De novo design of Ras isoform selective binders

Area of Science:

• Molecular Biology
• Oncology
• Bioinformatics
• Protein Engineering

Background:

• The proto-oncogene Ras regulates critical intracellular pathways, with four major isoforms (KRAS4A, KRAS4B, HRAS, and NRAS) exhibiting high sequence homology.
• Investigating individual Ras isoform functions is crucial due to their varied associations with different cancers.
• A significant limitation in Ras research is the scarcity of isoform-specific binding reagents, primarily due to challenges in targeting their divergent C-termini.

Purpose of the Study:

• To overcome the lack of isoform-specific reagents for Ras proteins.
• To design and validate novel binding molecules that specifically target the C-termini of Ras isoforms.
• To enable the dissection of distinct biological and pathological roles of individual Ras isoforms.

Main Methods:

• Utilized deep learning-based methods for *de novo* design of protein binders.
• Engineered Ras isoform-specific binders (RIBs) targeting the unique C-terminal regions of KRAS4A, KRAS4B, HRAS, and NRAS.
• Validated the specificity and functionality of designed RIBs *in vitro* and in cellular models.

Main Results:

• Successfully designed and generated Ras isoform-specific binders (RIBs) targeting the Ras C-terminus.
• Demonstrated high specificity of RIBs for their cognate Ras isoforms both *in vitro* and within cells.
• Observed that RIBs disrupt Ras membrane localization and inhibit Ras activity, confirming their functional impact.

Conclusions:

• Deep learning enables the *de novo* design of highly specific protein binders for challenging targets like Ras C-termini.
• The developed RIBs are valuable tools for distinguishing the functions of Ras isoforms in biological processes.
• These RIBs hold potential for advancing research into Ras-driven diseases, including various cancers.