Abstract
VHHs (also known as nanobodies) are important therapeutic antibodies. To prolong their half-life in bloodstream, VHHs are usually fused to the Fc fragment of full-length antibodies. However, stability is often the main challenge for their commercialization, and methods to improve stability are still lacking. Here, an in silico pipeline is developed for analyzing the stability of an anticancer VHH-Fc fusion antibody (VFA01) and designing its stable variants. Computational modeling is used to analyze the VFA01 structure and evaluate its conformational stability, disulfide bond reduction state, and aggregation and degradation tendency. By building mechanistic models of aggregation and degradation, the hotspot residues affecting stability: C130, F57, Y106, L120, and W111 are identified. Based on them, a series of VFA01 variants are designed and obtained a variant M11 (C130S/W111F/F57K) whose stability is significantly enhanced compared to VFA01: there are no visible particles in solution, and the change rate of DLS average hydrodynamic size, SEC HMW%, and CE-SDS purity are improved by 6.2-, 3.4-, and 1.5-fold, respectively. Both antigen-binding activity and production yield are also improved by about 1.5-fold. The results show that our computational pipeline is a very promising approach for improving the protein stability of therapeutic VHH-Fc fusion antibodies.