For the first time, a systematic protein interaction map—or interactome—has been constructed for a plant. In a pair of papers published online today (July 28) in Science, researchers from the Arabidopsis Interactome Mapping Consortium (AIMC) present their data from an extensive effort to map the pairwise interactions of over 2,700 proteins expressed within the cells of Arabidopsis thaliana, and show that pathogens target the most active proteins during infection.
The map shows that “there are few proteins that are highly connected,” said Christian Landry, an assistant professor of Biology at Laval University, who did not participate in the research. “This kind of structure gives robustness to the network because if you target proteins randomly, you are more likely to hit peripheral proteins” and not significantly disrupt cell function.
The effort also serves to “put proteins that we don’t know about into a molecular context,” added Pascal Braun, chair of AIMC and an author on both papers. While there has been much attention paid to sequencing projects, the genome tells researchers little about the functions of proteins that actually drive cellular processes, he said. “We need to know what the proteins do and how they interact. This is the first time we have done this at a systematic level for any plant.”
In the first study, Braun and colleagues expressed 8,000 Arabidopsis proteins—representing 30 percent of the plant’s protein-coding genes—in yeast cells. The yeast cells were engineered to replicate when two isolated proteins came together, providing a visual cue for the protein interactions. The researchers tested all pairwise combinations of the 8,000 proteins, one couple at a time, and verified interactions among about 2,700 of them.
From the results, the team constructed a map of all the interactions, which revealed nodes of dense interconnectivity centered around a relatively small number of proteins. These highly connected proteins, known as hubs, “are important for keeping everything together,” Braun said. “The system collapses if you take down the hubs.” The vast majority of the proteins, however, had only a handful of connections. Removing any of these peripheral proteins from the cell would be unlikely to significantly disrupt any major cellular processes.
The researchers also looked at the impact of these networks on evolution. The protein products of duplicated genes, for example, might be expected to take on different functions, as one can maintain the original task while the other is free to accumulate mutations. But the researchers found that most gene duplicates in Arabidopsistended to interact with many of the same proteins, even though those duplicates had originated more than 700 million years ago, suggesting that the interactome somehow reduces the freedom of duplicated proteins to diverge.
In the second study, Jeffrey Dangl of the University of North Carlolina at Chapel Hill and his colleagues paired genes from the Arabidopsis interactome map and additional immune system proteins with effector proteins from known Arabidopsis pathogens—the bacteria Pseudomonas syringae and the fungus-like oomycete Hyaloperonospora arabidopsidis. Interestingly, rather than directly interacting with the immune system proteins, the effectors bound to Arabidopsis hub proteins, indicating that the pathogens induce immune responses indirectly.
Furthermore, despite the 2 billion years of evolution separating the two pathogens, they shared the strategy of targeting the hub proteins, and even interacted with 18 of the same major hubs. When the researchers knocked out those 18 genes individually in Arabidopsis plants and exposed the plants to the effector proteins, 15 of the mutants had an altered immune response.
“This is an unheard-of validation rate,” said Braun. “Even though [the pathogens] have independently evolved and their mode of attack is different, they are still converging on the highly connected proteins.” This suggests that the pathogens have evolved to target proteins that participate in more than one immune response or pathogen recognition pathway.
The results point to ways researchers might protect plants from disease or engineer them to be better crops, for example, said Braun. “Plants play a fundamental role in food security, energy security, and climate change,” he said, and understanding how proteins interact in the cell could lead to the discovery of proteins with biotechnological applications.
But advancing plant disease resistance is not as simple as just removing these key proteins, because it not only alters the immune response, but would be detrimental to other plant cell processes. “The real challenge is to determine what the targeted plant proteins are doing in the innate immune response and how modifying the targets will modify the disease,” said Cyril Zipfel, group leader in molecular and microbe interactions at The Sainsbury Laboratory.