Scripps Research Captures First Structural Images of RNA Interference Machinery in Action

Scientists at Scripps Research have captured the first high-resolution structural images of the human RNA interference molecular machinery in its slicing-ready state, identifying the precise atomic interactions that determine when and where the machinery cuts. The findings, published in Nature Structural & Molecular Biology, could provide a structural basis for designing more effective siRNA drugs.

RNA interference is a natural cellular mechanism for controlling whether specific genes are active. Seven RNA interference drugs have already received FDA approval, with more in development, but the molecular details of how the system executes its cuts had remained poorly understood. For drug developers, a key challenge is that thousands of different siRNA sequences could theoretically shut down any given disease-causing gene—and without being able to predict which ones will succeed, development has relied heavily on trial and error. 

“RNA interference has become a powerful platform for treating disease, with more drugs entering the pipeline every year,” said senior author Ian MacRae. “But until now, we’ve been designing these drugs somewhat in the dark. This work finally gives us the structural picture we need to understand what makes a good therapeutic siRNA and how to design better ones.”

First author Sucharita Sarkar and co-author Luca Gebert used cryo-electron microscopy to capture the atomic-level arrangement of Argonaute 2—the protein at the center of the RNA interference machinery—in the moments just before it cuts its RNA target. Stabilizing this fleeting state required engineering a special set of mutations. “The trick was that we had to come up with a special set of mutations that would stabilize this active conformation,” said Gebert.

The resulting structure revealed that the guide-targeted RNA duplex is physically distorted inside Argonaute 2, positioning the precise chemical bond to be cleaved directly within the protein’s active site. Two previously overlooked amino acids, Lysine709 and Arginine710, play key roles in this process. Lysine709 acts as a molecular checkpoint, moving into cutting position only after the duplex deformation is triggered. Arginine710 fine-tunes catalytic efficiency by sensing a specific position in the target RNA, explaining a long-standing but poorly understood rule in siRNA design.

“For the first time, we could see exactly where two previously overlooked amino acids sit in the active site—and their positions redefine how we understand Argonaute 2 catalysis," said Sarkar.

The findings suggest that siRNA sequences and chemical modifications that allow the paired RNA to more readily adopt the distorted shape should favor activation, while those that rigidify the central region are expected to impair it. MacRae described this as opening the door to rational design—engineering siRNA sequences based on structural principles rather than trial and error. “Understanding the mechanism at this level should allow us to design better drugs from the start, which could expand the range of diseases we're able to treat with RNA interference,” he said.