Chinese scientists have developed an artificial intelligence-driven platform that can design synthetic immune receptors for plants, creating a new approach to protecting crops against diseases. The technology is designed to recognize proteins produced by specific viruses, bacteria, fungi and oomycetes and trigger an immune response inside the plant. The research was published in the peer-reviewed journal Science in July 2026.
The breakthrough could change how disease-resistant crops are developed. Traditional breeding often depends on finding useful resistance genes in nature and incorporating them into crop varieties, a process that can take years. The new platform combines AI-guided protein design with directed evolution to build programmable plant immune receptors, potentially reducing the design stage to weeks and allowing defenses to be tailored to emerging pathogens.
What Is The New Plant Immunity System?
The research team from the Chinese Academy of Sciences’ Institute of Genetics and Developmental Biology and biotechnology company Qi Biodesign developed what the researchers call programmable synthetic plant immune receptors, or SPIRs.
Plants already have sophisticated immune systems. They use receptors to recognize molecules or proteins associated with invading pathogens. Once a threat is detected, the plant can activate defensive pathways to limit infection.
The problem is that natural resistance is not always broad or durable. Pathogens evolve, while modern agriculture often relies on genetically similar crops planted across large areas. A pathogen capable of overcoming a particular resistance mechanism can therefore spread rapidly.
The new approach attempts to make the recognition component programmable.
How The Technology Works
The researchers designed binding modules that can recognize selected pathogen proteins and inserted them into an existing plant immune-receptor framework. When the engineered receptor recognizes its target, it can activate the plant’s immune response.
The system was tested against proteins associated with multiple classes of plant pathogens, demonstrating that the same general design strategy could be adapted to different biological threats.
| Feature | New SPIR Platform |
|---|---|
| Technology | Programmable synthetic plant immune receptors |
| Design assistance | AI-guided protein design |
| Pathogens tested | Viruses, bacteria, fungi and oomycetes |
| Recognition | Specific pathogen proteins |
| Immune activation | Engineered plant receptors trigger defense |
| Optimization | AI design + in-planta directed evolution |
| Recognition | Can be highly target-specific |
| Multiplexing | Receptors can be stacked |
| Demonstrated outcome | Disease resistance in transgenic plants |
| Research publication | Science, July 2026 |
AI Could Shrink The Development Timeline
One of the most important aspects of the research is the potential speed advantage.
Conventional resistance breeding generally requires researchers to identify naturally occurring resistance genes, introduce them into suitable crop backgrounds and evaluate whether the resulting plants maintain the desired characteristics. The process can take years and may become difficult when pathogens evolve faster than breeding programs can respond.
AI-guided protein design changes the starting point. Instead of relying entirely on naturally occurring receptor combinations, researchers can computationally design binding modules aimed at specific pathogen proteins.
The research team said custom receptors could potentially be engineered within weeks rather than years, although that should not be interpreted as meaning a new commercial crop can be developed and released within weeks. Field testing, breeding, safety assessment and regulatory processes would still take substantially longer.
From Natural Resistance To Designed Resistance
Traditional approach
Natural resistance gene
↓
Breeding
↓
Crossing and selection
↓
Testing
↓
Disease-resistant variety
Programmable approach
Identify pathogen protein
↓
AI-guided receptor design
↓
Synthetic immune receptor
↓
In-plant optimization
↓
Disease-resistance testing
The key change is that researchers gain greater control over the recognition component of plant immunity.
Researchers Tested Hundreds Of Synthetic Receptors
The scale of the experiments provides another indication of the platform’s potential.
According to reporting on the research, scientists created and tested 391 synthetic receptor variants designed to recognize proteins associated with 14 viruses, two bacteria, one fungus and one oomycete.
The peer-reviewed study reports that the synthetic receptors recognized pathogen proteins from viral, bacterial, fungal and oomycete sources. The researchers also improved receptor performance by combining AI-guided protein design with directed evolution carried out inside plants.
| Research Measure | Result |
|---|---|
| Synthetic receptor variants tested | 391 |
| Viruses targeted | 14 |
| Bacteria targeted | 2 |
| Fungi targeted | 1 |
| Oomycetes targeted | 1 |
| Major design methods | AI + directed evolution |
| Receptors | Programmable synthetic receptors |
| Disease resistance | Demonstrated in transgenic plants |
The numbers do not mean that 391 commercial crop varieties have been created. Rather, they describe the experimental receptor designs used to test whether the platform could recognize a broad range of pathogen-derived targets.
The System Can Recognize Live Viral Infections
The significance of the research goes beyond laboratory protein recognition.
The Science paper reports that virus-targeting SPIRs responded to infectious clones of plant viruses and that transgenic plants expressing the synthetic receptors showed disease resistance.
That is an important distinction because recognizing an isolated pathogen protein is not necessarily equivalent to protecting a living plant from infection. Demonstrating activity against infectious viral material provides stronger evidence that the engineered receptors can function as part of a plant’s defense system.
The researchers also found that synthetic receptors could be stacked for multiplexed recognition. This could eventually allow plants to carry multiple recognition mechanisms targeting different pathogen proteins rather than relying on a single resistance trait.
Why Crop Diseases Are A Major Agricultural Risk
Plant pathogens are a persistent threat to global food production. Crops can lose productivity when infections damage leaves, roots, fruits or seeds, while severe outbreaks can make entire production areas economically unviable.
A major challenge is pathogen evolution. A resistance gene that works effectively against one pathogen population may become less useful as the pathogen changes.
The problem can be amplified by monoculture, where large areas are planted with genetically similar varieties. Researchers behind the new platform specifically highlighted the vulnerability created when rapidly evolving pathogens encounter genetically uniform agricultural systems.
This creates a race between pathogen evolution and crop protection technology.
| Challenge | Traditional Limitation | Potential SPIR Advantage |
|---|---|---|
| New pathogen | Resistance discovery can be slow | Target-specific receptor design |
| Pathogen evolution | Existing resistance may be overcome | New receptors could be designed |
| Multiple pathogens | Multiple resistance genes may be needed | Receptors can potentially be stacked |
| Limited natural resistance | Depends on available genetic diversity | Synthetic designs expand options |
| Breeding time | Often measured in years | Design stage could take weeks |
| Disease outbreaks | Response can lag pathogen spread | Potentially faster design cycle |
Potential Impact On Agriculture
If the technology can be translated successfully from controlled experiments into reliable field performance, it could give plant breeders a new tool for developing disease-resistant varieties.
The most important potential application could be rapid response to newly emerging pathogens. Instead of waiting for a naturally resistant plant variety to be discovered, scientists could potentially identify a pathogen protein and design a receptor aimed at recognizing it.
That could be particularly valuable for crops with limited genetic diversity or diseases for which conventional resistance breeding has struggled.
The technology could also reduce reliance on some chemical disease-control approaches if engineered resistance proves sufficiently effective. However, it is too early to conclude that synthetic immunity will replace pesticides or conventional breeding.
The Technology Still Has Major Hurdles
Despite the promising results, the platform is still a research technology rather than an immediately deployable agricultural product.
A receptor that works in laboratory or greenhouse experiments must perform reliably across different varieties, environments and pathogen populations before it can become commercially useful. Researchers must also establish how stable the engineered traits remain across generations and whether they create unintended effects on plant growth or development.
Regulatory requirements will also differ across countries depending on how the technology is implemented. The pathway could involve genetic engineering, genome editing or other breeding methods, with the regulatory classification depending on the final crop and modification.
Most importantly, the researchers have demonstrated a platform—not a finished universal immunity system for every crop disease.
The Bigger Picture
The significance of the breakthrough lies in the shift from discovering plant immunity to designing it. AI-guided protein engineering could give agricultural scientists a way to build new pathogen-recognition mechanisms when natural resistance is unavailable or insufficient. The Science study provides experimental evidence that these synthetic receptors can recognize targets from multiple pathogen classes and contribute to disease resistance in engineered plants.
For agriculture, this could eventually create a more responsive model of crop protection. Disease resistance could become increasingly programmable, with researchers designing and combining immune receptors according to the pathogens threatening particular crops. The technology is not yet ready to provide instant protection against agricultural epidemics, but it represents a significant step toward that possibility.
Looking Ahead
The next challenge will be translating the laboratory results into stable, effective crop varieties that can withstand real-world agricultural conditions. Researchers will need to test the approach across more crops, pathogen strains and growing environments while determining how effectively the synthetic receptors perform over multiple generations. Regulatory and commercial development will also determine how quickly the technology can move beyond research settings.
If those hurdles can be overcome, programmable plant immunity could become an important part of future crop protection. The combination of AI protein design, synthetic biology and plant breeding could allow scientists to respond to emerging diseases more quickly than traditional approaches, potentially strengthening food security as pathogens continue to evolve.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
