AI-Designed Immune Receptors Help Chinese Scientists Speed Up Crop Disease Resistance

Scientists from the Chinese Academy of Sciences' Institute of Genetics and Developmental Biology and the biotechnology company Qi Biodesign have developed an AI-guided platform to engineer custom synthetic immune receptors for plants, aiming to speed up the response to emerging crop diseases. The work was published in the peer-reviewed journal Science last month. The programmable receptors can be introduced into plants to help them detect proteins made by specific bacteria, viruses and fungi, triggering an immune response that neutralises the threat, the team said.

Plant breeding programmes typically introduce resistance genes that encode receptors capable of recognising proteins secreted by pathogens; when a receptor binds to its target, it can trigger an immune response. 'However, this resistance is often rapidly overcome by fast-evolving pathogens, particularly in modern agricultural systems characterised by large-scale monoculture of genetically uniform crops,' the team said. Identifying new receptors to counter emerging pathogens can normally take years because the natural diversity of plant immune receptors is limited.

To build faster, custom receptors, the researchers used artificial intelligence to design proteins that bind to pathogen proteins, then spliced these AI-designed binders into a natural rice immune receptor. 'These synthetic receptors successfully conferred recognition and immune activation against proteins from diverse plant pathogens, including 14 viruses, two bacteria, a fungus and an oomycete,' the team said, referring to oomycetes as fungus-like microorganisms that can cause destructive plant diseases. Some of the pathogens the receptors worked against target vital crops such as potatoes and tomatoes.

The team tested 391 synthetic receptors and found that about 18% triggered robust and specific immune responses, about 58% displayed varying levels of autoimmunity, and nearly 24% showed no immune activity against their targets. According to the paper, a lack of activity could result from improper protein expression in the plants, failure of the designed binders to engage their target pathogen proteins, or an inability to trigger the physical shape change needed to activate immunity.

Because crops are often threatened by multiple pathogens at once, the team also developed a strategy to stack several synthetic receptors together, which it said 'offers a scalable and multiplexed framework for engineering broad-spectrum and durable disease resistance in crops.' The team acknowledged that designing synthetic receptors for certain pathogens remains challenging and that the overall design success rate is still low, but said identifying new natural receptor scaffolds and expanding AI-guided binder design could improve the system.