Avimer Bio and CNIO announce JACS publication detailing computational methods to build self-assembling protein cages with atomic precision for targeted medicine.
San Diego, CA (PRUnderground) October 1st, 2026
AVIMER BIO AND CNIO ADVANCE PROTEIN CAGE DESIGN FOR TARGETED MEDICINE
Avimer Bio, in collaboration with structural biologists at the Spanish National Cancer Research Centre (CNIO), announced a new publication in the Journal of the American Chemical Society (JACS) describing advances in building self-assembling protein cages with atomic precision. The work addresses a longstanding challenge in designing protein assemblies with predictable shapes.
The Technical Achievement
The research addresses how to build predictable protein architectures when building blocks bend as they assemble. Chief Scientific Officer Todd Yeates introduced geometric methods for creating protein cage nanoparticles decades ago at UCLA. Those earlier methods required precise shapes, making structural flexibility a hurdle for designers.
The team developed computer algorithms that predict how spiral-shaped protein segments, called alpha helices, bend. This allows scientists to use the natural flexibility of protein molecules during design. The approach yielded cubic protein cages with minimal mutations to four trimeric protein blocks, twelve subunits in total, reaching molecular masses over 600 kilodaltons while maintaining stability in solution.
“Protein cages require many subunits to fit together in precise arrangements. We used mathematics and mechanical design to identify architectures where the proteins’ natural flexibility helps them adopt the shapes needed for assembly. Combining those principles with modern AI methods gave us a way to design around that flexibility and opens new possibilities for building therapeutic protein assemblies.”
— Todd Yeates, Ph.D., Chief Scientific Officer and Co-Founder, Avimer Bio
Structural Validation
High-resolution cryo-electron microscopy performed at CNIO in Madrid, led by Pablo San Segundo-Acosta and Roger Castells-Graells, confirmed that the physical protein cages matched computational models. Cryo-EM structures reached resolutions of 3.0 to 3.9 angstroms. Differences between the models and measured structures stayed as low as 2 angstroms across the protein backbone. The results confirmed the predicted patterns of helix bending.
Clinical Applications and Advantages
Designed protein cages present multiple copies of therapeutic proteins in controlled arrangements to engage cell receptors. Certain receptors must gather in groups to activate a signaling pathway. A programmable cage controls that grouping beyond the two binding arms on a conventional antibody. Future designs can display different targeting proteins together to direct activity toward selected cell populations.
Potential applications include autoimmune conditions like rheumatoid arthritis and ulcerative colitis, where engaging selected immune cells helps control inflammation. In oncology, therapeutic proteins on a cage can gather receptors that stimulate immune cells to attack tumors.
The platform centers on two distinct properties:
-
● Few changes to natural protein sequences: The approach preserves much of the starting protein sequence. Applied to human building blocks, it supports designs that retain high similarity to human proteins, providing a rationale for evaluating immune response.
-
● Single protein subunit composition: Each cage builds from a single type of protein subunit. That simplifies production and maintains consistent composition compared to particles requiring several distinct chains.
-
Research Team and Methodology
Robin Aglietti and Peter Bowers conducted the design work at Avimer Bio alongside Todd Yeates. Roger Castells-Graells directed the structural biology team at CNIO during the cryo-electron microscopy analysis.
Publication Details
The complete study, “Design and Structure of Protein Cages Based on Helical Fusion and Machine Learning,” is available in the Journal of the American Chemical Society at https://doi.org/10.1021/jacs.6c11491
About Avimer Bio
Avimer Bio designs programmable protein architectures for targeted therapeutics. Based in San Diego, the company develops computational methods to build protein assemblies that control how therapeutic molecules engage cells and their receptors. For more information, visit www.avimerbio.com
Media Contact:
Adam Sragovicz
adam@avimerbio.com
