Hidden Nitrogen-Fixing Bacteria Reveal a Secret Evolution Inside Casuarina Root Nodules

For decades, microbiologists believed they knew exactly which bacterium lives inside the nitrogen-fixing root nodules of Casuarina trees. Every strain ever cultured from these nodules, whether the tree grew in Australia, Asia, Africa or the Americas, belonged to a single species, Frankia casuarinae, with genomes sharing at least 98 percent average nucleotide identity. That uniformity was puzzling, because early DNA studies of nodules collected in Australia had hinted at at least seven distinct genotypes. A new international study, published in Plant and Soil, has now resolved the paradox in a striking way: the true diversity of Casuarina’s microbial partners was hiding in plain sight, invisible to standard culturing because many of the dominant symbionts have lost the very metabolic machinery needed to grow on laboratory plates.

The research team, led by Katharina Pawlowski of Stockholm University with collaborators in Australia, Germany, France, Singapore and elsewhere, bypassed cultivation entirely. They collected nodules from Casuarina equisetifolia in Singapore and French Polynesia, Casuarina oligodon in the highlands of Papua New Guinea, and Casuarina glauca and Casuarina cunninghamiana in New South Wales, Australia, all within or near the natural range of the genus. Instead of trying to coax bacteria out of the nodules, the researchers sequenced the total DNA of the nodule metagenomes directly and reconstructed metagenome-assembled genomes, or MAGs, from twenty nodule samples. This approach, previously refined for other Frankia groups, allows scientists to read the genomes of organisms that have never been grown in isolation.

The results overturned the long-standing assumption of a single microsymbiont species. Phylogenomic analysis based on 138 conserved single-copy marker genes, combined with average nucleotide identity calculations and digital DNA-DNA hybridization, revealed that the nodules hosted two clearly separated lineages. One lineage corresponded to the familiar Frankia casuarinae, represented by genotypes from Singapore, Papua New Guinea and one French Polynesian sample. The other lineage, found in all Australian samples and three of the four French Polynesian samples, fell below the accepted species thresholds of roughly 95 to 96 percent ANI and 70 percent dDDH, and was formally proposed as a new species, Candidatus Frankia pacifica, a name reflecting its distribution across islands and coastal regions of the Pacific.

Even more remarkable than the discovery of a hidden species was what the MAGs revealed about genome evolution. Sixteen of the twenty reconstructed genomes were markedly smaller than those of cultivated strains, measuring only 3.7 to 4.3 megabase pairs compared with the 5.0 to 5.6 megabase genomes of cultured F. casuarinae. One genome had lost more than a quarter of its genetic material relative to the type strain. A statistical comparison confirmed that this size difference was highly significant, and the pattern is a textbook signature of genome erosion, the progressive degradation of DNA that occurs when a microbe becomes increasingly dependent on its host and no longer needs genes for independent life.

The erosion followed a precise molecular pattern centered on hydrogenases. Frankia’s nitrogenase enzyme releases hydrogen gas as an unavoidable byproduct of nitrogen fixation, and [NiFe] hydrogenases recycle that hydrogen to recover energy. Symbiotic Frankia strains carry up to three types of these enzymes. The type 1h hydrogenase, encoded by a gene cluster called synton-1, functions during saprotrophic growth, scavenging atmospheric hydrogen to fuel respiration when carbon is scarce. The type 2a and type 1f hydrogenases, by contrast, are expressed at higher levels inside the plant and are thought to be essential for symbiotic nitrogen fixation. In the new MAGs, it was specifically the type 1h hydrogenase that had been damaged or lost, often through stop codons, truncations, or operons shattered into fragments scattered across the genome by transposase activity.

The details were striking. In every MAG from beach-grown C. glauca in New South Wales, the hupD1 maturation protease gene was missing and the large hydrogenase subunit gene hupL1 contained at least one premature stop codon. The inland C. cunninghamiana genotypes had lost functional copies of hypB1, hupL1 and hypF1. The three Papua New Guinea genotypes had deleted the entire synton-1 region. Even the two Singapore genotypes, whose overall genome sizes looked normal, had lost the structural genes of the type 1h enzyme, showing that erosion can begin before a genome visibly shrinks. A second hydrogenase, the cytosolic type 3b enzyme encoded by synton-4, was also eroded in most Australian, French Polynesian and Papua New Guinean MAGs, though it remained intact in the Singapore samples, hinting at a possible link to salt stress at coastal sites.

Comparative genomics added further evidence that the two lineages are genuinely distinct organisms rather than mere variants. The two least-eroded MAGs of Candidatus F. pacifica, from Moorea and Rurutu, contain a gene encoding a bifunctional enzyme that fuses isocitrate lyase and malate synthase, the two catalytic steps of the glyoxylate shunt, a bypass of the tricarboxylic acid cycle. Bifunctional versions of this enzyme had previously been known only from the single-celled alga Euglena gracilis and the roundworm Caenorhabditis elegans. No cultivated F. casuarinae strain possesses the glyoxylate shunt at all, and the fact that the shunt is absent from all eroded genomes of both lineages suggests it was lost early in the transition toward symbiotic dependence. Conversely, genes for a secreted polyhydroxybutyrate depolymerase, possibly involved in degrading the suberin of nodule cell walls, and an azoreductase gene for detoxifying electrophilic quinones were found only in F. casuarinae.

The biogeographic picture that emerges is one of ancient divergence and parallel evolution. Because the four French Polynesian MAGs are the most dissimilar from one another, and because one of them still belongs to F. casuarinae, the authors conclude that the two species likely separated in French Polynesia, with Candidatus F. pacifica subsequently spreading to Australia and diversifying into coastal and inland subclades that may reflect coevolution with C. glauca and C. cunninghamiana respectively. One French Polynesian genome, CeTa4Puna’auia, occupies an ambiguous position between the two species, plausibly close to their point of separation. The pattern mirrors what has been documented in alder-infective Frankia, where uncultivable, spore-forming genotypes with eroded genomes outcompete saprotrophic strains in the field, suggesting that evolution toward obligate symbiosis has arisen independently in multiple Frankia lineages.

The practical implications could be considerable. Casuarina trees are planted worldwide as shelter belts, fuelwood sources and phytoremediation agents because they tolerate salinity, drought, flooding and heavy metals while fixing nitrogen on marginal soils. The findings suggest that inocula based on crushed local nodules, which contain the eroded, highly competitive symbionts, may outperform inocula of cultivable F. casuarinae strains for reforestation and coastal protection projects. The study also issues a broader warning to microbiology: by focusing on what grows in a petri dish, researchers had painted a misleading picture of Frankia diversity and its adaptation to specific hosts and environments. Whether similar hidden eroded symbionts lurk in the nodules of cluster-3 Frankia hosts, which carry the largest genomes in the genus, remains an open question that direct metagenomic sequencing is now poised to answer.

Subject of Research: Hidden Frankia biodiversity and genome erosion in Casuarina root nodule symbioses

Article Title: In the nodule or on the nodule? The hidden Frankia biodiversity in Casuarina nodules across continents

Article References: Obaid, N. B., Patyi, A., Berckx, F., Klages, K., Bernal-Gómez, M., Lavello, A., Jentzen, J., Brachmann, A., Wibberg, D., Blom, J., Kalinowski, J., Mehrabi, S., Kennedy, I. R., Normand, P., Mathesius, U., & Pawlowski, K. (2026). In the nodule or on the nodule? The hidden Frankia biodiversity in Casuarina nodules across continents. Plant and Soil. https://doi.org/10.1007/s11104-026-09140-z

Image Credits: AI Generated

DOI: 10.1007/s11104-026-09140-z

Keywords: Frankia, Casuarina, actinorhizal symbiosis, nitrogen fixation, genome erosion, metagenome-assembled genomes, hydrogenase, Candidatus Frankia pacifica, root nodules, symbiotic evolution, biogeography, Plant and Soil