Junyu Liu, Wanlin Chen, Ben Cristofori-Armstrong, Theo Crawford, Kai-En Chen, Peng Xie, Rainbow Wing Bo Chan, Yifei Zhu, Mimi Golder, Anneka Pereira Schmidt, Jennifer D Naughton, Nicholas D Condon, Åsa Andersson, Farhad Dehkhoda, Kirsten L McMahon, Tabea Klasfauseweh, Ashvriya Thapa, Hue Tran, Poanna Tran, Sina Jami, Lotten Ragnarsson, Sebastian G B Furness, Jennifer R Deuis, Brett M Collins, Wai-Hong Tham, Indira Prasadam, Irina Vetter, Mehdi Mobli
Selective molecular recognition of membrane proteins is challenging because they contain few solvent-exposed extracellular epitopes, which often depend on their native environment for structure, making them difficult to isolate faithfully for binder discovery. Here, we show that antigen-detected NMR is well suited both to characterizing the folding of engineered minimal epitopes from the human voltage-gated sodium channel NaV1.7 and to selecting binders that recognize their solvent-exposed surfaces. Isotope labelling of the antigen enables NMR resonance assignment to assess retained local secondary structure, while 15N titration and zz-exchange mapping provide binding and interface information. Combined with AlphaFold2 complex prediction, this creates a practical method for screening and ranking candidate binders. The approach was further validated by a high-resolution x-ray structure of an antigen-nanobody complex. Applying this workflow identified R4C8, a subtype- and species-selective nanobody whose binding to the extracellular surface of human NaV1.7 is supported by zz-exchange mapping, modelling, and cellular recognition, and which has minimal effects on channel gating. R4C8 detected NaV1.7 in engineered cell lines and in primary osteoarthritis-derived chondrocytes, providing a useful tool for selective target detection. These results show how antigen-detected NMR can support peptide engineering and structure-guided protein binder selection against minimal epitopes.