Bacterial blight, caused by the pathogen Xanthomonas oryzae pv. oryzae, commonly abbreviated Xoo, has re-emerged as a major threat to rice production worldwide. For decades, breeders have known that most cloned resistance genes against this disease, collectively known as Xa genes, trace back to wild relatives of rice or represent recessive alleles of susceptibility genes. That pattern suggested something counterintuitive: during domestication, functional disease resistance may have been actively lost rather than gained. Because the two major rice subspecies, indica and japonica, differ markedly in disease resistance and are reproductively isolated, understanding how resistance genes evolved during subspecies formation became critical for any serious attempt to balance durable resistance with high yield in modern breeding programs.

The Nature study, led by Lin and colleagues, addressed this question through a combination of map-based cloning and genome-wide association studies covering 1,945 rice accessions. From the indica variety Shuangkezao, the researchers identified a novel nucleotide-binding site and leucine-rich repeat protein, or NLR, which they designated XA48. Functional validation using CRISPR-Cas9 knockout and complementation confirmed that Xa48 confers race-specific, lifelong resistance to Northeast Asian strains of Xoo. Unlike many known Xa genes that become effective only in adult plants, Xa48 works from the seedling stage onward, a property that greatly enhances its value for breeding applications where early infection can devastate a crop.

The molecular story behind XA48 is as elegant as it is instructive. Its cognate target is XopG, a conserved and ancient metalloprotease effector deployed by diverse Gram-negative bacteria. When Xoo injects XopG into a rice cell, the XA48 receptor binds the effector directly through its coiled coil domain and oligomerizes into a calcium-permeable resistosome channel at the plasma membrane. Calcium ions flood into the cytoplasm, igniting the immune signaling cascade that culminates in hypersensitive cell death, the hallmark of effector-triggered immunity, or ETI. This mechanism parallels the well-characterized ZAR1 and Sr35 resistosomes, which likewise function as cation channels. However, the authors caution that the high-resolution structure of the XA48-XopG complex remains unsolved, and because functional XopG orthologs are concentrated in Northeast Asian Xoo populations, XA48-mediated resistance may be geographically limited in its usefulness.

Perhaps the most surprising discovery lies downstream of XA48. Using a yeast two-hybrid screen, the team identified two vascular plant one-zinc-finger transcription factors, OsVOZ1 and its homolog OsVOZ2, as direct interactors of the receptor. These transcription factors act as repressors: they suppress jasmonate signaling and negatively regulate both pattern-triggered immunity, the cell-surface basal defense known as PTI, and XA48-mediated ETI. Upon recognizing XopG, the XA48-XopG complex promotes metalloprotease-dependent degradation of OsVOZ1 and OsVOZ2 in the nucleus. With the repressors removed, expression of the OsJAZ genes drops and jasmonate-mediated defense responses are unleashed. Whether XopG directly cleaves the VOZ proteins or recruits an endogenous rice protease or E3 ligase to do so remains an open question.

The role of VOZ transcription factors turns out to be strikingly context-dependent. In rice blast resistance mediated by the NLR receptor Piz-t, both OsVOZ1 and OsVOZ2 contribute positively to immunity, and silencing both compromises Piz-t-mediated resistance. Yet in the bacterial blight system, the same proteins act as negative regulators of basal immunity and XA48-mediated ETI. This contrast suggests that VOZ proteins function as versatile signaling hubs whose output is redirected by different NLR-effector complexes in a pathogen-specific manner, a nuance that will matter for any attempt to manipulate these factors in breeding.

The evolutionary heart of the study comes from population genomic analysis of more than 2,451 rice accessions, spanning wild rice and cultivated varieties. The data revealed a remarkable asymmetry: functional Xa48 alleles exist almost exclusively in indica and have been nearly eliminated from japonica. During domestication and improvement, the frequency of functional Xa48 rose from 7.8 percent to 24.9 percent in indica, a clear signature of positive selection. In temperate japonica, by contrast, the allele collapsed from 1.0 percent to 0.3 percent, evidence of strong negative selection. The fate of a single resistance gene thus diverged dramatically between two subspecies of the same crop species.

The explanation lies at a second locus. The OsVOZ1 gene comes in two major haplotypes, one encoding an alanine, designated OsVOZ1A, and another encoding a serine, designated OsVOZ1S. Indica retains both, whereas japonica carries almost exclusively the OsVOZ1A version. When the researchers introduced Xa48 into japonica varieties carrying OsVOZ1A, the resulting plants suffered severe reductions in seed setting and grain yield across multiple seasons and locations. No such penalty appeared when Xa48 was introduced into indica backgrounds, or into japonica lines engineered to carry the OsVOZ1S allele. In other words, combining XA48 with OsVOZ1A creates a genetic immune incompatibility that cripples reproduction, which neatly explains why japonica rice purged Xa48 during domestication: the fitness cost outweighed the defensive benefit wherever XopG-positive pathogens were rare.

This incompatibility fits a broader phenomenon known as hybrid necrosis or hybrid weakness, in which novel genetic combinations trigger autoimmunity even in the absence of pathogens. Rice offers precedents, including the two-locus Hwc1/Hwc2 interaction that causes F1 hybrid weakness through constitutive accumulation of salicylic acid and jasmonic acid and upregulation of pathogenesis-related genes. The XA48-OsVOZ1A case adds to a growing catalog in which NLR receptors, paired with incompatible genetic backgrounds, undermine reproductive fitness. It also reframes how breeders should think about resistance genes: their value depends not only on pathogen recognition but on compatibility with the host’s developmental and reproductive programs.

Crucially, the study converts this evolutionary insight into a practical breeding strategy. Lin and colleagues engineered japonica lines to co-express Xa48, which mediates intracellular ETI, together with Xa21, the famous pattern-recognition receptor gene that detects the conserved RaxX peptide and confers cell-surface PTI. The resulting Xa21Xa48 plants resisted 30 diverse Xoo strains, including those virulent against either gene alone, and showed enhanced field resistance without yield loss. By reconstituting the XA48-OsVOZ1S module, pairing a functional NLR with a reproductively compatible VOZ allele, the team restored wild-rice-like broad-spectrum resistance in a modern japonica cultivar. The findings reinforce a revised model in which PTI potentiation is indispensable for NLR-mediated immunity during bacterial infection, uniting the two major branches of the plant immune system in a single engineered module. The broader lesson, the commentators argue, is that the future of crop protection lies not in deploying single resistance genes but in deliberately reconstructing entire immune networks, guided by an understanding of how natural and artificial selection have shaped them.

Subject of Research: Asymmetric selection of the rice XA48-OsVOZ1 immune module and synthetic engineering of broad-spectrum bacterial blight resistance

Article Title: An NLR-VOZ seesaw: from asymmetric selection to synthetic broad-spectrum resistance

Article References: Su, Y., Zang, S., Sun, T., You, C., & Que, Y. (2026). An NLR-VOZ seesaw: from asymmetric selection to synthetic broad-spectrum resistance. Stress Biology, 6(1), Article 52. https://doi.org/10.1007/s44154-026-00334-0

Image Credits: AI Generated

DOI: 10.1007/s44154-026-00334-0

Keywords: rice, bacterial blight, NLR immune receptor, XA48, effector-triggered immunity, pattern-triggered immunity, Xa21, OsVOZ1, asymmetric selection, broad-spectrum resistance, growth-defence trade-off, Xanthomonas oryzae