Abstract Nanobodies offer unique advantages in biomedical and biotechnological applications due to their smaller size, ability to bind challenging epitopes, and affordable production using recombinant technology. However, challenges in large-scale production, stability, and solubility limit their widespread use. To address this, we use artificial intelligence tools to optimize the scaffold region of nanobodies. We apply our approach to four nanobodies against clinically relevant targets: the cytokine tumor necrosis factor alpha, the chemotherapeutic drug methotrexate, the pancreatic biomarker amylase, and the placental hormone chorionic gonadotropin. For all the nanobodies tested, we improve stability, production, and intracellular stability while maintaining antigen-binding affinity. Our results thus demonstrate the potential for using AI-driven protein engineering to enhance the properties of nanobodies, offering insights into the interplay between stability, solubility, and antigen binding. Given the high conservation of the scaffold, we propose some mutations that could directly transfer to other nanobodies, providing an easy-to-implement, generalizable engineering strategy.
Pejenaute et al. (2025) studied this question.