Microbial biomining from asteroidal material onboard the international space station

TL;DR

This study explores using microbes to extract valuable elements from asteroid material in space, showing that microgravity affects how well this process works.

Contribution

The study demonstrates the feasibility of microbial biomining in microgravity and identifies specific microbial metabolic changes under these conditions.

Findings

Penicillium simplicissimum enhanced the release of palladium and platinum in microgravity compared to non-biological leaching.

Non-biological leaching was more effective in microgravity for many elements than on Earth.

Microgravity altered microbial metabolism, increasing carboxylic acid production and other potentially useful molecules.

Abstract

Expanding human space exploration necessitates technologies for sustainable local resource acquisition, to overcome unviable resupply missions. Asteroids, some of which rich in metals like platinum group elements, are promising targets. The BioAsteroid experiment aboard the International Space Station tested the use of microorganisms (bacteria and fungi) to extract 44 elements from L-chondrite asteroidal material under microgravity. Penicillium simplicissimum enhanced the release of palladium, platinum and other elements in microgravity, compared to non-biological leaching. For many elements, non-biological leaching was more effective in microgravity than on Earth, while bioleaching remained stable. Metabolomic analysis revealed distinct changes in microbial metabolism in space, particularly for P. simplicissimum, with increased production of carboxylic acids, and molecules of potential biomining or pharmaceutical interest in microgravity. These results demonstrate the impact of microgravity on bioleaching, highlighting the need for optimal combination of microorganisms, rock substrate, and conditions for successful biomining, in space and Earth.

Figures

The BioAsteroid experiment.A NASA astronaut Michael Scott Hopkins performs the insertion of the experiment containers in KUBIK (credits: ESA/NASA); B The six hardware units inserted into the KUBIK onboard the ISS (credits ESA/NASA); C BioAsteroid logo, created by Sean McMahon (University of Edinburgh); D Flow diagram of the experiment. After preparation, samples were integrated into the experimental units (EUs) together with the medium and the fixative. The EUs were either launched to the ISS (blue oval), where they were installed in KUBIK incubators and subjected to microgravity (µg) or kept

Characterization of the L-chondrite meteorite.A Backscatter electron image (BSE) and single elemental mapping for some of the major elements present in a pristine representative fragment of the L-chondrite used in this experiment (scale bar: 100 µm); B Phase analysis of a similar pristine fragment of the L-chondrite; C Photographic image of a pristine fragment analysed by Raman spectroscopy; D Composite Raman map with forsterite in red, enstatite in blue, and luminescence in green. Typical spectra are displayed for E forsterite, F enstatite, and G luminescence signal. Reference spectra (black)

Scanning electron microscopy (SEM) images of the L-chondrite fragments.Secondary electron SEM images are shown here and are representative of samples in the two gravity conditions (ISS for microgravity, Earth for terrestrial gravity). Images were acquired at varying magnifications to allow the visualisation of the features of interest. Scale bars are indicated in each figure, ranging between 2–20 µm.

Analysis of biomining for 44 elements on the ISS and on Earth.For each element, represented in order of atomic number, bioleaching data are shown as percentage differences with the non-biological control. Upper panels indicate data from the ISS samples, lower panels from the Earth samples. Circles indicate mean values; error bars indicate standard error; vertical grey lines indicate 100% values (i.e., same bioleaching efficiency as non-biological samples).Fig. 5Platinum group element (PGEs) biomining.In panels (A) and (B), data for the ISS experiment are shown. For each element, data are repor

Metabolomic analysis of microorganisms during microbe-meteorite interaction in space and on Earth.A Principal component analysis (PCA) representing all the samples; B PCA showing ISS samples only; C PCA showing Earth samples only. Ellipses correspond to 95% confidence intervals. D Heatmap showing the top 25 features measured in the metabolomics analysis, clustered by gravity condition and organism. Each column represents a single sample. Numbers of the left (1-6) represent visual clusters suggesting metabolomic patterns dependent on the organism and/or the gravity condition. Heatmap legend is

Full text

Introduction

Establishing a sustainable human presence in space requires harnessing local resources through in situ resource utilization (ISRU). This approach aims to reduce cost- and energy-intensive resupply from Earth and enable sustainable manufacturing of materials and products in situ^1–5^. Microorganisms have been proposed as the foundation for a ‘bio-factory’ on Mars^6,7^, leveraging their integral role in Earth-based industries, such as food, drugs, chemical feedstock, and bioplastics. Similar versatility could be harnessed in space^8^, unlocking sustainable approaches for human space exploration^4^.

Biomining^9–12^ is potentially one of these technologies. Microorganisms are used to catalyse the breakdown of rocks and the release of useful elements, accelerating the acquisition of the required elements, exploiting mine waste tailings, and avoiding the use of environmentally damaging toxic compounds such as cyanides^13^. While acidophilic autotrophs are commonly employed to bioleach sulfidic ores^10,14,15^, heterotrophic microorganisms, including bacteria and fungi^16–18^, are also effective in bioleaching through proton, organic acid, or complexing compound release^19^.

Biomining holds promise for extracting elements and compounds from local materials in space^2,20,21^. For example, the bacterium Sphingomonas desiccabilis has been used to extract rare earth elements and vanadium from basalt rock under microgravity and Martian gravity on the International Space Station (ISS)^22,23^. Bacterial leaching of copper has been demonstrated in microgravity^24^, and other elements were bioleached from lunar and Martian simulants on Earth^25,26^. Beyond element extraction, microbial interactions with regolith could facilitate soil formation and nutrients release for life support systems, further enhancing the self-sustainability of space habitats^4^. Meteorites such as carbonaceous chondrite, can even support microbial growth^27–29^.

Asteroids, including those in the asteroid belt and near-Earth asteroids^2,20^, offer a potential source of essential materials for future human space settlements (e.g. water, hydrogen, carbon, metals)^30–32^. Some asteroids contain precious metals, including platinum-group elements (PGEs)^31,33,34^ such as palladium and platinum, which are indispensable to high-tech industries^35^. Despite uncertainties surrounding the economic feasibility of asteroid mining^36^, the high market value of PGEs on Earth highlights a great potential role for in-space manufacturing^20,37^. On Earth, PGE bioleaching has been successfully accomplished using both chemolithotrophic^38^ and heterotrophic microorganisms^39^.

The BioAsteroid experiment tested the ability of heterotrophic microorganisms (the bacterium S. desiccabilis^22,23,40,41^ and the fungus Penicillium simplicissimum^12,42–46^) to catalyse the release of technologically and economically important elements from L-chondrite, a common type of meteorite^31^, under microgravity conditions onboard the ISS (Fig. 1). An artificial microbial consortium, a common strategy in terrestrial biomining^11,47^, was also tested using a mixture of these two organisms. The study targeted 3 PGEs and 41 additional elements of industrial interest, comparing microbial versus abiotic leaching, and evaluating the effects of microgravity. A thorough metabolomic analysis provided invaluable insight into how microorganisms respond metabolically to space conditions during microbe-mineral interactions. This experiment demonstrates proof of principle of microbial-driven transformation of asteroidal material for future human space exploration and settlement in a self-sustainable fashion.Fig. 1. The BioAsteroid experiment.A NASA astronaut Michael Scott Hopkins performs the insertion of the experiment containers in KUBIK (credits: ESA/NASA); B The six hardware units inserted into the KUBIK onboard the ISS (credits ESA/NASA); C BioAsteroid logo, created by Sean McMahon (University of Edinburgh); D Flow diagram of the experiment. After preparation, samples were integrated into the experimental units (EUs) together with the medium and the fixative. The EUs were either launched to the ISS (blue oval), where they were installed in KUBIK incubators and subjected to microgravity (µg) or kept for incubation on Earth for the terrestrial gravity control (Earth g, yellow oval). Steps in green were part of the experimental time period (19 days). Storage passages were omitted for brevity.

Results

Meteorite characterisation

To determine the minerals and elements available for leaching, we thoroughly characterised the L-chondrite used in this experiment (Fig. 2, Tables 1, 2).Fig. 2. Characterization of the L-chondrite meteorite.A Backscatter electron image (BSE) and single elemental mapping for some of the major elements present in a pristine representative fragment of the L-chondrite used in this experiment (scale bar: 100 µm); B Phase analysis of a similar pristine fragment of the L-chondrite; C Photographic image of a pristine fragment analysed by Raman spectroscopy; D Composite Raman map with forsterite in red, enstatite in blue, and luminescence in green. Typical spectra are displayed for E forsterite, F enstatite, and G luminescence signal. Reference spectra (black) are displayed for the minerals.Table 1XRD analysis of pristine meteorite fragments, indicating percentage mineral composition (mean±st. error); n = 3Mineral%Chemical formulaForsterite ferroan47.4 ± 1.6(Mg_0.82_Fe_0.18_)(Mg_0.092_Fe_0.098_)(SiO_4_)Enstatite, ordered29.5 ± 0.4MgSiO_3_Anorthite12.1 ± 1.6Ca(Al_2_Si_2_O_8_)Melilite, syn5.7 ± 0.5Ca_8_Al_2_Mg_3_Si_7_O_28_/8CaO•Al_2_O_3_•3MgO•7SiO_2_Troilite5.4 ± 0.1FeSTable 2ICP-MS and ICP-OES analysis of pristine meteorite fragments, indicating the concentration of major and trace elements (mean±st. error); n = 3ICP-MSICP-OESElementµg/gElementmg/gMo6.032 ± 2.510Mg159.399 ± 18.262Ce1.122 ± 0.032Fe84.502 ± 8.279Nd0.809 ± 0.027S28.555 ± 2.956Pb0.551 ± 0.079Mn25.440 ± 2.561La0.437 ± 0.012Ca13.897 ± 1.389Dy0.248 ± 0.009Na8.001 ± 0.797Gd0.196 ± 0.006Al7.786 ± 2.838Ir****0.186 ± 0.017****Ni5.746 ± 0.312Yb0.168 ± 0.005K1.457 ± 0.091Pr0.164 ± 0.005P1.113 ± 0.130Er0.161 ± 0.005Cr1.015 ± 0.059Sm0.145 ± 0.005Ti0.677 ± 0.101Hf0.112 ± 0.005Co0.152 ± 0.023Pt****0.091 ± 0.009****Zn0.079 ± 0.006Ho0.054 ± 0.002Cu0.063 ± 0.004Eu0.053 ± 0.002Sr0.018 ± 0.002Rh****0.051 ± 0.005****Cd0.015 ± 0.001Th0.049 ± 0.001Zr0.013 ± 0.001Tb0.037 ± 0.001Ba0.008 ± 0.001Ta0.032 ± 0.001Pb0.007 ± 0.000Lu0.025 ± 0.001Ag0.000 ± 0.000Tm0.025 ± 0.001U0.019 ± 0.001Os****0.017 ± 0.004****Hg0.004 ± 0.001Tl0.003 ± 0.000Pd****0.002 ± 0.000Ru0.001 ± 0.000Elements are listed in order of abundance in the meteorite. Concentrations in bold indicate PGEs.

X-ray Diffraction Spectroscopy (XRD) analysis revealed that the dominant crystalline mineral phases (>5%) were olivine and enstatitic pyroxene (Table 1), typical of L-chondrites^31^. Minor components included feldspar (anorthite), melilite (sometimes associated with calcium–aluminium-rich inclusions in chondritic meteorites) and iron sulfides (toilite). Although XRD does not reveal solid metal inclusions, backscatter electron microscopy (BSE) combined with energy dispersive spectrometry (EDS) elemental mapping identified iron-nickel inclusions (Fig. 2A, B). Raman spectra were recorded at an array of points across the surface to identify minerals and map their distribution (Fig. 2C–G). Forsterite and enstatite were detected, as expected, and were heterogeneously distributed through the material^48,49^.

Bulk elemental analysis by inductively coupled plasma mass spectrometry (ICP-MS), and optical emission spectroscopy (ICP-OES; Table 2) showed that magnesium was the most abundant element (159.399 ± 18.262 mg/g), followed by iron (84.502 ± 8.279 mg/g), sulfur (28.555 ± 2.956 mg/g), manganese (25.440 ± 2.561 mg/g), calcium (13.897 ± 1.389 mg/g), sodium (8.001 ± 0.797 mg/g), and aluminium (7.786 ± 2.838 mg/g). The results are consistent with elements and mineral phases identified by BSE and XRD (Table 1, Fig. 2B).

Microbe-mineral interaction

To investigate the effect of microgravity and local mineral composition on microbe-mineral interaction, we performed Scanning Electron Microscopy (SEM) using secondary electron imaging (Fig. 3) and EDS (Supplementary Figs. 1–3).Fig. 3. Scanning electron microscopy (SEM) images of the L-chondrite fragments.Secondary electron SEM images are shown here and are representative of samples in the two gravity conditions (ISS for microgravity, Earth for terrestrial gravity). Images were acquired at varying magnifications to allow the visualisation of the features of interest. Scale bars are indicated in each figure, ranging between 2–20 µm.

Both the bacterium and the fungus were observed interacting with the meteorite material under both microgravity and terrestrial gravity. S. desiccabilis formed a contiguous biofilm across many areas of the rock’s surface in both the ISS and the Earth samples. Qualitatively, no gravity-dependent pattern for biofilm formation or cellular morphology were observed. P. simplicissimum successfully developed mycelia on the meteoritic rock fragments in both gravity conditions (Fig. 3), with no evident qualitative morphological change. Similar results were observed for the Consortium samples, with the bacterium and the fungus interacting in a similar fashion in both gravity conditions and forming a mixed filamentous (P. simplicissimum) and rod-shaped cell (S. desiccabilis) biofilm on the rock surface. Across all samples, bacterial (Supplementary Fig. 1), fungal (Supplementary Fig. 2), and Consortium (Supplementary Fig. 3) cells showed a tendency to interact with minerals bearing principally magnesium, oxygen, silicon (silicates), and less frequently with iron, copper, sulfur, chromium, manganese and other metals, in both gravity conditions. This is in accordance with the rock composition (Fig. 2, Tables 1, 2).

Microbial bioleaching of PGEs onboard the ISS

The concentration of 44 elements in the liquid fraction were measured by ICP-MS to assess elemental dissolution from the meteorite rock (Fig. 4, Supplementary Data 1). Statistical analysis (Supplementary Table 1, Supplementary Data 1) identified 22 elements as potentially relevant for further investigation. We initially focused our analysis on the three PGEs ruthenium (Ru), palladium (Pd) and platinum (Pt) (Fig. 5, Supplementary Tables 2, 3).Fig. 4. Analysis of biomining for 44 elements on the ISS and on Earth.For each element, represented in order of atomic number, bioleaching data are shown as percentage differences with the non-biological control. Upper panels indicate data from the ISS samples, lower panels from the Earth samples. Circles indicate mean values; error bars indicate standard error; vertical grey lines indicate 100% values (i.e., same bioleaching efficiency as non-biological samples).Fig. 5. Platinum group element (PGEs) biomining.In panels (A) and (B), data for the ISS experiment are shown. For each element, data are reported as: A percentage of non-biological control (%NB); B percentage extracted from the total concentration in the meteorite (%M). For A and B, diamonds indicate mean values, error bars indicate standard error. For (C) and (D), values are shown per element analysed, and per gravity condition (ISS = microgravity, green circles, Earth = 1 x g, orange circles). Data from the ISS and Earth experiment are shown (C) as percentage differences with the non-biological control (%NB); D as percentage extracted from the total concentration in the meteorite (%M). ISS data in C,D are the same as A,B. For C and D, circles indicate mean values; error bars indicate standard error; yellow vertical lines in panels A and C indicate 100% values (i.e., same bioleaching efficiency as non-biological samples).

ANOVA (Supplementary Table 1) indicated that the microbial species influenced the dissolution of all the PGEs analysed. Within the ISS samples, p-values of Tukey pairwise comparisons for ruthenium, palladium and platinum were >0.05 in all pairwise comparisons (Supplementary Table 2), with two exceptions: platinum dissolution differed for S. desiccabilis versus P. simplicissimum, and versus the Consortium, suggesting distinct effects of these organisms on platinum extraction.

To evaluate microbial enhancement of leaching under microgravity, metal concentrations were normalised to non-biological controls (%NB) for the ISS samples (Fig. 5A, Supplementary Table 3). P. simplicissimum showed enhanced mean leaching for all three PGEs, while S. desiccabilis had little to no effect, or even reduced leaching. The Consortium displayed %NB similar to P. simplicissimum alone, except for palladium, where it showed intermediate values between the fungus and the bacterium (Fig. 5A, Supplementary Table 2). In addition, elemental concentrations for ruthenium, palladium and platinum in the ISS were compared to their concentration in the meteorite (%M), to calculate the proportion leached from the rock under both non-biological and biotic conditions under microgravity (Fig. 5B, Supplementary Table 3). For all 3 PGEs, mean extraction was more efficient in the presence of P. simplicissimum, with 19.29%M of ruthenium, 11.91%M palladium and 0.29%M of platinum leached from the meteorite. The Consortium showed similar values for ruthenium and platinum, but not palladium, where extraction dropped to 3.74%M.

Microbial bioleaching of PGEs on Earth

Similarly to the ISS samples, we assessed whether the organisms enhanced PGEs leaching under terrestrial gravity (Fig. 5C, Supplementary Tables 1–3). On Earth, p-values of raw concentration comparisons for ruthenium, palladium and platinum were ≤ 0.05 in all pairwise comparisons between biological and non-biological samples, except for P. simplicissimum in the case of ruthenium and platinum. Comparisons between biological samples had p-values > 0.05.

Compared to non-biological samples (%NB), enhanced bioleaching of ruthenium and platinum was observed with S. desiccabilis and P. simplicissimum, both alone or in consortium, with increases spanning between 142.2 ± 19.9%NB to 218.9 ± 67.5%NB. In contrast, palladium leaching was reduced by the presence of the microbial species. The total amount of PGEs leached from the rocks (%M) are reported in Fig. 5D and Supplementary Table 3.

Effect of microgravity on microbial-mediated PGEs bioleaching

In addition to the influence of the microorganisms, ANOVA results revealed that gravity influences the dissolution of all the PGEs analysed (Supplementary Table 1). The interaction between the variables gravity and organism had an effect for palladium and platinum, but not ruthenium. For ruthenium and palladium, all pairwise comparisons of %M and %NB values between microgravity and terrestrial gravity yielded p-values > 0.05, except for %NB of the Consortium for palladium (p = 0.009). For platinum, the comparison of %M in P. simplicissimum returned a p-value of 0.007, although the corresponding %NB comparison had p-value > 0.05. Conversely, for S. desiccabilis the %NB comparison produced a p-value = 0.01, while for %M was >0.05. All other comparisons produced p-values > 0.05 (Supplementary Tables 2, 3, Supplementary Data 1, Fig. 5C, D).

Bioleaching of other elements on ISS and Earth

Among the 22 elements identified by ANOVA as potentially relevant (Supplementary Table 1), and beside PGEs, 15 showed at least one biologically-relevant pairwise comparison with p-value ≤ 0.05, either when comparing the raw concentrations or the %NB (Supplementary Data 1).

In the ISS samples, only phosphorus reported a p-value of 0.049 for P. simplicissimum versus non-biological control. All other pairwise comparisons returned p-values > 0.05 (Supplementary Fig. 4, Supplementary Table 4, Supplementary Data 1). However, trends were evident from the fold-change between mean raw concentrations, using a ≥1.5-fold threshold (Supplementary Table 5). Compared to non-biological controls, copper leaching was reduced by S. desiccabilis, while phosphorus leaching increased in the presence of P. simplicissimum, alone and in Consortium. The fungus (alone and in Consortium) also enhanced bioleaching of phosphorus, vanadium and copper compared to the bacterium. No difference was present comparing the fungus and the Consortium.

On Earth, pairwise comparisons with p ≤ 0.05 (Supplementary Data 1) indicated increased bioleaching for 7 elements with S. desiccabilis (potassium, manganese, iron, nickel, strontium, zirconium and molybdenum), 2 elements with P. simplicissimum (vanadium and manganese), nine elements with the Consortium (potassium, vanadium, manganese, iron, nickel, strontium, zirconium, barium, europium). Comparing average concentrations (≥1.5-fold increase threshold, Supplementary Table 5, Supplementary Data 1), S. desiccabilis surpassed P. simplicissimum iron, cobalt, and molybdenum leaching, while the fungus increased copper leaching compared to the bacterium and the Consortium. The Consortium did not improve bioleaching compared to the individual organisms.

To highlight the effect of gravity on bioleaching, comparisons between ISS and Earth conditions were performed within each microbial treatment (Supplementary Fig. 4, Supplementary Tables 4, 5, Supplementary Data 1). P-values ≤ 0.05 for raw concentration comparisons were found only for lutetium in P. simplicissimum. When comparing %NB p-values, the Consortium showed p ≤ 0.05 for sodium, copper and zinc. All other pairwise comparisons, either for the raw concentrations or for the %NB, were >0.05. Fold-change trends (≥1.5) from mean concentrations indicate S. desiccabilis and the Consortium leached copper more effectively on Earth compared to ISS, and lutetium more effectively on ISS compared to Earth (Supplementary Table 5). P. simplicissimum alone showed higher average concentration in microgravity compared to Earth gravity for iron, cobalt, nickel, zirconium, molybdenum, europium, and lutetium, but higher on Earth for copper.

Effect of microgravity on abiotic leaching

To evaluate the effect of the gravity condition on abiotic (i.e., non mediated by the microorganisms) leaching, we compared non-biological samples in microgravity (ISS) and Earth gravity (Earth).

For PGEs (Fig. 5D, Supplementary Tables 2, 3), palladium and platinum comparison reported a p-value ≤ 0.05, but not ruthenium, with mean extraction from the rock of 14.8 ± 2.2%M in microgravity versus 6.6 ± 2.1%M under terrestrial gravity for ruthenium (2.2-fold increase on ISS), 2.2 ± 0.6%M on ISS versus 29.5 ± 6.7%M on Earth for palladium (13.6-fold increase on Earth), and 0.2 ± 0.0%M in microgravity versus 0.13 ± 0.02%M under terrestrial gravity for platinum (1.8-fold increase on ISS).

Beyond PGEs, additional 9 elements reported a p-value ≤ 0.05 for comparisons of non-biological controls between gravity conditions. Enhanced leaching in microgravity was observed for, in order of atomic number, aluminium (6.8-fold increase; ISS: 1.52 ± 0.95%M, Earth: 0.22 ± 0.12%M), scandium (3.4-fold; concentration in the rock was not detected. Absolute values in the leachate: ISS = 0.020 ± 0.008 ng/mL, Earth = 0.006 ± 0.002 ng/mL), iron (4.3-fold; ISS: 0.12 ± 0.01%M, Earth 0.03 ± 0.01%M), cobalt (2.5-fold; ISS: 0.50 ± 0.09%M, Earth: 0.21 ± 0.07%M), nickel (2.4-fold; ISS: 0.15 ± 0.01%M, Earth: 0.06 ± 0.01%M), strontium (1.8-fold, ISS: 0.39 ± 0.01%M, Earth: 0.21 ± 0.04%M), molybdenum (2.9-fold, ISS: 0.03 ± 0.01%M, Earth: 0.01 ± 0.00%M) and erbium (2.4-fold, ISS: 0.003 ± 0.0004%M, Earth: 0.001 ± 0.0004%M). Enhanced abiotic leaching under terrestrial gravity was observed for sodium (1.4-fold increase; ISS: 0.57 ± 0.07%M, Earth: 0.82 ± 0.11%M). Silicon also showed a similar result (1.2-fold increase; ISS: 0.0097 ± 0.001%M, Earth: 0.011 ± 0.001%M), but was excluded by further interpretation due to detection issues in the meteorite matrix (see Materials and Methods, Supplementary Tables 4, 5, Supplementary Data 1).

Metabolomics of biomining microorganisms in microgravity

To determine whether microgravity alters the metabolome of biomining microorganisms, and thus contributes to their distinct behaviour in space, we conducted a metabolomics analysis of the liquid fraction recovered after the experiment. Principal component analysis (PCA) of all samples, from both microgravity (ISS) and Earth gravity (Earth), showed substantial overlap across the groups, except for a single outlier (an ISS S. desiccabilis sample), indicating no dominant component that could allow the discrimination of the samples based on gravity condition or organism (Fig. 6A). PC1 and PC2 taken together explained 60.5% of the observed variance. Volcano plot analysis (Supplementary Fig. 5) showed a larger number of features that were up- or downregulated in space compared to Earth, particularly in fungus-containing samples.Fig. 6. Metabolomic analysis of microorganisms during microbe-meteorite interaction in space and on Earth.A Principal component analysis (PCA) representing all the samples; B PCA showing ISS samples only; C PCA showing Earth samples only. Ellipses correspond to 95% confidence intervals. D Heatmap showing the top 25 features measured in the metabolomics analysis, clustered by gravity condition and organism. Each column represents a single sample. Numbers of the left (1-6) represent visual clusters suggesting metabolomic patterns dependent on the organism and/or the gravity condition. Heatmap legend is on the left side of the figure, indicating red colours for higher concentrations and blue for lower concentrations of features in each specific sample. Features with long names were indicated with their chemical formula, with asterisks showing the complete names at the bottom of the figure.

To better appreciate gravity-independent metabolomic signatures, PCAs was performed separately for ISS (Fig. 6B) and Earth (Fig. 6C) samples. Within the ISS group, S. desiccabilis samples clustered partially with non-biological controls, except for the noted outlier. The fungus-containing samples, both monocultures and the consortium, formed distinct clusters from the non-biological controls and the S. desiccabilis samples. PC1 and PC2 together explain 75.5% of the variance. In contrast, PCA analysis of the Earth samples (Fig. 6C) shows 4 separate clusters, reflecting unique metabolic profiles for S. desiccabilis, P. simplicissimum and the Consortium under terrestrial gravity conditions.

A heatmap of the top 25 detected features (Fig. 6D) highlighted 6 distinct clustering patterns, indicated by numbers on the left side of the heatmap, associated with both gravity condition and/or the organism. Notably, Cluster 1 comprised metabolites strongly associated with the presence of the fungus, either in a single culture or in Consortium, with elevated abundances in microgravity (ISS). These trends were consistent with the broader pattern observed in the heatmap showing the top 70 features (Supplementary Fig. S6).

Discussion

This study provides proof of concept for microbial bioleaching of 44 elements from asteroidal rock (L-chondrite) in microgravity. Specifically, the work demonstrated that P. simplicissimum can enhance the extraction of critical elements, including PGEs, in space. In particular, palladium extraction increased 5.5-fold relative to abiotic controls in microgravity, with leaching yields reaching 549.3 ± 234.4% of the non-biological control. Ruthenium and platinum also showed moderate enhancement, along with phosphorus, an element critical for both industrial and life support systems. These results point to the feasibility of using fungi to facilitate element recovery from asteroid for ISRU. High variability was observed, likely due to differences in microbial growth rates and the heterogeneous nature of the meteorite. Small sample volumes and limited replicates, inherent to spaceflight experiments, may have further amplified this variability.

Bioleaching performance varies substantially by organism. P. simplicissimum was the most effective bioleaching organism under microgravity, while the bacterium S. desiccabilis performed similarly or worse than non-biology controls for most PGEs. This might be due to its production of biofilms and extracellular polysaccharides^50^, which have been shown to protect metal and concrete surfaces from biotic and abiotic corrosion^51–53^, despite generally considered beneficial for bioleaching^14,54^. Although SEM imaging confirmed the presence of a biofilm, their sporadic nature does not allow us to link biofilm growth to leaching yield. These findings may be consistent with results from our previous space biomining experiment BioRock^22,23^, where S. desiccabilis extracted rare earth elements (REEs) and vanadium more effectively under higher gravity regimens. The difference in substrate might also be important. Vanadium extraction from the basalt in BioRock was 184.92 ± 75.33%NB of the control in microgravity^22^, compared to 71.40 ± 13.97%NB from the L-chondrite here. Only 2 REEs (europium and lutetium)^23,55^ were bioleached in this study, compared to the full REE range in BioRock^23^.

When a microbial consortium was used, the leaching yield for all three PGEs and phosphorus exceeded those of non-biology controls in microgravity, though variability remained high. The data suggest the fungus was the primary contributor for ruthenium and platinum, while for palladium, the fungus alone outperformed the consortium, possibly due to the antagonistic effects of the bacterium.

Gravity clearly influenced leaching outcomes, both biotic and abiotic. Abiotic leaching in microgravity changed for 11 elements. Nine, including platinum (1.8-fold), increased under microgravity, while two decreased. Palladium leaching dropped 13.6-fold in space, further highlighting P. simplicissimum’s compensatory role. These changes may reflect altered fluid dynamics in microgravity^56^, such as reduced convection, leading to local saturation of the leached elements around the rock surface. However, this would not explain why this was only observed for certain elements.

Bioleaching outcomes were also notably different between Earth and space. Although bacterial extraction of palladium remained below the non-biological sample, it increased 13.6-fold in space. For the fungus, iron, cobalt, nickel, zirconium, molybdenum, ruthenium, europium, lutetium and platinum yields were higher in microgravity, while yields for 1 element (copper) declined. For S. desiccabilis and the Consortium, leaching of copper was enhanced on Earth, while lutetium was enhanced on the ISS (Supplementary Table 5).

To investigate potential mechanisms behind these differences, we examined microbial metabolomic profiles under Earth and space conditions. Data on microbial metabolomics under real or simulated microgravity are scarce and are mostly focused on bacterial rather than fungal metabolites^57^. Secondary metabolites of interest in previous experiments were mostly antibiotics, but also molecules such as poly-β-hydroxybutyrate (a polyester)^57,58^.

Our analysis demonstrated the role of microgravity in shaping the metabolic profile of bioleaching microorganisms, revealing gravity-dependent shifts in metabolite composition, some of which may be linked to bioleaching activity. Although key organic acids known to support leaching (e.g., citric, oxalic, malic, gluconic and glucuronic acids)^45,59,60^ were not detected, we identified other carboxylic acids, siderophore-associated molecules, and molecules of potential pharmaceutical and bioplastic production interest (Supplementary Table 6). PCA analysis showed greater metabolic diversity under terrestrial gravity, suggesting greater adaptability in terrestrial conditions with possible minor contributions from contaminants (Supplementary Note 1, Supplementary Fig. 11).

These data highlight the complexity of predicting optimal conditions for biomining, particularly in the context of space. While extraction yields in this study were lower than those typically observed in terrestrial biomining with P. simplicissimum^44–46^, the fungus was still able to extract 11.9% of the palladium present in the meteorite. This is a promising outcome, considering the differences between our experimental setup and those in industrial scale operations (e.g., ore pretreatments and leachate optimisation). The lower yields likely reflect the lower concentration and accessibility of metals in meteorites compared to industrial substrates (e.g., salt solutions, industrial ash). However, in the context of in situ resource utilization (ISRU) for space missions, maximising the use of extraterrestrial resources at hand, including asteroids or basalt, becomes critical^21–23^. In this context, microbial systems offer unique advantages: not only can they enhance extraction in specific cases, but they may also stabilize yields when abiotic leaching is reduced. For example, based on the current palladium market value of 36.39/g (source [www.kitco.com](http://www.kitco.com), retrieved on 27 June 2025), the palladium extracted under microgravity by *P. simplicissimum*, using a fully scaled up bioleaching process (i.e., 1000 m^3^ tank) and our experimental conditions would amount to approximatively ~10. While this is economically negligible, removing the fungus would result in an economic loss of 545%. Although space biomining is unlikely to be profitable in the near term, especially when judged against terrestrial economic standards, its value lies in enabling resource self-sufficiency in space. With improved technologies and upscaling, microbial extraction could become a key component of sustainable life-support systems in long-duration missions. Importantly, our results suggest that selecting optimal microbial species, rock types, and environmental conditions will be critical for effective space biomining.

Finally, we observed direct interaction of fungal mycelium and bacterial biofilms with extraterrestrial minerals in microgravity, on the ISS. They were frequently associated with magnesium-, oxygen- and silicon-harbouring minerals (silicate minerals), rather than sulfur- and iron- ones (sulfide minerals). This could suggest a preference for the greater elemental diversity of silicate minerals compared to the sulfides, or an avoidance of elements with potentially higher toxicity. No major morphological differences were observed between microgravity and terrestrial gravity. To our knowledge, this is the first demonstration of metabolically active microbe-meteorite interaction and mycelium formation on extraterrestrial materials in space. These results have implications beyond biomining, either in space or on Earth, including for rock weathering processes that could release nutrients^2,4,5,20,61^, such as phosphorus, potassium and iron, helping to establish sustainable life support systems beyond Earth.

Methods

Strains and medium

The microbial species used for this work were the bacterium Sphingomonas desiccabilis CP1D^62^, a Gram-negative, non-motile and non-spore-forming bacterium, first isolated from soil crusts in the Colorado plateau^40^ with demonstrated capacity to extract metals during spaceflight^22,23^, and the fungus Penicillium simplicissimum DSM 1078 (DSMZ), an Ascomycota known for its capacity to perform bioleaching^42–45^.

The medium used for this experiment was a solution of 50% v/v R2A medium^63^, chosen to encourage bacteria to extract nutrients from the meteorite. Five millilitres of medium were used in this experiment for each sample. The medium composition was (g L^−1^): yeast extract 0.25; peptone 0.25; casamino acids 0.25; glucose 0.25, soluble starch 0.25, Na-pyruvate 0.15; K_2_HPO_4_ 0.15; MgSO_4_.7H_2_O 0.025 at pH 7.2.

The fixative selected to prevent degradation of the biological portions and to stop microbial growth after the end of the experiment was RNAlater (Thermo Fisher), an aqueous and non-toxic storage solution compatible with the astronaut safety requirements on the ISS. One millilitre of fixative was used for each sample, with a final volume ratio of 1:5 fixative-medium.

Rock samples

The extraterrestrial rock sample used for this work was the Northwest Africa (NWA) 869 meteorite, a L3-6 chondrite regolith breccia^64,65^. A portion of the meteorite was crushed into irregular pieces of approximately 1–3.5 mm of diameter. Rock fragments were aliquoted in samples of 0.79 ± 0.14 g (mean±st. dev.) each and sterilised by dry-heat sterilisation in a hot air oven (Carbolite Type 301, UK) for 4 h at 250 °C.

Average surface area of the meteorite fragments was measured by gas adsorption analysis (Quantachrome, Nova Touch), to measure the surface available to the microorganisms for bioleaching and biofilm formation. The average surface area for the meteorite’s fragments was 1.94 ± 0.18 m²/g (mean ± standard error, n = 3), hence each sample provided ~1.5 m^2^ of surface available.

Sample preparation for the spaceflight

Single strain cultures of each species were desiccated on sterile rock samples

For S. desiccabilis CP1D, an overnight culture of the microorganism was grown in 100% v/v R2A medium at 20 °C–22 °C. When the culture reached stationary phase (OD_600_ = 0.88 ± 0.09, corresponding to 4.7 ×10^10^ CFU/mL), crushed sterile meteorite was soaked in 1 mL of the bacterial culture for S. desiccabilis and Consortium cultures and samples were air-dried at ~20 °C–25 °C in a laminar flow hood under sterile conditions.

The mycelium of a 7-day old pre-culture of P. simplicissimum (50 mL) was dissolved by sonication (Microson ultrasonic cell disruptor, Misonix) with continuous pulse at setting 3 for 2 min, and then filtered through a sterile cotton bud to remove larger bids of mycelium and obtain a homogeneous fungal solution. This procedure did not alter fungal viability (data not shown). One mL (containing ~6 ×10^6^ CFU/mL) of the resulting liquid fraction was used to inoculate the sterile crushed meteorite samples P. simplicissimum and Consortium in sterile 6-well plates, and these were air-dried overnight at ~20 °C–25 °C with a sterile procedure within a laminar flow-hood.

Non-biological controls were sterile crushed meteorite samples without cell inoculation

After preparation, all samples were stored at room temperature (~20 °C–25 °C) until integration in the BioMining Reactors (BMRs).

Flight experimental setup

Flow diagram summarising the BioAsteroid experiment setup is available in Fig. 1. Sample, medium and fixative integration into each Experiment Unit (EU)^66^ (KEU-RK, from Kayser Italia) was performed under aseptic conditions before the launch. Each EU was composed of two BioMining Reactors (BMRs). A complete description of the EU can be found in the literature^66^. After integration, the culture chamber contained the meteorite fragments on one side, held in place behind an aluminium grid to avoid dispersion of the rock pieces in the culture chamber, and a semipermeable silicone rubber membrane, to allow gas diffusion, on the remaining five sides. A small sterile piece of cotton ball was inserted between each rock sample and the EU back cover, to protect the rock pieces from excessive shaking during the rocket launch and space operations. Each BMR was connected to a 5 mL medium reservoir and a 1 mL fixative reservoir that were activated at the appropriate time. Each EU was integrated into an Experiment Container (EC, KIC-SLA-E3W, Kayser Italia), equipped with temperature loggers (installed in one EUs; Signatrol SL52T sensors, Signatrol, UK) and accelerometers (in all EUs on ISS). A total of 12 samples in 6 EUs for the flight experiment and 18 samples in 9 EUs for the Earth samples were prepared on different timelines. After integration of the 6 flight EUs between September 29th and October 2nd, 2020, the experiment was shipped to NASA Kennedy Space Centre (Florida, USA), while being stored at room temperature, and launched to the International Space Station (ISS) on a SpaceX Falcon-9 rocket (Commercial, Resupply Mission 21 mission) on December 6th, 2020. On arrival to the ISS, the samples were stored at room temperature (23.0 °C, temperature loggers) until installation into the microgravity (non-centrifuged) slots within the two KUBIK (ESA) incubators (5 and 6, Fig. 1) aboard the ISS, previously set to 20 °C, on December 20th, 2020, when the automatic timeline of the EUs was activated and medium (5 mL) was injected in consecutive manner to each culture chamber. All crew activities were performed by NASA astronaut Michael S. Hopkins (Fig. 1). Samples grew for 19 days at 19.5 °C (logged data). At the end of the experiment, 1 mL of fixative was automatically injected into the culture chambers on January 8th, 2021) and hardware was cold stored at 1.5 °C–11.5 °C (logged data). On orbit, the EUs were stored in the refrigeration unit of the MELFI (Minus Eighty degrees Laboratory Freezer for the ISS) hardware (i.e., not frozen), and were downloaded to Earth in cold storage bags (NASA-supplied passive temperature controlled facilities), in the SpaceX CRS-21 Dragon capsule (the same vehicle as for upload). Samples were shipped in cold storage to the University of Edinburgh (UK), where samples were retrieved after 12 days from the fixative injection.

Of the 9 EUs containing 18 Earth samples, 2 EUs, containing 4 non-biological controls, were prepared on March 16th, 2021, while the remaining 14 EUs were integrated on April 19th, 2021. All the Earth samples were subject to analogous procedures and conditions to those occurring in the flight hardware, with incubation at 20 °C in a laboratory incubator (Memmert). Medium and fixative were injected by manual manipulation of the appropriate screws. Fixative injection occurred after 19 days from the medium injection, similarly to the flight experiment. After fixative injection, EUs were stored at 8 °C for 12 (samples prepared in March) or 14 (samples prepared in April) days, until sample retrieval. The difference in storage timing was due to technical reasons and did not affect the results (data not shown).

Post-flight sample recovery

Samples were recovered separating the culture liquids, the meteorite fragments, the metal grids and the membranes.

Liquid cultures were treated differently depending on the species. While non-biological controls and S. desiccabilis samples did not require pre-treatments, liquid samples containing the fungus were homogenised by sonication (Microson ultrasonic cell disruptor, Misonix) with continuous pulse at setting 3 for 60 s and then filtered through a sterile commercial cotton ball, to dissolve the mycelium and obtain a homogeneous solution. An aliquot of 1 mL was recovered from each liquid sample and immediately stored at –80 °C for the metabolomics analysis, 0.5 mL were collected for pH measurement, 0.2 mL were collected for colony forming unit (CFU) and optical density analysis. 0.25 mL were collected for flow cytometry analysis and treated as described below, the remaining aliquot of the liquid cultures were treated for ICP-MS analysis as described in the dedicated section.

An aliquot of the rock fragments was recovered, washed once with sterile water and air-dried at ~20 °C–25 °C in a laminar flow hood under sterile conditions. These samples were analysed by XRD and Raman (data not shown). Other representative aliquots of rock fragment were stored in 4% v/v formaldehyde at 4 °C to preserve biofilms and mycelia. Remaining rock fragments were treated for scanning electron microscopy analysis as described below.

Final cell concentration

Final cell concentration was measured from the liquid fraction of the samples using three distinct methods: (i) measurement of the turbidity of the culture by spectrophotometric analysis; (ii) flow cytometry; (iii) counts of colony forming unit (CFU).

• (i)Optical density (OD) was measured at wavelengths (λ) of 600 nm and 530 nm from 100 µL of the liquid fraction of each sample. Traditional OD_600_ was used to measure final cell concentration. However, due to the iron bioleaching from the L-chondrite, the liquid fraction had a strong orange/red coloration for some samples, which influenced the measurement. For this reason, and because of the presence of the fungus^67^, non-standard OD_530_ was also measured.
• (ii)Flow cytometry was measured with a LSR Fortessa machine (BD Biosciences). Equipped with a 405 nm laser, detecting the emissions of Calcofluor White Stain (fungi-specific dye, Sigma Aldrich) binding through a 450/50 nm band pass filter, and a 488 nm laser with a 530/30 nm filter to excite BacLight Green Bacterial Stain (bacteria-specific dye, Invitrogen), as were forward (FSC) and side (SSC) scattering. A volume of 250 µL of the liquid fraction of each sample was washed once with a filter-sterile solution of Tween 80 at 0.1% v/v in PBS, then cells were fixed for 15 min at room temperature in a filter-sterile solution of 1% v/v formaldehyde in PBS. Finally, the liquid was removed and replaced after centrifugation for 5 minutes at maximum speed, with 250 µL of filter-sterile PBS. Samples were stored at 4 °C until analysis. To have an estimation of final cell concentration, a volume of 100 µL of the liquid fraction of the samples, appropriately diluted in the diluent (PBS filtered with a 0.22 µm nylon filter), were acquired at a flow rate of 2–3 µL/s, and all events were counted. Samples were stained with BacLight Green at a final concentration of 0.1 µM, Calcofluor White at a final concentration of 0.25 µg/mL, both or none. When possible, each sample was measured twice per dye (i.e., 2 technical replicates). Appropriate gating was constructed using the software BD FacsDiva 8.0.1, to distinguish bacterial from fungal cells (gating strategy is reported in Supplementary Fig. 7). Events in Bacteria and Fungi gates were counted and considered as single cells, to reconstruct final cell concentrations, expressed as cells/mL.
• (iii)To measure colony forming units (CFU), serial dilutions of the liquid fraction were prepared, and 6–10 spots of 10 µL (for a total volume of 60–100 µL, respectively) of each dilution were spotted on R2A solid medium. These were incubated at room temperature for 2–5 days, until single colonies became visible. These were counted from the lowest dilution in which they were clearly distinguishable, and colonies of each spot, for each sample, were summed. Final CFU concentration (CFU/mL) was then calculated with the formula: [(total colonies) x dilution) / total volume].

The 3 methods described above were compared building a growth curve for each organism (S. desiccabilis or P. simplicissimum) for 19 days (the timeframe of this experiment) in a separate ground-based experiment, and measuring cell concentration at each datapoint with the 3 techniques (Supplementary Figs. 8–9).

As CFU assay showed a potential bacterial contamination of some of the samples (see Supplementary Note 1), the genomic DNA of the affected samples, as well as that of the isolated contaminant species, was extracted with DNeasy PowerLyzer Microbial Kit (QIAGEN) to assess if the contamination was present in the original samples, or introduced later. The V3-V4 region of the 16S rDNA was amplified by PCR using the universal primers 341 F/805R^68^, and the Q5 HighFidelity DNA polymerase (NEB). Prior to the addition of the DNA and the primers, the PCR master mix was treated with 0.02% v/v of DNAse I 1 U/mL (Zymo) at room temperature for 15 minutes, followed by DNAse I deactivation at 75 °C for 15 min, following the manufacturer’s instructions, to ensure complete decontamination of the master mix^69^. The PCR was performed following the manufacturer’s instructions, using a T_m_ = 60 °C and 30 cycles. Amplicons were checked on a 1.5% w/v agarose gel, and sent for Sanger sequencing with primers 341 F and 805 R to an external facility (MRC PPU DNA Sequences and Services, Dundee, UK). The sequences obtained were compared with sequences from the GenBank database using BLASTN (NCBI), and EZBioCloud (CJ Bioscience), for the bacterial identification.

Inductively Coupled Plasma Mass Spectrometry (ICP-MS)

One millilitre of the liquid fraction of each sample was treated with nitric acid (final concentration 4% v/v), and the samples were analysed by ICP-MS, to determine concentrations of the elements bioleached from the meteorite. ICP-MS analysis was also carried out on the medium (50% v/v R2A) and fixative (RNAlater).

All samples were analysed for a variety of elements using an Agilent 8900 ICP-MS instrument employing an RF (radio-frequency) forward power of 1550 W, RF matching of 1.8 V, with Argon gas flows of 1.02 L/min and 0.90 L/min for nebuliser and auxiliary flows, respectively. Sample solutions were taken up into a micro mist nebuliser by peristaltic pump at a rate of approximately 1.2 mL/min. Skimmer and sample cones were made of nickel. The instrument was operated in spectrum multi-tune acquisition mode (three replicates runs per sample) for the three isotopes ^101^Ruthenium, ^105^Palladium and ^195^Platinum using Helium mode with a flow rate of He 5 mL/min. To calibrate the instrument, a multi-element calibration standard containing each element was prepared using 1000 mg /L single-element standards (SPE Science, Canada) diluted with 2% v/v HNO_3_ (Aristar grade, VWR International, United Kingdom). The limits of detection for each element in He mode were 0.005, 0.005 and 0.007 µg/L for ^101^Ruthenium, ^105^Palladium and ^195^Platinum respectively. Raw ICP-MS data (determined in μg/L) was converted to obtain the absolute quantity of a given element in the culture chamber, taking into account dilution factors applied during ICP-MS analysis.

To determine elemental concentrations in the L-chondrite material, between 25 and 50 mg of homogenised pristine sample (x3) was added to Savillex Teflon vessels. Rock standards (georem standards BCR2, BHVO1 and B-EN) were prepared in the same way. Two blanks were included (i.e., sample without L-chondrite). Three millilitres of double distilled HNO_3_, 2 mL HCl and 0.5 mL HF was added to each of the vessels. HF was added after the other acids to prevent disassociation, formation and precipitation of aluminium fluorides. The HF addition is a necessary step in this protocol, however it compromises the detection of silicon from the rocks, due to its volatilisation. Samples were placed on a hot plate for digestion overnight (temperature of 100 °C–120 °C) and checked for complete digestion. Samples were evaporated on the hot plate. Five millilitres of 1 M HNO_3_ was added to each vessel. Lids were added and the samples returned to the hot plate for a second digestion step. Samples were further diluted with 2–5% v/v HNO_3_ for ICP-MS analysis. Analysis was carried out on a high resolution, sector field, ICP-MS (Nu AttoM). The ICP-MS measurements for elements were performed in low resolution (300), in Deflector jump mode with a dwell time of 1 ms and 3 cycles of 500 sweeps. Data were reported in micrograms of element per gram of chondrite.

Scanning Electron Microscopy and elemental mapping

Representative samples of rock (~0.3 g) with or without microbial growth were stored in a solution of 3% v/v glutaraldehyde in 10 mM HEPES buffer, pH 7.0 for 5 days at 4 °C. After this period, stepwise dehydration with graded series of 10, 30, 50, 70, 90, and 100% v/v ethanol was performed for 10 min each. Samples were stored at 4 °C prior to drying with liquid carbon dioxide in a Polaron E3100 critical point dryer to preserve cell morphologies. Samples were then affixed to SEM aluminium stubs (Agar Scientific) using a small quantity of conductive carbon glue (Agar Scientific) and coated with 20 nm of gold with a sputter coater (Denton Vacuum) to enhance conductivity for secondary electron imaging.

Further samples were mounted in epoxy resin and polished before carbon coating (Denton BTT-IV carbon evaporation coater) for backscatter electron (BSE) imaging and EDS element mapping. Samples were stored in plastic boxes to prevent dust contamination prior to imaging and analysis using a Carl Zeiss SIGMA HD VP field emission SEM with an Oxford Instruments AZtec EDS system at the School of GeoSciences, University of Edinburgh.

Raman

Raman spectra were recorded with a fibre optic Raman probe and 785 nm stabilized diode laser (Ocean Insight). The probe was mounted to a motorized X-Y-Z translation stage and scanned across the sample surface. Raman spectra were recorded at ca. 0.1 mm lateral resolution and the probe height was adjusted in Z at each point to maximize the Raman signal. The resulting maps were analysed by comparing the Raman peaks at each spectrum to mineral Raman spectra from the RRUFF database to assign a mineral intensity. The broad background fluorescence intensity is the sum of the entire spectrum from 200–2000 cm^−1^. The intensity of sharp luminescence peaks is found by summing the spectral region between 1200–600 cm^−1^.

Metabolomics analysis

Polar and non-polar metabolites were analysed using liquid chromatography coupled to high resolution mass spectrometry. Polar metabolites were prepared by diluting the samples a ratio 1:5 (sample/buffer) in extracting buffer (40% v/v MeOH, 40% v/v MeCN and 20% v/v H_2_O) prior to injection. Non-polar metabolites were enriched bz bi-phasic extraction using ethyl-acetate. Metabolites were extracted by vortexing the tubes for 20 min with subsequent spinning down. The organic layer was evaporated and reconstituted in 50%(v/v MeOH- 50% H_2_O v/v prior to injection.

During metabolite analysis, a pHILIC column (Merck, Germany) was used to separate polar metabolites, and a Luna C18 (Phenomenex, United States) to separate non-polar metabolites. An Ultimate 3000 HPLC (Thermo Fisher Scientific, Germany) coupled to a Q-Exactive mass spectrometer (Thermo-Fisher Scientific, Germany) operated in polarity switch mode was used. Pooled samples, chemical standards and procedure blanks were also analysed. Detailed description of the methods are included in literature^70–72^.

Peak detection and integration from Raw data were performed using Compound Discoverer 3.2 (Thermo-Fisher Scientific). An automatic filter set was applied initially to remove features of low quality. Features marked as background signals (with a retention time below one minute, or whose annotated mass diverged by >5 ppm from measured mass) were removed. Features with at least two partial matches on reference databases [mzCloud (HighChem LLC), mzVault (Thermo-Fischer Scientific), ChemSpider (Royal Society of Chemistry), and a list of known standards] and full fragmentation data were considered appropriate for further analysis. Partial matches were not discounted as inconsistencies in database entries may affect the match strength without invalidating the annotation. If a feature passed filtering solely based on partial matches, its predicted structure was manually confirmed to be identical to that of the database entry, to ensure correct annotation. Finally, features of insufficient total signal area were removed. Once filtered, all features significantly associated with a condition of interest were identified via differential analysis, with significance determined by p ≤ 0.05. Peaks with a maximum intensity < 4×10^6^ counts were removed to ensure sufficient separation from background signals. Features were then assessed individually in greater detail via the metrics produced by Compound Discoverer 3.2. Data analysis and figures were produced using the open source MetaboAnalyst 5.0 program^73^.

X-Ray Diffraction (XRD)

Pooled samples were analysed using X-ray Diffraction at the School of Geosciences, University of Edinburgh. Fragments of pristine L-chondrite were gently crushed in a mortar and pestle into a powder. The powder (~1 g each) was mounted on clean plastic slides. Care was taken to use as little compressional force as possible to minimise preferred mineral grain orientation. The samples were fed into a Bruker D8-Advance X-ray Diffractometer, using a 2-theta configuration in which the X-rays were generated by a Cu-anode X-ray tube operating at 40 kV and a tube current of 40 mA. Diffracted X-rays were detected using a sodium iodide scintillation detector. The samples were scanned from 2 to 60 degrees two theta with a scan rate of 0.02° per second. Resultant diffractograms were compared to the International Centre for Diffraction Data (ICDD) diffractogram database library (2012 issue) using the EVA analysis package. Typically, this procedure gives a detection limit for crystalline phases of approximately 1 wt.%. To quantify mineral abundances in the samples, the diffractograms were subject to Rietveld analysis using the TOPAS software package. This involved identifying the mineral assemblage present by comparing peak positions and heights with those in the powder diffraction database. The TOPAS program then generated a ‘model’ diffraction pattern, calculated from an initial estimated mineral assemblage. The differences between the two are reduced iteratively, which typically takes around 100 iterations, until the model and observed patterns converge, revealing the amounts of the minerals in wt.%.

Data analysis and figure production

Statistical analysis of bioleaching was performed using RStudio 2023.03.0 Build 386 and Microsoft Excel for Microsoft 365 MSO (Version 2303 Build 16.0.16227.20202) 64-bit. ICP-MS leaching data of the liquid fractions were analysed using a two-way ANOVA, followed by Tukey post-hoc analysis (n = 3 for ISS samples, n = 4 for biological Earth samples, and n = 6 for non-biological Earth samples). Bioleaching data normalised to the non-biological samples (%NB) were compared using a two-tail Student’s t-test, which allowed comparisons of %NB values of same-organism samples between gravity conditions. To reveal the effect of gravity on overall leaching, we compared the raw concentrations of element extracted (ng/mL, equivalent to comparing %M) by performing Tukey pairwise comparisons between ISS and Earth samples harbouring the same organism(s). To analyse the effect of gravity on the organisms, we performed the same comparisons using a Student’s t-test on the concentrations normalised to the non-biological controls (%NB), which allowed to remove the effect of abiotic leaching in our samples.

Figures were produced using RStudio 2023.03.0 Build 386 and Inkscape 1.1.

Supplementary information

Supplementary information Supplementary data 1