Many of the bacteria that wreak havoc on crops and threaten our food supply use a common strategy to induce disease: They inject combinations of harmful proteins directly into plant cells. For 25 years, biologist Shengyang He and his senior research associate Shinya Nomura have been studying the set of molecules that plant pathogens use to cause disease in hundreds of crops around the world, from rice to apple orchards.
Now, three collaborative research groups have finally found answers to how these molecules make plants sick, as well as ways to undo them.
Relevant research results were published in the journal Nature on September 13.
Researchers in the lab study key ingredients in this deadly cocktail, a family of injected proteins called AvrE/DspE, which cause diseases ranging from brown spot in beans to bacterial spot in tomatoes to fire blight in fruit trees.
Since its discovery in the early 1990s, people who study plant diseases have taken a keen interest in this protein family. They are key weapons in bacterial arsenals; eliminating them in the laboratory can render otherwise dangerous bacteria harmless. But despite decades of effort, many questions about how they work remain unanswered.
The researchers found that many proteins in the AvrE/DspE family can suppress a plant's immune system or form water-soaked black spots on plant leaves - the first sign of infection. They even know the basic sequence of amino acids, which join together like beads on a string to form proteins. But they didn't know how the string of amino acids folded into its three-dimensional shape, so they couldn't easily explain how they worked.
Part of the problem is that the proteins in this family are so large. General bacterial proteins may only have 300 amino acids, while proteins of the AvrE/DspE family have 2,000 amino acids.
The researchers looked for other proteins with similar sequences for clues, but found none with a known function.
"They are strange proteins," he said. So they turned to a computer program called AlphaFold2, released in 2021, which uses artificial intelligence to predict the three-dimensional shape of a given string of amino acids.
Researchers know that some members of this family help the bacteria hide from the plant's immune system. But when they saw the protein's three-dimensional structure for the first time, they discovered another role.
"When we first saw this model, it was completely different from what we imagined," said study co-author Pei Zhou, a professor of biochemistry at Duke University.
Researchers studied AI predictions of bacterial proteins that infect crops such as pears, apples, tomatoes and corn and found that they all had similar three-dimensional structures. They appear to fold into a small mushroom with a cylindrical stem, like a straw.
The predicted shape matched closely with images captured using cryo-electron microscopy of the bacterial protein that causes fire blight on fruit trees. Viewed from top down, the protein looks very much like a hollow tube.
This got the researchers thinking: Perhaps bacteria use these proteins to punch holes in plant cell membranes and "force the host to drink water" during the infection process.
When bacteria enter a leaf, one of the first areas they come into contact with is the space between cells, called the cytoplasm. Normally, plants keep this area dry to allow for the gas exchange needed for photosynthesis. But when bacteria invade, water accumulates inside the leaves, creating a moist and comfortable paradise for them to feed and reproduce.
Further study of the predicted three-dimensional model of the fire blight protein revealed that while the outside of the straw-like structure is water-resistant, its hollow core has a special affinity for water.
To test the water channel hypothesis, the research team collaborated with Duke University biology professor Dong Ke and his lab postdoc and co-first author Felipe Andreazza. They added genetic readouts for the bacterial proteins AvrE and DspE to frog eggs, using the eggs as cellular factories to make these proteins. Place the frog eggs in diluted saline. Too much water will cause the eggs to swell rapidly and rupture.
The researchers also tried to unblock these bacterial proteins by blocking their channels. Nomura focused on a type of tiny spherical nanoparticles called PAMAM dendrimers. Such dendrimers have been used in drug delivery for more than two decades and can be made into particles with precise diameters in the laboratory.
"Our hypothesis was that if we found chemicals with the right diameter, we might be able to plug the pores," he said.
After testing particles of different sizes, they found one they thought was just the right size to block the aquaporins produced by the fire blight pathogen Erwiniaamylovora.
They took frog eggs that could synthesize this protein and watered them with PAMAM nanoparticles, so that the water no longer flowed into the frog eggs. They don't swell.
They also treated Arabidopsis plants infected with the pathogen Pseudomonas syringae, which causes bacterial spots. The channel-blocking nanoparticles stopped bacterial growth, reducing pathogen concentrations in plant leaves up to 100-fold.
These compounds are also effective against other bacterial infections. The researchers did the same experiment on pear fruits, which were exposed to the bacterium that causes fire blight, but the fruits never showed symptoms—the bacteria did not make them sick.
"It was a long process, but it worked," he said. "We're very excited about this."
The researchers say the findings could provide new ideas for controlling many plant diseases. 80% of the food we eat is produced by plants. However, more than 10% of global food production - crops such as wheat, rice, corn, potatoes and soybeans - are lost to plant pathogens and pests every year, costing the global economy as much as $220 billion.
The research team has applied for a provisional patent on this method. The next step is to figure out how this protection works by looking in more detail at how channel-blocking nanoparticles and channel proteins interact, said Zhou and co-first author Jie Cheng, a doctoral student in Zhou's lab.
"If we can image these structures, we can better understand and design better crop protection solutions," Zhou said.