Diazotrophic growth of free-living Rhizobium etli: Community-like metabolic modeling of growing and non-growing nitrogen-fixing cells

Abstract

Rhizobium etli, a nitrogen-fixing bacterium, grows both in symbiosis (with plants) and in free-living state. While most metabolic models focus on its symbiotic form, this study refined the existing iOR363 model to account for free-living growth. By addition of a biomass formation reaction followed by model curation growth was simulated using various N-sources (NH₃, NO₂, and NO₃). At fixed succinate uptake rate (4.16 mmol/gDWC/h), ammonia yielded the highest growth rate of 0.259 h ⁻ ¹. To represent free-living N-fixing R. etli, a novel two-member community-like model, consisting of both growing and differentiated non-growing N-fixing cells with ammonia exchange, was developed. The XFBA approach, based on community Flux Balance Analysis (cFBA), was formulated to maintain fixed abundances rather than assuming equal growth rates. With a non-growing:growing abundance ratio of 1:9 in community, N-fixation resulted in lower growth rate of 0.1933 h ⁻ ¹ due to the high energy demand of N₂ assimilation compared to ammonia. Sensitivity analysis revealed that increased abundance of N-fixing cells from 5 to 30% led to decreases of 10% in N2-fixation and 25% in growth rate of growing member. Furthermore, Principal Component Analysis identified oxidative phosphorylation, TCA cycle, and glycolysis as key pathways differentiating flux distributions across N-sources. At high uptake of oxygen, causing nitrogenase inactivity, cytochrome bd oxidase was activated to scavenge oxygen, though at the cost of lower growth rate (by 12% per mmol increase in O2 uptake/gDWC/h). This study provided a platform to obtain insights to free-living state of R. etli which may have applications for other diazotrophs.

Fig 1. Comparison of metabolic reaction distribution in R. etli models: iOR363 vs. the curated model developed in this study.

Fig 2. Metabolic map of the R. etli model.

Blue boxes represent alternative carbon and nitrogen sources. The full names of the abbreviated intracellular metabolites are provided in metabolite list in the S1 file of supplementary material.

Fig 3. Schematic representation of free living R. etli: growing cells undergo cell division, with a portion differentiating into non-growing, nitrogen-fixing cells.

Fig 4. Flux distribution in N₂ fixing community of R. etli with ammonia as the sole exchanged metabolite between members.

Numbers shown next to arrows show flux values in mmol/gDWC/h.

Fig 5. Flux distribution of one of optimal solutions in N₂ fixing community of R. etli with allowed exchange of NH₃, Glutamine, Glutamate, Aspartate, and Alanine.

Numbers shown next to arrows show flux values in mmol/gDWC/h.

Fig 6. Effect of aspartate exchange between the growing and non-growing members of the R. etli community on (A) community growth rate (

µ₁ + µ₂) and N₂ fixation, and (B) NH₃ uptake by the growing cell. Negative aspartate fluxes indicate transport from the non-growing to growing member, while positive values indicate the reverse. Also, negative NH₃ fluxes represent its export by the growing member to outer space of community, whereas positive values denote NH₃ uptake by the growing from non-growing member.

Fig 7. Effect of abundance of N-fixing cells on community behavior.

µ denotes growth rate, 1 and 2 refer to growing and N-fixing cells, respectively, and * shows differentiation.

Fig 8. Effect of high oxygen uptake rate on the activation of cytochrome bd for scavenging excess oxygen and its impact on the growth rate, simulated by the R. etli model at fixed succinate uptake of 4.16 mmol/gDWC/h.

Fig 9. Results of PCA comparing flux distributions of R. etli under various N-sources.

A: Explained variance diagram, B: Score plot, and C: Loading plot from PCA results comparing different nitrogen source substrates. Numbers in loading plot indicate reaction numbers as presented in supplementary S1 file.