Protein without farms: What comparative genomics reveals about “Power-to-Food” microbes

Conventional agriculture is increasingly incompatible with planetary boundaries, such as land and water demand, greenhouse-gas emissions, and disruption of the nitrogen cycle. Hydrogen-oxidizing bacteria (HOB) enable a scalable “power-to-food” approach in which aerobic gas fermentation turns CO₂ and renewable H₂, along with N₂ in some strains, into protein-rich biomass, largely decoupling protein production from arable land and climate variability. The same chemistry is attractive for closed-loop space life support, where crew CO₂ and electrolysis-derived H₂ can be recycled into edible biomass. Here, we compare two leading HOB chassis strains, Cupriavidus necator H16 and Xanthobacter sp. SoF1, using standardized re-annotation, orthology-based comparison, pathway reconstruction, and safety-oriented genome screening. Importantly, SoF1 is the production strain for Solar Foods’ Solein®, a dried microbial biomass ingredient, which is approved as a novel food in Singapore and has a self-affirmed GRAS status in the United States. H16 has a larger, multipartite genome of 7.41 Mb split across two chromosomes and the pHG1 megaplasmid, whereas SoF1 is more compact at 4.91 Mb and encoded on a single replicon. Both encode Calvin–Benson–Bassham CO₂ fixation and multiple [NiFe]-hydrogenase systems supporting growth on CO₂/H₂, but nitrogen economy differentiates the hosts. SoF1 encodes a complete nitrogen-fixation module ( nifHDK ) and nitrate-assimilation genes, whereas H16 lacks nif and instead encodes nitrate/nitrite respiration for oxygen-limited flexibility. Safety screening revealed no evidence of canonical virulence determinants, integron or plasmid-linked antimicrobial resistant (AMR) cassettes, or high-confidence foodborne exotoxins under strict thresholds. These results convert genome-level features into actionable design constraints for selecting and engineering food-grade HOB, strengthening robust air-to-protein bioprocesses on Earth and informing a blueprint for closed-loop, space-compatible protein production.