Genome-Scale Metabolic Modelling of Lifestyle Changes in Rhizobium leguminosarum

Carolin C. M. Schulte, Vinoy K. Ramachandran, Antonis Papachristodoulou, Philip S. PoolemSystems 2022 · DOI

Abstract

Biological nitrogen fixation in rhizobium-legume symbioses is a cornerstone of sustainable agriculture. To establish a mutualistic relationship, rhizobia must transition from free-living soil bacteria to growth within infection threads, finally differentiating into nitrogen-fixing bacteroids. This study presents iCS1224, a high-quality genome-scale metabolic model for Rhizobium leguminosarum bv. viciae 3841. By integrating transcriptomic, proteomic, and metabolomic data, the researchers mapped the metabolic flux across these distinct life stages. The model identifies key metabolic requirements—such as histidine synthesis in bacteroids and myo-inositol catabolism during infection—providing a framework for interpreting complex experimental datasets and identifying targets for future agricultural engineering.

Introduction

The symbiotic relationship between rhizobia and legumes is one of nature's most efficient systems for converting atmospheric nitrogen into plant-available ammonia. For Rhizobium leguminosarum, this process involves a dramatic lifestyle shift. The bacteria must survive in the competitive soil of the rhizosphere, navigate infection threads into plant roots, and finally differentiate into specialized, non-dividing bacteroids.

While genomic sequencing has identified the "parts list" for these bacteria, understanding how these parts work together to drive metabolism has remained a challenge. Previous studies provided snapshots of gene expression but lacked a systemic view of metabolic flux—the actual flow of nutrients and energy. To address this, researchers developed iCS1224, a genome-scale metabolic model (GSM) designed to simulate the bacterium's internal chemistry across its entire life cycle.

Developing the iCS1224 "Digital Twin"

The construction of iCS1224 involved a rigorous multi-step process. The authors combined automated drafts from databases like KEGG and MetaCyc with homology-based templates and extensive manual curation based on decades of biochemical literature. The resulting model encompasses 1,224 genes, 1,257 reactions, and 984 metabolites.

To ensure the model's reliability, it was validated against several high-throughput datasets:

  • Carbon Source Utilization: The model's predictions were compared against Biolog phenotype microarrays covering 190 different carbon sources, achieving 89.9% accuracy.
  • Gene Essentiality: Using INSeq (insertion sequencing) data, the model predicted which genes are vital for survival with 91.0% accuracy.
  • Flux Analysis: Quantitative metabolic fluxes predicted by the model showed excellent agreement with experimental $^{13}$C labeling data, proving the model is quantitatively precise.

Mapping the Metabolic Journey

By applying the RIPTiDe algorithm to integrate RNA-Seq and proteome data, the researchers "pruned" the general iCS1224 model into three context-specific versions representing the rhizosphere, undifferentiated nodule bacteria, and mature bacteroids.

1. The Rhizosphere: Survival and Preparation

In the soil surrounding plant roots, the model predicted that the synthesis of leucine, polyphosphate, and AICAR (a purine precursor) is essential. Interestingly, the model suggests these pathways are active even when some of these nutrients might be available in the environment, highlighting their role as metabolic "must-haves" for establishing a foothold near the host plant.

2. The Nodule: The Invasion Phase

As bacteria enter the plant and begin to divide within the nodule, their metabolic profile shifts. The model identified myo-inositol catabolism as a primary energy driver during this undifferentiated stage. This suggests that the bacteria use this sugar-like compound as a primary energy source specifically while they are invading and differentiating within the plant tissues.

3. The Bacteroid: Specialized Nitrogen Fixation

Once the bacteria differentiate into mature bacteroids, their primary role is nitrogen fixation. While it is well-established that bacteroids consume dicarboxylates (like malate) provided by the plant, iCS1224 revealed a more diverse diet, including the utilization of xylose and glycolate. This helps explain previously described gene expression patterns that were not fully understood.

A particularly surprising finding involved amino acid metabolism. While most amino acid synthesis pathways are downregulated in bacteroids, histidine synthesis remains highly active. The model, supported by proteomic evidence, suggests this is because the specialized proteins required for nitrogen fixation (Nif and Fix proteins) are unusually rich in histidine, necessitating a dedicated internal supply.

Implications for Sustainable Agriculture

The iCS1224 model represents a significant leap forward in our ability to predict and manipulate rhizobial metabolism. By identifying the specific metabolic "bottlenecks" and requirements of each life stage, researchers can now look toward designing "designer" rhizobial strains. These strains could be engineered to be more competitive in the soil or more efficient at fixing nitrogen, potentially reducing the global reliance on synthetic chemical fertilizers.

Limitations and Future Directions

While iCS1224 is a powerful tool, the authors note that it relies on the assumption that transcript levels (RNA) correlate perfectly with enzyme activity, which is not always the case due to post-translational regulation. Additionally, when multiple nutrients share the same catabolic pathway, it can be difficult for the model to distinguish exactly which one is being utilized. Future work will focus on using this model as a template for other economically important rhizobia and refining flux predictions through more complex regulatory data.

Summary updated

Abstract

Biological nitrogen fixation in rhizobium-legume symbioses is of major importance for sustainable agricultural practices. To establish a mutualistic relationship with their plant host, rhizobia transition from free-living bacteria in soil to growth down infection threads inside plant roots and finally differentiate into nitrogen-fixing bacteroids. We reconstructed a genome-scale metabolic model for Rhizobium leguminosarum and integrated the model with transcriptome, proteome, metabolome, and gene essentiality data to investigate nutrient uptake and metabolic fluxes characteristic of these different lifestyles. Synthesis of leucine, polyphosphate, and AICAR is predicted to be important in the rhizosphere, while <i>myo-</i>inositol catabolism is active in undifferentiated nodule bacteria in agreement with experimental evidence. The model indicates that bacteroids utilize xylose and glycolate in addition to dicarboxylates, which could explain previously described gene expression patterns. Histidine is predicted to be actively synthesized in bacteroids, consistent with transcriptome and proteome data for several rhizobial species. These results provide the basis for targeted experimental investigation of metabolic processes specific to the different stages of the rhizobium-legume symbioses. <b>IMPORTANCE</b> Rhizobia are soil bacteria that induce nodule formation on plant roots and differentiate into nitrogen-fixing bacteroids. A detailed understanding of this complex symbiosis is essential for advancing ongoing efforts to engineer novel symbioses with cereal crops for sustainable agriculture. Here, we reconstruct and validate a genome-scale metabolic model for Rhizobium leguminosarum bv. <i>viciae</i> 3841. By integrating the model with various experimental data sets specific to different stages of symbiosis formation, we elucidate the metabolic characteristics of rhizosphere bacteria, undifferentiated bacteria inside root nodules, and nitrogen-fixing bacteroids. Our model predicts metabolic flux patterns for these three distinct lifestyles, thus providing a framework for the interpretation of genome-scale experimental data sets and identifying targets for future experimental studies.
BibTeX
@inproceedings{schulte2022genome,
  author = {Carolin C. M. Schulte and Vinoy K. Ramachandran and Antonis Papachristodoulou and Philip S. Poole},
  title = {{Genome-Scale Metabolic Modelling of Lifestyle Changes in Rhizobium leguminosarum}},
  booktitle = {mSystems},
  year = {2022},
  doi = {10.1128/msystems.00975-21},
}
Figure