The research, led by Albin Teulet and Sebastian Schornack, focuses on a bacterial delivery apparatus called the Type III Secretion System, a syringe-like molecular machine that many disease-causing bacteria use to inject effector proteins directly into host cells. Rhizobia possess this system too, and they use it to ship a family of proteins known as Nodulation Outer Proteins, or Nops, into the cells of their legume hosts. In most rhizobia, these effectors fine-tune the symbiosis, helping the bacteria evade plant immune defenses or adjusting the balance of the partnership. But in a few remarkable lineages, Nop effectors do something far more dramatic: they can trigger nodule organogenesis on their own, bypassing the classical nodulation signals entirely and forcing the plant to build the nitrogen-fixing home the bacteria need.

What has frustrated researchers until now is that these effector proteins are almost impossible to interpret by looking at their amino acid sequences alone. They have diverged so radically over evolutionary time that standard computational searches, which hunt for telltale sequence signatures of known protein functions, come up essentially empty. The result has been a catalogue of mysterious proteins with no obvious purpose. Teulet and Schornack attacked this problem with a different strategy: instead of comparing sequences, they compared shapes, using the AlphaFold2 artificial intelligence system to predict the three-dimensional structures of rhizobial effectors and then searching structural databases for proteins with matching folds.

This approach, sometimes called structural proteomics, paid off spectacularly. The analysis revealed that rhizobial effectors are not amorphous blobs of uninterpretable sequence but modular machines, assembled like molecular Lego from a repertoire of 22 distinct structural units. Each unit is a recognizable domain with its own likely biochemical activity, and different effectors combine these units in different arrangements. This modular architecture explains both the bewildering sequence diversity of the proteins and their functional versatility: evolution can shuffle, duplicate and fuse these building blocks to generate new effectors with novel combinations of capabilities, all while the underlying structural logic remains conserved and detectable.

The most striking discovery concerns what those building blocks actually do. Many Nop effectors, the study found, harbor domains that bind nucleic acids, the DNA and RNA molecules at the heart of gene expression. Some carry modules that resemble transcription factors, the proteins that switch plant genes on and off. Others contain domains suggestive of post-transcriptional regulation, the ability to intercept and modify RNA messages after they leave the nucleus. Perhaps most surprising of all, some effectors appear to include RNA-dependent RNA polymerase domains, enzymes capable of copying RNA molecules, a function associated in plants with gene silencing and antiviral defense. In other words, these supposedly gentle symbiotic proteins look, structurally, like a cryptic arsenal of genetic regulators.

The evolutionary implications sharpen when the comparison extends beyond rhizobia. The same structural units identified in the symbionts turn up in specific plant pathogens, including gall-inducing bacteria of the genus Pantoea, which form tumor-like growths on their hosts. Crucially, in these pathogens the predicted structural units align with DNA-binding domains that have already been validated experimentally in the laboratory. That correspondence gives the computational predictions real weight: if the same fold in a pathogen is a proven DNA-binding module, the matching fold in a rhizobial effector very likely performs a comparable function. Symbiont and pathogen, it seems, have converged on the same molecular strategy for pulling the levers of host biology.

Among the newly characterized components, one stands out for what it implies about how bacteria might seize control of plant development. The researchers discovered a domain they named BPN, short for B3 and PUA-like nucleic acid binding. Its structure closely mimics the B3 domain, a DNA-binding module found in a large family of plant transcription factors that regulate processes including seed development and hormone responses. A bacterial protein carrying a structural mimic of a plant-specific regulatory domain points to a direct and elegant mechanism of hijacking: the effector could enter the plant nucleus, bind the same DNA sequences that the plant’s own B3 transcription factors target, and either compete with them or redirect their activity, rewiring developmental gene networks from the inside.

This finding reframes the boundary between mutualism and pathogenesis. Plant pathogens manipulate host development to create feeding structures and galls; rhizobia manipulate host development to create nodules. Both, according to the new work, may deploy modular effector proteins that act as direct genetic modulators rather than merely signaling from outside the cell. The difference lies not in the mechanism but in the outcome: one interaction costs the plant, the other, in most contexts, pays dividends in the form of fertilizer-grade nitrogen. Yet the shared toolkit suggests that symbiosis and disease are variations on a single evolutionary theme, with the same domain-fusion strategy recycled across dramatically different lifestyles.

The study also highlights how artificial intelligence is transforming the study of fast-evolving proteins. Sequence-based methods have long been the workhorse of molecular biology, but they fail precisely where evolution has scrambled the text while preserving the structure. Structure prediction changes the calculus: folds persist far longer than sequences, so proteins that look unrelated at the letter level can be recognized as cousins at the shape level. For effector biology in particular, a field dominated by rapidly diversifying virulence and symbiosis factors, AlphaFold2-mediated searches offer a way to annotate proteins that have resisted decades of conventional analysis, converting a catalogue of unknowns into a map of testable functions.

The road ahead is experimental. Structural predictions generate hypotheses, and each of the 22 identified modules now invites targeted biochemical testing: does the predicted DNA-binding domain actually bind DNA, does the RNA polymerase domain actually copy RNA, does the BPN mimic genuinely compete with plant B3 factors in the nucleus? Answering those questions will clarify how rhizobia that bypass classical nodulation signaling actually accomplish the feat, and may reveal general principles of how bacteria, symbiotic or pathogenic, reprogram host gene expression. For agriculture, the stakes are tangible. Understanding the molecular grammar of nodule initiation could eventually help engineers nitrogen-fixing symbioses in crops that currently lack them, a long-sought goal with the potential to reduce fertilizer dependence worldwide. What began as a puzzle about uninterpretable bacterial proteins has opened a window onto the deepest layer of the legume-microbe conversation: the direct manipulation of the plant’s own genetic script.

Subject of Research: Structural characterization of Type III effector proteins in symbiotic rhizobia and their predicted roles as transcriptional and post-transcriptional modulators of host plant gene expression

Article Title: Type III effectors of symbiotic Rhizobia include diverse predicted transcriptional and post-transcriptional modulators

Article References: Teulet, A., & Schornack, S. (2026). Type III effectors of symbiotic Rhizobia include diverse predicted transcriptional and post-transcriptional modulators. PLOS Biology, 24(10), e3004038. https://doi.org/10.1371/journal.pbio.3004038

Image Credits: AI Generated

DOI: 10.1371/journal.pbio.3004038

Keywords: rhizobia, Type III secretion system, Nop effectors, AlphaFold2, structural proteomics, nodule symbiosis, transcription factors, B3 domain, nucleic acid binding, plant pathogens, Pantoea, host gene regulation

The legume-rhizobium symbiosis is one of biology’s most intricate cooperative arrangements. After mutual molecular recognition, soil bacteria called rhizobia attach to root hairs and travel down infection threads into the cells of a developing nodule. There, each bacterium is engulfed in a plant-derived membrane, becoming a symbiosome, and then differentiates into a bacteroid, the specialized form capable of reducing atmospheric nitrogen to ammonia. In temperate legumes such as Medicago, pea and clover, nodules are indeterminate, meaning they retain a persistent apical meristem that continuously generates new cells arranged in a spatial gradient of developmental zones. In the infection zone, rhizobia are released into host cells; in the narrow interzone, bacterial division halts and both partners undergo dramatic differentiation, including multiple rounds of genome multiplication without cell division; and in the nitrogen-fixing zone, elongated bacteroids packed around central vacuoles carry out the actual chemistry of nitrogen reduction under the microaerobic conditions maintained by leghemoglobin.

This elaborate program requires the coordinated activation of hundreds to thousands of genes, and the legume-specific transcripts induced during it were christened nodulins in 1980. Early nodulins are switched on before nitrogen fixation begins, while late nodulins dominate the mature, functioning nodule. NOD25 belongs firmly to the late category. In Medicago truncatula, the gene contains thirteen exons, and its expression is first detectable four days after rhizobial inoculation, rising sharply by seven days and remaining elevated throughout the nodule’s active life. Laser-capture microdissection coupled to RNA sequencing showed that MtNOD25 transcripts are most abundant in the interzone and nitrogen-fixing zone, precisely where bacteroid differentiation and nitrogen reduction occur. Quantitatively, its expression exceeds not only several nodule-specific cysteine-rich peptide genes but even that of leghemoglobin 1, long considered the benchmark of nodule abundance.

Structurally, NOD25-like proteins are unlike almost anything else in the plant proteome. They combine conserved N-terminal and C-terminal regions with a central core built from repetitive modules, each encoded by a separate exon, a genomic architecture that arose through exon shuffling. The N-terminal sequence functions as a signal peptide, demonstrated experimentally by the targeting of an MtNOD25-GFP fusion to the symbiosome and by immunogold staining in alfalfa nodules. Proteomic work had also identified MtNOD25 among the proteins of the symbiosome membrane fraction, the very interface between plant and bacterium. Everything about the gene’s expression pattern, localization and evolutionary conservation pointed toward an essential role in the symbiotic conversation between the two partners.

To find out what that role might be, a team at the HUN-REN Biological Research Centre in Szeged, Hungary, led by Péter Kaló, turned to CRISPR/Cas9 editing. Because no mutant alleles of NOD25 existed in any available insertion or deletion collections of Medicago truncatula, the researchers designed a guide RNA targeting the first exon and delivered the editing construct into roots via Agrobacterium rhizogenes-mediated hairy root transformation. From the edited roots they regenerated whole plants and recovered seeds from four independent mutant lines. Three of these, Mtnod25-1, Mtnod25-3 and Mtnod25-4, carried small insertions or deletions that introduced premature stop codons, truncating the protein before its modular core and C-terminal domain. The fourth, Mtnod25-2, was more dramatic: a 1,304-base-pair deletion removed most of the promoter and the first two exons, abolishing transcript production entirely and classifying the line as a true knockout.

What followed was a systematic search for any trace of a symbiotic defect, and the search came up empty. Four weeks after inoculation with either Sinorhizobium meliloti or Sinorhizobium medicae, mutant plants were indistinguishable from wild type in growth habit, producing dark green leaves with no sign of nitrogen starvation and bearing elongated pink nodules of normal size and morphology. Spinning disc confocal microscopy of SYTO13-stained nodule sections revealed the characteristic zonation of mature indeterminate nodules, with infection, interzone and nitrogen-fixing zones all properly organized and densely colonized. Bacterial positioning and morphology matched wild type at every magnification. Scanning electron microscopy of high-pressure-fixed nodules confirmed that cells in the nitrogen-fixing zone were packed with elongated bacteroids oriented toward the central vacuoles, in sharp contrast to the dnf7-2 control mutant, which lacks the NCR169 peptide and whose nodules contain only undifferentiated rod-shaped bacteria and empty symbiotic cells.

The functional assays were equally unambiguous. Shoot dry weight, a standard proxy for symbiotic effectiveness, was essentially identical between Mtnod25 mutants, the Mtncr068 control line and wild-type plants inoculated with S. medicae, and the mutants even slightly outperformed wild type with S. meliloti. Under non-symbiotic conditions with supplemental nitrogen, all lines grew identically, confirming intact nitrogen assimilation. Finally, the team checked whether the loss of NOD25 might trigger hidden stress responses by measuring the expression of MtPR10, a pathogenesis-related marker, and MtCP2, a senescence-associated cysteine protease. Neither gene was activated in mutant nodules, unlike in the nad1-3 control mutant, in which defense-like and senescence pathways run unchecked. By every cellular, ultrastructural, physiological and molecular measure, symbiotic nitrogen fixation remained fully functional without NOD25.

The evolutionary analysis conducted alongside the mutagenesis deepens the mystery. NOD25-like sequences were found exclusively in the Hologalegina clade of cool-season legumes, and the modular repeat architecture of the central core is restricted to the Inverted Repeat-Lacking Clade, or IRLC, the same group of legumes characterized by terminal bacteroid differentiation and by the rapidly evolving family of secreted NCR peptides that drive that differentiation. Nine distinct repeat motifs were identified across the family, with striking clade specificity: Motifs 1 and 2 occur almost exclusively in the Vicioid clade, Motif 3 is confined to Medicago and Trigonella, and Motif 9 appears only in Lathyrus, Pisum and Vicia. More distantly related species such as Lotus japonicus and Anthyllis vulneraria possess homologous sequences lacking any internal repeats, suggesting that the ancestral NOD25-like protein was a simpler, non-repetitive form and that the repeats arose by exon shuffling after the IRLC diverged. AlphaFold 3 structure predictions, with confidence scores below 0.5 outside the signal peptide, indicate that these proteins likely contain substantial intrinsically disordered regions, a feature often associated with flexible interaction surfaces in host-microbe recognition systems.

How can a gene be so highly expressed, so precisely localized and so evolutionarily conserved, yet dispensable for the process it accompanies? The authors point to precedents: the early nodulin ENOD12 was shown thirty years ago to be non-essential in alfalfa, though in that case redundant proline-rich genes could compensate, whereas no homolog of MtNOD25 exists that might take its place. One tantalizing clue lies in the conserved RK/RLLL motif of the C-terminal region, which resembles the canonical cleavage sites recognized by subtilisin-like serine proteases. In Brassicaceae, subtilase-mediated processing of PROSCOOP precursor proteins generates bioactive signaling peptides that regulate defense and root development, and subtilases have been implicated in legume symbioses and found in soybean peribacteroid membrane fractions. The researchers propose that NOD25 may be produced as an inactive proprotein within the symbiosome, activated by proteolytic cleavage only under specific stresses such as pathogen attack, drought, salinity or temperature extremes, conditions absent from standard laboratory growth. Such a context-dependent function would reconcile the protein’s strict conservation with its invisibility in greenhouse assays. For now, NOD25 joins the growing list of nodule proteins whose true roles await the right environmental trigger, a reminder that in symbiosis, as in much of biology, abundance and necessity are not the same thing.

Subject of Research: Functional analysis of the NOD25 nodulin gene in the nitrogen-fixing root nodule symbiosis of Medicago truncatula

Article Title: Loss of Nod25 function does not affect nitrogen-fixing symbiosis in Medicago truncatula

Article References: Biró, J. B., Domonkos, Á., Güngör, B., Farkas, A., Mohammadi Eghbash, E., Kiss, G. B., & Kaló, P. (2026). Loss of Nod25 function does not affect nitrogen-fixing symbiosis in Medicago truncatula. Plant Molecular Biology, 116(5), Article 101. https://doi.org/10.1007/s11103-026-01766-y

Image Credits: AI Generated

DOI: 10.1007/s11103-026-01766-y

Keywords: Medicago truncatula, nitrogen-fixing symbiosis, nodulin-25, CRISPR/Cas9, rhizobia, symbiosome, IRLC legumes, bacteroid differentiation, nodule development, exon shuffling, subtilases, plant molecular biology

The research focused on Arachis batizocoi, Arachis duranensis, and Arachis magna, wild species that grow naturally in South American drylands and represent part of the genetic reservoir from which cultivated peanut, Arachis hypogaea, ultimately descended. Wild relatives of crops are increasingly recognized as valuable sources of traits that were lost during domestication, from disease resistance to drought tolerance. The new study extends that logic below ground, asking whether the symbiotic bacteria that partner with wild peanuts in their native habitats might outperform the rhizobia that currently colonize agricultural fields. These bacteria, members of the genus Bradyrhizobium, infect legume roots and trigger the formation of nodules, specialized organs in which the microbes convert atmospheric nitrogen into a form the plant can use, a process known as biological nitrogen fixation.

To characterize this hidden diversity, the researchers isolated bacteria from root nodules collected at distinct sites across the Brazilian Drylands, a region characterized by prolonged droughts, nutrient-poor soils, and high temperatures that impose strong selective pressures on both plants and their microbial partners. The team then applied phylogenetic analysis based on the 16S rRNA gene, the standard molecular tool for mapping bacterial evolutionary relationships. Rather than revealing a single homogeneous population, the analysis clustered the wild Arachis bradyrhizobia into five distinct genetic groups, a clear signal that these soils harbor multiple evolutionary lineages of peanut-associated symbionts. Notably, the sequencing did not identify clonal strains, meaning that the isolates were not duplicates of one another or of previously cataloged strains, underscoring how underexplored this microbial resource remains.

This kind of genetic diversity matters for practical reasons as much as for evolutionary ones. Rhizobial symbiosis is highly specific: particular bacterial strains pair effectively with particular host genotypes, and compatibility is governed by an intricate molecular dialogue involving signaling molecules exchanged between root and microbe. Peanut is especially interesting in this regard because it is a crack-entry legume, meaning that its bacteria invade the root through breaches at the base of lateral roots rather than through the classic infection threads seen in legumes such as clover or soybean. That distinctive infection route, highlighted in transcriptomic studies of the peanut symbiosis, suggests that the rules governing which Bradyrhizobium strains succeed in nodulating peanut may differ from those in other legumes, and that wild Arachis species may maintain symbiont communities shaped by millions of years of coevolution with their local soil conditions.

The decisive question, of course, was whether any of this wild diversity translates into benefits for the crop. The researchers inoculated cultivated peanut plants with the wild strains under controlled conditions and compared their performance against uninoculated controls and nitrogen-fertilized plants. The results were striking. Wild strains significantly enhanced root and shoot dry biomass, promoted heavier nodulation, increased nitrogen accumulation in plant tissues, and improved overall plant responsiveness compared with unfertilized, uninoculated plants. In other words, bacteria harvested from the nodules of wild peanut cousins growing in neglected dryland soils proved capable of forging highly productive partnerships with a modern crop variety.

One strain stood out from the pack. ESA 584, isolated from Arachis batizocoi, consistently outperformed the other isolates, delivering the highest values for nodule biomass, total nitrogen content, and biomass allocation across the inoculated plants. That a single wild-derived strain could dominate such a broad set of performance metrics is significant for inoculant development, where consistency across traits is a prized but rare quality. Commercial inoculants must be registered and must meet stringent guarantees of efficacy, and candidate strains typically undergo years of screening before reaching farmers’ fields. ESA 584 now joins a short list of elite Brazilian Bradyrhizobium strains with documented agronomic promise for peanut, several of which have already been shown in earlier studies to enhance peanut drought tolerance and yield in tropical drylands.

Perhaps the most commercially compelling result concerns the comparison with mineral fertilizer. While shoot nitrogen concentration remained highest in the fertilized treatments, the plants inoculated with strains derived from Arachis batizocoi accumulated total nitrogen at levels comparable to the nitrogen-fertilized plants. This distinction between concentration and total accumulation is technically important: a fertilized plant may carry a higher percentage of nitrogen in its leaves, but an inoculated plant can fix enough atmospheric nitrogen to build an equivalent overall nitrogen pool while also channeling more resources into growth and nodule function. For farmers, that translates into the possibility of obtaining much of the crop’s nitrogen budget from the air rather than from purchased fertilizer, cutting input costs and reducing the environmental footprint associated with nitrogen production and runoff.

The Brazilian Drylands are an ideal hunting ground for such microbes precisely because their environmental stresses act as natural filters. Bradyrhizobia surviving in soils that cycle between severe drought and intense rainfall, and that are often acidic or nutrient-poor, must tolerate conditions that would defeat more delicate strains. Previous surveys of rhizobia across Brazilian ecosystems, from the semi-arid peanut regions of the northeast to the Amazon savanna and the dry forests of Caatinga, have repeatedly revealed that wild and native legumes harbor unexpectedly diverse and effective bacterial communities. The new study reinforces the pattern and adds a specific, economically important target: the wild gene pool of peanut itself. Because wild Arachis species already contribute resistance traits to modern peanut breeding, including resistance to root-knot nematode and stem rot, pairing below-ground microbial resources with above-ground genetic improvements could compound the benefits.

The implications extend beyond peanut. Biological nitrogen fixation is one of agriculture’s most valuable ecosystem services, and the search for elite rhizobial strains is a global enterprise, with researchers cataloging highly adapted Bradyrhizobium species nodulating diverse legumes across Africa and the Americas. The Brazilian team’s work demonstrates a template that can be replicated for other orphaned or underappreciated crop wild relatives: sample the nodules of wild populations across their native range, resolve their phylogenetic diversity, and test their symbiotic performance against the cultivated form. In doing so, scientists can convert biodiversity surveys into concrete tools for sustainable intensification, particularly in regions where fertilizer access is limited and soils are marginal.

For now, the wild bradyrhizobia of the Brazilian Drylands remain primarily a research resource, with their DNA sequences deposited in GenBank for the wider community to examine. But the trajectory is clear. ESA 584 and its relatives offer a glimpse of a peanut crop that feeds itself, drawing nitrogen from the atmosphere through partnerships forged in the wild and refined by evolution under some of the toughest conditions the tropics can offer. As the researchers conclude, wild peanut-associated Bradyrhizobium strains are both diverse and efficient in association with cultivated peanut, making them a promising source of new inoculants. In the race to feed a growing population while shrinking agriculture’s environmental bill, the humble bacteria of wild root nodules may prove to be among the most valuable allies yet discovered.

Subject of Research: Diversity and symbiotic efficiency of Bradyrhizobium strains from wild Arachis species in the Brazilian Drylands and their potential as peanut inoculants

Article Title: Wild Arachis bradyrhizobia from distinct sites in the Brazilian Drylands: biodiversity and symbiosis with peanut (Arachis hypogaea, Fabaceae) crop

Article References: Wild Arachis bradyrhizobia from distinct sites in the Brazilian Drylands: biodiversity and symbiosis with peanut (Arachis hypogaea, Fabaceae) crop. (n.d.). https://doi.org/10.1007/s44473-026-00252-z

Image Credits: AI Generated

DOI: 10.1007/s44473-026-00252-z

Keywords: Bradyrhizobium, Arachis hypogaea, wild peanut relatives, biological nitrogen fixation, rhizobia, Brazilian Drylands, root nodules, peanut inoculants, 16S rRNA phylogeny, sustainable agriculture, Arachis batizocoi, symbiosis