Rewriting the code of plant immunity

New approaches to engineering plant immunity can more broadly protect crops against pathogens and disease, once they get through regulatory hurdles.

Plants are constantly fending off attacks from bacterial and fungal pathogens — and for crops the onslaught is ramping up as environmental changes accelerate the emergence of major diseases. Bacteria, fungi and other infections can lead to crop death or decreased yields. Pesticides can provide protection against certain bacteria or fungi but pose environmental risks and may be toxic to humans. With emerging protein engineering technology and a greater understanding of the intricacies of the plant immune system, biotech researchers are finding an alternative: generating plants and crops that can defend themselves using intrinsic immune responses.

“With emerging protein engineering technology and a greater understanding of the intricacies of the plant immune system, biotech researchers are finding an alternative: generating plants and crops that can defend themselves using intrinsic immune responses.”

Plants combine cell-surface and intracellular immune receptors to mount an immune response against pathogens. The first line of plant defense is pattern recognition receptors (PRRs), which recognize and bind molecules from the pathogen at the plasma membrane, triggering downstream immune responses. Intracellular pathogen recognition, the plant’s second line of defense, is largely mediated by a diverse family of nucleotide-binding and leucine-rich-repeat receptors (NLRs). NLRs detect proteins (effectors) secreted by pathogens inside the cell; binding of an NLR to an effector will initiate a signaling cascade leading to cell death. Some NLRs function as individual genetic units, recognizing specific cognate pathogen effectors, while others work as sensors in signaling networks, sometimes also including PRRs. Making these receptors stronger or reactive with a wider range of pathogen effectors would lead to stronger plant defense systems, similarly to immune cell engineering in humans.

Engineering plant NLRs is not a new idea — over a decade ago, receptor recognition was reprogrammed through both targeted and random mutagenesis. For example, NLRs could recognize additional pathogenic proteins or launch a stronger immune response following infection by mutating amino acids on the receptor’s binding site¹. By swapping entire NLR binding domains, scientists expanded plants’ recognition or signaling behavior², as was the case with transgenics created to express resistance genes from other species³. But, outside of closely related plant species, transferring NLR genes is not entirely effective because NLR structures and functions are highly variable across different species as a result of their rapid evolution⁴.

Now, researchers and companies have abandoned using mutagenesis in favor of AI prediction tools and exchanged transgenics for precise gene editing technologies. For example, Resurrect Bio raised $2 million in seed funding in 2023 to generate its own AI tool that will predict plant–pathogen interactions and is aiming to ‘resurrect’ or engineer NLRs that have become dormant or lost⁵.

More recently the focus has shifted from precisely engineering plant defense systems to strategies that confer a more moderate but broader defense through targeting the cell-surface PRRs⁶. As recently published in Nature Biotechnology, the C terminus of the PRR RLP23 has been engineered to enhance the resistance of tomato, rice and poplar to fungal, bacterial and oomycete pathogens. Engineering PRRs avoids some threat of evolving resistance, as these receptors tend to recognize more conserved regions across a broader range of pathogens⁷.

Protein prediction tools can be used to engineer specific binders to known effector proteins. Pikobodies are bioengineered intracellular immune receptors generated from the Pik-1–Pik-2 NLR pair, in which the recognition domain is replaced with a nanobody specifically raised against a target molecule. Thus, the receptor is reprogrammed to trigger an immune response against any pathogen effector that the nanobody can bind inside the cell⁸. Nelarix plans to apply this technology in crops to protect them against major pathogens such as fungi, oomycetes, obligate intracellular bacteria and viruses. Another company, Ohmic Biosciences, is looking to engineer proteins that directly inhibit effector proteins using a combination of protein engineering and synthetic biology. With receptor engineering, a major challenge is evolution: effectors are quick to evolve or may be dropped entirely by the pathogen. Using a range of inhibitors or pikobodies in combination (called “gene stacking”) may help.

To enhance the immune response, it is crucial to know which microbes are recognized by which receptors, and knowledge about the structure and function of NLRs and PRRs has primarily been from a small number of crops and model plant species. A recent Science paper laid out a strategy integrating bioinformatics, biochemistry and synthetic biology to map the recognition landscape of a subgroup of plant immune receptors — over 13,000 receptors from 285 angiosperm genomes⁹. The authors were able to discover an unknown immune receptor that has variants in many plants used for food today and engineer new versions that recognize different pathogens. Larger machine learning models trained on more data will be able to help identify and optimize regions of these receptors for modifications.

The arms race between pathogenic proteins and immune receptors will always exist, but computational tools combined with larger protein datasets may help predict how pathogens will evolve, allowing companies to adapt their technologies quickly in response. Other technologies may emerge, including ones that target translation instead of protein binding.

As with any new crop technology, regulation will present hurdles. It is easier to work with native plant proteins than to introduce novel ones, although both would have to be tested and shown to have minimal impact on healthy plant microbiomes and not to compromise yield. Any new protein introduced would also need to be ruled out as a potential allergen to humans. Field trials will take time, but the technology is only going to improve while we wait for these new plants to bear fruit (or seed).