July 16, 2024 | Caixuan Liu, Kejia Wu, Hojun Choi, Hannah Han, Xulie Zhang, Joseph L. Watson, Sara Shijo, Asim K. Bera, Alex Kang, Evans Brackenbrough, Brian Coventry, Derrick R. Hick, Andrew N. Hoofnagle, Ping Zhu, Xingting Li, Justin Decarreau, Stacey R. Gerben, Wei Yang, Xinru Wang, Mila Lamp, Analisa Murray, Magnus Bauer, David Baker
This study presents a method for designing high-affinity binders for intrinsically disordered proteins (IDPs) and intrinsically disordered regions (IDRs) using RFdiffusion, a deep learning-based approach that samples both target and binder conformations. The method starts from the target sequence alone and does not require prespecification of the target geometry. The approach was used to generate binders for several IDPs and IDRs, including Amylin, C-peptide, and VP48, with binding affinities ranging from 3.8 nM to 100 nM. The Amylin binder inhibits amyloid fibril formation and dissociates existing fibers, enabling the enrichment of Amylin for mass spectrometry-based detection. For IDRs such as G3bp1, common gamma chain (IL2RG), and prion, binders were designed to beta strand conformations, achieving binding affinities of 10-100 nM. The IL2RG binder colocalizes with the receptor in cells, enabling new approaches to modulating IL2 signaling. The approach is widely applicable for creating binders to flexible IDPs/IDRs spanning a wide range of intrinsic conformational preferences. The binders were validated using various techniques, including bio-layer interferometry (BLI), circular dichroism, and mass spectrometry. The study also demonstrates the utility of the approach in applications such as amyloid fibril dissociation, mass spectrometry-based detection, and colocalization imaging in mammalian cells. The results show that the designed binders have high specificity for their intended targets and can be used as affinity reagents for therapeutic and diagnostic applications. The study highlights the potential of RFdiffusion in designing binders for IDPs and IDRs, which are important biomarkers in clinical care and biomedical research.This study presents a method for designing high-affinity binders for intrinsically disordered proteins (IDPs) and intrinsically disordered regions (IDRs) using RFdiffusion, a deep learning-based approach that samples both target and binder conformations. The method starts from the target sequence alone and does not require prespecification of the target geometry. The approach was used to generate binders for several IDPs and IDRs, including Amylin, C-peptide, and VP48, with binding affinities ranging from 3.8 nM to 100 nM. The Amylin binder inhibits amyloid fibril formation and dissociates existing fibers, enabling the enrichment of Amylin for mass spectrometry-based detection. For IDRs such as G3bp1, common gamma chain (IL2RG), and prion, binders were designed to beta strand conformations, achieving binding affinities of 10-100 nM. The IL2RG binder colocalizes with the receptor in cells, enabling new approaches to modulating IL2 signaling. The approach is widely applicable for creating binders to flexible IDPs/IDRs spanning a wide range of intrinsic conformational preferences. The binders were validated using various techniques, including bio-layer interferometry (BLI), circular dichroism, and mass spectrometry. The study also demonstrates the utility of the approach in applications such as amyloid fibril dissociation, mass spectrometry-based detection, and colocalization imaging in mammalian cells. The results show that the designed binders have high specificity for their intended targets and can be used as affinity reagents for therapeutic and diagnostic applications. The study highlights the potential of RFdiffusion in designing binders for IDPs and IDRs, which are important biomarkers in clinical care and biomedical research.