads and quality trimmed using Seqtk with the default settings (62). Quality trimmed reads were searched for similarity against protein encoding genes in the genome of M. okinawensis IH1 (31) by searching reads via BLASTx against the predicted proteome as a database. Read mappings to specific genes were analyzed as a matrix of mapped reads per gene per sample as described previously (61). Differential expression of protein encoding genes and rRNA from the M. okinawensis genome across treatments was plotted and analyzed using DESeq2 (63), using raw read counts per sample as input with the default settings, in Rstudio. The SpecE8 program in the Geochemist’s Workbench 2025 software version 18.0 using the thermodynamic database thermo.dat (64) was used to calculate the aqueous CO₂ concentrations and calcium carbonate mineral saturation states across a range of pH (9, 9.5, 10, 10.5, and 11).

In this study, five different sets of experiments were performed (table S1). Experiment 1 for mineralogical analyses (Fig. 1 and fig. S2), experiment 2 for abiotic H₂ production in the Enceladus simulant (Fig. 2A), experiment 3 for stable isotope labeling (Fig. 2B and fig. S3), experiment 4 to test for growth of M. okinawensis in the Enceladus simulant (Fig. 3, A to C, and 4 and fig. S4) and experiment 5 for gene expression analysis (Fig. 3, D to F; figs. S5 to S7; and table S2). Figure S1 gives an overview of all analytical tools used. More details can be found in the Supplementary Materials.

Supplementary Materials

This PDF file includes:

Supplementary Methods

Figs. S1 to S7

Tables S1 to S3