Modified heated dynamics and deep mutational scanning enhance anti-CXCR2 antibody affinity

Colorectal cancer (CRC) is a leading global malignancy, with many stage IV cases requiring surgery and chemotherapy. However, chemoresistance limits treatment efficacy. Targeting resistance pathways such as CXCR2 with therapeutic antibodies offers a promising solution. This study harnessed deep mutational scanning guided by modified heated coarse-grained molecular dynamics (CGMD) simulations, to enhance the binding affinity of the HY29-1 antibody toward CXCR2. The structural stability and intermolecular binding of the variable heavy (vH) and light (vL) chains of the HY29-1 antibody fragment (Fv-HY29-1) alone and complexed with CXCR2 were assessed using modified heated CGMD simulations under thermal stress conditions. Using a modified heated CGMD protocol, the Fv-HY29-1 antibody fragment, modeled via ClusPro (C7) exhibited remarkable conformational stability, maintaining complete (100%) complexation with CXCR2 over a 70 ns simulation. Meanwhile, a benchmark model of the established Fv-Fab14-canine parvovirus capsid complex retained 100% stability under similar conditions. Extended deep mutational scanning over a 1000 ns trajectory pinpointed critical framework residues in the vL chain (A43, L47, and F98) as key determinants of paratope-epitope binding. Rational substitutions (A43P, L47V, F98W) significantly enhanced binding affinity, improving from -36.02 kcal/mol to -94.09 kcal/mol, approximately two-fold increase in binding strength. This marked affinity gain translated into strengthened CXCR2 engagement, highlighting the promise of structure-guided antibody engineering in optimizing therapeutic interactions. This study demonstrated that a 1 μs modified heated simulation with coarse-grained models and deep mutational scanning enables cost-effective optimization of antibody binding affinity and will guide future design of high-affinity antibodies across diverse antigen targets.