Art Courtesy of Kara Tao
Plant-animal hybrids are here, and they are exactly what they sound like. In the Sainsbury Laboratory in Norwich, UK, wild tobacco plants have been engineered to produce “pikobodies,” synthetic proteins that can recognize and attack pathogens expressing fluorescent proteins. These pikobodies are made by taking rice-derived receptors and swapping in antibody fragments originating from camelid mammals, specifically llamas and alpacas. In a proof-of-concept study—carried out by postdoctoral scientist Jiorgos Kourelis of the group headed by Sophien Kamoun—the team discovered that these pikobody-producing tobacco plants could successfully stave off viral invaders.
This discovery arrives at a crucial moment: in the past year alone, wars, climate change, and trade routes in flux have ferried pathogens around the globe in dangerously unprecedented ways. Meanwhile, as a wheat blast devastates crop yields in Africa and Asia, and scientists sound alarm bells to food security worldwide, protection against plant diseases is more important than ever.
While current mechanisms of disease resistance in agriculture are mostly chemical (think pesticides, fungicides, and a host of other -ides), with pikobodies, genetic treatments may replace our reliance on chemical treatments. “We can generate made-to-order resistance genes against virtually any pathogen,” Kourelis said.
The idea is certainly an imaginative if not unbelievable one, almost like science fiction. But it didn’t originate out of the blue. Scientists have long been interested in the integrated domain (ID) of plant receptors, which is responsible for recognizing pathogen effectors and triggering an immune response. In one key study by French scientist Stella Cesari, engineering the ID of Pik-1—a receptor that normally attacks fungal invaders—could allow tobacco plants to gain specificity and bind new sequences. Ever since then, Kamoun has wondered if he could engineer Pik-1 to recognize other pathogens.
The second piece of the puzzle—introducing the animal antibody—came four years ago, with Kourelis’ arrival at the lab. During a lab meeting, Kourelis proposed “A General Solution for Plant Pathology”—a title which, of course, caught the attention of everyone in the room.
As Kourelis noted, one problem with plant immune defenses is that they lack mobile or specialized cells for attacking viruses, unlike animals. Plants instead rely on nucleotide-binding, leucine-rich receptors (NLRs), which can recognize diverse pathogen components and activate the immune response. But NLRs are limited in what they can recognize—and their scope is hard-coded by DNA, which cannot evolve as fast as rapidly-changing viruses.
In contrast, mammals can create and proliferate specific antibodies for virtually any virus they are exposed to. Not only do these antibodies target, mobilize, and kill viral particles, but the animal also retains them well after the infection, in case of future threats—the same principle which guides vaccines. “Basically, you could potentially build a disease resistance gene against any plant pathogen by exploiting the adaptive immune system of animals,” Kamoun said.
Enter the llamas. Kourelis suggested taking camelid mammal nanobodies—the fragment of antibodies which actually binds to the pathogen—and fusing them to Pik-1. In this way, plants’ integrated domains could serve as a scaffold to trigger the immune response, while mammalian antibodies could let them recognize a host of other pathogens.
To turn their concept into practice, Kourelis still needed a framework for where and how to engineer the receptors. His answer was bolstered by an earlier project by colleague and Ph.D.student Aleksandra-Ola Bialas, who investigated how the Pik-1 receptor arose fifty million years ago. Critically, by looking at the evolutionary origins of Pik-1, Bialas’s work helped delimit the boundaries of the domain that Kourelis was so interested in.
“So fifty million years ago, nature actually did this engineering and integrated the domain into this [Pik-1] receptor,” Kamoun said. “And if you understand how nature has done it, you could repeat it in the lab.”
Even with Bialas’s contribution, the process to pare down potential candidates was arduous. It required iterations of revisiting basic science, tweaking existing combinations, and, at some point, sheer brute force, Kourelis recalled. “Sometimes we were like, ‘Let’s drop in as many nanobodies as we can and see if some of them work,’” Kourelis said. “‘And then, if a few work, even if they don’t work great, let’s understand what’s going on there.’”
Eventually, he engineered eleven pikobody candidates to recognize fluorescent proteins. They were vetted for autoimmune responses and cell death responses until only four pikobodies remained. After introducing a live Potato virus X, which was engineered to express fluorescent proteins, two pikobodies were found to halt viral spread substantially. And if used together, these two pikobodies proved to be even more effective.
Now, Kourelis and Kamoun are looking toward future applications for pikobodies. They are both enamored with the potential of “designer” domains, with each plant tailored with pikobodies against diseases that might threaten it. They even speculated that, with advanced artificial intelligence, computers may someday design the binding domains, circumventing the need for antibodies altogether.
For Kourelis, one of the challenges lies in accounting for—and staying ahead of—pathogen evolution. As he looks to apply his model and develop a toolkit of new receptors, he hopes to identify target sequences that will be long-lasting. If they can pick target sequences that are unlikely to change, then the pathogen is less likely to evolve to resist that receptor. “Then again, I like to say, ‘Never bet against the pathogen,’” Kamoun said.
In the book-lined office from which Kamoun and Kourelis take this Zoom interview, it is not difficult to envision them collaborating in a laboratory. Kourelis fields a question. Kamoun modifies Kourelis’s answer, then raises another point, which reminds Kourelis of another idea. Even as they talk to me, their conversation with each other never ceases.
Their dialogue—revising, rephrasing, circling—reflects the cycles it took Kourelis to get to his top-performing pikobodies. But this is their process and, perhaps, the reason they are successful. Kourelis emphasized multiple times that they regard themselves as scientists first, and engineers second. No matter what science-fiction universe plant-animal hybrids and pikobodies seem to resemble, for Kourelis and Kamoun, these ideas are firmly rooted in listening to what already exists—to plants, to other scientists, to each other—before they create something new.
“There’s a more fundamental understanding in seeing how, in nature, these proteins have evolved, how these domain integrations have happened,” Kourelis said. “You have to understand before you can take the next step. You have to understand nature by making it.”